Upload 913 IFPRI-matched files to data/part_2

data/part_2/0002306736.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"c96151f36918c59c16a5ffc0dfa6a2a0","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/35741c4a-b431-44f0-8516-c50f8bc3fe00/retrieve","description":"","id":"-418022776"},"keywords":[],"sieverID":"5d2cdb50-d334-4a9d-a953-de8dcb892103","pagecount":"4","content":"District Nutrition Profiles (DNPs) are available for 707 districts in India. They present trends for key nutrition and health outcomes and their cross-sectoral determinants in a district. The DNPs are based on data from the National Family Health Survey NFHS-4 (2015NFHS-4 ( -2016) ) and NFHS-5 (2019NFHS-5 ( -2021)). They are aimed primarily at district administrators, state functionaries, local leaders, and development actors working at the district-level.  Note: NA refers to data unavailable for a given round of NFHS/Census.• What are the trends in infant and young child feeding (early initiation of breastfeeding, exclusive breastfeeding, timely initiation of complementary feeding, and adequate diet)? What can be done to improve infant and young child feeding? • What are the trends in IFA consumption among pregnant women in the district? How can the consumption be improved?• What additional data are needed to understand diets and/or other determinants?  Note: NA refers to data unavailable for a given round of NFHS/Census.• How can the district increase women's literacy, and reduce early marriage, if needed?• How does the district perform on providing drinking water and sanitation to its residents? Since sanitation and hygiene play an important role in improving nutrition outcomes, how can all aspects of sanitation be improved? • How can programs that address underlying and basic determinants (education, poverty, gender) be strengthened?• What additional data are needed on food systems, poverty or other underlying determinants? Note: NA refers to data unavailable for a given round of NFHS/Census.• How does the district perform on health and nutrition interventions along the continuum of care? Does it adequately provide both prenatal and postnatal services to women of reproductive age, pregnant women, new mothers and newborns? • How has access to health and ICDS services changed over time (food supplementation, health and nutrition education and health checkups)?","tokenCount":"300","images":["-418022776_1_1.png","-418022776_1_2.png","-418022776_1_3.png","-418022776_1_4.png","-418022776_1_5.png","-418022776_1_6.png","-418022776_1_7.png","-418022776_1_8.png","-418022776_1_9.png","-418022776_1_10.png","-418022776_1_11.png","-418022776_1_12.png","-418022776_2_1.png","-418022776_2_2.png","-418022776_2_3.png","-418022776_2_4.png","-418022776_2_5.png","-418022776_2_6.png","-418022776_2_7.png","-418022776_2_8.png","-418022776_2_9.png","-418022776_2_10.png","-418022776_2_11.png","-418022776_2_12.png","-418022776_3_1.png","-418022776_3_2.png","-418022776_3_3.png","-418022776_3_4.png","-418022776_3_5.png","-418022776_3_6.png","-418022776_3_7.png","-418022776_3_8.png","-418022776_3_9.png","-418022776_3_10.png","-418022776_3_11.png","-418022776_3_12.png","-418022776_4_1.png","-418022776_4_2.png","-418022776_4_3.png","-418022776_4_4.png","-418022776_4_5.png","-418022776_4_6.png","-418022776_4_7.png","-418022776_4_8.png","-418022776_4_9.png","-418022776_4_10.png","-418022776_4_11.png","-418022776_4_12.png"],"tables":["-418022776_1_1.json","-418022776_2_1.json","-418022776_3_1.json","-418022776_4_1.json"]}

data/part_2/0011908196.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"a29e4b146ece3a996e20be7677f752a5","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/badb28ae-6870-45f2-9541-aab0369f66b5/retrieve","description":"The migrant selection literature concentrates primarily on spatial patterns. We integrate two workhorses of the labor literature, the Roy and search models, to illustrate the implications of migration duration for patterns of selection. Theory and empirics show that temporary migrants are intermediately selected on education, with weaker selection on cognitive ability. Longer migration episodes lead to stronger positive selection on both education and ability because the associated jobs involve finer employee-employer matching and offer greater returns to experience. Networks are more valuable for permanent migration, where search costs are higher. Labor market frictions explain observed complex network-skill interactions. When considering migrant selection, the economics literature has largely focused on patterns by area of origin. However, the duration of migration episodes–temporary versus permanent–is another important determinant of selection. We integrate two workhorses of the labor literature, the Roy model and a search model, to illustrate the implications of migration duration for patterns of self-selection. We provide theoretical and empirical evidence showing that, because short-term migration episodes have less scope for skill-based matching and greater need for screening, temporary migrants are more likely to display intermediate selection on education, with weaker selection on underlying cognitive ability. Longer term migration episodes, in contrast, allow for finer employee-employer matching and greater returns to experience, leading to stronger positive selection on both education and cognitive ability among permanent migrants. Networks are also found to be more valuable for permanent migration, where search costs tend to be higher. However, we also provide evidence of complex network-skill interactions, driven primarily by labor market frictions.","id":"-1187811762"},"keywords":["migration","search costs","networks","Pakistan"],"sieverID":"43a4fe94-a375-4bb9-84a3-e2297c12b593","pagecount":"45","content":"established in 1975, provides evidence-based policy solutions to sustainably end hunger and malnutrition and reduce poverty. The Institute conducts research, communicates results, optimizes partnerships, and builds capacity to ensure sustainable food production, promote healthy food systems, improve markets and trade, transform agriculture, build resilience, and strengthen institutions and governance. Gender is considered in all of the Institute's work. IFPRI collaborates with partners around the world, including development implementers, public institutions, the private sector, and farmers' organizations, to ensure that local, national, regional, and global food policies are based on evidence. IFPRI is a member of the CGIAR Consortium.Trade -the movement of goods and services-has long been central to economic development, both in theory and in practice. But as the geographic mobility of workers continues to rise, migration-the movement of individuals and families-will be equally central to economic development. Remittances now far exceed official aid flows (Ratha et al. 2014), and the skill composition of migrants has increasingly greater consequences for destination markets (Card 1990; Borjas 2005; Card 2005; Borjas  2006; Ottaviano and Peri 2012). But, for migration unlike for trade, individuals must move with their goods, making the migration decision much more complex. Therefore, to better understand the impact of migration on both spatial and economic inequality, we must consider more deeply the migration decision itself.The extant literature still largely fails to make substantive distinctions among migrants, despite drastically varied motives and experiences. In this paper, we take an initial step toward modeling heterogeneity among migrants, even those originating from the same area in the same time period, by focusing on the duration of migration episodes. We consider two types of migrants: temporary migrants, who tend to make short trips and then return to the area of origin, and permanent migrants, who tend to move over longer distances with no intention to return. Since short-term migration is typically motivated by transitory market fluctuations, the characteristics of short-term migrants will vary more with local economic conditions and business cycles. Their investment decisions, including those concerning human capital, will be based on the returns in local-rather than distant-labor markets (Dustmann 1993), and their households may be inclined to substitute human capital for social capital investments (McKenzie and Rapoport 2010). In contrast, characteristics of long-term migrants are more likely to be stable over time, shifting only when the underlying market structure changes. The gains to migration will primarily accrue at the destination, rather than returning to the area of origin. Because gains and search costs are amplified when migration is long term, there is a greater role for complementarities between worker ability and networks (Montgomery 1991; Loury  2006; Beaman and Magruder 2012).In this paper, we describe patterns of selection into temporary and permanent migration. The simultaneity of migration, education, and other investment decisions creates a stark trade-off between a comprehensive, more descriptive assessment of self-selection and a narrowly focused identification of a specific causal relationship. We opt for the former because the characteristics of migrants are the principal determinant of the effects of migration on both sending and receiving areas. In contrast, identifying the underlying causes of selection would be more relevant for understanding the consequences of, for example, education policy on migration decisions.We use data from a unique panel survey of rural households in Pakistan spanning 22 years (1991-2013) to study the drivers of migration during this time period. Our survey offers an array of detailed information on worker attributes, including education, cognitive ability (digit span and ravens test scores), and physical ability (height). To account for differences in search costs, we incorporate demographic and wealth information from 1991 to avoid simultaneity bias. We also include a measure of migrant networks, based on the movement of all respondents in an individual's community of origin (but not in the individual's household) since 1991. Because migration by network members is not contemporaneous, this strategy helps alleviate concerns about general equilibrium effects while still providing a good characterization of network size and scope. Since networks affect the costs and benefits of migration, we include a comprehensive set of interaction terms between migrant networks, education, and ability.To formalize the differences between permanent and temporary migration, we embed a classic Roy  (1951) model with heterogeneous moving costs, as in (Chiquiar and Hanson 2005; McKenzie and Rapoport  2010; Orrenius and Zavodny 2005), into a standard search model (Lippman and McCall 1976). Individuals choose between migration opportunities with either a permanent (lifetime) wage offer or a temporary (single-period) wage offer. The difference in the duration of these opportunities implies important distinctions not only in the decision process but also in patterns of self-selection, as well as in the data required to fully characterize migration decisions. Long-term employment affords greater opportunity for employer-employee matching and returns to experience, but these opportunities also require a more costly search process. Furthermore, while decisions about short-term moves can be adequately characterized by a static per-period problem, decisions about long-term moves are characterized by an optimal stopping problem. This fact intimates that recent (for example, those from the previous five years) migration histories may not adequately characterize individual migration choices. Finally, combining temporary and permanent migrants, and conflating the selection processes for the two types of moves, may be problematic if the composition of temporary versus permanent migrants is changing over time or is responsive to changing economic conditions.Our empirical findings reveal positive selection on cognitive ability for both permanent and temporary migrants, driven by individuals with dense migrant networks. With regard to schooling, having a tertiary education increases the likelihood of permanent migration, while having a secondary education increases the likelihood of temporary migration, but only for those with weak migrant networks. Consistent with the model, these effects are stronger for permanent migrants, given the greater scope for employer-employee matching and returns to experience. Similarly, networks provide greater assistance for permanent migration, where search costs tend to be higher. But in the presence of moderate to dense networks, workers with more education are actually less likely to move overall, indicating that networks both draw high ability types and dissuade the highly educated. Taken together, these findings suggest that although networks can reduce search costs, particularly for high-ability workers, there are limits to how much they can help. Using an alternative data source (the Pakistan Labor Force Surveys), we show that our findings on education are consistent with a scenario in which networks alert highly educated workers of a low elasticity of demand for skilled labor at destinations, rather than a story of competition within thin markets (Calvo-Armengol 2004).The remainder of the paper is organized as follows: Section 2 presents our theoretical framework, describing the choice between migrating permanently, migrating temporarily, or remaining at the area of origin. Section 3 describes our dataset, and Section 4 outlines our empirical approach. Section 5 presents our main results and robustness checks. Finally, Section 6 concludes.A potential migrant faces a number of destination choices with different combinations of wages, search costs, and moving costs. Conceptually, we often narrow the decision process to a binary choice of whether to take the single best migration option or not. Yet it is becoming increasingly clear that migration options, even in developing countries, are much more heterogeneous and complex than such a simplified framework can explain. In this section, we lay out a basic search model to illustrate the key differences between permanent and temporary migration. We then embed features of a classic Roy  (1951) model to consider how patterns of self-selection may differ with the duration of migration.Consider two different types of migration that can be, and often are, employed by different individuals in the same period or by the same individual in different periods. 1 The first involves a permanent employment opportunity and, with it, a permanent change in residence. This type of migration often occurs over longer distances and requires more up-front investment. The second involves a temporary employment opportunity, necessitating a temporary change in residence but an eventual return to the area of origin. This type of migration typically occurs over shorter distances and is more frequently used as a short-term diversification or risk-coping mechanism, requiring relatively little upfront investment (Bryan, Chowdhury,  and Mobarak 2014). Migration is no longer a binary choice; in each period, an individual considers both opportunities and selects the one that maximizes expected lifetime utility. Given differences between the two types of opportunities, we expect different patterns of selection as well.To formalize this choice process, we employ a standard search model as in Lippman and McCall  (1976). We assume an individual faces a fixed working life, with the length of this period, N, taken as given. Upon reaching adulthood, the individual receives a lifetime wage offer, w 0 , in his or her tehsil of origin (home)-the home wage.2 He or she also has employment opportunities outside of the home tehsil. A temporary employment opportunity with wage w s is available in each period but for only a single period 3 , akin to entering a spot market for labor at the destination. In this case, the individual observes a wage distribution for temporary employment opportunities with cumulative distribution function G t (•). This distribution may vary from period to period,4 with the mean of these temporary wage distributions drawn from a known distribution function,Γ(•). The individual must then incur a search cost, c s , in order to acquire a specific wage draw, and once this cost has been incurred (that is, the individual has moved to the temporary labor market), the temporary wage offer must be accepted. These assumptions are motivated by the observation that migrants often observe labor market shocks at the destination and then engage in a search to secure a specific position and wage after migrating.A permanent offer from outside the home tehsil can also be obtained in each period t ∈ {1,N} by incurring a search cost, c pt . Permanent employment provides a lifetime wage, w p , drawn from a known distribution with cumulative distribution function F (•). In this case, search costs are incurred before the wage draw is received, consistent with the notion that permanent migrants typically identify a (set of) specific employment opportunities before migrating. We assume that the costs for permanent migration are increasing over time (that is, dc pt /dt > 0), reflecting the notion that the search begins with destinations for which the individual has the best information but expands over time to include areas about which less is known. 5 In contrast, the search for temporary migration opportunities is generally concentrated among a small number of destinations that involve similar transportation and transaction costs, and for which the individual has comparable information. We further assume that once a permanent wage offer is accepted, search concludes, consistent with the existence of moving costs and liquidity constraints. Moving costs can be paid out of an initial asset endowment, but financing a second move is prohibitively expensive. 6 Temporary migration relocation may also be costly, but is much less expensive than a permanent move.To identify the optimal migration strategies, we assume that the individual chooses the option in each period that maximizes expected discounted lifetime income, net of search costs. At time t, the individual may work in one of three jobs: home (0), permanent migration, (p), or temporary migration (sfor a single period). While in each job, he or she may also choose whether or not to search for another job. However, once a permanent job has been accepted, the search for permanent offers concludes. The individual may also recall (return to) any previous lifetime employment opportunity.More formally, if the individual has not yet accepted a permanent wage offer, he or she considers the following choices, with the associated value functions:1. Work in the home tehsil and do not search:2. Work in the home tehsil and search:3. Accept a permanent wage offer:4. Accept a temporary wage offer and do not search: E[w s ] -c s + βϕ(w p , t + 1) or 5. Accept a temporary wage offer and search:where β denotes the discount rate, w p denotes the highest permanent wage offer to date, including w 0 , ϕ(w p ,t) denotes the maximum expected payoff to the individual given that he or she has received-but not accepted-a maximum wage offer w p a s of time t, and ϑ (w p ,t) denotes the expected lifetime utility given that a permanent wage offer w p h as been accepted at time t. If a permanent wage offer has already been accepted, then only options (3) and ( 4) are available,7 and the value function for (4) becomesBy comparing the above value functions, we see that a temporary wage offer will be accepted whenwhile a permanent wage offer will be accepted whenThat is, a permanent wage offer will be accepted when the difference between the current wage offer and the home wage, net of search costs, exceeds the expected benefit of continued search, while a temporary offer will be accepted any time the expected earnings, net of search costs, exceed the current wage. The above two expressions highlight the key motivation for distinguishing temporary and permanent migration: the differing behavior of reservation wages over time. The reservation wage for temporary migration is constant over time. In each period, the individual simply evaluates earnings at home (or in another permanent job) relative to the distribution of earnings in the temporary employment market. There are no dynamic considerations involved in the decision because temporary migration is, by definition, a transitory event. In contrast, the search for permanent migration unfolds sequentially, and search costs increase over time as the search expands to include destinations for which the individual has less or worse information. A permanent migration decision is therefore an optimal stopping problem-rather than a single-shot choice-involving sequential search. Moreover, because permanent migration is very costly (irrevocable, in this case), timing is central to the decision process, since the option value of continued search decreases as the number of periods remaining in the individual's working life decreases (Lippman and McCall 1976).One consequence of this distinction between permanent and temporary migration is the need to consider different reference periods. Because the reservation wage for temporary migration offers is time invariant, inferences based on migration activity in a single period of time (for instance, the last five years) will yield similar results irrespective of which \"snapshot\" in time is considered. In contrast, the reservation wage for permanent offers decreases over time because search costs increase over time and because, with a finite horizon, the expected value of continued search decreases over time. The likelihood of permanent migration therefore changes over time, and the age composition of the sample will affect both the number of migrants (within a given time frame) and, potentially, the patterns of self-selection that are observed. Thus, to accurately characterize permanent migration, data on lifetime migration histories is needed, particularly during the early adult years, when permanent migration is more likely.Search costs also differ for permanent and temporary migration. Costs for temporary migration are paid after the migration decision is made, whereas those for permanent migration must be incurred before a wage draw is received. Consequently, higher search costs reduce the likelihood of temporary migration but may either increase or decrease the cumulative probability of permanent migration, conditional on T . Because search costs for permanent migration are sunk, once a wage draw is received, higher costs for continued search encourage acceptance of the current offer, increasing the likelihood of permanent migration, but also discourage continued search, 8 increasing the likelihood that search concludes before an acceptable offer is obtained. The net effect, then, remains an empirical question. 8 The search for a permanent wage offer will cease when the expected value of continued search no longer exceeds the cost. For offers w ≥ y t , the search will continue, where y t is determined by equating the value functions for (1) and (2):There may also be some individuals for which either search costs or the home wage are sufficiently high that search for another permanent offer never commences (analogous to the discouraged worker effect in models of unemployment).To consider the effect of the wage distribution on migration decisions, consider alternative wage distributions F 0 (•) and 0 (•) that first-order stochastically dominate F (•) and (•). Ceteris paribus, the likelihood of receiving an acceptable wage offer will be higher under these alternative distributions for both permanent and temporary migration. But it is important to note that the wage distributions themselves may also reflect variation in search costs. However, observationally, we cannot distinguish between direct effects on the wage distribution (for example, individuals with more education face a higher wage distribution) versus effects of search costs operating via the wage distribution (for example, higherquality information about migration opportunities eliminates the lower tail of the wage distribution, thereby increasing the mean, without reducing the fixed cost of acquiring information), because both alter the expected returns to migration. Thus, we continue to discuss \"search costs\" primarily as the costs associated with identifying migration opportunities while noting that any discussion of wage distributions and expected returns may implicitly involve search costs as well.We next consider the effect of individual heterogeneity-ability, education, and access to migrant networks-on migration decisions. Note that this is not a causal model of the determinants of migration. We use this framework only to illustrate how different types of migration may result in different patterns of selection. Following Chiquiar and Hanson (2005), suppose wages and search costs are functions of individual characteristics as follows:δ is similar to a skill price but generalized to be a function of ability (A), education (E), and migrant networks (N). π represents the exogenous component of search costs (for instance, related to geographic remoteness), and θ is the component of search costs affected by migrant networks. Then, empirically, the effects of ability, education, and networks on migration depend on both the effect of the characteristic on the wage distribution (search costs) and the effect of the wage distribution (search costs) on the migration decision. This formulation is purposely very general because the lack of existing research comparing permanent and temporary migration provides no clear reasons a priori for specifying explicit relationships between individual characteristics and the two different types of migration. Nonetheless, we can provide some preliminary discussion about these relationships, which we will test in the empirical results section.Consider first formal education. Education will affect the distributions of not only permanent and temporary wage offers but also the home wage offer. The relative returns to education in the migrant labor market(s) will then determine migration decisions. We may even observe a nonmonotonic relationship between education and migration if the relevant education-earnings profiles intersect. As in Chiquiar and  Hanson (2005), this result depends not only on the effect of education on wages but also on the relationship between education and search costs. That is, if the returns to education are higher at the area of origin, we will observe strictly negative selection of migrants when π = 0. With only fixed migration costs, and provided the mean difference in wages offsets search costs, less educated individuals will relocate to destinations with lower returns to education. We can observe intermediate selection (migration of only workers in the middle of the education distribution) only when θ > 0, creating the possibility that those in the middle of the education distribution have the greatest incentive to migrate (that is, because they have both relatively low migration costs and less to gain from higher returns to education at the area of origin).9 Indeed, much of the observed relationship between education and migration will likely reflect the effect of education on search costs, given that our data include migration between a large set of locations in the same country, as well as both rural-urban and rural-rural migration, with no clear ordering of the returns to education between origin and destination points.Migration will similarly be influenced by the relative returns to ability between migrant labor markets. However, formal education confers a credential that cognitive ability, on its own, does not, and this may be particularly useful when seeking work opportunities outside the origin tehsil. Comparing the effects of education and ability allows us to learn about the importance of screening in migrant labor markets. Put another way, if education is simply a proxy for ability then education should provide no additional return, conditional on ability. The relationship between ability and search costs conditional on education may also be less strong than that between education and search costs conditional on ability. For example, formal education may convey familiarity with the bureaucratic or administrative processes involved in migration, as well as enhanced knowledge of outside labor markets. It is difficult, empirically, to distinguish between the effects of ability on migration that come through its effects on search costs versus its effects on wage distributions. However, as mentioned above, intermediate selection of migrants would be indicative of a relationship between ability and search costs.Finally, we consider the impact of migrant networks, which can be quite complex. Generally, access to information about migration opportunities and destination labor markets will increase the expected returns to migration, suggesting that networks may directly affect both search costs and wage distributions. Networks may affect fixed search costs, providing benefits across education and ability groups, particularly to those in more isolated markets. Conversely, networks may be selective, meaning either they are composed of individuals with certain characteristics or they support only individuals with certain characteristics. In both cases, the benefits of network access would tend to be concentrated among certain types of individuals, amplifying or dampening the effects of other characteristics, depending on the type of information or support the network provides. In this case, it is more difficult to discern whether network benefits operate through search costs or wage distributions-for example, referrals reduce the time spent locating employment opportunities but may also improve the set of opportunities accessible to the migrant.From our model, we observe that differences in the duration of migration episodes generate important differences in patterns of migration. Short-term migration is a static decision involving only a simple comparison of current wage offers, whereas long-term migration involves dynamic considerations. This difference implies not only that age profiles will differ for each type of migration but also that lifetime migration histories should be examined because permanent moves are more likely to have occurred at younger ages. Moreover, permanent and temporary migrants are likely to differ in demographic composition (age), earnings potential at the destination (long-term versus short-term employment), and remittance behavior. These differences may help explain inconsistent findings in the existing literature regarding the effect of remittances on household investment and well-being. 10 Additionally, the duration of migration episodes may affect patterns of selection. Searching for permanent positions is more costly than is searching for temporary positions. As a result, factors affecting search costs may have a proportionally larger effect on permanent migration. The relationships between education and migration, and between ability and migration, will also be stronger for employment opportunities that make greater use of either skills or screening. In the case of permanent migration, the search is more deliberate, so there is more scope for effectively matching skills with positions, suggesting a strong relationship between education, particularly higher education, and permanent migration. Conversely, the search in a spot market for temporary employment likely involves minimal matching, but employers may be more inclined to utilize observable skills such as education, particularly lower and middle levels, as a screening device.Ability is more difficult to observe than education; it may thus be a less effective screening mechanism for temporary employment, where employees are typically hired with minimal evaluation. However, these factors do not necessarily imply there is no role for ability in temporary migration. Provisional employment opportunities may also be more prevalent in a spot market, allowing (unproven) workers to demonstrate their ability on the job, whereas separation costs (for instance, unemployment benefits) may inhibit such arrangements for longer-term positions.With regard to migrant networks, the information they provide is likely both more cost saving and more useful when a search is conducted over longer distances, both physical and social. For example, economic shocks-such as a construction boom-in potential markets for temporary migration may be easy to observe in comparison with the long-term returns to education/ability that would be relevant for permanent migration. Temporary migrants are also more likely to consider the same set of potential destinations repeatedly, whereas permanent migrants will search sequentially, over increasingly distant markets. Thus, we might expect networks to have a larger effect on permanent migration than on temporary migration. However, another important feature of networks is their capacity to provide referrals. This may involve either facilitating skill-based matching between employers and potential employees in long-term positions, or simply filling temporary labor shortages with network members, or both. With regard to referrals, we may also observe interesting interactions between migrant networks and education or ability, which will depend on the behavior of networks in assisting different types of workers in finding permanent versus temporary positions.Our main source of data is a survey carried out in Pakistan during September 2013-July 2014 that tracked all individuals in a set of households last surveyed in 1991 as part of the International Food Policy Research Institute's (IFPRI's) Pakistan Panel Survey (PPS) (1986-1991). 11 We refer to this latest followup, 22 years later, as the Pakistan Panel Tracking Survey (PPTS). The PPTS survey team visited all 726 households surveyed in 1991, which we refer to as the PPS households. They are spread across five districts: Attock, Faisalabad, and Toba Tek Singh (in the Punjab province); Badin (in the Sindh province); and Lower Dir (in the Khyber Pakhtunkhwa, or KPK, province). Using original (that is, 1991) rosters of all PPS households, the survey team completed a tracking roster documenting all original household members' current whereabouts. Any original member of a PPS household who was alive and residing incountry at the time of the PPTS was eligible for tracking. Once PPS households were contacted and a tracking roster completed for each of them, a current household roster was constructed for each PPS household and each \"split-off\" household formed by an original PPS household member. We complement data from the 1991 PPS and the 2013-2014 PPTS with data from a tracking survey carried out in 2001 by the Pakistan Institute for Development Economics (PIDE 2001); Nayab and Arif (2012)  describes this dataset. This survey team visited each of the original 726 PPS households and completed a tracking roster that noted whether each original PPS household member was present, not present but still considered a household member, or no longer a household member. For those who were no longer members, the survey recorded whether their new location was in Pakistan or abroad.We study the permanent and temporary migration behavior between 1991 and 2013 of male original PPS household members aged 22-60 at the time of the PPTS. These individuals were alive but under age 38 in 1991, allowing us to focus on migration of young, working-age adults. Individuals who joined the PPS household after 1991 or who are members of split-off households are not in our sample. The final sample for analysis consists of 1,346 adult men at risk of migrating permanently during 1991-2013.Figure 3.1 illustrates our success rates in tracking male original members of PPS households who were aged 22-60 at the time of our 2013-2014 PPTS. Of 1,888 men, 208 were international migrants and thus ineligible for tracking. An additional 154 were members of households for which no original member could be traced. This represents a household attrition rate of 4 percent, which is comparable to that of other large panel surveys (Thomas, Frankenberg, and Smith 2001). Excluding these two groups, 180 of 1,526 individuals attrited from the survey between 1991 and 2013-2014, just under 12 percent. Tracking concluded prematurely in August 2014 due to security conditions in the field, which also prevented any tracking of international migrants.12  Individual attrition is of particular concern in the context of a migration study because the majority of attritors are migrants as well. To gauge the severity of this problem, we report in Table A.1 the 1991 characteristics of tracked and untracked respondents. We find few significant differences across groups, except that untracked respondents originate from somewhat wealthier and better-educated households. In contrast, individuals who attrited along with their full households differ greatly from tracked respondents and are worse off overall. They are younger, and their households have less education and wealth, are larger in size, and have higher dependency ratios. These observable characteristics suggest that individuals attriting with their full households are motivated to migrate more by distress than as part of a forwardlooking optimization strategy, distinguishing them from other types of migrants. In light of this observation, and given our model's focus, we omit this group from the analysis, with the caveat that our results cannot be generalized to the case of full household migration.Based on reports from the household head, we can also determine whether an untracked migrant has moved permanently or temporarily. Of the 180 individuals that could not be tracked, 124 were permanent migrants (69 percent), 29 (16 percent) were temporary migrants, and 27 (15 percent) were located in the original 1991 tehsil but refused to respond to the survey. Attrition is largely from permanent migrants, and a larger portion of permanent migrants (46 percent, versus 11 percent for temporary migrants) could not be tracked. This finding is consistent with permanent migrants traveling longer distances, both geographic and social, and therefore being more difficult to locate. The tracking rate for temporary migrants is also higher because those who were reported as temporarily away in earlier survey rounds have returned and did not require special tracking in 2013-2014. In terms of observable characteristics, untracked permanent migrants are slightly younger and come from more educated households, compared to tracked permanent migrants. Untracked temporary migrants are younger and come from smaller, wealthier households when compared to tracked temporary migrants. Given these differences, we perform robustness tests for sample attrition.Our primary interest lies in understanding what leads an individual to migrate, permanently or temporarily, versus not at all. However, since we lack complete data on migration histories, we focus on an individual's first move since 1991. Therefore we divide individuals in our sample into three mutually exclusive, exhaustive categories: those that never migrate, those whose first move was a temporary one (hereafter, temporary migrants), and those whose first move was a permanent one (hereafter, permanent migrants). 13 We define permanent migration as no longer being considered a member of the original PPS household at the time of the 2013-2014 PPTS. We further require a move to be out of the origin tehsil to count it as migration, to avoid capturing local marital moves. Permanent migration may have occurred any time during 1991-2013 but, by construction, can only occur once. We define temporary migration as an individual's being temporarily away-but still considered a PPS household member-at the time of any of our three visits : 1991, 2001, and 2013-2014. 14 The requirement that temporary migrants still be considered members of the PPS household distinguishes them from permanent movers. As an additional safeguard, to ensure that we do not code a permanent migration episode as a temporary one, we further require that an individual who was temporarily away at one of these three points in time had not permanently left the PPS household at the time of a subsequent survey. Specifically, in the case of individuals temporarily away (but still considered household members) in 1991, we require that they still be household members at the time of the 2001 survey. In the case of individuals temporarily away (but still considered household members) in 2001, we require that they still be considered household members at the time of the 2013-2014 survey.15 About 10.9 percent of our sample for analysis is coded as permanent migrants and 18.1 percent as temporary migrants.Labor market potential and search costs are functions of individual skill and ability. To capture skill, we include three categories of education: completed primary education, completed secondary education, and completed higher secondary or tertiary education, with no education or less than primary education as the reference category. 16 We also include a measure of ability that is independent of education: the Digit Span test score. We compute z-scores based on the number of correct responses to 16 forward and backward Digit Span questions. Respondents are provided a set of numbers and asked to repeat the same numbers in the same and opposite order. This tests an individual's attention, memory, and-for the backward Digit Span test-the higher-order cognitive ability to invert the order of information.The overall Digit Span z-score is our primary measure of cognitive ability, but we also examine the robustness of our results to several alternative measures, also converted into z-scores: Digit Span forward, Digit Span backward, standing height, and components of Raven's Progressive Matrices tests (Raven, Raven, and Court 2000). Height has been positively associated in other settings with measures of both cognitive and noncognitive abilities (Vogl 2014; Schick and Steckel 2015) and may visually signal this ability. Raven Test Scores measure abstract reasoning; individuals are shown a series of patterns (matrices) and are asked to select the missing element from a set of eight possibilities.Access to community networks can reduce the costs of migration through the provision of housing, knowledge of labor market conditions, and referrals, among other factors (Carrington, Detragiache, and  Vishwanath 1996; Munshi 2003; McKenzie and Rapoport 2007, 2010). Following Massey, Goldring, and  Durand (1994) and McKenzie and Rapoport (2010), we measure an individual's community migration network using the share of all original PPS village members (excluding members of one's own household) who migrated during 1991-2013-whether permanently or temporarily. We include both male and female migrants because migrants of either gender-and individuals in their new householdsmay assist migrants. This formulation of networks helps to mitigate simultaneity bias because, although it focuses on migrants originating from the same village, it incorporates migration activities over a long time period. As a result, networks are less likely to be correlated with either local economic conditions in the area of origin at the time of migration or composition effects in the labor market driven by recent migration trends. We also examine the sensitivity of our findings to alternative network definitions.We use distance of the PPS village to a primary (that is, paved and high-quality) road in 1994 to account for the effect of search costs on migrant selection (Survey of Pakistan 1994). 17 In selected cases where we lack village GPS coordinates, we assign to the village the centroid of its tehsil, computed from a map of tehsil boundaries (OCHA and PCO 2011) . Additionally, we control for the amount of nonland durable assets owned by the PPS household in 1991, which reflects the household's ability to finance search and moving costs.Our main objective is to test how selection differs for permanent and temporary migrants. Given the limited empirical work on this topic, our model allows for flexible relationships between education, ability, networks, and migration. Migrant selection is modeled via a linear probability model18 with separate regressions for permanent and temporary migration:where Y ihl is an indicator for whether individual i in household h migrated (temporarily or permanently) from village l during 1991-2013; X ih is a vector of individual and household factors that affect the decision to migrate (age categorical variables; an indicator for having ever been married; and tercile categorical variables for the number of household members, number of child members, and value of durable assets in the associated 1991 PPS household); E is a vector of education level indicators; A is a measure of individual ability-our primary measure of interest being the Digit Span z-score; C is search costs, proxied by tercile categorical variables indicating the distance from location l to a primary road in 1994; and N captures the community migrant network, measured by tercile categorical variables for the share of all original PPS village members who migrated during 1991-2013. Given the cross-sectional nature of the analysis, we also include province fixed effects (γ).To the extent possible, we mitigate endogeneity concerns by using values from 1991, before migration occurred. Including a direct measure of cognitive skill also helps to reduce ability bias for schooling. Still, we cannot purport to estimate causal effects of education and migrant networks, given the simultaneity of these decisions and migration. We would argue, however, that descriptive analysis is essential to both future research and policymaking. First, there is little to no empirical evidence on how permanent and temporary migrants differ and to what extent. Second, in order to understand how expanding opportunities for migration will affect the composition of migrant cohorts and the impact on destination economies, it is critical to fully understand descriptive patterns of selection. 19 Similarly, understanding selection patterns is essential to estimating the impact of migration on origin communities and households. 20 Our analysis thus provides an important starting point for deepening understanding of the migration decision and its impact on local economies.Although we use the same specification for permanent and temporary migration in order to facilitate comparison, there are differences in the underlying relationships being estimated. As noted above, permanent migration is a sequential decision, while temporary migration can be represented by a static problem. Thus, for permanent migration, we look at the individual's entire migration historythat is, the cumulative probability that the individual has accepted a permanent migration opportunity by time T . 21 In contrast, for temporary migration, we look at the probability that an individual has21 This can be written asandwhere D is the period in which the search concludes (search costs equal the option value of continued search), determined as follows: accepted a temporary employment opportunity in one of three time periods-1991, 2001, or 2013-2014. 22 When we look at the Pakistani Labor Force Survey, we find approximately 2 percent of males aged 22-60 reported as \"temporarily away\" in each year, compared with 18 percent of our sample that has engaged in temporary migration in at least one of the three time periods. This comparison suggests that our measure does not substantially undercount temporary migrants, though we do miss temporary migration episodes in the intervening years. Since the temporary migration decision is essentially the same in each period, however, our results should not be affected by the choice of periods or by limiting our attention to only three periods. Nonetheless, we test this assumption more carefully below.Summary statistics are presented in Table 5.1. Permanent migrants are older than nonmigrants, and the opposite is true for temporary migrants. Both temporary and permanent migrants appear to have certain advantages for migration. They are, on average, closer to a primary road and have denser migrant networks relative to their nonmigrant counterparts. Both migrant types also have higher cognitive scores than nonmigrants, and a greater proportion of temporary migrants have completed secondary education or more. Among permanent migrants, a greater proportion have also completed education beyond secondary school, but the proportion completing secondary school is comparable to that of nonmigrants. Permanent migrants also come from households with greater asset wealth per capita, while the opposite is true for temporary migrants, though the latter come from larger households, on average. Table 5.2 provides ordinary least squares results from regressions of permanent (columns 1-4) and temporary (columns 5-8) migrant selection. We estimate four stepwise regressions, starting with the inclusion of indicators for the province of the household member's origin village and the individual's age and marital status (columns 1, 5). We then add individual education and cognitive ability (columns 2, 6). Next, we introduce the individual's 1991 PPS household demographics, wealth level, and distance to a primary road (columns 3, 7), and finally we add community migrant networks (columns 4, 8). The final column provides the difference in each estimated coefficient across temporary and permanent migrant selection equations, comparing columns 4 and 8. The p-value from a test for the statistical significance of this difference is presented in brackets below the difference.The likelihood of permanent migration is increasing in age, consistent with the decreasing reservation wage in our model. Interestingly, the oldest group in our sample has the highest likelihood of permanent migration, and we see no evidence of leveling off, which would occur if the search for permanent wage offers ceased at some age below the upper bound of our sample (60). In contrast, we observe a flattening of the age profile for temporary migration because the increasing likelihood of permanent migration supersedes the (constant) likelihood of temporary migration. 23 These coefficients are stable across specifications and consistent with the raw sample means, suggesting minimal correlation between age and other regressors.Our theoretical model allows both search costs and the returns to migration to vary with worker attributes. As a first pass, we enter education and ability linearly in the migrant selection equations. Cognitive ability positively affects both forms of migration; a 1 standard deviation increase in an individual's Digit Span z-score increases the likelihood of both permanent and temporary migration by 6.7 percentage points relative to sample means of 11 percent and 18 percent for permanent and temporary first movers, respectively. The similarity of the point estimates is striking although, proportionally, the effect is considerably larger for permanent migration. This is consistent with higher rewards for cognitive skills in long-term than in short-term employment opportunities (for example, due to employee-employer match quality, firm-or industry-specific human capital, and so on.). In contrast, education has a statistically insignificant and small impact on migration. This finding suggests that education is strongly correlated with innate cognitive ability-consistent with a standard ability-bias story-and likely other covariates as well (for instance, wealth, household composition). Together, these estimates suggest that conditional on innate cognitive ability, there is little to gain in migrant labor markets from formal education, at least in this context. This implication may indicate either low-quality schooling or limited demand for educational signaling and screening in migrant labor markets.  Next, we examine how search costs and the ability to finance them affect migrant selection, drawing on the parameter estimates of distance to a primary road, initial household wealth, demographics, and networks. First, residing in a location very far from a primary road (in the third tercile) is associated with a statistically significantly lower probability of migrating permanently (-0.048), compared with a statistically insignificant increase in the probability of migrating temporarily (0.094). Geographic isolation substantially increases the search costs for permanent offers, reducing this type of migration. In contrast, lucrative, less distantly located temporary opportunities seem to mitigate higher transportation costs.Assets owned by the 1991 PPS households are found to have no significant effect on either permanent or temporary migration. These estimates suggest that credit and liquidity constraints do not play a substantial role in migration decisions. 24 This finding stands in contrast to the raw means, which indicated that permanent migrants originated from significantly wealthier households, and suggests substantial correlation between initial wealth and other covariates (for instance, age, education, household composition), particularly among permanent migrants.Temporary migration alone is positively influenced by the 1991 PPS household size. One might expect larger households to facilitate migration by allowing its fixed costs to be spread over more people. However, that is assuming the origin household benefits from the migration decision. Temporary migration may substantially benefit the origin household, given the relatively high likelihood of remittance receipts and the eventual return of the migrant (Stark and Lucas 1988). In contrast, the benefits of permanent migration may accrue predominantly to the migrant; as a result, certis paribus, having more household members is less likely to influence the migration costs faced by the migrant.Finally, we observe that migrant networks significantly increase the likelihood of both types of migration, but only when networks are sufficiently dense (in the third tercile). This suggests that as expected, networks assist new migrants in locating employment, either by reducing search costs or by improving the distribution of prospective offers. We cannot reject that being in the third tercile of networks has the same impact on both types of migration. However, given the means of the two migration variables, proportionally this effect is far larger for permanent migration, which we would expect to have higher overall search costs.Our model suggests important complementarities between networks and individual characteristics. We thus add interactions of network terciles with the education, ability, and road access variables. Conducting separate F tests to examine the joint significance of search costs and their interaction with networks (see Table 5.3), we find evidence of marginally significant complementarities between network size and search costs for permanent migration (though not for temporary). The search costs-captured by the distance to a primary road, a measure of geographic isolation-that previously reduced selection into permanent migration are almost perfectly offset by the presence of dense (third-tercile) networks, indicating that strong migrant networks substantially mitigate search costs. For the bottom tercile of networks, however, search costs significantly reduce permanent migration. Because only very dense networks mitigate the search costs associated with permanent migration, and only for those who are the most geographically most isolated, networks seem to confer (costly) information about job opportunities-particularly for long-distance, permanent moves. In contrast, if networks simply provided referrals, we would not expect their benefits to vary with the search costs implied by an individual's location. Further, our estimates suggest that moderate networks are insufficient to mitigate search costs, nor do networks significantly reduce search costs for those who are less geographically isolated.   Complementarities between networks and skill are strong and offer perhaps the most interesting motivation for empirically distinguishing between temporary and permanent migration. As shown in Table 5.3, high-ability individuals are more likely to engage in permanent migration only when they have access to dense (in the third tercile) migrant networks. Again, this effect is much larger in magnitude for permanent migration, which tends to entail greater returns on ability. Conditional on having a network size within the third tercile, a 1 standard deviation increase in the Digit Span score increases the likelihood of permanent migration by 14.7 percentage points, compared with the sample mean of 18.1 percent. But in the absence of strong networks, ability has no significant effect on either type of migration. Nor do networks confer benefits irrespective of ability; the direct effect of networks on those with low ability is much smaller in magnitude and not statistically significant. These nuanced impacts may reflect strategic behavior on the part of the network, such as an effort to safeguard network quality by providing costly job information and referrals only to high-ability workers. Alternatively, strong networks may provide information, but only high-ability people can benefit from it.Interactions between education and networks are more complex. In our basic specification (Table 5.2), education did not significantly affect either type of migration. But when we add interactions with networks, we find statistically significant evidence of positive selection among permanent migrants and intermediate selection among temporary migrants-though only in the case of weak migrant networks. In contrast, for moderate or dense networks, we find significant negative complementarities between networks and formal education that exceed the direct effects of education for both temporary and permanent migration. Those with more than secondary education are 13.2 percentage points more likely to engage in permanent migration, all else equal. But if they also have access to moderate (second-tercile) or dense (third-tercile) networks, they are 5.0 and 8.1 percentage points, respectively, less likely to migrate permanently than an individual with less than primary education and weak networks. In the case of temporary migration, individuals with secondary education are 17.1 percentage points more likely to migrate when they have weak networks. But if they also have access to moderate (second-tercile) networks, they are 3.6 percentage points less likely to migrate.What emerges from these results is a nuanced role for networks in the job search process. Without strong networks, ability has no significant effect on migration, while the opposite is true for education. These opposite findings are intuitive; while ability may be the main determinant of worker productivity, education is much more readily observable to employers. Therefore, our results suggest that, when entering a new labor market, those with weak networks must rely more heavily on credentials than on underlying ability. Put differently, potential migrants cannot rely on innate ability without either networks as a complement or credentials as a signal. Conversely, moderate and dense networks assist in migration for high-ability workers yet appear to deter the migration of highly educated workers.The results offer interesting insights about the dynamics of migrant labor markets and networks in Pakistan. First, it is possible that the returns to tertiary (secondary) education are actually low for permanent (temporary) migrants. Individuals with weak networks may have less accurate priors about these returns, resulting in more migration than is optimal. Second, networks may be less willing to assist the highly educated in pursuit of migrant employment. This could occur if there are local concerns about \"brain drain\" or if there is internal competition within a thin labor market (Calvo-Armengol 2004) such that additional migration of educated workers will significantly reduce expected wages. While a complete analysis of the returns to education is beyond the scope of this paper, the latter is consistent with cursory evidence from data in several Pakistani Labor Force Surveys. In Table 5.4, we report regressions of wages and unemployment on age and education for males aged 22-60. Tertiary education provides no wage gains above and beyond those of secondary education and carries a significantly higher probability of unemployment. Although studies describing the labor markets in Pakistan are limited, recent case studies suggest a trend of increased tertiary education in conjunction with overqualification in occupational choices (Farooq, Ahmed, and Ali 2008; World Bank 2013). This situation leads to low returns and high unemployment. Below, we explore how network quality influences these relationships.  1986-1991 (IFPRI 1991), PIDE 2001 (Nayab and Arif 2012), and Pakistani Panel Tracking  Survey 2013-2014 (IFPRI 2014). Note: District-clustered standard errors in parentheses. * p < 0.1 ** p < 0.05; *** p < 0.01. District and year fixed effects included.We first provide regression estimates including untracked migrants to demonstrate the sensitivity of our estimates to their exclusion. For untracked individuals, we lack two explanatory variables: education 25 and the Digit Span score. However, we have data from their origin households on other variables. Our estimates are reasonably stable when attritors are added to the sample (Table A.3). 26 These findings suggest that the results may not be driven by sample attrition, with one exception pertaining to the age profile of permanent migrants. Once attritors are included, we observe no significant differences in permanent migration across age groups, although our main sample displays a significant positive relationship with age.We next examine how our results may be affected by selective attrition using the inverse probability weights recommended by Fitzgerald, Gottschalk, and Moffitt (1998). 27 The weighted regression version of Table 5.3 is shown in Table A.5. Inferences are similar when accounting for attrition, with the exception that the negative effect of being 35 to 44 years old gains significance in the temporary migration regression. The weighted specification appears to improve the overall precision of the coefficient estimates with negligible consequence on interpretation.We replicate the basic permanent and temporary migration regressions (omitting interaction variables) with alternative measures of cognitive ability: forward Digit Span, backward Digit Span, total Raven score (sections A, B, and D), Raven A, Raven B, Raven D, and height z-score. The ability parameter and standard error estimates from each specification are reported in Table 5.5. In none of the models can we reject that the estimated parameters on the ability measures are statistically different across permanent and temporary migration-consistent with our findings in Table 5.2 using the overall Digit Span. Variation in the more challenging component of the Digit Span test-backward memorization-appears to be driving the ability estimates. Scores from section A of the Raven test corroborate a positive effect on both forms of migration, although it is only statistically significant in the temporary migration equation. Height-a visual measure of ability-does not predict either form of migration, controlling for education. This result suggests that nutrition and educational outcomes are likely jointly determined by the household (Vogl  2014). It also suggests that migrant labor markets do not reward \"brawn\" over \"brains\" (Pitt, Rosenzweig,  and Hassan 2012). 25 We are missing education levels for a fraction of the untracked migrants because many were young when last surveyed in 1991. For those with some education in 1991, we are unable to infer what their completed education levels were in 2013-2014. 26 The 180 untracked individuals in Figure 3.1 can be stratified into 124 permanent migrants, 29 temporary migrants, and 27 nonmigrants based on additional information on the tracking rosters. We add 124 permanent migrants and 27 nonmigrants to the original permanent migrant sample (originally 1,103 observations). We add 29 migrants and 27 nonmigrants to the temporary migrant sample (originally 1,198 observations). 27 We estimate restricted and unrestricted (with supervisor indicators and village attrition rate as instruments) probit regressions to formulate inverse probability weights. Table A.4 shows the unrestricted results. We next use the alternative measures of ability in the interacted model. Table A.6 measures ability with the three Digit Span scores and height, while Table A.7 measures ability with Raven scores. In each table, we display the estimated parameters and standard errors for the ability, education, and network variables. Overall, these estimates are consistent with our main findings above. The direct effect of cognitive ability is not statistically significant for any measure. The positive and significant interaction between ability and networks is quite robust to the measures derived from the Digit Span test, but not when using components from the Raven test or height. The direct effects of education are also found to be positive and significant, as are the negative and significant interactions between education and networks. Both sets of effects are robust to all specifications of cognitive ability. However, the coefficients related to more than secondary education are significant only when overall or forward Digit Span is used, implying stronger correlation between high levels of education and more challenging cognitive tests (backward Digit Span, Raven).These findings, though consistent with Kaestner and Malamud (2014), may be specific to prevailing labor market conditions. That is, the Digit Span test is largely considered a \"simple\" task that measures short-term recall and working memory, while Raven's Progressive Matrices measure fluid intelligence involving inductive reasoning (Fry and Hale 1996). The Digit Span, therefore, may be more relevant for work requiring the learning and repetition of simple to moderately difficult tasks without higher-order cognitive processing, while the Raven test would be more relevant for highly skilled positions involving a great deal of inductive reasoning. Moreover, formal education may have a larger effect on test scores for fluid intelligence than those for memory (Jaeggi et al. 2008), suggesting that the effect of Raven scores will be minimal when controls for education are also included. In light of these differences, future surveys should consider carefully the choice of cognitive tests.Our last set of specifications explores the implications of using narrower definitions of the migrant network: focusing only on network members who (1) have an above-median Digit Span score, (2) have completed secondary education or higher, and (3) are above the sample median age. The results for permanent migration (Table A.8) are very similar in sign and significance to our main results (Table 5.3), suggesting that network-skill interactions are not driven by the quality of the network. Network quality does magnify the effects, with point estimates considerably larger. However, the direct effect of education is no longer statistically significant and is much smaller in magnitude. This may point to a correlation between network quality and formal education.The findings for temporary migration are sensitive to the choice of network definition. A positive interaction between ability and networks is now evident only for high-ability networks, and the point estimates are twice as large as in Table 5.3. High-ability networks appear to strongly support temporary migration for high-ability workers, but high-education networks do not do the same, nor do they deter migration among the highly educated. Taken together, these results suggest that with regard to temporary migration, it is the low-ability, low-education network members that deter migration among the highly educated. 28 Finally, as a proxy for the tenure of the network contacts, we exclude from the network community members who are below the sample median age. 29 These contacts have more experience in the destination location to improve a potential migrant's employment prospects, and are arguably less vulnerable to pressures of competition. We find that older contacts continue to drive the permanent migration of highly able workers. Yet the negative selection previously observed among highly educated workers with larger networks is smaller in magnitude (though imprecisely estimated). These findings confirm that the network-education dynamics are likely driven by the inelastic demand for skilled labor in permanent migrant labor markets. However, tenured networks can soften the selection effects associated with highly educated workers. 28 One possible explanation for this result is that low-skilled networks may understate the returns to education in temporary employment opportunities. Educated workers will formulate beliefs based on this information, and given their pessimistic income expectations they may be less likely to migrate. Given a similar skill profile, those with weak networks will have noisy but unbiased beliefs about the returns to education. 29 Migrant age is used as a proxy for migrant tenure (years away from the origin household) because we lack data on the year and duration of temporary migration episodes, as in Munshi (2003) and Beaman (2012).Existing literature on migration largely fails to separately model two distinct types of migrants: temporary and permanent. We address this by utilizing a search model to illustrate the two migration decisions and then using a unique panel dataset with detailed migration information from rural Pakistan spanning 1991-2013 to test the model. Patterns of selection are found to vary with the duration of the migration episode, consistent with differences in search costs and the scope for employer-employee matching and returns to experience.We find clear evidence of positive selection on skill among permanent migrants, suggesting that skill is a key factor in obtaining good permanent wage offers. Networks appear to facilitate job matching in the permanent migrant labor market for those with strong cognitive skills, perhaps to improve network quality or perhaps because other labor markets are relatively easy to enter without network assistance. Conversely, temporary migrants display much weaker selection on skills, consistent with the idea that since temporary migration opportunities are located through a spot market, it is difficult to signal skill, resulting in relatively low returns. The links between migration, education, and networks, however, are found to be much more nuanced and highly dependent on prevailing labor market conditions. We also find that search costs-proxied by distance to a primary road-significantly reduce permanent migration, though only in the absence of strong networks. Strong networks nearly perfectly offset the negative impacts of search costs on permanent migration. Overall, these findings underscore the importance of studying temporary and permanent migration separately. The duration of the migration episode has important implications for the type of individual that chooses to migrate, meriting greater attention in the literature.    ","tokenCount":"9898","images":["-1187811762_1_1.png","-1187811762_16_1.png"],"tables":["-1187811762_1_1.json","-1187811762_2_1.json","-1187811762_3_1.json","-1187811762_4_1.json","-1187811762_5_1.json","-1187811762_6_1.json","-1187811762_7_1.json","-1187811762_8_1.json","-1187811762_9_1.json","-1187811762_10_1.json","-1187811762_11_1.json","-1187811762_12_1.json","-1187811762_13_1.json","-1187811762_14_1.json","-1187811762_15_1.json","-1187811762_16_1.json","-1187811762_17_1.json","-1187811762_18_1.json","-1187811762_19_1.json","-1187811762_20_1.json","-1187811762_21_1.json","-1187811762_22_1.json","-1187811762_23_1.json","-1187811762_24_1.json","-1187811762_25_1.json","-1187811762_26_1.json","-1187811762_27_1.json","-1187811762_28_1.json","-1187811762_29_1.json","-1187811762_30_1.json","-1187811762_31_1.json","-1187811762_32_1.json","-1187811762_33_1.json","-1187811762_34_1.json","-1187811762_35_1.json","-1187811762_36_1.json","-1187811762_37_1.json","-1187811762_38_1.json","-1187811762_39_1.json","-1187811762_40_1.json","-1187811762_41_1.json","-1187811762_42_1.json","-1187811762_43_1.json","-1187811762_44_1.json","-1187811762_45_1.json"]}

data/part_2/0038174167.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"3d9841a242de7467f71b3c43d73d77b1","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/6521a9e4-6208-4946-9641-5f872d1feb4c/retrieve","description":"","id":"-651956210"},"keywords":[],"sieverID":"2996fe53-a8ff-41c4-9868-73a5af248f17","pagecount":"4","content":"District Nutrition Profiles (DNPs) are available for 707 districts in India. They present trends for key nutrition and health outcomes and their cross-sectoral determinants in a district. The DNPs are based on data from the National Family Health Survey NFHS-4 (2015NFHS-4 ( -2016) ) and NFHS-5 (2019NFHS-5 ( -2021)). They are aimed primarily at district administrators, state functionaries, local leaders, and development actors working at the district-level.  Note: NA refers to data unavailable for a given round of NFHS/Census.• What are the trends in infant and young child feeding (early initiation of breastfeeding, exclusive breastfeeding, timely initiation of complementary feeding, and adequate diet)? What can be done to improve infant and young child feeding? • What are the trends in IFA consumption among pregnant women in the district? How can the consumption be improved?• What additional data are needed to understand diets and/or other determinants?  Note: NA refers to data unavailable for a given round of NFHS/Census.• How can the district increase women's literacy, and reduce early marriage, if needed?• How does the district perform on providing drinking water and sanitation to its residents? Since sanitation and hygiene play an important role in improving nutrition outcomes, how can all aspects of sanitation be improved? • How can programs that address underlying and basic determinants (education, poverty, gender) be strengthened?• What additional data are needed on food systems, poverty or other underlying determinants? Note: NA refers to data unavailable for a given round of NFHS/Census.• How does the district perform on health and nutrition interventions along the continuum of care? Does it adequately provide both prenatal and postnatal services to women of reproductive age, pregnant women, new mothers and newborns? • How has access to health and ICDS services changed over time (food supplementation, health and nutrition education and health checkups)?","tokenCount":"300","images":["-651956210_1_1.png","-651956210_1_2.png","-651956210_1_3.png","-651956210_1_4.png","-651956210_1_5.png","-651956210_1_6.png","-651956210_1_7.png","-651956210_1_8.png","-651956210_1_9.png","-651956210_1_10.png","-651956210_1_11.png","-651956210_1_12.png","-651956210_2_1.png","-651956210_2_2.png","-651956210_2_3.png","-651956210_2_4.png","-651956210_2_5.png","-651956210_2_6.png","-651956210_2_7.png","-651956210_2_8.png","-651956210_2_9.png","-651956210_2_10.png","-651956210_2_11.png","-651956210_2_12.png","-651956210_3_1.png","-651956210_3_2.png","-651956210_3_3.png","-651956210_3_4.png","-651956210_3_5.png","-651956210_3_6.png","-651956210_3_7.png","-651956210_3_8.png","-651956210_3_9.png","-651956210_3_10.png","-651956210_3_11.png","-651956210_3_12.png","-651956210_4_1.png","-651956210_4_2.png","-651956210_4_3.png","-651956210_4_4.png","-651956210_4_5.png","-651956210_4_6.png","-651956210_4_7.png","-651956210_4_8.png","-651956210_4_9.png","-651956210_4_10.png","-651956210_4_11.png","-651956210_4_12.png"],"tables":["-651956210_1_1.json","-651956210_2_1.json","-651956210_3_1.json","-651956210_4_1.json"]}

data/part_2/0061512578.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"56ad90b7d2441db9965740c2700d3df0","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/d8faafae-478f-4b82-a161-8d416fa69409/retrieve","description":"Agricultural productivity increases are one of the desired outcomes from sensible food security and agricultural policies. Increased productivity might lead to improved welfare of rural populations through several pathways. First, increased productivity ensures higher food availability and higher incomes at the farm household level. Second, increased food availability leads to lower prices of agricultural products and higher real wages, to the benefit of poor net buyers and wage laborers respectively. Third, a well-performing agricultural sector has important economic multiplier effects on the vibrancy of the off-farm rural economy. However, policy makers are often hampered by a lack of information on agricultural productivity, the constraints that farmers face, and the levers that they can use to improve productivity.","id":"939806089"},"keywords":[],"sieverID":"afa0bf4b-f7c6-49e2-9f3c-92448007e517","pagecount":"62","content":"The report was written by Bart Minten (IFPRI) and Chencho Dukpa (CoRRB). It benefitted from discussions with other members of the MoAF-IFPRI project team: Nicholas Minot (IFPRI), Kailash Pradhan (CoRRB), Phub Dem (DAMC), and Nidup Peljor (PPD). The authors would like to acknowledge the data management assistance of Reno Dewina (IFPRI). They would like to thank for the support and assistance received from the IFPRI office based at New Delhi and Washington DC (USA), the Ministry of Agriculture and Forests (MoAF) and the Council for Renewable Natural Resources Research of Bhutan (CoRRB). The authors are grateful to the National Statistical Bureau (NSB) which generously provided data from the Population Census, the Bhutan Living Standards Surveys, and other datasets. They would also like to thank Karpo Dukpa (Statistical Investigator, PPD) for providing the 2009 RNR Census data for analysis and Vaishali Dassani and Task Digital for coordinating the design, the formatting, and the printing of this booklet. Finally, the paper benefited from constructive feedback from participants at an interim workshop on 14 January 2010 and the final project workshop on 2 July 2010, where earlier versions of the paper were presented.However, any opinions stated in this report are only those of the authors and do not necessarily reflect the policies or opinions of the SDC, the RGB, or IFPRI.Agricultural productivity increases are one of the desired outcomes from sensible food security and agricultural policies. Increased productivity might lead to improved welfare of rural populations through several pathways. First, increased productivity ensures higher food availability and higher incomes at the farm household level. Second, increased food availability leads to lower prices of agricultural products and higher real wages, to the benefit of poor net buyers and wage laborers respectively. Third, a well-performing agricultural sector has important economic multiplier effects on the vibrancy of the off-farm rural economy. However, policy makers are often hampered by a lack of information on agricultural productivity, the constraints that farmers face, and the levers that they can use to improve productivity.In this study, we look at patterns and determinants of agricultural productivity and technology adoption in Bhutan. We also look more in detail at the effects of road infrastructure and remoteness on agriculture given the pledge of the new government to improve connectivity of the geogs in the country. We rely on data from the Renewable Natural Resource (RNR) Census fielded in the beginning of 2009 by the Policy and Planning Division (PPD) of the Ministry of Agriculture and Forests (MoAF). Data were collected concerning land holdings and tenure, crop and livestock productions, forest production and utilization, agricultural inputs, accessibility, quantities marketed and prices received, and farming constraints. 57,606 households were interviewed, covering about 93% of all the households in Bhutan that are involved in farming, livestock, or forestry activities.Several important insights emerge from our analysis. First, agriculture is characterized by a low level of usage of modern inputs and mechanization. Chemical fertilizer and plant protection chemicals are used by 33% and 16% of the farmers respectively and only 10% of the farmers use power tillers for plowing. The low level of modern technologies leads to relatively low productivity. Our results show that there are important productivity effects and high rates of returns of the increased use of improved agricultural technologies such as chemical fertilizer or plant protection chemicals. While not studied in this report in particular, there might be further benefits to changes in seed markets and to a bigger emphasis on hybrid seeds (e.g. for the rice sector) which until now are little used in Bhutan but which have shown to lead to significantly higher yields in neighboring India.Second, there are strong spatial patterns in agricultural performance and technology adoption. Some of these effects might be related to inherent infrastructural or institutional characteristics but some of these are also strongly related to climatic and soil differences. The results indicate that the Western and Westcentral part of the country are doing well in terms of productivity per household or per unit of land while the Eastern part and the Southern part of the country seem to be doing less well. The spatial analysis also shows strong differential spatial constraints in agriculture with crop damage due to wildlife and plant diseases more prevalent in the Eastern part of the country and land shortage and water constraints more complained xiv about in the Western and Southern part of the country. This thus calls for spatially differentiated agricultural development strategies.Third, roads as well as the quality of land (access to irrigation) are strongly related with productivity and technology adoption. Controlling for other factors, household production is 20% lower for those households that have to walk for more than 1 day to a motorable road compared to those households located close to a motorable road. Distance to roads further show important effects on the likelihood to adopt different types of improved technologies, despite the policy of pan-territorial pricing of inputs. However, the results also show that investments in infrastructure are a necessary, but not sufficient, condition as adoption rates and agricultural productivity might still be low in areas that are well-connected. In simulations presented in this report, we show that the costs of road investment are overall probably prohibitively high to justify towards stimulation of the agricultural sector. Complimentary investments in other infrastructure and in research and extension might improve agricultural productivity more efficiently.Fourth, the size of the farm is in general small in Bhutan (the average size of cultivated land is evaluated at 1.2 hectares per farm) and the distribution of land in the international context is rather equal. Smaller farms depend relatively more on the lower quality dryland while bigger farms have better access to irrigated land as well as orchards. Especially orchards are in the hands of larger farmers as we find that almost 60% of all the orchard land is held by 10% of all orchard farmers. Interestingly, despite their lower land quality, smaller farmers are able to achieve much higher land productivity than larger farmers as their land productivity is evaluated to be almost three times as high.Fifth, a major problem in agricultural production in Bhutan is crop damage by wildlife and plant diseases. Respectively 25% and 20% of the farmers consider these as the most important constraints to their agricultural activities. They are ranked higher than problems with irrigation or labor. Especially maize, an important crop in the East, seems to suffer most of the losses due to wildlife. We further find that areas kept in fallow are quite significant. While fallow land would be part of a shifting cultivation pattern, the major reported reason for why farmers leave land fallow is linked to potential wildlife damage. The results also show that 3,400 hectares of wetland are not used, an important consideration for policy makers given their aim to improve the rice selfsufficiency in the country.Several different pathways can be followed to improve agricultural performance in Bhutan. The adoption of better agricultural technologies can be enhanced by ensuring productive investments, such as roads, irrigation investments, and availability of appropriate agricultural inputs developed by a suited R&D system. Another pathway is the subsidization of agricultural inputs such as chemical fertilizers, electricity, and irrigation facilities. Experiences in other countries have shown that the returns to the productive investments are significantly higher than to subsidies and thus productive investments have generally higher rates of returns and are thus preferred.Agricultural productivity increases are one of the desired outcomes of sensible policies to improve food security and reduce poverty in rural areas in developing countries. Increased productivity might lead to improved welfare of rural populations through several pathways. First, increased productivity ensures higher food availability and higher incomes at the farm household level. Second, increased food availability leads to lower prices of agricultural products and higher real wages, to the benefit of poor net buyers and wage laborers respectively. Third, a well-performing agricultural sector has important economic multiplier effects on the vibrancy of the off-farm rural economy. The strong effect of agricultural performance on poverty is typical for a large number of developing countries, as shown for example by Christaensen et al. (2010), Datt andRavallion (1998), andDe Janvry andSadoulet (2002). They show that poverty reduction elasticities from agricultural growth are significantly larger than for other sectors of the economy, usually because of large employment of poor people in the agricultural sector.Growing emphasis on increasing the investments to improve agricultural productivity worldwide is one of the consequences of the global food crisis at the end of first decade of 2000, as policy makers realized that the era of cheap food might be coming to an end if required attention is not given to the agricultural sector. However, policy makers are often hampered by a lack of information on agricultural productivity, the constraints that farmers face, and the levers that they can use to improve productivity. These themes are the subject of this report.Bhutan is characterized by a largely rural population that makes its livelihood mainly from agriculture (about two-thirds of Bhutan's population does so). This rural population often also lives in rather remote areas hampering its access to agricultural input and output markets. It is estimated that almost half of the rural population lives in villages that are more than an hour walking distance to a motorable road. The government is committed to improving their lot by trying to connect all geog centers of the country. Given the importance of access to infrastructure for the agricultural performance of small and poor farmers, and thus for policy purposes, we present an analysis on the effect of remoteness on technology adoption and agricultural performance.To study the topics of agricultural performance, agricultural productivity, and agricultural technology adoption, we rely on recent data, collected in the beginning of 2009, from an agricultural census in Bhutan. The aim of the census was to interview all rural households that are involved in farming, livestock, or forestry and inquire about their activities in the year prior to the survey. Relying on these data, we will use different analytical methods to study this theme and we report the results of cross-tabulations, regression analysis, and mapping exercises.The overall purpose of the study is to provide an updated analysis on agricultural productivity and technology adoption in Bhutan. As the intended audience of this report is diverse, including researchers, policy makers, donors, and other stakeholders in agricultural and rural development in Bhutan, technical discussions alternate with less technical and more descriptive analysis and mapping exercises. It is hoped that this makes the report accessible to all.The structure of the document is as follows. In Section 2, we present the data and the methods that are used to analyze these data. Section 3 discusses the results of the productivity analysis. In Section 4, we look at four different inputs that are important in the agricultural production process, i.e. land, fertilizer, plant protection chemicals, and agricultural equipment. Section 5 describes the different perceived constraints in agriculture as reported by the rural households. An analysis on crop damages will also be presented. In Section 6, we present the results of an analysis on the association of remoteness with agricultural performance. We finish with the summary and conclusions in Section 7.We rely on data from the Renewable Natural Resources (RNR) Census that was fielded in the beginning of 2009. This RNR Census was coordinated by the Policy and Planning Division (PPD) of the Ministry of Agriculture and Forest (MoAF). About 700 enumerators and 60 supervisors were employed to implement the survey. 57,606 households were interviewed, covering 93% of all households in Bhutan that are involved in farming, livestock, or forestry activities. The data that were collected concerned land holdings and tenure, crop and livestock productions, forest production and utilization, agricultural inputs, accessibility, quantities marketed and prices received, and farming constraints. In the analysis, extrapolation coefficients were used that allow for the estimation of these statistics at the national level.In the report, we mostly rely on the information that was collected regarding crop production and agricultural inputs. Most of the descriptive statistics will be presented by farm size (5 quintiles ranging from the smallest size (quintile 1) to the largest size (quintile 5)) and by distance to a motorable road (for five categories, going from less than 1 hour, 1-3 hours, 3-6 hours, 6 hours-1 day, and more than 1 day). In the productivity and technology adoption analysis, we only use information from those households that reported land cultivation. As some households did not report this information (8,066 households out of 57,606), the reported statistics here are slightly different than those reported in the official census report (MoAF, 2010). Some of the variables are also mapped in the report. As we only had access to an older ArcView map of Bhutan and as some of the geogs have been split since the creation of the maps, statistics were calculated at the level of these old geogs and are presented that way.Assumptions had to be made for the construction of some of the variables used in the analysis. First, all information from annual and perennial crop production is included in our production and productivity measures. However, livestock production is excluded from this valuation. To allow aggregation of production over different crops for the estimation of overall productivity measures, production was valued at actual household level prices or if retained for home consumption at average geog level prices.1 Second, no areas were reported for fruit production and assumptions, based on expert opinions of extension agents and other stakeholders, were used on the number of trees per unit of land as to be able to aggregate the area under fruits trees with other crops. Third, to allow crop information to be used in regression analysis in a manageable manner, it was aggregated at different levels. Following the convention in the Ministry of Agriculture and Forests, they include cereals (barley, buckwheat, finger millet, foxtail millet, maize, paddy irrigated, paddy upland, sweet buckwheat, and wheat), vegetables (asparagus, broccoli, cabbage, carrot, cassava, cauliflower, collacacia, cucumber, egg plant, guards, lady finger, potato, pumpkin, radish, spinach, squash, sweet potato, tomato, tapioca, turnip, cabbage), spices (cardamom, chili, garlic, ginger, onion), grain legumes (beans, dhal, mung bean, peas, rajma bean), and oilseeds (ground nut, mustard, sesame, soyabean, sunflower). Fourth, a variable was constructed that approximates the risk of crop damage due to wildlife attacks that farmers in a particular geog are exposed to. To do so, the share of farmers in a geog that ranked this as one of the three most important constraints for his agricultural activities was calculated.We rely on three types of statistical and econometric analysis. First, to model agricultural productivity at the household level, a production function can be written as follows, under the neutrality assumption:Where Y is the quantity of output, X is a vector of quantities of variable inputs, Z is a vector of quantities of fixed inputs, and E is a vector of household characteristics and location. For the empirical analysis, the production function is expressed in a widely-used and convenient Cobb-Douglas form, so thatTaking the logarithm of both sides of the equation, this form will be used in the production function analysis. Given potential endogeneity issues with input use, they would best be instrumented. However, given the lack of potential instruments in the Census dataset, this could not be done and we thus have to keep in mind this potential caveat in the interpretation of the results.Second, we look at the determinants of modern technology adoption, such as power tillers, chemical fertilizer use, and plant protection chemicals. The probability of adoption is explained by a vector of covariates W, composed of a vector of quantities of fixed inputs X and a vector of household characteristics and location (reflecting the prices of inputs and outputs). Adoption is modeled as a probit model where adoption takes the value 1 if the technology is applied by the household and 0 if not. The specification can be written as follows:where F denotes the cumulative standard normal distribution and b are parameters.Third, in the Propensity Score Matching (PSM) approach to analyze the effect of remoteness on agricultural activities, a probit regression is first run where the dependent variable equals one if the household lives in an area that is less than 1 hour away from a motorable road and zero when it is living in a more remote location. Control variables include land areas cultivated, quality of land, a risk factor, livestock ownership, and locational dummies. After the estimation of these probit models, a propensity score is obtained that allows for the matching of treatment households (those living close to a motorable road) and control households (those living in remote areas). We use the common support approach to assure that the density of the treatment and comparison group overlap for a large extent. To ensure that the common support is fulfilled, only results are used that pass the balancing tests. We further match treatment and control group by using the kernel PSM (through Stata's psmatch2 command). In this type of matching, the least remote group is matched with a weighted average of control households with greater weight given to the households that have similar characteristics (Heckman et al., 1997). Heckman et al. (1997) and Smith and Todd (2005) show the benefits from these matching techniques over other ones.We first present descriptive statistics on agricultural productivity at the national level and we cross-tabulate these statistics by quintiles of farm size (based on cultivated land). Table 1 shows these levels and their variation over quintiles. We find that the largest farms have production levels that are almost four times as high than the smallest farms. The smaller farms are able to obtain more output per unit of land as productivity levels are almost four times as high for the small compared to the biggest farms. It seems that most of the productivity changes between farm size quintiles are driven by a change in crop composition as yield levels for main cereals are not significantly different between small and large farmers (Table 1). If anything, they are higher for the larger farms. However, when we look at the share of different crops in the output of small and big farmers, smaller farmers specialize significantly more in non-cereal crops, that are more highly valued, on average. Non-cereal crops make up 65% of the production of the small farmer. This compares to only 44% for the largest farmers. Agricultural production shows a strong spatial component as shown on Map 1, 2, 3, 4, and 5. Map 1 illustrates where most of the farms are found in Bhutan. 2 The number of farms per geog is especially higher in the Eastern part of the country and is low in the Western, and more in particular Northwestern, part of the country. Map 2 shows how the value of agricultural production (including all crops and fruits, but excluding livestock) varies in the country. Despite farms being more concentrated in the East, production levels per geog are more evenly distributed. It indicates that households that are located in the Western and Central-western part of the country are able to realize much higher production levels than other households (especially compared to those in the East) (Map 3). A large number of households grow potatoes and apples in this part of the country. It seems that relatively higher prices and production for these crops helps to increase relative incomes for these farmers. Land productivity shows rather similar patterns (Map 4). Given that some of these spatial differences in productivity are driven by crop mix and price differences, we try to control for these factors by looking at the same crop (rice) in Map 5 where the geogs are differentially shaded by the level of rice yield. It seems that especially the West-central part of the country is able to realize the highest rice yields. This is partly explained by the fact that these areas are most secure in water supplies as rice fields are irrigated by rivers and do not depend on sometimes erratic rainfall. Table 2 shows the results of a production function regression that was estimated using the method outlined in section 2. In the first three columns, we present the results of a reduced form regression where the value of agricultural output per household is regressed on distance to a motorable road, the share of the wetland and orchards in total cultivated land area, a risk variable, and dzongkhag dummies. The results show the large effects that each of these factors have on agricultural production of the household. As expected, the value of agricultural production goes gradually down the further households are located from a motorable road. Households that live at a distance from a motorable road of 1-3 hours, 3-6 hours, 6 hours to a day, and more than one day have 4%, 3%, 16%, and 21% lower agricultural production respectively than those households that live close to a motorable road. Wetlands are in general the better quality lands. An increase of the share of the wetland by 1% in the total land area increases the value of agricultural production per household by 0.92%, illustrating the benefits that the improvement of irrigation infrastructure might have on agricultural performance. We also see a lot of inherent differences between dzongkhags. The dzongkhag dummies show that especially those dzongkhags at higher altitudes and disadvantaged because of a difficult agricultural environment have overall significant lower agricultural production levels, ceteris paribus. For example, the dzongkhag of Gasa shows the lowest agricultural production levels of all dzongkhags.In a second specification, we add the use of several agricultural inputs that might explain levels of agricultural production. These inputs include land, chemical fertilizer, plant protection chemical use, and farm yard manure. All these are expressed in logarithms and the resulting coefficients can thus be interpreted as elasticities. The distance effects stay highly significant in this specification, partly indicating the important price effects that these distances have on agricultural production valuation. However, the coefficients decrease in magnitude, compared to the first specification, showing that remoteness has lower effects when one controls for input use (and indicating the strong relationship of inputs with remoteness). Wetlands and orchards continue to show high positive effects on agricultural production levels per household, compared to dryland. A 1% increase in the share of wetland and orchards leads to an increase of agricultural production by 0.9% and 0.3% respectively. Higher risk levels due to attacks by wildlife lead to significantly lower production, ceteris paribus.All the inputs in the agricultural production process have the expected sign and are highly significant. A doubling of the cultivated area leads to a 47% increase of agricultural production. Interestingly, the higher the share of land leased-in, the lower the productivity. It might be that lower investments are done on those lands that are leased out by owners, leading to lower productivity levels. The use of other inputs is also highly significant. A doubling of the use of chemical fertilizer, farm yard manure, and plant protection chemicals leads to an increase of the value of agricultural production by 4%, 3%, and 6% respectively. Evaluated at the mean, this shows that an additional use of a kg of chemical fertilizer leads to an extra output of 58 Nu, significantly above the cost of fertilizer.3 An extra kg of farm yard manure leads to 2.3 Nu of extra output. Plant protection chemicals even show a higher rate of return. Part of the -at least in economic terms -underuse of plant protection chemicals might be explained by the religious Buddhist beliefs that killing of any life is sinful and that the use of the chemicals that achieve such effects is especially discouraged by the widely respected monastery bodies in Bhutan. The results further show that use of mechanized plowing leads to significantly higher output levels. Households that use mechanized plowing have about 30% higher production levels. On top of the better technology for tilling, power tillers are also often used to irrigate fields and this variable might thus reflect more reliable and better access to water resources, potentially explaining the unexpected large size of the coefficient. Livestock ownership also shows strong positive effects on agricultural productivity, on top of manure use and better plowing. The strong complementarities between the crop and livestock sector have been documented in several other settings (e.g. McIntire et al., 1992).Access to quality land is very important for almost any agricultural activity. Land tenure security and investments in land quality such as irrigation might have large effects on agricultural performance (e.g. Deininger, 2003). Table 3 shows land ownership statistics in Bhutan, using the reported thram records. Most of the rural people in Bhutan own land as about 99% of the households that were interviewed in the RNR census reported to be a land owner. An average household owns 1.6 hectares of land. Of all the land owned by households, 66% is classified as dryland, 21% as wetland, 10% as orchards, and 4% as kitchen gardens. 63% of the land owned by farmers is cultivated by them, 8% is leased out, and almost 30% is left fallow. Bigger land owners cultivate less of their owned land, they relatively leave more land fallow (34% of their land), and they lease out more (9%).In Figure 1, we compare the different types of land over time, based on the published data from the RNR census of 2000 and 2009 (based on official data as verified from thram records). We see an overall decrease of wetland areas in Bhutan of about 11%, from 21,870 hectares to 19,522 hectares. This is seemingly driven by the increasing urbanization, especially in the urban areas around Paro and Thimphu, and by the conversion of wetlands to drylands because of the drying up of some springs/streams. However, the dryland areas are also shown to have decreased over time. Questions in the RNR Census of 2000 were asked on the land under shifting cultivation (tseri/pangshing) and on dryland areas. While farmers did previously not own land that was cultivated under shifting cultivation practices (this was property of the state), this was changed by the Land Act of 2007 and was re-classified under dryland in 2008. In our calculations, shifting cultivation land in 2008 was added to the reported dryland as to allow comparison over time. The results show that the area in dryland agriculture has still decreased by 12% over time, from 74,469 hectares in 2000 to 65,665 hectares in 2008. On the other hand, the area declared to be under orchards has increased from 8,603 hectares to 9,714 hectares, an increase of 13%. This increase is possibly linked with the policy of the government to encourage farmers to grow orchards on tseri land as this is presumed to be more environmentally friendly. Land used for kitchen gardens is shown to be relatively less important in Bhutan.Map 6 shows how the average farm sizes differ between the different geogs. It shows overall that the Southern part of the country is characterized by significantly larger farm sizes than the rest of the country. Map 7 shows the size of the land that is left fallow per farm in each geog. Fallow practices, as part of shifting cultivation or because of other reasons, are particularly prevalent in the East and the South of the country. Table 3 shows total areas left fallow in Bhutan. It is especially the less valuable dryland that is left fallow. The data from the RNR Census show however that even wetlands are left fallow and it is estimated that 3,400 hectares of wetland in Bhutan were not used in 2008. Farmers seemingly lack the appropriate incentives to cultivate these lands. This is an important finding given the emphasis on rice self-sufficiency by policy makers in the country and hence the proposed policy measure of providing per acre subsidies to stimulate wetland cultivation.Reasons were further asked in the RNR Census on why farmers leave land fallow. The results of those farmers that answered this question are reported in Table 3. The most important reason for leaving land fallow are wildlife problems (47%), the distance from the residence (37%), lack of irrigation (21%), and the un-productivity of the land (18%). As farmers were allowed to give multiple reasons, the answers do not add up to 100%. When the reasons are compared for different farm size classes, it seems that the small farmers leave land relatively more fallow for house construction purposes (13% of the farmers). However, this reason seems to be overall less important as only 2% of the farmers report this as a major reason for them not to cultivate their land.Table 4 shows land cultivation statistics by farm size (only calculated for those households that reported cultivated land in the Census). The quintile of the smallest farms cultivates on average 0.2 hectares while the largest quintile cultivates just over 3 hectares. The share of leased-in land is relatively more important for the larger farms where it makes up 15% of their land. This compares to 11% for the smallest farms. Overall land statistics show that big land cultivators hold the more valuable wetland and orchards and we see a gradual decline of the dryland share from small to large. While dryland makes up 74% of total land cultivated for the smallest farmers, this is only as high as 58% for the biggest farms. Large farms are much more likely to be involved in orchard cultivation as orchards make up 18% of the land area for the larger farms. This is only as high as 4% for the small land cultivators. When we look at the distribution of orchards, we find that almost 60% of all the orchard areas are held by 10% of all orchard cultivators.  Figure 2 shows the level of inequality in the distribution of land ownership and land cultivation in Bhutan by use of a Lorenz-curve on land areas cultivated and owned. 4 There are few differences between the distribution of land ownership and land cultivation, indicating that land tenure markets do not seem to contribute towards making the land distribution more equal or unequal. Figure 2 shows that 35% of the cultivated and owned land is cultivated by 10% of the farmers. On the other side, it also illustrates that 50% of the farmers own and cultivate about 20% of all land. Maps 8, 9, 10, 11, and 12 illustrate some spatial land patterns in Bhutan, based on the 2009 RNR Census data. Map 8 illustrates the importance of leased in-land for the different geogs. The lease market is much more active in the center of the country compared to the East, West, and North. In the RNR Census, questions were asked on the land area owned, based on the thram records that farmers had to show. The results of these are shown on the next four maps. Map 9 shows total agricultural land area in each geog. It shows that there is very little agricultural land in the Northern geogs, except for the Bumthang dzongkhag, and that geogs turn darker, indicating more agricultural land, the more South they are located in the country. Large agricultural areas are found in the most Southern geogs of the country. Map 10 shows land used towards dryland cultivation. The patterns are similar to the previous map indicating the large importance of dryland areas in agricultural land. Map 11 indicates the importance of wetlands in central, south-central and south-western geogs. There are few geogs with large wetland areas in the East of the country. Map 12 illustrates that most of the orchards are largely situated in the Southern and Western part of the country.A multi-variate regression was run to see which factors are associated with rural land leasing (Table 5). The results show that the further households are located from a motorable road, the higher the share of land that is leased in and thus the more active leasing markets are. Of the land that is leased in, wetlands are more likely to be used in leasing transactions, significantly more than drylands. The more land a farmer owns, the less likely he is to lease in more land. The regression results show that there are further significant differences in lease market activity between the different dzongkhags. Bhumtang is the dzonkhag where the highest level of leasing is going on. The lowest level is in the dzongkhag of Gasa.A similar regression for leasing was used to explore factors that are associated with the share of land being left fallow. Farmers that own more land are more likely to leave more land fallow. A bigger share of the highly Orchard area from thram records information (RNR Census 2009)valuable orchard and wetland is less likely to be associated with higher shares of fallow land (compared to the share of dryland in total cultivated land). Distance to a motorable road does not seem to have a major effect on land being left fallow, e.g. households that live at more than 1 day from a motorable road have 4% less land in fallow than households living close to a road. The size of the dzongkhag dummies show that it is especially the Eastern dzongkhags that have more land in fallow, ceteris paribus.Fertilizers are important for assuring sustainable soil fertility and thus agricultural productivity. This can be achieved by applying locally produced manure or compost or imported chemical fertilizer (as Bhutan does not produce its own chemical fertilizer). In Bhutan, chemical fertilizer is distributed by Druk Seed, a parastatal unit at the time of the survey. It organizes the import of chemical fertilizer in the country, mostly from India, and assures pan-territorial pricing of it by subsidizing transport costs to the private retailers/commission agents that sell chemical fertilizer in the country-side. Druk Seed was recently converted from an independent parastatal to an arm of the Ministry of Agriculture and Forests, presumably because of the perceived increasing emphasis of the parastatal on profits and less on its social mandate.Table 6 shows the use of different fertilizers by farm size. We present the share of farmers that adopt the particular inputs and the quantities used per farmer, using zero for those farmers that do not apply the particular fertilizer. Two-thirds of the farmers report to use farm yard manure. Most of this farm manure is home-made and few farmers rely on markets to obtain the manure. This lack of markets in farm manure is typical for a lot of developing countries. One-third of the farmers in Bhutan use chemical fertilizers on their plots. Interestingly, this chemical fertilizer is relatively more used by smaller farmers than by big farmers, i.e. 43% of the smallest quintile compared to only 17% of the largest farmers report to use chemical fertilizer on their farm. The most used chemical fertilizer is urea, applied by 29% of the farmers. Second comes NPK with 11% of the farmers. Only 4% and 0.4% of the farmers use phosphate or any other chemical fertilizer respectively. It was found in 2000 that 27%, 3%, and 1% of the farmers used urea, NPK, and phosphate respectively (MoA, 2000). This shows an increase in adoption rates but in the Census of 2000, chemical fertilizer use was only asked for cereal crops and given that chemical fertilizers are often used on non-cereal crops, it seems that there has been little dynamics in the share of farmers that use chemical fertilizer over time.Map 13 shows where most of the chemical fertilizer is being used in Bhutan. Few of the farmers in the South and North-west of the country used fertilizers and most of the fertilizer used in Bhutan is concentrated among the farmers that are situated at medium altitude. The highest share of chemical fertilizer adopters is seen in the dzongkhags of Bumthang and Punakha (in the center), and of Paro and Thimphu (in the Western part). Map 14 illustrates that the highest share of the farmers that use farm yard manure for their crops is found in the Western half of the country. We thus see significant overlap of intensive chemical and farm yard manure users. A simple correlation coefficient between farm manure and chemical fertilizer was calculated to see to what extent farmers that use farm manure might substitute it with chemical fertilizer. The results show that there is a positive correlation coefficient (0.24) indicating that these households that these households that do (not) use chemical fertilizer do also (not) use farm manure for the fertilization of their fields.In Table 7, we look at the determinants of fertilizer use by using a probit model where adopters of chemical fertilizer have a dependent variable 1 compared to 0 for the non-adopters. Two models are presented, one short model with crop dummies and one without. The results show that areas that are far out are significantly less likely to adopt chemical fertilizer. Controlling for other factors, bigger land owners are more likely to use chemical fertilizers. Households that own more livestock are less likely to use chemical fertilizer, presumably because they rely relatively more on manure of their own. A higher share of leased-in land does not lead to higher or lower uses of chemical fertilizer use. Controlling for all these factors, four dzonkhags have higher probabilities of chemical fertilizer use than the default dzonkhag of Thimphu. They are the dzonkhags of Paro, Bumthang, Trashigang, and Trashiyangtse. In the model specification where we include crop dummies, we note that households that are cultivating some particular crops are much more/less likely to use chemical fertilizer. Potatoes and apple dummies are most significant and have the highest coefficients, indicating that households that cultivate these two crops are much more likely to use fertilizer than any other households. Farmers that cultivate irrigated rice, oilseeds, radish and other vegetables as well as chili and spices are also more likely to use fertilizer. On the other hand, maize cultivators and most of the farmers that grow fruits are significantly less likely to use chemical fertilizer.In Table 8, we present the same models for the adoption of farm yard manure (as for chemical fertilizer) by the household. Distance to a motorable road has not much of an impact as could be expected given that it is almost exclusively locally produced by the household itself. Livestock ownership is a very significant determinant of farm yard manure use, reflecting the lack of manure markets to make up for lack of access. Farm yard manure is also more (less) likely to be used on wetlands (orchards). The crop dummies reveal that it is more likely that manure is used on irrigated paddy and other cereals, grain legumes, potatoes and other vegetables, and apple. Orange farmers are less likely to use it.Plant diseases are a major risk in Bhutanese agriculture (see Section 5) but the application of plant protection chemicals is not widespread although they seem to have a large pay-off (Section 3). Table 9 shows that 16% of the farmers in Bhutan report to use plant protection chemicals. It is especially more prevalent for larger farmers: 18% of the largest quintile uses plant protection chemicals. This compares to 9% of the smallest quintile. The most used plant protection chemicals are insecticides (7% of the farmers) and herbicides (7% of the farmers). Anecdotic evidence suggests that the use of weedicites is especially prevalent in paddy cultivation as manual weeding in commercial areas has increasingly been replaced by chemicals to deal with the important bog pond weed (Potamogetom distinctus), locally know as shochum.When adoption rates are compared with the data of 2000, it seems that there has been a decline in the number of households that use plant protection chemicals. However, data are not easily comparable as adoption statistics were only calculated at specific crop levels in 2000 while overall household level questions were asked in 2009. In 2000, it was reported that 22% of the paddy cultivators in Bhutan used herbicides on their fields. Given that a large number of farm households in Bhutan cultivate paddy (about 49% of all rural households), it seems however that application rates might have come down over time. It is unclear why this is the case, but it might potentially be partly explained by the decline in wetland cultivation. In Table 10, we use a similar regression format as for chemical fertilizer as to look at the different factors that are associated with the probability of higher plant protection chemical use. Similar results as in the case of chemical fertilizer emerge. Farmers that are located further away from the motorable road are less likely to use plant protection chemicals; bigger farmers are more likely to use these chemicals; farmers that have more wetland and orchards are more likely to use chemicals while leased-in land leads to higher likelihoods of applications. When we look at the crop dummies, we see that those households that cultivate especially irrigated paddy, potatoes, other vegetables, apple and oranges are more likely to use these types of chemicals.Map 15 shows for each geog the percentage of households that used plant protection chemicals in the year of the survey. It is especially the farmers in the South and the North-western part of the country that use less chemicals. Map 16 shows that the use of insecticides is higher in the South and in mid-altitude geogs that are involved in commercial agriculture such as the cultivation of apples and potatoes.When agriculture modernizes, farmers give up traditional ways of doing agriculture and switch to more mechanized forms of agriculture. This often allows them to achieve higher levels of labor productivity. Noted changes typically seen involve first a move from manual labor to cattle traction and then from cattle traction to tractors and power tillers (Binswanger, 1986). The innovations that are adopted depend often on the particular constraints that are faced. Countries that face labor or land constraints usually adopt these technologies that most successfully address these particular constraints (Hayami and Ruttan, 1985).Questions were asked in the RNR Census on the ownership of different 'serviceable machinery and equipment' and the type of ploughing adopted by farmers. We present the results of these statistics by farm size in Table 11. Farms are overall very little mechanized in Bhutan. Only 3% of the farms own a power tiller and a low 0.2 % of the farms own a tractor. As could be expected, larger farms are more mechanized than smaller ones. 5% of the largest quintile owns a power tiller. This compares to only 1% for the smallest quintile. Most of the plowing is done by means of bullocks (manual plowing counts for only 1% of the farms): 89% of the farmers report that they do their ploughing in this way. Although smaller farmers do not own power tillers, there are no big differences between the use in power tillers between large and small farms, indicating a very active rental market from larger to smaller farms: 8% of both the smaller and larger farms reported to use power tillers.Map 17 shows the spatial adoption of power tillers for ploughing in Bhutan. Some very strong spatial patterns show up. Power tillers are widely adopted in the North-central (the dzongkhags of Bumthang and Wangdue) and the Western part of the country (the dzongkhags of Thimphu and Paro). On the other hand, few farmers in the Southern and Eastern part use power tillers.Table 12 illustrates in a multi-variate regression framework the factors that are associated with the adoption of power tillers. Distance to the motorable road is an important determinant of adoption of power tillers. When comparing the different types of land, power tillers are less likely to be used on wetlands compared to drylands. Leased-in land does not lead to any significant difference in adoption. Households that own livestock rely less on power tillers, possibly because they use their own animal traction. Higher risks of crop damage by wildlife leads to a lower likelihood of adoption. In the version of the regression where crop dummies are included, we see that power tillers are much more likely to be used by those households that cultivate potatoes and other vegetables. Households that cultivate maize, irrigated rice and fruits are in general less likely to use power tillers. The dzongkhag dummies confirm the spatial patterns and show to what extent households that are located in North-central and Western dozngkhags of the country are more likely to adopt power tillers. Farmers in the RNR census were asked to rank the constraints that they faced in their agricultural activities in the year prior to the survey. The results are shown, by farm size, in Table 13. Crop damage by wildlife was mentioned as the most important constraint to farming by one quarter of the farmers. This constraint was reported in an equal extent by small and big farmers alike. The second most reported constraint was crop damage by insects and plant diseases, mentioned by about 20% of the farmers. Then comes insufficient irrigation supply and labor shortage, both mentioned by 16% of the farmers. For the two latter constraints, there is a significant difference between small and big farmers. The bigger farmers complain significantly more about lack of sufficient irrigation. The smaller farmers complain more about shortages of labor (as well as shortage of land).Maps 18 through 21 show the shares of the farmers for each geog that mentioned four items (crop damage by wildlife, lack of irrigation, land shortage, and damage by insects/plant diseases) as one of the three major constraints for agricultural production. Map 18 shows the perceived importance of damage by wildlife on crop production. This constraint is especially mentioned by farmers in the East and the South of the country. Lack of irrigation facilities is stated in the Center and the South-West of the country as a major constraint especially in those areas where wetlands are prevalent (Map 19). Lack of land is rarely mentioned as a constraint in the Eastern part of the country while especially farmers in the North and the mountainous areas complain about this (Map 20). Complaints about plant diseases and insects are most prevalent in the most Eastern part of the country (Map 21). The results thus show strong spatial differentiation in perceived constraints in agricultural production. This then calls also for spatially differential agricultural development strategies to address these constraints. Given the importance of crop damage by wildlife, some follow-up statistics are presented below. Information was collected in the RNR census on the area that was damaged by attacks of wildlife and domestic animals for five major crops (barley, maize, paddy, potato, and wheat). It is estimated that an average 0.07 hectares are affected by crop damage due to wildlife, leading to 112 kgs of crop losses per household (Table 14). The losses are especially high for maize where 71 kgs are lost per household. Paddy and potato come second and third. At the bottom of Table 14, it is shown which type of animal is perceived to have done most damage to crops. Wild boars do most of the damage (65%). Monkeys come second at 21%. Damage by wild boars is especially important for the smallest quintile as 76% of them complain about them. This compares to 52% for the bigger farms. Damage by monkeys is more important for the biggest quintile, possibly because of the relatively higher importance of orchards for these farmers as monkeys might focus relatively more on the fruits from these.Map 22 illustrates the spatial pattern of crop damages. To show the importance of crop damage, the share of the households that complained about crop damage is mapped. There are no clear spatial patterns in reported crop damage but it seems that especially people in the South of the country report most damage. The two most affected dzongkhags are those of Haa and Zhemgang. Map 23 shows the importance of the different types of animals causing the damage. The results show that the altitude areas are more affected by damages by wild boars. Monkeys on the other hand seem to cause most damage in the South of the country. We will focus our analysis in this section on analyzing what the effect is of remoteness on agricultural performance and technology adoption indicators. We will do so using different methodologies. In the first part, we will rely on simple descriptive statistics. In the second part, we will try to control for confounding factors, using multivariate regression analysis and propensity matching scores. In the third part, we will compare the benefit of road infrastructure investments with other interventions.We first present the data on the extent of remoteness in Bhutan and its spatial distribution. A remoteness number per geog was constructed by taking the mean average distance that each household in that geog has to walk to get to a motorable road. Map 24 shows the spatial variation of this indicator. The rural households in mid-altitude regions are seemingly least remote. Households in the higher mountain ranges, the Eastern region, and the Southwestern and Southeastern region have to walk most far of all before they can get to a motorable road.Table 15 present simple descriptive statistics of land cultivation and agricultural performance by distance to a motorable road. The areas cultivated are slightly higher for the households that live further out. Land becomes more valuable when closer to major roads reflecting the effect of distance on the marketing costs of outputs and inputs. Farm sizes thus generally tend to become smaller and land tends to become more intensively cultivated when located closer to a road (Fafchamps and Shilpi, 2003;Jacoby and Minten, 2009). This is also seen in Bhutan. Table 16 further shows that there is a strong association between the quality of the land and the distance to a motorable road. While the share of wetland in total cultivated land is 35% for those farmers living close to a motorable road, this is as low as 14% for those households far out. This is because governments usually tend to construct roads in areas that have higher agricultural potential. However, because of better road connections, some farmers might also invest more in land and upgrade it.There are few differences in land lease market activity by distance to a motorable road. In normal situations, leasing markets tend to be more active in these areas that are more commercialized (and closer to a road) (Deininger, 2003). Surprisingly, leasing in Bhutan is very active in the areas further out. It might be that because of emerging economic opportunities elsewhere and of more difficult living situations, people might be migrating out of these areas, leading to more active rental land markets as these migrants might be unwilling or unable to sell land.Households that are located close to a main road are able to obtain production levels per household that are 50% higher than the ones that are very far out. The differences in productivity levels per unit of land are 63% higher. Different phenomena are going on. First, surprisingly and in contrast with studies in other countries (e.g. Fafchamps and Shilpi, 2003), there are few differences between the relative importance of cereals versus non-cereals by distance to a motorable road. Cereals account for 51% of the value of production for households close by and far out. However, within cereals, we see a shift in the relative importance of rice versus maize where the higher-valued rice is much more important for households close by, driven by the higher availability of wetlands. Second, productivity of cereals declines by distance to a motorable road, but only slightly so: rice yields are on average only 16% higher for the households that are located close to the road. Third, prices for rice and potato do not differ very much over distance to a motorable road, possibly driven by the stabilizing effect (because of pan-territorial pricing) of the Food Corporation of Bhutan. It thus seems that the higherWalking distance in hours 8 -18 4 -8 2 -4 1 -2 0 -1 Average distance of households to a motorable road (RNR Census 2009) productivity levels for households close by are mostly driven by crop composition effects and relatively less by changes in prices or yields. Table 16 presents all agricultural technology adoption statistics by distance to a motorable road. As could be expected, modern inputs are much less adopted by farmers that live further out as the costs of carrying these modern inputs to more remote locations make them significantly more expensive while prices for outputs are usually also less rewarding in these far out areas which further reduces the incentives for their use. It is estimated that only 8% of the households that are located more than a day from a major road use fertilizer. This compares to 42% for those farmers located close to a road. Interestingly, also farm yard manure is applied less in more remote areas. Only 46% of the farmers report to use it in these areas. This compares to 65% of the households overall. However, the low use of manure in remote areas seems partly compensated by more use of leaf mould (or other leaf litter from forest floors). As is the case with chemical fertilizer, farmers that are located further from the motorable road are much less likely to use plant protection chemicals. Table 16 shows that while 21% of the households that live less than 1 hour away from a major road use plant protection chemicals, this was as low as 4% for those households that had to walk for longer than a day to reach a major road. Access to roads shows also a strong association with the adoption of agricultural equipment. 17% of the farmers that live less than 1 hour from a main road use power tillers for plowing. This compares to only 1% for the most remote households (living more than a day from the main road). Ownership of agricultural equipment shows similar declines. The most common modern equipments, i.e. power tillers, rice mill sets, and plant protection equipment, are owned by 5%, 6% and 3% respectively of the farms in the least remote villages. This compares to 0.3%, 1.7% and 1.0% respectively of the most remote farms.The results here thus illustrate the importance of transportation costs in the adoption of modern technology. However, it also shows that better access to roads is only one factor among many that leads farmers to adopt modern technologies as even for those farmers that are located close to motorable roads, adoption rates are still low. In the previous section, we have illustrated the large effects that distances to motorable roads have on technology adoption and on agricultural performance by relying on simple cross-tabulations. As variation between different agricultural indicators can be explained by a number of other factors other than remoteness, we use two techniques that attempt to control for some confounding factors that are at our disposal in the RNR Census, i.e. multivariate regression analysis and Propensity Score Matching. First, we present the analysis using multivariate regression models. The results show that, except for rice prices, all outcome variables (agricultural productivity per household and per unit of land, rice yields, prices of rice and potato) are significantly and negatively affected by remoteness (Table 17). Agricultural productivity per unit of land is for example 67% lower for the most remote households compared to the ones closest to a road. The effects on rice yield directly are much smaller and we only see a decline of 7% between these two categories. While potato prices decline by distance, the effects are rather small. We see no effect at all of distance on rice prices probably because of the presence of FCB shops that sell rice at equal prices throughout the country (pan-territorial prices) which might lead to spatial uniformity of local paddy prices.Table 17 further shows the results on technology adoption by using these different distance categories as independent regressors. Again, we see strong statistically significant impacts of these categories on technology adoption and the effects become stronger the further a household is located from a motorable road. In the case of manure application, we only see a large negative effect in the most remote category. The three other categories show even higher manure applications than in the households located near to a main road. Coefficients for chemical fertilizer are much larger than for plant protection chemicals illustrating the bigger difficulty -and thus relatively higher costs -of transporting the former inputs, due to its bulkiness, when no roads are available. In the Propensity Score Matching (PSM) analysis, we repeat the same exercise for the different outcome and technology adoption variables (Table 18). In this case, we club all households that are located more than 1 hour from a motorable road in a remote category (the control households). Households close to the road, the nonremote households, are the treated households. The results are robust compared to the multivariate regression results and show strong effects on all agricultural performance variables, except rice prices, and all technology adoption variables, except the use of manure. As would be expected, differences are smaller when we look at comparable households (under ATT (Average Treatment Effect for the Treated)) as the overall characteristics of non-remote households are different of remote households. By comparing similar households, we get at a better picture of the true impact of remoteness. The results using this method show that the percentage of households that adopt chemical fertilizer, plant protection chemicals, and power tillers is respectively 13%, 5%, and 11% lower in remote locations.Using the analysis from the previous sections, we will compare in this section the impact on agricultural production and on rice production of a program that will aim to improve the access of rural households to roads with two other policies that are being considered by the Bhutanese government, i.e. a wetland protection program and an irrigation investment program. While the data that we have at our disposal do not allow us to perform any calculations that take all relevant factors into account and the calculations are thus crude and simplistic, they however allow us to indicate some order of magnitude, as well as spatial differentiated, benefits of these different interventions.In a first policy simulation, the government would embark on a rice intensification program, on currently used wetland, through irrigation investments and through assuring better access to improved seeds and better technologies. We assume that through such a program, rice yields would increase on average by 20%. The value of total agricultural production in the country would go up by 7.9% (as rice makes up about 40% of the value of total output) (Table 19). Map 25 shows the geogs that would benefit from such a program, i.e. those geogs where current rice production is situated.In a second policy simulation, the government would give a flat cash subsidy to all farmers as to use their wetlands for rice production (as is done for example in China). We assume that this subsidy would be high enough as to give all farmers enough incentives to cultivate rice on all fallow wetland. We assume that similar yields on these reconverted wetlands are achieved as in the other wetlands of that geog. In such a scenario, total rice production and total agricultural production would increase by 17.3% and 6.8% respectively. Map 26 shows that the benefits from such an intervention would be mostly for those geogs that have currently high levels of fallow wetland.In a third policy simulation, roads would be improved in such a way that all rural households would now live no further away than one hour of a motorable road. We use the regression coefficients from the multi-variate regression to simulate the impact of this change on agricultural and rice production. While agricultural production at the national level would go up by 3.5%, rice production would go up by 2.8%. Rice production would be less affected by such a program compared to agriculture overall because rice is currently relatively more done in well-connected areas. Map 27 shows that the impact of such a program would mostly be felt in the South and the East of the country.While these calculations are crude, they however help to make some strategic points. Road improvements would have a positive effect on agriculture but its impact would be small as only a relatively small number of farmers in Bhutan would benefit from it (those that are located at more than 6 hours walk from a motorable road accounting for 18% of the rural population, as these are shown to be highly affected by remoteness). The lesser impact of roads in Bhutan on agriculture might not be that surprising given the high subsistence level and the little reliance on imported inputs in Bhutan. On the other hand, there are of course many other benefits from road construction, especially towards the stimulation of the non-farm economy and towards access to social services that are not captured in these calculations (Jacoby and Minten, 2009). To justify road construction in a cost-benefit analysis, one would also need to take into account how many farmers would be reached by the construction of these roads. In this report, we looked at agricultural productivity and technology adoption by farmers in Bhutan. This is an important topic given the strong link of agricultural performance and rural poverty. For example, Christiaensen et al (2006;2010) evaluated the elasticity of poverty reduction with respect to agricultural GDP in low-and middle-income countries at -1.60 and -2.24, indicating that a 1% agriculture growth would lead to a reduction of poverty by 1.6% and 2.2% for these two types of countries. These results show the needed emphasis on agriculture as a vehicle for sustainable poverty alleviation. Several important insights emerge from our analysis. First, agriculture is characterized by a low level of usage of modern inputs and mechanization. Chemical fertilizer and pesticides are used by 33% and 16% of the farmers respectively and only 10% of the farmers use power tillers for plowing. The low level of modern technologies leads to relatively low productivity. Our results show that there are important productivity effects and high rates of returns of the increased use of improved agricultural technologies such as chemical fertilizer or plant protection chemicals. While not studied in this report in particular, there might be further benefits to changes in seed markets and to a bigger emphasis on hybrid seeds (e.g. for the rice sector) which until now are little used in Bhutan but which have shown to lead to significantly higher yields in neighboring India.Second, we document in this report the strong spatial patterns in agricultural performance and technology adoption. Some of these effects might be related to inherent infrastructural or institutional characteristics but some of these are also strongly related to climatic and soil differences. The results indicate that the Western and West-central part of the country are doing well in terms of productivity per household or per unit of land while the Eastern part and the Southern part of the country seem to be doing less well. The spatial analysis also shows strong differential spatial constraints in agriculture with crop damage due to wildlife and plant diseases more prevalent in the Eastern part of the country and land shortage and water constraints more complained about in the Western and Southern part of the country. This thus calls for spatially differentiated agricultural development strategies.Third, we document the impact of roads as well as the quality of land (access to irrigation) on productivity and technology adoption. Controlling for other factors, household production is 20% lower for those households that have to walk for more than 1 day to a motorable road compared to those households located close to a motorable road. Distance to roads further show important effects on the likelihood to adopt different types of improved technologies, despite the policy of pan-territorial pricing of inputs. However, the results also show that investments in infrastructure are a necessary, but not sufficient, condition as adoption rates and agricultural productivity might still be low in areas that are well-connected. In simulations presented in the report, we show that the costs of road investment are overall probably prohibitively high to justify towards stimulation of the agricultural sector. Complimentary investments in other infrastructure and in research and extension might improve agricultural productivity more efficiently.Fourth, the size of the farm is in general small in Bhutan (the average size of cultivated land is evaluated at 1.2 hectares per farm) and the distribution of land in the international context is rather equal. Smaller farms depend relatively more on the lower quality dryland while bigger farms have better access to irrigated land as well as orchards. Especially orchards are in the hands of larger farmers as we find that almost 60% of all the orchard land is held by 10% of all orchard farmers. Interestingly, despite their lower land quality, smaller farmers are able to achieve much higher land productivity than larger farmers as their land productivity is evaluated to be almost three times as high.Fifth, a major problem in agricultural production in Bhutan is crop damage by wildlife and plant diseases. Respectively 25% and 20% of the farmers consider these as the most important constraints to their agricultural activities. They are ranked higher than problems with irrigation or labor. Especially maize, an important crop in the East, seems to suffer most of the losses due to wildlife. We further find that areas kept in fallow land are quite significant. While fallow land would be part of a shifting cultivation pattern, the major reported reason for why farmers leave land fallow is linked to potential wildlife damage. The results also show that 3,400 hectares of wetland are not used, an important consideration for policy makers given their aim to improve the rice selfsufficiency in the country.While Bhutan would benefit from better agricultural performance for the improvement of food security and the alleviation of poverty, some of the conventional modern technologies used in other countries have shown important negative externalities on the environment. This is especially important in Bhutan given the unique bio-diversity and special eco-systems that are found in the country. Any program that leads to agricultural intensifications thus has to take into account the (negative) effects that such a program might have on the environment and alternative improved technologies (as for example composting techniques that are still not very widely used) with less damaging effects on the environment should be considered. While the low use of modern inputs is seemingly associated with lower productivity and low agricultural incomes, it could however also be used towards Bhutan's advantage. There is a growing demand for organic produce worldwide (UNCTAD, 2004) and certification efforts combined with more aggressive sales of a Himalayan Bhutan brand might allow Bhutan to tap into these growing premium markets to the benefit of its farmers.Several different pathways can be followed to improve agricultural performance in Bhutan. The adoption of better agricultural technologies can be enhanced by ensuring productive investments, such as roads, irrigation investments, and availability of appropriate agricultural inputs developed by a suited R&D system. Another pathway is the subsidization of agricultural inputs such as chemical fertilizers, electricity, and irrigation facilities. Experiences in other countries have shown that the returns to the productive investments are significantly higher than to subsidies (Fan et al., 2008) and thus productive investments have generally higher rates of returns and are thus preferred.This document is a first step of analyzing technology adoption and agricultural productivity in Bhutan. We had the advantage to be able to use updated data available at the national level. The disadvantage of the methodology used is that we had limited number of variables to work with which did not allow us to go in great analytical detail in productivity analysis. To make meaningful further analysis on the agricultural situation, it would be useful for Bhutan to have access to a detailed nationally representative household survey -including information among others on access in and use of input markets (e.g. improved seeds, chemical fertilizer, plant protection chemicals), credit, extension, knowledge of improved technologies, tenure security and land arrangements, access and reliability of irrigation, the functioning of output markets, storage behavior and consumption patterns -that would allow generating updated and relevant information on some of the important questions that policy makers are currently struggling with.","tokenCount":"11383","images":["939806089_1_1.png","939806089_1_2.png","939806089_1_3.png","939806089_1_4.png","939806089_1_5.png","939806089_62_1.png","939806089_62_2.png","939806089_62_3.png","939806089_62_4.png","939806089_62_5.png"],"tables":["939806089_1_1.json","939806089_2_1.json","939806089_3_1.json","939806089_4_1.json","939806089_5_1.json","939806089_6_1.json","939806089_7_1.json","939806089_8_1.json","939806089_9_1.json","939806089_10_1.json","939806089_11_1.json","939806089_12_1.json","939806089_13_1.json","939806089_14_1.json","939806089_15_1.json","939806089_16_1.json","939806089_17_1.json","939806089_18_1.json","939806089_19_1.json","939806089_20_1.json","939806089_21_1.json","939806089_22_1.json","939806089_23_1.json","939806089_24_1.json","939806089_25_1.json","939806089_26_1.json","939806089_27_1.json","939806089_28_1.json","939806089_29_1.json","939806089_30_1.json","939806089_31_1.json","939806089_32_1.json","939806089_33_1.json","939806089_34_1.json","939806089_35_1.json","939806089_36_1.json","939806089_37_1.json","939806089_38_1.json","939806089_39_1.json","939806089_40_1.json","939806089_41_1.json","939806089_42_1.json","939806089_43_1.json","939806089_44_1.json","939806089_45_1.json","939806089_46_1.json","939806089_47_1.json","939806089_48_1.json","939806089_49_1.json","939806089_50_1.json","939806089_51_1.json","939806089_52_1.json","939806089_53_1.json","939806089_54_1.json","939806089_55_1.json","939806089_56_1.json","939806089_57_1.json","939806089_58_1.json","939806089_59_1.json","939806089_60_1.json","939806089_61_1.json","939806089_62_1.json"]}

data/part_2/0080719722.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"7f5053633cd04a440481a0ee4c6d9986","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/13c96651-d6e9-41dd-8a03-e3b8513412f2/retrieve","description":"If fundamental climate change mitigation and adaptation goals are to be met, international climate negotiations must include agriculture. Agriculture and climate change are linked in important ways, and this brief focuses on three: (1) climate change will have large effects on agriculture, but precisely where and how much are uncertain, (2) agriculture can help mitigate climate change, and (3) poor farmers will need help adapting to climate change. As negotiations get underway in advance of the meeting of the 15th Conference of Parties of the UN Framework Convention on Climate Change in Copenhagen in December 2009, this brief suggests negotiating outcomes for both mitigation and adaptation funding that will support climate change goals while enhancing the well-being of people who manage and depend on agriculture, especially in the developing world.","id":"760860306"},"keywords":[],"sieverID":"f05cd08a-2bae-423c-bb7c-4d9c8b3f9487","pagecount":"2","content":"I f fundamental climate change mitigation and adaptation goals are to be met, international climate negotiations must include agriculture. Agriculture and climate change are linked in important ways, and this brief focuses on three: (1) climate change will have large effects on agriculture, but precisely where and how much are uncertain, (2) agriculture can help mitigate climate change, and (3) poor farmers will need help adapting to climate change. As negotiations get underway in advance of the meeting of the 15th Conference of Parties of the UN Framework Convention on Climate Change (UNFCCC) in Copenhagen in December 2009, this brief suggests negotiating outcomes for both adaptation and mitigation that will support climate change goals while enhancing the well-being of people who manage and depend on agriculture, especially in the developing world.Even with the best efforts to mitigate greenhouse gases (GHGs), it is inevitable that poor farmers will be affected. The goal is to find and fund the most cost-effective ways to help the poor adapt to the changes, a daunting task because of uncertainty about the magnitude of possible changes, their geographic distribution, and the long lead times needed to implement adaptation efforts.A pro-growth, pro-poor development agenda that supports agricultural sustainability also contributes to climate change adaptation. Adaptation is easier when individuals have more resources at their command and operate in an economic environment with the flexibility to respond quickly to changes. If, as seems likely, the effects of climate change will fall disproportionately on poor farmers, a policy environment that enhances opportunities for smallholders will also be good for climate change adaptation. Such an environment would include more investment in agricultural research and extension, rural infrastructure, and access to markets for small farmers. Funding should support these kinds of policy changes and investments in institution building.Many changes to management systems that make them more resilient to climate change also increase carbon sequestration. Conservation tillage increases soil water retention in the face of drought while also sequestering carbon below ground. Small-scale irrigation facilities not only conserve water in the face of greater variability, but also increase crop productivity and soil carbon. Agroforestry systems increase above-and below-ground carbon storage while also increasing water storage below ground, even in the face of extreme climate events. Properly managed rangelands can cope better with drought and sequester significant amounts of carbon. Project-and program-based funding schemes that support adaptation should also be able to draw on mitigation resources.Even without climate change, greater investments in agricultural science and technology are needed to meet the demands of a world population expected to reach 9 billion by 2050. Many of these people will live in the developing world, have higher incomes, and desire a more diverse diet. Agriculture science-and technology-based solutions are essential to meet those demands.Climate change places new and more challenging demand on agricultural productivity. It is urgent to pursue crop and livestock research, including biotechnology, to help overcome stresses related to climate change such as heat, drought, and novel pathogens. Crops and livestock are needed that respond reasonably well in a range of production environments rather than extremely well in a narrow set of climate conditions. Research is also needed on how dietary changes in food animals, including pasture improvements, can reduce methane emissions.One of the key lessons of the Green Revolution is that improved agricultural productivity, even if not targeted to the poorest of the poor, can be a powerful mechanism for alleviating poverty indirectly Improvements in water productivity are critical, and climate change, by making rainfall more variable and changing its spatial distribution, will exacerbate the need for better water harvesting, storage, and management. Equally important is supporting innovative institutional mechanisms that give agricultural water users incentives to conserve.Investments in rural infrastructure, both physical (such as roads, market buildings, and storage facilities) and institutional (such as extension programs, credit and input markets, and reduced barriers to internal trade) are needed to enhance the resilience of agriculture in the face of the uncertainties of climate change.Agriculture is an intensely local activity. Crop and livestock productivity, market access, and the effects of climate are all extremely location-specific. Yet national and global efforts to collect and disseminate data on the spatial nature of agriculture, especially over time, are poorly developed. Countries have reduced funding for national statistical programs, and remote sensing systems are still inadequate to the task of monitoring global change. Understanding agriculture-climate interactions well enough to support adaptation and mitigation activities based on land use requires major improvements in data collection and provision.Today, agriculture contributes about 14 percent of annual GHG emissions, and land use change including forest loss contributes another 19 percent. The relative contributions differ dramatically by region. The developing world accounts for about 50 percent of agricultural emissions and 80 percent of land use change and forestry emissions.The formal inclusion of REDD (Reducing Emissions from Deforestation and forest Degradation) in the current negotiations is a result of a new appreciation of the importance of this source of GHGs and initial findings of low-cost opportunities to reduce them. Significant challenges remain, however. What are the best approaches to dissuade poor people from cutting down trees and converting other lands to unsustainable agricultural practices and to instead encourage them to adopt technologies and management strategies that mitigate carbon, methane, and nitrous oxide emissions? The tasks ahead include identifying and supporting the most appropriate approaches in farmers' fields and monitoring their implementation.Agriculture has huge potential to cost-effectively mitigate GHGs through changes in agricultural technologies and management practices. Changing crop mixes to include more plants that are perennial or have deep root systems increases the amount of carbon stored in the soil. Cultivation systems that leave residues and reduce tillage, especially deep tillage, encourage the buildup of soil carbon. Shifting land use from annual crops to perennial crops, pasture, and agroforestry increases both above-and below-ground carbon stocks. Changes in crop genetics and the management of irrigation, fertilizer use, and soils can reduce both nitrous oxide and methane emissions. Changes in livestock species and improved feeding practices can also cut methane emissions. Mitigation funding programs arising from the negotiations should thus include agriculture.It is much easier to monitor 1,500 U.S. coal-fired power plants than several million smallholder farmers who rely on field, pasture, and forest for their livelihoods. Nonetheless, promising technologies exist for reducing the costs of tracking the performance of agricultural mitigation programs. For example, microsatellites can be used for frequent, high-resolution land cover imaging, inexpensive standardized methods are available to test soil carbon, and simple assessment methods can adequately quantify the effects of management technologies on methane and nitrous oxide emissions. These monitoring technologies and others require funding.Agricultural production differs qualitatively from other sources of GHGs in that the sources are individually small, geographically dispersed, and often served by inadequate physical and institutional infrastructure. Cost-effective payment mechanisms to encourage agricultural mitigation must reflect these differences. Beyond the traditional schemes developed under the Kyoto Protocol, the negotiating outcome should allow and encourage alternatives that take advantage of these differences, exploiting activities beyond project-specific funding. Examples include land retirement contracts, one-time payments for physical infrastructure investments that have long-term mitigation effects, and payments for institutional innovations that encourage mitigating behavior in common property resources.Agricultural activities around the world are responsible for almost 15 percent of annual GHG emissions. They could be an important sink for emissions from other sectors and are likely to be altered dramatically by climate change. Agriculture also provides a living for more than half of the world's poorest people. The ongoing negotiations to address climate change provide a unique opportunity to combine low-cost mitigation and essential adaptation outcomes with poverty reduction. n","tokenCount":"1270","images":[],"tables":["760860306_1_1.json","760860306_2_1.json"]}

data/part_2/0081629656.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"93e1eb4c47963f87c243ecf2367b2d32","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/ad00d478-14b5-43ab-97e2-12d0c32a0dd0/retrieve","description":"","id":"-1401715296"},"keywords":[],"sieverID":"e6cafc2b-8a8f-495d-8126-0479395ae2c2","pagecount":"81","content":"Land degradation is advancing at an alarming rate in Sub-Saharan Africa, particularly in the form of desertification in dryland areas, soil erosion and deforestation in hillsides, and loss of soil fertility in many cropped areas. The degradation of fragile drylands and hillside areas is particularly worrying because it is often irreversible, or can only be reversed at high cost. While natural forces such as climate change, drought, floods and geological processes contribute to land degradation, the most important contributing factor in Sub-Saharan Africa is human activity. The key driving force is thought to be a nexus of poverty, rapid population growth, and inadequate progress in increasing crop yields. Poor rural people in their quest for food and other livelihood needs are increasingly (i) expanding cultivation into forests, steep hillsides, and other fragile areas that are easily degraded; (ii) reducing fallow periods to the point where soils are inadequately rejuvenated; (iii) pursuing land management and cultivation practices that deplete soils of their nutrient and organic matter content and promote acidification;(iv) overgrazing pasture areas; and (v) cutting but not replanting sufficient trees for fuelwood and other purposes. New or improved technologies that can increase yields and hence reduce the pressure to expand the crop area are either not available, or not economic or appropriate under the subsistence conditions of small-scale farmers.The dimensions of these resource degradation problems are large and are growing.Land degradation is now thought to affect two-thirds of the total cropland of Africa, and one-third of the pasture land. About three-fourths of these degraded lands lie in dry regions. Moreover, about 3.7 million hectares of closed forest are lost each year, and this rate is accelerating. There is accompanying loss of biodiversity, increased siltation, and flooding of rivers that threaten downstream uses such as dams and irrigated agriculture, and changes in regional an global climate.This degradation is matched by worsening poverty and food insecurity in rural areas, and by unabated increases in the population. Each year, another 20 million people are added to Africa's population, leading to a doubling every 20 years. About half of all Africans currently live in poverty, and about 160 million poor live in dryland and hillside areas. Africa is the only major region where poverty is expected to increase in the next decade. Food availability is already below the UN recommended minimum, yet food production per capita continues to decline (by 0.8% per year during 1981-92). IFPRI's projections for the year 2020 suggest worsening poverty and food insecurity problems in Africa, unless there is a significant change in strategy, with the total food gap increasing from 8 million tons per year today to about 30 million tons per year by 2020. And these food gaps are based on estimated purchasing power, not on the amounts of food that Africans would need to be properly fed. With poor economic growth and worsening social problems, the pressure on natural resources, particularly in fragile areas, seems destined to worsen. Migration to cities will provide some relief, but even so, population densities in many fragile areas are likely to increase during the next several decades. As governments and aid agencies confront the increasingly urgent need to develop a broad range of types of fragile lands in Africa, there is a need for a new consensus on how this can best be achieved. An evaluation of past and recent experiences in developing fragile lands, and synthesizing the lessons learned for the future would be a useful input into achieving a new consensus.The broad goal of the conference was to contribute towards balanced agricultural growth and food security in Sub-Saharan Africa, through formulation of strategies for sustainable resource management and poverty alleviation in the fragile lands. In pursuit of this goal, the conference focused on the following objectives:Foster common understanding of the issues and challenges in the development of fragile lands to solve the problem of food security and poverty alleviation;C Identify constraints to successful management of fragile lands;C Formulate recommendations for designing appropriate policies and strategies for fragile lands to meet the livelihood needs of people while conserving and sustainably managing the natural resource base; andIdentify appropriate follow-up activities for implementing the agreed upon policies and strategies (for research, capacity strengthening, policy formulation, NGOs and future investment).The Entebbe Conference was planned and managed by an advisory committee comprised of:Dr. Bruno Ndunguru (SACCAR)Prof. Chris Akello-Oguto (Technoserve)Dr. Suresh Babu (IFPRI)Dr. Peter Hazell (IFPRI)Ms. Anne de Ligne (EC)Dr. Matthias Magunda (NARO)Mr. Juergen Richter (DSE)Professor Joseph Mukibi, the Director General of NARO, and the president of the conference welcomed the participants to Uganda and to the conference, and introducedProfessor Bruno Ndunguru, the Director of SACCAR and the conference chairman.Professor Ndunguru introduced the objectives of the conference, its format, and the expected outcome. He emphasized the need for coming up with a future plan of action at the end of the workshop which should be presented to key policy decision makers. Mr.Herbert Beck, representing the Ambassador of Germany to Uganda, stressed that given the increased demand for food and agriculture commodities for subsistence and commercial purposes, sustainable use of natural resources will be vital in the years to come. According to him, developing sustainable agriculture production systems is the key challenge for the fragile lands of Sub-Saharan Africa. Thus, the core question of the conference was how to maximize growth in agricultural production for poverty reduction while minimizing the damage to the natural resource base. Following him, Dr. PeterHazell of IFPRI welcomed the gathering on behalf of IFPRI. Giving a brief introduction to IFPRI and its research program in less-favored areas, he highlighted the urgent need to increase Africa's per capita food production. Unfortunately, while total food production has increased moderately in recent years, population size has grown faster and food production per capita has declined. Dr. Hazell emphasized that the increased food production in the high-potential areas will not be sufficient to redress the food needs of many millions of rural poor. Additional food production will have come from vast areas of less-favored areas including the fragile hillsides and low rainfall areas that are the primary focus of this conference.The conference was officially opened by the Honorable Dr. Israel Kibirige Sebunya, Minister of State for Agriculture on behalf of the Minister of Agriculture and Vice President of Uganda, Honorable Madam Speciosa Kazibwe. He welcomed the participants and emphasized the importance of the timing of the conference given that land degradation is advancing at an alarming rate. In Sub-Saharan Africa, particularly due to desertification in dry areas, soil erosion and deforestation in hillsides and loss of fertility in many cropped areas, the productivity of crops has been going down. Because degradation of fragile dry areas and hillsides is often irreversible, he highlighted the need for developing technological and policy solutions for reversing the trends, thereby increasing food security and reducing rural poverty.THE BACKGROUND PAPERS Six plenary papers were commissioned for the conference. Their titles and authors are listed below in their order of presentation. In addition to the plenary papers, five poster papers were also displayed.Dambo Irrigation System: Indigenous Water management for Food Security in Zimbabwe. Ruth Meinzen-Dick, IFPRI; Godswill Makombe, Colorado State University.Poverty: A Case of Tanzania's SACCOs. Andrew Temu, Sokoine University.Soil Conservation and Environment Recovery in Machakos District of Kenya. Francis Gichuki, University of Nairobi.Soil Conservation in the Highlands of Uganda. Julius Zake, Makerere University; Mathias Magunda, Kawanda Agricultural Research Institute.Conservation of Fragile Wetlands in Uganda, Mathias Magunda, Kawanda Agricultural Research Institute; Julius Zake, Makerere University.A brief summary of the plenary papers follows.The fragile lands of Sub-Saharan Africa are facing a worsening social and environmental crisis. Fragile lands include hillsides and mountains prone to soil erosion that can produce good crops under certain conditions; sub-humid savannahs prone to soil acidification; humid lowlands prone to soil nutrient leaching; tropical lowlands that have low and unstable rainfall; semi-arid regions prone to wind and desertification; and those areas that may have pockets of productive soils but are inaccessible to and need rural infrastructure to make them economically worthwhile for farmers to increase production.The worsening problems of stagnant agricultural production, increasing poverty and natural resource degradation in the fragile areas of Sub-Saharan Africa require urgent action by governments and development agencies to provide solutions that integrate the livelihood needs of local people with the sustainable management and conservation of natural resources. Without such action, governments and donors are likely to have to spend increasing amounts of resources on crisis relief, safety nets and environmental protection and rehabilitation.The conventional wisdom that governments and donors should continue to focus most investment in high-potential, often irrigated areas for higher agricultural growth is being increasingly challenged, since this focus has failed to redress poverty amongst the vast majority of Africa's smallholders. Reaching the poor through additional investments in the commercial sector is also in doubt, particularly in many countries in Sub-Saharan Africa that have limited land that is suitable for expanding commercial farming and limited non-farm opportunities. However, some of the recent political, social, institutional, and economic transformations, have provided Sub-Saharan Africa with improved opportunities to achieve increased growth, reduced poverty, and the sustainable use of natural resources.Achieving poverty alleviation and the sustainable use of natural resources does not have to mean slow growth in the agricultural sector. While there often seems to be a tradeoff among these objectives in the short-run, recent experiences in the sustainable Land fragility is an elusive concept that relates to the potential mismatch between human use of land and its biophysical conditions, leading to resource degradation. Such a mismatch is conditioned by a host of climatic, biophysical, social, economic and technology factors, and even then may change over time. Fragility is a difficult concept to operationalize for development planning, and it is more useful to think instead in terms of the suitability of different \"locations\" rather than land for specific economic activities over time. Spatial analysis of patterns and processes related to the environment, demography and infrastructure can provide significant benefits to policy makers by highlighting potential rural development opportunities. Many policy, market, and institutional variables exhibit spatial aspects that are useful for designing programs for the development of the fragile land areas.Formulating good development strategies is a complex matter for fragile lands.One of the complexities is the dynamic interaction among natural and human-induced factors and the stocks and flows of natural and other resources in the fragile lands.Obtaining an improved vision, literally and metaphorically, of such interactions can bring order and structure to the policy formulation process. Order and structure arise not only from the existence of a geographically referenced universe, but also because geographical information systems (GIS) can integrate both spatial and seemingly non-spatial information in their database systems. Thus, the spatial framework can be catalytic in providing both a common focus for multi-disciplinary inputs to designing and testing development strategies, as well as common formats for the integration of multi-discipline information. This presents opportunities for time and cost savings, as well as for shaping more effective strategy options, particularly if the visual simplicity of this media can serve to increase transparency in the policy formulation process and thereby make it more accessible to a greater range of stakeholders.Developing capacity to acquire, generate, manage and interpret spatial information, and feed it into the strategy development process would ultimately improve the likelihood of favorable social and environmental outcomes in the fragile lands. The visual presentation of data in spatial formats allows a fresh interpretation of agricultural and resource-related issues that face fragile areas. An overview of the relative spatial distribution of key variables may offer some insights into causality, as well as point to issues of data resolution and quality for designing policies and programs in the fragile lands. Most importantly, the approach offers a geographic perspective of the feasibility and attractiveness of targeting of specific strategy options for reducing food insecurity and poverty in fragile areas of Sub-Saharan Africa.Many Sub-Saharan African countries have made serious efforts to implement the policy reforms needed to achieve macroeconomic stability. In cases where stabilization measures were accomplished by deregulation and liberalization of prices and markets, significant supply responses over pre-reform levels have been achieved. However, the sustainability of agricultural growth, even in those countries where economic recovery and growth have been impressive, remains questionable. While macroeconomic and sector-specific reforms are necessary for economic recovery, they are not sufficient for sustained long-term agricultural growth. At best, they create an enabling economic environment for increased agricultural production. But in many Sub-Saharan African countries, agriculture's supply responses is constrained by poor rural infrastructure and high marketing costs, and by the absence of improved technologies. Overcoming these constraints will require significant public and private investments in rural areas.However, partly as a result of the financial retrenchments associated with structural adjustment programs, public investments in agriculture and rural areas have declined sharply in many Sub-Saharan African countries in recent years. This downward trend will need to be reversed if the full promise of the policy reforms is to be captured.Policy reforms also have an impact on natural resource management practices in rural areas. These effects can be complex and difficult to observe, but the available evidence suggests that many of the changes have been favorable and have contributed towards improvements in the condition of natural resources. Investments in land conservation and soil fertility improvements seem more likely to be successful given the poor infrastructure and high transport and marketing costs encountered in rural areas.Market liberalization policies can easily lead to increases in the prices of key inputs like fertilizers in these areas, as well as reduced access to input supplies and market outlets.Additional public investments are needed in many backward regions if they are to successfully compete in liberalized economics and to redress their poverty and environmental problems.The policy and institutional requirements for sustainable development depend structures which make states more accountable to local communities and their needs.Consistency in tenure institutions and authority is also important for fostering tenure security.When introducing technologies where robust institutions are in place, it is preferable to adapt technologies to be compatible with institutions rather than the reverse.Policy frameworks are needed to strengthen common property institutions which encourage the evolution of local collective action by devolving partial or complete authority for resource management to local communities and which actively facilitate the conditions for collective action. Institutions which can organize and balance competing claims, mediate conflicts, and promote stable and sustainable use patterns will tend to be the most resilient.In order to identify the key problems facing fragile lands, the workshop participants were organized into two working groups. The terms of reference for the working group discussions were as follows:1.What are the major problems of fragile lands?2.What are the reasons for these problems, and how are they interlinked?What are the consequences of these problems?4. What has (not) been done in the past to solve the problems?Based on discussions and lessons learned from the paper presentations, it became apparent that fragile lands could not be restricted to a uniform set of characteristics.Some may be more susceptible to degradation than others. Rather, fragile lands encompass both those lands which are highly susceptible to degradation (yet may have so far been exposed to relatively little exploitation) as well as those lands which initially may have had favorable characteristics in terms of resource endowments, climate and fertile soils, but have come under heavy exploitation such that now their status can be considered fragile in that they are unable to serve the long-term needs of the populations which depend on the land and its resources. Uncertainty and risk, whether environment-related or having to do with markets and prices, may create incentives to exploit resources unsustainably. This is especially true for the poor, whose emphasis on meeting immediate consumption needs may lead to practices which produce the highest return in the short run, but accelerate levels of degradation in the long run. Concerns over risk may also diminish adoption of technologies and natural resource management practices which have beneficial implications for the environment, if they are more sensitive to variable conditions or if payoffs are realized in the long term. Despite high risks, it is rare for rural communities in developing countries to have access to formal safety nets to protect them from risk. This can be attributed to policies which ignore the importance of insurance as well as a lack of public investment in infrastructure and other means of encouraging insurance market development.Although informal mechanisms of risk-coping may be in place, they are frequently insufficient to cope with large-scale or covariate risks or they may be absent or weak due to inadequate local institutions to support them.Local institutions and organizations have been shown to be crucial not only for managing risk, but also managing natural resources in general. The collective action necessary to ensure efficient, equitable, and sustainable resource management may not be sufficient if policies are lacking for encouraging local organizational development (or which may even go so far as deterring organization), participation costs are high, human capacity is weak, and populations are either too low to support collective action or too high to assure a reasonably low level of resource conflicts.Institutions for defining property rights to resources are also very important.Insecure property rights can contribute to decreased incentives for investment in long-term, environmentally sound technologies. Conflicts over land use may also arise from insecure tenure and lead to worsening degradation. Inheritance practices which involve subdividing smaller and smaller parcels of land have a tendency to induce higher levels of agricultural intensification in order to meet subsistence needs, thereby putting added pressure on fragile lands. They may also result in less secure forms of tenure for women, implying negative implications for resource management.  The result of many of these negative endogenous processes is not only worsening resource degradation, but also increasing poverty for those who reside in fragile land areas. As a result, the gap between those living in high potential areas and those occupying low potential areas widens and equity implications become more pressing.Unfortunately, this is not the end of the road since poverty itself can contribute to the perpetual cycle of resource degradation by eliciting higher discount rates on net returns to resource investments and through higher adversity to risk. High cost technologies and technologies involving higher risks may also be resisted in favor of more affordable and reliable methods that may not carry the environmental benefits of improved practices and technologies.African countries in the past to solve the problems identified. It was noted that actions have been undertaken in response to these problems at the technology, institutional and policy levels and have had varying degrees of success. However, limited time precluded much discussion among the groups of the performance of various measures. Whereas unfavorable natural conditions cannot be altered, actions have been undertaken to try to overcome their deleterious impact, such as non-traditional crops which are adapted to fragile landscapes, small-scale irrigation schemes to overcome limited water supplies, and crop insurance initiatives to enable people to cope with severe weather uncertainties.Efforts to curtail growing population pressures typically center on the promotion of family planning methods and education on 'family life' issues although programs which support women's employment, entrepreneurship, and self-reliance may also have an effect on lowering birth rates.More recently, research and NGO activities have called attention to improving the performance of local organizations through efforts to empower them and encourage selfreliance, use of participatory approaches such as Participatory Rural Appraisal methods, and decentralization of natural resource management authority and responsibilities. In a few cases, this has included policies aimed at devolving property rights to resources to local populations. In Kenya, this went so far as privatization via land adjudication and consolidation, although its intended impact on improving tenure security is questionable.Many African countries have initiated reviews of their land policies in the interest of addressing insecure tenure issues, in some cases raising awareness of gender issues. In addition to the above measures, resettlement schemes and land use reform measures have also been undertaken.To different degrees, country-level efforts have been put in place that create a more favorable policy environment to effectively deal with issues affecting fragile lands.Several African nations have undertaken economic reforms in the form of market liberalization and structural adjustment programs which have included reduced taxes on export products. In addition to the policies mentioned earlier in this section, other new or reform measures have included integrated rural development projects, small-holder development projects, decentralization of government authority, increased regional integration, formal recognition of NGOs, and greater consideration of gender including the creation of ministries to address women's issues. Public investments have been madein infrastructure improvements such as rural roads, irrigation, construction of water reserves, increased regional integration, and the coordination of input, production and marketing chains, with higher levels of investment often enabled by increased government borrowing. Subsidies continue to play a large role in supporting policy initiatives.In The terms of reference for the policy intervention group were as follows:1.Which policies have been in place to specifically address the problems of fragile lands?2.Which of these policies have worked? Why have they worked or not? How can one make them work?3.What has been the impact of macro and sectoral policies, e.g., impact of market liberalization on fragile lands (past and future)?In an effort to foster more sustainable resource use and management and to improve the natural status of fragile environments, African countries have put in place many policies directly addressing resource degradation and poor resource conditions. In addition to the above policy options, the group went on to discuss the impact of macro and sectoral policies on fragile lands. These policies included: exchange rate and market liberalization, financial deregulation, trade liberalization, investment policy, and other monetary and fiscal policies related to stabilization and structural adjustment. The impacts were assessed to be improved prices for tradeable goods, higher costs for imported inputs, more stable prices and increased availability of goods and services, improved incentives to invest, and market creation. However, such policies can lead to exploitation and food insecurity if they are not monitored and tailored to soften negative impacts on the poor.The terms of reference for the groups that worked on institutional intervention for the fragile lands included:1.Which institutions have been put in place to address problems of fragile lands?2.Which of these institutions have worked? Why have they worked or not? How can one make them work?3.What has been the impact of national institutional issues on fragile lands (past and future)?In addressing the terms of reference focused on institutions, Group C defined 'institutions' in the sense of organizations, whereas Group D defined 'institutions' as sets of rules and norms guiding human interaction. The discussions arising from each are therefore separated.Group C organized its discussion by listing various organizations according to their international, regional, national and local status and describing their particular functions. They then distinguished features of these organizations which worked well versus those which were less effective or did not work.Those identified under international institutions included FAO, IFAD, World At the regional level, organizational funding provides a mechanism for cooperation, but is seldom targeted to fragile lands given their limited capacity to address fragile land issues both technically and financially. Low priority is typically accorded to structuring policy and legal frameworks. Regional activities for fragile lands also tend to be poorly coordinated, partly because there are so many regional integration entities around. On the positive side, these organizations bring researchers, policymakers and farmers together and provide a panel for discussion among them, thereby giving voice to local interests. They further provide a forum for discussion of African issues and enable consensus building.The effectiveness of national organizations, such as ministries, environmental institutions, NGOs, universities, extension services, and farmer cooperatives were seen to vary depending on their functions. Ministries rated high on policy formulation, but had a mixed reputation for policy implementation and decentralizing authority. They maintained a poor record of establishing an enabling policy and institutional environment while mechanisms for monitoring the effect of policies were seen to be virtually nonexistent. Mixed reviews were also assigned to universities in their research capacity and to extension services with respect to the education and training they provide. Group D's discussion began by agreeing on a working definition for institutions as 'rules, norms and traditions designed to shape human actions', as opposed to organizations. Since formal rules encompassed mainly policies and laws, the group chose to focus on informal institutions. After identifying several institutions which operated in fragile land areas, the group highlighted the problems such institutions were designed to address, the country or region where they were practiced, their operational level, the degree of success they had in addressing the problem(s), and the reasons for success or failure.Common property institutions, which operate throughout Africa, are a means for communities to manage natural resources while maximizing resource benefits in a risky environment. Their degree of success is mixed. Whereas they can be a means of protecting vulnerable populations during shocks, such as drought, they may break down as a result of population growth and increased competition for land. Under these circumstances, intense resource pressures can create incentives to 'cheat'. In other cases, government policies, such as nationalizing resources, can undermine common property institutions.Many village-and community-level institutions operate around the management of particular resources, such as water, rangelands, trees, and soils. In some instances, they may evolve into specific organizations, such as water user associations and catchment area committees. The general assessment was that where these institutions had evolved locally and authority and management rested at the community level, they were largely successful although some experienced problems with long-term maintenance and strong pressures on resources. However, where there existed a high degree of government intervention and imposed rules, such as communal grazing and livestock watering schemes in East Africa, the tendency was for resources to be degraded. This was also the case with government programs in Tanzania and Uganda for mandatory terrace-making, group farming, river basins, and some afforestation schemes. Another problem relates to property rights being shifted from local users to the state, thereby removing incentives for communities to manage resources sustainably.The success of informal savings and credit groups operating in many African countries is attributed to factors such as the trust built among group members, mutual accountability, the locus of ownership and control being the group, small group size, voluntary membership, low transaction costs, and the production of individual benefits.Traditional labor pooling institutions designed to address manpower shortages are similarly based on cooperation and trust-building that arises from social pressures to participate. Likewise, livestock sharing which takes place in Ethiopia among small groups enables communities to cope with scarce land, labor and feed resources.Youth groups and migration institutions were also discussed. Youth groups in Burkina Faso and Mali are instrumental in slowing rural-urban exodus, addressing unemployment and providing development assistance to villages. In many cases, they have performed well as a result of the motivation derived from members being in the same age bracket. However, incentives to migrate to find employment have sometimes undermined their resilience. Migration, whether to urban areas in search of employment or other rural areas in search of better land, is a response to population and resource pressures, declining land productivity and a lack of alternative income sources.Migration institutions have had mixed success depending on the extent to which new areas can absorb incoming populations and their impact on resource productivity. In the short-term, they can relieve pressure on over exploited resources as well as channel money into rural areas via remittances.The group then turned its attention to what we have learned from observing institutional dynamics over time and what appear to be the criteria for their success in addressing environmental and poverty issues. In the case of increasingly scarce resources, two distinct trends emerge: either 1) institutional arrangements evolve to address an widening array of complex issues, or 2) traditional institutions break down,often leading to open access. Institutional success depends in large part on the degree to which they are conceived and controlled at the local level and on the extent to which they are based on trust, local ownership, and social control; externally imposed institutions are generally weak and unsustainable in practice. It is also important that individual benefits accompany social benefits given that participatory approaches involve high transaction costs for individuals. Mutual interdependency and homogeneity among individuals also increases the probability of local institutions succeeding by expanding the need for cooperation and lowering the likelihood of conflict. However, more attention needs to be directed to how to scale-up community institutions and how to create networks of farmers in order to enhance institutional impact.The terms of reference for the field trip to the Mabira forest reserve were as follows:1. What problems of fragile lands were observed in the field?What were the underlying causes?Which solutions have been found, in terms of technology, institutions, policy?What impact has been observed?The groups identified numerous problems observed in the field. Surprisingly few actually concerned resource degradation. Those that did mentioned poor soil fertility, degradation of roads and pest problems emanating from the forest. Rather, most problems centered on the conflict between state foresters and local villagers on who should manage the forest.It was observed that encroachment into the protected Mabira Forest by villagers for purposes of firewood collection and clearing for crop cultivation was problematic.When the forest was designated by the state as an ecotourism project, rights of neighboring villages to forest resources were appropriated, leaving them only with the right to collect fallen branches and kindling. Given the shortage of farmland and population growth in the area, this has become a source of tension between locals and the state. The project is administered in a very 'top-down' fashion by state foresters, with limited cooperation and interaction between villagers/farmers and forestry management.In fact, the community's role in forest management remains very unclear while policies for management of ecotourism are not transparent. As a result, the project fails to address farmers' needs and problems while providing them with few benefits, either economically or environmentally. Since its inception, ecotourism in the forest area has provided very few economic opportunities for the villagers in an environment that offers virtually no non-farm income alternatives to begin with. The benefits for forest protection are questionable as well in that there are not many forest guards and enforcement of forest policy is weak. There seemed also to be a lack of awareness on the part of farmers of the importance of forest conservation. Likewise, the project is deficient of any monitoring or impact assessment.Other problems noted were the lack of knowledge of existing agricultural technologies by farmers as a result of poor extension and technical assistance delivery.Yield-enhancing technologies are especially needed given the constraints farmers face in applying land extensification strategies as a result of the forest project. The high input prices and low producer prices farmers face exacerbate the productivity problem.Other observations revealed that the groups were not clear whether the lands they saw were actually fragile or high-potential. For the most part, the group felt the soils and land cover were good with seemingly high production potential. Similarly, it was not apparent that farmers were actually poor, although there did appear to be signs of poverty in the periphery of the Mabira Forest. Farmers appeared to be knowledgeable while the presence of few children may have indicated reduced population growth.Despite the extent of the problems, several measures had been undertaken to address them. The Mabira Forest Committee was established to deal with issues related to the management of the project and involved some of the influential members from the surrounding local communities. Opportunities for generating income from the forest and revenue sharing strategies with the community are being explored. On the protection side, awareness creation campaigns were organized which stressed conservation education and the importance of protecting the forest. Enhanced enforcement of forest policy involved evicting encroaching farmers, boundary demarcation and zoning of the forest. Planting and replanting activities aimed to improve tree densities and rejuvenate encroached areas.Forest rejuvenation efforts proved successful in tree regeneration, stimulating healthy forest growth, and reinitiating ecological diversity. However, organized campaigns instilled only a limited degree of awareness among villagers while efforts to incorporate local interests into management of the forest still had a way to go. Interests of villagers and foresters remained polarized and could potentially lead to social conflict, particularly since farmers felt resentful over the loss of land. Benefits generated by the forest included a certain degree of ecotourism and money to build some schools. Yet, it seemed that those who were poor remained poor and did not benefit much from the project.While the working groups recognized the importance of forest protection, they emphasized that more needed to be done to address poverty and growth. They recommended that awareness creation on the importance of forest/environment conservation be community-based and imparted from the primary school level upwards with the involvement of influential members from local communities. More education on family planning was also needed. With respect to forest management, they suggested administrative responsibilities be shared and that joint ownership and management of the forest be promoted. More proactive management of the forest by villagers was viewed as a means of increasing incomes. Improving yields necessitated more adaptive research and extension efforts aimed at increasing soil fertility so that farmers were less inclined to encroach on the forest areas. Extension services also need to be more active in making new technologies available to farmers. Fencing crop boundaries was suggested as a way to keep forest animals from invading crops. In terms of improving income, farmers' share of benefits need to be increased and marketing channels improved. Another possibility was to diversify into new activities, including raising livestock, pigs and poultry. Finally, it was noted that institutions to protect the water flow systems outside the forest needed to be developed.The groups who visited the wetlands focused primarily on three problems:C lack of appropriate technology, C perceived labor shortages and weak labor organization, and C land tenure insecurity.Technology deficiencies led to such problems as water logging, cutting fuelwood for firing bricks, and scooping land to gather material for brick-making. Causes stemmed from the high investment cost of developing wetlands for agriculture, insufficient diffusion of information on technologies, limited access to capital, inexperience in intensive land use, and lack of alternatives for construction materials, drainage, and sources of fuel. In the case of rice farming, lack of fertilizer technology means there is no nutrient replacement in the soil. Obtaining organic matter is costly to obtain due to the distance from town. It was recommended that an extension package be developed for the wetlands which is people centered, economically viable, and environmentally sound.Efforts to make people aware of the package and widespread dissemination should also be undertaken.Although at the outset there appeared to be a labor shortage given the widespread use of prison labor by farmers, the group thought this perception might arise from labor allocations in these newly developed areas still being unstable. Also, farmers may prefer prison labor to hired labor if the cost is comparable since the fact that the former were accompanied by guards kept supervision costs down. It was decided that no intervention was needed in this area, but rather the labor market would likely develop on its own.Whereas there did not appear to be a labor shortage in the wetlands, there were issues of land tenure security. The \"Environmental Act\" instituted by the government prohibits 'unsustainable' use of wetlands while there is substantial ambiguity regarding who owns these lands, if anyone. The farmers occupying the land now do not have title and it appears that they do not have the legal right to acquire one. The complexity and uncertainty emanating from the legal system on land rights have implications for tenure security and whether sufficient incentives exist to make sustainable investments in the wetland areas. Action was recommended to clarify and harmonize land policies so that legislation on property rights and environment were compatible. Once established, effective mechanisms for implementing and enforcing property rights need to be put in place.The final session of the workshop sought to build on the discussions and ideas generated from previous working group session in order to formulate a series of recommendations directed at policymakers, researchers, and donor organizations. Rather than dividing into the original four groups, conference participants opted whether to join either the group developing recommendations on highlands or the one addressing The argument was made that lack of research is not a problem; there is ample information. Rather, there is a problem of insufficient information dissemination and extension. Also, there is a lack of appropriate knowledge on how to manage resources.It was suggested that another means be used for categorizing issues, which distinguished between ends (e.g., poverty) and means (e.g., infrastructure and investment). Of the issues identified, food insecurity and resource degradation are 'ends'. Everything else could be considered a 'means'. Caution was expressed with respect to having too narrow an approach and focusing only on agriculture. Growth is also an end, but is not achieved only by addressing agricultural problems.Poverty alleviation and food security were cited as the overarching goals, conditioned on sustainable natural resource management, whereas the priority problems (related to the 'means' for attaining the primary goals) selected by the group included: Once the group reconvened, a rapporteur from each small group presented the recommendations they formulated. Others from the larger group were charged with assessing: 1) are the recommendations concrete enough? 2) is there specificity with regard to countries? 3) to whom are we recommending?The group assessed the pros and cons of different alternatives, addressing the government as the client.Land registration (with and without a cadastral survey) targeting areas of higher population density. Where there is no cadastral survey, local institutions need to be codified.Resettlement programs to address land fragmentation. Negative effects are likely where resettlement is forced. Instead, voluntary incentives are needed, although there may be high costs associated with this.Government marketing of inputs. The private sector is not likely to fill in when subsidies are removed. May need to phase out subsidies gradually.Government provision of output and price information. The cost to government is low, but institutional capacity needs to be developed.Formulation of investment policy which provides incentives (e.g., tax breaks) and provides 'one-stop-shop' investment information.Institution of credit policy for deregulating rural financial institutions and encouraging the development of rural financial institutions (e.g., initial assistance, seed money).Institution of forest policy that advocates benefit/income-sharing between governments and local communities for forest management. Also, policy should allow for private use of degraded land for woodlots and agroforestry.The highland group identified three investment sources: 1) farmers, 2) private sector, and 3) government.1) Farmer investment in agriculture.Problem: lack of appropriate rural financial institutions in rural areas due to 1) high transaction costs of rural lending and lending to poor; 2) incompatibility of conventional banking practices with small farmer realities (e.g., using land as collateral).Recommendations:Development of suitable rural financial institutions (RFIs) and banking practices which lower the transaction costs of banking and incorporate alternative collateral mechanisms so as to match the socio-economic realities of local people (e.g., savings mobilization).Institute rural finance policy which is supportive of:-alternative rural financial institutions, including informal institutions, -financial intermediation (as opposed to purely credit) approach -savings mobilization, including for non-formal RFIs, especially where it has traditionally been restricted, -institutionalizing non-traditional forms of collateral, -restructuring the formal financial sector based on best-practices demonstrated by the informal sector (so as to lower transaction costs), -liberalizing interest rates on savings and loans, -liaisons between larger, formal financial institutions and local informal institutions.Undertake research on non-formal rural financial practices.Establish training programs designed to enhance the capacity of formal financial institutions to adapt to more appropriate practices for rural clients.Direct funding to best-practice rural financial institutions (including NGOs), particularly for purposes of building loan portfolios and capital funds.2) Private sector investment in agriculture.Problem: How to get the private sector to invest in rural areas to generate alternative income sources and foster other means of addressing poverty alleviation.Recommendations:Increase government investment in infrastructure improvement,Promote policies which give preferential treatment to small scale industries (e.g., milling, food-processing, tailoring),Formulate policies which enhance the purchasing power of the poor, so C Promote agroprocessing and forestry related industries.3) Government investment in agriculture.Problem: Although public investment is necessary to promote private investment by farmers and private sector industries, governments often lack the financial resources and are immersed in political conflicts and constraints. Also poor returns to infrastructure investment can act as a disincentive (chicken and egg problem), although smaller projects are likely to be more viable.Recommendation: Increased public investment in smaller, higher return infrastructure projects.Recommendations:Revive technologies that are 'on the shelf' (not implemented) and establish why they were not disseminated or adopted,Strengthen the capacity to do research in the highlands (e.g., research on technology adoption in these environments which takes into account local realities and is multi-disciplinary in its approach),Create an enabling environment for researcher/extension/farmer linkages: information dissemination.Recommendations:Initiate resettlement programs (for temporary relief) by reserving land for resettlement purposes and designating areas for reclamation (e.g., insectinfested or disease-ridden areas).Establish a land tax to encourage land to be released in the highlandshowever, this requires land markets and titles.Initiate education programs and incentives geared toward slowing population growth.Intensify production via targeted subsidies, irrigation, infrastructure development and maintenance, facilitating access to markets.Promote off-farm employment, such as agroprocessing, and provide support services for management.Recommendations:Improve market integration via constructing roads, reducing transportation costs, improving access to inputs and marketing of outputs.Improve access to public servicesIncrease investment in rural roads and communication, electrification, healthcare provision, and schools -depending on an assessment of the returns, investing the greatest shares of resources so as to reap the highest returns.The group then reacted to the recommendations presented and offered suggestions as outlined below.Emphasis was put on the need to harmonize environmental and agricultural policies and integrate across sectoral policies, i.e., cross-sectoral policy linkages.The point was made that land registration is needed only where population pressures are high, and that cadastral surveys may be appropriate only where they are economically feasible. Generally, they tend to be needed in areas where there is disputed ownership. In this case, adjudication is also needed. Adjudication is not needed, however, if there are minimal disputes and widespread recognition of rights.The importance of implementing good macroeconomic policies was noted although debate ensued as to whether structural adjustment policies are appropriate for fragile areas. Rather, one needs to account for the uniqueness of specific areas. There was some disagreement on whether government marketing of inputs to remote areas was a good policy. However, it was agreed that this depended on the state of infrastructure and road access and that the policy was meant to be only short-term until infrastructure caught up.Concern was expressed that limited government budgets may preclude the adoption of recommendations on public investment. However, recommendations for private-sector investment needed to be pushed further.More clarity was needed to know to whom the recommendations should be targeted. It was suggested that the recommendation for private sector research should be moved under the specific research area.The group suggested that more meat be added to the recommendations. can also be a guide for where there is a need to divest from research.It was mentioned that there is much sensitivity to the issue of reduced population growth. Often countries and communities perceive a need to increase their populations.Greater specification was needed as to when diminishing returns are realized as well as carrying capacity reached. One proposal was made to treat population pressure as one of the goals, thereby invoking a Critical Square. --Formulate policies which:1.Recognize savings mobilization by non-formal financial institutions, especially where savings mobilization has been illegal for them.Continue efforts to liberalize interest rates,Institutionalize non-traditional forms of collateral.--Governments and donors should take action to:4. Undertake research on non-formal rural financial practices in hillsides/highlands areas,Establish training programs designed to enhance the capacity of formal financial institutions to adopt these practice for small scale clients in highland areas.Direct funding to best-practice rural financial institutions (including NGOs), especially for loan portfolios and capital funds.1.Establish an inventory of all research undertaken to address problems of hillside areas, assess their adoption and appraise reason for them not having been accepted.Strengthen the capacity to do research, particularly in the following areas: --In areas of high population density and low-medium agricultural potential where market access is already good, infrastructure investments should focus on maintenance of existing road networks as well as supplementary irrigation.The question was raised as to how realistic the proposal was to give public funds to farmers to commission their own research as a way of achieving more demand-driven, site-specific research. Are governments going to invest in research? If farmers don't see the full value of public goods, will they under invest? Rather, change needs to take place in the incentive structure within research. Research institutions need to be reformed to make them more participatory and to do site-specific work.Also, while market policies can be good for countries, they can have negative impacts. Modifiers may be needed to buffer negative impacts.The Highlands group noted the inclusion of recommendation for government subsidy of transportation costs to buffer the impact of marketing policies. With respect to research, they noted that while it should be publically funded, greater involvement of communities was needed, with part of the money for research going to farmers organizations to enable them to be in the driver's seat. Tanzania offers an example of demand-driven research whereby farmers participate in identifying priorities and funds are allocated to rural areas to address priorities.Concern was expressed on the potential effectiveness of resettlement programs in terms of having to establish new eco-systems and the likelihood of substantial opportunity and moving costs. However, the group responded that infrastructure can be used as an incentive for resettlement with the idea being to induce resettlement in places where there is scope for agricultural potential.A final comment was made on the need to put less emphasis on research and to more actively examine new ways of funding and prioritizing farmer-driven extension activities.The group preceded its recommendations by articulating an overall vision and related strategies, as follows:Vision: To increase human well-being in fragile lands without degrading the natural resource base.  C How can we structure monitoring and evaluation so that it feeds back into the existing regulatory framework?Another comment centered on the need for different approaches tailored to the different environmental realities of highlands and drylands. In the highlands, there are terrain problems, population is high, and production is intensive. In the drylands, populations are low and production is extensive. In the highlands, road construction is difficult because of the terrain and non-conventional means of transportation may be needed. It was suggested that perhaps the groups had not thought sufficiently about the particularities of these regions.It was argued that GIS as discussed in the Wood presentation can be used for this purpose. The question was then whether governments should invest in an Africa-wide GIS database. If so, this would have to be built up and needs lots of micro-level data. To do this effectively, capacity needs to be built in African countries to collect data and to use it.Thanks were extended by the chairperson, Bruno Ndunguru, to the participants, interpreters, facilitators, and logistical and back-up persons. He commended the excellent work of the facilitators, the support from NARO, the hotel management, and the government of Uganda for the peace enjoyed during the meetings. He noted that it was a participatory meeting, whereby all participants had a collective responsibility for the recommendations, for which they took full ownership. While there would be gaps and issues not fully addressed, he expressed hope that the participants would use these shortcomings to make improvements in the future. Ndunguru thanked DSE and the Advisory Committee for the meeting, with a special 'thank you' to Juergan Richter of DSE. Thanks were also extended to the farmers who participated in hosting the field trip.The chair person reviewed the meeting's objectives to assess whether the group had achieved them:Overall goal: Balanced agricultural growth and food security together with sustainable resource management and environmental sustainability. Ndunguru announced that the outlines of the presentations would be available that afternoon.Concluding remarks were delivered by the Prof. J.K. Mukiibi of NARO: \"The opening ceremony emphasized following up on workshop recommendations. We have all worked hard over the past four to five days, even nights, and produced an impressive set of recommendations and notes. Some of the areas we have emphasized are that there should be improved macro policies and marketing in both the highlands and drylands, land and environmental policies, rural development, technology transfer, infrastructure and financial investment, both in the highlands and drylands. The government of Uganda is doing a lot in these areas. Macroeconomic policies are especially important.Also critical are marketing and land issues. There has been a debate over the past two to three weeks over proper utilization of land. Rural financial institutions are important because without them, we revert to traditional agriculture.\"Being the head of an agency promoting technology adoption, Dr. Matthias Magunda of NARO assured the participants that he had noted these recommendations carefully and their importance, saying that they are problems the Uganda government is trying to address. These recommendations he felt would reassure government officials that they are on course in the area of fragile lands.Having been sent by the Minister to represent him at the closing ceremony, he read the statement the Minister had asked him to read: \"This conference is timely given the dimensions of the problems in the region. It is structured in a way that optimizes sharing of experience. The methodology of the conference is a productive way to tap expertise. Identification of key issues of fragile land is very important, as is the linkage of problems and causes. These have contributed greatly to the policy triangle, not only by identifying policies, but also in evaluating why they have worked or not worked and how they can work. This analysis is important before advancing forward in attacking fragile land problems. Key strategies are to achieve agricultural intensification and poverty alleviation without compromising environmental sustainability. These strategies are right on target, and those involved in these activities are commended. As for the field trip, the nearest fragile lands to Kampala were selected. However, those further out are more fragile. We are very grateful that the objectives of the conference have been fulfilled and appeal to those involved in the follow-up activities to use the recommendations of the conference in designing intervention policies and programs. We extend thanks to IFPRI for tackling global and regional problems and for facilitating the workshop and also express our thanks to DSE and the EU for funds. You are always welcome to Uganda.\"","tokenCount":"8445","images":[],"tables":["-1401715296_1_1.json","-1401715296_2_1.json","-1401715296_3_1.json","-1401715296_4_1.json","-1401715296_5_1.json","-1401715296_6_1.json","-1401715296_7_1.json","-1401715296_8_1.json","-1401715296_9_1.json","-1401715296_10_1.json","-1401715296_11_1.json","-1401715296_12_1.json","-1401715296_13_1.json","-1401715296_14_1.json","-1401715296_15_1.json","-1401715296_16_1.json","-1401715296_17_1.json","-1401715296_18_1.json","-1401715296_19_1.json","-1401715296_20_1.json","-1401715296_21_1.json","-1401715296_22_1.json","-1401715296_23_1.json","-1401715296_24_1.json","-1401715296_25_1.json","-1401715296_26_1.json","-1401715296_27_1.json","-1401715296_28_1.json","-1401715296_29_1.json","-1401715296_30_1.json","-1401715296_31_1.json","-1401715296_32_1.json","-1401715296_33_1.json","-1401715296_34_1.json","-1401715296_35_1.json","-1401715296_36_1.json","-1401715296_37_1.json","-1401715296_38_1.json","-1401715296_39_1.json","-1401715296_40_1.json","-1401715296_41_1.json","-1401715296_42_1.json","-1401715296_43_1.json","-1401715296_44_1.json","-1401715296_45_1.json","-1401715296_46_1.json","-1401715296_47_1.json","-1401715296_48_1.json","-1401715296_49_1.json","-1401715296_50_1.json","-1401715296_51_1.json","-1401715296_52_1.json","-1401715296_53_1.json","-1401715296_54_1.json","-1401715296_55_1.json","-1401715296_56_1.json","-1401715296_57_1.json","-1401715296_58_1.json","-1401715296_59_1.json","-1401715296_60_1.json","-1401715296_61_1.json","-1401715296_62_1.json","-1401715296_63_1.json","-1401715296_64_1.json","-1401715296_65_1.json","-1401715296_66_1.json","-1401715296_67_1.json","-1401715296_68_1.json","-1401715296_69_1.json","-1401715296_70_1.json","-1401715296_71_1.json","-1401715296_72_1.json","-1401715296_73_1.json","-1401715296_74_1.json","-1401715296_75_1.json","-1401715296_76_1.json","-1401715296_77_1.json","-1401715296_78_1.json","-1401715296_79_1.json","-1401715296_80_1.json","-1401715296_81_1.json"]}

data/part_2/0083727155.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"95290e15ff0e7f78912b00f1e9fd143f","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/ea3a7f9a-5f6d-48f5-9057-0fa124332bf1/retrieve","description":"as part of IFPRI EGYPT SEMINAR SERIES","id":"1810204768"},"keywords":[],"sieverID":"c063645b-f36b-4f94-8409-db9ff1acc317","pagecount":"32","content":"Internal policies are closely linked to establishing a democratic political system that respects human rights principles and based on the rule of law; for:• establishing a democratic society characterized effective executive institutions,• strengthening and supporting the role of the Parliament in legislation and supervising the executive authority,• eliminating the phenomena of nepotism and favoritism,• establishing an effective civil society that supports decentralization, thus, alleviating the burden on the central government and increasing opportunities for political participation,• supporting and enabling political parties as a guarantee of democracy,• activating and enabling the political system and its institutions to cope with global developments,• establishing a free and pluralistic society that empowers youth and women, fights terrorism, and achieves a prompt justice.Setting the target (where we want to be)  Combating the phenomena of slums and insecure areas. Encouraging the use of green and sustainable building methods. ","tokenCount":"143","images":["1810204768_1_1.png","1810204768_1_2.png","1810204768_1_3.png","1810204768_2_1.png","1810204768_2_2.png","1810204768_2_3.png","1810204768_3_1.png","1810204768_3_2.png","1810204768_3_3.png","1810204768_4_1.png","1810204768_4_2.png","1810204768_4_3.png","1810204768_4_4.png","1810204768_5_1.png","1810204768_5_2.png","1810204768_5_3.png","1810204768_5_4.png","1810204768_6_1.png","1810204768_6_2.png","1810204768_6_3.png","1810204768_6_4.png","1810204768_6_5.png","1810204768_6_6.png","1810204768_6_7.png","1810204768_6_8.png","1810204768_6_9.png","1810204768_6_10.png","1810204768_6_11.png","1810204768_6_12.png","1810204768_6_13.png","1810204768_6_14.png","1810204768_6_15.png","1810204768_6_16.png","1810204768_6_17.png","1810204768_6_18.png","1810204768_6_19.png","1810204768_7_1.png","1810204768_7_2.png","1810204768_7_3.png","1810204768_7_4.png","1810204768_8_1.png","1810204768_8_2.png","1810204768_8_3.png","1810204768_8_4.png","1810204768_9_1.png","1810204768_9_2.png","1810204768_9_3.png","1810204768_9_4.png","1810204768_9_5.png","1810204768_9_6.png","1810204768_9_7.png","1810204768_9_8.png","1810204768_9_9.png","1810204768_9_10.png","1810204768_9_11.png","1810204768_10_1.png","1810204768_10_2.png","1810204768_10_3.png","1810204768_10_4.png","1810204768_11_1.png","1810204768_11_2.png","1810204768_11_3.png","1810204768_11_4.png","1810204768_11_5.png","1810204768_12_1.png","1810204768_12_2.png","1810204768_12_3.png","1810204768_12_4.png","1810204768_12_5.png","1810204768_13_1.png","1810204768_13_2.png","1810204768_13_3.png","1810204768_13_4.png","1810204768_13_5.png","1810204768_14_1.png","1810204768_14_2.png","1810204768_14_3.png","1810204768_14_4.png","1810204768_14_5.png","1810204768_15_1.png","1810204768_15_2.png","1810204768_15_3.png","1810204768_15_4.png","1810204768_15_5.png","1810204768_16_1.png","1810204768_16_2.png","1810204768_16_3.png","1810204768_16_4.png","1810204768_16_5.png","1810204768_17_1.png","1810204768_17_2.png","1810204768_17_3.png","1810204768_17_4.png","1810204768_17_5.png","1810204768_18_1.png","1810204768_18_2.png","1810204768_18_3.png","1810204768_18_4.png","1810204768_18_5.png","1810204768_19_1.png","1810204768_19_2.png","1810204768_19_3.png","1810204768_19_4.png","1810204768_19_5.png","1810204768_20_1.png","1810204768_20_2.png","1810204768_20_3.png","1810204768_20_4.png","1810204768_20_5.png","1810204768_21_1.png","1810204768_21_2.png","1810204768_21_3.png","1810204768_21_4.png","1810204768_22_1.png","1810204768_22_2.png","1810204768_22_3.png","1810204768_23_1.png","1810204768_23_2.png","1810204768_23_3.png","1810204768_24_1.png","1810204768_24_2.png","1810204768_24_3.png","1810204768_24_4.png","1810204768_24_5.png","1810204768_25_1.png","1810204768_25_2.png","1810204768_25_3.png","1810204768_25_4.png","1810204768_25_5.png","1810204768_26_1.png","1810204768_26_2.png","1810204768_26_3.png","1810204768_26_4.png","1810204768_26_5.png","1810204768_27_1.png","1810204768_27_2.png","1810204768_27_3.png","1810204768_27_4.png","1810204768_27_5.png","1810204768_27_6.png","1810204768_27_7.png","1810204768_28_1.png","1810204768_28_2.png","1810204768_28_3.png","1810204768_28_4.png","1810204768_28_5.png","1810204768_29_1.png","1810204768_29_2.png","1810204768_29_3.png","1810204768_29_4.png","1810204768_30_1.png","1810204768_30_2.png","1810204768_30_3.png","1810204768_30_4.png","1810204768_30_5.png","1810204768_31_1.png","1810204768_31_2.png","1810204768_31_3.png","1810204768_31_4.png","1810204768_31_5.png","1810204768_31_6.png","1810204768_32_1.png","1810204768_32_2.png","1810204768_32_3.png","1810204768_32_4.png","1810204768_32_5.png","1810204768_32_6.png"],"tables":["1810204768_1_1.json","1810204768_2_1.json","1810204768_3_1.json","1810204768_4_1.json","1810204768_5_1.json","1810204768_6_1.json","1810204768_7_1.json","1810204768_8_1.json","1810204768_9_1.json","1810204768_10_1.json","1810204768_11_1.json","1810204768_12_1.json","1810204768_13_1.json","1810204768_14_1.json","1810204768_15_1.json","1810204768_16_1.json","1810204768_17_1.json","1810204768_18_1.json","1810204768_19_1.json","1810204768_20_1.json","1810204768_21_1.json","1810204768_22_1.json","1810204768_23_1.json","1810204768_24_1.json","1810204768_25_1.json","1810204768_26_1.json","1810204768_27_1.json","1810204768_28_1.json","1810204768_29_1.json","1810204768_30_1.json","1810204768_31_1.json","1810204768_32_1.json"]}

data/part_2/0096539937.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"77456dbaf44bdb75557366b98161b442","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/a61b686e-7876-4cd7-9f86-a3605f9f565a/retrieve","description":"","id":"-1838491462"},"keywords":[],"sieverID":"e1bff2f5-f01a-4b77-b6eb-180535c0c5b7","pagecount":"4","content":"District Nutrition Profiles (DNPs) are available for 707 districts in India. They present trends for key nutrition and health outcomes and their cross-sectoral determinants in a district. The DNPs are based on data from the National Family Health Survey NFHS-4 (2015NFHS-4 ( -2016) ) and NFHS-5 (2019NFHS-5 ( -2021)). They are aimed primarily at district administrators, state functionaries, local leaders, and development actors working at the district-level.  Note: NA refers to data unavailable for a given round of NFHS/Census.• What are the trends in infant and young child feeding (early initiation of breastfeeding, exclusive breastfeeding, timely initiation of complementary feeding, and adequate diet)? What can be done to improve infant and young child feeding?  Note: NA refers to data unavailable for a given round of NFHS/Census.• How can the district increase women's literacy, and reduce early marriage, if needed?• How does the district perform on providing drinking water and sanitation to its residents? Since sanitation and hygiene play an important role in improving nutrition outcomes, how can all aspects of sanitation be improved? • How can programs that address underlying and basic determinants (education, poverty, gender) be strengthened?• What additional data are needed on food systems, poverty or other underlying determinants? Note: NA refers to data unavailable for a given round of NFHS/Census.• How does the district perform on health and nutrition interventions along the continuum of care? Does it adequately provide both prenatal and postnatal services to women of reproductive age, pregnant women, new mothers and newborns? • How has access to health and ICDS services changed over time (food supplementation, health and nutrition education and health checkups)?","tokenCount":"269","images":["-1838491462_1_1.png","-1838491462_1_2.png","-1838491462_1_3.png","-1838491462_1_4.png","-1838491462_1_5.png","-1838491462_1_6.png","-1838491462_1_7.png","-1838491462_1_8.png","-1838491462_1_9.png","-1838491462_1_10.png","-1838491462_1_11.png","-1838491462_1_12.png","-1838491462_2_1.png","-1838491462_2_2.png","-1838491462_2_3.png","-1838491462_2_4.png","-1838491462_2_5.png","-1838491462_2_6.png","-1838491462_2_7.png","-1838491462_2_8.png","-1838491462_2_9.png","-1838491462_2_10.png","-1838491462_2_11.png","-1838491462_2_12.png","-1838491462_3_1.png","-1838491462_3_2.png","-1838491462_3_3.png","-1838491462_3_4.png","-1838491462_3_5.png","-1838491462_3_6.png","-1838491462_3_7.png","-1838491462_3_8.png","-1838491462_3_9.png","-1838491462_3_10.png","-1838491462_3_11.png","-1838491462_3_12.png","-1838491462_4_1.png","-1838491462_4_2.png","-1838491462_4_3.png","-1838491462_4_4.png","-1838491462_4_5.png","-1838491462_4_6.png","-1838491462_4_7.png","-1838491462_4_8.png","-1838491462_4_9.png","-1838491462_4_10.png","-1838491462_4_11.png","-1838491462_4_12.png"],"tables":["-1838491462_1_1.json","-1838491462_2_1.json","-1838491462_3_1.json","-1838491462_4_1.json"]}

data/part_2/0097661653.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"2d9ca50a5e5147b0541f2e838a4e6f02","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/b5324b65-2c11-4cb2-aeb5-2ba8a2ca745d/retrieve","description":"The System-wide Program for Collective Action and Property Rights (CAPRi) sponsored an International Conference on Policy and Institutional Options for the Management of Rangelands in Dry Areas, May 7-11, 2001 in Hammamet, Tunisia. The conference focused on institutional aspects of rangeland management and brought together policy makers and researchers from North Africa, Sub-Saharan Africa and West Asia to discuss sustainable rangeland production strategies and livelihood of pastoral communities in dry areas. This conference summary paper contains summaries of the CAPRi sponsored research findings on institutional options for rangeland, policy makers’ interventions and reactions as well as the synthesis of discussion groups. These working groups evaluated outcomes of policies and institutions guiding rangeland management in terms of their impact on livelihoods and environmental sustainability, and explored alternative policies and institutional strategies in light of their capacity to reduce poverty and enhance food security.\" -- Author's Abstract","id":"-1615607129"},"keywords":["Rangelands","institutions","natural resource management","property rights","collective action","case studies","Africa","sub-Saharan Africa","West Asia","North Africa"],"sieverID":"4fe0fdfe-ef26-4081-bc2f-410a6768504c","pagecount":"76","content":"In many African and West Asian countries there is particular concern about the degradation and, in some regions, continued desertification of rangeland areas, and the social, economic, and environmental impact resulting from these processes. However, until recently governments and development agencies accorded semi-arid rangeland areas relatively low priority and most interventions have concentrated on technical solutions to improve range productivity.There is a debate in the literature as to how much degradation of rangelands in arid and semi-arid regions is due to unpredictable changes in rainfall patterns, and how much is due to misuse by agro-pastoral populations. Nonetheless, low and declining productivity, increased impoverishment and vulnerability of pastoral peoples, and the increase in conflicts in these regions is still considered to be caused by inappropriate land use policies, multiple and contradictory legal systems (state, customary/religious) over pastoral resources, population pressures, and the disruption of pastoral production strategies and mechanisms that govern herder-farmer relationships -in addition to low and erratic rainfall patterns. Different types of tenure reform, ranging from privatization to common property to state ownership arrangements have been explored to support the improvement of rangelands and the development of pastoral communities, as have other institutional reforms such as the reorganization of pastoral communities into cooperatives and pastoral associations.Results from these reforms differ from country to country. Understanding their impacts on livestock production and livelihood strategies of herding communities requires systematic evaluation in order to draw lessons for designing adequate policy and institutional frameworks. There is a general consensus amongst researchers and development practitioners on the need to reconcile the different institutional approaches to pastoral development (e.g. various property rights, mobility, access options), to enhance the enabling environment under which livestock producers operate, and to promote greater participation of local communities in the management of natural resources.In response to these critical policy issues, the System-wide Program on Collective Action and Property Rights (CAPRi) of the Consultative Group on International Agricultural Research (CGIAR) organized an international conference to review the results of research and identify key policy recommendations. The conference was held under the patronage of the Minister of Agriculture of Tunisia, the conference host country. The case studies presented at the conference were undertaken by three CGIAR Other local national institutions as the Institution for Agricultural Research and High Education of Tunisia, played an active role in the organization of the event. The support and encouragement of the Tunisian Ministry of Agriculture, the high interest among regional institutions, as well as the participation of policymakers from 12 African and West Asian countries indicates the breadth of interest in the issues discussed at the conference.The conference brought together over 50 participants from Algeria, Burkina Faso, Jordan, Ethiopia, Iraq, Kenya, Eritrea, Morocco, Niger, Syria, Tunisia, Uganda; including representatives from pastoral organizations, local and national government ministries, national, regional, and international research institutions and organizations working in the fields of agriculture, natural resource management and policy formulation. The broad goal of the conference was to contribute to sustainable rangeland production strategies and livelihood of pastoral communities, especially in African and West Asian countries, through the participatory formulation of strategies for sustainable range management.More specifically, five objectives for the conference were defined as follows:• Presentation of the principal results and conclusions of CAPRi-sponsored research on rangeland management to policymakers and others involved in rangeland management.• Discussion of current government policies, and the practical policy and implementation issues faced in setting rangeland policy in those countries.• Identification of the appropriate medium-and longer-term roles of rangelands in contributing to poverty eradication and food security.• Initiation of dialogue, through three working group sessions, amongst participants on the key issues identified, possible solutions and implications for future economic, social and environmental policies.• Evaluation of the consequences, in terms of impact on livelihoods and environmental sustainability, of alternative institutional options and strategies for different types of rangelands and livestock production systems. This paper summarizes the paper presentations and discussions, as well as recommendations developed by the conference working groups.This International Conference addresses important aspects of rangelands management and Tunisia has a special interest in the promotion of rangelands and development of policy options for improving the management of these areas.So far, the major focus of researchers and of research centers has been on technological aspects, but technical solutions alone have failed to solve the problems of rangelands management. The understanding of social and historical processes, current local organizations and the creation of associations for developing rangelands are crucial to achieving positive results.In Tunisia, there are around 2500 water associations, whose members are elected every year and whose work is supervised by the government. Tunisian rangelands support 30% of the national livestock population under very unstable climatic conditions especially with respect to rainfalls and a 3-year drought is not a rare event. The development of rangeland should integrate many aspects of the social and economic life of the country. Education, and in particular education for women, potable water supplies, infrastructures such as roads and local health care units, are part of an integrated plan for development of rangelands. It is also important to recognize the variability and diversity of local circumstances and take them into account when formulating policies.Today, Tunisia's agricultural sector satisfies the national milk and vegetables demand and exports its products to the European Union. However, the big advances experienced in the agricultural sector have not occurred in the management of rangelands. For this reason further research is needed. Technological research should go hand in hand with institutional and social research to address rangeland management problems in an integrated fashion.Tunisia's research and development planning for the future relies on the following key aspects:• Rangeland projects should be integrated with projects in other sectors.• Extension, in the form of training and farming consultation, should be an integral component of any development project and projects should improve the functioning of local government institutions, communities and herder associations to achieve greater efficiency.• The study of conflicts and the development of mechanisms for conflict resolution are crucial for successful implementation of development projects and improvement of rangeland management.• With respect to climatic variables the protection of rangelands during drought years is a crucial aspect, which has major implications for the vulnerability of rangeland populations that represent the poorest section of Tunisia's society.The National Institute for Agricultural Research of Tunisia (INRAT) has recently started a 7-year rangeland management project in the southern part of the country that focuses on the following:• Modernization• New technology for water• Creation of localized service centers• Education campaignsThe project is based on scientific research of a multidisciplinary team that includes agronomists, economists and sociologists. The promotion of rangeland development is today a priority for the Tunisian government. In the past, research was done in laboratories and pastoralists' needs were often neglected. Another mistake consisted of addressing rangeland issues separately from other sectors and activities.Ongoing and recent projects are trying to achieve the integration between rainfed agriculture, rangeland management and other sectors by designing an overall development policy. The development of rangelands is crucial for decreasing the vulnerability of the poorest section of Tunisia's population and alleviating poverty. These new integrated approaches and policy recommendations formulated at this International Conference will be Tunisia's best weapon to combat poverty. The country case studies focused partly on testing the hypothesis of whether different range management systems (state, community, cooperative) improve the welfare of pastoral households. This research built on the country reviews of the policy and legal environments under which pastoral communities make their decisions and RRAs in 10-15 communities in each country. The data were used to characterize the pastoral communities, range management options and the constraints of pastoral communities. This exercise was followed by an in-depth pasture characterization to evaluate range productivity and floristic composition under each management option; and in-depth household surveys to evaluate the effects on these options on household feed expenditures.Household data were collected amongst 292 households in Jordan, 325 households in Morocco, 265 households in Tunisia and 3-year monitoring data on 69 households in the Jub-Jamaa community in Syria. Econometric analysis was used to evaluate the effects of the different range management options on the welfare of the pastoralists and the strategies they use to access additional grazing resources. The preliminary results of the analyses are presented in the different country papers.In the West Asia and North Africa region (WANA), small ruminants contribute to a large proportion of farmers', nomadic and semi-nomadic herders' income. In the 1950s, livestock production depended mainly on rangelands that provided 70 percent of the feed needs of small ruminants. But at present, natural grazing has declined to 10-25 percent, due to the continuous increase of flock numbers and removal of vegetation through plowing or for fuelwood. To address some of the loss in rangeland productivity, governments of the Mashreq and Maghreb (M&M) countries carried out numerous policy and institutional reforms along with technological innovations. Even though many of these countries tried to enhance the decision-making environment of pastoral institutions, it is difficult to find a balance between the rights and roles of traditional pastoral communities and those of the state and its institutions. In most cases, policy and institutional reforms weakened pastoral institutions. The institutional reforms can be classified into three main approaches.The first approach consisted of state appropriation of rangeland resources and was used by the majority of the M&M countries, as governments assumed that they were better equipped to manage rangeland resources. Along with tenure reforms, traditional tribal communities were reorganized into cooperatives. However, traditional institutions continued informally to manage range resources, although they did not have any legal rights over these resources. Such actions led to conflicts and disputes. In recent years, more emphasis is being placed on encouraging the participation and involvement of communities in the management of their resources (e.g. IFAD, AFESD, FAO and UNCCD projects in Jordan and Syria), but a legal framework to support such efforts is lacking.The second policy option consisted of strengthening customary tribal claims.Under this option, pastoral communities have full control over their resources and continue to use traditional mechanisms and rules to define access and resource use for all community members. This framework, however, does not address intercommunitiy access options and by confining livestock grazing on tribal resources reduces actual mobility.The third option is privatization with titling, which has been tried mainly in A number of different herding communities reorganization policies have been implemented in the M&M countries:These cooperatives, prevailing in most WANA countries, co-opted the roles traditionally played by pastoral communities and institutions. They proved to be unpopular due to the separation between traditional rules and production systems, and rules governing the functioning of cooperatives and their resources.Failure of the previous type of cooperatives encouraged some herders to organize their own cooperatives and request land from the government to improve and manage.The main benefit, compared to state-driven cooperatives, is that they offer better security of tenure to their members, in addition to new services such as health and feed provision.However, more exclusive decision-making authority on access and use of cooperative pastures is needed to prevent government institutions from issuing grazing licenses to non-cooperative members.These have been created to enhance the managerial role of local institutions and maintain customary access and use rules. They provide security of tenure over pastureland and mere local control over resource access and use.This strategy, mainly used in Tunisia, involves placing non-privatized tribal pastureland under the control of the Forest Services to improve the range and manage its utilization. All community members pay a fee to access the range or cut fodder until the Forest Services recoups its investments, at which point the community reacquires control over the pasture.Most M&M governments view pastoral resources as state property, while the pastoral communities consider them as their territory. Poorly defined tenure rights often lead to conflicts and equity issues. Those who advocate devolution policies suggest that the success of range management depends on the extent to which pastoral communities are granted full control over access and use of the resources and on the assurance of benefiting from improvements. This case study analyzes the different range management options implemented by the Tunisian government in the central and southern regions to improve the availability of feed resources and enhance the welfare of pastoral households. Most new institutional and tenure policies were introduced in the central region where the privatization process was more advanced. Four types of management regimes were identified in Tunisia:1. The tribal system prevails in the majority of the ranges located in southern Tunisia. These ranges have not been privatized but the weakening managerial role of tribal institutions has led to crop encroachment and appropriation of the best pastoral areas by agriculturalists. State development intervention in some of these areas includes development of roads and watering points.2. The private system emerged from the privatization of tribal rangelands. The Pasture and Livestock Office (OEP), a parastatal agency, is in charge of range improvement activities in these areas, such as promoting the development of cactus plantations. The main problem associated with this management option is land fragmentation.3. The government sponsored cooperative system is relatively new, and involves organizing pastoral communities and devolving range management to local communities. The experience of the World Food Program (WFP) cooperatives has not been an overall success due to the limited role played by cooperative members.4. The co-management system is in place on the residual tribal pastures that have not been privatized in central Tunisia. Under this system the community cedes control of overgrazed pastures to the Forest Services for pasture improvement. In exchange for the improvement the Forest Services charges grazing fees. The community may reclaim its rights once improvement costs have been fully recovered. The main problems facing this option are associated with strong state intervention and weak local participation.Rapid rural appraisals, range productivity data collection and in-depth household surveys were conducted under these different management options. Econometric analysis was conducted to evaluate the effects of range management options on total household feed expenditures. The preliminary results show that, compared with the tribal system, the co-managed and privately managed reserves reduce household feed expenditures by 33% and 9% respectively, while cooperative reserves increase household feed expenditures by 62%. The results reflect the changes occurring in rural Tunisia.The performance of the co-managed reserves depends on the management quality of the Forest Services and the ability of community members to pay for grazing or cutting forage. Co-management could be the best option for providing additional feed resources while also improving the resource base. However, these are preliminary results and generalizations may be misleading since problems facing rangeland management in Tunisia are diverse and complex.All the sites studies, regardless of the type of management option, are facing similar problems such as animal and human population pressures, scarcity of grazing resources, and weak participation of communities in the management of their common resources. In addition, there are regional differences between central and southernTunisia due to the extent to which privatization policies have been implemented. In central Tunisia, where the privatization process is very advanced, major problems include unequal access to grazing resources, overexploitation, and projects that introduce inappropriate technologies. The main problems in southern Tunisia where tribal systems prevail, are poorly defined property rights and consequent land encroachments and resource degradation. This situation is also fueling many inter-and intra-community conflicts. Suggested policy options include the development of coherent range management policies in integrated development projects that would organize and empower communities in the management of range resources as well as provide services and infrastructure.Mustapha Guellouz Director General of the Pasture and Livestock Office in Tunisia President of the Council of Enterprises Social changes after Tunisian independence generated structures and human relations that are different from those that prevailed under the tribal system. The dislocation of land tenure regimes provoked by the phenomenon of privatization or appropriation of the collective tribal lands is the prominent feature of this period. In central Tunisia pastures are indeed relics of tribal lands that were not privatized, while in the south the land tenure system is not well-defined and litigations between right holders and users are frequent and persistent.This study shows that these changes led to the development of weak management institutions whose role remains unclear or incompletely defined. For example, the WFP cooperatives that were created to manage pastures in tribal lands that were not individualized and the management councils that have an implementation oversight role in the allocation of tribal collective lands lack an adequate legal framework.Will the present initiative to develop Agricultural Development Groups (GDA)for the management of pastures have positive effects? Will the GDA motivate the participation of the beneficiaries?Besides the need for appropriate technological packages for different pasture conditions, (e.g. Pasture and Livestock Office program) there is the need to develop professional structures that emanate from the beneficiaries on the basis of their tribal cultural heritage. The designation of a single coordination center for all the intervention programs that will link to these structures will assure a stronger engagement of the populations.The present report lends support to the process undertaken by the Tunisian Forest Services through pasture development plans and projects, and to the conclusions of the workshops on institutional and political aspects of range management held in Kairouan (June 28-29, 2000) and Tataouine (2000). With the help from the donor community, the Moroccan government launched ambitious programs for the improvement of major rangeland areas. These programs cover entire agro-ecological zones, are holistic in vision and try to address in a comprehensive way the problems regarding rangelands.The purpose of this study is to evaluate the impact of different institutional options introduced with the aim to enhance rangeland management. To capture the diversity of agro-ecologies and range management options, three zones were considered:the high plateau of the Eastern Atlas (or Oriental Atlas), where range cooperatives have been created according to tribal membership; the Middle Atlas, where traditional tribal rangeland management is reported to face severe difficulties; and the Central High Atlas, where the tribal management system continues to play an important role in the management of the community pastures.The study is based on quantitative and qualitative data from rapid rural appraisals (RRA) that were conducted along transects in each of the three regions to characterize production systems and range management options. The RRA was followed by an indepth household survey on 325 households. Econometric analysis was used to evaluate the effects of different range management options on total household feed expenditures.Except in the High Atlas, tribal management systems are playing a limited role in the management of their community pastures. In the Middle Atlas and the Oriental Atlas where cooperatives have been introduced, many people have an inadequate understanding of the functioning of cooperatives and there is a general tendency not to respect the rules governing the use of cooperative reserves.Preliminary results of the econometric analysis suggests that.compared to households that relied mainly on tribal non-improved (or unmanaged) pastures, households with access to tribal cooperatives face 3.4% lower feed expenditures, and households with access and involvement in actual management tribal pastures (agdals of the High Atlas) face 10% lower feed expenditures. In comparison pastures under government management (Forest Services) demand 11% higher feed expenditure per household. These results suggest that in the Oriental Atlas, where the tribal management is eroding due to the increasing sedentarization of pastoral households, the cooperative reserve could be an important option. However, in the Central High Atlas, where traditional management systems continue to effectively manage access and use of the pastures, it is important to keep these systems in place. This does not mean that the Moroccan government should not intervene in the Central High Atlas, but that development action should be taken to improve the general performance of the system without disrupting existing management institutions. Moreover, range improvement should be based on knowledge of pastoral societies, their customs, their institutional arrangements, etc. Technological solutions alone will not solve the problems of developing pastoral zones. Often pastoral institutions need to be strengthened to enable them to implement improved range management practices. For this to be effective pastoral communities need to be involved in the elaboration and implementation of development projects.Morocco has adopted principles of integration and participation in the preparation of agricultural development projects. Law No. 33-94 on the improvement of rainfed perimeters was promulgated in 1995 and its application is based on four principles:1. sector integration in development efforts; Although differences in productivity between government reserves and herder-driven cooperatives are marginal, given the high transaction costs associated with fencing and guarding government reserves, herder-driven cooperatives are likely to be more efficient in managing rangeland reserves.Baker Qudah Ministry of Agriculture, Jordan.The total surface area of the Kingdom of the Jordanian Hashimate is 90,000 square km. Rangelands cover about 90% of the total area and receive less than 200 mm of annual rainfall. These rangelands are confronted with overgrazing, recurrent droughts and desertification leading to reductions in forage production.Various studies and experiences of rangeland protection show that rangelands are resilient if correctly managed. Degradation in these areas is mainly due to unclear property and grazing rights. The study presented concludes that the cooperative system should be applied to improved rangelands and foster environmental sustainability.However, the study focuses on areas where property rights are well defined and not disputed. The longer Bedouin households stay outside their own sites, the more likely they have to resort to supplemental feeding. Reflecting trends in the countryside, Bedouin livestock production systems are becoming increasingly dependent on purchased feed supplies. As lack of adequate water and marketing infrastructures further weakens Bedouin production systems, Bedouin households are developing strategies such as selling part of their flock to purchase irrigated fields, exiting the livestock industry by investing in the transportation business, or migrating to the Arabian Peninsula. Some of these strategies may work in the short run, but may not be sustainable in the long term.The improvement of Bedouin livelihood strategies will depend on the extent to which adequate policy, institutional, and technical options are identified and used with full participation of the communities.Mahmud Nuhayyer Director General of the Rangeland Project. Ministry of Agriculture and Agrarian Reform, SyriaThe Syrian steppe represents 55% of the total Syrian surface area or 10.2 million hectares, with average annual rainfalls of less than 200 mm. In the past the dominant One of the problems facing the cooperatives is that their boundaries do not coincide with tribal boundaries. In order to proceed with the steppe development, boundaries must be the same. unique, some impacts are consistent among all three study sites:• The team rarely observes formal rules on stock densities, land allocation or mobility. Nonetheless, effective collective action in other natural resource management activities (i.e. waterpoint maintenance, soil erosion control measures, seasonal access restrictions, restrictions on settlement locations, etc.) has a significant impact on pasture use and land management in all three regions.• High rainfall variability is often associated with either lower stock densities or greater mobility, or both, but is never associated with higher stock densities as would be the case if livestock were predominantly used as a source of savings or as a drought survival strategy-often held assumptions by researchers and policymakers alike.• Heterogeneity in terms of social and/or wealth differentiation has a negative impact on the capacity of a community to cooperate and often leads to greater stock densities and more land allocated to private uses.• Greater profitability of livestock products generally has a positive impact on capacity to cooperate.• Greater number of members and the degree to which community resources are shared with other communities appear to make cooperation more difficult, and often leads to higher stock densities, though estimated impacts are less pronounced than the effect of other variables. The Borana people are the predominant ethnic group on the Borana Plateau in southern Ethiopia. Though traditionally transhumant pastoralists, they have recently increased their reliance on crops. Rainfall in the region is bimodal and averages between 353 mm to 873 mm; variability is high, with coefficients of variation ranging from .21 to .68. Anecdotal evidence implies that the vulnerability of pastoralist households to drought is increasing; stock levels increase dramatically during good rainfall years but plummet when rainfall is poor, indicating that the drought cycle is becoming more pronounced. In recent years, there has also been a dramatic increase in land allocated to crops and land allocated to pastures that are either privatized or accessible to only a small sub-group of people. Nonetheless, the Borana are still highly dependent on access to common grazing lands, which provide the predominant source of forage and, importantly, which also provide a mechanism to reduce risk associated with poor rainfall in one area by allowing for mobility. Because many of the land resources are used and managed in common, it is hypothesized that one of the key determinants of the productivity and sustainability of the systems is the ability of community members to cooperate over the use and maintenance of these resources.In this paper, the authors develop indicators of cooperation and examine factors affecting these indicators. They then use these indicators to determine the impact of cooperation on stock densities and land allocation patterns. Results indicate that cooperation is positively related to factors that increase the profitability of livestock, but negatively related to the total number of households, the use of community pastures by non-community members, and heterogeneity of wealth within the community.Furthermore, stock densities are negatively related to the index of cooperation as expected. Stock densities are also lower in areas with more highly variable rainfall, indicating that high variability reduces the number of livestock held, a result which is inconsistent with the hypothesis that households build greater stockholdings in highly variable environments in order to survive a drought with more animals or as a source of savings. Finally, results from the land allocation estimations give evidence to support the notion that more land is privatized for pasture where levels of cooperation are lower.Given the importance of mobility and the poor suitability of most land for cropping, measures to offset increasing stock densities and privatization of land should focus on improving the capacity of communities to cooperate and mitigate the impact of heterogeneity on that capacity, and on improving market access to improve cooperation and increase incentives to sell stock in good as well as poor rainfall years. Results also highlight the need to search for alternative policy mechanisms that mitigate the impact of drought that do not simultaneously increase incentives to augment herd levels in nondrought years.Edmealem Shitaye Head of Pastoral Extension Team, Agricultural Extension Department Ministry of Agriculture of EthiopiaThe authors of the paper are to be congratulated for the excellent work on the Borana Plateau. Nonetheless, the Borana Plateau is only one area of pastoralist activity in Ethiopia, and thus the study might not be representative for the status of rangeland management throughout the country.Also the economic importance of environmental variability on crop and forage production is not highlighted enough in the study. In particular, rainfall variability has a more significant negative effect on crop production compared to livestock production, and more detailed attention should be given to changes in rainfall patterns and duration of drought conditions. One important aspect that is not mentioned in the paper is the role of rainfall variability on livestock market performance. Similarly the linkages between grain availability and livestock production for the local market also deserves more attention.Care needs to be taken when considering the hypothesis that stocking densities do not have significant effect on rangelands, a hypothesis that might be too general. This might not always be true without consideration of particular time frames and species of In Niger, key climatic characteristics include the high relative rainfall variability and recently increased frequency of droughts. Livestock mobility is often seen as one of the most valuable risk mitigation strategies, as it enables herders to improve both mean output as well as decrease output fluctuations associated with both spatial and temporal variability in rainfall. Broadly speaking, land tenure in this region consists of a mix of quasi-private and essentially common property, allowing for both fixed agricultural production and mobile livestock production.Unlike the case for Ethiopia and Burkina Faso, it was not able to develop a proxy index of cooperation based on observed features of existing community structures, rules, regulations and activities in other areas of natural resource management. Thus, an index was developed based directly on exogenous variables thought to help or hinder collective action: degree of ethnic heterogeneity and distribution of farm sizes, number of members, degree of use of community land by neighbors and transhumants in the dry and in the rainy season, and the extent of migration of household heads for wage work. The factor analysis resulted in two primary factors, both of which were hypothesized to hinder cooperation. The first captured heterogeneity within the community and total households, while the second factor captured pressure on resources by neighbors -but not transhumants -as well as total number of members. These factors were then used as explanatory variables in an econometric model of mobility, stock densities, and land allocation.There are three main conclusions to be drawn from the analysis. The first is that even when there are no formal \"rules\" or regulations regarding stocking rates on common pastures, factors associated with capacity to cooperate at the community level do impact decisions on stocking rates and on mobility. In communities with relatively high scores on the constructed non-cooperation indices, mobility is reduced and overall stock densities are much higher. Though difficult to address directly through policy measures, the results reinforce the notion that devolution of management of resources must consider the capacity of communities to cooperate. The results do support the notion that measures will have to be developed to offset the negative impacts of heterogeneity-in terms of wealth and ethnicity-on the ability of the community to cooperate. External pressures on the resource and the number of households, which are more highly correlated with the second index of non-cooperation, also affect mobility and stock density, but the estimated effects are smaller than those associated with the first index.Second, relative prices favoring livestock actually increase the share of land allocated to crops. This indicates that in these communities, the value of crops (i.e.through use of residues as animal feed) is quite high in livestock activities. It would be ideal to be able to combine this information with studies identifying factors associated with off-take rates; results from this study alone, however, indicate that increasing relative prices for livestock will likely not have a large effect on stock densities per se, but the response is likely to be increasingly intensified animal production and stronger crop-livestock linkages.Finally, the impact of rainfall variability is quite pronounced for stock densities, but has no impact either on mobility or on percent of land allocated to crops. A priori, it would seem reasonable that mobility would be related to rainfall variability. The discrepancy may in part be due to the fact that mobility, by definition, is a flexible response to actual rainfall, whereas stock densities and the percent of land allocated to crops are less flexible and thus depend more on longer-term indicators of variability and mean rainfall realizations. Thus, the measure of long-term mean rainfall and variability used in this study might not adequately capture incentives for mobility in the particular year studied. However, consistent with results from the study undertaken in Ethiopia, there is a strong negative impact on stock densities particularly in communities where rainfall variability is relatively high. This result is important, because many drought mitigation and preparedness measures are predicated on the belief that programs that offset the impact of rainfall variability on animal productivity will lead to lower stock densities. The results do not support this belief; rather, it is likely that stock densities would increase in response to measures directly aimed at reducing the impact of poor rainfall on animal productivity. Unfortunately, the policy conclusion is thus that measures to mitigate the impacts of drought must simultaneously consider measures to increase off-take or otherwise reduce stock densities.The new context of globalization as well as the recent agreements that Niger has signed with the UEMOA (West African Economic and Monetary Union) and the CEDEAO (Economic Community of the West African States) countries has led our government to give priority to the development of crop and livestock production. This emphasizes livestock production in particular because Niger has a comparative advantage in this sector with respect to other countries.The countryside offers the following favorable conditions for livestock production:• 62 million hectares of pasture areas, especially in pastoral zones.• A diversity of forage species with high nutritive value, particularly in the lower rainfall regions.Nonetheless, evidence on the status of rangelands shows a worrying process of degradation, which is manifest through:• Decreased plant density and proliferation of denuded soils.• The disappearance of species desirable to animals, notably perennial species.• The invasion of grasslands by undesirable species such as sida cordifolia, pergularia tomentosa, boerhavia spp, zornia glochidiata.The herders and some pastoralists attribute this degradation to climatic changes.Though rainfall has indeed played an important role in the degradation process, the comparison of areas without pastures in the Sudanian zone with areas in the pastoral zone show that livestock, through the forage selection process, has also played an important role in this degradation process, perhaps a more important role than decreased rainfall. This situation is surely the consequence of the traditional management system in Niger in which livestock has largely free access to most areas. While this system is efficient for adding value from the sparse but nutritious forage species in the Sahel, overgrazing leads to the excessive selection of the more nutritive forage species.However due to the absence of other viable approaches to optimal rangeland management, the government has not been able to introduce modifications. This calls for more research regarding appropriate alternative approaches, research in line with the work of Dr. McCarthy and her colleagues.Research work has rarely been carried out on the management of pastoral resources, a domain in which the government needs a great deal of information in order to set up appropriate programs to sustainably improve rangeland productivity. Lack of information has been one reason why the Niger government has not yet effectively responded to instances of land privatization now occurring in the pastoral zones, which unfortunately continue. The option of privatization merits reflection where several factors might favor this process:• Demographic pressures that lead to the exploitation of arable land for agricultural activities in areas originally used for animal production. Land privatization might actually slow down this process of converting pastures to cropland.• The successful role played by State ranches as reserves during the drought periods, which is in part due to herders' respecting limited access because it is considered as \"private property\" of the state.• Transhumance constraints in parts of the Sahelian sub-region call for systems that can maintain livestock herds with limited mobility.The last option deserves deeper analysis in order to make it effective in terms of sustainable production and resource conditions. While waiting for a broader analysis, the government decided to create the Secrétariat du Code Rural de Commission Foncière in order to look more closely at property rights issues and the resolution of land conflicts. This body, though just beginning, can constitute an important mechanism to manage lands. The question remains as to whether and what mechanisms can be used efficiently to manage land in pastoral zones.By discussing the experiences of each country at this workshop we can learn about different options that might be used as 'pilot' programs in our country. The Sahelian zone of Burkina Faso has traditionally been characterized as being overwhelmingly geared towards livestock production heavily reliant on mobility.However, with the process of sedentarization of the population, the region is more accurately depicted as agro-pastoral, though livestock products still comprise the largest share in combined value of cash income and home consumption. Nevertheless, most pastoral land is still \"owned\" in common, which means that the success of provision and management of most natural resources relies on cooperation between villagers.Villagers' decisions depend also on the presence of external actors (state, projects and NGO's) in this region such as the PSB/GTZ project, co-managed by the government and the German Technical Cooperation Agency. Traditionally oriented toward supporting local desertification control and natural resource management (NRM) through the 'gestion des terroirs' approach, the project changed focus in the mid-90s by putting emphasis on the institutional side of NRM.The purpose of the paper is twofold: first, to determine the external and internal factors influencing the way NRM institutions work and second, to identify how these institutions' performance affects the level of resource use, observed in this study through stock densities and land allocation. For this purpose, the paper relies on a survey conducted in 48 villages of the provinces of Oudalan and Seno, stratified on the basis of their entry date in the PSB/GTZ program, to include sites where the project had employed only a technical approach, sites with institutional interventions, and control sites in which the project had not yet begun to work.A general authority -traditional chief and/or official delegate (RAV) -is present in every village, though several other institutions are almost always present, including: general village associations (men, women, mixed), producers associations (farmers, or herders), and water and tree management associations. The main activities operated by those institutions are water source maintenance and management, erosion control, reforestation, and agro-pastoral zoning. Rules that govern NRM in these villages concern the pastoral as well as the agricultural zone (e.g. calendar for animals to enter or be removed from cultivated fields), restrictions or prohibitions on harvesting hay and/or forest products, and water use rules and regulations (e.g. health and hygiene norms, access conditions).Since the analysis consists of comparing institutions at the community level, the authors built several indicators by aggregating the institutional information at the village level. These indicators were then classified according to structure (% rules monitored and enforced by the chief only, % institutions that work at the supra-community level), conduct (number of institutions, rules, activities in the village), and performance (participation in meeting, in working activities, presence of conflicts, rule violations). A correlation matrix was computed between village characteristics and NRM institutional indicators.The following patterns could be identified through the correlation coefficients: i) the presence of projects and education are positively and significantly correlated with Investigating the links between cooperation and resource use, we find that stock densities and percentage of land allocated to crops are explained by non-cooperation (i.e. higher degrees of disagreement lead to higher stock densities and higher proportion of land dedicated to crops), population density (i.e. the greater the pressure, the greater the State, which proceeds with successive reforms without clarifying the previous or current policies on land tenure, etc. Also, NGOs and projects often do not build on the existing institutional structures, but rather create new structures under their control in order to realize their development objectives. This paper reviews the history and policy outcomes of Uganda's rangelands which are concentrated in the \"cattle corridor\" of Uganda. The main use of rangelands in Uganda is grazing by wild and domestic animals, which provides the cheapest source of nutrients for ruminants in Uganda. Rangelands support about 90% of the national cattle population, mainly kept by pastoral and agro-pastoral communities. About 85% of the total marketed milk and beef in the country is produced from indigenous cattle that thrive on natural rangeland pasture. Yet, most of the farmers remain poor and are increasingly experiencing food insecurity.From an environmental point of view, rangelands in Uganda constitute very fragile ecosystems, subject to desertification due to drought, overgrazing, deforestation, poor farming practices and soil erosion. Poverty coupled with a rapidly increasing population exacerbates these factors.Since colonial times policies have fallen short of recognizing pastoralism (livestock farming) as an economic activity. The tendency has been to introduce policies geared at the sedenterization of pastoralists. Apart from the 'crop production bias' favoring agriculture over pastoralism, two recent policies have further marginalized pastoralists. One is the development of a very strong environment-oriented pressure group which has caused the displacement of the encroachers on gazetted land, such as forest reserves. The existence of these closed areas within the cattle corridor has created management problems for pastoralists, as well as for forestry and wildlife authorities.Secondly, the development of tourism, although a very important source of foreign exchange, has further marginalized the interests and the rights of pastoralists whose land has been turned into national parks, wildlife reserves or wildlife sanctuaries.There are three main land tenure systems in Uganda:Customary tenure is the most prevalent tenure system throughout the pastoral and non-pastoral areas of Uganda. It is the most egalitarian tenure system but generally does not foster investments for maintenance of the resource.Private property has resulted from a high level of individualization of the communal pastoral land throughout the entire corridor, but tends to lead to a reduction in the available grazing land.State property includes national forest reserves, national parks, game reserves, wild life sanctuaries and community wild life areas. Generally, traditional rights of existing populations have been neglected by policy decisions regarding these lands.The shift from customary to private and state property has triggered a number of regional conflicts, as well as disrupted traditional management rules. Local communities have often lost control over rangeland resources.Current policies continue to concentrate on sedentarization of pastoralists, rehabilitation of the animal sector (e.g. veterinary support), provision of water, attention to gender issues, and agricultural modernization. New areas that need policy intervention include the problem of livestock overstocking on rangelands, inadequate water supply, insufficient market facilities, disease and pest control, as well as low investment in extension, infrastructure, and research on rangelands. However, the most important factor is probably the lack of institutional support to local communities.Comprehensive national policies that recognize the multiple use characteristics of rangeland resources and reduce coordination problems among agencies are needed.Policy approaches need to involve communities in the planning and implementation of programs. Decentralization of power should favor marginalized pastoral people, though this may be challenging given their isolation and difficulty of integrating them into the policy arena. Major research efforts are needed to increase productivity of rangelands and identify viable income generating activities for pastoralists.There is a need for a new approach to dryland and pastoral development in Uganda. Four important areas for policy indicated in the paper should be highlighted.The first is a role of pro-active government policy toward pastoral property rights.With respect to this issue in Uganda there is the need for an over-all land use allocation strategy toward agro-pastoral and other uses (conservation etc.), and the possibility of using conservation areas as fall-back resources without threatening conservation objectives. The government must recognize the existing trade-off between conservation and production. With regard to the implementation of the land law (1998 Land Act), it is important to investigate the applicability of this law that seeks to solidify private claims to agricultural lands. The establishment of government ranches is generally not an appropriate policy approach, except possibly as multi-use conservation, buffer zone or transition from other ownership types.The second area regards the guidelines for macro, environmental and trade policy and their effects on pastoral areas. Food security and international relations issues, as well as decentralization strategies, all affect pastoralist livelihood strategies. With respect to decentralization, apart from obvious advantages it can also carry risks for pastoralists when there is a lack of contact with central planning authorities, as is evident from the experience of the terroir approach in Niger.The third area relates to the possible existence of an \"optimal fuzziness\" in land use planning and property rights. Major issues in this respect are the zoning within national parks and forests, the need to set aside areas for rest and regeneration, the need for agricultural-livestock integration and intensification of production strategies and the need for mobility during drought times.The last area relates to collective action strategies. There is a strong inter-relation between collective action and property rights. Collective action should be a higher priority for policy than property rights per se. Especially in case of high variability of rainfall, collective action should not be seen only as bonding within a group. It is important to consider collective action across groups and collective action to integrate into markets. It is thus important to look at broader political agendas and international relations. It is also important to keep in mind that the final goal of research on pastoral areas is to arrive at appropriate policy implementation. The study was conducted at landscape and regional levels in southern Ethiopia, where the Booran pastoral production system comprised the Golbo (lowlands), the Dirre (Plateau) and the Liiban production systems (hereafter also referred to as regions). By involving traditional range scouts in evaluating landscape and regional level environmental changes, the study challenges the notion that IEK is mythical and could not meet scientific rigor. The use of common soil and vegetation indices allows comparisons of land degradation assessments between IEK of the pastoralists and ecological techniques.Evaluation by traditional range scouts (TRSC) and range ecologists (RE) on changes in range conditions and trends showed high correlations. IEK was effectively used to determine landscape suitability and potential grazing capacity of individual landscapes and at regional levels. The study shows different perceptions in interpreting grazing suitability and potential grazing capacity. While grazing capacity is an inherent property of individual landscapes, management decisions have impact on grazing suitability. Both TRSC and RE made comparable predictions on threats to range conditions and trends, but interpreted landscape stability differently. We suggest that integrating IEK in the ecological methods would help identify important perceptions of the pastoralists on effects of land use on local landscapes. Moreover, the value of IEK should also be considered when monitoring landscape level changes as well as when assessing degradation of the grazing lands. We hope the information in this paper will motivate policy makers to incorporate IEK of the pastoralists into decisions on landscape level range rehabilitation.Presenter Tahar Telahigue International Fund for Agricultural Development (IFAD), Rome co-authored with Abdelhamid Abdouli International Fund for Agricultural Development (IFAD), Rome This paper presents IFAD's project approach to rangeland management and development (findings and recommendations are based on IFAD projects in Morocco, Jordan, and Syria) which focuses on empowering local communities to become the main players in the management process. At least three inter-related tools have been used by IFAD for the empowerment of herder communities:1. Involvement of the herders through introduction of participatory approaches.Beneficiary participation is key to the success of conservation-oriented projects.NGOs have an important role to play in testing, identifying and experimenting with new alternatives and technologies that can contribute to sustainable rangelands management by the herders themselves.2. Policy dialogue with governments for the promotion of appropriate land rights: Granting long-term grazing rights to local communities is important for conservation management, yet a very complex process. Acknowledgement of local land users' rights and the integration of customary land-tenure arrangements within new administrative structures is a pre-requisite for any long-term sustainable investment activity for the rehabilitation and management of the rangelands.The benefits from rangelands conservation activities are of a long-term nature, while poverty compels herders to engage in conservation activities that produce substantial, quick returns at low cost. Therefore conservation projects need to contemplate compensation of foregone income at least during the initial years.Difficulties arise if new approaches are adopted under pressure from external donors and without complete commitment of the government. Implementation is also more difficult where there is increasing stratification and diversification of herders' income, because of conflict of interests. In particular these questions were addressed:1. What are the key issues (problems, questions) regarding access, use, ownership claims, and management of resources?2. What are the strengths and limits regarding poverty alleviation?3. What are the strengths and limits regarding environment sustainability? SUMMARY OF SESSION 1:One of the key issues regarding state management regards the capacity of the state to effectively control access and use of rangeland resources. It was felt that management capacity is often quite limited due to lack of local knowledge necessary for good management. The state may also face higher costs in enforcement than would more local level authorities, and may instead be forced to rely on such costly measures as fencing and paying for guards. There would also need to be mechanisms for the identification of violators and enforcement of fines, and again, such mechanisms may be much more costly for the state. Also, poor management by the state may make previously \"common-pool\" resources open access with consequent negative impacts on productivity and an increase in conflicts. In fact, inappropriate or inadequate management capacity by the state may lead to use rates anywhere along the spectrum of far too little to far too much.State ownership may also have an adverse effect on the rights of the poor. This may be particularly true where the state's objectives are mainly conservation and/or to promote productivity of the relatively wealthy livestock owners. Again, because of lack of local knowledge, the state may inadvertently deny access to those who had previously had at least some degree of access to the resource. Pure conservation goals may be more likely to be achieved by the state, but at a very high cost if local community members are not involved with setting and attaining those goals, or if members are not compensated in any form when the government completely restricts access. Links between users and government have generally been very poor, many user groups have been alienated by various state agencies. Furthermore, where there are differences between those who consider themselves predominantly crop farmers versus pastoralists, the state has often sided with the crop farmers. Lack of criteria -or lack of transparency of criteria -by which the state allocates rights, promotes uncertainty and often conflict among users. This is also true in land use planning at the national level when decisions are made by centralized state agencies regarding land classification into different systems (national park, national reserves, national forests, etc.)Many participants felt that there was a role for the state to play in providing fallback reserves -though not all agreed that state was the best level to manage reserves.Also, the state may have a role to play in undertaking specific large-scale investments, and in gathering and disseminating the best available technical information. In most instances, there is a hierarchy of access rights with different individuals or groups negotiating access in response to such factors as very poor rainfall or loss of wage-earning income. Many systems are flexible and capable of responding to criseseither experienced by just one individual or by the whole communities. Thus, the riskspreading role of common rangelands is considered very important, particularly for poor and marginalized groups. This brings one to the issue of devolution policies, and how they might retain a hierarchical structure with rights and responsibilities matched to the appropriate levels of governance to ensure flexibility and equity. It was suggested that consideration of policy instruments such as \"long-term\" leases (99 years) would be best to promote long-term interests by community members, but at the same time retain certain roles for the state.However, problems of identifying which local institutions best represent the interests of all community members was also mentioned by many participants; equitable distribution of access to and control over community-based resources may be compromised when powerful local elite assert their power, especially if they co-opt community-based institutions that are given legitimacy by the state. Furthermore, equity in access does not necessarily mean equity in use, since the difference between large and small scale herders may be very large -and large scale herders will benefit disproportionately. Also, boundary conflicts may often arise. The flexibility of the system, which is beneficial for risk management may inadvertently lead to more conflicts. Conflicts of interests may also weaken management under common property, but it was felt that many of these conflicts could be managed by the different groups, although legitimate conflict management mechanisms need to be developed.It was noted that there are many other resources on pastureland, and community members must also find mechanisms to allocate access to and perhaps restrictions over such resources as wood for fuel, hay, soil for cropping, etc.Finally, participants also raised the issue of the capacity of local level institutions to withstand changes in the external environment, especially changes that promote overuse by certain groups or that increase incentives to privately appropriate land. In fact, the very flexibility that is seen as very valuable for responding to temporary crises and to offset variability in production and income, was also mentioned as being a factor that may make it easier for certain groups or individuals to privatize and/or mis-use the resource. Also, with very large changes, local level institutions may not be able to adapt conflict resolution mechanisms to adequately handle pressures for land use change.In many cases, privatization of common grazing lands leads to unequal distribution of grazing resources, and there is the risk of excessive fragmentation.Privatization also makes mobility more difficult, if not impossible, increasing the riskiness of livestock production. It also may lead to inappropriate practices that are undertaken simply to stake claims on rangeland -for example, inappropriate planting of poorly adapted but cheap tree species or cropping on marginal and fragile areas.It was thought that management of private property, however, is likely to be \"cheapest\" -allowing for quick decision making. Private property may also facilitate innovations and adaptations, which may have spillover benefits for local populations.It is important to understand which factors accelerate change towards privatization, in order to adequately manage land-based resources. Also, under all regimes, there needs to be better integration between technicians, land-use planners in government ministries, and the users themselves.During the second session, participants built on the issues discussed in the first session to identify the conditions under which each property rights institution is effective in reaching goals related to the efficient, equitable and sustainable management of rangeland resources. Secondly, acknowledging that optimal management likely requires participation of different actors at different levels (individuals, communities, local, regional, national -and sometimes international -governments), the groups discussed the roles and tasks of different actors in achieving effective management.Participants thought that at least partial state control might be the best option in the following contexts:• Environments with very fragile ecosystems or areas with important bio-diversity characteristics. Even here, however, partnerships with local users are considered necessary to achieve goals of rehabilitation and sustainability. Federations of user groups may play a key role in interacting with governments.• Managing large-scale water catchment areas, where communication, coordination and cooperation would otherwise have to be undertaken by many distant and disparate communities.• Where large-scale investments and basic infrastructure must be made to rehabilitate rangelands or protect environmental amenities; or in other investments where returns will not be realized until some time in the future.As before, participants emphasized the fact that the role of compensation to users who will now be denied access or restricted in activities must be addressed by the state in a transparent way, particularly when the state's goals are to achieve benefits at the supracommunity level. The state should also consider appropriate policies to aid newly restricted herders to adopt new practices.Some participants also thought the state had a role in managing reserve areas designated to be used only during drought years; and again, where required large-scale investments would be very risky for communities to undertake alone. Managing In all cases, the state's activities should be limited to those activities where there is ease of monitoring and enforcement, i.e. where information is readily accessible. There is no need to engage in micro-management, since that requires local knowledge and constantly updating information on changing local conditions.To summarize, it was thought that the state's direct involvement in the management of rangeland resources is best reserved for enforcing temporary use;undertaking large and risky investments; protecting and rehabilitating heavily degraded and/or fragile ecosystems; and managing situations of heavy conflict. Clearly, there is a key role for the state to play in land use policy, regulation, and legislation; establishment of a guiding framework of rules and regulations, to give legitimacy to local institutions where appropriate, and devise adequate compensatory mechanisms where access will be denied or restricted. may be made conditional on responsibility in use and management, but these conditions must be fairly negotiated and clearly understood by all participants.• Identify appropriate local institutions, and give clear criteria for this identification, so that the institution is then seen as both credible and legitimate. Revitalize traditional institutions where appropriate, but only where these institutions are also considered appropriate according to the criteria. Legislation should also be adopted to legally recognize common property rights. Legitimacy is most easily established where concerns of all users are voiced and listened to. Local empowerment is part of the strategy for successful devolution.• Take a lead role in facilitating cooperation when this is required across many different communities (i.e. for mobility or for management of water catchment areas), and make sure that weaker groups are not marginalized (e.g. poor pastoralists and women).• Design policies to promote fair and credible conflict management mechanisms by empowering local communities.• Carefully consider large-scale, strategic investments that generate large public goods, and undertake those that yield benefits to users, particularly both now and in the long run. Some participants also felt that establishing drought reserves is still an important policy to be undertaken and enforced by the state.• Take an active role in disseminating information related to rangeland management and livestock production -and integrate indigenous knowledge with knowledge from other sources.• Develop a distinct set of contingency plans and social safety nets in the event of serious droughts and/or other catastrophes.• Ensure state to state coordination for the management of the transboundary pastures.• Foster diversification and income generation sources for pastoral zones in order to reduce the vulnerability of poor pastoralist communities.In the arid and semi-arid areas of North Africa, Sub-Saharan Africa and West Asia, rangeland management issues remain of critical importance in ensuring both equitable and sustainable development in a highly variable environment. Participants emphasized the risk-reducing role of mobility and access to a wide range of pastoral resources. Nonetheless, increasing population pressures and policies that increased uncertainty over access and use rights over rangelands resources and favored sedenterization of pastoral populations and crop production. Consequently, many regions have experienced degradation of rangeland resources and increased vulnerability of pastoralists' livelihoods. Conference participants identified the following critically important issues:• Maintaining mobility while simultaneously ensuring that community investments in the management of common pastures accrue primarily to community members.The role of drought reserve areas, restrictions on access, and which institutions should make and enforce decisions were also considered.• Identifying the appropriate institutional mix: one that balances the flexibility of more informal systems of access, use and management rights with institutional arrangements that offset the high level of conflicts, greater opportunities for private appropriation, greater ease with which households justify circumventing restrictions, and permanent encroachment that often accompany such flexible systems.• Designing legal frameworks to resolve uncertain property rights, which are seen as the root cause of degraded and unproductive rangelands.The papers presented and discussed at the conference highlight four important points related to the issues presented above:1. Communities can and do cooperate over the management of resources, either through cooperative societies or more informal traditional mechanisms. Building on local knowledge and traditional structures to create more formal structures, i.e. herder cooperatives, is likely to lead to better management than relying on informal mechanisms only, particularly in regions subject to large changes in population pressures, weakened traditional institutions, market access, etc. However, certain communities and cooperatives are more successful than others.2. The number of members, heterogeneity in wealth and ethnicity, and the extent to which community resources are shared with others all negatively affect the capacity to cooperate, whereas greater profitability of livestock activities often improves cooperative capacity.3. Both long and short distance mobility is important to increase livestock productivity in all years, as well as to reduce vulnerability under poor rainfall conditions. Mobility is also a function of the capacity of the community to cooperate; lower cooperation leads to lower mobility.4. Areas with relatively high rainfall variability have lower stock densities because of the added risk to production. Policies and programs that successfully mitigate impacts of drought may in fact induce dramatic and unsustainable increases in stock levels. Drought mitigation strategies must be developed that reduce vulnerability of herders to drought but that do not lead to large increases in stockholdings.Policy recommendations reached by conference participants focus strongly on identifying appropriate roles, rights and responsibilities of government and local level institutions in rangelands management. Community participation is considered a necessary prerequisite for sustainable management, since local knowledge is required for technical aspects of management and local institutions have better information on which to base management decisions and enforcement mechanisms. However, the state still has a role to play, particularly in ensuring that local institutions represent interests of all community members (and not only the wealthy elite), and, in many cases, in helping to create legitimate conflict resolution mechanisms. Some participants remain skeptical of devolving responsibility of rangelands management to communities, but the empirical evidence does support the fact that communities can and do manage use of pastoral resources, though investments may remain lower than socially optimal. Thus, the state may have a role to play in undertaking larger scale investment projects whose benefits are realized only in the long term. Even here, however, evaluating the benefits and costs to such a project must be done in conjunction with the community, and arrangements for cost sharing may also be considered.One reason for the skepticism arises from the impact of heterogeneity of resource users -in terms of wealth levels, ethnicity, education, access to credit, and access to nonfarm income sources -which participants strongly felt hindered collective action in the management of natural resources. The negative impact of heterogeneity on collective action was also borne out by research results. Thus, there remains a knowledge gap in understanding which specific policy mechanisms may best alleviate the negative impact of heterogeneity of local and regional interests.  ","tokenCount":"10333","images":["-1615607129_1_1.png","-1615607129_1_2.png","-1615607129_1_3.png","-1615607129_1_4.png","-1615607129_1_5.png","-1615607129_1_6.png","-1615607129_1_7.png"],"tables":["-1615607129_1_1.json","-1615607129_2_1.json","-1615607129_3_1.json","-1615607129_4_1.json","-1615607129_5_1.json","-1615607129_6_1.json","-1615607129_7_1.json","-1615607129_8_1.json","-1615607129_9_1.json","-1615607129_10_1.json","-1615607129_11_1.json","-1615607129_12_1.json","-1615607129_13_1.json","-1615607129_14_1.json","-1615607129_15_1.json","-1615607129_16_1.json","-1615607129_17_1.json","-1615607129_18_1.json","-1615607129_19_1.json","-1615607129_20_1.json","-1615607129_21_1.json","-1615607129_22_1.json","-1615607129_23_1.json","-1615607129_24_1.json","-1615607129_25_1.json","-1615607129_26_1.json","-1615607129_27_1.json","-1615607129_28_1.json","-1615607129_29_1.json","-1615607129_30_1.json","-1615607129_31_1.json","-1615607129_32_1.json","-1615607129_33_1.json","-1615607129_34_1.json","-1615607129_35_1.json","-1615607129_36_1.json","-1615607129_37_1.json","-1615607129_38_1.json","-1615607129_39_1.json","-1615607129_40_1.json","-1615607129_41_1.json","-1615607129_42_1.json","-1615607129_43_1.json","-1615607129_44_1.json","-1615607129_45_1.json","-1615607129_46_1.json","-1615607129_47_1.json","-1615607129_48_1.json","-1615607129_49_1.json","-1615607129_50_1.json","-1615607129_51_1.json","-1615607129_52_1.json","-1615607129_53_1.json","-1615607129_54_1.json","-1615607129_55_1.json","-1615607129_56_1.json","-1615607129_57_1.json","-1615607129_58_1.json","-1615607129_59_1.json","-1615607129_60_1.json","-1615607129_61_1.json","-1615607129_62_1.json","-1615607129_63_1.json","-1615607129_64_1.json","-1615607129_65_1.json","-1615607129_66_1.json","-1615607129_67_1.json","-1615607129_68_1.json","-1615607129_69_1.json","-1615607129_70_1.json","-1615607129_71_1.json","-1615607129_72_1.json","-1615607129_73_1.json","-1615607129_74_1.json","-1615607129_75_1.json","-1615607129_76_1.json"]}

data/part_2/0101391738.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"c299aa1efc9f67a844ef7b7548e76b68","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/eb41ac33-0290-4d25-afba-8c2856c7657c/retrieve","description":"De nombreux efforts ont été récemment mis en oeuvre pour élucider les liens entre l'agriculture, la nutrition et le bien-être (1–6) et voir dans quelle mesure les programmes d’agriculture axés sur la nutrition peuvent améliorer ces résultats (7). Toutefois, il est nécessaire de mieux comprendre les liens associés au contexte entre l'agriculture, la nutrition et le bien-être afin d’élaborer des programmes et des politiques visant à surmonter les obstacles de ce contexte pour faciliter l'utilisation de l'agriculture à cet effet. Pour le comprendre da-vantage, nous avons mené une étude qualitative avec l'évaluation de l'impact du programme dénommé « Creating Homestead Agricul-ture for Nutrition and Gender Equity (CHANGE) » (Créer une Agriculture Familiale pour la promotion de la Nutrition, l’Équité et le Genre) de Helen Keller International’s (HKI) au Burkina Faso (8,9). Dans le cadre du volet qualitatif, nous avons cherché à mieux comprendre comment les membres de la communauté définissent le bien-être, comment ils classent leur communauté en termes de bien-être ac-tuellement (2015) et 5 ans auparavant (2010), comment cela était représenté dans l'agriculture, la santé et la nutrition et ce qu'il faudrait pour passer d'un niveau de bien-être à l'autre. L'objectif général est que ces résultats soient utilisés par les chargés de la mise en oeuvre des programmes et les décideurs politiques pour affiner les programmes et les politiques agricoles afin de mieux exploiter l'agriculture en améliorer les résultats en matière de nutrition et de bien-être, en particulier chez les femmes et les jeunes enfants au Burkina Faso.","id":"150303740"},"keywords":[],"sieverID":"9dc843f8-9f70-4aa0-af9c-8d79e62515ee","pagecount":"12","content":"heureux comme éléments clés du bien être alors que pour les hommes c'est l'accès aux services de santé qui était déterminant pour le bien-être.De nombreux efforts ont été récemment mis en oeuvre pour élucider les liens entre l'agriculture, la nutrition et le bien-être (1-6) et voir dans quelle mesure les programmes d'agriculture axés sur la nutrition peuvent améliorer ces résultats (7). Toutefois, il est nécessaire de mieux comprendre les liens associés au contexte entre l'agriculture, la nutrition et le bien-être afin d'élaborer des programmes et des politiques visant à surmonter les obstacles de ce contexte pour faciliter l'utilisation de l'agriculture à cet effet. Pour le comprendre davantage, nous avons mené une étude qualitative avec l'évaluation de l'impact du programme dénommé « Creating Homestead Agriculture for Nutrition and Gender Equity (CHANGE) » (Créer une Agriculture Familiale pour la promotion de la Nutrition, l'Équité et le Genre) de Helen Keller International's (HKI) au Burkina Faso (8,9). Dans le cadre du volet qualitatif, nous avons cherché à mieux comprendre comment les membres de la communauté définissent le bien-être, comment ils classent leur communauté en termes de bien-être actuellement (2015) et 5 ans auparavant (2010), comment cela était représenté dans l'agriculture, la santé et la nutrition et ce qu'il faudrait pour passer d'un niveau de bien-être à l'autre. L'objectif général est que ces résultats soient utilisés par les chargés de la mise en oeuvre des programmes et les décideurs politiques pour affiner les programmes et les politiques agricoles afin de mieux exploiter l'agriculture en améliorer les résultats en matière de nutrition et de bien-être, en particulier chez les femmes et les jeunes enfants au Burkina Faso.Pour atteindre les objectifs de l'étude, nous avons utilisé des données qualitatives recueillies dans le cadre de Discussions Centrées en Groupes (DCG) dans 28 des 60 communautés qui ont pris part à l'évaluation du programme CHANGE. Les DCG étaient menées séparément pour les femmes et les hommes. Toutefois, dans quatre communautés, nous n'avons pas pu terminer les DCG des hommes en raison d'un manque de participation. Ainsi, nous avions un total de 28 DCG de femmes et 24 DCG d'hommes. Les données ont été analysées en codant les réponses selon des thèmes communs et analysées séparément pour les hommes et les femmes, puis combinées. Les résultats combinés sont présentés. Les différences entre les réponses des hommes et des femmes sont mises en évidence.La plupart des répondants ont défini le bien-être comme le fait d'être en bonne santé (n=46/52 (89%)), de manger suffisamment ou de bien manger (n=46/52 (89%)), avoir de bonnes relations avec leur famille (n=40/52 (77%)) et avoir assez d'argent et de moyens pour subvenir aux besoins de leur famille (n=29/52 (56%)). Dans près d'un tiers des DCG, la paix et la sécurité sont également citées comme étant des composantes significatives du bien-être (n=16/52 (31%)). En outre, quelques DCG ont aussi qualifié l'accès aux services de santé, le bonheur, les bonnes récoltes et le fait d'avoir des enfants comme étant des aspects clés du bien-être. En général, les DCG ont révélé que les réponses des hommes et des femmes sont largement similaires, bien que les femmes soient plus susceptibles de mentionner l'importance de bonnes relations avec leur famille comparées aux hommes (respectivement, n=23/28 (82%) et n=17/24 (71%)) alors que, dans leurs discussions, les hommes citaient plutôt l'importance du fait de bien manger ou d'en disposer suffisamment) comme élément essentiel au bien-être en comparaison aux femmes (respectivement, n=23/24 (96%) et n=23/28 (82%)). Les Discussions ont aussi mis en exergue que les femmes étaient beaucoup plus disposées à souligner l'importance du fait de disposer de plus d'argent et d'êtreEn 2010, les participants aux DCG ont estimé que près de 60 % des personnes vivaient dans des conditions de bien-être minimales (niveau 1), avec environ 27% au niveau intermédiaire et 20% au niveau supérieur de bien être (niveau 3) (Figure 1) Cependant, cela a varié selon les villages participants. Par exemple, deux DCG de femmes ont affirmé que cinq ans auparavant (en 2010) tous les habitants de leur village figuraient au plus bas niveau de bien-être et d'autres ont évalué cette proportion à environ 75%. Plusieurs autres DCG ont décrit une répartition plus équilibrée des ménages entre les trois niveaux de bien-être. Enfin, dans un village, les répondantes du DCG des femmes interrogées ont estimé que plus de 40% des personnes vivaient dans le niveau de bien-être le plus élevé, tandis que le DCG des hommes du même village a estimé que cette proportion n'était que de 20 %.Alors que la grande majorité des DCG semblent ne pas avoir constaté d'évolution majeure de la situation au cours de ces cinq dernières années, l'exercice d'affectation des ménages aux différents niveaux donne une autre version. Selon cet exercice, les conditions s'étaient généralement améliorées dans les villages participants, avec la proportion de ménages appartenant à chacun des trois niveaux s'équilibrant en 2015 par rapport à 2010 lorsque la distribution était beaucoup plus fortement pondérée en faveur des ménages appartenant au niveau le plus bas. Près de 20% de DCG ont dit que les conditions s'étaient améliorées, ce qui était conforme aux résultats de l'exercice d'allocation.Interrogés sur la description des ménages à chaque niveau de bien-être, les ménages du niveau le plus bas ont généralement été caractérisés comme étant confrontés à des difficultés, ceux du niveau intermédiaire pouvaient satisfaire leurs besoins minimums, tout en ayant quelques difficultés, alors que ceux du troisième niveau (supérieur) disposaient de ce dont ils avaient besoin (Figures 2 et 3  2). Quelques DCG ont décrit le manque d'argent, le manque de bonheur et le manque d'harmonie au sein de leur foyer comme étant les caractéristiques principales des ménages du niveau moyen. Les participants de quelques DCG ont aussi mentionné que certaines personnes de ce niveau souffrent de problèmes de santé ou de mauvaises récoltes (Figure 2). D'autre part, la majorité des DCG ont révélé que les personnes du niveau moyen disposent d'une quantité suffisante de nourriture et qu'elles peuvent satisfaire leurs besoins minimums, y compris leurs besoins en termes de santé (Figure 3). Enfin, les ménages au niveau de bien-être le plus élevé ont été décrits comme n'ayant aucune difficulté. Ces ménages sont considérés comme étant capables de subvenir à leurs besoins, ayant des biens et suffisamment de nourriture, et comme étant heureux et vivant en harmonie au sein de leur foyer (Figure 3). Au vu des réponses, diverses nouvelles pratiques agricoles ont été mises en évidence par les participants des DCG, notamment au cours des cinq dernières années. Au titre des pratiques les plus courantes, on peut citer l'utilisation accrue d'intrants agricoles et les différentes techniques de gestion des terres. Les évolutions les plus fréquentes dans l'utilisation des intrants ont porté sur l'utilisation accrue de pesticides (n=32/52 (62%)), de fumier ou de fosses à fumier (n=32/52 (62%)) et d'engrais (n=26/52 (50%)). Les répondants ont avancé l'existence de changements dans la gestion de leurs champs ces cinq dernières années, notamment l'utilisation accrue d'animaux de trait (n=38/52 (73%)), de charrues (n=16/52 (31%)), de cordons pierreux (n=27/52 (52%)) et de semis en ligne (n=17/52 (33%)). Près d'un tiers des groupes de discussion ont signalé que les ménages disposent désormais de potagers (n=19/52 (37 %)) et que l'accent est davantage mis sur la conservation et la gestion des forêts (n=20/52 (39 %)). Par ailleurs, entre 5 et 15% des DCG ont déclaré avoir modifié leurs pratiques en matière de techniques de repiquage, de désherbage, de culture de nouvelles variétés, de semences anticipées dues à la rareté des pluies et à la technique consistant à inverser chaque année le sens du labourage, de gestion de l'eau et d'utilisation de fumier organique. Certains groupes de discussion ont expressément mentionné que HKI leur avait conseillé de planter des légumes à feuilles vertes. Les réponses recueillies de la part des DCG des hommes et des femmes ont été généralement similaires. Néanmoins, il ressort des discussions des femmes qu'elles ont insisté sur le recours aux charrues à traction animale (n=8/12 (67%)), l'utilisation d'engrais (n=16/26 (62%)) et l'importance accrue accordée à la conservation et à la gestion des forêts (n=12/20 (60%)), alors que les hommes ont plutôt évoqué les pratiques de désherbage et l'utilisation d'intrants agricoles.Parmi les nouvelles pratiques, les changements majeurs concernent l'utilisation d'animaux de trait (n=28/52 (54%)) et de pesticides (n=23/52 (44%)). En outre, les répondants des DCG ont aussi cité les changements dans l'utilisation des engrais (n=20/52 (39%)), des charrues (n=18/52 (35%)), du fumier (n=13/52 (25%)), des cordons de pierre (n=12/52 (23%)) et des charrues à traction animale (n=12/52 (23%)) comme significatifs. Contrairement aux changements dans les pratiques agricoles, très peu de changements ont été mentionnés dans les DCG concernant les pratiques d'élevage. Les deux changements qui ont été relevés sont relatifs aux pratiques d'alimentation du bétail (n=22/52 (42%)) et à celles de soins vétérinaires (n=21/52 (40%)). Les DCG des femmes, par rapport aux hommes, ont tendance à se concentrer davantage sur l'importance de l'utilisation des animaux de trait (n=19/28 (68%) et n=9/24 (38) et des pesticides (respectivement, (n=14/28 (50%) et n=9/24 (38%)). En revanche, dans les DCG, les hommes sont plus susceptibles de souligner l'intérêt du changement lié à l'utilisation des charrues qui a été signalé dans la moitié de leurs entretiens contre moins d'un quart de ceux des femmes (respectivement, (n=12/24 (50%) et n=6/28 (21%)).Les raisons pour lesquelles les personnes interrogées ont estimé que ces changements étaient significatifs concernaient en grande partie leur contribution à l'augmentation de la productivité. Par exemple, elles ont déclaré que l'utilisation de cordons pierreux avait permis d'éviter la perte de fumier ou d'engrais due à la pluie ou à l'arrosage et donc de protéger les sols (n=8/52 (15%)), tandis que l'utilisation d'engrais, y compris le fumier, enrichissait les sols et augmentait leurs récoltes et leur productivité. Selon les répondants l'utilisation de pesticides a aussi contribué à alléger le travail en réduisant le temps nécessaire au désherbage. Enfin, les participants ont estimé que l'utilisation de charrues à traction animale a permis d'accélérer leur travail et de cultiver de plus grandes superficies (n=4/52 (8%)), ce qui SUMMARY | APRIL 2010 est également perçu comme ayant contribué à l'augmentation de la productivité. Hormis l'importance des changements liés à l'augmentation de la productivité, les DCG ont révélé que d'aucuns ont aussi souligné le rôle des potagers familiaux leur permettant de disposer de légumes frais pour la consommation.A la lumière des DCG Il est apparu que les types de biens agricoles utilisés étaient disparates selon les niveaux de bien-être, les différences les plus importantes se situant entre le niveau le plus bas et les deux autres (Figure 4). Au niveau le plus bas, l'intrant agricole le plus souvent mentionné était la pioche (daba  En général, les réponses des DCG des hommes étaient similaires à celles des femmes, bien que celles-ci aient été plus nombreuses que les hommes à expliquer que les individus appartenant au niveau inférieur de bien-être avaient plus de connaissances que ceux des autres niveaux, à savoir : ils sont au courant de l'utilisation du fumier organique (n=4/28 (14%) contre n=1/24 (4%)); et de l'utilisation des engrais (n=2/28 (17%) contre (n=0/24 (0%)).Les répondants des DCG ont indiqué qu'il était plus difficile pour les personnes du niveau de bien-être le plus bas de mettre en pratique les nouvelles techniques qu'elles avaient apprises, à moins qu'elles ne bénéficient d'un soutien pour utiliser ces nouvelles pratiques n=14/52 (27%)). De même, au niveau intermédiaire, les personnes interrogées ont estimé qu'il serait difficile pour les agriculteurs de ce niveau de bien-être d'adopter de nouvelles techniques (n=15/52 (29%)).Toutefois, plus de la moitié des DCG (n=35/52 (67%)) estiment que les personnes appartenant au niveau de bien-être supérieur peuvent adopter ou utiliser de nouvelles pratiques parce qu'elles disposent d'argent ou de moyens.Au-delà de cette caractérisation générale, des points plus spécifiques concernant l'application de nouvelles pratiques agricoles ont émergé. Par exemple, certaines DCG ont mentionné que les personnes du niveau inférieur utiliseraient de nouvelles pratiques si celles-ci ne coûtaient rien ou si elles en avaient les moyens (n=9/52 (17%)); et 8-15% d'entre elles ont signalé que les individus du niveau intermédiaire appliqueraient de nouvelles pratiques agricoles, notamment l'utilisation d'engrais, de charrues, d'animaux de trait, de pesticides et la plantation en rangs, tandis que ceux du niveau supérieur adopteraient l'utilisation de pesticides (n=13/52 (25%)) et d'engrais (n=11/52 (21%)) et dans une moindre mesure, l'utilisation de cordons pierreux, d'animaux de trait, de charrues et la plantation en rangs (entre 6-12% des DCG).En général, les réponses des DCG de femmes sont similaires à celles des hommes, bien que les premières soient plus susceptibles de mentionner que les agriculteurs du niveau plus bas adopteraient quelques nouvelles pratiques (n=6/28 (21%)) alors que les hommes insistent sur le fait que cette catégorie là (niveau 3) ne peuvent pas mettre en oeuvre des pratiques nécessitant des moyens (n=6/24 (25%)). Comparées à celles des femmes, les réponses des hommes ont également plus tendance à indiquer que les individus appartenant au niveau le plus élevé peuvent adopter de nouvelles pratiques puisqu'ils disposent de l'argent ou des moyens ( respectivement, n=20/24 (83%) et n=15/28 (54%)).  également mentionné que ceux du niveau moyen ont recours à d'autres moyens de subsistance employés par des gens plus modestes, mais l'utilisation de ces moyens n'est pas aussi courante. À l'instar du niveau le plus faible, les femmes étaient plus sujettes à évoquer une éventuelle réduction de la fréquence ou de la quantité de nourriture, alors que les hommes parlaient plutôt d'emprunt de nourriture. Certains répondants aux DCG ont noté que les plus pauvres et de condition modeste donnent la priorité à leurs enfants en vue d'obtenir de la nourriture en période de pénurie. Pour le niveau le plus bas, cela a été mentionné exclusivement dans les DCG des féminins et pour le niveau moyen, par les DCG des hommes. Pour les plus nantis (niveau supérieur), la grande majorité des groupes de discussion ont jugé que les périodes de pénurie ne constituaient pas un défi pour eux. Quelques groupes de discussion ont déclaré qu'en cas de crise, ils réduiraient la quantité de certains types d'aliments, comme l'huile, pour y faire face.Causes de régression en matière de santé et de nutrition infantile (6-24 mois) Des participants aux groupes de discussion ont soutenu qu'il existe deux principales voies de régression en matière de nutrition chez les enfants ne variant pas en fonction du niveau : maladie et manque de nourriture. Aux niveaux inférieur et intermédiaire, les participants ont discuté de la régression des enfants en cas de disette ou d'insécurité alimentaire dans leur foyer. Certaines DCG de femmes ont également expliqué que la faible productivité agricole pourrait entraîner une dégradation de la santé ou de l'état nutritionnel chez les enfants du niveau le plus bas. Les réactions pour les niveaux supérieur et intermédiaire sont similaires à celles du niveau le plus bas. Quelques groupes de discussion des hommes ont également mentionné que la paresse pourrait entraîner une dégradation de santé et de nutrition chez les enfants.La maladie et le manque de nourriture étant reconnus comme causes principales de la dégradation en matière de santé et d'état nutritionnel des jeunes enfants, nourrir les enfants de manière adéquate, leur procurer des compléments alimentaires et les conduire au centre de santé en cas de maladie comptent parmi les solutions les plus couramment proposées dans le cadre de l'amélioration de la santé et de l'état nutritionnel des jeunes enfants. Une autre solution éventuelle évoquée est le fait que les parents travaillent plus pour améliorer leur condition de vie en général. Les participants aux DCG, hommes et femmes, ont décrit diverses manières dont les femmes soutiennent leurs conjoints dans leur quête de bien-être. Le plus courant consiste à les épauler aux travaux agricoles, plus fréquemment évoqué dans les DCG de femmes que dans celles des hommes ( respectivement, n=16/28 (57%) et n=10/24 (42%)). Les autres voies par lesquelles les DCG estimaient que les femmes pouvaient soutenir leur conjoint en vue de progresser concernent le soutien de leur conjoint dans toutes leurs activités (n=18/52 (35%)), le soutien financier (n=17/52 (33%)) ou l'aide aux dépenses du ménage (n=20/52 (38%)). Quelques DCG ont déclaré que les femmes pouvaient également aider leur conjoint en s'acquittant de ses tâches pendant son absence, en exerçant des activités génératrices de revenus, en veillant à la garde des enfants (mentionné uniquement par les hommes) et en assurant un cadre familial harmonieux (mentionné uniquement par les femmes).Les réactions apportées par les DCG sur la capacité des femmes à progresser à tous les niveaux ont été semblables à celles des hommes et sont axées sur les efforts et l'apprentissage de nouvelles compétences et techniques. Celles-ci étaient également similaires à tous les niveaux de bien-être. L'une des plus grandes différences liées à l'évolution des femmes à chaque échelon était que les répondantes aux DCG des femmes comparés aux DCG des hommes étaient beaucoup plus enclins de mentionner les rôles des femmes dans l'élevage, l'agriculture, le jardinage et le commerce pour aider leurs ménages à progresser.Les facteurs les plus courants qui entravent la progression des femmes à chaque niveau sont les mêmes que pour les hommes : la paresse ou le refus de travailler, la maladie ou les dépenses liées à la santé. Néanmoins, les groupes de discussion des femmes ont évoqué de nombreux autres obstacles spécifiques à la progression des femmes, tels que le contrôle de leur revenu par le mari et le manque de soutien de la part de leur conjoint, les dépenses domestiques ou familiales et la malédiction. Bien qu'ils diffèrent des problèmes vécus par les hommes et qu'ils soient plus souvent abordés pendant les discussions consacrées aux femmes, ces obstacles ne varient pas en termes de niveau de bien-être.À la question de savoir dans quelle mesure les hommes pouvaient aider leurs épouses à progresser en termes de bien-être, la réponse la plus fréquente, tant du côté des femmes que des hommes des DCG, consiste à dire que les conjoints devraient soutenir leurs épouses dans toutes leurs activités. Même si c'est la réponse la plus récurrente dans les deux types de DCG, presque deux fois plus de groupes de discussion de femmes que d'hommes l'ont mentionnée ( respectivement, n=20/28 (71%) and n=9/24 (38%)). Les hommes, contrairement aux femmes, étaient plus susceptibles de déclarer qu'ils pourraient soutenir leurs épouses avec un appui financier à leurs activités génératrices de revenus ( respectivement, n=10/24 (42%) et n=4/28 (14%)). Les autres moyens par lesquels les groupes de discussion ont expliqué que les hommes pouvaient aider leurs épouses à progresser consistaient principalement à les soutenir dans leurs activités liées à l'agriculture, y compris le jardinage, ou dans leurs autres activités génératrices de revenus comme le commerce. Les participants aux DCG ont décrit divers changements qui ont pu contribuer à l'évolution positive du bien-être, telles que l'augmentation de la productivité agricole et l'augmentation du nombre de légumes disponibles pour la consommation. Plusieurs groupes de discussion ont déclaré avoir constaté des changements majeurs concernant les pratiques agricoles durant ces cinq dernières années, contribuant ainsi à alléger la charge de travail et à accroître la productivité. Parmi ces changements, figurent des modifications apportées à l'utilisation des animaux de trait, des engrais et des pesticides, qui sont autant de facteurs plus souvent associés aux ménages vivant à un niveau intermédiaire ou supérieur comparé à ceux du niveau inférieur. En dépit de la diversité des évolutions agricoles, très peu de changements dans les pratiques d'élevage ont été mentionnés par les DCG. Selon des répondants, il était nettement plus facile pour les personnes appartenant au niveau supérieur d'adopter de nouvelles pratiques agricoles. Quant aux personnes du niveau le plus faible, voire moyen dans certains cas, elles ne pourront prétendre à l'adoption de nouvelles pratiques que si les changements ne nécessitent des fonds ou si les moyens nécessaires leur sont fournis.De manière générale, les plus grands défis auxquels hommes et femmes sont confrontés pour favoriser le progrès de leur ménage sont la paresse, le refus de travailler, la maladie et/ou les dépenses en matière de santé. Quant aux femmes, diverses difficultés liées à la dynamique du couple ont été abordées, notamment l'absence de soutien et le contrôle des revenus par le conjoint. Les plus défavorisés ont aussi été reconnues comme ayant des défis importants liés à l'accès aux ressources et aux fonds servant à améliorer leur condition. Certains groupes de discussion ont jugé nécessaire d'aider les ménages à adopter de nouvelles pratiques agricoles et à accroître leur productivité. Les discussions ont révélé que l'agriculture et la hausse de productivité agricole constituent une approche décisive pour l'amélioration des conditions de vie des ménages dans cette région du Burkina Faso. En revanche, les foyers les plus démunis sont susceptibles de solliciter une aide financière ou des intrants pour accroître leur productivité. De plus, il apparaît que l'autonomisation des femmes ou l'équité des genres peut être améliorée en ce sens que les femmes reçoivent un surcroît de soutien de leur époux dans leur activités domestiques et génératrices de revenus et qu'elles aient davantage de contrôle sur les revenus générés.Il est également probable que le soutien aux ménages en période de crise alimentaire ait un impact, car les enfants et les autres personnes appartenant aux niveaux le plus bas et intermédiaire sont exposés à la faim en période de disette et à la dégradation de leur santé et de nutrition. Par ailleurs, les risques liés à la santé entraînent une charge supplémentaire pour les ménages déjà éprouvés. Ce soutien pourrait revêtir plusieurs formes, notamment des espèces, des bons ou la provision directe de nourriture, et pourrait être destiné aux enfants et à tout autre membre vulnérable du ménage, comme les femmes enceintes ou allaitantes.","tokenCount":"3586","images":["150303740_1_1.png","150303740_1_2.png","150303740_1_3.png"],"tables":["150303740_1_1.json","150303740_2_1.json","150303740_3_1.json","150303740_4_1.json","150303740_5_1.json","150303740_6_1.json","150303740_7_1.json","150303740_8_1.json","150303740_9_1.json","150303740_10_1.json","150303740_11_1.json","150303740_12_1.json"]}

data/part_2/0112905894.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"ddb4463629cecf9a9362653f1a0cb857","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/c288c161-e573-4269-bb6f-ca3fcd4d158c/retrieve","description":"En las últimas décadas, el mundo ha hecho un progreso impresionante para mejorar la calidad de vida de millones de personas; sin embargo, todavía sigue inconclusa la tarea de garantizar la seguridad alimentaria a los más pobres de una manera sostenible. La explosión demográfica, la expansión urbana, la desnutrición y la mala salud persistentes, las tierras agrícolas degradadas y el agua escasa, la carencia de poder de las mujeres, la globalización acelerada y la rápida aparición de nuevas tecnologías - todos estos y muchos otros factores están influenciando este esfuerzo continuo. Este libro compila docenas de resúmenes y artículos para presentar las perspectivas de los expertos sobre estos tópicos vitales. Producidas como parte de la iniciativa \"Visión de la alimentación, la agricultura y el medio ambiente en el año 2020,\" del Instituto Internacional de Investigaciones sobre Políticas Alimentarias, las piezas coleccionadas aquí ofrecen una representación completa de los temas de política que el mundo debe abordar si ha de superar la pobreza, el hambre y la degradación ambiental; y le señalan además el camino a las acciones de política que deben ejecutarse para lograr estos objetivos.","id":"-209618447"},"keywords":[],"sieverID":"72a5de9d-fcbd-4faa-b229-73763fa9b6cd","pagecount":"5","content":"futuro no deberían llevar el mundo a hacerse complaciente sobre la habilidad de los programas de planificación familiar para tener un impactostá la tierra encaminada hacia una masiva sobrepoblación para el año 2020 y más allá? O más bien está la población del mundo destinada a declinar debido a los horrores de la enfermedad y la guerra? Podrán ser alimentadas todas las bocas en los años venideros, sin importar cual sea el tamaño de la población?Los expertos parecen estar de acuerdo en que para el año 2020 la población se incrementará de los seis mil millones de personas que hay actualmente a cerca de siete mil seiscientos millones. Sin embargo, para después de ese año las predicciones varían enormemente. Por ejemplo, las Naciones Unidas ofrecen tres posibles predicciones de población para después del 2020, que van desde cinco mil seiscientos millones hasta diecisiete mil quinientos millones de personas en el año 2100.Aunque el futuro lejano puede ser incierto, en la actualidad hay un incremento de población en el mundo en desarrollo de 75 millones de personas por año. Los diseñadores de política deben reconocer esta realidad y a la vez poner atención a las realidades que hay detrás de los números de crecimiento de la población en el futuro más distante.Sin embargo, según los investigadores, los diseñadores de política no deben concentrarse sólo en el tamaño absoluto de la población, sino dirigir su atención a las implicaciones de los números, las cuales variarán aún más enormemente que los números mismos. A qué regiones geográficas serán distribuidos los números? Tendrán las difer-entes regiones la capacidad-o la \"capacidad de carga\"-para sostener la población en términos de alimentación e ingreso? Será contrarrestado el crecimiento de la población, por ejemplo, por las nuevas tecnologías que incrementan la producción de alimentos para cubrir las necesidades de alimentos? O los incrementos en la producción de alimentos harán crecer aún más a la población, la cual a su vez agotará la oferta de alimentos? Cuál será la distribución por edades de la población? La mejor calidad de vida, los servicios sociales y los programas de planificación familiar, ayudarán a reducir substancialmente las tasas de fertilidad? 0RYLPLHQWR GH SREODFLyQ YHUVXV FUHFLPLHQWR GH SREODFLyQ Una de las fuerzas más importantes detrás del crecimiento de la población es el poderoso impacto de los incentivos de política para reacomodar la gente alrededor del globo. De acuerdo con Steve Vosti, un investigador del IFPRI, \"Hay una distinción importante entre crecimiento de población y movimiento de población. La políticas deben examinarse para ver si están induciendo la migración de grandes cantidades de personas hacia áreas donde la capacidad de carga es la más baja-es decir, las áreas que tienen menos capacidad para sostener grandes cantidades de gente.\"\"Si hay millones de personas moviéndose en un periodo de cinco años hacia áreas agrícolas marginales,\" continuó Vosti, \"es posible prever que las políticas que afectan los movimientos de población podrían tener un impacto más inmediato sobre la base de recursos de estas áreas, que políticas que afectan la fertilidad familiar. Las tasas de mortalidad decrecientes también pueden resultar en incrementos de población que sobrepasen el efecto de las políticas dirigidas a desacelerar las tasas de fertilidad.\"En parte, precisamente, en respuesta a las políticas que afectan el movimiento de población, es que grandes cantidades de gente se han establecido durante las décadas pasadas en regiones costeras y ribereñas, de acuerdo con Thomas Merrick, asesor en asuntos de población del Banco Mundial. \"Esto causa problemas en términos de recursos acuíferos, por ejemplo, en el Oriente del Mediterráneo. El buen gobierno y el buen manejo, planeados con buena anticipación, son críticos para lidiar con estas presiones. Los gobiernos no deben distorsionar el sistema de incentivos a fin de reducir artificialmente los costos reales de la tierra y del uso del agua.\"Sin embargo, los incrementos en la densidad de población no tienen por qué enviar una señal de catástrofe Maltusiana inminente. A medida que la población crece, de acuerdo con Peter Hazell del IFPRI, las capacidades de carga de las tierras actualmente productivas pueden aumentarse a través de tecnologías agrícolas ambientalmente sostenibles. En contraste con la tecnología de la Revolución Verde, en que las mismas mejoras podían aplicarse en todas partes, muchas tecnologías sostenibles son específicas para cada lugar y requieren además esquemas de manejo más complejos. Sin embargo, tales tecnologías, como cultivos resistentes a pestes, podrían incrementar los rendimientos de los cultivos haciéndole poco daño al ambiente. De hecho, algunos investigadores sostienen que si se invierte en ellas ahora, las tecnologías agrícolas podrían incrementar suficientemente la producción de alimentos como para alimentar a todo el mundo-sin importar cuál de las predicciones se cumple.Muchos de los pobres del mundo viven en áreas marginales, tales como laderas y márgenes de bosque, donde las perspectivas de incrementar la productividad en los alimentos son limitadas. Los expertos dicen que se requiere más investigación sobre cómo aumentar la capacidad de carga de esta tierra; aunque, mientras tanto, pueden tomarse acciones que incluyan, por ejemplo, el establecimiento de \"encadenamientos de mercado.\"\"En áreas marginales, con frecuencia la gente trata de subsistir con el alimento que producen en sus propias huertas,\" dijo Vosti. \"Sin embargo esto es extremadamente exigente sobre su pedazo de tierra, y con frecuencia no sostenible en el largo plazo. Lo que se necesitan son 'encadenamientos de mercado'-es decir, la capacidad de producir un cultivo comercial o extraer un producto de su tierra y venderlo en el mercado por dinero a fin de comprar en vez de producir alimentos. El café, por ejemplo, crece bien en algunas zonas marginales donde los cultivos alimenticios no crecen bien. Sin embargo, debe existir un buen mercado para café-o sea, un mecanismo mediante el cual el café puede ser intercambiado por comida a tasas razonables a lo largo del año.\"Finalmente, Vosti señala la importancia de considerar la proporción de la población en edad de trabajar versus la proporción que no puede trabajar. \"Si una proporción grande de la población es mayor de 60 años y menor de 10 años, puede haber problemas reales en términos de la habilidad de la gente que trabaja y de los gobiernos para mantener el ingreso y la producción de alimentos a un nivel suficiente para sostener los componentes 'dependientes' de la población.\"Aquellos que están mirando sólo a los números de población, no a los aspectos morales, pueden preguntar: Una vez que la gente está bien alimentada, bien sea sobre tierras marginales o productivas, no incrementará esto el crecimiento de la población y exacerbará el estado de congestión del mundo y la devastación ambiental? Así como la construcción de nuevas autopistas parece crear aún más carros y embotellamientos de tráfico, llevará el incremento en la oferta de comida a más incrementos de la población?No tanto, dicen los investigadores. \"Si la comida está disponible a precios al alcance de la gente pobre-si hay suficiente 'disponibilidad de comida'-esto será un factor crítico para incrementar el bienestar general de la familia y, por lo tanto, contribuir a reducciones en la mortalidad de infantes y niños,\" dijo Vosti. \"Cuando la gente está comiendo mejor y se encuentran mejor, tienden a invertir más en cada niño, en vez de invertir en un mayor número de niños, y las tasas de fertilidad de la familia tienden a disminuir.\"También contribuyen a una mejor calidad de vida y, por lo tanto, a tasas reducidas de fertilidad, las inversiones en programas sociales, tales como en educación, particularmente para mujeres.\"Gracias a décadas de inversión en la gente, hemos tenido resultados favorables en nuestras tasas de alfabetismo, tasas de fertilidad y tasas de mortalidad infantil,\" dijo la Doctora Rebeca Grynspan Mayufis, exvicepresidenta de Costa Rica y miembro del comité asesor internacional de la iniciativa Visión 2020. \"La educación ha estado en el centro de nuestras preocupaciones desde el siglo XIX y es el principal factor que explica dónde estamos hoy. Tenemos educación obligatoria y gratuita para muchachas y muchachos y hemos tenido una expansión reciente de la educación secundaria para ambos sexos. Nuestra tasa de alfabetismo es ahora del 95 por ciento, y esto hace a la gente más receptiva a los mensajes de planificación familiar.\" En los años sesenta, el país redujo la tasa de crecimiento de su población en más del 33 por ciento.3ODQLILFDFLyQ IDPLOLDU Otra fuerza crítica en el control y reducción del crecimiento de la población son los programas de planificación familiar. De acuerdo con Robert Engelman, director del programa de población y ambiente de Population Action International, los altos números de población predichos para el futuro no deberían llevar el mundo a hacerse complaciente sobre la habilidad de los programas de planificación familiar para tener un impacto.\"El crecimiento de la población no es un patrón mecánico, fácilmente predecible,\" dijo Engelman. \"Hay una combinación de 'comodines' buenos y malos que podrían cambiar tremendamente el actual patrón de crecimiento de la población. Las tasas de nacimiento podrían bajar más rápido de lo esperado o las tasas de mortalidad podrían reversar su descenso histórico y subir. Los números que nos ofrecen los demógrafos no son predicciones sino estimaciones basadas en supuestos, y las estimaciones bajas son tan plausibles como las altas.\"Según Engelman, una de las principales metas de la Conferencia Internacional sobre Población y Desarrollo en el Cairo, y de las organizaciones de población en todo el mundo, es poner en práctica políticas para que se alcance en todo el mundo un nivel de fertilidad de \"reemplazo\"-estimado en 2.1 niños por familia. La tasa de fertilidad promedio es ahora de 4 niños por familia. El dice que esta reducción podría alcanzarse satisfaciendo toda la demanda por servicios de planificación familiar insatisfecha, cerrando la brecha del género en la educación y mejorando el estatus de las mujeres en la familia y en el lugar de trabajo.\"Las comunidades de agricultura y población tienen una causa común,\" dice Engelman. \"Es en el interés de la comunidad de desarrollo agrícola apoyar la planificación familiar y los programas sociales porque estos programas harán su trabajo de alimentar el mundo más fácil. De la misma manera, es en el interés de la comunidad de población que la gente esté bien alimentada, saludable y en capacidad de tomar decisiones acerca de sus propias vidas.\"Sin embargo, de acuerdo con Vosti, \"Bajar las tasas de fertilidad será un proceso lento, y millones de bocas nuevas tendrán que ser alimentadas mientras tanto. Sin embargo, lo más probable es que estas nuevas bocas nazcan en zonas menos capaces de alimentarlos, y, por lo tanto, hay tres alternativas para prevenir el hambre: incrementando sosteniblemente la productividad agrícola, destruyendo el ambiente natural o suministrando ayuda alimentaria masiva. La única alternativa razonable es la primera, y la hora para comenzar fue ayer.\"","tokenCount":"1776","images":[],"tables":["-209618447_1_1.json","-209618447_2_1.json","-209618447_3_1.json","-209618447_4_1.json","-209618447_5_1.json"]}

data/part_2/0156742065.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"2013d6673d6dda418720f7f13d0ab22f","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/0f0264e4-6a02-40b7-bace-abfc4e3726f4/retrieve","description":"Data are essential both to understand the level of malnutrition in a country and to develop strategies to address it. Reliable and recent data are important, as are data that can be compared over time to assess whether strategies are yielding results. Finally, having data available at relevant geographic or administrative units for policy decisions is also critical.","id":"-66614830"},"keywords":[],"sieverID":"b9b451e5-3cfb-4019-b42e-2f182534b52c","pagecount":"8","content":"Data are essential both to understand the level of malnutrition in a country and to develop strategies to address it. Reliable and recent data are important, as are data that can be compared over time to assess whether strategies are yielding results. Finally, having data available at relevant geographic or administrative units for policy decisions is also critical.Recently, the World Health Assembly, of which India is a member, endorsed a set of six global nutrition targets (Exhibit 1) to be used globally at the national and subnational levels to track progress in reducing malnutrition (WHO 2014). These targets reflect a set of nutrition outcomes that, if improved, will have significant impact on short-and long-term human development outcomes. Since the adoption of these targets, the Global Nutrition Report (IFPRI 2014(IFPRI , 2015) ) has attempted to document levels of and trends in these targets for several countries.Unfortunately, in India, the landscape of data sources on levels and trends in nutrition is complex. While several surveys on nutrition are delivered intermittently, it is unclear which, if any, could yield the required data to track the World Health Assembly targets.To enable a better understanding of the nutrition data currently available in India, a team from POSHAN assessed the geographic scope, frequency, availability, EXHIBIT 1 World Health Assembly global nutrition targets and related indicators (WHO 2014) Since 1992, eight unique and multiround populationbased household surveys on health have been conducted in India. The authors performed a review of these surveys and narrowed the list to those that generated data on nutritional outcomes; were representative at least at the district, state, or national levels; and were administered in more than one state. The team found six surveys that met these criteria (Exhibit 2).The team then assessed each of the surveys according to the following criteria and related questions:▶ Geographic scope: Did the survey cover a representative sample across India so that national estimates could be derived? Did the survey generate data that would give statistically representative estimates at the state or district levels?▶ Frequency: How often was the survey conducted? The findings described below are summarized in Exhibit 3.National. Among the six surveys reviewed, the NFHS, DLHS, IHDS, and RSOC were designed to provide estimates of nutritional status for the entire country. However, among the most recent rounds of these surveys, only three-NFHS 3, IHDS 2, and RSOC-were administered in all the states of India.The IHDS 2, however, was a panel survey that followed up households previously surveyed in the IHDS 1 in 2004. Therefore, only the NFHS 3 and RSOC provided nationally representative estimates.State. The NFHS, DLHS, AHS, and RSOC were designed to generate estimates of nutritional status by state. The DLHS used to be administered in all states, but the most recent round, DLHS 4, was not administered in nine states with among the highest rates of undernutrition (Bihar, Rajasthan, Madhya Pradesh, Odisha, Chhattisgarh, Uttar Pradesh, Uttarakhand, Jharkhand, and Assam). The AHS was only administered in a subset of high child mortality states (Bihar, Rajasthan, Madhya Pradesh, Odisha, Chhattisgarh, Uttar Pradesh, Uttarakhand, Jharkhand, and Assam). Therefore, the NFHS 3 and the RSOC were the only two surveys that provided state estimates of nutritional status for all states.EXHIBIT 2 Major nutrition-relevant surveys included and excluded from the assessment Among the surveys that have been conducted more than once, the number of years between each round has ranged from as low as 1 year (AHS 1 to AHS 2) to as high as more than 9 years and counting (NFHS 3 to the current NFHS 4, which is still in progress). The average length of time between each survey round has been 7.0 years each for the NFHS and IHDS, and 4.5 years for the DLHS.As shown in Exhibit 4 , for the latest rounds of completed surveys, data reports and data sets are either readily available in the public domain or unavailable with no information on when a report might be forthcoming, if at all. The availability of each survey's documentation (manuals, questionnaires, sampling information) also varies substantially.The NFHS 3 and IHDS 2 are the only surveys to have complete documentation and data sets publicly and easily available. The NFHS 3 and HUNGaMA are the only surveys to have complete reports, while the remaining surveys have either summary factsheets or no report at all.Among the surveys that have been conducted at a national scale, the RSOC, conducted in 2013-2014, currently offers the most recent source of anthropometric data for the entire country and all states. However, the RSOC's findings are currently only available in the form of fact sheets for India and 29 states, limited information on the survey questionnaires and data collection methods is available, and the data are not released in the public domain.All of the six surveys have varied considerably in the type of data collected, with a few surveys collecting data on all six of the World Health Assembly nutrition targets (Exhibit 5).The NFHS has consistently been the most comprehensive survey on nutrition in India, having collected data on a broad set of indicators on various topics related to nutrition, including all six nutrition targets in the last two rounds. The first two rounds of the DLHS had a very limited focus on nutrition with only the DLHS 2 collecting data on child weight. More recently, however, the DLHS 4 gathered data on all six of the targets. The AHS previously focused on a limited set of indicators related to maternal and child mortality targets, such as birth rate, mortality rates, and health services utilization; it added nutrition indicators to a subsample in its most recent round and collected data on four of the six nutrition targets. Both rounds of the IHDS, HUNGaMA, and RSOC gathered data on three, four, and five of the targets, respectively.Given the need for the World Health Assembly to track all countries' progress in the same manner, the team reviewed the surveys to see if they were using the same measures as those used globally. The team also reviewed the surveys from round to round to determine if the same measures were used for each round to determine whether the survey could report trends over time. The team reviewed the following three measures: reference age group for anthropometry, target group of survey respondents (which affects the indicators' denominators), and the definition of the exclusive breastfeeding indicator.Reference age group for anthropometry. All of the most recent rounds of the six surveys used the age reference group of less than five years, which conforms with the global nutrition targets for child anthropometry.However, among the surveys that have been conducted more than once, the NFHS used a different reference age group each round, making comparisons among each round challenging without an analysis of the raw datasets. The IHDS used the same reference group for child anthropometry (< 5 years and 8-11 years) in both rounds. However, because it surveyed the same households in both rounds, it is statistically inappropriate to use this survey to estimate trends in national prevalence.Target respondent group. The NFHS 3 and 4 were the only surveys to use the same target respondent group as the one specified in the global nutrition targets (all women 15-49 years of age).The IHDS and AHS are the only surveys that have consistently used the same target respondent group for their surveys, with both using ever-married women of age 15-49, thus enabling internal comparisons from round to round. The NFHS and DLHS used different survey respondent groups across rounds, thus limiting internal comparisons, especially regarding women's nutrition.Definition of exclusive breastfeeding. Some surveys (NFHS, RSOC, and HUNGaMA) have used the currently accepted 24-hour recall-based indicator 1 (WHO 2010), while other surveys (DLHS) have estimated EBF based on recall up to the date of the survey by mothers of children less than six months of age.In summary, there are important differences across surveys in reference age groups for anthropometry, target respondents, and, in some cases, indicator definitions. In spite of these issues, a comparison of data between the NFHS 3 and RSOC currently allows one to see a trend in global nutrition targets over time for India.The lack of regularly updated, readily accessible, and comparable data sources on nutrition can greatly limit effective nutrition policymaking in India. targets or the country's progress in meeting the targets over time. Based on our review of the multiple available data sources, we conclude the following:▶ To understand the current nutritional status of India at the national and state levels, the NFHS 3 and RSOC provide data on the greatest number of global nutrition targets for the greatest number of states.▶ To understand the nutritional status of India over time at the national and state levels, the NFHS 3 and the RSOC are similar enough to be compared. This is because they share a similar geographic scope, collect data on the same global nutrition targets, and have the recommended reference group for child anthropometry. However, the comparability of the target survey respondents (all versus evermarried) is a potential limitation.▶ To understand the current nutritional status of India at the district level, the DLHS 4 generated data to track all six global nutrition targets, albeit for a limited geographic scope of 336 out of 640 districts in India. The AHS CAB generated data on four of the targets, and for only a subset of districts across India. Assessment of trends at the district level remains a massive challenge, as neither one of these district-level surveys has collected data on all six global nutrition targets using a consistent approach over time.Based on the findings from this review, the research team has the following recommendations for the nutrition policy community in India:▶ Establish guidance for a single survey that consistently collects data on all six World Health Assembly global nutrition targets and other essential variables.▶ Create a mechanism for more frequent (at least every three years) data collection to ensure that comparable data on these targets are available for strategic decisionmaking and benchmarking of progress at the national, state, and especially district levels.▶ Use the same indicator definitions, reference age groups, and target group of women respondents in each survey round, so that reliable comparisons can be made over time.▶ Release survey reports, questionnaires, and raw data to the public domain rapidly and simultaneously.","tokenCount":"1716","images":[],"tables":["-66614830_1_1.json","-66614830_2_1.json","-66614830_3_1.json","-66614830_4_1.json","-66614830_5_1.json","-66614830_6_1.json","-66614830_7_1.json","-66614830_8_1.json"]}

data/part_2/0170972332.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"a418f8b0d7208267938d7a5164166bda","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/fb35a8db-5493-459e-a4bd-b65489809928/retrieve","description":"","id":"1328226587"},"keywords":[],"sieverID":"7274e9d7-cd6c-4121-860a-f6c3af261c33","pagecount":"6","content":"hina has achieved remarkable economic and agricultural growth over the past three decades. This growth lifted rural household incomes and transformed the structure of the economy (Fan, Qian, and Zhang 2006). Agriculture in particular has played a crucial role in China's success in achieving food security and reducing poverty. Furthermore, agricultural output has continued to rise in recent years. Grain production has reached new highs, and modern hybrids have boosted yields of major crops such as rice and maize. These agricultural developments emerged from a series of policy reforms, infrastructural improvements, and investments in agricultural research and development (R&D).China stepped up its agricultural R&D spending after the turn of the millennium, ending a period of stagnation in the 1990s. Total public investment in agricultural R&D doubled from 2001 to 2008, reaching 14.0 billion yuan or 4.0 billion PPP dollars (both in constant 2005 prices) (Figure 1). Note that unless otherwise stated all dollar values in this note are based on purchasing power parity (PPP) exchange rates. PPPs relect the purchasing power of currencies better than standard exchange rates because they compare the prices of a broad range of local goods and services-as opposed to internationally traded ones.Government research agencies accounted for 84 percent of public funds for agricultural R&D in 2008, while the remaining 16 percent were directed to the higher education sector (Table 1). That same year, the public sector employed some 43,000Note: \"Total staing\" refers to researchers that are nationally classiied as scientists and engineers.full-time equivalent (FTE) agricultural researchers (those classiied nationally as scientists and engineers); 62 percent stafed government research agencies. Private-sector agricultural R&D data were unavailable for 2008. But two years earlier, in 2006, the private sector's contribution was estimated at 16 percent of total agricultural R&D spending in the country (Hu et al. 2011). Consistent information on funding and human resources for agricultural R&D is diicult to obtain for China due to the sheer number of agencies involved as well as the complexity of oversight and funding structures. Various estimates of agricultural R&D investment are reported in the literature, and these are often diicult to reconcile. The data here pertain to primary agriculture-crops, forestry, livestock, isheries, and agricultural services-as well as the more general area of water conservation. Agricultural machinery and food processing-two categories that are often considered part of the agricultural sector-are excluded from the current dataset to enable crosscountry comparisons.The core of China's public agricultural research system is formed by an array of agricultural research agencies at the national, provincial, and prefectural levels. 2 The main national agricultural research agency is the Chinese Academy of Agricultural Sciences (CAAS). Other key national institutes are the Chinese Academy of Fishery Sciences (CAFS) and the Chinese Academy of Tropical Agricultural Sciences. These, and others, report to the Ministry of Agriculture (MOA). Their focus is on basic research and technologies that address key national priorities and challenges.The Chinese Academy of Sciences (CAS) is the nation's foremost research institution in natural sciences and technologies. CAS undertakes agricultural research as well as overseeing multiple institutes, such as the Institute of Genetics and Developmental Biology, the Institute of Geographic Sciences and Natural Resources Research, the Institute of Botany, the Institute of Zoology, the Institute of Microbiology, and the Institute of Subtropical Agriculture. CAS is administered by the Ministry of Science and Technology (MOST).Each provincial government oversees its own provincial academies of agricultural sciences. In contrast to the national agencies, provincial institutes concentrate on applied research tailored to the agroecological challenges within their provincial boundaries. Prefectures have their own agricultural research institutes as well, which similarly focus on adaptive research of local relevance. Extension falls under the provincial Departments of Agriculture and activities take place at the county level. Links between extension and research institutes or universities are not well-developed (Fan, Qian, and Zhang 2006).It is unclear just how many public agencies are active in agricultural research in China. However, MOA tracks the number of institutes under its own authority and under provincial and prefectural departments of agriculture. At the end of 2007, it counted 1,105 research institutes. Of these, 59 institutes were administered by MOA, 454 were provincial institutes, and 592 were prefectural institutes. These numbers difer somewhat from those of the mid-1980s, when 84 national institutes were in operation alongside 414 provincial institutes and 624 prefectural ones. Thus the country has seen a reduction in the number of agencies at the national and prefectural level, and an increase in provincial-level agencies.Provincial and prefectural institutes tend to be relatively small, averaging, respectively, about 50 and 20 researchers. National institutes are generally larger, employing 100 researchers on average. Though the individual institutes at the lower levels are smaller, their collective R&D capacity is greater. Together, provincial and prefectural institutes accounted for 88 percent of government agricultural research investment in 2007, relecting the system's high degree of decentralization (MOA various years).The share of the higher education sector in total public agricultural research investments, while relatively small at 16 percent, has nonetheless grown rapidly since 2001. In absolute terms, this represents an investment of 2.2 billion yuan or 0.7 billion PPP dollars (in 2005 constant prices). Much of the recent growth in spending went to noncrop research areas, such as livestock and forestry (NBS and MOST various years). The higher education agencies employed some 17,000 FTE researchers in 2008, 38 percent of the country's total public agricultural researchers (NBS and MOST various years).Both multidisciplinary and agriculture-speciic universities are active in agricultural research. In 2007, China had 54 agricultural universities or colleges. Each province has at least one agricultural university, and there are various other agriculture-related colleges as well. National agricultural universities were administered by MOA up until 2000, when they were transferred to the Ministry of Education. Provincial departments of education oversee the remaining agricultural universities and colleges (Fan, Qian, and Zhang 2006)  The data in this brief are derived from secondary sources, or were estimated. More information on data coverage is available at asti.cgiar.org/china/datacoverage.  Additional agricultural R&D resources are available at asti.cgiar.org/china.The private sector has become increasingly active in agricultural R&D in China. From 1995 to 2000, private investment rose from an estimated 3 percent of total agricultural research expenditure to 9 percent (Pray and Fuglie 2001). At the start of this period most of the private funding came from foreign sources. Later, however, national actors grew more involved, though most of these enterprises were still partially stateowned. A nationwide survey by MOA found that private-sector spending had reached 16 percent of total agricultural research expenditure in 2006, totaling 2.0 billion yuan or 0.6 billion PPP dollars (both in 2005 prices) (Hu et al. 2011). The origin of the funding changed also from the 1990s, with domestic enterprises now accounting for almost all of the expenditure. Moreover, this igure does not include investment in food processing, which is left out of the scope of agricultural research in this note for the purpose of international comparisons (as it is also excluded from international calculations of AgGDP). The size and growth of R&D investment in the food processing industry has been substantial and totaled 1.4 billion yuan or 0.4 billion PPP dollars (both in 2005 prices) in 2006. If the subsector were included, it would constitute 42 percent of all private agricultural R&D investment in China (Hu et al. 2011).The research focus of private actors difers from that of public agencies. Private enterprises typically invest in research areas where intellectual property rights are more strongly enforced. They are thus better able to secure potential proits from new technologies. Most private investment in agricultural R&D has been directed towards livestock research, with smaller shares going to crops and isheries (Hu et al. 2011).The State Council Steering Group for Science, Technology, and Education coordinates science and technology (S&T) at the national level. S&T policy and its implementation are primarily the responsibility of MOST, though others may be involved as well. Some of these are, for example, the National Development and Reform Commission (NDRC), CAS, the Chinese Academy of Engineering (CAE), and line ministries such as the MOA and the National Natural Science Foundation of China (NSFC). Also inluential are the Ministry of Finance, the Ministry of Commerce, and to a lesser extent, the Ministry of Personnel and the State Intellectual Property Oice (OECD 2008). In 2007, total government expenditure on R&D across all sectors was 89.4 billion PPP dollars (in 2005 prices), equivalent to 1.3 percent of GDP. Of this general R&D expenditure, the agricultural sector comprised 4.2 percent in that year (NBS and MOST 2008).During the \"Cultural Revolution\" from 1966 to 1976, China's agricultural R&D system was nearly destroyed. After 1978, the government adopted policies to reestablish agricultural R&D agencies and, subsequently, to improve the efectiveness of the R&D system. Signiicant outcomes of this early period were the patent system, policies promoting commercialization of research, and competitive funding schemes. Reforms after 1999 continued the emphasis on research commercialization, along with a sharper focus on promoting innovative capacity and high-tech, large-scale agricultural production systems. In April 2001, the State Council released its \"Development Plan for Agricultural S&T 2001-2010. \"Four key areas of that plan were structural transformation of the agricultural and rural sector, increased agricultural revenue, environmental protection, and international competitiveness.Post-2007 reforms addressed issues of eiciency, duplication, and proitability. Innovation in agricultural S&T was promoted, a supply-chain approach to research was adopted, and new funding mechanisms were established to further partnerships between research institutes, universities, and industry.The government has reduced barriers to private-sector investment in agricultural research as well. In the past, stateowned enterprises had enjoyed favored status. Private investment was discouraged, both outright and by a lack of clear regulatory structures for intellectual property rights and foreign ownership of joint ventures (Pray and Fuglie 2001). Following the reforms of the 1990s, some agricultural research institutes became commercial enterprises, and commercial agriculture-related enterprises began to invest in research.China The 1990s reforms to improve eiciency led to a drop in government staing levels from an average of 122 employees per research institute in 1986 to 85 in 2007 (the number of institutes also fell slightly, as mentioned earlier). Average researcher qualiications improved, however. The share of staf classiied as scientists and engineers increased from one-third of all active research staf (researchers and research support staf ) in 1986 to three-quarters in 2008 (NBS and MOST various years). These scientists and engineers generally held a BSc degree or higher. Of all government R&D personnel, 12 percent held a doctorate, 29 percent held a master's degree, and 59 percent held a bachelor's degree in 2009. Women comprised one-third of the research staf that same year.The agricultural research output of government agencies also grew since the reforms. The number of papers published rose considerably, to more than 23,000 in 2007 from about 7,000 in 1986. Some 630 books were published and 575 patents were awarded. Looking more closely at the share of published books, national institutes contributed 35 percent, provincial institutes 50 percent, and prefectural institutes 13 percent. In the case of patents, national, provincial, and prefectural institutes accounted for 31, 50, and 18 percent, respectively (MOA various years).In recent years, universities have further enhanced their to conduct research by recruiting faculty globally. In private research facilities, 13 percent of the agricultural researchers were qualiied to the MSc or PhD level in 2006 (Hu, Liang, and Huang 2009).In addition to researchers, public research institutes, universities, and private enterprises employ technicians, other research support staf, and administrative staf. In 2008, government agencies employed 7,583 technicians and other research support staf, or 0.28 research support staf per researcher. The research support staf ratio was much lower in the higher education sector, at just 0.04 (NBS and MOST 2009). Universities typically have fewer research support staf, as their primary mandate is education rather than research.Government research institutes derive their funding from diferent sources than private enterprises (Figure 2). Most public research institute funding comes from government grants, the share of which increased from 55 percent in 1990-95 to 86 percent in 2006-07. Government grants are awarded as core funds to be applied towards salaries and beneits or as project funds obtained through competitive schemes. The share of project funds in MOA grants to agricultural research institutes has increased steadily over the years.Private enterprises earn most of their income through commercial activities such as the sale of goods and services. These funding sources accounted for about 90 percent of their income in 2006-07, up from 70 percent in 1996-2000.Bank loans have declined in prominence as a funding source for both government institutes and business enterprises. In 2006-07 they accounted for less than 1 percent and 6 percent of income, respectively.Funding for agricultural R&D in China underwent substantial reform after 1985, which however rendered it increasingly complex. Prior to these reforms, funding was delivered through ive-year government plans (Huang, Hu, and Rozelle 2004). Research staf numbers, rather than institute performance, determined funding allocations. The reforms encouraged research institutes to establish commercial companies and promoted competitive funding through NSFC, MOA, and other government agencies and foundations. It also stimulated collaborative eforts with international organizations and foreign agencies. The new policies rewarded performance by ofering inancial incentives for researchers (Fan 2000). Competitive funding greatly increased due to the reforms, rising from zero in 1985 to some 30 percent by 1998, and further to 41 percent by 2006 (Huang and Hu 2008).At the national level, NDRC authorizes yearly ministerial budgets, including the budgets of MOST and NSFC. S&T funding for the national research agencies, such as CAS and CAAS, is then channeled through MOST, MOA, and related ministries according to the S&T plan. Prior spending patterns and political motivations inluence budgets. Local governments fund the provincial and prefectural institutes. These institutes also receive funds from the national institutes when undertaking collaborative research projects. Research priority setting and budget allocation processes are often not formal or transparent within the ministries and institutes (Fan, Qian, and Zhang 2006).MOA and the Ministry of Finance provide other funds as well for speciic purposes. Some project funds are allocated to attract leading advanced technology from abroad. A number of new funds were created in 2006 to support sustainable innovation within research academies and institutes. Another recent initiative for agricultural R&D involved establishment of an innovation system for major agricultural commodities. Ten agricultural products were included in 2007, with coverage expanding to 50 products in 2009. The initial three-year phase ofered 967.5 million yuan (in current prices) for research on key technologies and their practical application.Allocation of resources across various lines of research is a signiicant policy decision. China's main public agricultural research focus is crops, which accounted for more than half of all research activity in 2008 (Figure 3). Following crops in terms of importance were agricultural services (15 percent), forestry (9 percent), livestock (6 percent), and water conservation (6 percent). In the higher education sector, researchers targeted livestock (19 percent),   Shares of total funding (%)Government Private enterprises 1 9 9 6 -2 0 0 0 1 9 9 6 -2 0 0 0 2 0 0 6 -0 7 2 0 0 6 -0 7 Note: Data for 1990-95 were unavailable for private enterprises.forestry (13 percent), and water conservation (10 percent).Remaining government and higher education researchers focused on isheries and biological sciences.A comparative indicator used to track agricultural R&D spending across countries and over time is the research intensity ratio, calculated as total public spending on agricultural R&D as a percentage of national agricultural output (AgGDP). In China, this ratio ranged between 0.3 and 0.5 percent during 1986-2008 (Figure 4). In 2000, which is the latest year for which global data are available, China's agricultural research intensity ratio was 0.4 percent. In other words, China spent $0.40 on agricultural research for every $100 of agricultural output. While substantially less than the 2.4 percent that high-income countries spent on average on agricultural research, it is more comparable to the 0.6 percent average for the developing world (Beintema and Stads 2010). As recently as 2008 China's agricultural R&D intensity of 0.5 remained below the generally recommended 1.0 percent for developing countries. However, in absolute terms, China's agricultural research spending far exceeds that of any other country except the United States. In 2000, China contributed 9 percent of the 25 billion PPP dollars spent on public agricultural R&D globally (in 2005 prices) (Beintema and Stads 2010). Moreover, China has signiicantly increased its agricultural R&D spending since that time, outpacing both Brazil and India (Figure 5).After three decades of reform, agricultural R&D in China has made considerable progress. Total public expenditures on agricultural R&D doubled from 2001 to 2008, and private expenditure on agricultural R&D grew at an even faster rate. Moreover, preliminary data for more recent years suggests that investments have continued to rise. Furthermore, the government's recently released 2012 Number 1 document indicates that agricultural technology remains high on the policy agenda (Huang 2012).Policy reforms have contributed greatly to the increased public and private investment. Measures have strengthened the patent system and diversiied R&D funding sources by introducing commercialization and competition. Agricultural researcher qualiications have risen as well. The share of scientists and engineers holding a bachelor's degree or higher is now signiicantly greater than in the 1980s. The productivity of government agencies has likewise improved, as evidenced by the rising number of patents and publications.Despite the progress achieved, problems remain in China's agricultural R&D system, and new challenges have emerged. Numerous ministries and agencies are involved in managing and conducting agricultural R&D. The resulting high level of decentralization limits coordination and has led to funding ineiciencies and duplication of research efort. In addition, due to the nature of the social welfare system, individual government institutes bear a substantial inancial burden in relation to their retirees. This problem is growing as the number of retirees rises. Innovation capacity is still limited as well, and is related to the relatively small share of researchers with postgraduate degrees. Most patents are for the adaptation of technology, rather than for new inventions; investment in basic research is still very low. Finally, commercialization of research continues to present both opportunities and challenges. In China, as elsewhere, it has proven diicult to strike an appropriate balance between market-oriented research and research that meets speciic developmental needs.1 This Country Note is based on the 2011 report \"Agricultural R&D as an Engine of Productivity Growth: China\" by Kevin Z. Chen and Yumei Zhang. Unlike other ASTI Country Notes, which are based on primary ASTI data, this study is based on secondary sources, supplemented by interviews with key researchers and policymakers. As is recognized in the literature, obtaining accurate data on agricultural R&D in China is challenging. Several ministries provide funding and oversee agricultural research, with each publishing its own statistical yearbooks.   6 1991 1993 1995 1997 1999 2001 2003 2005 2007  ","tokenCount":"3132","images":[],"tables":["1328226587_1_1.json","1328226587_2_1.json","1328226587_3_1.json","1328226587_4_1.json","1328226587_5_1.json","1328226587_6_1.json"]}

data/part_2/0173922901.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"044ebd65e5e29249d882712ff2abbd52","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/15c550ed-2e23-43d9-a78b-351c157982cc/retrieve","description":"Ces dernieres decennies, de nombreux pays en developpement ontatteint des taux de croissance agricole impressionnants. L'Asie, par exemple, dans les annees 60, etait menacee par la faim et d'inanition massive, et elle est aujourd'hui auto-suffisante en denrees de base, bien que sa population ait plus que double. En depit de cette reussite, de graves preoccupations perdu rent pour I'avenir. La faim et la malnutrition persistent dans de nombreux pays, souvent en raison de sChemas anterieurs de croissance agricole, insuffisants ou inaptes 8 apporter des avantages adequats aux pauvres.","id":"1709439419"},"keywords":[],"sieverID":"ee360ffa-85eb-49dc-8993-127c4ead5f54","pagecount":"2","content":".RECAPITULATIF 2020 N° 59Vision 2020 pour I'Alimentation, l'Agriculture et l'Environnement TRADUIT DE L'ANGLAIS MARS 1999Ces dernieres decennies, de nombreux pays en developpement ont atteint des taux de croissance agricole impressionnants. L'Asie, par exemple, dans les annees 60, etait menacee par la faim et d'inanition massive, et elle est aujourd'hui auto-suffisante en denrees de base, bien que sa population ait plus que double. En depit de cette reussite, de graves preoccupations perdu rent pour I'avenir. La faim et la malnutrition persistent dans de nombreux pays, souvent en raison de sChemas anterieurs de croissance agricole, insuffisants ou inaptes 8 apporter des avantages adequats aux pauvres. Les augmentations escomptees de la demande agricole, associees 8 la croissance demographique et a I'augmentation des revenus par habitant exigeront la poursuite de I'accroissement de la productivite agricole, bien que tout indique que la croissance des rendements connait un ralentissement et que les perspectives d'expansion plus ample des superficies cultivees et irriguees soient limitees. En outre, les problemes environnementaux, associes 8 I'agriculture, pourraient, s'ils ne sont pas jugules, menacer les futurs niveaux de la productivite agricole et imposer des coats importants dans Ie domaine de la sante et de I'environneme nt, a I'echelon national et international.Pour la plupart des pays en developpement, la poursuite de la croissance agricole est une necessite et non pas une option. Mais cette croissance ne peut mettre en danger la base sous-jacente de ressources naturelles, ni imposer des externalites coOteuses aux autres. II lui faut egalement etre equitable pour qu'elle puisse appuyer I'attenuation de la pauvrete et I'insecurite alimentaire. Ces trois objectifs (croissance agricole, attenuation de la pauvrete et durabilite environnementale) ne sont pas necessairement complementaires, et leur realisation simultanee ne peut etre prise pour un acquis. Bien qu'une grande partie soit tributaire des circonstances sociales, economiques et agroecologiques specifiques, un degre de complementarite sera plus probable lorsque Ie developpement agricole est (1) generalise et qu'il englobe les petites et moyennes exploitations agricoles, (2) axe sur Ie marche, (3) participatif et decentralise et (4) entraine par Ie changement technologique valorisant la productivite factorielle, sans degrader la base de ressources. Cette croissance peut reduire les prix alimentaires tout en relevant les revenus agricole, elle est 8 forte intensite d'emploi et elle releve la demande effective de bien et de services non alimentaires, notamment dans les petites vi lies et les centres commerciaux. Grace a la reduction de la pauvrete et la promotion de la diversification economique dans les zones rurales, elle attenue egalement les pressions exercees sur la base de ressources naturelles pour assurer des moyens d'existence.Les imperatifs d'un developpement agricole generalise sont relativement bien compris et il convient de ne pas les oublier dans cette recherche contemporaine de la durabilite environnementale. Puisqu'ils sont de cette importance, il convient de les noter ci-dessous.Dans les annees 50 et 60, les decideurs et les experts du developpement agricole se sont principalement interesses 8 la croissance, et les enseignements tires de cette experience pourraient etre recapitules sous forme de cinq termes clef de la croissance agricole.Innovation: des systemes nationaux soli des de recherche agricole et de vulgarisation (des secteurs public et prive) afin de produire et de diffuser des technologies de valorisation de la productivite.Infrastructure: notamment de bons systemes routiers et de transports.Intrants : des systemes de prestation efficientes pour les services agricoles, notamment pour les intrants agricoles modernes, I'agrotransformation, I'eau d'irrigation et Ie credit.Institutions: des marches efficients et liberalises fournissant aux exploitants agricoles un acces aise aux marches nationaux et internationaux, et des institutions publiques efficaces offrant des services clef lorsque ces derniers ne peuvent incomber au secteur prive.Incitations: des politiques macro-economiques, commerciales et sectorielies propices ne penalisant pas I'agriculture.Dans les annees 70 et 80, les decideurs et les specialistes du developpement ont commence 8 s'axer sur les moyens permettant d'avoir recours au developpement agricole afin de reduire la pauvrete et I'insecurite alimentaire, tout en contribuant a la croissance. Les enseignements degages de cette epoque se resumeraient en six \"rectificateurs des disparites\"de la croissance agricole : 1.    ","tokenCount":"662","images":["1709439419_1_1.png","1709439419_2_1.png"],"tables":["1709439419_1_1.json","1709439419_2_1.json"]}

data/part_2/0182616382.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"ab7d333682df3d88265696c61f30831b","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/705f39fe-77d8-48f3-8800-6f031544fdb0/retrieve","description":"This report describes the baseline data collected from 1,835 men and women respondents in 998 households in two irrigation sites in the central dry zone in Myanmar to help diagnose, design, and test interventions to enhance the Myanmar Agricultural Development Support Project’s impacts on gender equality and nutrition. Baseline data show large gender gaps, in which fewer women than men achieved adequacy in all 11 indicators of empowerment. Eighty-nine percent of women versus 64 percent of men respondents were not empowered, and 66 percent of dual-adult households have gender gaps. The main contributors of disempowerment among women were high tolerance and acceptance of intimate partner violence, lack of work balance, and low membership in groups, especially influential groups. Although 95 percent of respondents owned smartphones, women were less likely than men to access Internet or social media through their phones. Thirty-nine percent of respondents received rice-related information and half received health-related information. Nine to 14 percent of respondents attended agriculture- or health-related training courses. Women were significantly less likely to receive agriculture and nutrition-related information and training than men. The dietary diversity score, a common indicator of diet quality and a good proxy for nutrition, is low in the sample. The individual dietary diversity score was 4.32, with no significant difference between women and men and no major differences between irrigation water users and other households. Dairy, nuts and seeds, eggs, vitamin-A-rich fruits and vegetables, and other fruits are not commonly or frequently consumed by a majority of respondents. Beans and dark leafy vegetables, which are relatively abundant in the study context, are consumed by only 38–48 percent of the respondents on a daily basis. Nutrition education highlighting dietary diversity can help the sample communities achieve better nutrition. Overall, most women and men in the sample communities employ good sanitation practices, but more people need to be sensitized on proper garbage disposal, drinking water treatment, and proper and more frequent handwashing.","id":"-993276048"},"keywords":["irrigation","agricultural productivity","gender","nutrition","crop diversification"],"sieverID":"070bfdfa-063c-401b-8f63-b21c5d7e7c90","pagecount":"58","content":"in 1975, provides research-based policy solutions to sustainably reduce poverty and end hunger and malnutrition. IFPRI's strategic research aims to foster a climate-resilient and sustainable food supply; promote healthy diets and nutrition for all; build inclusive and efficient markets, trade systems, and food industries; transform agricultural and rural economies; and strengthen institutions and governance. Gender is integrated in all the Institute's work. Partnerships, communications, capacity strengthening, and data and knowledge management are essential components to translate IFPRI's research from action to impact. The Institute's regional and country programs play a critical role in responding to demand for food policy research and in delivering holistic support for country-led development. IFPRI collaborates with partners around the world.vii List of Tables Table 1.   18. Proportion (%) of survey respondents by attitude about self-efficacy and gender norms .......... Table 19. Proportion (%) of survey respondents by leadership and engagement in the community.......... Table 20. Proportion (%) of survey respondents who answer nutrition questions positively or correctly . Table 21. Proportion (%) of survey respondents by proper handwashing practices ................................. Table 22. Proportion (%) of sample households by water source and sanitation practices ....................... Table 23. Proportion (%) of survey respondents consuming different essential food groups (24-hour food recall) ..       In May 2019, the Japanese Scaling Up Nutrition (SUN) Trust Fund, the Agricultural Development Support Project (ADSP), and the International Food Policy Research Institute (IFPRI) initiated a research project (called SUN study hereafter) that aims to diagnose, design, and test interventions to enhance gender equality and nutrition impacts within a sample of the ADSP coverage areas. ADSP (P147629) runs from 2015 to 2022 and aims to increase crop yields and cropping intensity in selected existing irrigation sites in Nay Pyi Taw, Bago East, Mandalay, and Sagaing regions. ADSP aims to improve irrigation infrastructure and provide technical support in irrigation governance and enhancing agricultural productivity and incomes of farmers. Starting in 2014, World Bank projects are encouraged to report on learning outcomes based on their nutrition and gender mainstreaming. The objective of the SUN study is to test delivery approaches to enhancing women's empowerment and the nutrition outcomes to feed into the learning agenda that can be used for future scale-up of ADSP project.Agriculture is a potential pathway to improved nutrition, and the need for investments that boost agricultural production, keep prices low, and increase incomes is undisputable. Evidence of the nutritional effect of agricultural programs, however, is inconclusive; studies have shown that increased income does not necessarily lead to improved dietary diversity and nutrition (Ruel and Alderman 2013). Various studies also have indicated the critical role of women's empowerment in food and nutrition security (Cunningham et al. 2014;Kumar et al. 2018;Malapit et al. 2018;Ruel and Alderman 2013;Ruel et al. 2018).We implement the empirical research in Myanmar, a country with huge agricultural potential but confronted by a high burden of malnutrition (MIID 2018;World Bank 2012). Myanmar performs worse in nutrition than most of its East Asian neighbors including those with lower income (World Bank 2012). The National Nutrition Center identified five undernutrition problems: protein energy malnutrition and high deficiencies in four micronutrients (iron, iodine, vitamin A, and vitamin B1), especially among women and children (MIID 2018). A study by Mahrt et al. (2019) shows that the country needs to look at the availability and affordability of nutritious foods and the role of nutrition education in addressing malnutrition.In Myanmar, the Multisectoral National Plan of Action for Nutrition (MS-NPAN) has been developed and was approved in 2018. MS-NPA highlights women's lack of empowerment and poor knowledge and education as major underlying causes of malnutrition. Key indicators in the MSNPAN include (1) women's participation in household decision making, (2) life expectancy ratio of females to males, and (3) women's education attainment of at least secondary level. The SUN study aims to measure and directly contribute to improving the first indicator, and its longer-term goal is to contribute indirectly to improving indicators #2 and #3. This SUN study aims to contribute to key results and planned activities under MS-NPAN including (1) increased capacity of families to improve their practices, (2) improved knowledge of nutrition and health care of women and children, and (3) increased role and capacity of women and their groups in making decisions for health and nutrition issues in families as well as in social or community matters and issues, among others. The consultative processes related to MS-NPAN also generated suggestions to integrate the MS-NPAN activities into agricultural programming; this SUN study aims to help inform the future integration of nutrition into agricultural programs in Myanmar. Some studies have documented the gender-based constraints in access to inputs, land, services, information, and participation in groups and community organizations, coupled with strong social norms and gender biases, that limit women's opportunities and productivity and increase their time burden. Such constraints and norms, in turn, affect their family's and children's nutrition and well-being (Akter et al. 2017;Faxon et al. 2015;MIID 2018;MS-NPAN 2018). Because of decades of military rule and oppression and a largely closed economy, Myanmar has mainly been deprived of research, including research on gender inequalities. Very little data are available from Myanmar, compared to other countries (Akter et al. 2018;Lambrecht and Mahrt 2020). In recent years, several project reports or case studies have looked at gender issues, but few large-sample datasets and rigorous studies capture the status of gender-based constraints and empowerment in the culturally diverse areas of Myanmar. This paper aims to provide empirical evidence on the status of women's empowerment, gender equality, and nutrition using innovative and tested measurements and methods for data collection and analysis. In particular, this paper uses the survey-based tool to quantify women's and men's empowerment through the project-level Women's Empowerment in Agriculture Index (pro-WEAI) launched in 2018 by IFPRI and its partners. Moreover, this paper uses good practice guidance in measuring the Minimum Dietary Diversity for Women of Reproductive Age (MDD-W) score, which is a good proxy for micronutrient adequacy in a population of women of reproductive age. For comparison, we extend this indicator to men.This paper provides the description of the baseline data collected under the SUN study. The specific research questions answered in this baseline study are as follows:• What livelihood strategies do the sample households in the irrigation catchment areas adopt?• What is the level of crop diversification, productivity, and commercialization among farmers in the focus irrigation catchment areas? • What is the level of empowerment between women and men in the focus irrigation catchment areas? • What are the sources of disempowerment among women and men in the study site?• What is the level of nutrition knowledge and practice and diet quality in the study site?• Are there differences in livelihoods, empowerment, and nutrition indicators between irrigation water users and other households? • What are areas to improve women's and men's empowerment and nutrition indicators?The rest of the report is structured as follows. Section 2 describes the sampling method for this baseline study. Section 3 presents the main findings. Section 4 summarizes the main findings and discusses their implications to the World Bank's ADSP and other similar projects.This study focuses on communities in the catchment areas of two irrigation sites, the Sinthe irrigation site in Tatkon township in Nay Pyi Taw region and the North Yamar irrigation site in Pale and Yinmarbin townships in Sagaing region (Figure 1). Both regions are in the central dry zones in Myanmar. These irrigation sites were the focus pilot sites of the ADSP by the time of the survey. Within these irrigation sites, three distribution channels (DYs) were the focus of ADSP in terms of its activities and were the focus of this baseline survey: DY2 in Sinthe site and DY5 and DY6 in North Yamar (Figure 2). These irrigation sites cover about 13,000 acres of irrigated area and are expected to increase coverage by 5,000 acres as a result of ADSP interventions (Table 1). North Yamar's main canal (ADSP's project area) has 10 DYs and one direct outlet (DO) covering 10,082 acres of irrigable area; the Sinthe irrigation site has five DYs and five DOs covering 7,761 acres as irrigable area (Table 1). According to data from the Department of Agriculture Land Management and Statistics (DALMS), in 2016-17 there were 5,769 farmers in 12 village tracts under the Sinthe irrigation scheme and 5,913 farmers in 16 village tracts under the Nother Yamar irrigation scheme.  A household survey was conducted in the study sites between November 2019 and February 2020. The household survey included both a household questionnaire and an individual questionnaire that was administered separately to a primary woman and a primary man decision maker within each household (usually the husband and wife). Scoping field work was undertaken from July 20 to August 4, 2019. A total of nine complementary community surveys and focus group discussions were conducted on September 9-28, 2019. A series of survey questionnaire pretests, particularly on women's empowerment modules, was conducted from October 22 to November 15, 2019, at the villages near the North Yamar and Sinthe irrigation sites.The irrigation catchment areas are defined as the villages surrounding or engaged in farming in DY2 in the Sinthe irrigation site and DY5 and DY6 in the North Yamar irrigation site. These villages were identified by the irrigation officers, water user group facilitators, and village heads. Figure 3 shows the sampling strategy used for the household survey. A total of 12 villages in Sinthe and 18 villages in North Yamar were selected. In each village, full listing of households in the village was conducted by the survey team supervisors and 29-32 sample households were selected on the basis of stratified random sampling.The listing process was supported by village leaders, hundred household leaders and water user group members in each village. A list of water user group members were received from Nippon Koei organization facilitators and was confirmed and checked again by teams on arrival to each village. Households were stratified into three types. The first two types are the irrigation water users (called irrigation households hereafter), defined as those with at least one irrigated plot, disaggregated into (1) water users who have started forming water user groups (WUGs) under ADSP and (2) water users whoare not yet part of a WUG. Because WUG formation has just started, we may not see differences in characteristics and activities between WUG and non-WUG members during baseline but expect to see some differences during the follow-up survey. The third type consists of other households with no irrigation plots, although some of them may have nonirrigated, upland plots used for crop cultivation (called other or nonirrigation households hereafter). The set-up of WUGs was concentrated in three villages. We oversampled these villages to ensure a sufficient sample of WUG members. In North Yamar site, I Yaung had 9 WUGs with 88 members, and 42 members were randomly selected. Min Kan Gyi had 25 WUGs and 222 members, and 65 members were randomly selected. Kyay Chaung (North) had nine WUGs and 195 members, and 50 members were randomly selected.Source: Authors' compilation. HH=household; WUG=water user groupsThe total sample is 998 households. This total consists of 340 households that have at least one member who is a member of a WUG (33 percent of the sample), 480 households that are irrigation water users but not yet members of WUGs (49 percent of the sample), and 180 other households with no irrigated plots (18 percent of the sample). The sample WUG members are roughly 40 percent of all the 840 WUG members identified in the two focus irrigation sites (Annex Table 1). The sample non-WUG irrigation users are roughly 23 percent in the focus villages; while sample nonirrigation-water-using households (other households) are roughly 6 percent in the focus villages. Overall, the sample households cover 16 percent of all households in the focus villages in the irrigation catchment areas.The survey consisted of both a household questionnaire and an individual questionnaire that was administered separately to the target woman and target man in the household. The household questionnaire contains basic information on household size and household members, land and crops cultivated, assets, and details on the value chains, including costs and labor. It was administered through a face-to-face interview with the most knowledgeable members or primary decision maker within the household and took, on average, one hour. The individual questionnaire contains information on the participation and experience in the ADSP program, other trainings and extension services, pro-WEAI modules, and gender attitudes. It was administered to the primary woman and man decision makers (interviewed separately) for about one hour each.Myanmar Survey Research (MSR), which was selected through a competitive bidding process, implemented the household survey from November 2019 to February 2020. A total of 28 women and 20 men were trained to be enumerators for the household survey. The survey was administered using tablets and a computer-assisted personal interviewing software called SurveyToGo (Dooblo) on which the interviewers were trained. As the data were collected, they were stored on a secure cloud-based server so that they were accessible to the research team for data checks.This study and all of its procedures were approved by IFPRI's Institutional Review Board as well as by the ADSP director and Irrigation and Water Utilization Management Department (IWUMD) under the Ministry of Agriculture, Livestock and Irrigation (MoALI) in Myanmar. The enumerators received training on research ethics, including the process for seeking informed consent. Prior to being interviewed, all participants provided oral informed consent. Consistent with standard practice, each household in the household survey was offered a package of coffee mix amounting to USD1.80 after the interview in appreciation for its time.We measured five sets of outcome indicators that pertain to (1) gender equality and women's empowerment, (2) nutrition, (3) crop productivity. ( 4) crop diversification, and (5) crop commercialization. We also measured demographic, socioeconomic, and other indicators that may affect the outcome indicators.We adopted the project-level Women's Empowerment in Agricultural Index (pro-WEAI), a survey-based tool to measure women's and men's empowerment and inclusion in agricultural development projects (Malapit et al. 2019). Unlike other empowerment measures (e.g., measures based on the Demographic and Health Surveys), which do not typically cover both men and women, pro-WEAI allows for direct comparison between women and men in the same household. The index consists of two subindices: the Three Domains of Empowerment index (3DE) and the Gender Parity Index (GPI). The 3DE aggregates women's and men's achievements across 12 equally weighted indicators that measure three types of agency: intrinsic, instrumental, and collective (Figure 4). Indicators of intrinsic agency are autonomy in income, self-efficacy, attitudes about intimate partner violence (IPV) against women, and respect among household members; indicators of instrumental agency are input in productive decisions, ownership of land and other assets, control over use of income, access to and decisions on financial services, freedom of movement and work balance; and indicators of collective agency are group membership and membership in influential groups. The GPI compares the achievements of women and men in the same household (for dual-adult households only). In this study, we adopted similar adjustments to those adopted by the pro-WEAI development team based on lessons learned from other ongoing pro-WEAI efforts. The first adjustment to the pro-WEAI (described in Malapit et al. 2019) was to exclude one indicator-freedom of movement-from the 12 indicators of the pro-WEAI. The missing indicator, freedom of movement, is meant to capture whether individuals can freely choose to move about within their communities and to neighboring communities; this indicator is currently undergoing further validation by the WEAI team. This indicator was also excluded in other applications of pro-WEAI in Benin, Malawi and the Philippines (Heckert et al. 2020;Malapit et al. 2020;Ragasa et al. 2020). Freedom of movement was also not identified as an issue during the scoping field visits and community surveys for this SUN study. The second adjustment was to relax the definition of asset ownership: adequacy was earlier defined as owning land and any three other assets and is now defined as owning land or any three other assets. The third adjustment was to streamline the self-efficacy indicator, from using eight statements to using three statements because of redundancies in the earlier eight statements encountered during the field team's training and pretest. The fourth adjustment was to expand the list the list of livelihoods covered in the inputs to production decisions, so they include not just agriculture-related activities. The heavy focus on agriculture activities has been the main criticism of the WEAI methodology (Malapit et al. 2019). The fifth adjustment was to simplify the calculation of autonomy in income. Annex Table 2 provides the definitions and adequacy cut-offs for each indicator.The study used the following indicators of nutrition and health. First, we computed a composite index and 12 indicators representing the exposure to knowledge and good practices on nutrition mainly based on the LEARN modules (LEARN 2015). The 12 indicators are binary variables based on survey responses (yes or no; agree or disagree) to a set of 12 questions or statements based on the LEARN modules (Table 2). Second, we asked respondents about their handwashing practices, sanitation facilities and practices, and water source. On the latter two indicators, we calculated both including and excluding water collection time (not deprived if water collection usually takes <30 minutes) and whether sanitation facility is not shared with other households. Other sanitation practices asked about were garbage disposal, drinking water treatment, and whether animals enter food preparation areas. Have heard that the most important time for good nutrition in a child's life is from when a woman becomes pregnant until a child's second birthday.Agreed with \"Like meat, fish, and eggs, legumes are also a body-building food.\"Chose \"Fathers support a pregnant or breastfeeding woman by helping her with chores, making sure she gets plenty of rest, or making sure she eats a variety of healthy foods.\" Disagreed with \"Breastfeeding women should avoid eating all kinds of beans, legumes, vegetables, fruits, eggs and meats because those foods could be harmful for the health of the mother and child.\" Disagreed with \"Rice porridge (tamingazi) is the only solid food that children aged six months to two years old need to be healthy.\" Access to Nutritious Foods Agreed with \"Families should avoid spending money on food which is not nutritious.\" Disagreed with \"Foods gathered from the wild are not as healthy as foods that are bought.\" Have heard that a home garden with good production methods can contribute significantly to household nutrition security.Agreed with \"We must take great care to ensure that the food preparation environment and the eating place are kept clean to protect our families from disease.\" Have heard that washing vegetables before they are cut and using a lid when boiling vegetables are practices that can make prepared food more nutritious. Source: LEARN 2015.Third, the MDD-W indicator for the women and men respondents was calculated. Dietary diversity is a commonly used indicator for diet quality and an international accepted proxy for nutrition. Martin-Prével et al. (2015) establish an MDD-W threshold of 5 out 10 food groups, which at the population level is associated with greater micronutrient adequacy of among women of reproductive age. In this study, we asked each respondent to list the food eaten the day before the interview (24-hour food recall, from the time of waking up to before going to sleep). This study follows the good practice guidelines in FAO and FHI360 (2016) in designing and conducting MDD-W. We used a list of 10 food groups as in FAO and FHI360 (2016) and adapted in the Myanmar context:• Grains, white roots and tubers, and plantains  FAO and FHI360 2016). Examples given to respondents during the survey are any vegetables or roots that are orange colored inside, such as pumpkin, carrots, squash, or sweet potatoes that are yellow or orange inside, or any fruits that are dark yellow or orange inside, such as ripe mango, ripe papaya, muskmelon, or ripe, deep yellow-fleshed or orange-fleshed bananas.Crop productivity is measured as the harvested crop in kilogram per acre of cultivated land per cropping season. Crop productivity is measured in terms of the biggest rice and non-rice parcels and on total harvest reported for the last 12 months divided by the corresponding total cultivated areas. Cropping intensity is defined as the sum of crop areas planted in wet, winter and dry seasons divided by the net area equipped by irrigation and drainage, in percent (World Bank 2015, page 45). Crop diversification is measured in two ways: (1) the number of different crops grown in the same parcels and across all parcels and ( 2) the Simpson's Diversity Index,, where \uD835\uDC50\uD835\uDC50 \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 is the share of the household's total cropland area that is planted with crop \uD835\uDC57\uD835\uDC57, \uD835\uDC57\uD835\uDC57 = 1, … , \uD835\uDC3D\uD835\uDC3D \uD835\uDC56\uD835\uDC56 . SID that is equal or closer to 0 means limited diversification and complete specialization; SID that is equal or closer to 1 means diversification. Crop commercialization is measured as the proportion of total harvest sold.This section describes the sample respondents and households. The communities are pretty homogenous. Almost all respondents are Burmese, with only four respondents who belonged to Shan, Kayin, or Mon ethnicities. All respondents practiced Buddhism.The average household size was 4.22 (Table 3). Eighty-eight percent of the sample households had both female and male adults (dual-adult households) and 12 percent had female adults only (no male adults). Most households were composed of a married couple, one or two of their parents or older members, and either no children or one child. Irrigation households had generally bigger household size than nonirrigation households. Despite cultural similarities, many differences in socioeconomic indicators between irrigation and nonirrigation households were observed. Most households (89 percent of irrigation households and 73 percent of nonirrigation households) had access to electricity (Table 4). Most households also had sufficient rooms and dwelling size and did not experience crowding, an indicator of household welfare (UN 2007). Few households, however, had improved cooking fuel or improved housing materials. Fewer other households than irrigation households had improved cooking fuel or improved housing materials. Overall, irrigation households are not deprived in 59 percent of housing indicators compared 46 percent of nonirrigation households. UN-Habitat defines crowding, and therefore deprivation, when there are more than three people per room (UN 2007). Crowding indicator excludes kitchens, toilets, corridors, balconies, and rooms used only for business. Improved cooking fuels include electricity, liquified petroleum gas, or biogas. Improved housing is a house that is not a hut and does not have natural or rudimentary floors, walls, or roof.In terms of individual respondents, the average age was 51 (Table 5). Ninety-two percent belonged to dual-adult households and 8 percent belonged to female-adult-only households. There was low level of education in the study site, with 46 percent of respondents having no formal schooling or only a few years in primary school. Only 11 percent of respondents were high school graduates or higher. Sample women had significantly lower educational attainment and literacy than men. Eighty-four percent of the respondents were married, 9 percent were single, and 7 percent were widow/divorced/separated. Most sample respondents (81 percent of men and 71 percent of women) were engaged in farming activities (Table 5). Qualitative interviews and focus group discussions revealed that farming was seen as \"family activity\" in which the husband, wife, and other adult household members help out with farming activities. Farming was reported as the main occupation by 82 percent of men respondents and 54 percent of women respondents.The average agricultural land owned or operated by the sample households within 12 months prior to the survey was 6.94 acres, ranging from 0.15 to 45 acres (Table 3). Among irrigation households, the average agricultural land owned or operated was 8.31 acres. Twenty-one percent of the nonirrigation households owned and cultivated some land in nonirrigated areas (upland areas). Among the nonirrigation households, the average land owned and operated was 0.68 acres. In terms of the type of documents for land ownership, 92 percent of households had some kind of legal documents (Table 6). Eighty-nine percent held a Form 7 (also called land use certificate), which gives farmers the legitimacy to sell or transfer their land or use their land as collateral for accessing credit. Aside from the farming, households in the irrigation catchment areas rely on other sources of income, including wage or salary employment, nonfarm enterprises, and remittances, Forty percent of sample households were engaged in wage or salary employment (Table 3). Thirty-three percent of irrigation households and 72 percent of nonirrigation households were engaged in wage or salary employment. The communities are roughly 4-5 kilometers or 20-30 minutes by motorbike away from the nearest town center, where household members work in wage or salary employment. Twenty-one percent of households had nonfarm enterprises they were managing or were engaged in. Twenty percent of irrigation households and 25 percent of nonirrigation households were engaged in nonfarm enterprises. The most common of these nonfarm enterprises were trading, dry good shops, and crafts or artisan activities.Men respondents more often reported farming as their main occupation as compared to women respondents (82 versus 54 percent) (Table 5). Nearly a third of female respondents (31 percent) but only 3 percent of male respondents reported homemaker (not engaged in any farm or nonfarm employment or enterprise) as main their employment. There is no difference in the proportion of women and men respondents reporting farm and nonfarm employment and own nonfarm enterprises as their main occupation: 5-6 percent of men and women respondents reported agricultural labor as their main occupation, and 9-10 percent of men and women respondents reported nonfarm employment or own enterprises as their main occupation (Table 5).Among irrigation households, the vast majority (93 percent) had farming as the main occupation of men respondents. Sixty-five percent of the wives also had farming as their main occupation, while 28 percent were homemakers and the remaining 7 percent had other main jobs (mainly small household nonfarm enterprise). Half of the women in irrigation households were engaged and helped out in farming.Among the nonirrigation households, 26 percent of men respondents had nonirrigation upland acreage and had farming as their main occupation, whereas 26 percent of them worked as daily farm labor. Among the women respondents, 13 percent had farming as their main occupation, 25 percent worked as farm labor, and the majority of them are homemakers.The study reveals some migration from the study sites. Thirty percent of households had some relatives working in other places and were receiving remittances from these relatives. Twenty-eight percent of irrigation households and 38 percent of nonirrigation households received remittances. Some households also received transfers (cash or in-kind) from government or nongovernmental organizations (NGOs).Eleven percent of irrigation households and four percent of nonirrigation households received transfers (Table 3).Livestock raising is not common in the study sites, except for draft animals (Table 7). Many farmers do not raise or own any more animals than are necessary to complete farming tasks. This section describes the production, cropping patterns, and commercialization of households that cultivated some crops in any of the growing seasons during the 12 months prior to the survey. Of the total agricultural land owned and operated reported by the sample households, 52 percent was cultivated during the post-monsoon season (October 2018 to mid-February 2019) (Table 8), 25 percent during the pre-monsoon season (mid-February to mid-May 2019) (Table 9), and 83 percent during the monsoon season (mid-June to late October 2019) (Table 10). During the 2018-19 post-monsoon season, chickpeas were the dominant crop, accounting for 78 percent of the cropland. During the 2019 premonsoon season, sesame and paddy were the main crops grown accounting for 47 percent and 43 percent of total acreage, respectively. During the 2019 monsoon season, paddy dominated the cropland area.  Cropping systems are quite diverse with different crops being grown across seasons (Table 11). Most farming households planted two or three different crops in the same parcel across different seasons, although monocropping (i.e., one crop per parcel of land) was practiced in 98 percent of all parcels in 2019. The average Simpson index of crop diversification (SID) was 0.457, which indicate relatively diverse cropping system across seasons. Fourteen percent of households with land cultivated only one crop per year (thus SID=0), about of households had SID less than 0.5, and the others had SID of at least 0.5. Crop intensity, measured as the sum of crop areas planted in wet, winter and dry seasons divided by the net area equipped by irrigation and drainage, was 1.4 in 2019. Fifty-four percent of irrigation HHs cultivated only for one cropping season (i.e., rainy or monsoon season), whereas the rest of the irrigation HHs cultivated two or three cropping seasons. On average, of the total 14.2 acres of gross area planted by an irrigation household, 8.1 acres was irrigated, which was about 57 percent. During postmonsoon season, 8 percent of cultivated parcels of the sample irrigation households were irrigated; monsoon season, 76 percent; and pre-monsoon, 71 percent.In terms of productivity, we asked for the quantity of harvest for each farming household in its biggest rice plot and biggest nonrice plot. Table 12 shows the harvested quantity in kilograms per hectare of land harvested for paddy and the most common crops grown per season. The average paddy yield during monsoon was 3.2 tons per hectare (t/ha); the average paddy yield during pre-monsoon season was 3.6 t/ha. The average yield of green gram was 0.6 t/ha. These averages were all comparable to the baseline estimates in the ADSP Project Appraisal Document (PAD) (World Bank 2015). The average yield of sesame was 0.23-0.27 t/ha, which was closer to a bad year harvest (0.11 t/ha) than a good year harvest (0.65 t/ha) described in Oo (2018). The figures in Table 12 also show wide variability in the crop yields and low levels achieved many farmers. These indicate the need to understand why there is wide variability and low yields achieved by many farmers in order to devise strategies to reduce this variability and improve the yields of many farmers with currently low levels. We also asked respondents roughly how much they harvested per crop across all parcels instead of the biggest parcel as in Table 12. Rice production was 2.9 t/ha across all parcels (Table 13), which was slightly lower than the biggest rice parcel mentioned above. Chickpea production was 0.68 t/ha, which was also lower than the biggest chickpea parcel above. The biggest difference was with sesameproduction was at 23 kg/ha across all parcels compared to more than 200 kg/ha in the biggest parcel.Being described as \"the gambling crop\" (Laung-ka-sar-thee-nan), sesame's yields are volatile and sensitive to weather changes. Therefore, sesame farming is considered risky. Compared to rice, agronomic researchers and policymakers have paid little attention to sesame and provided scant support to grow markets or promote value-added investments (Proximity Designs and Studio D Radiodurans 2019). Nonetheless, sesame is the second largest staple crop grown in Myanmar, both in terms of acreage and its estimated 500,000 farmers (Proximity Designs and Studio D Radiodurans 2019). While some product is directed for domestic consumption as oil and snacks, the country is now the sixth-largest exporter in the global market, with nearby China, South Korea, and Japan being the primary destination for 76 percent of overall product (Proximity Designs and Studio D Radiodurans 2019). More attention can be given to sesame. Most producing households sell their harvest (Table 14). For example, 72 percent of households that grew rice in the last 12 months have sold their harvest. On average, a farmer sold about 28 percent of the total harvested rice. For chickpeas, a farmer sold about 59 percent of total harvest on average. For sesame, a farmer sold about 51 percent of total harvest on average. The pro-WEAI includes 11 indicators measured across three domains of empowerment (Table 15). The overall proportion of respondents achieving empowerment, as defined by the 72 percent threshold (adequacy in 8 out of 11 indicators), was low for both women and men. Only 47 percent of women and 75 percent of men achieved empowerment. Fifty-three percent of women and 25 percent of men were not empowered. Among women, fewer women in dual-adult households achieved adequacy than women in women-adult-only households in almost all indicators, except attitude about IPV and respect among household members (Annex Figure 2).Another component of the pro-WEAI is the GPI, which reflects women's achievements relative to men in the same household. Only 55 percent of all dual-adult households achieved gender parity (and 45 percent have gender disparity in empowerment). The average empowerment gap, or the average percentage difference in empowerment between men and women in households that did not achieve gender parity, was 0.21. Both the 3DE score and the GPI score are used to calculate the pro-WEAI score, which is 0.78 for the total sample. Figure 5 shows the absolute contributions of each of the 11 indicators to disempowerment. Inadequacy in attitudes about IPV and lack of participation on groups and in influential groups are the top contributors to disempowerment for women and men. These inadequacies were significantly worse for women. Most women and men had high tolerance and acceptance of IPV against women. Only 8 percent of women and 19 percent of men believed that husbands are not justified in hitting their wives regardless of circumstances, based on responses to six scenarios, similar to the Myanmar Demographic and Health Survey, narrated to the respondents. More details are in Annex Figures 1 and2.Lack of membership in groups, especially in influential groups, was the second top contributor for women and men (Figure 5). Other leading contributors to disempowerment for women were lack of access to and decisions on financial services, lack of work balance, and lack of intrahousehold harmony or respect among household members. Table 16 shows the proportion of women and men respondents achieving adequacies in 11 pro-WEAI indicators. In all indicators except control over use of income, autonomy in income, and ownership of land and other assets, fewer women achieved adequacies. The largest differences in adequacies between women and men were in indicators of collective agency, particularly group membership (34 percentage points [pp], with men more likely to achieve adequacy) and membership in influential groups (33 pp difference). The next-largest differences in adequacies between women and men were in indicators of instrument agency, in particular, access to and decisions on financial services (13 pp difference). The difference between men and women in inputs to productive decisions is also relatively large (9 pp difference). Among the intrinsic agency indicators, attitudes about IPV against women show the largest difference in adequacy between women and men (11 pp difference, with men more likely to achieve adequacy). Women were more tolerant and accepting than men of IPV against women. Women and men in irrigation and nonirrigation households were generally similar in terms of empowerment and main contributors of empowerment (Table 17), with a few exceptions. First, tolerance and acceptance of IPV are worse among both women and men in other households than in irrigation households. Second, access to and decisions on financial services was much worse for both women and men in other households than in irrigation households. This difference results from the availability of credit from the Myanmar Agricultural Development Bank (MADB) given to rice farmers but not available to other households. Third, work balance was worse for women in the other households than for women in irrigation households, and the same for men in both the irrigation and other households. Work balance for women in both irrigation and other households was worse than for men in these households. Fourth, there was no difference in group membership between women in irrigation and in other households, whereas group membership of men in other households was much lower than of men in irrigation households.  18 shows strong sense of self-efficacy among both women and men respondents. Almost all respondents strongly agreed or agreed with statements indicating self-efficacy (#1-3). A majority of respondents also agreed or strongly agreed with statements in favor of gender equality. For example, almost all women and men respondents strongly agreed with the statements that women should be encouraged and supported to be leaders or entrepreneurs. Similarly, nearly all women and men respondents strongly agreed that husbands should help wives with household chores like cooking and taking care of children. Disagreement did arise about whether women can own land or assets or have their names in the title. This disagreement mirrors the strong culture in Myanmar that land and assets should be in the name of the head, which is often the man in the household. For gender sensitization, these gender norms on equality in land and asset ownership could be emphasized. Source: IFPRI/World Bank/MSR 2020. Respondents were read the following statements and were asking to say whether they SD=strongly disagree, D=disagree, N=neither agree or disagree, A=agree, or SA=strongly agree on each of those statements.Table 19 shows the level of engagement of respondents in community affairs and their level of comfort about speaking in public. Nineteen percent of men and 46 percent of women said they were not comfortable speaking in public. Moreover, a majority of women and men did not feel they had influence in what was happening in the community. Fifty-four percent of men and 67 percent of women felt they could not change anything in the community if they really wanted to. Building self-confidence, negotiation skills, and leadership skills should therefore also be emphasized during sensitization efforts, especially for women. Source: IFPRI/World Bank/MSR 2020.This section describes respondents' scores on ( 1) nutrition knowledge and exposure; (2) health practices, including handwashing practices, and hygiene and sanitation practices; and (3) dietary diversity.Community health services in the study sites are available. Each village has its own community health center, with at least two community health officers. Almost all young children (6-59 months) in the sample households were vaccinated and received vitamins and deworming medicine, mainly from the community health centers.We asked selected questions indicating knowledge, exposure, and practice related to the four main themes of LEARN: nutrition basics, family nutrition, access to nutritious foods, and food preparation (LEARN 2015). Table 20 shows the percentage of respondents who answered correctly according to the good nutrition-related practices in each of these themes. Higher figures mean better nutrition exposure and knowledge. Women and men respondents scored high on three themes-food preparation, access to nutritious foods, and nutrition basics-but scored very low in family nutrition. The themes that the majority of respondents got wrong, therefore indicating weak nutrition knowledge, are the following:• Rice porridge (tamingazi) is NOT the only solid food that children aged six months to years old need to be healthy. Children six months to two years old need diverse foods, such as egg, soft meat, fish, fruits, and vegetables, to be healthy. • Breastfeeding women should eat beans, legumes, vegetables, fruits, eggs, and meats because diverse foods are good for the health of the mother and child. • Many vegetables and fruits gathered from the wild can be as healthy as foods that are bought.  In general, a high proportion of women and men respondents practiced frequent handwashing in the number of scenarios presented to them (Table 21). The survey found no difference in handwashing practices between irrigation and nonirrigation households but did find differences in handwashing practices between women and men. Fewer men than women practiced good handwashing practices with water and soap than. Targeting men with health and nutrition messaging can help in this regard. Moreover, the frequency of handwashing, particularly after handling animals, and with the use of both water and soap, can be further improved.  22). Past water, sanitation, and health projects had built boreholes and tubewells in the communities. Other households had their own boreholes or tubewells. Most households also had an improved sanitation facility not shared with other households, although 16 percent of irrigation households and 27 percent of nonirrigation households share a sanitation facility with other households or use an unimproved sanitation facility. Overall, the majority of households in the focus communities have improved water sources, handwashing areas, and improved sanitation; and they practiced good handwashing and other hygiene and sanitation practices. The areas that need further improvement and to be included in health and nutrition messaging are (1) treating drinking water, for example by boiling; (2) proper garbage disposal, for example not burning garbage or throwing it into rivers, streets ,or open spaces; and (3) ensuring that no domestic animals enter the kitchen or food preparation areas (Table 22). Number of sample households 998 818 180 Source: IFPRI/World Bank/MSR 2020. Note: Water source pertains to cooking water. Improved water sources include piped water, boreholes or tubewells, protected dug wells, protected springs, rainwater, and packaged or delivered water. Improved sanitation facilities include flush toilets, ventilated improved pit latrines, composting toilets, and pit latrines with slabs. Water treatment methods include boiling, chemical treatment, filtration, and solar disinfection. HH=households.The MDD-W indicator, a commonly used indicator of diet quality and a good proxy for nutrition, is low at 4.71 out of 10 (Table 23), which falls below the MDD-W of threshold of 5. No significant difference was found in dietary diversity between women and men. More women than men ate nuts and seeds, vitamin A-rich fruits and vegetables, and other fruits on a daily basis. Women and men in irrigation households had higher dietary diversity scores than those in nonirrigation households. More women and men in irrigation households ate nuts and seeds, meat and fish, and vitamin A-rich fruits and vegetables on a daily basis than those in nonirrigation households.We looked closely at consumption of dairy, a good source of calcium; however, in the study area (as in the whole of Myanmar), only 1 percent of sample individual adults consume dairy on a daily basis (Table 23). Promoting more dairy consumption in the study area would be difficult for several reasons:1. Lactose intolerance is common in much of Asia and could also be an issue in Myanmar. Although young children can often tolerate dairy, lactose intolerance is common in adulthood. 2. Much of the dairy that is available is not especially helpful. Sweetened condensed and evaporated milk brands are commonly available but can contain palm oil and sugar, which are both health hazards. Quite a few drinks with dairy are available, but they contain a low percentage of dairy and/or a high percentage of sugar. Nondairy creamer is readily available but contains no calcium. 3. In the study by Mahrt et al. (2019) on the cost of the recommended diet, the authors found that consuming the recommended amount of dairy (about 1 cup) would cost approximately 20 percent of the total cost of a healthy diet-similar to the cost of staples and the cheapest meat/egg/fish sources. Meeting this cost would require a major adjustment in spending habits. 4. A more realistic approach to improving calcium intake given existing consumption patterns, particularly given the short time frame, might be to promote the consumption of small fish with head and bones as well as dark leafy greens. Small fish eaten whole, dried fish, and fish-based products are good alternatives, culturally appropriate, and of high bioavailability (like milk). Along with low dairy consumption, the survey found that nuts and seeds, eggs, vitamin A-rich fruits and vegetables, and other fruits are not consumed on a daily basis by three-quarters of the respondents (Table 23). Even beans and dark leafy vegetables, which are relatively abundant in the study context, are consumed by only 52-62 percent of the respondents on a daily basis. Consumption of these food groups should be emphasized during nutrition education.About a third of respondents received information on farming rice and other crops, and about half received information on nutrition, health, or sanitation (Table 24). Fourteen percent of men and women respondents had attended training on rice farming, and 11 percent had attended training on irrigation management. Men were more likely than women to receive agriculture and nutrition-related information and training-31 percent of women versus 49 percent of men received agricultural information and 9 percent of women versus 20 percent of men attended agricultural training. The main sources of information on farming rice and other crops were another farmer or a neighbor, reported by 43-48 percent of the men and women respondents (Table 25). Other main sources were agricultural seminars or trainings, TV, and input companies, reported by 13-25 percent of the respondents. Other sources were township agriculture offices, radio messaging, the Internet or social media, extension workers, and brokers or agents, reported by 3-10 percent of the respondents. Women were less likely than men to receive agriculture-related information from all sources, except from other farmers/neighbors and brokers/agents. The largest gaps were in receiving information from township agriculture offices, Internet/social media, radio, TV, and agricultural trainings. The main sources of nutrition-related information were extension workers and TV, reported by 35-39 percent of men and women respondents. The second main source was another farmer or neighbor, reported by 21 percent of respondents. The other main sources were radio messaging; health talks, particularly those with the Department of Health; and Internet or social media. Women were more likely than men to receive nutrition-related information from another farmer or neighbor and extension workers. Men were more likely to receive nutrition-related information from radio, TV, and Internet or social media.Most respondents said they followed or acted upon the information received (Table 25). This finding is consistent with responses on training lessons being followed and adopted by more than 90 percent of training participants (Table 26). Of those not following received advice, the most common reasons were that (1) the information was not relevant to their farming, enterprise, or growing season; (2) they did not trust the information; (3) they did not understand the information; and (4) they did not have enough resources and spare time to implement training lessons (Table 27). Only about half of training participants, however, said they had learned something new from the training attended. Most respondents who attended training shared the information received with their spouse or other household members (Table 26).  Almost all respondents owned a smartphone and had access to a cell phone network (Table 28). Only a quarter, however, reported having tried to access the Internet via cell phone. Sixty-six percent of respondents had heard of agricultural information available through Internet or phone. Twentytwo percent participated on social media, particularly Facebook. Women respondents had significantly less access to these technologies than men respondents did. The largest gaps were in accessing the Internet via cell phone, social media participation, and having heard of agricultural information available through Internet or phone. Forty percent of the men and women respondents were members of a WUG (Table 29). More men were members of WUGs than women. For those not members, 40 percent said they wanted to become members. Significantly more men than women would like to be a member of a WUG. Respondents had generally positive perceptions on the usefulness of WUGs. Ninety-two percent of respondents thought that WUGs are either useful or very useful, and 8 percent thought they are not useful. The main topics during WUG meetings were water usage or drainage/tunnel formation or repair, the WUG itself, and agriculture or agricultural practices in general (see annex tables 9 and 10). Twenty-nine percent did not know or remember the topic of the meeting. Most respondents said they found the meetings useful and that they learned about irrigation systems, how to reduce water wastage, how to use dam water systematically, the importance of water dam repair, fairness and unity in accessing water, and about agriculture and improving practices.This report provides a baseline study on the status of gender equality, crop diversification, and nutrition of a sample of 998 households and 1,835 men and women in two pilot irrigation catchment areas in the Myanmar Agricultural Development Support Project. The main findings are summarized and the major learning outcomes are discussed below.1 Women were less likely than men to receive agriculture-and nutrition-related information and training. Women were also less likely to access information and communication technologies, such has smartphones, the Internet, and social media. Extension and training programs should involve both men and women rather than targeting only household heads, who are often men. A more commonly used channel for reaching most women with agricultural information was through other farmers or neighbors. Men and women had different channels for accessing nutrition-related information. Women were more likely to receive nutrition-related information from another farmer or a neighbor and extension workers, whereas men were more likely to receive nutrition-related information from radio, TV, and the Internet or social media.Results show large gender gaps, in which fewer women than men achieved adequacy in all 11 indicators of empowerment. A total of 53 percent of women versus 25 percent of men respondents were not empowered; and 45 percent of dual-adult households have gender disparities. Strategies to empower women and reduce gender gaps in the irrigation catchment areas could be strengthened in ADSP activities.The main contributors to disempowerment among women and men were attitudes about intimate partner violence and lack of group and influential membership. Women and men had high tolerance for and accepting attitudes about intimate partner violence against women. Gender-transformative approaches that emphasize zero tolerance of domestic violence should be promoted in the households and communities. Both men and women within the households should be targeted in these gendertransformative approaches.Fewer women achieved adequacy in membership in groups and influential groups than men. Women were less likely than men to be members of water user groups (WUGs) and to participate in WUG meetings and activities. Women were also less likely than men to be part of any group or association. Greater participation of women should be promoted in groups and organizations. Women-only groups should be promoted, as well as women's participation and leadership in mixed-gender groups. A good starting point is by enhancing women's confidence and organizational management and leadership skills. Skills development trainings should be provided in ways that make them accessible and convenient for women, without worsening women's time burden. Strategies to provide childcare services during trainings and other activities could be explored.Other leading contributors to disempowerment for women were lack of access to and decisions on financial services, lack of work balance, and lack of intrahousehold harmony or respect among household members. More women and men in other households did not achieve adequacy in access to and decisions on financial services than those in irrigation households. More men in other households did not achieve adequacy in group and influential group membership than men in irrigation households. More women in other households did not achieve adequacy in work balance than women in irrigation households.Lack of work balance for men and women, especially for women, is a major issue. Labor-saving technologies in farming can help lessen the time burden for both women and men. Gender balance in childcare and family responsibilities, greater support by fathers, and shared household responsibilities can be promoted through gender awareness activities in the households and communities.The priority food groups that should be promoted in the study area and in future nutrition education are (1) nuts and seeds, (2) eggs, (3) vitamin A-rich fruits and vegetables, (4) dark leafy vegetables, (5) other fruits, (6) beans, and (7) small fish with head and bones. Almost no one in the study group consumes dairy because of its limited availability and cost, people's preference, and in some cases lactose tolerance. A realistic approach to improving calcium intake given the existing consumption patterns and the short SUN project time frame, is to promote the consumption of small fish with head and bones as well as dark leafy greens. A diet with a greater diversity of food groups must be emphasized in the nutrition messaging, as the consumption of more food groups is associated with a greater likelihood of meeting nutritional needs. With development community fears of a potential health crisis following the health and economic crises from the COVID-19 pandemic, there is a critical need to focus on emphasizing the importance of diet quality, sanitation, and health.The majority of respondents incorrectly answered questions on the following themes, therefore indicating weak nutrition knowledge. The following are examples of themes that should be promoted in the nutrition messaging, along with the other messages in the LEARN module:• Rice porridge (tamingazi) is NOT the only solid food that children aged six months to two years old need to be healthy. Children aged six months to two years old need diverse foods, such as eggs, soft meat, fish, fruits, and vegetables, to be healthy. • Breastfeeding women should eat beans, legumes, vegetables, fruits, eggs, and meats because diverse foods are good for the health of the mother and child. • Many vegetables and fruits gathered from the wild can be as healthy as foods that are bought. (3) ensuring that no domestic animals enter the kitchen or food preparation areas; and (4) more frequent handwashing using soap and water, particularly after handling animals. Fewer men than women were practicing good handwashing practices with water and soap. Targeting men with health and nutrition messaging can help in this regard.Farmers were growing diverse crops in their farms. More diverse crops can be grown for several cropping seasons once the irrigation facility is improved. Productivity of rice and chickpeas in the past three seasons were comparable to the baseline estimates by ADSP; and the productivity of sesame is very low. The yield levels show wide variability across farmers. The current levels suggest the need for a better understanding of why there is wide variability and low yields achieved by many farmers in order to devise strategies improve the situation.This paper has provided rich baseline data and descriptive analysis on the SUN study site. Further analysis are underway that analyzes in-depth the drivers of increased incomes, nutrition, women's empowerment, and gender equality in these focus areas. Finally, an intervention around gender and nutrition behavioral change communication that incorporates the lessons from this baseline paper is being designed and soon to be piloted. ","tokenCount":"8989","images":["-993276048_1_1.png","-993276048_1_2.png","-993276048_1_3.png","-993276048_1_4.png","-993276048_10_1.png","-993276048_11_1.png","-993276048_11_2.png","-993276048_15_1.png","-993276048_27_1.png","-993276048_53_1.png","-993276048_54_1.png","-993276048_54_2.png","-993276048_55_1.png","-993276048_55_2.png","-993276048_56_1.png","-993276048_56_2.png","-993276048_57_1.png","-993276048_57_2.png"],"tables":["-993276048_1_1.json","-993276048_2_1.json","-993276048_3_1.json","-993276048_4_1.json","-993276048_5_1.json","-993276048_6_1.json","-993276048_7_1.json","-993276048_8_1.json","-993276048_9_1.json","-993276048_10_1.json","-993276048_11_1.json","-993276048_12_1.json","-993276048_13_1.json","-993276048_14_1.json","-993276048_15_1.json","-993276048_16_1.json","-993276048_17_1.json","-993276048_18_1.json","-993276048_19_1.json","-993276048_20_1.json","-993276048_21_1.json","-993276048_22_1.json","-993276048_23_1.json","-993276048_24_1.json","-993276048_25_1.json","-993276048_26_1.json","-993276048_27_1.json","-993276048_28_1.json","-993276048_29_1.json","-993276048_30_1.json","-993276048_31_1.json","-993276048_32_1.json","-993276048_33_1.json","-993276048_34_1.json","-993276048_35_1.json","-993276048_36_1.json","-993276048_37_1.json","-993276048_38_1.json","-993276048_39_1.json","-993276048_40_1.json","-993276048_41_1.json","-993276048_42_1.json","-993276048_43_1.json","-993276048_44_1.json","-993276048_45_1.json","-993276048_46_1.json","-993276048_47_1.json","-993276048_48_1.json","-993276048_49_1.json","-993276048_50_1.json","-993276048_51_1.json","-993276048_52_1.json","-993276048_53_1.json","-993276048_54_1.json","-993276048_55_1.json","-993276048_56_1.json","-993276048_57_1.json","-993276048_58_1.json"]}

data/part_2/0182735851.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"11ecd4b929e1f68b5cac445ddcf4788b","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/2945fedd-b7fb-4838-b730-67fc9ca09f8a/retrieve","description":"Weather shocks and natural disasters, it has been argued, represent a major threat to national and international security. Our paper contributes to the emerging micro-level strand of the literature on the link between local variations in weather shocks and conflict by focusing on a pixel-level analysis for North and South Sudan at different geographical and time scales between 1997 and 2009. Temperature anomalies are found to strongly affect the risk of conflict. In the future the risk is expected to magnify in a range of 21 to 30 percent under a median scenario, taking into account uncertainties in both the climate projection and the estimate of the response of violence to temperature variations. Extreme temperature shocks are found to strongly affect the likelihood of violence as well, but the predictive power is hindered by substantial uncertainty. Our paper also sheds light on the vulnerability of areas with particular biophysical characteristics or with vulnerable populations.","id":"-1340965979"},"keywords":["weather shocks","violent conflict","vulnerability","Sudan"],"sieverID":"4ee42814-dde7-415c-a3ea-756622a702c4","pagecount":"78","content":"established in 1975, provides evidence-based policy solutions to sustainably end hunger and malnutrition and reduce poverty. The Institute conducts research, communicates results, optimizes partnerships, and builds capacity to ensure sustainable food production, promote healthy food systems, improve markets and trade, transform agriculture, build resilience, and strengthen institutions and governance. Gender is considered in all of the Institute's work. IFPRI collaborates with partners around the world, including development implementers, public institutions, the private sector, and farmers' organizations, to ensure that local, national, regional, and global food policies are based on evidence. IFPRI is a member of the CGIAR Consortium.Climate change and natural disasters, it has been argued, represent a major threat to national and international security by increasing resource scarcity and competition and inducing health problems (Homer-Dixon 1994, 2007;Sachs 2005;Steinbruner, Stern, and Husbands 2012).1 So far, these claims lack consensual empirical support and would deserve a more careful investigation of the specific channels linking climatic phenomena and conflict events (Salehyan 2008;Scheffran et al. 2012;Steinbruner, Stern, and Husbands 2012). Quantitative assessment of the climate-conflict nexus has largely been initiated by Miguel, Satyanath, and Sergenti (2004), who seminally used rainfall shocks on income growth to assess how the risk of conflict may increase in Africa south of the Sahara (SSA) when the opportunity cost to fight decreases. 2 Since then, Burke et al. (2009) found that temperature variations increase the risk of conflict in SSA and, interestingly, temperature is the only significant climatic variable when included in the model along with rainfall. They suggest that earlier findings, including Miguel, Satyanath, and Sergenti (2004), about increased conflict due to lack of rainfall, might have been partly capturing the effect of higher temperature. 3 The importance of temperature is also in line with the results by Zhang et al. (2007Zhang et al. ( , 2011) ) showing that temperature variations are correlated with the frequency of wars in Europe and China in the preindustrial period. Offering an alternative approach to the one based on weather variations, Hsiang, Meng, and Cane (2011) further investigated the relationship between climate change and global patterns of civil conflicts. They exploited the dominant interannual mode of the modern climate, the El Niño-Southern Oscillation, to show that conflict is more likely during El Niño years (warmer and dryer in the continental tropics) relative to La Niña years.A recent paper by Klomp and Bulte (2012) revisits these cross-country analyses through a battery of robustness checks and finds little evidence linking global and local weather shocks and conflict. In line with the general review on the conflict literature by Blattman and Miguel (2010), Klomp and Bulte (2012, 26) call for moving beyond the conventional country-year focus and embracing shorter time intervals and subnational regions. The lack of robustness in previous cross-country findings may indeed result from the inability of country-level variables to capture the dynamics of local conflict events (for a discussion, see Buhaug andLujala 2005 andBuhaug andRod 2006). Using subnational units of analysis would allow us to overcome this shortcoming while preserving the robustness of the econometric approach recently advanced by scholars looking at the links between climatic variations and violence in SSA. 4 On the one hand, Harari and La Ferrara (2012) exploit the grid-cell-level (1 degree over 1 degree) annual variation to study the relationship between weather shocks and conflict in Africa. They show that negative weather shocks (proxied by a new drought index), occurring during the growing season of the main crops, significantly increase the incidence of conflict. On the other hand, Raleigh and Kniveton (2012) and O'Loughlin et al. (2012) propose regional analyses focused on East Africa. Based on geo-referenced (2.5 degree) data, Raleigh and Kniveton (2012) assess the link between rainfall anomalies and conflict events in Uganda, Ethiopia, and Kenya and find that the frequency of violent events increases in periods of extreme rainfall variations. O'Loughlin et al. (2012) evaluate the role not only of precipitation, but also of temperature changes at a finer resolution (grid of 1 degree) using similar data and including nine entire countries (Burundi, Djibouti, Eritrea, Ethiopia, Kenya, Rwanda, Somalia, Tanzania, and Uganda). They show that wetter deviations from the precipitation norms decrease the risk of conflict, whereas warmer than normal temperature raises the risk.Our paper is complementary to such studies, in particular the above-referenced ones about East Africa, even if they do not specifically include Sudan in their research area. But beyond the exclusive focus on Sudan and the finer grid-cell resolution (0.5 degree)5 , our paper differs by adopting a more restrictive methodological approach. Our main results are estimated using a fixed-effects framework at both pixel and quarterly levels. This framework not only isolates the impact of climatic variations from specific control variables-such as the population, the distance to the nearest urban center (Raleigh and Kniveton 2012), the crop production index, and the infant mortality rate (O'Loughlin et al. 2012)-but also from all the other characteristics that are time constant at the grid-cell level. Although we will discuss the way our results change when a different framework is chosen, we consider an approach based on celllevel fixed effects more likely to reduce estimation bias compared to an approach based on adding a limited number of potentially endogenous control variables (Angrist and Pischke 2009). In this respect, our methodology is closer to Maystadt, Eckers, and Mabiso (2013), who, using monthly and regional variations and focusing on the role of the livestock market as a channel of transmission, found a strong relationship between more frequent and intense temperature-based droughts and the occurrence of violent conflict in Somalia.Furthermore, our paper contributes to the spatially disaggregated strand of the literature on the links between weather shocks and conflict by assessing the validity of the analysis at different geographical and time scales and by reviewing a series of proxies for weather shocks previously used in the literature in an inconsistent way. More precisely, we estimate the relationship between local warming and violent conflict between 1997 and 2009 in North and South Sudan, controlling for time dummies, grid-cell fixed effects, and area-specific time trends. Following Burke et al. (2012), we then project the changes in violent conflict by 2030 under different climate models and scenarios and show that the uncertainty in the projections increases when the extreme nature of temperature shocks is captured. The available data do not allow us to test the different channels linking weather shocks to conflict, but we assess how some characteristics magnify or reduce the strength of such relationship. Therefore, our discussion about the heterogeneous effects sheds light on the vulnerability of areas with particular biophysical characteristics or with vulnerable populations.Sudan is known for having experienced two civil wars after independence in 1956, but it actually has a long-lasting history of repeated conflict events starting well before independence. Like many African conflicts, the Sudan conflict took its roots in the colonization period. 6 Most scholars agree that the divide between the north and the south was fueled by the British colonizers who favored social and economic investment in the north under the so-called Southern Policy implemented between 1920 and 1947 (Ali, Elbadawi, and El-Bathani 2005). After independence, such structural divide was exacerbated by the northern elite that came into power and led to 17 years of civil war (known as the first civil war) between the north and the south. A peace settlement, the Addis Ababa Peace Accord, was reached in 1972, but the then president, Nimeri, aggravated grievances in the south by redesigning the border to include oilproducing areas in the northern territory, by grabbing land through the development of mechanized farming, and by exploiting the divisions between various groups within the south. As a result, the Sudan People's Liberation Army (SPLA) was created in 1983 with external support from Ethiopia. The second Sudanese civil war was then triggered as a continuation of the first civil war and lasted until 2005, when it ended with the signature of the Comprehensive Peace Agreement that paved the way for a referendum in January 2011 and for the independence of South Sudan in July 2011. Although the exact figures are a subject of debate (Duffield 2001), the dramatic history of violence in Sudan resulted in more than 1.9 million civilian deaths between 1983 and 1998 (more than 600,000 since 1993, according to Burr 1998) and about 5 million displaced people (United Nations Environment Programme, UNEP 2007).Behind this national scene and the description of the civil war as an opposition between the north and the south, local conflict events also multiplied within North and South Sudan (Johnson 2011). The exploitation of resources, once the source of warfare financing, became a warfare objective in itself. 7 At the same time, conflict events evolved from ethnic tensions between the north and the south to local or regional conflicts increasingly reported to be linked to environmental factors. The study by UNEP (2007, 70) was certainly instrumental in maintaining that \"competition over declining natural resource was one of the underlying causes of the conflict\" and in pointing to four specific conflict-contributing categories of natural resources: \"oil and gas reserves, Nile waters, hardwood timbers, rangeland, and rain-fed agricultural land (and associated water points).\" In particular, in marginalized areas, conflict was intensified by the expansion of large semimechanized farms and the subsequent loss of access to land for both smallholders and pastoralists (Keen and Lee 2007). Keen and Lee (2007, 17), for example, reported that the area of land taken up by rain-fed semimechanized agriculture increased from about 2 million feddans (that is, about 0.84 million hectares) at the beginning of the 1970s to 14 million feddans (that is, about 6 million hectares) by 2003.In addition, pastoralist and agropastoralist communities have been increasingly under pressure by the combination of population growth and more frequent and intense droughts. In Sudan, agriculturethat accounted for 30-40 percent of GDP between 1996 and 2010 (Benke 2012)-remains extremely vulnerable to droughts, whereas the climatic conditions appear to have become harsher to cope with. According to UNEP (2007), an estimated 50-to 200-kilometer southward shift of the boundary between desert and semidesert has occurred since the 1930s, and the remaining semidesert and low rainfall land are at considerable risk of further desertification. Thus, the vulnerability of semiarid areas to climatic stresses and shocks is more likely to intensify in the decades to come.However, the link between resource scarcity and conflict is far from being trivial. Scholars and policymakers have equalized resource scarcity to an incentive for conflict (Homer-Dixon 1994), especially for Sudan and pastoralist communities (UNEP 2007;Hendrickson, Armon, and Mearn 1996), but detrimental weather shocks may also reduce the value of the resources that are fought over. In particular, Butler (2007) and Kevane and Gray (2008) argued that weather patterns only weakly corroborated the claim that climate change caused the Darfur conflict and concluded that the United Nations overestimated the case. Certainly, there is still a need to understand which conditions make the link between resource scarcity and conflict hold in one direction or another. That is the main objective of our empirical analysis.We combine climatic and conflict data for each 0.5 degree grid-cell (i) of Sudan and for each quarter (t) from 1997 until 2009 to examine the relationship between weather shocks (Weather i,t ) and conflict occurrence (Conflict i,t ). Accordingly, we estimate the following baseline equation:The dependent variable, Conflict i,t , is given by the quarterly sum of violent conflict events by grid-cell (i). Our main variable of interest, Weather i,t , seeks to capture weather deviations and extreme events at the grid-cell (i) and quarter (t) levels. Since there is no consensus, yet, on the best way to assess the impact of weather shocks on socioeconomic outcomes, we propose a series of proxies most likely to capture deviations from normal conditions and the nonlinearity induced by extreme events. First, we apply the anomaly transformation: precipitation and temperature quarterly data are transformed into anomalies, that is, deviations from the long-term quarterly mean, divided by the long-run quarterly standard deviation.8 Such anomaly transformation has become standard and is frequently adopted in the economic literature (for example, Maccini and Yang 2009;Barrios, Bertinelli, and Strobl 2010;Marchiori, Maystadt, and Schumacher 2012;Dell, Jones, and Olken 2012;Harari and La Ferrara 2012). In addition, we introduce the quadratic term of the weather anomalies as a first indication of nonlinearity. Second, similar to Schlenker, Hanemann, and Fisher (2006); Schlenker and Roberts (2009); and Harari and La Ferrara (2012), we isolate the component of climate variability that is relevant for agriculture interacting the weather variables with an indicator identifying the growing period by state (De-Pauw and Wu 2012). 9 In particular, we try to capture extreme events that could lead to yield losses by defining a dummy for positive and negative deviations happening during the growing period above one (or two) standard deviation(s). Furthermore, we consider also a cell-specific threshold for extreme events and introduce a dummy equal to one for deviations below 15 (or 10 or 5) percent and above 85 (or 90 or 95) percent of the grid-cell-specific distribution in the growing period (as in Brückner 2010 andBurke, Gong, andJones 2011). Finally, to assess more accurately temperature shocks on agriculture, we follow the approach introduced by Schlenker, Hanemann, and Fisher (2006) that suggests exploiting daily data to compute degree-days transformation. Based on agronomist literature specifically for SSA, Schlenker and Lobell (2010) define a lower threshold at 10 degrees Celsius and a higher threshold at 30 degrees Celsius. The two variables are considered together to capture the nonlinearity of temperature shocks (see also Schlenker and Roberts 2009). The first variable, \"moderate degree-days,\" provides the sum of degreedays above the lower threshold of 10 degrees Celsius and below the upper threshold of 30 degrees Celsius. 10 The second variable, \"extreme degree-days,\" sums the number of degree-days above the upper threshold of 30 degrees Celsius. The two variables are expressed in degree-days per quarter (or the concerned period in the robustness checks) and then transformed into anomalies as in Dillon, Mueller, and Salau (2011). Similar to Schlenker, Hanemann, and Fisher (2006), these variables are interacted with our state-level indicator of the growing period, and a quadratic term is introduced to capture nonlinear effects. As a last robustness check, we follow the approach of Harari and La Ferrara (2012), and we compute for each grid-cell quarter the Standardized Precipitation-Evapotranspiration Index (SPEI), a multiscalar drought index that offers the advantage of being based on both precipitation and temperature.We estimate equation (1) with a linear least squares specification because nonlinear models with fixed effects yield inconsistent slope estimates due to the incidental parameter problem (King and Zeng 2001;Greene 2004). To be able to draw causal inferences, we introduce in the equation grid-cell fixed effects (α i ) and time dummies (φ t ). Therefore, we investigate how climate changes (compared to the pixel mean) affect the frequency of conflict events within each grid-cell (compared to the mean). In addition, we augment the specification by introducing a county-specific time trend (t i,t ) 11 and the night-lights density (X i,t ). The former is included to reduce the threat of spurious parallel trends, whereas the latter is used as a proxy to capture changes in economic activities potentially unrelated to climate. 12 However, we cannot reject the hypothesis that county-specific time trends' and night-lights density's get rid of interesting variations in the relationship between weather shocks and violent conflict, and consequently, we will show how our results change when we exclude such variables. 13  The introduction of grid-cell fixed effects is the main difference from the approach proposed by O' Loughlin et al. (2012), who prefer to introduce highly aggregated country dummies in their main estimations and who present a similar methodology just as a robustness check (O'Loughlin et al. 2012, Supporting Information, 5). Interestingly, when the authors use grid-cell fixed effects in place of country fixed effects (but without replacing yearly dummies by quarterly or monthly ones), precipitation anomalies do not affect the risk of conflict, whereas additional support is found for the role of hotter than usual temperatures in predicting greater conflict. We believe our approach is more likely to control for unobserved (time-constant) characteristics that may bias the estimated relationship between weather shocks and conflict at the grid-cell level. Nevertheless, we acknowledge that the use of fixed effects at a disaggregated level has recently been questioned by some scholars because the effects absorb most of the variation, making the identification rely on slight margins. Dealing with climatic variables, this might lead to the amplification of measurement errors, as Fisher et al. (2012) point out. Thus, we will show that our results are confirmed even when we use the more commonly preferred random-effects estimation with dummies at the state level (suggesting that our findings are not driven by measurement errors).Moreover, we cluster the standard errors at the county level to reduce potential problems generated by time and spatial dependency within Sudanese counties. In addition, bearing in mind the recent debate on the importance of explicitly modeling such dependency in the process itself (Harari and La Ferrara 2012), we will control for serial and spatial correlation by transforming equation (1) into a simple dynamic model (using the Arellano-Bover/Blundell-Bond estimator) and into a dynamic model with the spatial lags of the independent variables included. This latter model has the advantage of being straightforward since adding the spatial lags does not involve serious econometric problems. The pitfalls of the model are that it does not allow correcting the standard errors for clustering at the county level, it does not take into account the fact that spatial correlation might also be present directly in conflict itself through cross-cell spillovers, and it offers estimates affected by a simultaneity bias. As Harari and La Ferrara (2012, 12) remark, in the typical case of positive covariance of spatial lags and independent variables, we will overestimate the interdependence effect and underestimate the cell-specific effect of weather shocks. Thus, the estimates of such model represent a credible lower bound for the effects of weather shocks on conflict.11 There are 117 counties in North and South Sudan. 12 Several papers have shown that the importance of environmental variables may be downplayed by the inclusion of political and economic variables (Raleigh and Urdal 2007). The use of grid-cell fixed effects and time dummies already reduces the importance of time-constant political or economic factors and of those factors that would affect equally over time the units of observations. Adding night-lights, the best proxy of economic activity at the local level, offers a further robustness check. Nevertheless, we cannot exclude that night-lights are capturing relevant variation or act as a bad control (Angrist and Pischke 2009). We therefore exclude that control (along with the time trend) to provide an upper-bound limit of the effect of extreme weather shocks on conflict and confirm that our findings do not change. The results are also similar when the lagged value of night-lights density is used. 13 See Results and Robustness Checks section.Finally, although our empirical strategy relies on quarterly variations in climatic variables to assess their effects on conflict, in section 4, we also exploit time-invariant local characteristics to evaluate the heterogeneous effects of the climate shocks and identify mitigating and exacerbating factors.Data on conflict events come from the Armed Conflict Location and Event Dataset (ACLED) presented by Raleigh et al. (2010). 14 ACLED is the most recent, detailed, and widely used conflict dataset developed by the International Peace Research Institute of Oslo (PRIO). It specifies the exact location, the date, and other characteristics of conflict events based on news and reports within unstable states. Given its nature, it might be affected by selection in reporting, a drawback common to conflict datasets not based on surveys, but such reporting bias is not likely to be systematically correlated with our weather indicators and should not constitute a major problem for our identification strategy. Another drawback of these data is the lack of information about the number of causalities, but the monthly frequency of violent events should give us a fair approximation of the local intensity of conflict. We focus on violent conflict events, comprising battle, defined as \"a violent interaction between two politically organized armed groups at a particular time and location\", and violence against civilians (one-sided violence), defined as \"deliberate violent acts perpetrated by an organized political group, typically either a rebel or a government force, on an unarmed non-combatant\" (ACLED Codebook version 2, 8 and 11). 15 In North and South Sudan, the number of 2,497 violent events represents the overwhelming majority of events (97 percent) reported in the ACLED dataset. Although our results do not depend on that restriction (results available on request), we exclude nonviolent events (establishment of rebel headquarters, nonviolent rebel presence, changes of territorial control without violence, and protests and riots) as they are not directly related to resource-based conflicts.Weather data are mainly generated from the University of East Anglia's (UEA) Climatic Research Unit (CRU) Time Series (TS) dataset, version 3.1. This dataset provides monthly mean temperature and precipitation from January 1901 at 0.5 degree grid resolution (equivalent a 50-kilometer grid resolution). However, the accuracy of these data has been questioned. As explained in Mitchell and Jones (2005, 702), values at the station level \"were interpolated onto a continuous surface from which a regular grid of boxes of 0.5 degree was derived and, in order to ensure that the interpolated surface did not extrapolate station information to unwarranted distances, 'dummy' stations with zero anomalies were inserted in regions where there were no stations.\" Thus, if the closest weather station with available data is too far, a long-term average value is used. The issue seems to be particularly important for precipitation data. 16 For North and South Sudan, Figures A.1.a and A.1.b illustrate the bias it introduces in the shape of the precipitation distribution. In spite of the critics, 17 most studies on the climate-conflict nexus use this dataset, not only at the cross-country level (Miguel, Satyanath, and Sergenti 2004;Burke et al. 2009;Harari and La Ferrara 2012) but also at the regional level (Kevane and Gray 2008;Raleigh and Kniveton 2012;O'Loughlin et al. 2012) and the national level (Theisen 2012;Maystadt, Eckers, and Mabiso 2013). The CRU dataset is so widely used because it has the advantage of providing precipitation and temperature data from 1901 and, consequently, allows correcting for deviations from long-term normal conditions. Given the consensus confirming that data from 1901 to about 1950 are not accurate for SSA, 14 See www.acleddata.com, downloaded in October 2012. 15 See http://strausscenter.org/codebooks/ACLED%202.0%20Codebook.pdf. 16 In our sample, out of 75,012 observations, there are only 18 observations with zero anomalies for mean temperatures, but there are, depending on the quarter, between 357 and 5,586 observations with zero anomalies for precipitations.17 Kudamatsu et al. (2012, 6), among other scholars, stated that such an interpolation method is problematic for exploiting variation within location over time since weather stations with consistent time-series observations in most African countries are few and far between. For example, Kevane and Gray (2008), investigating the relationship between rainfall shocks and conflict in Darfur, noticed that rainfall station data for Darfur had been collected since the early 1990s only in three or four main towns. Similarly, some climatologists claim that only data based on satellite estimates can really cover the entire African continent at a suitably detailed resolution. Others, such as Lobell (2013), stress that the measurement errors resulting from the interpolation method may be particularly problematic for data on precipitation.anomalies have been computed based on a long-term reference period starting in 1949 as in O'Loughlin et al. (2012). In addition, considering the criticism expressed about data based on weather stations, we test the robustness of our analysis with an alternative satellite-based dataset covering the period from 1997 to 2009 and provided by the POWER project of the National Aeronautics and Space Administration (NASA) of the United States. 18 Beyond offering a robustness check on the quality of the East Anglia data, these data also offer us the possibility to compute the degree-days variables based on daily data. Moreover, such data are based on a larger pixel size and thus allow us to show that our results are not affected by the so-called modifiable areal unit problem and, in particular, by the scale problem, \"which is the variation in numerical results occurring due to the number of zones used in the analysis, and hence the possibility of obtaining different results for different resolutions\" (Harari and La Ferrara 2012, 27). Finally, SPEI comes from SPEIbase, version 2.0, a global dataset with a spatial resolution of 0.5 degree latitude/longitude and temporal coverage between 1901 and 2009, based on the routine programmed by Vicente-Serrano et al. (2010). 19 Compared to other multiscalar drought indexes, SPEI has the advantage of taking into account the joint effects of precipitation, potential evaporation, and temperature and therefore offers a more accurate measure of \"effective\" rainfall. Table A.1 summarizes the names, the construction, and the sources of all the weather variables used.Furthermore, we collect geo-referenced data on various geographical, economic, and social timeinvariant characteristics. Geographical data are similar to the variables employed by Dorosh et al. (2012), who explained to a greater extent the algorithms used for the estimations. Data on agroecological zones are based on the calculations of the FAO and the International Institute for Applied Systems Analysis, which combine data on land resources (climate, soil, and terrain) with a mathematical model for the estimation of potential biomass (Fischer et al. 2001). Crop-type data are drawn from the Spatial Production Allocation Model (2000, version 3, release 6) of the IFPRI. The IFPRI Spatial Production Allocation Model (You, Wood, and Wood-Sichra 2009) generates highly disaggregated, crop-specific production data by a triangulation of any and all relevant background and partial information. This includes national or subnational crop production statistics, satellite data on land cover, maps of irrigated areas, biophysical crop suitability assessments, population density, secondary data on irrigation and rainfed production systems, cropping intensity, and crop prices. This information is compiled and integrated to generate \"prior\" estimates of the spatial distribution of individual crops. Priors are then submitted to an optimization model that uses cross-entropy principles and area and production accounting constraints to simultaneously allocate crops into the individual pixels of a Geographic Information System database. The result for each pixel (notionally of any size, but typically from 1 to 100 square kilometers) is the area and production of each crop produced, split by the shares grown under irrigated, high-input rainfed, and low-input rainfed conditions (each with distinct yield levels). Data on road infrastructure are largely based on UNEP (2005) data, 20 urban centers are identified using the Global Rural-Urban Mapping Project (2000) data from the Center for International Earth Science Information Network (CIESIN), 21 and travel times are estimated based on an algorithm taking into account road quality, slope, biophysical characteristics of the land, and other factors (Thomas 2007). Data on population come from the fourth version of the African Population Database (UNEP/CIESIN 2004); 22 in particular, we use the population from 1990, which is based on intercensual 1983 through 1993 growth rates at the county level (or at the state level for the areas not enumerated). 23 Geo-referenced yearly information on night-lights density comes from the database presented by Henderson, Storeygard, and Weil (2012), and data on livestock density (head/square km, 2005) are drawn from the Gridded Livestock of the World (Wint and Robinson 2007). Information on the location of ethnic groups is based on the University of Zurich's Geo-referencing of Ethnic Groups dataset that relies on maps from the classical Soviet Atlas Narodov Mira. 24 More specifically, we use anthropological studies to classify the different ethnic groups according to their main type of livelihood: pastoral (including nomad and seminomad groups), agropastoral, or mostly based on agriculture. 25 Information about the distance to a major river or a lake comes from the Yale Geographically Based Economic Dataset (G-Econ, version 4.0) introduced by Nordhaus et al. (2006). 26  The descriptive statistics of the conflict-and weather-based variables are given in Table A.2. Figure A.2.a shows the time variation for the whole sample of the conflict data, along with the time variations of the first two climatic variables, that is, temperature anomalies and temperature shocks greater than one standard deviation, happening during the growing period. The time variation in the occurrence of conflict events is a case in point with major peaks corresponding to the main events reported by Johnson (2011). 27 Figure A.2.b presents the maps illustrating the location of violent events and the geographical variation of the aggregated values of the two climatic variables, chosen for presentation purposes. 28 These figures constitute a first indication of a time and spatial correlation between temperature-related shocks and the frequency of conflict. However, on this basis, we cannot infer a causal relationship. The cross-country literature warns us about potential bias due to unobserved heterogeneity. Favorable climatic conditions (for example, moderate temperature) have been associated with better institutions (Acemoglu, Johnson, and Robinson 2002;Easterly and Levine 2003;Rodrik, Subramanian, and Trebbi 2004), faster transition out of agriculture (Diamond 1997;Masters and McMillan 2001), and hence, stronger economic growth (Sachs and Warner 1997;Nordhaus 2006;Dell, Jones, and Olken 2012). Although certainly reduced, the same concern applies within a country. For example, the preference for a more temperate climate by colonizers and its impact on institutions may well explain the differences in local governance between locations of North and South Sudan. Similarly, we cannot exclude that the time correlation is driven by common factors generating spurious correlations between weather shocks and conflict. The introduction of grid-cell fixed effects, time dummies, and county-specific parallel trends in the regression analysis should drastically reduce these threats to causal inference.(Khartoum, Sudan: Census Office, 1993).24 See http://www.icr.ethz.ch/data/other/greg, downloaded in October 2012. 25 Pastoral groups include nomad (Baggara/Shoa Arabs) and seminomad (Karamojo, Teso, Zagawa, and Tubu) groups. Agropastoral groups include Hamitic (Lotuko, Bari, and Murle) and Nuba (Dago, Kadugli-Krongo, Koalib-Tagoi, and Temaini) tribes. The other groups are considered mainly reliant on agriculture.26 See http://gecon.yale.edu/data-and-documentation-g-econ-project, downloaded in October 2012. 27 For example, the five major peaks correspond to particularly conflictive times in North and South Sudan. The first peak in 1997 (quarter 1) corresponds to operations conducted by the Ethiopian army in collaboration with SPLA and the operation launched in Central Equatoria; in 1999 (quarter 2), the resurgence of violence between the government of Sudan (GOS) and SPLA followed the agreement between GOS and Eritrea not to support each other's rebel movement; in 2002 (quarter 2), the number of violent events surged as a result of the agreement by the GOS to allow the Ugandan army to pursue the Lord's Resistance Army in Sudan and the intensification of fighting (including bombing) in Bhar al-Ghazal and Upper Nile as well as in the south; in 2008 (quarter 2) and 2009 (quarter 1), fighting between SPLA and government militias intensified along the Kordofan-northern Bahr al-Ghazal border as well as in Unity State and around the town of Malakal (in Upper Nile).28 Violence frequently occurred in Darfur (around the three major cities of Nyala, El Fasher, and Geneina); in the bordering state with Uganda (Eastern Equatoria) given the recurrent involvement of the Lord's Resistance Army in South Sudanese conflicts; in South Kordofan, Upper Nile, and Jonglei provinces (around the oil fields close to the town of Bentio); and in the Eastern part of North Sudan (Blue Nile and Kassala).Table A.3 summarizes the results of estimating equation (1) including only temperature indicators. Based on these results, a change in temperature anomalies of one standard deviation increases the frequency of violent conflict by 31 percent (partial effect expressed as a share of the mean value of violent conflict). In addition, a change of one standard deviation in moderate temperature shocks during the growing period increases conflict by about 21 percent (for the variables \"Temp> 1 s.d.,\" \"Heat Shock> 1 s.d.,\" or \"Heat Shock Pctile85\"). A change of one standard deviation in extreme temperature shocks during the growing period increases conflict by 27 percent (for \"Temp> 2 s.d.\" or \"Heat Shock> 2 s.d.\") or by 31 percent (for \"Heat Shock Pctile90\"). These partial effects (obtained based on regressions 1 to 9 of Table A.3) reveal an interesting pattern according to which the impact of weather shocks, happening during the growing period, increases when our proxies capture more extreme events. This pattern is also confirmed when moderate and extreme events are distinguished with the use of degree-days thresholds. Introducing the quadratic terms, moderate temperature shocks reduce conflict by about 12 percent, whereas extreme temperature shocks exacerbate violence by about 4 percent (see regressions 10 and 11 of Table A .3). This result also confirms that the choice of scale for the units of analysis does not drive our findings, given that the pixel size used in the NASA POWER dataset is much larger. All the partial effects relative to Table A.3 are summarized in Table 3.1, column 2.When estimated without time trends and night-lights density, the effects remain essentially unchanged (see Table A.4 and column 4 of Table 3.1). Adding rainfall-related variables indicates that precipitation variations do not affect the frequency of conflict and do not alter the coefficients of the temperature-related variables (see Table A.5) or the relative partial effects (see column 6 of Table 3.1). These results clearly point to the role of temperature shocks in explaining variations in violence in North and South Sudan and are in line with recent evidence on the impact of temperature shocks on agricultural income in both developed (Schlenker and Roberts 2009;Lobell et al. 2013) and developing countries (Schlenkler and Lobell 2010;Lobell et al. 2011). However, they may still be sensitive to the choice of the specification adopted to estimate equation (1).Therefore, we investigate the robustness of our results (1) to other proxies, functional forms, and data sources for precipitation shocks; (2) to other modeling choices, including the use of state fixed effects and state-level controls similar to that of O'Loughlin et al. (2012) or Raleigh and Kniveton (2012);(3) to other levels of aggregation, that is, monthly and yearly levels; and (4) to explicitly modeling time and spatial dependency. First, we may wonder whether the fact that variations in precipitation do not affect violence is related to the lack of accuracy of the adopted proxies for precipitation shocks. In line with Harari and La Ferrara (2012), precipitation shocks may be better captured by variations in SPEI. Our results suggest that it is not the case for North and South Sudan, without altering the impact of temperature shocks (see Table A .6). We also exclude the possibility that the lack of impact of precipitation shocks may be driven by the absence of a time-lagging effect since including the time lags does not alter the coefficients of the temperature indicators and changes only slightly the partial effects (see Table A.7). 29 Still, the explanatory dominance of temperature shocks in contrast to rainfall shocks might just be due to larger errors in measuring precipitation. For example, Lobell (2013) shows that the interpolation method used in the UEA CRU-TS dataset substantially underestimates the impact of precipitation on crop yields. Therefore, we collected information regarding the exact location of the weather stations employed by the University of East Anglia for its Sudan database to check that the interaction terms between the weather indicators and the distance from the nearest weather station were indeed not significant. 30 This is a first indication that such measurement errors do not play much of a role in driving our results. In addition, we test the importance of precipitation shocks using the alternative NASA POWER dataset, and even in this case, we confirm the superiority of temperature shocks in explaining variations in violent conflict (see Table A .8). All the partial effects related to Tables A.6, A.7, and A.8 can be found in Table A.9.Second, we test the robustness of our results to the use of fixed effects at the highest level of aggregation (26 states in North and South Sudan), similar to O'Loughlin et al. ( 2012) and Raleigh and Kniveton (2012). Our results remain robust to such a modeling choice (see Table A.10); the coefficients of interest have the same direction and significance but are larger in magnitude, pointing to a possible upward bias. Adding the population by grid-cell (transformed into logarithm), an urban dummy, and the distance to roads and to international borders (transformed into logarithm) as control variables, like O'Loughlin et al. ( 2012), provides the expected signs without altering our main results (see Table A .11). Although introducing potential selection bias and changing the external validity of the results, we also implement a model with state fixed effects excluding the cells that never experienced violent conflict, similar to Raleigh and Kniveton (2012). Our results remain unaltered and the control variables proposed by the authors have the expected signs (see Table A.12). All the partial effects based on Tables A.10, A.11, and A.12 are summarized in Table A.13.Third, we confirm the validity of our findings using alternative time aggregations. When we estimate equation ( 1) at the monthly level, our results maintain the same signs (see Table A .14) and offer conclusions similar to the ones based on the estimates at the quarterly level, with the main difference that the introduction of two lags is needed to obtain effects of comparable magnitudes (see Table A.15)-as expected. At the yearly level, our results are equally confirmed (see Table A .16). Table A.17 shows all the partial effects related to Tables A .14, A.15, and A.16.Fourth, our results are robust to taking into account possible serial and spatial correlations. Table A.18 and column 8 of Table 3.1 present the estimates and the partial effects of the dynamic panel model, whereas Table A.19 shows that even adding the spatial lags of the variables of interest does not change the previous findings. These spatial lags are obtained by multiplying the vector of observations by the matrix W, a normalized spatial matrix of order 1 (but the results appeared to be robust also to the choice of a matrix of order 2). There are spatial spillovers for temperature shocks, but they are sufficiently close to zero and do not alter the partial effects (see column 10 of Table 3.1).Therefore, the estimation of equation ( 1) indicates that extreme temperature shocks increase the frequency of conflict in North and South Sudan. Such impact appears to be robust to other proxies, functional forms, and data sources for precipitation shocks; to other modeling choices; to other levels of aggregation; and to explicitly modeling time and spatial dependency. With a view to our projection exercise, the most reduced-form set of estimations whose partial effects are given in column 4 of Table 3.1 (excluding night-lights and time trends) can be considered an upper-bound limit of the effect of extreme weather shocks on conflict. On the contrary, the most structured set of models, that is, the dynamic model with spatial lags, represents the lower-bound limit of such an impact.Projecting changes in the incidence of violent events under future climate change is far from trivial. Previous research has tended to rely on selected climate models and to overlook climate uncertainty in future temperature (and rainfall) changes. To incorporate that uncertainty into our projections, we follow the approach recommended by Burke et al. (2012) and applied in their study on conflict in SSA (Burke et al. 2009). Thus, we incorporate our estimated responses of conflict to climate (see Table 3.1) with climate projections for the corresponding SSA subregion (Sahel) from 20 climate models and three scenarios resulting from the World Climate Research Project Program's (WCRP) Coupled Model Inter-comparison Project phase 3 (CMIP3). 31 Such models provide the expected change in temperature between 2030 and 1980-1999, expressed in degrees Celsius, for each model and scenario. To calculate the expected temperature in 2030 we summed the expected change at the quarter level and the average quarterly temperature for the period 1980-1999. Temperature anomalies in 2030 are then estimated considering the same long-term mean and standard deviation described in the previous section. A major assumption is that no adaptation behaviors or policies in addition to the ones already incorporated in our estimates will take place by 2030. Given the relatively short timeframe, the assumption seems reasonable.We apply the same method used by Burke et al. (2012) by distinguishing between climate and regression uncertainties. Climate uncertainty results from the difference in the various predictions for the expected change in monthly temperatures given by three possible scenarios and 20 possible climate models for each scenario. The expected impact on violent conflict is calculated by multiplying the expected percentage change in temperature anomalies by the coefficient showed in Table A.3. Regression uncertainty results from taking into account the variability given by the standard errors of the estimated coefficient. To quantify that uncertainty, we bootstrap 10,000 times the specification regressing temperature anomalies on violent conflict and we multiply the percentage change in temperature anomalies given by the median of the three scenarios by the coefficients obtained by bootstrapping. Total uncertainty results from taking into account both climate and regression uncertainty. According to Figure 3.1, all models predict more frequent violence as a result of projected temperature anomalies, with a 23.8 percent median projected increase by 2030. Such predictions are based on the partial effects obtained in our baseline model. Such additional increase would vary in a range between 21 and 30 percent for the lower-bound (dynamic model with spatial lags) and upper-bound estimates (model excluding night-lights and time trends). Similar to Burke et al. (2012), climate uncertainty is a larger concern than regression uncertainty in predicting the changes in violence by 2030. Taking the ratio of the difference between the 95 th and 5 th percentiles for the climate-uncertainty-only projections and the regression-uncertainty-only projections, we find that climate uncertainty is about 1.4 times larger than regression uncertainty. We also perform similar projections using proxies for more extreme temperature shocks. 32 As summarized in Table 3.2, the median projected impact is 10.2 percent when the shock is defined as a temperature deviation above one standard deviation, whereas the expected impact increases to 39.6 and 75.9 percent for more extreme shocks (above two standard deviations and above 95 percent of the pixelspecific distribution, respectively). However, the level of both climate and regression uncertainties does not allow us to give much interpretation to the projected changes in conflict due to changes in other proxies than temperature anomalies. Such a limit in our analysis confirms the difficulty of projecting how expected increases in temperature will translate into more frequent and more intense extreme events (Hansen, Sato, and Ruedy 2012;Rhines and Huybers 2013), hence the challenge to make predictions on the socioeconomic consequences of future extreme events. Our empirical analysis clearly points to the negative role of extreme temperature shocks in North and South Sudan, but due to limited data availability, it cannot describe the channels through which such shocks affect conflict. To partially cover this gap, we exploit the heterogeneity in the impact of weather variables and identify mitigating and exacerbating factors of the relationship between weather shocks and violence. 33  The worst detrimental effect of climate change on African economies usually relates to decreased crop yields, in particular for maize, sorghum, millet, groundnut, and cassava (Jones and Thornton 2003;Lobell et al. 2008;Lobell and Burke 2010;Schlenker and Lobell 2010;Blanc 2012). The literature links this detrimental impact to a robust predictor of conflict: the relative change in income and, consequently, in the opportunity cost to participate in violence (Miguel, Satyanath, and Sergenti 2004;Burke et al. 2009;Blattman and Miguel 2010). However, in North and South Sudan, the presence of these crops is not a significant exacerbating factor (see Tables A.20.a through A.20.e). On the contrary, as summarized in Table 4.1, we find a mitigating impact when weather shocks occur in areas with a large share of land occupied by millet production (see Table A.20.c). This impact can be explained by the low sensitivity of the crop to temperature variations (thanks to its high threshold temperature, set in the agronomy literature at 35 degrees Celsius). Moreover, it produces a low but steady yield with little fertilizer input, 34 and it grows well in arid and semiarid environments, requiring less water compared to other grains. 35 Table 4.1 and, in more detail, Tables A.20.a through A.20.e indicate that with the exception of millet, the presence of particular crops does not affect the relationship between weather shocks and conflict in Sudan-in line with the low percentage of national income that is on average derived from crops.In accordance with the importance of livestock for livelihoods in North and South Sudan, our analysis points to three main significant factors. All partial effects are summarized in Table 4.1. First, Tables A.21.a and A.21.b indicate that the interaction terms with the proxies for the presence of pastoralist livelihoods (using goat densities or the presence of pastoral and agropastoral ethnic groups) constitute exacerbating factors and confirm the vulnerability of livestock to temperature shocks (Thornton et al. 2009). Second, mitigating roles are found in Tables A.21.c and A.21.d for water availability and irrigation systems (when we interact our proxies of weather shocks with a dummy for grid-cells near a major river36 and with the share of irrigated land). The importance of water availability is not surprising, in particular in lowland areas where shocks on the limited amount of water have been reported to generate conflict about property rights and competition between pastoralists and other farmers. Interpretation about the role of irrigation systems is less obvious. The mitigating role of irrigation may point to a social benefit associated with the private benefits found for such investments (Lipton, Litchfield, and Faurèsc 2003;Smith 2004). However, our results should be taken with caution since we do not shed light on the potential of new investments that, it has been argued, are relatively limited in pastoralist areas (You et al. 2011;Headey, Taffesse, and You 2013). Third, Table A.21.e shows that the impact of weather shocks on conflict is largely mitigated by road and market accessibility (being within two hours' reach from the nearest human settlement of 50,000 or greater population). Such result can be explained by market access's facilitating destocking and restocking process and thus helping herders to smooth the detrimental impact of extreme weather shocks. 33 As pointed by Gleditsch (2012, 6), \"one of the lessons that the large N-community could learn from proponents of case studies is the emphasis on interaction terms.\" Nevertheless, we cannot rule out that these interaction terms may be endogenous to conflict. We therefore consider this exercise primarily interpretative.34 El-Dukheri, Damous, and Khojali (2004, 56) reported that between 1992 and 2004 millet had a mean yield of 99 kilograms/feddans, with standard deviation equal to 17, while sorghum had a mean yield of 201 kilograms/feddans, with standard deviation equal to 45 (data from the Sudan Statistics Department, General Administration of Planning and Agricultural Economics). 35 In particular, in Sudan, millet can grow in sandy soil (Goz land), whereas the other crops need to be cultivated in clay soil or near watercourses (Wadis land). These sandy areas (mostly in Darfur and Kordofan) are classified as marginal lands, unsuitable to and unfavorable for the cultivation of other crops. In case of shocks or of a negative yield-per-feddan trend, it's easier to increase the cultivated area of millet and therefore to keep the same total production amount. Authors' estimation based on ACLED and UEA CRU-TS. The sources for the interaction terms are detailed in the Data part of Section 3. Notes: Table 4.1 reports only the coefficients that were statistically significant in the models' equations. The values of the interaction terms \"Main Crop is Millet,\" \"Pastoral and Agropastoral Groups,\" \"Near to Major River,\" and \"Market Accessibility\" are calculated for the dummy = 1. The values of the interaction terms \"Goat Density\" and \"Share of Irrigated Land\" are calculated for the median value. A description of the weather variables is given in Last, we want to point to recent evidence suggesting that the coping strategies that had traditionally been adopted in arid and semi-arid areas of the Horn of Africa are progressively breaking down due to different mutually reinforcing factors, such as population growth, spread of pests (for example, prosopis julifora), limited mobility, and fragmentation of grazing land (Lybbert et al. 2004;Devereux 2006;McPeak, Little, and Doss 2011). Our results on the security consequences of the vulnerability of these areas make action even more urgent. As critically reviewed by Headey, Taffesse, and You (2013), the existing literature in the field suggests not only improving the resilience of the livestock sector through improved veterinary services, access to credit, provision of emergency feed, and better access to water but also supporting income diversification, in particular through education investments.Our analysis sheds light on the importance of enhancing resilience to weather shocks in North and South Sudan, in particular in arid and semiarid lowland areas, and therefore calls for more decisive and coordinated action to help herders better cope with shocks. Initiatives aimed at reducing vulnerability in the Horn of Africa should include support in destocking and restocking processes at times of drought through improved access to markets; development of insurance and credit markets, especially weather insurance schemes; and supply of income diversification opportunities through investment in irrigation (when profitable) and in education services adapted to a mobile population. Nevertheless, our analysis is limited in drawing clear policy recommendations. Understanding the returns on investment, also for conflict resilience, is certainly a path for further research.                                  ","tokenCount":"8248","images":["-1340965979_1_1.png","-1340965979_22_1.png","-1340965979_27_1.png","-1340965979_27_2.png","-1340965979_30_1.png","-1340965979_31_1.png","-1340965979_31_2.png","-1340965979_32_1.png"],"tables":["-1340965979_1_1.json","-1340965979_2_1.json","-1340965979_3_1.json","-1340965979_4_1.json","-1340965979_5_1.json","-1340965979_6_1.json","-1340965979_7_1.json","-1340965979_8_1.json","-1340965979_9_1.json","-1340965979_10_1.json","-1340965979_11_1.json","-1340965979_12_1.json","-1340965979_13_1.json","-1340965979_14_1.json","-1340965979_15_1.json","-1340965979_16_1.json","-1340965979_17_1.json","-1340965979_18_1.json","-1340965979_19_1.json","-1340965979_20_1.json","-1340965979_21_1.json","-1340965979_22_1.json","-1340965979_23_1.json","-1340965979_24_1.json","-1340965979_25_1.json","-1340965979_26_1.json","-1340965979_27_1.json","-1340965979_28_1.json","-1340965979_29_1.json","-1340965979_30_1.json","-1340965979_31_1.json","-1340965979_32_1.json","-1340965979_33_1.json","-1340965979_34_1.json","-1340965979_35_1.json","-1340965979_36_1.json","-1340965979_37_1.json","-1340965979_38_1.json","-1340965979_39_1.json","-1340965979_40_1.json","-1340965979_41_1.json","-1340965979_42_1.json","-1340965979_43_1.json","-1340965979_44_1.json","-1340965979_45_1.json","-1340965979_46_1.json","-1340965979_47_1.json","-1340965979_48_1.json","-1340965979_49_1.json","-1340965979_50_1.json","-1340965979_51_1.json","-1340965979_52_1.json","-1340965979_53_1.json","-1340965979_54_1.json","-1340965979_55_1.json","-1340965979_56_1.json","-1340965979_57_1.json","-1340965979_58_1.json","-1340965979_59_1.json","-1340965979_60_1.json","-1340965979_61_1.json","-1340965979_62_1.json","-1340965979_63_1.json","-1340965979_64_1.json","-1340965979_65_1.json","-1340965979_66_1.json","-1340965979_67_1.json","-1340965979_68_1.json","-1340965979_69_1.json","-1340965979_70_1.json","-1340965979_71_1.json","-1340965979_72_1.json","-1340965979_73_1.json","-1340965979_74_1.json","-1340965979_75_1.json","-1340965979_76_1.json","-1340965979_77_1.json","-1340965979_78_1.json"]}

data/part_2/0191075220.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"4b77779bf2b9cbb50be652b4cc1bbbe1","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/475ba90b-efd8-45a4-8f8c-6d3c6b929369/retrieve","description":"The role of chemical fertilizers for increased agricultural production, particularly in developing countries such as India, is well estab-lished. Some argue that fertilizer was as important as seed in the Green Revolution (Tomich et. al. 1995), contributing as much as 50 percent to the yield growth in Asia (Hopper 1993, FAO 1998). Others have found that one-third of worldwide cereal production is due to the use of fertilizer and related factors of production (Bumb 1995).","id":"2050765791"},"keywords":[],"sieverID":"dbd76098-a6a3-4de2-b403-e4e30e14d953","pagecount":"41","content":"The options for increasing food production are limited by the availability of land and water and the increasing population, among other factors. Fertilizers can play an increasingly important role in agricultural production as the opportunity to bring new area under cultivation diminishes and the majority of Indian soil becomes deficient in many macro-and micronutrients. The application of essential plant nutrients, particularly macro-and micronutrients, in the optimum quantity and the right proportion, by using the correct method and time of application and efficient and environmentally sound management, is the key to increasing and sustaining agricultural production. Therefore, it is important to understand fertilizer use behavior and efficiency over time and space, the changing structure of fertilizer markets, the policy environment, and the role of various factors influencing fertilizer consumption. This paper is an attempt to address some of these issues.The role of chemical fertilizers for increased agricultural production, particularly in developing countries such as India, is well established. Some argue that fertilizer was as important as seed in the Green Revolution (Tomich et. al. 1995), contributing as much as 50 percent to the yield growth in Asia (Hopper 1993, FAO 1998). Others have found that one-third of worldwide cereal production is due to the use of fertilizer and related factors of production (Bumb 1995).For the past four decades, India has relied on increasing crop yields to supply an ever-increasing demand for food. According to Ministry of Agriculture data, total food grains production rose from about 102 million tons in the triennium ending (TE) 1973(TE) -1974 to about 253 million tons in TE2012-2013, a 148 percent increase (GoI 2013). Meanwhile, the total area under food grains, which accounted for nearly three-fourths of the total cropped area in early 1970s, declined to 63.6 percent in TE2011-2012 and total area under food grains declined from 125 million hectares (ha) in the 1970s to 122 million ha in the 2000s. This dramatic increase in food grains production was the result of a 133 percent increase in crop yields between TE1973-1974and TE2011-2012. During the past two decades, India has lost 2 to 3 million ha of net sown area to nonagriculture purposes.Food security has been and will continue to be one of the major challenges confronting the world, including India, as the country faces the challenge and pressure to feed more than 1 billion people today. The agricultural policy has focused on increasing productivity and modern inputs, such as high-yielding variety (HYV) seed, chemical fertilizers, and irrigation, and subsidies that supported intensive farming played an important role during the 1970s and 1980s. Trends in fertilizer consumption and cereal production in India, shown in Figure 1, clearly indicate that the increased consumption of fertilizer has been a dominant factor underlying increases in crop production in the country. However, the association between fertilizer use and cereals production has weakened over time; for example, the correlation coefficient between fertilizer consumption and cereals production increased from 0.88 during the first phase of the Green Revolution (1965-1966 to 1970-1971) to 0.95 in the second phase (1980s) but declined during the 1990s (0.86) and 2000s (0.83).India is the second largest consumer of fertilizers in the world, with an estimated consumption of 28.1 million tons in 2010, after China (49.8 million tons). It accounted for 15.8 percent of the world's consumption of N, 19.9 percent of phosphatic (P2O5), and 12.7 percent of potassic (K2O) nutrients in 2008 (FAI 2012). At the onset of the Green Revolution in 1966-67, consumption of fertilizers was about 1 million tons and increased to 2.26 million tons in 1970-71, which further increased to 12.73 million tons in 1991-92. The rapid expansion of irrigation, spread of HYV seed, introduction of Retention Price Scheme (RPS), distribution of fertilizers to farmers at affordable prices, expansion of dealers' networks, improvement in fertilizer availability, and virtually no change in farmgate fertilizer prices during the 1980s were major reasons for the increase in fertilizer consumption from 1971 to 1990. During the 1990s, total fertilizer consumption fluctuated between 12.15 and 16.8 million tons, with the exception in 1999-00, when fertilizer consumption was more than 18 million tons. In the past decade, fertilizer consumption increased at a faster rate, and total fertilizer consumption reached a record level of 28.1 million tons during 2010-11 and marginally declined to 27.8 million tons in 2011-12.Figure 3.1.1 shows the share of primary nutrients in total fertilizer consumption. Nitrogenous fertilizers accounted for 62.4 of total nutrient consumption in the country during the 2000s. The share of N was 78.5 percent in the 1950s, and declined to 68.6 percent in the 1960s, to 67.9 percent in the 1970s, and to 65.7 percent in the 1980s. However, the share of N increased to 67.9 percent in the 1990s, then fell to 62.4 percent in the 2000s. for P fertilizers, the share increased from 13.5 percent in the 1950s to 21.4 percent in the 1960s, then marginally declined during the 1970s and again picked up during the 1980s (24.1 percent). During the 1990s the share of P in the total consumption declined to 23.6 percent and then increased during the 2000s to 26.3 percent. Likewise, the share of K increased from 8 percent in the 1950s to 11.4 percent in the 1970s, declined to 10.2 percent in the 1980s, and fell further to 8.5 percent in the 1990s. The share of K increased to 11.4 percent in the 2000s. The rise in the share of N and the decline in the share of P and K fertilizers SUMMARY | APRIL 2010 during the 1990s was mainly because of slow growth in the consumption of P and K fertilizers compared with N fertilizers due to the decontrol of P and K fertilizers and the relatively high increase in their prices vis-à-vis N fertilizers, which remained almost stable during the decade. Concerned with the problem of increasing imbalance in the use of primary nutrients, the government introduced a concession scheme on the sale of decontrolled P and K fertilizers to farmers in the mid-1990s, but still prices of these fertilizers were higher than nitrogenous fertilizers. In the late-990s and early 2000s, the government hiked the concession rates for P and K fertilizers, which led to an increase in their consumption and a higher share in total fertilizer use during the 2000s. The consumption of N fertilizers increased from about 4.5 million tons in TE1983-84 to about 16.6 million tons during TE2011-12 (about 266 percent increase), P consumption increased from about 1.5 million tons to about 7.5 million tons (418 percent increase), and K consumption from 726 thousand tons to 3.2 million tons (346 percent increase). The share of N declined from 66.9 percent to 60 percent, while the share of P increased from 22.2 percent to 28.2 percent and K consumption from 10.8 percent to 11.8 percent during the past three decades. The government introduced Nutrient-Based Subsidy (NBS) of subsidized fertilizers in 2008 and prices of all fertilizers were refixed by benchmarking to prices of urea, DAP, and MOP, which led to a reduction in prices of complex fertilizers, a decline in the share of N (from 63.9 percent in 2007-08 to 60.6 percent in 2008-09), an increase in the share of P (from 24.4 percent to 26.1 percent), and a decline in the share of K (from 11.7 percent to 13.3 percent). However, the government partially decontrolled the fertilizer sector and introduced the NBS scheme on phosphatic and potassic fertilizers in April 2010, which led to a very steep increase in P and K prices and a marginal decline in the share of P (from 28.6 percent in 2010-11 to 28.5 percent in 2011-12) and a significant decline in K fertilizers (from 12.5 percent in 2010-11 to 9.3 percent in 2011-12). On the other hand, the share of N consumption increased from 58.8 percent in 2009-10 (pre-NBS) to 62.3 percent in 2011-12 (post-NBS).Table 3.2.1 shows the major fertilizer products consumed in India. Urea is by far the most widely used product. Together with other straight nitrogen fertilizers, such as ammonium sulphate (AS) and ammonium chloride (ACl), they make up nearly half the total market share. NP/NPK complex fertilizers (excluding DAP) are the second largest products, accounting for 20 percent of the market share, followed by DAP (18 percent) and SSP (7.5 percent) during 2011-12. The share of NP/NPK complex fertilizers, which witnessed a declining trend between 1981-82 and 1991-92, increased during the post-reforms period and more particularly during the past five years (after introducing Nutrient-Based Subsidy of fertilizers in 2008). Urea accounts for about 80 percent of India's total nitrogen consumption, and the other nitrogenous fertilizers-calcium ammonium nitrate (CAN), ammonium sulphate, and ammonium chloride-account for only a 1 percent share during 2011-12. In the case of phosphatic fertilizers, DAP accounts for about 59 percent of the total phosphorus consumption, single superphosphate accounts for 10 percent, and NK/NPK complex fertilizers account for 30.9 percent. The main reason for the predominant share of these two products (urea and DAP) is that the subsidy/concession was available on these products. Under earlier pricing regimes, the price of nutrients in complex fertilizers and other decontrolled fertilizer products were higher than the price of the same nutrient in other straight fertilizers, such as urea, DAP, MOP, and SSP. This led to a comparatively higher usage of straight fertilizers vis-à-vis complex fertilizers. However, in order to promote balanced use of fertilizers and provide more choice to farmers, the government took a positive step and introduced the Nutrient-Based Subsidy (NBS) scheme to cover other products, including complex fertilizers, in June 2008. This policy intervention increased the choice of products within three primary nutrients as well as a more balanced use of fertilizers in terms of the N:P:K ratio. NBS significantly reduced the price of complex and other fertilizers, which led to some improvement in the share of complex fertilizers in total nutrient consumption.  Fertilizer consumption in India is highly skewed, with wide interregional, interstate, interdistrict, and intercrop variations. Intensity has generally been higher in the southern (215.5 kg/ha) and northern regions (187.4 kg/ha) and lower in the eastern (119.1 kg/ha) and western regions (104.1 kg/ha). The sustained growth in fertilizer use during the past three to four decades is quite apparent in all regions. However, some of these regional averages are heavily influenced by individual state fluctuations. For example, during SUMMARY | APRIL 2010 2011-12 in the western region, Gujarat had a high rate, at 155.6 kg per hectare, while Rajasthan had a very low rate, at 62.3 kg per hectare. Similarly, in the northern region, Punjab had a very high rate, at 243.6 kg per hectare, while Himachal Pradesh had a low rate, at 55.2 kg per hectare (Figure 3.2.3). In the south, fertilizer consumption varied from about 674 kg per ha in Puducherry to 112.9 kg in Kerala. Similar variations are quite apparent in the eastern region as well (less than 10 kg in the northeastern states to 184 kg in Bihar). Out of 32 states/union territories (UTs) for which fertilizer consumption data are available, 22 states/UTs had less than the national average consumption and the remaining 10 states had higher than the national average during 2011-12. The average intensity of fertilizer use in India at the national level is still much lower than in other developing countries, but there are many disparities in fertilizer consumption patterns both between and within regions of India. During the triennium ending (TE) 1986-87, only three districts were using more than 200 kg per hectare of fertilizer and another 12 districts were consuming between 100 and 150 kg per hectare. In contrast, about 60 percent of the districts were using less than 50 kg of fertilizer (N+P+K) per hectare. However, the number of districts in the high-fertilizer use category (>200kg/ha) increased significantly during the second half of the 1990s and 2000s. In TE1999-00, out of 470 districts, 31 districts (6.6 percent) were using more than 200 kg per hectare, while about one-third of the districts were consuming less than 50 kg. Between TE2002-03 and TE2011-12, the number of districts consuming more than 200 kg/ha more than tripled, from 36 in TE2002-03 to 135 (24.7 percent) in TE2011-12. The increasing number of districts consuming consistently higher amounts of fertilizer (>200 kg/ha) is a cause for concern because it may lead to environmental degradation, particularly land and water resources. On the other hand, about 20 percent of the districts still use less than 50 kg/ha of fertilizer. Therefore, there is a need have a two-pronged strategy, one to monitor districts with a high intensity of consumption and take corrective actions to reduce adverse effects on environmental resources and the second to promote fertilizer consumption in low-use districts to improve crop productivity.Understanding the share of different crops in fertilizer consumption is a key component of fertilizer market analysis and a prerequisite to the development of sound fertilizer demand forecasts. It is generally expected that the major benefit of fertilizers goes to the areas having better access to technology, irrigation facilities, and infrastructure, and growing fertilizer-intensive crops like rice, wheat, sugarcane, fruits, and vegetables. Table 3.3.1 shows the trends in fertilizer usage in India by various crops/crop groups. It shows that in 2006-07, rice was the largest user of fertilizer (about one-third of the total consumption), followed by wheat (24.2 percent). Rice and wheat accounted for more than 60 percent of total fertilizer consumption in the country in 1995-96, and the share declined to 56.8 SUMMARY | APRIL 2010 percent in 2006-07. Fruits, vegetables, and sugarcane combined represented another 11 percent of fertilizer use. Cotton accounted for about 5.6 percent of total use. In all the years, rice was the dominant crop fertilized. Fruits and vegetables appear to be increasing in importance. Fertilizer intensity, measured as the average kg per hectare, does not follow exactly the same pattern across crops; intensity tends to be higher on sugarcane (234.9 kg/ha), vegetables (253.8 kg/ha), cotton (183 kg/ha), and fruits (158.6 kg/ha) and lower on cereals (rice 129.2 kg/ha and wheat 162.6 kg/ha) and pulses (Table 3.3.2). It is evident that farmers growing input-intensive crops are the main beneficiaries of fertilizer use.    in 1991-92, 1996-97, 2001-02, and 2006-07. It is evident from the table that the average use of fertilizers was higher with small and marginal farmers compared to medium and large farmers. The average fertilizer consumption per hectare of gross cropped area (GCA) was the highest (139.74 kg) on marginal farms and the lowest on large farms (67.64 kg) in 2006-07. A similar trend was observed during 1991-92, 1995-96, and 2001-02. Moreover, there has been a significant increase in fertilizer intensity per hectare of GCA on all farm size holdings during the periods 1991-92 and 2006-07. However, the increase was the largest (95.9 percent) on small farms, followed by marginal holdings (93.5 percent), and the lowest (47 percent) was on large farms. Because the fertilizer subsidy is universal and not targeted at a particular category of farmers, we computed the average subsidy rate (Rs./ton) based on the total fertilizer subsidy and the actual consumption of nitrogenous, phosphatic, and potassic fertilizers. It is estimated that the average fertilizer subsidy per ha of cropped area is significantly higher in cases of small and marginal farmers compared with large farmers because average fertilizer consumption is also higher on small and marginal farms. Sharma and Thaker (2010) have reported that the benefits of fertilizer subsidies are not restricted to only resource-rich states but have spread to other states also. It is worth mentioning that the benefits of fertilizer subsidy have spread to unirrigated areas because the share of area treated with fertilizers has increased on unirrigated lands. Likewise, the share of unirrigated areas in total fertilizer use has also increased during the past decade, indicating that fertilizer subsidies have benefited farmers in rainfed areas.As is evident from the Table 3.3.4, with a share of just over 6 percent in total holdings, medium and large farmers consumed about 25 percent of total fertilizers used in the country in 2006-07. Semi-medium farmers accounted for about 11 percent of holdings, but consumed 22 percent of total fertilizers. On the other hand, small and marginal farmers, which constitute 82.5 percent of total holdings, consumed nearly 53 percent of total fertilizers. However, when we look at relative shares of different farm size groups in area operated and fertilizer used, the picture changes dramatically. For example, in 2006-07, the share of small and marginal farmers in gross cropped area was 44.4 percent, and they consumed about 53 percent of total fertilizer used in the country. On the other hand, the share of medium and large farmers in gross cropped area was nearly one-third, but they consumed about one-fourth of total fertilizers. Significantly, 73.6 percent of gross cropped area on marginal farms and 76.6 percent on small farms were fertilized compared with only 58 percent on large farms in 2006-07. The share of fertilized area to gross cropped area has increased on all farm sizes, but the increase was higher on small and semi-medium farms than on large farms. However, small farms witnessed some decline in the share of fertilized area to gross cropped area between 2001-02 and 2006-07. The subsidy has encouraged greater use of fertilizers in general and small and marginal farmers in particular. The N:P:K ratio was a little skewed toward N in the mid-1970s but started improving in the late 1970s and 1980s and reached a level of 5.9:2.4:1 in 1991-92. This improvement was due to the government's tight controls on fertilizer prices, sales, and distribution during the 1980s, when fertilizer prices remained unchanged. However, in August 1992, prices, distribution, and movement of phosphatic and potassic fertilizers were decontrolled while urea remained under statutory price control. The subsidy on phosphatic and potassic fertilizers was withdrawn, resulting in a sudden increase in retail prices of these fertilizers. For example, the price of DAP in terms of nutrient content increased from Rs. 7.57 per kg of P2O5 in July 1991 to about Rs. 12 in August 1992 and reached a level of Rs. 19.45 in rabi 1995-96. Similarly, the price of MOP in terms of nutrient content (K2O) increased from Rs. 2.83 per kg in July 1991 to Rs. 7.50 in August 1992 and reached a level of Rs. 7-8 in rabi 1995-96. On the other hand, the retail price of urea was reduced by 10 percent. The retail price of DAP and urea was in the ratio of 1.5:1 in 1991-92. MOP and urea prices were in the ratio of 0.56:1. However, with the decontrol of P and K fertilizers, the ratio of retail prices of DAP and urea widened to 2.4:1 in rabi 1991-92 and the ratio of MOP and urea also distorted to 1.6:1. The NPK use ratio got distorted significantly from 5.9:2.4:1 during 1991-92 to 9.5:3.2:1.0 in 1992-93. The share of N, P, and K in total fertilizer consumption was 63.2, 26.1, and 10.7 percent, respectively, in 1991-92. The N share increased to about 71 percent in 1993-94, while the share of P declined to 21.6 percent and that of K to 7.3 percent.To correct the imbalance in use of N, P, and K fertilizers, the government implemented a scheme of ad hoc concession on the sale of decontrolled fertilizers to farmers from October 1, 1992, but still there was significant disparity in prices of N, P, and K fertilizers, which led to more use of N and less use of P and K, resulting in more imbalance in use (the NPK ratio reached a level of 10.0:2.9:1.0 in 1996-97). Concerned with this deteriorating NPK ratio, the government announced a substantial increase in concession on P and K fertilizers, effective from July 6, 1996. The rate of concession on indigenous diammonium phosphate (DAP) was raised by three times from Rs. 1,000 per ton to Rs. 3,000 per ton. A concession to the extent of Rs. 1,500 per ton was extended to imported DAP to bring its selling price on par with indigenous DAP. Similarly, the concession on muriate of potash (MOP) was increased from Rs. 1,000 per ton to Rs. 1,500 per ton. The rate of concession on single super phosphate (SSP) was also enhanced from Rs. 340 to Rs. 500 per ton. Further increases in concessions on phosphatic and potassic fertilizers in subsequent years and an increase in price of urea in February 1997 led to improvement in the NPK ratio and reached a level of 4.3:2.0:1.0 in 2009-10, which was very close to the desired ratio of 4:2:1.To ensure a balanced use of fertilizers, the government moved toward a Nutrient-Based Subsidy (NBS) regime instead of a product pricing regime. The government introduced an NBS policy for P and K fertilizers on April 1, 2010, and market prices of all fertilizers (except urea) were to be determined by market forces based on a demand-supply situation. The subsidy was also given on sulphur and additional subsidy on micronutrients, namely, zinc and boron. Imports of all P and K fertilizers were placed on Open General License SUMMARY | APRIL 2010 (OGL) but import of urea remained canalized. Exclusion of urea from NBS and decontrol of P and K fertilizers led to imbalanced application of nitrogen vis-à-vis phosphatic and potassic fertilizers. The retail price of urea remained stagnant (Rs. 4830/ton) between 2002-03 and 2010, increased marginally to Rs. 5310 per ton in April 2010, and recently increased to Rs. 5360. The increase was about 11 percent during the past decade.On the other hand, prices of decontrolled P and K fertilizers increased significantly due to a reduced level of subsidy, higher world prices of raw materials, and depreciation of Indian currency. The retail price of DAP and MOP remained constant (Rs. 9350/ton for DAP and Rs. 4455/ton for MOP) in the pre-NBS period, from February 2003 to March 2010, but the subsidy kept on changing depending on the cost of production and import parity prices After the NBS policy was introduced in April 2010, which moved from a \"fixed-price-floating subsidy\" regime to a \"fixed-subsidy-floating price,\" the prices of phosphatic and potassic fertilizers registered a sharp increase. For example, price of DAP more than doubled between March 2010 and June 2012, from Rs. 9,350 per ton to more than Rs. 24,000 per ton, while the subsidy declined from Rs. 19,763 per ton in 2011-12 to Rs. 14,350 per ton in 2012-13. In the case of MOP, prices witnessed a very sharp increase in the post-NBS period and the price of MOP increased from Rs. 4,455 per ton in March 2010 to about Rs. 17,000 per ton in June 2012, an increase of about 280 percent (Sharma 2013). The average price of urea in India is one of the lowest (US$98/ton) compared with prices in the US (US$503), China (US$348), Pakistan (US$344), and Bangladesh (US$250). The current ratio of international prices of DAP, MOP, and urea is around 1.4:1.1:1; however, the ratio is much distorted (4.5:3.1:1) in India due to policy distortions. The subsidy on decontrolled P and K fertilizers has also witnessed a declining trend during the past three years, and NBS rates for N, P, and K have declined by about 23, 42, and30 percent, respectively, in 2013-14 compared to 2011-12 (Figure 3.4.1). All these developments have led to worsening of the NPK ratio; it reached a level of 6.7:3.1:1 in 2011-12 (post-NBS period) and became even worse (8.7:3.4:1) in 2012-13. The NPK ratio, which is a measure of balanced use of fertilizer, shows wide interregional and interstate disparities and consumption ratios. While existing variation from the ideal ratio (4:2:1) was nominal in the south (3.9:2.2:1) and eastern regions (4.2:1.8:1), it was very wide in the north (20.4:6.8:1) and western regions (7.3:4.0:1). The state-wise consumption ratio of N and P in relation to K shows that the greatest degree of N:P:K imbalance was seen in Rajasthan (34.9:15.9:1), followed by Haryana (27.2:9.8:1) and Punjab (26.8:8.5:1) in 2011-12. It is also interesting to note that the NPK ratio has deteriorated in almost all states in the post-NBS period, which is a cause for concern.Fertilizer is a key component in improving agricultural production and productivity, but fertilizer use efficiency has become much more important in the market-driven economy. Fertilizer use efficiency can be expressed in several ways. Mosier et al. ( 2004) described four SUMMARY | APRIL 2010 agronomic indices commonly used for nutrient use efficiency, partial factor productivity (kg crop yield per kg nutrient applied), agronomic efficiency (kg crop yield increase per kg nutrient applied), apparent recovery efficiency (kg nutrient taken up per kg nutrient applied), and physiological efficiency (kg yield increase per kg nutrient taken up). Crop removal efficiency (removal of nutrient in harvested crop as percent of nutrient applied) is also commonly used to explain nutrient efficiency. Available data and objectives determine which term best describes nutrient use efficiency. Since data on farm trials are very limited, we used physical returns data (kg of crop required to buy 1 kg of N/P/K) to understand trends in nutrient use efficiency.Figure 3.4.2 shows the relationship between fertilizer nutrient prices and paddy and wheat prices during the past three decades. The figure shows that farmers had to sell more quantity of paddy rice to buy 1 kg of P than for N and K fertilizers. In 1981-82, farmers had to sell 5.07 kg of paddy to buy 1 kg of P through DAP, 4.44 kg of paddy to buy 1 kg of N through urea, and 1.89 kg of paddy to buy 1 kg of K through MOP. However, with the steady increase in the procurement prices of crops over the years and almost stable fertilizer prices during the 1980s, the profitability increased for all three nutrients. The profitability of P and K use declined significantly after the decontrol of the prices of these fertilizers in 1992, and farmers needed 4.6 kg of paddy to buy 1 kg of P and 2.78 kg to buy K compared with 3.29 kg and 1.23 kg, respectively, in 1991-92. However, after the reintroduction of subsidy on P and K fertilizers as well as a significant increase in output prices, the profitability of fertilizer use increased significantly and farmers needed 1.15 kg of paddy to buy 1 kg of N, 1.89 kg for P, and 0.84 kg for K in 2010-11. After the NBS scheme for P and K fertilizers was implemented in 2010, with unprecedented increases in the prices of these fertilizers, the profitability of P and K use declined significantly (3.67 kg of paddy to buy 1 kg of P and 1.86 kg for 1 kg of K) during 2011-12. An almost similar trend was observed in the case of wheat. Inadequate application of potassium and phosphates combined with over-application of nitrogen is a serious problem in intensive agricultural production systems. It leads to large N losses, environmental pollution, and low nitrogen use efficiency. Although ferti-SUMMARY | APRIL 2010 lizer consumption has increased significantly during the past four decades, the corresponding yield increase per unit of nutrient has diminished over the years (Samra andSharma 2011, Benbi andBrar 2011). The response ratio (kg grain/kg nutrient) in food grain crops in irrigated areas in India has substantially declined. The fertilizer response ratio in irrigated areas in the country declined from 13.4 kg of grain per kg of nutrient in 1970 to 3.7 kg of grain per kg nutrient in 2005 (Figure 3.4.3). While only 54 kg per ha was required to produce around 2 tons/ha in 1970, about 218 kg per ha is being added to sustain the same yield level now. The impaired soil health due to imbalanced fertilizer use, along with less use of organic manure, is mainly responsible for declining fertilizer response and crop productivity. There is a need to improve nutrient use efficiency for both economical and environmental reasons. Although the nutrient response ratio has declined during the past two to three decades, financial profitability of fertilizer use has improved (Figure 3.4.4). For example, gross financial returns from rice per rupee invested in nitrogen has increased from Rs. 2.64 in 1971-72 to Rs. 6.36 in 2001-02, and from Rs. 1.71 to Rs. 2.40 during the same period in the case of P. An almost similar trend was observed in the case of wheat. The continuous application of higher amounts of N, lower doses of P, and organic manure has led to the emergence of secondary and micronutrient (Zn, B, Fe, Mn, Mo) deficiencies in Indian soils. As a result, the rate of response of crops to applied fertilizers, factor productivity of crops, and nutrient use efficiencies have declined over the years. Deficiencies of essential elements in Indian soils and crops started emerging in the 1950s and as food production increased, the number of elements becoming deficient in soils also increased (Figure 3.5.1). Analysis of more than 25 thousand soil samples revealed widespread deficiency of Zn (49 percent) followed by S (41 percent), B (33 percent), Fe (12 percent), Mo (13 percent), Cu (3 percent), and Mn (5 percent) (Figure 3.5.2). While 130 districts were deficient in sulphur in the 1990s, the number has crossed to 240 in recent years. The states having serious deficiency in S include Himachal Pradesh (84 percent), Kerala (81 percent), Rajasthan (65 percent), Andhra Pradesh (56 percent), and Jharkhand (51 percent). Recognizing the importance of secondary nutrients and micronutrients, the government included sulphur under the NBS scheme and an additional Rs. 300 and Rs. 500 per ton for subsidized fertilizer fortified with boron (B) and zinc (Zn), respectively, was provided to encourage their application along with primary nutrients. During 2013-14, the subsidy on nitrogen was reduced to Rs. 20.875 per kg from Rs. 24 per kg in 2012-13. Similarly, the subsidy for phosphate was cut to Rs. 18.679 per kg from Rs. 21.804 per kg, while the subsidy on potash was reduced to Rs. 18.333 per kg from Rs. 24 per kg. However, the subsidy for sulphur was kept unchanged at Rs. 1.677 per kg for the 2013 fiscal year.Fertilizer production in India has grown, from very low levels after independence (38.7 thousand tons in 1951-52) and still low levels in the 1960s and early 1970s (1.24 million tons) to a total production of 16.6 million tons in 2011-12. There are about 145 fertilizer plants in operation in the country, which comprises 29 urea, 19 DAP and NP/NPK complex, 85 SSP, 10 ammonium sulphate (AS), one calcium SUMMARY | APRIL 2010 ammonium nitrate (CAN), and one ammonium chloride unit (FAI 2012). Currently, India produces various kinds of both nitrogenous and phosphatic fertilizers domestically, which include straight nitrogenous fertilizers (urea, ammonium sulphate, ammonium chloride, and calcium ammonium nitrate), straight phosphatic fertilizers (single super phosphate, and NP/NPK complex fertilizers, like diammonium phosphate (DAP). Potassic fertilizers are not manufactured domestically due to lack of commercially viable indigenous reserves of potash, the main raw material.At the time of independence in 1947, total fertilizer capacity in the country was about five thousand tons each of N and P2O5 with an investment of Rs. 5.9 crore. The capacity of nitrogenous fertilizers remained stagnant during the 1950s and early part of the 1960s. The real growth of the nitrogenous sector started only after the mid-1960s. During the period from 1969 to 1974, ten urea plants based on naphtha as feedstock were set up. The N capacity increased more than fourfold from 470 thousand tons at the end of the third fiveyear plan to 1,947 thousand tons in the fourth five-year plan due to more focus on agricultural development and the introduction of high-yielding varieties of rice and wheat in the mid-1960s (Table 4.1.1). The capacity creation was much faster during the fourth, fifth, and sixth five-year plans. The introduction of a retention price scheme (RPS) in the late 1970s contributed to this increase in N capacity. However, there has not been much capacity addition and it has remained stagnant at about 12-13 million tons during the past decade due to lack of an encouraging policy framework. Capacity utilization has increased considerably from around 67 percent during the fifth year of the five-year plan to 95.8 percent at the end of the tenth plan and reached 100 percent during 2010-11. Production shares are distributed slightly differently, due to sector-specific capacity utilization and efficiencies. The capacity utilization in N is considerably high in all sectors, but public units have relatively lower capacity utilization (91 percent) compared with the private sector (96 percent) and the cooperative sector (Table 4.1.2).  For nitrogenous fertilizer capacity, the public sector share has been declining over time. In the early 1970s the public sector accounted for about 62 percent of nitrogenous fertilizer capacity. The private sector held a share of about 28-29 percent and the cooperative sector about 8-9 percent. With policy changes toward greater investment in the private sector induced by the introduction of RPS in 1977, the public sector share started to decline and that of the private and cooperative sectors improved. As of November 2012, the public sector share was 27.1 percent, the private sector share was about 46.4 percent, and the cooperative share was 26.5 percent.Urea is the largest straight nitrogenous fertilizer in terms of capacity and in 2012 accounted for 78.9 percent of installed capacity, followed by DAP (9.7 percent) and NP/NPK (4.4 percent). Small quantities of other straight nitrogenous fertilizers, such as ammonium sulphate, calcium ammonium nitrate, and ammonium chloride, are also produced, but their share in total N capacity is small (2.4 percent).Table 4.1.3 shows sector-and product-wise capacity of the Indian fertilizer industry. The private sector share is higher for urea, ammonium sulphate, SSP, and complex fertilizers. There has been no capacity addition between 2001-02 and 2011-12 in almost all products except for some addition in urea and complex fertilizers. This additional capacity has been created mainly in the private and cooperative sectors. There has been a decline in the capacity of ammonium sulphate, ammonium chloride, and SSP fertilizers during the period.Source: FAI 2012.In the early years, the N capacity was based almost entirely on coke oven gas. By the 1970s, naphtha had become the most common feedstock, a position that was taken over by natural gas later on. In the 1970s, due to a shortage of naphtha for the fertilizer sector, coal and fuel oil raw material stock-based plants for producing urea and ammonia were set up. In 1981-82, naphtha was the major feedstock for N fertilizers, accounting for a 47.7 percent share, followed by fuel oil (22.7 percent) and natural gas (14.4 percent) (Figure 4.1.1). However in the late 1970s, with the discovery of gas fields off the west coast and on-shore in the northeastern parts of the country, the feedstock policy was amended in 1975-76 and new capacities were added in the 1980s and 1990s. Most of the capacity addition in the nitrogenous fertilizer sector was in natural gas feedstock-based units due to a new pricing scheme that sought to promote the use of natural gas, the efficient and comparatively cheaper feedstock, for urea production and encouraged naphtha/fuel oil/LSHS- based units to switch over to using gas as feedstock. Consequently, the share of natural gas increased to 65.4 percent, followed by external ammonia (15.5 percent) and naphtha (9.6 percent) in 2012. Natural gas is the preferred feedstock for urea production because it is a clean fuel and energy source. However, its availability, even to existing gas-based plants, has been under severe pressure because demand for gas is quite competitive since it serves as a major input to electricity generation and provides the preferred input to many other industrial processes. From the mid-1990s, the supply of gas to the fertilizer sector has decreased (42 percent in 1995-96 to about 26 percent in 2007-08) despite an initial allocation to meet the full requirements (Sharma and Thaker 2010). Consequently, gas-based units started facing a supply shortage and had to meet the shortfall using naphtha. Against the total requirement of 36.33 Million Metric Standard Cubic Meter Per Day (MMSCMD of gas for the existing gas-based fertilizer units, the actual average supply was 27.29 MMSCMD, a shortfall of about 24.8 percent (GoI 2012a). The nitrogenous fertilizer sector has suffered during the past decade because there has not been any addition to its capacity.At present, the fertilizer sector gets gas under the priority sector at a price decided under the Administrative Price Mechanism (APM) of the government. However, there is a pressure to give equal priority to the power sector in gas allocation as well as increase gas prices. Currently, APM and New Exploration Licensing Policy (NELP) gas are priced at US$4.2 per mmBtu, with gas from pre-NELP blocks costing between US$3.5 and US$5.73 per mmBtu. The basic price of imported gas is around US$14.17 per mmBtu. As per government estimates, India's fertilizer sector requires 62 MMSCMD and is expected to reach 113 MMSCMD in 2014-15. However, the Cabinet Committee on Economic Affairs (CCEA) has approved the Rangarajan Committee formula, which would lead to an increase in domestic gas prices to US$8.4/mmbtu for 2014-15 and to over US$10/mmbtu from 2015-16 onward. The move will have a significant impact on the domestic fertilizer industry.The government has also announced a New Investment Policy (NIP) for urea. Under the New Investment Policy 2012, in order to facilitate fresh investments in the urea sector, a system of a floor price and a ceiling price for the amount payable to urea units calculated based on the delivered gas price (inclusive of charges and taxes) to respective urea units was introduced. The floor and ceiling price of each urea unit is operative with respect to the computed Import Parity Price (IPP). The IPP defined for urea under the investment policy of 2008 is the average C&F price without any applicable custom duties and handling and bagging charges at the port. If the SUMMARY | APRIL 2010 computed IPP (payable) is between the floor and the ceiling price for that gas cost, it is the IPP (payable) that will be used. If the IPP (payable) is above or below the ceiling or the floor, respectively, it is the ceiling or floor price that will be acceptable as the case may be. Table 4.1.4 shows the base price for gas, the floor and ceiling prices for urea, and the increase in floor price for an increase in gas price beyond USD$14/mmbtu for different categories of investment. The capacity of phosphatic fertilizers in the country remained stagnant during the 1950s and early part of the 1960s. However, the capacity more than doubled from 274 thousand tons at the end of the third five-year plan to 581 thousand tons in the fourth five-year plan (Table 4.1.5). The capacity creation was much faster during the third, fourth, and fifth five-year plans. The new capacity addition during the eighth five-year plan was much less (from 2,716 thousand tons at the end of the seventh plan to 2,948 thousand tons at the end of the eighth plan). The main reason for this was decontrol of phosphatic fertilizers in 1992. Investment in the P sector picked up during the ninth plan but again became stagnant during the tenth plan. The total capacity addition during the tenth plan was 422,000 tons versus 2,301 thousand tons during the ninth plan. As of November 1, 2012, installed capacity of phosphate (P) nutrients was 6,242.9 thousand tons and production was 4,363.7 thousand tons. Capacity utilization of phosphatic fertilizers in the country has increased considerably, from around 71 percent during the fifth five-year plan to 86 percent at the end of the sixth plan. However, capacity utilization witnessed some decline during the seventh fiveyear plan. The long-term trend of a progressive step-up in capacity utilization suffered a setback in the wake of the partial decontrol of phosphatic fertilizers in 1992-93, and capacity utilization reached a level of 68.5 percent in 1993-94. However, with the introduction of a concession scheme, it was revived in 1994-95 and reinforced in 1995-96, when the capacity utilization attained the level of 90.7 percent. The capacity utilization was at an all-time high in 1997-98, at 100 percent, but witnessed a declining trend since 2007-08 and was about 70 percent in 2011-12. Private and cooperative units have higher capacity utilization compared with the public sector, but those declined between 1991 and 2012 (Table 4.1.6). Over the years, the public sector has lost its share to the private and cooperative sectors. About two-thirds of phosphatic fertilizer capacity is in the private sector. In 2012, 66.4 percent (61 percent in 1991-92) of installed capacity was held by private-sector units. The cooperative sector accounted for 27.4 percent (11.0 percent in 1991-92) and the public sector for only 6.2 percent (28.1 percent in 1991-92). The public sector has lost its share in production while cooperatives have increased their share significantly during the past two decades. Public units have lower capacity utilization and their share in production is only 5.4 percent, while the share of the private and cooperative sectors in phosphatic fertilizer production is 64.1 and 30.5 percent, respectively. There has been a substantial reduction in capacity utilization in all sectors between 1991-92 and 2012.DAP constituted about 51.6 percent of the total P2O5 capacity of about 7 million tons in 2012. SSP is the only straight phosphatic fertilizer manufactured in India, and it constituted about 21 percent of the total phosphate capacity. The remaining 27 percent of phosphate capacity was constituted by NP/NPK fertilizers (other than DAP).The raw materials and intermediates for phosphatic fertilizers are rock phosphate, sulphur, ammonia, phosphoric acid, and sulphuric acid. India meets a large part of its requirements in the phosphatic sector through imports of phosphatic raw materials/intermediates, such as rock phosphate and phosphoric acid. India imported 7.5 million tons of rock phosphate, 2 million tons of phosphoric acid, and 1.8 million tons of sulphur during 2011-12. In addition, India imports significant quantities of finished products, such as DAP fertilizer. India's share in the global trade of rock phosphate is about 21.3 percent, because the indigenous production is extremely limited. India's indigenous production of phosphoric acid is also very low and the country imports more than half of the global trade in phosphoric acid and uses 11-12 percent of world consumption. Sulphur is the main feedstock for phosphatic fertilizers and accounts for nearly half of the total capacity. The share of sulphur has remained almost stable during the past two and half decades, but the share of imported phosphoric acid, which is the most important feedstock, has increased significantly (from 26.9 percent in 1981-82 to 54 percent in 2011-12). The share of other raw materi-SUMMARY | APRIL 2010 als/intermediates has declined significantly. The share of imports in total feedstock supply for phosphatic fertilizers is quite high. Therefore, a high dependence on imports of raw materials exposes the Indian phosphatic industry to external factors such as high variability in prices.The fertilizer consumption in India has generally exceeded the domestic production in both nitrogenous and phosphatic fertilizers except for a few years. The entire requirement of potassic fertilizers is met through imports because India does not have commercially viable sources of potash. India mainly imports urea, DAP, and MOP. During the 1950s and 1960s, about two-thirds of the domestic requirement of N fertilizers was met through imports. Total imports of N fertilizers increased from 97 thousand tons in the 1950s to 482.4 thousand tons in the 1960s and 923.2 thousand tons in the 1970s (Table 4.2.1). The level of P imports was very low in the 1950s, and it increased significantly during the 1960s and 1970s. With the introduction of high-yielding varieties of wheat and rice in the mid-1960s, fertilizer imports increased significantly in 1966-67 and thereafter. Fertilizer imports increased dramatically in 1977-78 and 1978-79, in 1984-85, and again in 1988-89 and 1989-90. However, during the 1990s imports were at low levels except in 1995-96 and 1997-98. There appears to be a cycle of about eight to nine years when imports jump significantly. During the past decade, due to low/no addition in domestic capacity coupled with a rise in demand for fertilizers, imports have increased significantly in the 2000s (Figure 4.2.1). India imported 12.4 million tons of NPK fertilizer nutrients in 2011-12 compare with less than 1 million tons in the early 2000s. The growth of imports was rather slow in the 1980s and 1990s and accelerated in the 2000s. Fertilizer imports increased significantly in 2005-06 and the trend continued thereafter. Along with the quantity, the value of imported fertilizer nutrients also increased significantly during the past few years due to increases in international prices of feedstocks and the cost of imported fertilizers.observed for nitrogenous and phosphatic fertilizers. However, in terms of volume of imports, N fertilizer imports declined during the 1980s, marginally increased during the 1990s, and dramatically increased in the 2000s, while phosphatic fertilizers imports have increased consistently over time.Figure 4.2.2 shows the growth rates in fertilizer imports during the period 1971-72 to 2011-12. Fertilizer imports grew at an annual compound rate of 9.2 percent during the 1970s and 3.9 percent the following decade. During the 1990s, the growth rate in fertilizer imports was almost zero due to a negative growth rate in N fertilizer imports. However, fertilizer imports grew at an annual compound rate of about 23 percent during the period 2001-02 to 2011-12. Nutrient-wise, import trends show a different pattern. After the introduction of high-yielding varieties in the mid-1960s, demand for N fertilizers increased and so India's imports grew at an annual compound rate of more than 11 percent. However, due to domestic capacity additions during the 1970s (because of the introduction of RPS), domestic production increased significantly and thus reduced dependence on imports; as a result, N fertilizer imports recorded negative growth rates during the 1980s and 1990s. However, due to uncertainty in the N fertilizer sector policy environment during the past decade, there was no capacity addition and therefore imports grew at a rate of 47.4 percent.In the case of P fertilizers, imports grew at an annual rate of 1.6 percent in the 1970s, decelerated to -7.2 percent in the 1980s, and rose to 0.2 percent in the 1990s. However, in the 2000s, P fertilizer imports increased at a rate of 34.8 percent. In the case of K fertilizers, because all demand is met through imports, imports have registered a steady growth rate of about 8-9 percent during the past four decades, with the exception of the 1990s, when imports increased at a rate of 5.2 percent. This deceleration in the growth of imports was mainly because of slow growth/reduction in consumption of K fertilizers due to decontrol of K fertilizers in 1992-93 and subsequent price increases. The main fertilizer products imported in India are urea (7.8 million tons), DAP (6.9 million tons), and MOP (about 4 million tons). Urea imports have increased significantly during the last six or seven years. This increase in imports and rising international prices of urea and other fertilizer products have led to a substantial increase in fertilizer subsidies in the country. Oman (36.2% percent, China (22 percent), Iran (17.2 percent), and Commonwealth of Independent States (CIS) (12.9 percent) were major exporters of urea to India during TE2011-12 (Figure 4.2.3).The government provides a subsidy to fertilizer companies for transportation of fertilizers on the basis of three components. ( 1) \"Primary movement\" is the movement of fertilizers from port or plant by rail to various rake points, and the cost is reimbursed on the basis of railway receipts; (2) \"secondary movement\" is transportation of fertilizers from the railway rake points to the district headquarters, which the government scrapped on April 1, 2012; and (3) the government provides freight subsidy on direct road movement of P and K fertilizers (except single super phosphate) from plant or port to district headquarters as per the actual distance up to a maximum of 500 km.Currently, more than 40 percent (up from about 13 percent in the early 2000s) of total fertilizer nutrients used in India are sourced through imports. The capacity to produce more fertilizer in the country is currently limited due to availability and/or cost of raw materials/feedstocks, and installed capacity has remained stagnant during the past decade. The recent increases in fertilizer prices due to rising demand and rising feedstock/raw material costs has led to renewed discussions about the role of imports versus domestic production and the role of fertilizer subsidies and government-controlled imports and distribution, usually through state enterprises.  Some studies have led to a common perception that domestic industry, particularly urea, has been overprotected and less efficient than imports. For example, Gulati (1990), Gulati and Sharma (1995), and Gulati and Narayanan (2003) calculated the implicit fertilizer subsidy accruing to industry/farmers and argued that about half of the subsidy goes to the fertilizer industry. Panagariya (2001) wrote an article on fertilizer subsidy in the Economic Times on February 28, 2001, in which he stated that the bulk of the fertilizer subsidy rewards the gross inefficiency of urea manufacturers, and thus all subsidies to fertilizer manufacturers must go and imports opened up. However, these arguments were based on the fact that international price of urea was very low and varied from US$70 to US$140 per ton between January 1998 and February 2001 and assumed that the import price of urea will remain at about US$150 per ton.Empirical evidence clearly shows that the perception of the domestic urea industry being overprotected and less efficient than imports does not hold true, as is evident from Figure 4.3.1, which shows that the average subsidy per ton of imported urea is much higher than indigenously produced urea. The average subsidy on imported urea varied from about Rs. 2136 per ton in 2001-02 to about Rs. 18000 per ton in 2008-09 and 2011-12. In contrast, the subsidy on domestic urea varied from Rs. 4183 per ton in 2002-03 to about Rs. 9020 per ton in 2008-09, much lower than for imported urea. The average subsidy on domestic urea was higher (Rs. 4233/ton) than imported urea (Rs. 2136/ton) in only one year (2001-02) during the past decade whenever India imported urea. Because domestic urea is cheaper and more competitive vis-à-vis imported urea, the government must encourage domestic production, which will insulate Indian farmers from highly unpredictable, cartelized, and volatile world fertilizer markets.Keeping in view the availability of fertilizers in the country and the subsidy paid thereon, the government has put the export of all fertilizers in the restrictive category in order to discourage exports and smuggling. However, large quantities of subsidized fertilizers, particularly urea, are exported illegally to neighboring countries like Nepal, Bangladesh, and Myanmar, where prices are much lower than Indian prices. For example, the price of urea in India is INR 5360 per ton, while it costs about INR 28800 per ton in Nepal (1 INR = 1.6 NPR). Some estimates show that 60-65 percent of fertilizers supplied in Nepal are through the channels of informal imports, mainly from India.Fertilizers are produced at about 140 locations in the country and distributed to farmers scattered throughout the length and breadth of the country in more than 600,000 villages by a network made up of the private and cooperative sectors and other institutional agencies. Some quantities are also sold through the manufacturers' own outlets. Figure 5.1.1 shows the present fertilizer distribution system in India. Private trade accounts for about 65 percent of the total fertilizer distributed in the country, followed by institutional agencies, including cooperatives, at 35 percent; marginal quantities are distributed through manufacturers' own outlets. Among institutional agencies, cooperatives are the main distribution organizations. The cooperative marketing structure varies from state to state (two to four tiers). The total number of fertilizer sale points in India is 269,175, out of which about 62,637 (23 percent) are cooperative and other institutional agency sale points; private trade controls the remaining 206,538 (77 percent). The number of fertilizer sale points increased up to the mid-2000s (from 66,576 in 1969 to 292,692 in 2006) and then declined to 258,718 in in 2008. The share of institutional agencies also declined, while the private-sector share increased over the years. On average, one fertilizer sale point covers more than two villages. The northeastern states-Bihar, Orissa, Himachal Pradesh, Madhya Pradesh, Jharkhand, Chhattisgarh, and Rajasthan-have a very thin spread of sale points. Distribution networks in these states require intensification. Railways are the major share of transportation. During 2011-12, railways moved about 75 percent of the fertilizers produced and/or imported in the country; about 25 percent was moved through road transport.government laboratories, with an annual analyzing capacity of 134,000 samples. The government laboratories invariably check the quality of imported fertilizers.The state governments are adequately empowered to draw samples of fertilizer anywhere in the country and take appropriate action against sellers of nonstandard fertilizer. The penal provision includes prosecution of offenders and, if convicted, up to seven years of imprisonment, in addition to cancellation of the authorization certificate and other administrative action (GoI 2011). During 2006-07, 2007-08, and 2008-09, the percentage of fertilizer samples declared nonstandard were 6.0 percent, 6.2 percent, and 5.5 percent, respectively. Payment of concession for P and K fertilizers and for single super phosphate (SSP) is made by the department responsible for quality certification in the state. Further, SSP units are required to produce monthly \"quality certificates\" issued by the state governments in which the units are located. The units are required to have a well-equipped laboratory to test SSP samples. SSP units are also required to bear a \"quality certified\" stamp on each bag released in the market.The major deficiencies that constrained sustained rapid growth in fertilizer production also constrain fertilizer supply: shortage of raw materials, intermediates, and feedstocks such as natural gas for urea production and rock phosphate and phosphoric acid for phosphates; lack of a consistent long-term policy; constraints on working capital for the distribution channels; and physical infrastructure problems in some regions of the country. Due to constraints in raw material availability, the share of indigenous production of fertilizers has been decreasing, while imports have risen.It is more energy efficient and cheaper to produce urea using natural gas as feedstock. However, due to declining supplies of natural gas, even the existing gas-based units may face shortages. Although the fertilizer sector has been treated as a priority for the allocation of low Administrative Price Mechanism (APM) gas, the proportion of gas for the fertilizer sector has been declining (Sharma 2013). At present, the availability of gas to urea units is around 41 MMSCMD compared with their requirement of 43.14 MMSCMD (GoI 2012a). The new investment policy for the urea sector based on the Import Price Parity benchmark announced in 2008 was expected to attract much-required investment, but no major investment has been made. Suitable amendments to the new investment policy are required to create a conducive, incentive-based environment for new investments in urea sector.In the case of phosphates, the paucity of domestic raw material constrains the attainment of self-sufficiency. Indigenous rock phosphate (the main raw material) supplies meet only 5-10 percent of the total requirement of P2O5. At present, most of the indigenous rock is used in SSP plants. The rock phosphate exploitable reserves in the country are limited, and it is expected that the country will continue to depend on imported rock phosphate for meeting its demand in the years to come. Sulphuric acid is an intermediate in the manufacture of P2O5 fertilizers. India does not have any reserves of sulphur, and only moderate quantities of sulphur are available as recovered from the oil and gas sector. Sulphur is mostly imported from Iran, United Arab Emirates, Saudi Arabia, Kuwait, Bahrain, and Qatar. The indigenous production of phosphoric acid (an intermediate for phosphate production) has remained stagnant during the past few years. Approximately 85 percent of the world production of phosphoric acid is for captive consumption and only 15 percent is traded in the international market. It is reported that the trade of phosphoric acid is not a free trade and more than half of the international trade is by way of long-term supply arrangements between the producers and the importers. Out of the total trade of approximately 5 million tons of phosphoric acid, India imports more than 2.5 million tons every year, which exposes the Indian fertilizer industry to volatile world markets. With the expected increase in demand for fertilizer, the import of intermediates and raw materials is expected to grow significantly in the coming years.As a critical input in agricultural production, fertilizer consumption is affected by both price and non-price factors. The factors that affect fertilizer consumption cover a wide range of issues at different stages of its use. These can be classified into three groups: economic factors, such as fertilizer prices, output prices, and other input prices; physical and technological factors like soil quality, fertilizer-use management, availability of other inputs, climate, extent of micronutrient deficiency, and imbalanced use of various fertilizer nutrients; and institutional factors, including inadequate credit availability for farmers and dealers, insufficient extension activities, inadequate infrastructure (roads, transportation), inadequate distribution facilities, domestic production, and nonavailability of quality fertilizers (Raju 1989). These factors have a significant influence on fertilizer use patterns, although their relative importance varies across farm size, region, season, and other location-specific characteristics. Several studies have attempted to examine the role of price and non-price factors in the growth of fertilizer use in India (Raju 1989, Kundu and Vashist 1991, Subramaniyan and Nirmala 1991, SUMMARY | APRIL 2010Sharma 1993, Sidhu and Sidhu 1993, Dholakia and Majumdar 1995, Sharma 1999, Schumacher and Sathaye 1999, Rabobank 2005, and Sharma and Thaker 2011a). Sharma and Thaker (2011a) reported that non-price factors such as irrigation and high-yielding varieties were more powerful in influencing demand for fertilizer compared with price factors. The price of fertilizer had an adverse effect on consumption and was more powerful than output price. The results suggest that to increase fertilizer consumption in the country, policymakers should prioritize non-price factors such as better irrigation facilities, high-yielding varieties, and easy access to credit over agricultural price policy as an instrument. Second, there is a need to keep fertilizer prices at an affordable level because the price is more powerful in influencing fertilizer demand than higher output prices and benefits for small and marginal farmers.As discussed earlier, governments in developing countries, including India, promote fertilizer use through various policy instruments, such as subsidies. The fertilizer price at both producer and farmer levels are determined directly or indirectly by the government, and such interventions generally have two basic objectives, to provide fertilizers to farmers at stable and affordable prices to increase agricultural production and to encourage domestic production by allowing fertilizer producers a reasonable return on their investments. To achieve this objective, the government introduced the Retention Price cum Subsidy Scheme (RPS), a cost-plus approach, for nitrogenous fertilizers in November 1977 and extended it to complex fertilizers in February 1979. Under RPS the retail price of fertilizers was fixed and was uniform throughout the country, and the difference between the retention price (adjusted for freight and dealer's margin) and the price at which the fertilizers were sold to the farmer was paid back to the manufacturer as subsidy. RPS did achieve its objectives of developing a large domestic industry, near self-sufficiency in fertilizer production, and increased consumption of chemical fertilizers, but it has not been free from criticism of fostering inefficiency and leading to a huge burden of subsidies. The mounting burden of subsidies compelled the policy planners to make a serious attempt to reform fertilizer price policy to rationalize fertilizer subsidies. As part of the economic reforms initiated in the early 1990s, the government decontrolled the import of complex fertilizers such as diammonium phosphate (DAP) and muriate of potash (MOP) in 1992, and extended a flat-rate concession on these fertilizers. Urea imports continue to be restricted and canalized. Based on the recommendations of various committees, including the High Powered Fertilizer Pricing Policy Review Committee (HPC) and the Expenditure Reforms Commission (ERC), a New Pricing Scheme (NPS) for urea units was implemented in a phased manner starting in April 2003 with the objectives of bringing transparency, uniformity, efficiency, and reduced cost of production. Similarly based on the recommendations of the Expert Group on P and K fertilizers, the policy for phosphatic and potassic fertilizers was implemented. The government implemented the Nutrient-Based Subsidy (NBS) policy on April 1, 2010, for phosphatic, potassic, and complex fertilizers and from May 1, 2010, for single super phosphate (SSP). Under the NBS, the market price is determined based on supply and demand factors, and the government pays a fixed subsidy. The main objective of all policy interventions has been to contain and target fertilizer subsidies.However, estimates of the fertilizer subsidy as per central government budgets over the years in the post-reforms era show that fertilizer subsidy has increased significantly. Table 6.2.1 presents the estimates of major subsidies, including the food and fertilizer subsidies in the post-reforms period (1991-92 to 2011-12). It is evident from the table that fertilizer subsidy has increased from Rs. 5185 crore in 1991-92 to Rs. 70012 crore in 2011-12, representing an increase of more than 13 times. Fertilizer subsidy in India as percentage of the GDP varied from 0.47 in 2002-03 to 1.9 percent in 2008-09 and declined to about 0.8 percent in 2011-12. However, a steep increase in the cost of inputs to fertilizer production, high import prices of fertilizers, and constant farmgate prices have led to a substantial increase in fertilizer subsidy in the recent period. Fertilizer subsidy increased by more than 5.5 times between TE2003-04 and TE2010-11, from Rs. 11853 crore to over Rs. 66000 crore. The share of fertilizer subsidy in total subsidies varied from about 25 percent in 2002-03 to about 59 percent in 2008-09. Fertilizer subsidy reached a peak of Rs. 99495 crore in 2008-09 and then witnessed a declining trend. After two consecutive annual decreases in 2009-10 and 2010-11, fertilizer subsidy increased during 2011-12, mainly due to a rise in world prices of fertilizers. Fertilizer prices in 2011 averaged 43 percent higher than in 2010. However, after introduction of the NBS scheme, fertilizer subsidy recorded a decline during 2012-13 and was budgeted to be at almost the same level during 2013-14. Most of the time it is argued that domestic fertilizer prices are higher than world prices and the domestic industry is protected from import competition. However, it is not always true and in order to establish the fact we compared domestic prices with international prices (Figure 6.2.1). It is evident from the figure that domestic prices were lower than international prices in the first half of the last decade, but the situation changed dramatically during the second half of the decade and world prices were much higher and more volatile than domestic prices.2 Share was computed from subsidy figures given in various issues of Expenditure Budget Vol. I, Ministry of Finance, Government of India. 3 Includes Rs. 385 crore fertilizer subsidy given to small and marginal farmers. 4 Total subsidy on imported and indigenous P and K fertilizers. 5 Subsidy figures for 2007-08 and 2008-09 include both cash and bonds for both urea and decontrolled fertilizers. 6 Data on subsidies on sale of decontrolled fertilizers for 2010-11, 2011-12, and 2012-13 are a total of imported and indigenous P and K fertilizers because separate data are not available after NBS.It is evident from Figure 6.2.1 that the retail price of DAP and MOP remained constant (Rs. 9350/ton for DAP and Rs. 4455/ton for MOP) in the pre-NBS period, from February 2003 to March 2010, but the subsidy kept on changing, depending on the cost of production and import parity prices. The average subsidy on DAP varied from Rs. 2134 per ton on indigenous DAP in 2003-04 to Rs. 36488 per ton in 2008-09 (Rs. 53056/ton was the highest, reached in September 2008) in the pre-NBS era. In the case of MOP, the average subsidy varied from Rs. 2822 per ton in 2003-04 to Rs. 22528 per ton in 2008-09 (Rs. 29804/ton was the highest, reached in March 2009). After the NBS policy was introduced in April 2010, the \"fixed-price-floating subsidy\" regime was changed to a \"fixed-subsidy-floating price\" regime, and the prices of phosphatic and potassic fertilizers registered a sharp increase, particularly during the past year. For example, the price of DAP more than doubled between March 2010 and June 2012, from Rs. 9350 per ton to more than Rs. 24000 per ton, while the subsidy declined from Rs. 19763 per ton in 2011-12 to Rs. 14350 per ton in 2012-13 (Figure 6.2.2). In the case of MOP, prices witnessed a very sharp increase in the post-NBS period and the price of MOP increased from Rs. 4455 per ton in March 2010 to about Rs. 17000 per ton in June 2012, an increase of about 280 percent.The government of India has recently (as of June 26, 2103) asked the fertilizer companies to reduce the retail prices because demand for these fertilizers is largely met from imports, and local prices should fall in line with the decline in international prices. This has again raised the issue of indirect government controls in fertilizer pricing in the country.The share of subsidy in the total cost (retail price + subsidy) of DAP fertilizer was the highest (79.6 percent) during 2008-09 and declined in the post-NBS era to about 40 percent during April-June 2012. In the case of MOP, the share of subsidy in the total cost was as high as 83.5 percent in 2008-09 and declined significantly during the past two years due to a reduction in subsidy under the NBS scheme (Figure 6.2.3). If the subsidy on fertilizers is withdrawn in one go, the market price of DAP would increase to over Rs. 38000 per ton and MOP to about Rs. 31000, which are very high and unaffordable even for large farmers.Urea is being currently sold to farmers at the maximum retail price (MRP) of Rs 5360 per ton. The difference between the MRP and the production/imported cost is paid by the government to producers. A comparison of the domestic cost of production with import parity prices (IPP) of urea during the period 2004-05 to 2011-12 clearly shows that the IPP has been much higher than the domestic cost of production (Figure 6.2.4). On the other hand, maximum retail prices have remained constant during 2004-05 and 2009-10 and marginally increased during 2010-11. The Indian urea industry is quite diverse and the average subsidy varied from Rs. 8998 per ton in pre-1992 gas-based plants to Rs. 25772 per ton in pre-1992 naphtha-based plants and Rs. 22736 per ton in FO/LSHS feedstock-based units. It may be observed that FO/LSHS units account for 11 percent of capacity, but their share in subsidy is 23 percent (Table 6.2.2). However, the government has advocated for rapid conversion of existing naphtha-based urea units into gas-based units because gas is a much more efficient and cheaper fuel than naphtha. This would help in containing fertilizer subsidies.  There is debate about whether the fertilizer subsidy benefits the farmers or the fertilizer industry (Gulati 1990, Gulati andNarayanan 2003). Furthermore, the benefits of the fertilizer subsidy are heavily tilted to large farmers growing water-intensive crops in a handful of states. As per the estimates by Gulati and Narayanan (2003), the share of farmers in the fertilizer subsidy increased from 24.54 percent in TE1983-84 to 75.62 percent in TE1995-96, with an average share of 67.5 percent for the period 1981-82 to 2000-01. The the rest went to the fertilizer industry. These estimates of the shares of fertilizer subsidy going to farmers and/or industry have been computed by comparing subsidy estimates through import parity price and farmgate prices of fertilizers with the amount of subsidy given in the central government budget. Some of the recent policy announcements, such as the government's intention to move to a system of direct subsidy transfer to farmers, are based on such findings, which are based on unrealistic assumptions. For example, the study assumes that India's entry into the world fertilizer market as an importer would not affect world prices and that world fertilizer markets are perfectly competitive. However, both of these assumptions are not valid (Sharma and Thaker 2010).The benefits of fertilizer subsidies are analyzed using two All India Input Survey Reports (1996-97 and 2006-07) by the Agricultural Census Division of the Ministry of Agriculture. It is evident from Table 6.3.1 that small and marginal farmers, on average, use more fertilizer per hectare of gross cropped area than do larger farmers. In 2006-07, the marginal farmers used twice as much fertilizer per hectare (140 kg/ha) than large farmers (68 kg/ha). For small farmers, the average fertilizer consumption was about 90 percent higher than for large farmers. Between 1996-97 and 2006-07, average fertilizer use had the highest increase for small farmers (55.4 percent), followed by semi-medium farmers (43.9 percent); the lowest was on large farms (32.2 percent). The data on fertilizer consumption show that small and marginal farmers use more fertilizer compared to large farmers. Small and marginal farmers, who accounted for 82.6 percent of the total operational holdings in 2006-07, had a 44.3 percent share in the gross cropped area (Table 6.3.2). On the other hand, the proportion of large farmers in total holdings was 1 percent and their share in gross cropped area was more than 10 percent. However, it is interesting to note that the share of small and marginal farmers in total fertilizer consumption was much higher (52.9 percent) than their share in gross cropped area (42.8 percent). For large farmers, the share in fertilizer consumption was lower (6.1 percent) than their share in total cropped area (10.2 percent). These results show that small and marginal farmers have a significant share in fertilizer subsidies (higher than their share in total cropped area). To assess the benefits of fertilizer subsidies in irrigated and unirrigated areas, Sharma (2013) analyzed fertilizer consumption trends between 1996-97 and 2006-07. The data showed that although farmers in irrigated areas use more fertilizer (172 kg/ha) than those in unirrigated areas (59 kg/ha), fertilizer consumption has increased at a much higher rate in unirrigated areas (64.5 percent) compared with irrigated areas (32.2 percent). A similar trend was observed in all farm size groups. Sharma and Thaker (2010) found that there was a high concentration of fertilizer subsidies in only a few states but over time the inequalities in subsidy distribution among states have declined sharply. The coefficient of variation in the share of states with total fertilizer subsidy declined from 96.5 percent in 1992-93 to 82.1 percent in 1999-00 and further to 76.7 percent in 2007-08. The coefficient of variation in the per hectare fertilizer subsidy at the state level was substantially lower and has declined more sharply, from 79.3 percent in 1992-93 to 51.9 percent in 2007-08. This has happened due to improvements in rural infrastructure and irrigation facilities, coverage of area under high-yielding variety seeds, easy access to fertilizers, affordable prices, and a shift in crop patterns toward fertilizer-intensive crops in some of the less developed states during the past decade. The benefits of fertilizer subsidy are not restricted to only resource-rich states but have spread to other states. The analysis in this section supports the argument that public spending to subsidise fertilizers is desirable because a larger share of the benefits is captured by small and marginal farmers, who use higher quantities of fertilizers and have a greater share in total fertilizer consumption. Because there is no targeting of fertilizer subsidies and all categories of farmers pay the same price, it can be inferred that small and marginal farmers receive a higher subsidy per hectare as well as a larger proportion of the total subsidy. These findings are corroborated by the fact that earlier studies and input surveys show a similar distribution of benefits (Sharma and Thaker 2010). However, as fertilizer subsidies have become financially unsustainable, significant fiscal savings can be made through better targeting of fertilizer subsidies and an affordable increase in fertilizer prices. Having explored the distribution of benefits of fertilizer subsidies in the country, one question remains unanswered. Will dismantling the subsidy adversely affect fertilizer consumption and thereby farmers' income?A simple exercise using cost of production data from the Commission for Agricultural Costs and Prices reports on Price Policy for Kharif and Rabi Crops for the Marketing Season 2012-13 (CACP 2011 and 2012) examined the impact of removing fertilizer subsidies on farm incomes. Sharma (2013) examined the changes in net income (gross value of output from main and by-product -cost C2 * ) and farm business income (gross value of output [main and by-product] -cost A2+Family Labor) (Table 6.4.1).fertilizer use in India remains much lower than most countries in the world, but in certain states/districts fertilizer use is consistently high. The number of districts consuming higher than 200 kg/ha has more than tripled, from 36 in TE2002-03 to 135 in TE2011-12.In many developed countries, there has been a decline in fertilizer use efficiency, and one of the major constraints to fertilizer use efficiency in India is an imbalance of applied nutrients, partly as the result of a difference in the price of nutrients and partly due to the lack of knowledge among farmers about the need for balanced fertilizer application. The N:P:K ratio was a little skewed toward N in the mid-1970s but started improving in the late 1970s and 1980s and reached a level of 5.9:2.4:1 in 1991-92. However, decontrol of P and K fertilizers and a steep increase in prices in 1992 resulted in a decline in their consumption and a consequent imbalance in fertilizer use. The NPK ratio, which was 5.9:2.4:1 during 1991-92, widened to 9.7:2.9:1.0 during 1993-94 and reached a level of 10.0:2.9:1 in 1996-97. However, due to the government's concerted efforts, such as increasing concessions on phosphatic and potassic fertilizers and marginally increasing the price of urea in 1997, the NPK ratio improved, and it reached a level of 4.3:2.0:1.0 in 2009-10. However, recent policy changes, such as the introduction of NBS in 2010 and a reduction in subsidies on P and K fertilizers in the post-NBS period, led to a worsening of the NPK ratio. It reached a level of 6.7:3.1:1 in 2011-12 and became worse (8.7:3.4:1) in 2012-13. There are also wide interregional and interstate disparities in N:P:K ratios.There is a high degree of inequality in fertilizer consumption among crops. Rice, wheat, and sugarcane are the prime beneficiaries, with rice being the largest user of fertilizer (about one-third of total consumption), followed by wheat (24.2 percent). Fruits, vegetables, and sugarcane combined represent another 11 percent of fertilizer use. Given the importance of food grains and the government's recent efforts to increase their production, these crops have the potential to stimulate fertilizer use. In addition, the rising demand for high-value crops (fruits and vegetables), due to increasing income levels, urbanization, and changing lifestyles, is also expected to increase the demand for fertilizer, as these crops are fertilizer-intensive.Fertilizer intensity measured as average kg per hectare does not follow the exact same pattern across crops; intensity tends to be higher on sugarcane (234.9 kg/ha), vegetables (253.8 kg/ha), cotton (183 kg/ha), and fruits (158.6 kg/ha) and lower on cereals (rice 129.2 kg/ha and wheat 162.6 kg/ha) and pulses. Farmers growing input-intensive crops are the main beneficiaries of fertilizer use.Fertilizer consumption also varies across farm sizes, but there is a fair degree of inter-farm size equity in fertilizer consumption. The share of small and marginal farmers in gross cropped area was 44.4 percent and they consumed 52.8 percent of the total fertilizer used in the country in 2006-07. On the other hand, the share of medium and large farmers in gross cropped area was nearly one-third and they consumed about 25 percent of the total fertilizer used in the country.The relationship between fertilizer nutrient prices and paddy and wheat prices during the past three decades reveals that with the steady increase in the procurement prices of crops and almost stable fertilizer prices, the profitability has increased for all three nutrients. However, profitability of P and K use declined significantly after the decontrol of their prices in 1992 and the introduction of NBS for P and K fertilizers in 2010. Furthermore, the response ratio (kg grain/kg nutrient) in food grain crops in irrigated areas in India has substantially declined during the past four decades, from 13.4 kg of grain per kg of nutrient in 1970 to 3.7 kg of grain per kg of nutrient in 2005. The continuous application of higher amounts of N, lower doses of P, and organic manure has led to the emergence of secondary and micronutrient (Zn, B, Fe, Mn, Mo) deficiencies in Indian soils.Fertilizer consumption in India has generally exceeded the domestic production in both nitrogenous and phosphatic fertilizers except for a few years. The entire requirement of potassic fertilizers is met through imports, as India does not have commercially viable sources of potash. During the past decade, due to low/no addition in domestic capacity coupled with a rise in demand for fertilizers, imports have increased significantly in the 2000s. The main fertilizer products imported in India are urea (7.8 million tons), DAP (6.9 million tons), and MOP (about 4 million tons). Urea imports have increased significantly during the past six or seven years. This increase in imports and rising international prices of urea and other fertilizer products have led to a substantial increase in fertilizer subsidies in the country. Oman (36.2 percent), China (22 percent), Iran (17.2 percent), and CIS (12.9 percent) were major exporters of urea to India during TE2011-12.To ensure an adequate supply of fertilizers in all regions/areas of the country, the distribution and movement of fertilizers is controlled under the Essential Commodities Act 1955 (ECA) to bridge supplies in underserved areas. Urea is under partial movement and distribution control of the government, and 50 percent of the indigenous production of urea is regulated by issue of movement orders to the manufacturers for dispatch to the states on a month-to-month basis, keeping in view the assessed requirement. Twenty percent of decontrolled fertilizers produced/imported in India are under movement controls.Imports of nitrogenous fertilizers are canalized through state trading enterprises, while imports of P and K fertilizers and raw materials/intermediates have been decontrolled and placed under Open General License (OGL). Currently, more than 40 percent (up from about 13 percent in the early 2000s) of total fertilizer nutrients used in India is sourced through imports. The capacity to produce SUMMARY | APRIL 2010 more fertilizer in the country is currently limited due to the availability and/or cost of raw materials/feedstocks, and installed capacity has remained stagnant during the past decade. On the fertilizer supply side, the major deficiencies that constrained sustained rapid growth in fertilizer production include shortage of raw materials, intermediates, and feedstocks, such as natural gas for urea production and rock phosphate and phosphoric acid for phosphates; lack of a consistent long-term policy; lack of working capital in the distribution channels; and poor physical infrastructure in some regions of the country. Due to constraints in raw material availability, the indigenous production of fertilizers has been decreasing, while imports have risen.Private trade accounts for about 65 percent of the total fertilizer distributed in the country, followed by institutional agencies, including cooperatives, at 35 percent; marginal quantities are distributed through manufacturers' own outlets. Among the institutional agencies, cooperatives happen to be the main distribution agencies.The demand for fertilizer depends on price factors, such as the price of outputs, the price of fertilizer, and the prices of other inputs that substitute for or complement fertilizer, and non-price factors, including production and market infrastructure. The non-price factors such as irrigation and high-yielding varieties were more powerful in influencing demand for fertilizer compared with price factors. Within price factors, the price of fertilizers had an adverse effect on fertilizer consumption and was more powerful than output price.The fertilizer subsidy has been one of the most hotly debated issues in the country over the past two decades. Fertilizers, after oil and food, account for the third biggest share of India's total subsidy bill, and several attempts have been made to contain subsidies. However, estimates show that fertilizer subsidy has increased significantly over the years in the post-reforms period, from Rs. 5185 crore in 1991-92 to Rs. 70012 crore in 2011-12, representing an increase of more than 13 times. Fertilizer subsidy in India as a percentage of the GDP varied from 0.47 in 2002-03 to 1.9 percent in 2008-09 and declined to about 0.8 percent in 2011-12. However, a steep increase in the cost of inputs to fertilizer production, high import prices, and constant farmgate prices have led to a substantial increase in fertilizer subsidy in recent years. For example, fertilizer subsidy increased by over 5.5 times between TE2003-04 and TE2010-11, from Rs. 11853 crore to over Rs. 66000 crore. It is well known that the fertilizer subsidy has helped increase the availability and consumption of fertilizers at affordable prices and thereby increased agricultural production, but it has also led to some unintended negative consequences, such as imbalanced use of nutrients, declining fertilizer use efficiency, adverse impacts on land and water resources in certain areas, and unsustainable levels of subsidy.There is a debate about whether fertilizer subsidy benefits farmers or the fertilizer industry and whether domestic industry is overprotected from world markets. The general perception that about one-third of the fertilizer subsidy goes to the fertilizer industry is misleading because the underlying assumptions do not hold true. The world fertilizer markets and trade flows are highly concentrated and volatile, and Indian imports have a significant impact on world prices. Moreover, with the shift from the earlier cost-plus based approach to import parity pricing (IPP), the Indian fertilizer industry has been exposed to world competition, which would drive out inefficient units. Empirical evidence also shows that the perception of the domestic urea industry being overprotected and less efficient than imports does not hold true because the average subsidy per ton of imported urea is much higher than that for indigenously produced urea.On the issue of whether fertilizer subsidy is distributed equitably across crops, states, and farm classes, the results indicate that it is concentrated in a few states. Interstate disparity in fertilizer subsidy distribution is still high, though it has declined over the years. Rice, wheat, sugarcane, and cotton account for about two-thirds of the total fertilizer subsidy. However, the study shows that fertilizer subsidy is more equitably distributed among farm sizes. The small and marginal farmers have a larger share in fertilizer subsidy in comparison to their share in cultivated area. The benefits of fertilizer subsidy have spread to unirrigated areas as the share of area treated with fertilizers and share of unirrigated areas in total fertilizer use have also increased. A reduction in fertilizer subsidy is, therefore, likely to have adverse impacts on farm production, income of small and marginal farmers, and unirrigated areas, because they do not benefit from higher output prices but do benefit from lower input prices.It is evident that withdrawal of subsidies will make farming unprofitable, particularly for small and marginal farmers and those in less developed states/regions. Therefore, there is a need for subsidizing fertilizers for small and marginal farmers as well as for less developed regions. There is a need to contain these subsidies without hurting millions of smallholders, including tenant cultivators who produce for self-consumption and have no or very small marketed surplus. These farmers do not benefit from high output prices but higher fertilizer prices would certainly reduce their income. Targeting and rationing are important tools to contain subsidies and ensure that they are largely provided to those farmers/regions/crops where fertilizer use is constrained by high prices, insufficient institutional credit support, and low productivity levels. Rationing, for example by limiting the volume of subsidized fertilizer that a farmer can get, is a better option compared to targeting and is also more acceptable politically and administratively. Rationing will provide proportionately greater benefits to small and marginal farmers compared to large farmers, and it will promote fertilizer consumption on small and","tokenCount":"13981","images":["2050765791_1_1.png","2050765791_2_1.png","2050765791_5_1.png","2050765791_6_1.png","2050765791_7_1.png","2050765791_11_1.png","2050765791_12_1.png","2050765791_12_2.png","2050765791_13_1.png","2050765791_13_2.png","2050765791_14_1.png","2050765791_14_2.png","2050765791_17_1.png","2050765791_17_2.png","2050765791_21_1.png","2050765791_21_2.png","2050765791_21_3.png","2050765791_22_1.png","2050765791_23_1.png","2050765791_23_2.png","2050765791_23_3.png","2050765791_23_4.png","2050765791_24_1.png","2050765791_25_1.png","2050765791_26_1.png","2050765791_30_1.png","2050765791_30_2.png","2050765791_30_3.png","2050765791_31_1.png","2050765791_31_2.png","2050765791_31_3.png","2050765791_32_1.png"],"tables":["2050765791_1_1.json","2050765791_2_1.json","2050765791_3_1.json","2050765791_4_1.json","2050765791_5_1.json","2050765791_6_1.json","2050765791_7_1.json","2050765791_8_1.json","2050765791_9_1.json","2050765791_10_1.json","2050765791_11_1.json","2050765791_12_1.json","2050765791_13_1.json","2050765791_14_1.json","2050765791_15_1.json","2050765791_16_1.json","2050765791_17_1.json","2050765791_18_1.json","2050765791_19_1.json","2050765791_20_1.json","2050765791_21_1.json","2050765791_22_1.json","2050765791_23_1.json","2050765791_24_1.json","2050765791_25_1.json","2050765791_26_1.json","2050765791_27_1.json","2050765791_28_1.json","2050765791_29_1.json","2050765791_30_1.json","2050765791_31_1.json","2050765791_32_1.json","2050765791_33_1.json","2050765791_34_1.json","2050765791_35_1.json","2050765791_36_1.json","2050765791_37_1.json","2050765791_38_1.json","2050765791_39_1.json","2050765791_40_1.json","2050765791_41_1.json"]}

data/part_2/0195442968.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"615ee549a04228131d98df149caeec74","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/049ace90-4247-4c25-b989-b013c8832167/retrieve","description":"This Data Note describes the coverage of a set of key nutrition and health interventions. The findings here are based on data from the National Family Health Survey 2015-2016. Indicators to capture the coverage of the interventions here are all calculated for women (15-49 years) with a child under five years of age. All analyses are based on the last-born child for each woman and indicator definitions are provided in Annex 1 of this Note. For some indicators, age categories may vary.","id":"-1706619775"},"keywords":[],"sieverID":"5808db57-1425-4a06-8e08-17934a19d9eb","pagecount":"12","content":"Nutrition-specific interventions are aimed at improving the proximal food, health, and care environment for women and children during the first 1000 days. They can help improve maternal and child health, improve diets and other nutrition practices, and reduce infections. These interventions span pregnancy, postnatal, and early childhood periods and include food and micronutrient supplementation, nutrition education and/or counselling, growth monitoring and promotion, as well as routine immunization, deworming, and care during illness. At 90% coverage, these interventions can contribute to 20% reduction in stunting and 61% reduction in severe wasting 1 . India's policy framework for health and nutrition is robust and includes most evidence-based nutrition and health interventions. Two large-scale national program platforms -the Integrated Child Development Services and the National Health Mission together provide the public sector delivery platforms with the mandate to deliver these interventions across the country. India's efforts at scaling up nutrition interventions are now also supported by the National Nutrition Mission. This Data Note describes the coverage of a set of key nutrition and health interventions. The findings here are based on data from the National Family Health Survey 2015-2016. Indicators to capture the coverage of the interventions here are all calculated for women (15-49 years) with a child under five years of age. All analyses are based on the last-born child for each woman and indicator definitions are provided in Annex 1 of this Note. For some indicators, age categories may vary.","tokenCount":"239","images":["-1706619775_1_1.png","-1706619775_2_1.png","-1706619775_2_2.png","-1706619775_3_1.png","-1706619775_3_2.png","-1706619775_4_1.png","-1706619775_4_2.png","-1706619775_5_1.png","-1706619775_5_2.png","-1706619775_6_1.png","-1706619775_6_2.png","-1706619775_7_1.png","-1706619775_7_2.png","-1706619775_8_1.png","-1706619775_8_2.png","-1706619775_9_1.png","-1706619775_9_2.png","-1706619775_12_1.png"],"tables":["-1706619775_1_1.json","-1706619775_2_1.json","-1706619775_3_1.json","-1706619775_4_1.json","-1706619775_5_1.json","-1706619775_6_1.json","-1706619775_7_1.json","-1706619775_8_1.json","-1706619775_9_1.json","-1706619775_10_1.json","-1706619775_11_1.json","-1706619775_12_1.json"]}

data/part_2/0239731777.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"1b4d28e38e713aab312a450513f96099","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/06d0b80b-f4bc-4eb9-96d1-aa208d67d98b/retrieve","description":"This summary note is an excerpt from the chapter on Madagascar that will appear in the peer-reviewed IFPRI monograph, East African Agriculture and Climate Change: A Comprehensive Analysis.","id":"-274896276"},"keywords":[],"sieverID":"25d83ae3-ae08-4a5f-b0a6-fe4b02b380ba","pagecount":"2","content":"Madagascar is an island in the southwestern Indian Ocean off the coast of Mozambique. The topography of the country is varied and uneven, with some regions dominated by plains and plateaus extending into vast delta areas.Average annual temperatures range between 23˚C and 27˚C. The eastern and northwestern coasts are dominated by southeasterly trade winds, which produce annual precipitation ranging from 2,000 mm to 3,700 mm. The central plateau and the western coast receive rain from the monsoon winds, with precipitation ranging from 1,000 mm to 1,500 mm per year. The south receives as little as 350 mm of rain in a year.Only about 5 percent of the land area is cultivated at any given time, of which 16 percent is irrigated. In addition to providing livelihoods for two-thirds of the population, agriculture contributes 29 percent of Madagascar's GDP.Most farmers practice subsistence agriculture, growing rice, cassava, bananas, maize, and sweet potatoes. Yields are generally low and not keeping up with population growth. For example, per capita rice production declined from 1.2 tons in 1975 to 0.9 tons in 2006, despite a 15 percent growth in rice area and and 35 percent increase in rice yield over the same period. Slash-and-burn (shifting) agriculture is a common practice, resulting in environmental degradation and loss of forest area.Extreme weather events threaten agricultural productivity. For example, in early 2000, a series of three particularly devastating cyclones affected more than a million people and caused nearly $85 million in damage to agricultural infrastructure.As a basis for our analysis, we used four downscaled global climate models (GCMs) from the IPCC AR4. For the northern part of the country, the models projected either no change or an increase in annual precipitation. In the south, rainfall either remains relatively unchanged or decreases (the range of increase or decrease in both regions falls within 50-200 mm). Because most of the country's key crops are rainfed, reduced rainfall has negative consequences for agricultural production. More irrigation investment is required or farmers may need to switch to crops that are suitable with less rainfall. Less rainfall would mean more cassava mosaic disease, which would reduce yield. In short, without adaptation or technological improvements, food insecurity may increase. Since the south already receives much less rain than the north, this reduction could be a significant blow to the cultivation of many of that region's crops.The climate models also project increased temperatures of 0.5-3°C. The CSIRO model predicts a median temperature increase of 1.3°C, which is considerably cooler than the increase of 2.1°C projected by the MIROC model. Furthermore, the results differ in  Copyright © 2012 International Food Policy Research Institute. All rights reserved. To obtain permission to republish, contact [email protected] their geographic variation. High temperatures promote evapotranspiration, thus reducing soil moisture and increasing soil degradation. High temperatures may also foster more pests and disease, as in the case of cassava mosaic disease.The Decision Support System for Agrotechnology Transfer (DSSAT) crop modeling software was used to compute yields for six crops in rainfed and irrigated systems under current temperature and precipitation regimes, and then again for the 2050 climate projections. The maps illustrate projected yield changes for rainfed rice. The results indicate remarkably similar changes across all GCMs. Each shows losses throughout the island. In the CSIRO model, most of the losses appear to be less than 25 percent, while in the MIROC model most losses are greater than 25 percent. But there is a noticeable area of yield gain near Antananarivo, which other maps suggest is devoted to irrigated rice. There is also a much smaller patch in the north that shows yield gain. Both of these patches are in high elevations with colder temperatures. Rice yields are hampered by the cold. These maps illustrate that with climate change rice yields could increase in selected locations. While it is difficult to determine whether the yield gains in these high-production areas will offset the losses, many of which occur in low-production areas, we can see that there will likely be winners and losers among rice growers as a result of climate change, and there may be pressure for new settlements or more intensive cultivation in those areas where productivity is expected to rise.For rainfed maize, the DSSAT results show climate change resulting in scattered areas of yield increase but with overall yield declines.The research used the IMPACT global model for food and agriculture to estimate the impact of future GDP and population scenarios on crop production and staple consumption, which can be used to derive commodity prices, agricultural trade patterns, food prices, calorie consumption, and child malnutrition. Three GDP-per-capita scenarios were used -an \"optimistic scenario\" with high per capita income growth and low population growth, a pessimistic scenario with low per capita income growth and high population growth, and an intermediate scenario. Per capita GDP increases in all scenarios, with the pessimistic scenario predicting a gradual rise to $650 by 2050, while the optimistic scenario projects that per capita GDP will increase to $1,740.The results indicate that rice yields will more than double between 2010 and 2050 as technology advance outweighs the negative effects of climate change. With harvested area projected to increase only slightly through that period, production is projected to more than double. Rising internal demand for the crop will lead to increased imports in the intermediate and optimistic scenarios despite increases in international prices of rice (by 55 percent -averaged across all scenarios and climate models -between 2010 and 2050); maize (by 101 percent) and cassava (5 percent.)For cassava, the analysis showed increased production, harvest area, and yields, as well as rising cassava prices. However, there is also a clear trend of increased imports between 2010 and 2050.For maize, yields rise by around 60 percent but the area under cultivation is projected to drop by around 40 percent. Both of these changes are true regardless of scenario and regardless of climate model, since there is virtually no variation between the results for any of these. This leads to a slight increase in production until 2020, and then a decline, leaving production in 2050 around 10 percent lower than in 2010. As in the case of cassava and rice, the country will experience declining net exports and face higher prices by 2050. As a result, Madagascar is projected to go from being an exporter to an importer of maize.A projected rise in available calories per capita will increasingly come from trade rather than domestic production.The IMPACT model projects that the number of malnourished children will rise slightly before it declines in both the optimistic and baseline scenarios. The numbers increase more dramatically in the pessimistic scenario and do not return to the 2010 levels by 2050.To enable farmers to adapt to the impacts of climate change, policymakers should:• fund agricultural research and extension institutions' development and promulgation of crop varieties better suited to a range of potential future climate conditions;• develop new varieties and substitutes for maize (including grains that are more resilient in hot, dry conditions, possibly sorghum or millet); and• improve existing weather forecasting systems and international cooperation on meteorological issues.","tokenCount":"1181","images":["-274896276_1_1.png","-274896276_1_2.png","-274896276_1_3.png","-274896276_1_4.png"],"tables":["-274896276_1_1.json","-274896276_2_1.json"]}

data/part_2/0240356654.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"f6fa385a37ca52d7c811558603181085","source":"gardian_index","url":"http://www.fao.org/3/ca5526en/ca5526en.pdf","description":"Fiscal policies have multiple effects on the economy and, therefore, on the agricultural sector and rural economies within it. This paper analyzes agricultural taxes and agricultural expenditures; it also looks into different forms of public policy support for agricultural producers and follows the Producer Support Estimates (PSE) methodology already applied to the Latin America and the Caribbean (LAC) region. This methodology combines fiscal support and other forms of support, such as those deriving from trade policies. Finally, this paper presents a summary of evaluation studies about different types of public expenditure for agriculture and draws some conclusions.","id":"240344457"},"keywords":null,"sieverID":"2efac83f-1b24-4c69-85f4-4ab1cc763daf","pagecount":"0","content":"Fiscal policies have multiple effects on the economy and, therefore, on the agricultural sector and rural economies within it. This paper analyzes agricultural taxes and agricultural expenditures; it also looks into different forms of public policy support for agricultural producers and follows the Producer Support Estimates (PSE) methodology already applied to the Latin America and the Caribbean (LAC) region. This methodology combines fiscal support and other forms of support, such as those deriving from trade policies. Finally, this paper presents a summary of evaluation studies about different types of public expenditure for agriculture and draws some conclusions.","tokenCount":"97","images":null,"tables":null}

data/part_2/0258339796.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"425147058c6b76655bd39a4f32cffd07","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/0af44cc9-1da0-4872-adc2-bfe315c18abd/retrieve","description":"Hunger and acute child malnutrition are increasingly concentrated in fragile countries and civil conflict zones. According to the United Nations, Yemen’s civil war has caused the world’s worst humanitarian crisis in recent history. We use high-frequency panel data and district fixed-effects and household fixed-effects models to estimate the impact of civil conflict on child nutrition. Our results indicate that an increase by one standard deviation in civil conflict intensity translates into an increase in the prevalence of acute child malnutrition by at least 0.7 percentage points if measured by weight-for-height z-scores and by at least 1.7 percentage points if measured by mid-upper arm circumference z-scores. In mid-December 2018, Yemen’s main warring parties agreed to a ceasefire for the contested port city of Hodeida and to allow humanitarian aid to be shipped in and distributed through protected corridors. While the recent agreements are an important, first step to tackle the humanitarian crisis, the road to a sustainable peace agreement will certainly be long and bumpy. Relative stability could soon open a window of opportunity for targeted interventions to support recovery in Yemen. Against this background, our analysis suggests that unconditional cash transfers can be an effective tool in situations of complex emergencies. Our estimation results show that cash transfers can mitigate the detrimental impact of lingering civil conflict on child nutritional status in Yemen on a large scale. Our results also reveal that the regularity of transfer payments influence the magnitude of the mitigation effect, as regular assistance allows beneficiary households to smoothen their food consumption and other demands influencing child nutrition outcomes.","id":"1260011009"},"keywords":["acute child malnutrition","unconditional cash transfers","civil conflict","fixed-effects"],"sieverID":"8856c407-34cc-4064-9136-0bedf770930a","pagecount":"33","content":"Hunger and acute child malnutrition are increasingly concentrated in fragile countries and civil conflict zones. According to the United Nations, Yemen's civil war has caused the world's worst humanitarian crisis in recent history. We use high-frequency panel data and district fixed-effects and household fixed-effects models to estimate the impact of civil conflict on child nutrition. Our results indicate that an increase by one standard deviation in civil conflict intensity translates into an increase in the prevalence of acute child malnutrition by at least 0.7 percentage points if measured by weight-for-height z-scores and by at least 1.7 percentage points if measured by mid-upper arm circumference z-scores.In mid-December 2018, Yemen's main warring parties agreed to a ceasefire for the contested port city of Hodeida and to allow humanitarian aid to be shipped in and distributed through protected corridors. While the recent agreements are an important, first step to tackle the humanitarian crisis, the road to a sustainable peace agreement will certainly be long and bumpy. Relative stability could soon open a window of opportunity for targeted interventions to support recovery in Yemen. Against this background, our analysis suggests that unconditional cash transfers can be an effective tool in situations of complex emergencies. Our estimation results show that cash transfers can mitigate the detrimental impact of lingering civil conflict on child nutritional status in Yemen on a large scale. Our results also reveal that the regularity of transfer payments influence the magnitude of the mitigation effect, as regular assistance allows beneficiary households to smoothen their food consumption and other demands influencing child nutrition outcomes.iii  Hunger and acute child malnutrition are increasingly concentrated in fragile countries and civil conflict zones (IFPRI 2018;von Grebmer et al. 2015). About 1.35 billion children and adolescents under the age of 18 live in a country affected by civil conflict, and almost 357 million of them live in a conflict zone (Bahgat et al. 2017). Currently, the most dramatic case in point is Yemen. Conservative estimates suggest that the death toll has reached nearly 50,000 people among active participants in battles between January 2018 and July 2018, and that there were at least 6,500 civilian casualties, including more than 1,600 children, between March 2015 and July 2018 (Fahim 2018). More than two million people were forced to flee their homes (UNHCR 2018). In October 2018, the United Nations Office for the Coordination of Humanitarian Affairs (UN-OCHA) warned that half the population-some 14 million people-could soon be on the brink of famine and completely relying on humanitarian aid for survival (Reuters 2018). According to the international NGO Save the Children, 84,700 children under the age of five may have died from acute malnutrition between April 2005 and October 2018, as a result of the conflict (BBC 2018a). Although the precise estimates are contestable, there is broad consensus that Yemen's current situation is extremely alarming in terms of widespread starvation, especially among young children.Evidence on the detrimental consequences of civil conflict on human capital, particularly when exposed at early ages, is strong (e.g., Alderman et al. 2006;Bundervoet and Verwimp 2009;Akresh et al. 2012aAkresh et al. , 2012b;;Minoiu and Shemyakina 2012;Domingues and Barre 2013;Dagnelie et al. 2018). However, rigorous empirical research on how to mitigate the adverse impact of civil conflict is limited. An exception is Tranchant et al. (2018), who assessed the effectiveness of food assistance in Mali. Using a quasi-experimental impact evaluation method, the authors find that food assistance has a protective effect among food-insecure population in conflict settings. Other researchers discussed the role of different interventions in fragile contexts. For example, Rossi et al. (2006) explore the role of health and nutrition interventions in the Democratic Republic of Congo, but the lack of baseline data or credible counterfactuals make causal inference about the effectiveness of the interventions examined difficult, if not impossible.Despite common concerns of cash diversion, including for war purposes, unconditional cash transfer programs have become increasingly popular in fragile contexts, including in active conflict zones, and have been employed in Afghanistan, Chechnya, Democratic Republic of Congo, Nigeria, Pakistan, Somalia, Syria, and Yemen (e.g., Economist 2018a;HPN 2012). The spread of mobile phones and mobile-money transfer systems, such as through the traditional Muslim hawala networks, has helped to leverage the implementation of cash transfers in conflict-prone settings. Operational evaluations of cash transfers in fragile and conflict-affected areas suggest that the risks of diversion, corruptible behavior, and insecurity concerns are largely overstated or at least are not greater than in the case of in-kind transfers, such as food assistance (Chene 2010;Gordon 2015;Bailey and Harvey 2015;Doocey and Tappis 2016). Moreover, cash transfers have been found to facilitate the reintegration of ex-combatants into their local communities (Gilligan et al. 2012;Annan and Blattman 2016) and, in some settings, reduce the risk of civil conflict in fragile communities (Crost et al. 2016). However, there is little scientific evidence on the effectiveness of unconditional cash transfers on food security and nutrition outcomes in fragile contexts and conflict zones. Increasing experience among program implementers suggests that cash transfers do have a substantial mitigation effect, counteracting the detrimental impact of civil conflict (ODI 2015;World Bank 2011).The objective of this paper is twofold. First, we quantify the detrimental impact of civil conflict increasing acute child malnutrition in Yemen. We exploit quarterly variation in conflict events at the district level to assess the impact on weight-for-height z-scores (WHZ) and mid-upper arm circumference z-scores (MUACZ) of children under five years of age. According to our knowledge, despite the alarming nutritional emergency, the detrimental impact on child nutrition has not been rigorously quantified. Our results indicate that an increase by one standard deviation in civil conflict intensity translates into an increase in the prevalence of acute child malnutrition in our sample by at least 0.7 percentage points, if measured by WHZ, and by at least 1.7 percentage points, if measured by MUACZ.Second, we assess the mitigation effect of the national cash transfer program of the Social Welfare Fund (SWF). Although our data do not offer an experimental design, the longitudinal nature of the dataset allows us to control for unobserved household-level heterogeneity and seasonal variations. We use data from the 2012-13 National Social Protection Monitoring Survey (NSPMS) that provides observations from four survey rounds over a period of one year. Households' beneficiary status was determined prior to the observation period of our study and remains fixed throughout that period, independent of the households' living conditions and the nutritional status of children in the household. The results of our household fixed-effects (FE) model estimations suggest that the cash transfer program mitigates the detrimental impact of civil conflict on child nutrition. This holds for children both in households that have been beneficiaries for a long time and in households that were newly enrolled. Modifications of our model estimations further suggest that the regularity of transfer payments matters for the size of the mitigation effect. Finally, we discuss the policy implications of our findings that may strengthen the mitigating role of cash transfers in fragile countries and civil conflict zones and particularly in Yemen.The rest of the paper proceeds as follows. Section 2 provides the context of our study. Section 3 presents the data and descriptive statistics. Section 4 explains the methods of the econometric analysis, and Section 5 presents and interprets the estimation results. Section 6 concludes and discusses the policy implications of our findings.Yemen's current civil war began with the 2011-12 revolution against President Ali Abdullah Saleh, who became the first president of a unified north and south Yemen in 1990. The Yemeni revolution shortly followed the Tunisian and Egyptian revolutions and other \"Arab Spring\" protests throughout the Middle East and North Africa (MENA) region. In Yemen, mass protests in early 2011 demanding better governance, people's voice, and economic opportunities evolved into civil conflict, including armed clashes between protesters and government security forces (WFP 2012). Political resistance and mass protests quickly spread from the capital, Sana'a, throughout the country, melding with three ongoing prolonged conflicts (Economist 2013;2014;2015;2017a). First, since 2004 military troops frequently battled rebels from the Houthi clan to regain full control over Yemen's northwest. The Houthi insurgency heated up in 2009 but quieted the following year after a ceasefire was signed (Taylor 2015). Second, in the south, since 2008 clashes of government security forces with protesters demanding political reform and partial secession (anew) from the north became more frequent. In parallel, government also struggled to control a range of jihadist groups, lawless tribes, secessionist efforts, and bandits in parts of the rural south (WFP 2012). Third, from 2009 Al-Qaeda in the Arabian Peninsula (AQAP) gained influence and launched terrorist attacks across Yemen, including in Sana'a.The 2011-12 revolution ended with the signature of a power transition agreement in early 2012 (World Bank et al. 2012). Abdrabbuh Mansur Hadi, who was Saleh's former vice president, was hastily elected as president in February 2012. The new government struggled to unite Yemen's fractious political landscape and to fend off threats both from Houthi militants and AQAP (Economist 2017a(Economist , 2017b(Economist , 2017c)). Although the number of conflict events declined in the second half of 2012 and in the first half of 2013, political instability remained, terrorist attacks continued, and violent clashes between different militia groups and with government security forces flared across the country in the second half of 2013 (Figure 2.1). In the fall of 2014, Houthi fighters swept into Sana'a and quickly gained control over strategic points in the city (Economist 2014;El-Naggar 2015). The Houthis refrained from an immediate coup d'état, but forced President Hadi to negotiate a 'unity government' with other political factions. In early 2015, Hadi along with his ministers resigned after his presidential palace and private residence came under attack. The Houthis declared themselves in control of the government, dissolved the Parliament, and installed the interim Supreme Revolutionary Committee. Hadi first escaped to the port city of Aden in Yemen's south, before fleeing to Riyadh in Saudi Arabia (BBC 2015). In Aden, Hadi declared that he remained the legitimate president of Yemen, proclaimed the city as the temporary capital, and called on loyal government officials and military officers to rally to him. Subsequently, Yemen's civil war erupts between loyalists of Hadi's government and the Houthis. A coalition of Arab countries led by Saudi Arabia began military operations against the Houthi fighters. Since early 2018, the Southern Transitional Council-a secessionist organization with its headquarters in Aden-joined the conflict in fighting against Hadi's government (Dahlgren 2018;Economist 2018).The Yemeni civil war has lasted for over three years and has had devastating consequences for the population, particularly for young children. The United Nations declared the crisis in Yemen as \"the world's worst humanitarian crisis\" (UN 2018) and warned of a looming large-scale famine. Basic infrastructure, including sanitation and drinking water systems, collapsed. A severe cholera epidemic began in September 2016, the largest documented cholera epidemic of modern times.Between September 2016 and March 2018, there were more than 1.1 million suspected cholera cases with 2,400 deaths due to the disease (Camacho et al. 2018). 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 On 13 December 2018, Yemen's main warring parties agreed to an immediate ceasefire for the contested port city of Hodeidah, which is the gateway for the bulk of humanitarian aid coming into the country (Wintour and McKernan 2018). After skirmishes immediately following the agreement, the ceasefire was reported to have taken effect on 18 December (BBC 2018b).The observation period of our empirical analysis is from July 2012 to October 2013. This falls within the transitional period of Yemen's government, when civil conflict was less intense before the outbreak of the current civil war (Figure 1). It was also the time when a window of opportunity opened for increasing and expanding public assistance to disincentivize individuals to engage in civil conflict for economic reasons and to protect people's living conditions from the impact of possible conflict intensification and other shocks (e.g., Berman et al. 2011;Blattman and Ralston 2015;Maystadt and Ecker 2014).In this analysis, we combine high-frequency, longitudinal household survey data with georeferenced conflict event data. The survey data are from the Yemen National Social Protection Monitoring Survey (NSPMS). The conflict event data were compiled by the GDELT Project (2013;2015). 1 We merged the conflict event data to the household survey data at the administrative district level and by survey round.The NSPMS was implemented by the United Nations' Children Fund (UNICEF), the International Policy Centre for Inclusive Growth (IPC-IG); Yemen's Ministry of Planning and International Cooperation; and Interaction in Development, with support from the Yemen NSPMS Technical Committee (IPC-IG et al. 2014a). The survey had two main objectives: (1) to examine the living conditions of poor households in Yemen after the 2011-12 revolution, and (2) to assess the targeting of the national cash transfer program of the Social Welfare Fund (SWF) and the program's impact on a variety of development indicators, including child nutrition.The SWF cash transfer program is an unconditional cash transfer program of the Government of Yemen, supported by the World Bank. It provides financial assistance to citizens who are temporarily or permanently unable to sustain themselves and whose families are not able to support them financially. The beneficiaries include socially-disadvantaged people, such as the elderly, disabled people, and orphans, and economically-disadvantaged people, such as unemployed men and single, widowed, or divorced women (IPC-IG et al. 2014a).The NSPMS was designed to provide nationally representative estimates of key indicators of households' living conditions (when using survey sampling weights) and to accommodate the assessment of the SWF cash transfer program (IPC-IG et al. 2014a). The survey oversampled at the district level both the poor population and SWF beneficiaries and potential beneficiaries. Data collection for the NSPMS started in October 2012 and was completed in September 2013. The survey provides data for a balanced sample of 6,396 households. Each of these households were interviewed on a quarterly basis over the 12 months of the survey period. The balanced sample provides household data from 19 governorates out of the 21 mainland governorates and for 218 districts out of the 331 mainland districts. Districts in the two northern governorates of Sa'ada and Al-Jawaf-the stronghold of the Houthi rebels-suffered complete attrition, as the enumerator teams were unable to visit the enumeration areas in later survey rounds because of serious security concerns (IPC-IG et al. 2014a). A detailed description of the sample design and sampling procedures can be found in IPC-IG et al. (2014b).The balanced sample includes 3,644 beneficiary households, who were defined as households that had ever received a SWF cash transfer payment, and 2,752 non-beneficiary households, who were never enrolled in the cash transfer program. The beneficiary households include 2,188 \"old beneficiary\" households and 1,175 \"new beneficiary\" households. These beneficiary groups were defined using a district-based census of poor people in 2008. Old beneficiaries are those who were already enrolled in the program at the time of the census, while new beneficiaries are those who did not receive benefits at that time, but were identified as potential beneficiaries. New beneficiaries were gradually enrolled in the program from 2011 onward. Some of the new beneficiaries received a first cash transfer payment in early 2011, but their payments were suspended during the 2011-12 revolution due to a lack of funds. Their payments resumed during the observation period of our study. Other new beneficiaries received their first payment after the first round of NSPMS data collection (IPC-IG et al. 2014a). Transfer payments for old beneficiaries were generally not suspended during the 2011-12 revolution.The beneficiary status of 281 beneficiary households (7.7 percent) is not specified in the released NSPMS dataset. However, our data checks strongly suggest that these households were misclassified as beneficiaries in the survey sample design or excluded from the program after the sample selection and during the survey. About one-fourth of them (24.6 percent) reported to have not received any transfer payment during the observation period; most of them (59.8 percent) received only one payment; and none of them received all four regular payments throughout the observation period of the survey. We therefore drop these households from our sample to reduce estimation biases from \"contamination\" of our beneficiary groups.Our analysis uses indicators of acute child malnutrition (i.e., \"wasting\"), as our study focus is on the immediate nutritional impact of civil conflict. Moreover, the observation period of the underlying data is too short to credibly track the impact on long-term child nutrition indicators that measure linear growth faltering (i.e., \"stunting\" or low height-for-age). There are two common child nutrition indicators of wasting: (1) weight-for-height and (2) mid-upper arm circumference (MUAC). Wasting implies a recent or continuing severe process of weight loss, which is often associated with acute starvation or severe disease (WHO 1995). Weight-for-height and MUAC measures are accurate predictors of mortality among children with severe acute malnutrition (Chiabi et al. 2017). The two indicators should be used as complementary-rather than alternative-measures to diagnose the prevalence of acute malnutrition (Grellety and Golden 2016a).For all survey rounds, the NSPMS included an anthropometry module that provides measurements of the body height and weight of all children who permanently lived in the interviewed households and were aged between 0 and 59 months at the time of the survey rounds. The module also provides MUAC measurements for these children if they were 6 months or older. Hence, these anthropometric data allow the tracking of the nutritional status of individual children over the 12-month observation period. We used the height and weight measurements in combination with information on child sex, age, edema signs, and positioning for height measurement to compute weight-for-height z-scores (WHZ) by applying a routine developed by Leroy (2011) for the Stata software package. We dropped children from our sample if they had missing WHZs in any survey round; if their WHZ in any round was outside a biologically plausible range (defined as -5 ≤ WHZ ≤ 5, following the classification of outlier measurements recommended by the World Health Organization); or if their height in any round was lower than in any previous survey round (as shrinking in children is biologically impossible). We use MUAC z-scores (MUACZs) as provided by the released NSMPS dataset. We do not consider the MUACZs of children if they had missing MUACZs in any survey round or their MUACZ in any round was outside a biologically plausible range (defined as -5 ≤ MUACZ ≤ 5). Children are classified as wasted if their WHZ or MUACZ is below -2; they are severely wasted if their WHZ or MUACZ is below -3.WHZ and MUACZ are the outcome variables of our econometric analysis. We therefore restrict our sample to households who have children under five years of age with valid WHZs in all survey rounds. Our final sample includes 36.5 percent of the households in the original balanced household sample (after dropping misclassified/terminated beneficiary households). It has four rounds of observations for 3,176 children (aged 0-59 months) from 2,232 households, of which 2,687 children (aged 6-59 months) from 1,905 households have valid MUACZ scores.Table 3.1 presents the prevalence of acute malnutrition and severe acute malnutrition among pre-school children in our sample for both indicators.2 Note that the standard deviations (SD) of WHZ and MUACZ in all rounds are near or even below 1.0, which gives us confidence in the quality of the anthropometric data (Grellety and Golden 2016b;Mei and Grummer-Strawn 2007). The average SD-across all four survey rounds-is 1.03 for WHZ and 0.95 for MUACZ. Our final sample has 1,084 beneficiary households, made up of 448 new beneficiary and 636 old beneficiary households, and 1,148 non-beneficiary households. The selection of beneficiary households based on eligibility criteria differed between old and new beneficiaries due to a series of reforms of the SWF cash transfer program that started in 2008. The most important reforms were (1) the enactment of an SWF law in early 2008 that specifies the eligibility criteria of the cash transfer program, and (2) the (re)assessment of all SWF beneficiaries and potential beneficiaries based on the above-mentioned census of poor people and the use of a proxy means test (PMT) formula that accounts for these criteria. While the eligibility criteria had been considered in part for the selection of beneficiary households before 2011, the PMT formula was first implemented in 2011 to select new beneficiaries. Due to the political sensitivity of excluding beneficiaries from the program, especially after the 2011-12 revolution, the reclassification of old beneficiaries through the use of PMT screening were not effectively enforced (IPC-IG et al. 2014a). Given these classification issues, our analysis mainly focuses on all beneficiary households, but we will show that the results are robust to both new and old beneficiary households considered separately.As a first data check, we used information from the NSPMS data to test mean differences in variables corresponding to the SWF cash transfer program eligibility criteria between the different groups of beneficiaries and non-beneficiaries. Annex 1 describes the eligibility criteria and presents the results of our analysis in detail. The results confirm that for most eligibility criteria beneficiary households are socially and economically disadvantaged compared to non-beneficiary households, on average. They also suggest that there are many households in the non-beneficiary group that appear to fulfill the eligibility criteria. Most importantly, the results indicate that there are no statistically significant differences in the means of most eligibility criteria variables between old and new beneficiary households, despite the selection of new beneficiaries through the PMT method. On average, old beneficiaries appear to be better targeted with respect to some social criteria, particularly for households with elderly people and single, widowed, or divorced women, whereas new beneficiaries appear to be better targeted in terms of total chronic poverty. Overall, the findings imply that the introduction of the PMT method has slightly shifted the targeting focus towards economic eligibility criteria rather than social eligibility criteria, while it appears not to have substantially improved general targeting effectiveness (IPC-IG et al. 2014a).The NSPMS provides information on the payment modalities of the SWF cash transfer program. By law, transfer payments were to be made quarterly (IPC-IG et al. 2014a). While cash transfers were administered to individual beneficiaries, beneficiaries of the same households typically received their payments at the same time. In 85.3 percent of all beneficiary households in our sample, there is only one registered beneficiary per household. The percentage is 90.8 percent for new beneficiary households and 81.4 percent for old beneficiary households. Source: Own estimated based on 2012-13 Yemen NSPMS data. Note: The frequencies are reported in percentage. The frequency is calculated by household as the count of the survey rounds when the interviewed household reported to have received a payment since the last interview and, for the first round, over the three-month period prior to the first interview.Table 3.2 shows the frequency of the received transfer payments among beneficiary households. It suggests that many beneficiary households received their payments irregularly. While all beneficiary households in our sample reported having received cash transfers, more than one in five households received only one or two payments during the observation period of our study. On average, new beneficiaries received 2.95 payments and old beneficiaries received 3.13 payments.The mean difference is statistically significant and reflect that some new beneficiaries were enrolled in the program after the first survey round.Since the 2008 SWF program reforms, the maximum monthly transfer amount per beneficiary was YER 4,000 (IPC-IG et al. 2014a), or about USD 18.64 in early October 2012. This is equivalent to USD 0.10 per day in a typical six-person household with one registered beneficiary. While this amount is small, focus group discussions revealed that beneficiaries valued the regularity of the transfer payments to cover regular expenses for basic needs, including food purchases, and to repay debts for purchases made on credit (IPC-IG et al. 2014a).The conflict event data are taken from the GDELT 1.0 Event Database, which provides data through 31 March 2013 (GDELT 2013) and the GDELT 2.0 Event Database, which provides data from 1 April 2013 onward (GDELT 2015). The GDELT Project compiles news media records of conflict events from countries around the world. The databases provide the number of conflict events per day, the \"importance\" (as proxy of the significance) of the event, and the location of the event, among other variables. We consider only conflict events that are classified as important events and for which the administrative district of the event location can be identified. For almost all reported events, the location is specified by the precise geographic coordinates of the event or at least the geographic coordinates of the capital of the district where the event took place. We used the GDELT event coordinates and a geographic map for Yemen to identify the districts of the event locations. We were able to identify the district of the event location for 99.5 percent of all important events in Yemen between 2005 and 2015. Then, we aggregated the number of events at the district level and by month per year. Finally, we merged this dataset with our NSPSMS sample at the district level and by year and month. Table 3.3 shows summary statistics for the exposure of the households in our sample to civil conflict and the intensity of civil conflict experienced by survey round. We report the statistics for the different groups of SWF cash transfer program beneficiaries and non-beneficiary households. Source: Own estimated based on 2012-13 Yemen NSPMS and GDELT data. SD = standard deviation Table 3.4 shows the results of mean difference tests for conflict exposure and intensity between the different household groups. The estimates clearly show that, although conflict exposure and intensity slightly vary among the different household groups, there are no statistically significant differences in these variables between the groups on average. This finding provides suggestive evidence that the coverage of the SWF cash transfer program is not correlated (positively or negatively) with the occurrence or intensity of civil conflict in our sample. The econometric analysis exploits the panel structure of our dataset. We use fixed-effects (FE) regression models to first estimate the nutritional impact of civil conflict and then the mitigation effect of the SWF cash transfer program. Hence, our analysis includes two main steps. First, we test our hypothesis that acute child malnutrition in Yemen increases with the number of conflict events that children experience. Second, we examine if unconditional cash transfers can help to mitigate the detrimental impact of conflict intensification on children's WHZ and MUACZ-the most common child nutrition indicators used for diagnosing acute child malnutrition. Specifically, we explore the mitigation effect of SWF cash transfer program for all beneficiaries compared to non-beneficiaries.We then check the robustness of our estimation results by separately considering only new and old beneficiaries in the treatment variable, given the changes in selection and enrollment modalities discussed above. In addition to exploring the mitigation effect that is associated with program participation, we estimate the marginal mitigation effect that is associated with changes in the regularity of received transfer payments among all beneficiaries.We estimate regression models with district FE and household FE, together with survey round FE. These models address potential endogeneity problems due to unobserved (time-constant) heterogeneity across district and households, respectively, as well as time-variant factors, such as seasonality. Introducing district or household FE helps to minimize potential estimation biases from spatial or inter-household correlations between child nutritional status and civil conflict or the SWF cash transfer program that may occur by coincidence or from unobserved or omitted factors driving both outcomes. The household FE model is more restrictive and can be expected to yield more robust estimation results, as it allows only for variations in variables at the household level or higher levels (such as district) over time. In all estimations, standard errors of the parameter estimates are clustered at the district level, relaxing the usual requirement that the observations be independent within clusters.The specifications of the district and household FE models have the following estimation form:where \uD835\uDC56\uD835\uDC56 refers to the individual child, ℎ refers to the child's household, \uD835\uDC51\uD835\uDC51 refers to the district where the household is located, and \uD835\uDC5F\uD835\uDC5F refers to the survey round. \uD835\uDC66\uD835\uDC66 \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 denotes the child's nutritional status per survey round, measured by WHZ or MUACZ. \uD835\uDC65\uD835\uDC65 \uD835\uDC51\uD835\uDC51\uD835\uDC56\uD835\uDC56 denotes the number of conflict events per district and by survey round that occurred since the last interview to the current interview or, for the first round, over the three-month period prior to the baseline interview. The variable measures the civil conflict intensity that the child experiences. For ease of interpretation of estimation results, we standardized the values of the variable to yield a standard deviation equal to one in our full sample. District FE enter the estimation through Ω \uD835\uDC51\uD835\uDC51 . Alternatively, household FE, denoted Ψ ℎ , are used. Survey round FE, denoted Φ \uD835\uDC56\uD835\uDC56 , are introduced in all specifications. \uD835\uDEFD\uD835\uDEFD 1 is the parameter to be estimated, and \uD835\uDF00\uD835\uDF00 \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 is the residual.Fixed effects control for unobserved heterogeneity likely to be correlated with the dependent variable, i.e., children's nutritional status, and the independent variable(s), i.e., civil conflict intensity in the first estimations and SWF cash transfers in following specifications. These controls are important for several reasons.First, survey round FE are crucial to account for seasonality effects in both household food insecurity-and hence child nutrition outcomes-and civil conflict (Abay and Hirvonen 2017;Shah and Steinberg 2017;Guardado and Pennings 2017). They also capture external shocks that affect all Yemenis similarly, such as food price spikes in the world market (Breisinger et al. 2011;Ivanic et al. 2012).Second, civil conflict events are likely to occur more frequently in poor areas where malnutrition is likely to be concentrated and where the opportunity costs to fight are expected to be low (Miguel at al. 2004;Dube and Vargas 2013;Maystadt and Ecker 2014). District FE control for such unobserved spatial heterogeneity. Because we may be concerned about possible changes in the sample population over time and household-level factors that influence child nutrition, we stepwise augment the basic estimation specification by introducing vectors that control for individual characteristics, \uD835\uDC4D\uD835\uDC4D \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 , and household characteristics at baseline, \uD835\uDC4D\uD835\uDC4D ℎ .Third, civil conflict violence has been shown to be targeted at wealthier households in countries like Burundi (Bundervoet 2010), Rwanda (Verpoorten 2009), Uganda (Blattman and Annan 2010), and the Democratic Republic of Congo (Dagnelie et al. 2018), implying a bias that may push the estimated response of socioeconomic outcomes to conflict toward zero. Household FE control for such unobserved, time-constant heterogeneity across households.Naturally, we cannot exclude that (unobserved) time-varying factors may act as confounding factors, provided they are correlated with both child nutrition outcomes and civil conflict. Weather shocks, for example, have been found to affect the risk of civil conflict violence (Hsiang et al. 2013;Maystadt and Ecker 2014;Maystadt et al. 2015) and children's nutritional status (Hoddinott and Kinsey 2001;Zivin and Neidell 2013;Kadamatsu et al. 2016). Therefore, weather shocks, denoted by \uD835\uDC4A\uD835\uDC4A \uD835\uDC51\uD835\uDC51\uD835\uDC56\uD835\uDC56 , are introduced as one step of our augmentation of the basic specification. The additional parameters to be estimated in the augmented specifications are \uD835\uDEFE\uD835\uDEFE 1 , \uD835\uDEFE\uD835\uDEFE 2 , and \uD835\uDEFF\uD835\uDEFF. 3The vector \uD835\uDC4D\uD835\uDC4D \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 controls for individual characteristics in each survey round. It includes variables identifying the child's sex and his/her age (in months) as linear and squared terms. The vector \uD835\uDC4D\uD835\uDC4D ℎ controls for time-constant, standard household characteristics that were measured in Round 1 of the NSMPS. It includes variables identifying household wealth, household size (as measured by the number of household members who permanently live in the household), and the sex, the age (in years), and the literacy status of the household head. Household wealth is measured by a household asset-based wealth index. To construct the index for our sample, we used principal component analysis and a large set of household asset variables, following the procedure proposed by IPC-IG et al. (2014c). All other variables are directly available from the NSMPS without considerable transformation.The vector \uD835\uDC4A\uD835\uDC4A \uD835\uDC51\uD835\uDC51\uD835\uDC56\uD835\uDC56 controls for weather shocks at the district level. It includes either temperature and precipitation anomalies, in one estimation specification, or the Standardized Precipitation Evapotranspiration Index (SPEI), in another specification. Both the anomalies and the SPEI indicate extreme weather events, such as droughts and floods. We constructed these extreme weather indicators from two weather data sources and merged them to our dataset at the district level and by survey round. We used georeferenced land surface temperature and precipitation data from the Moderate Resolution Imaging Spectroradiometer (MODIS) database of the US National Aeronautics and Space Administration (NASA) (Wan 2015) and georeferenced precipitation data from the Climate Hazards Group InfraRed Precipitation with Station (CHIRPS) database of the Climate Hazards Group at the University of California Santa Barbara (Funk 2015). To obtain one monthly observation per district from the georeferenced data, we applied a series of geoprocessing procedures. Most notably, we used the Spline spatial interpolation method (Mitas 1988) to impute missing observations at the spatial raster level, and the Zonal Statistics function in the ArcGIS software package to calculate district-level averages from the spatial raster data. 4We used the monthly, district-level MODIS data to calculate average temperature anomalies and the monthly district-level CHIRPS data to calculate cumulative precipitation anomalies over a three-month observation period. The three-month observation period includes the observation of the survey month of the NSPMS and the preceding two months. Anomalies are generally calculated as the deviation of the current month value from the long-term monthly mean, divided by the monthly long-term standard deviation over the same time period (Dell et al. 2014;Marchiori et al. 2012;Barrios et al. 2010). Our reference period is 2001 to 2015. We also used monthly district-level MODIS data to construct the SPEI based on three-month periods. SPEI is a multi-scalar drought indicator that includes both rainfall and temperature observations and has emerged as one of the most comprehensive drought indexes over recent years (Begueria et al. 2010;Cook et al. 2014). 5 The index is based on a monthly weather-water balance, i.e., precipitation minus potential evapotranspiration (PET). SPEI measures the monthly difference between precipitation and PET and is expressed as a standardized Gaussian variate with a mean of zero and a standard deviation of one.To estimate the mitigation effects of participation in the SWF cash transfer program and the regularity of cash transfer payments on child nutrition, we augment the specifications of the district FE model and household FE model by stepwise introducing a respective treatment variable and an interaction term of the treatment variable with civil conflict intensity. The specifications of the district FE model have the following estimation form:where \uD835\uDC61\uD835\uDC61 ℎ is the treatment variable that is time-constant throughout all survey rounds. The additional parameters to be estimated are \uD835\uDEFD\uD835\uDEFD 2 and \uD835\uDEFD\uD835\uDEFD 3 .In a first set of estimations, the treatment variable, \uD835\uDC61\uD835\uDC61 ℎ , identifies the program beneficiary status of a household. This variable is binary, and its parameter estimate, \uD835\uDEFD\uD835\uDEFD 2 , indicates the general effect of program participation on child nutrition independent of the intensity of civil conflict experienced. As explained in the previous section, the beneficiary status of households is defined prior to the observation period of our study. Replacing district FE by household FE in the estimations controls for unobserved selection into the program. We augment both the district FE model and household FE model with a term that captures the interaction between the treatment variable and the conflict events variable, \uD835\uDC65\uD835\uDC65 \uD835\uDC51\uD835\uDC51\uD835\uDC56\uD835\uDC56 * \uD835\uDC61\uD835\uDC61 ℎ . The parameter estimate of the interaction term, \uD835\uDEFD\uD835\uDEFD 3 , if positive, indicates a mitigation effect of program participation on child nutrition which counteracts the negative nutritional impact of civil conflict intensification.An identification concern might be the possibility that beneficiaries and non-beneficiaries are on different development paths in the absence of the SWF cash transfer program. However, given the very low level of pro-poor economic growth and human capital investment in Yemen and the relatively good targeting of beneficiaries in terms of total chronic poverty documented in the previous section, we can reasonably conjecture that improvements in living conditions among beneficiaries is much slower than those among non-beneficiaries, even with the benefits from the SWF cash transfer program. Furthermore, given our focus on the mitigation effect of the program, it is even less obvious to conjecture opposite trend differentials between beneficiaries and nonbeneficiaries in conflict affected and non-affected districts. The lack of correlation between the coverage of the SWF cash transfer program and the occurrence or intensity of civil conflict in our sample corroborates this identifying assumption (Table 3.4). Finally, we show that our estimation results for the effect of program participation among all beneficiaries compared to non-beneficiaries are robust to restricting the group of beneficiaries to only new beneficiaries and old beneficiaries, respectively.In a second set of estimations, we replace the binary treatment variable with a continuous variable that counts the number of SWF cash transfer payments that a household received over the observation period of our study. Hence, the parameter estimate, \uD835\uDEFD\uD835\uDEFD 2 , indicates the marginal general effect of a change in the regularity of cash transfer payments independent of the intensity of civil conflict experienced. The parameter estimate of the interaction term, \uD835\uDEFD\uD835\uDEFD 3 , if positive, indicates, at a given payment frequency, the marginal effect of changes in the payment regularity on child nutrition that mitigates the negative nutritional impact of civil conflict intensification. The second set of estimations is motivated by anecdotal evidence on the beneficial effects of regular transfer payments among SWF program beneficiaries (IPC-IG et al. 2014a). Irregularity of payments can potentially affect how beneficiary households spend the cash transfers, with implications for child nutrition outcomes.In the following, we first establish that there is a robust detrimental impact of civil conflict increasing acute child malnutrition in Yemen. Then, we explore the mitigation effects of the SWF cash transfer program and the regularity of cash transfer payments.Table 5.1 shows that increasing civil conflict intensity is associated with deteriorating child nutrition, increasing the risk of acute child malnutrition. The parameter estimates of the conflict event count variable are highly statistically significant in all estimations for child WAZ and MUACZ. The estimated effect sizes for both nutrition outcomes are similar across the different specifications, which indicates the robustness of our estimation results. The estimated effect of civil conflict on child WAZ ranges from -0.033 in the household FE model estimation (with SPEI) to -0.038 in a district FE model estimation. The estimated effect of civil conflict on child MUACZ is larger. It ranges from -0.056 in the household FE model estimation to -0.064 in a district FE model estimation. These estimates imply that an increase in the conflict event count variable by 1 SD decreases children's WAZ by at least 0.033 SDs and their MUACZs by at least 0.056 SDs, on average. One SD in the non-standardized conflict event count variable varies by around 40 conflict events per quarter (from 64 events prior to Round 1 to 16 events prior to Round 3). To put this number into perspective, the average child in our sample was exposed to about seven conflict events in his/her home district per quarter-or 28 events over the entire observation period of this study (July 2011 to October 2012). More than 4 percent of all children were exposed to more than 47 conflict events in at least one quarter. The average number of conflict events per quarter across Yemen was 6,579 events during the observation period; it was almost twice that much during the first two quarters of 2015 (12,209 events), which marked the onset of the current civil war. An average decrease in child WHZ by 0.033 SDs and child MUACZ by 0.056 SDs translates into an increase in the prevalence of child wasting among the sampled children by 0.7 percentage points, if measured by WHZ, and 1.7 percentage points, if measured by MUACZ. A decrease in average child WHZ and MUACZ by 0.1 SDs translates into an increase in the child wasting prevalence by 1.7 percentage points and 2.8 percentage points, respectively. The prevalence of child wasting in our sample averaged across all four survey rounds is 8.7 percent for WHZ and 14.9 percent for MUACZ (see also Table 3.1).Our estimation results in Table 5.2 suggest that the SWF cash transfer program indeed mitigates the detrimental impact of civil conflict on child nutrition. The parameter estimates of the term that captures the interaction between the conflict intensity variable and the program participation variable are positive and highly statistically significant (with p-values smaller or equal to 0.001) in the district FE model estimations with child WHZ as dependent variables and the household FE model estimations with child MUACZ as dependent variables. The parameter estimates of the interaction term in the household FE model estimations for child WHZ and the district FE model estimations for child MUACZ are also positive, but just above the highest conventional statistical significance level of asterisking at a p-value of 0.1; all are statistically significant at least at a p-value of 0.15. The parameter estimates of all variables of interest are quite similar across the different specifications of both the district FE and household FE models (within each estimation panel). These results further increase our confidence in the robustness of these estimation results. The estimates indicate that the mitigation effect is considerably larger for MUACZ as an indicator of child nutrition outcomes than for WHZ. These estimation results are also robust to restricting the group of beneficiaries to new and old beneficiaries, shown in Tables 5.3 and 5.4, respectively. The parameter estimates of the interaction term consistently suggest that the mitigation effect is somewhat larger among new beneficiaries than old beneficiaries. Moreover, the parameter estimates are statistically significant in the estimations of the household FE model-the methodologically most robust one-for both child WHZ and MUACZ at least at the 10 percent level. Thus, our estimation results provide strong statistical evidence for a mitigation effect of the SWF cash transfer program that counteracts the detrimental impact of civil conflict on child nutrition.  The parameter estimates of the program participation variable are small and far from being statistically significant in all model estimations. This provides suggestive evidence that the SWF cash transfer program is not targeted at malnourished children and hence that our identification strategy is not violated by a potential bias associated with such (unobserved) targeting. Further, this finding is consistent with the results of the impact assessment of the SWF cash transfer program by IPC-IG et al. (2014a). That research team used a quasi-experimental impact evaluation method, i.e., propensity score matching, to estimate the general effect of household program participation on children's nutritional status and the prevalence of child malnutrition among all beneficiary households, i.e., the 'average treatment effect on the treated'. They find no statistically significant result for any indicator of nutritional status or malnutrition prevalence, considering measures of short-term, long-term, and composite nutrition outcomes. However, their analysis does not account for children's exposure to civil conflict intensity. The mitigation effect of the program that we found remained undetected and possibly was averaged out in their analysis as it ignored spatial and temporal heterogeneity in conflict occurrence and intensity. Furthermore, our estimation results in Table 5.5 suggest that the regularity of SWF cash transfer payments matters for the mitigation effect. The parameter estimates of the term that captures the interaction between the conflict intensity variable and the payment frequency variable are positive and statistically significant in the district FE model estimations for child WHZ and the household FE model estimations for child MUACZ, like in the estimations of the models having the binary program participation variable as treatment variable. Thus, these estimates indicate that the mitigation effect tends to be larger if transfer payments are received more regularly. Regular assistance allows beneficiary households to smoothen their food consumption and other demands, influencing child nutrition outcomes. The findings of our analysis stress the detrimental impact of civil conflict on child nutrition in Yemen.Our estimation results show that conflict intensification decreases weight-for-height z-scores (WHZ) and mid-upper arm circumference z-scores (MUACZ) of children under five years of age and, hence, increases the risk of acute child malnutrition. An increase by one standard deviation in conflict intensity translates into an increase in the prevalence of wasting in our sample by at least 0.7 percentage points, if measured by WHZ, and by at least 1.7 percentage points, if measured by MUACZ.Finding a political resolution of the current civil war in Yemen is an absolute priority to tackle what has been recognized as currently the world's worst humanitarian crisis. Although the latest agreements signed under UN oversight in mid-December 2018 to halt fighting in the contested port city of Hodeida and to allow the future deployment of UN-supervised neutral forces and the establishment of humanitarian aid corridors constitute a first important step, the road to a sustainable peace agreement will certainly be long and bumpy. Civil conflict can be expected to continue for some time, although, hopefully, with declining intensity. Building resilience to civil conflict and violence-facilitating shocks in fragile states is challenging (Breisinger et al. 2015) Against this background, our analysis suggests that unconditional cash transfers can be an effective tool in complex emergencies. This finding confirms the practical experience of program implementers in several fragile countries and civil conflict zones (e.g., HPN 2012;ODI 2015). Despite the institutional and logistical challenges encountered after the 2011-12 revolution in Yemen, the SWF cash transfer program was able to reach vulnerable households. Beneficiary households on average were more socially and economically disadvantaged compared to non-beneficiary households. Program coverage could certainly have been improved, as it was found that some of the non-beneficiary households in our study appear to have been eligible for the assistance. Nonetheless, being able to target the most in need, especially due to the program expansion started in 2011, was a notable achievement in such challenging conditions. A major implementation challenge seen throughout the observation period of our study was the irregularity of payments.Perhaps the most critical opportunity that was missed during the period of a transition in government between 2012 and 2014, preceding the onset of the current civil war, was the inability to increase and expand public assistance to disincentivize individuals to engage in civil conflict for economic means and to protect people's living conditions from the impact of conflict intensification and other shocks (e.g., Berman et al. 2011;Blattman and Ralston 2015;Maystadt and Ecker 2014). International donors could have played a greater role in providing financial support. Instead, SWF stopped cash transfer payments at the end of 2014 due to a lack of funds. After 2½ years of civil war, the World Bank stepped in to restart payments in October 2017 (ReliefWeb 2018).While the observation period of our study falls within the post-revolution transitional period, our findings are highly relevant for designing and implementing interventions what will hopefully be an upcoming post-war transitional period. Given the humanitarian crisis, there has never been a more urgent need for assistance in Yemen's recent history than today. Another window of opportunity for intervention should not be missed, as the consequences for Yemen's population will be even more devastating.Our estimation results also suggest that unconditional cash transfers can mitigate the detrimental impact of lingering civil conflict on child nutrition in Yemen on a large scale. This is consistent with findings from a forthcoming impact evaluation study of a pilot program of conditional cash transfers with a nutrition training component in three districts of Hodeidah governorate. Kurdi et al. (2018) document that household food security and child nutrition deteriorated between January 2015 and July 2017, plausibly as a result of the civil war. The authors find statistically significant and positive program effects on the consumption of a range of nutritious foods. These effects were largest among the poorest tercile of beneficiary households. They also find statistically significant and positive effects on the reduction of children diagnosed with moderate and severe malnutrition and on child weight-for-height z-scores and height-for-age z-scores (identifying child stunting) among the poorest tercile of beneficiary households.Finally, our estimation results suggest that the regularity of cash transfer payments matters for the mitigation effect to effectively counteract the detrimental impact of civil conflict. The mitigation effect tends to be larger the more regular payments are received, as regular assistance allows beneficiary households to smooth their food consumption and other demands influencing child nutrition outcomes. This finding implies that payment regularity, such as on a monthly or at least on a quarterly basis, should be given high priority in the implementation of cash transfer programs by addressing the institutional and logistical challenges associated with irregular payments.More broadly, our findings provide additional evidence of the beneficial role of cash transfers in civil conflict settings found in other contexts, such as in India (Fetzer 2018) or Iraq (Crost et al. 2016). Our study also complements recent work by Tranchant et al. (2018), who find that food assistance has a protective effect among food insecure populations experiencing civil conflict in Mali. Understanding the relative efficiency of food assistance versus (both unconditional and conditional) cash transfers in fragile countries and during civil conflict is an important area of future research that can help humanitarian and development assistance organizations in strategizing and further improving their support to affected populations.The 2008 SWF law defines household eligibility for SWF cash transfer payments according to social and economic categories (IPC-IG et al. 2014a). In the social category, a household is eligible if a household member fulfills one of the following criteria:1. Disabled, which can be fully and permanently; partially and permanently; and fully or partially temporarily. The common, practical identification is that the person is unable to work either permanently or temporarily due to a physical or mental disability or chronic disease;2. Orphaned, for children and adolescents less than 18 years of age and college students or students of technical education between 18 and 25 years and whose parents are either dead or disappeared (and therefore cannot support them financially);3. Elderly, for women above 55 years of age and men above 60 years of age.In the economic category, a household is eligible if a household member fulfills one of the following criteria:4. Single and widowed or divorced, not remarried women aged 18 years and older and widowed or divorced, not remarried women younger than 18 years with a child, whose breadwinner is absent from the household for any reason and does not provide financial support;5. Unemployed man aged 18-60 years, who does not have a job in the public or private sector and whose total income is below the SWF cash transfer level (see above).In addition to these individual-based eligibility criteria, household eligibility is assessed based on legal conditions for assistance and household chronic poverty status (IPC-IG et al. 2014a). The legal conditions are that the individual or any other family member has (a) currently no other source of income that can compensate for not receiving SWF assistance and (b) no relative who is legally obliged to provide financial support. Household chronic poverty status was determined by relating information on household assets and wealth-related characteristics of the household head from the 2008 census of poor people with corresponding information of a previous, representative household budget survey to obtain income-based poverty classifications. Households were classified into poor and non-poor, and, within the group of poor households, into extremely poor, moderately poor, and vulnerable to poverty.We used information from the NSPMS data to create variables that correspond to the eligibility criteria for the SWF cash transfer program, while the categorical variable of household chronic poverty was directly available from the obtained dataset. Table A1.1 shows summary statistics for these variables among all households, all beneficiary households, new and old beneficiary households, and non-beneficiary households in our sample in Round 1-the baseline of our study. Table A1.2 shows results of mean difference tests for the eligibility criteria variables between the different household groups, which provide indication of the accuracy of program targeting.   The statistics confirm that, for most eligibility criteria, beneficiary households are socially and economically disadvantaged compared to non-beneficiary households, on average. The only exception are households with orphans, for which there is no statistical difference between the means of beneficiary and non-beneficiary households. This finding also holds for the groups of new and old beneficiaries separately. Compared to non-beneficiary households, both new and old beneficiary households are more likely to have disabled, elderly, or single female household members; to have less income from non-SWF sources; and to be poor or extreme poor. In addition, new beneficiary households are more likely to have an unemployed male household member than non-beneficiary households. Old beneficiary households are more likely to have above-average income from non-SWF sources and are more likely to be vulnerable to poverty-instead of moderately poor-than non-beneficiary households. Both results point to targeting issues in terms of economic eligibility criteria. Moreover, the summary statistics suggest that there are many households in the non-beneficiary group that may be eligible for SWF program benefits. For example, 22.6 percent of the non-beneficiary households in our sample have an elderly person, and 11.9 percent were classified as extremely poor and 28.7 percent as moderately poor.Finally, the results also suggest that there are no statistically significant differences in the means of most eligibility criteria variables between old and new beneficiary households, despite the selection of new beneficiaries through the PMT method. On average, old beneficiaries appear to be better targeted with respect to some social criteria-specifically households with elderlies and single, widowed, or divorced women. In contrast, new beneficiaries appear to be better targeted in terms of total chronic poverty. Overall, the findings imply that the introduction of the PMT method seems to have only slightly shifted the targeting focus toward stronger weights on economic eligibility criteria rather than social eligibility criteria, while it appears to not have substantially improved overall targeting effectiveness (IPC-IG et al. 2014a).  ","tokenCount":"8949","images":["1260011009_1_1.png"],"tables":["1260011009_1_1.json","1260011009_2_1.json","1260011009_3_1.json","1260011009_4_1.json","1260011009_5_1.json","1260011009_6_1.json","1260011009_7_1.json","1260011009_8_1.json","1260011009_9_1.json","1260011009_10_1.json","1260011009_11_1.json","1260011009_12_1.json","1260011009_13_1.json","1260011009_14_1.json","1260011009_15_1.json","1260011009_16_1.json","1260011009_17_1.json","1260011009_18_1.json","1260011009_19_1.json","1260011009_20_1.json","1260011009_21_1.json","1260011009_22_1.json","1260011009_23_1.json","1260011009_24_1.json","1260011009_25_1.json","1260011009_26_1.json","1260011009_27_1.json","1260011009_28_1.json","1260011009_29_1.json","1260011009_30_1.json","1260011009_31_1.json","1260011009_32_1.json","1260011009_33_1.json"]}

data/part_2/0273265745.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"053e89f47bc5a035f5c31376394da52d","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/710adce5-81bd-4b4c-a92c-7afaf2c9506d/retrieve","description":"Poor dietary quality is a significant risk factor for stunting and micronutrient deficiencies among young children and globally one of the leading causes of premature death and disease (Arimond & Ruel, 2004; Forouzanfar et al., 2015). Dietary quality is typically proxied by diversity of the consumed diet. Foods with similar nutritional qualities are first grouped together and dietary diversity is measured by the number of different food groups consumed in a certain time interval. For example, the World Health Organization recommends that children 6-23 months consume at least from four food groups (out of seven) every day. Based on this metric, Ethiopian children in this age range consume one of the least diversified diets in sub-Saharan Africa (Hirvonen, 2016) with only 14 percent meeting the WHO recommendation (CSA & ICF, 2016). Recent analysis of the timing of growth faltering of young children suggests that poor complementary feeding practices, including poor dietary quality, is an important risk factor for stunting in Ethiopia (Hirvonen, Headey, Golan, & Hoddinott, 2019). The available evidence suggests that diets are monotonous also at the household level. For example, in 2011, the average Ethiopian household consumed only 42 kg of fruits and vegetables in a year per adult equivalent (Hassen Worku, Dereje, Minten, & Hirvonen, 2017) – far below the World Health Organization’s recommendation of 146 kg per year (Hall, Moore, Harper, & Lynch, 2009). This report is structured as follows. In the subsequent section, we describe the data used in this report. In section 3, we assess the consumption of nutritious foods among vulnerable groups: young children and mothers. In section 4, we assess the production of nutritious foods in the region. In section 5, we study the availability of nutritious foods in rural markets. In section 6, we assess the affordability of nutritious foods in the region. Section 7 concludes and summarizes the findings.","id":"192175356"},"keywords":[],"sieverID":"48c9b6bc-3ac8-4fd4-ab77-4e7f3f7767ed","pagecount":"41","content":"as well as the colleagues at the Food and Nutrition Coordination Office at the Ministry of Agriculture for useful comments on the earlier versions of this document.Alive & Thrive is an initiative to save lives, prevent illness, and ensure healthy growth and development through improved breastfeeding and complementary feeding practices. Good nutrition in the first 1,000 days, from conception to 2 years of age, is critical to enable all children to lead healthier and more productive lives. Alive & Thrive is scaling up nutrition through large-scale programs in several countries in Asia and Africa and through strategic technical support and the dissemination of innovations, tools, and lessons worldwide.Alive & Thrive is funded by the Bill & Melinda Gates Foundation and the governments of Canada and Ireland. The initiative is managed by FHI 360. The Alive & Thrive initiative, managed by FHI 360, is currently funded by the Bill & Melinda Gates Foundation, Irish Aid, the Tanoto Foundation, and UNICEF. Disclaimer This publication has not been peer reviewed. Any opinions stated in this publication are those of the author(s) and are not necessarily representative of or endorsed by IFPRI.Poor dietary quality is a significant risk factor for stunting and micronutrient deficiencies among young children and globally one of the leading causes of premature death and disease (Arimond & Ruel, 2004;Forouzanfar et al., 2015).Dietary quality is typically proxied by diversity of the consumed diet. Foods with similar nutritional qualities are first grouped together and dietary diversity is measured by the number of different food groups consumed in a certain time interval. For example, the World Health Organization recommends that children 6-23 months consume at least from four food groups (out of seven) every day. Based on this metric, Ethiopian children in this age range consume one of the least diversified diets in sub-Saharan Africa (Hirvonen, 2016) with only 14 percent meeting the WHO recommendation (CSA & ICF, 2016). Recent analysis of the timing of growth faltering of young children suggests that poor complementary feeding practices, including poor dietary quality, is an important risk factor for stunting in Ethiopia (Hirvonen, Headey, Golan, & Hoddinott, 2019). The available evidence suggests that diets are monotonous also at the household level. For example, in 2011, the average Ethiopian household consumed only 42 kg of fruits and vegetables in a year per adult equivalent (Hassen Worku, Dereje, Minten, & Hirvonen, 2017) -far below the World Health Organization's recommendation of 146 kg per year (Hall, Moore, Harper, & Lynch, 2009).Such monotonous diets are likely to increase the risk of various chronic diseases. For example, a recent global study indicated elevated mortality as well as major cardiovascular disease risk when energy intake from carbohydrates exceed 60 percent (Dehghan et al., 2017). This is particularly worrying for Ethiopia as recent estimates by the Ethiopian Public Health Institute suggest that 60-80 percent of the energy intake of children and adults comes from carbohydrates (Misganaw et al., 2017). Meanwhile, cardiovascular diseases are among the top causes of premature mortality in the country (Misganaw et al., 2017). Another indication of limited dietary quality is that micronutrient deficiencies in Ethiopia are widespread. Nearly 60 percent of young children are anemic (CSA & ICF, 2016) and nearly 15 percent suffer from Vitamin A deficiency (EPHI, 2016). Recent research from Ethiopia further highlights the importance of fruit and vegetable consumption among pregnant and lactating women. Data from health centers in rural Oromia, Zerfu, Pinto, and Baye (2018) finds that pregnant women who consumed fruits and dark green leafy vegetables more frequently were less likely to experience adverse pregnancy outcomes. Moreover, analyzing samples of breastmilk from mothers in rural Amhara, Abebe, Haki, Schweigert, Henkel, and Baye (2018) find very low concentrations of Vitamin A in milk.These issues are well-acknowledged by the government of Ethiopia. The national nutrition programme sets out ambitious plans to increase the year-around availability, access, and consumption of nutritious foods (GFDRE, 2016a). Through its 'Seqota Declaration of 2015,'Ethiopia further announced its ambitious goal of ending child malnutrition by 2030 (GFDRE, 2016b). These nutrition strategies are endorsed by several ministries highlighting that the emerging consensus that combating under-nutrition and poor diets requires multi-sectoral efforts (Menon & Frongillo, 2018).At the core of these efforts is the urgent need to transform food systems in Ethiopia to support healthier diets (Gebru et al., 2018). The concept of food system captures the food environment (affordability and accessibility), consumer preferences, and the food supply system that is formed of production, storage, transportation, processing, and marketing (Scott, 2017). It is now well acknowledged that the dietary choices of individuals are determined by the surrounding food system.Against this backdrop, in this report, we aim to gain a better understanding of the food sub system1 in the Amhara region of Ethiopia by analyzing the affordability, accessibility, and consumption of nutritious foods. 2 The region hosts more than 20 million people out of which most reside in rural areas and rely on agricultural production as their main livelihood (CSA, 2018a). About 47 percent of the children under 5 in Amhara are chronically under-nourished (or stunted) (Hirvonen et al., 2019)  This report is structured as follows. In the subsequent section, we describe the data used in this report. In section 3, we assess the consumption of nutritious foods among vulnerable groups: young children and mothers. In section 4, we assess the production of nutritious foods in the region. In section 5, we study the availability of nutritious foods in rural markets. In section 6, we assess the affordability of nutritious foods in the region. Section 7 concludes and summarizes the findings.All the analyses in this report are based on secondary data collected by the Central Statistical Agency (CSA) of Ethiopia or the International Food Policy Research Institute (IFPRI). This section describes these data sources.We use the 2016 Demographic and Health Survey (DHS) data for Ethiopia to analyze child diets.This survey is nationally as well as regionally representative and was implemented by the CSA with technical assistance from the ICF. The interviews took place between January 18, 2016, and June 27, 2016. A total of 16,650 households were interviewed in all regions of Ethiopia and 1,902 households in Amhara. Apart from rich information on various health outcomes and socioeconomic characteristics, the DHS collects information about complementary feeding practices of young children.We use the Productive Safety Net Program (PSNP) and Feed the Future (FtF) evaluation data sets to assess diets of adult women (mothers) as well as market availability of nutritious foods.In 2016, the International Food Policy Research Institute was tasked by the Bill and Melinda Gates Foundation to evaluate the impact of the nutrition-sensitive components of the PSNP. To this end, we fielded a baseline survey in PSNP localities in Amhara, Oromia, SNNP, and Tigray.This baseline survey was administered in two parts. The first baseline survey took place in March 2017. In this round, 2,635 households with a child less than 2 years of age were interviewed. Out of these, 661 households were located in the Amhara region. Roughly half of the sampled households benefitted from the PSNP and the other half identified themselves as poor but did not benefit from the program. The second baseline survey took place roughly six months later in August 2017. In this round, the survey teams visited the same households that were interviewed in the March round. A total of 2,569 households were interviewed, indicating a dropout (attrition) rate of 3 percent. The first endline survey took place in March 2019. Again more than 2,500 households were interviewed in the same localities.In this report, we use the August 2017 round to assess women's dietary diversity in PSNP localities. To assess market availability of nutritious foods, we use the March 2019 survey round.The reason for using this round is that we improved the market questionnaire by including more food items, especially with respect to animal-sourced foods.These PSNP surveys are geographically widespread having been administered in 264 kebeles in 88 woredas in the four regions (66 kebeles and 22 woredas from Amhara). Despite this, the surveys focused on localities in which the PSNP is operational and therefore these data sets are not representative of the country, nor any of the four regions.In our attempt to address this limitation about representativeness, we append the analysis with the FtF survey that was based on similar survey instruments as the BMGF-PSNP survey. The FtF survey data sets were collected between September and October 2018 in Amhara, Oromia, SNNP, and Tigray regions. These data sets were collected to obtain post-intervention (endline) information in localities that received investments aimed at improving agricultural production and nutrition under the Feed the Future (FtF) program funded by the United States Agency for International Development (USAID). The sample is large -3,890 households -and widely distributed, the survey having been implemented in 151 kebeles in 51 woredas. In Amhara, 1,076 households from 66 kebeles in 22 woredas were interviewed. As with the BMGF-PSNP survey, it is important to note that while these data sets are representative of the zone of influence within which the FtF program operates, they are not nationally or regionally representative.  This report provides the total annual crop output in 10 out of the 11 administrative regions of the country. Only the production in the capital, Addis Ababa, is not reported. The crop output for meher (long rainy season and the main cropping season for most part of the country) and belg (short rainy season) are reported separately.This report provides the annual estimates of the livestock population and livestock production. The sample sizes are large -typically containing more than 40,000 rural households. These reports are based on nationally and regionally representative data collected by the CSA each year.We digitalized these reports. For crop output, we aggregated the total annual production for each crop in each region during meher and belg seasons as well as the total crop output produced by the commercial sub-sector. For livestock products, we took the total annual milk (cow and camel), eggs and honey produced. The CSA reports give the number of different livestock types slaughtered. These numbers were converted to kilograms of beef, sheep, goat, camel, and poultry meat using conversion factors to account for edible portions only (FAO, 1972).The annual crop and livestock production was then converted into energy (kilocalories) using the Ethiopian food composition tables (EPHI, undated). We also used FAO reject estimates which correspond to weight estimations of edible portions of foods produced to calculate energy and nutrients supplied. The total energy was then divided by the total population for that specific year based on estimates by the CSA (2013). The energy and nutrients produced were expressed on a per day basis by dividing by 365 to enable comparison with daily requirements.For the affordability analysis, we used monthly retail price data collected by the CSA. The original purpose of these data sets is to calculate the official consumer price index (CPI) in the country. These price data are collected from about 110 markets in all regions of Ethiopia with the number of markets approximately proportional to the region's size in terms of population. The CSA enumerators visit these markets every month and collect price data for more than 400 food and non-food items. For each item, the enumerators target three price quotations from different traders. For more information about this survey, see Headey, Nisrane, Worku, Dereje, and Taffesse (2012) and Bachewe, Hirvonen, Minten, and Yimer (2017).Child diets are described using the World Health Organization recommended infant and young child feeding (IYCF) indicators, which are widely accepted and used to capture optimal feeding practices in populations (WHO, 2010). Here we focus on indicators related to adequate consumption of high-quality complementary foods that are dense in key micro and macronutrients.We assess this using the protocols recommended by the World Health Organization (WHO, 2008).The DHS-2016 survey asked mothers a series of yes/no questions about foods consumed in the past 24 hours by children 6-23 months. These responses were grouped into the following categories:1. Grains, roots, and tubers (e.g., barley, enset, maize, teff, and wheat);2. Legumes and nuts (e.g. chickpea, beans, groundnut);3. Dairy products (milk, yogurt, cheese); 4. Flesh foods (meat, poultry, and fish products); 5. Eggs; 6. Vitamin A-rich fruits and vegetables (e.g. mango, pumpkin, carrots);7. Other fruits and vegetables (e.g. onions, tomatoes, bananas).This yields a count ranging in value from zero to seven. WHO (2008) recommends that children in this age range consume daily at least from four food groups out of the above seven food groups.This number was selected to maximize the likelihood that in addition to the staple food (grain, root, tuber), the child consumed animal-source foods as well as fruits and vegetables in the previous day (WHO, 2008).This relatively simple indicator is highly correlated with more detailed measures of food intake (Ruel, 2003) as well as with children's micronutrient intakes 4 . Moreover, several studies show strong correlations with this dietary diversity score and longer term measures of children's nutritional status (e.g. child stunting prevalence) in a number of countries including Bangladesh, Ethiopia, India, and Zambia (Jones et al., 2014).In Amhara, the average 6-23 month old child consumes from 1.4 food groups and consequently, only 2.9 percent of the children meet the criteria for minimum dietary diversity (at least four food groups in the last 24 hours). Children's diets are somewhat more diverse in urban areas; 2.6 percent of the rural children and 5.1 percent of the urban children in Amhara met the minimum dietary diversity. Meanwhile, 27 percent of the children consumed only from one food group. For these children, the dietary diversity would have to increase by three food groups in order to meet the minimum recommended dietary diversity. About 35 percent were two food groups shy from meeting the recommendation while 11 percent of the children were missing just one food group. The remaining 3 percent of the children consumed from four or more food groups.  We wondered how dietary diversity changes as children age.  There is an emerging consensus among researchers and practitioners that income growth alone will not address poor dietary quality (Ruel, Alderman, Maternal, & Group, 2013). To illustrate this, we regressed child dietary diversity on an aggregated wealth measure based on household's ownership of different durable assets. Figure 3.4 shows how children originating from wealthier households do consume a more diverse diet but even in the richest households, the average child consumes only from two or three food groups. We assessed maternal diets using FAO and FHI 360 (2016)  This yields a dietary diversity score ranging from 0 to 10. Minimum Dietary Diversity (MDD-W) is met if the mother consumed from five or more food groups during the 24-hour window.The average mother in the BMGF-PSNP sample consumed from 2.5 food group (out of ten).Figure 3.5 shows the full distribution of mothers' dietary diversity score in PSNP woredas in August 2017. As before, the bars falling short from five food groups are marked with gray color.We see that very few (0.5 percent) of the mothers met the minimum dietary diversity of five food groups. Meanwhile, 7 percent of the mothers consumed only from one food group. These women would have to increase their dietary diversity by four food groups in order to meet the minimum recommended dietary diversity. About 46 percent of the women were three food groups shy from meeting the recommendation, while 42 percent needed two food groups to meet (MDD-W).About 5 percent of the women were missing just one food group.  From animal-sourced foods, less than 4 percent consumed dairy, 2 percent consumed flesh foods (poultry, fish, meat), and less than 1 percent eggs. Less than 6 percent of the women consumed Vitamin A-rich dark green leafy vegetables and only 1 percent other Vitamin A-rich fruits and vegetables. Hardly anyone consumed other types of fruits.    In this section, we examine the agricultural production in the region from a nutrition perspective.Using CSA's agricultural sample survey data, we disaggregate the production into the seven-food groups used to measure children's dietary diversity in section 3.1. To capture the foods that are not categorized in the seven-food groups (e.g. red peppers, garlic, coffee, sugarcane, and honey),we add an additional food group and called it 'other foods.' We then convert the amounts produced to calories using the Ethiopian food composition tables (EPHI, undated) and FAO reject estimates which correspond to weight estimations of edible portions of foods produced to calculate energy supplied. The total energy is then divided by the total population for that specific year reported in CSA (2013). Finally, the energy produced is expressed on a per day basis by dividing by 365 to enable comparison with daily requirements. 6   Table 4.1 shows the daily total energy production per capita per these food groups. First, the total calorie production calories increased by 20 percent between 2011 and 2015. This increase was largely driven by the production of grains, roots, and tubers. In this food group, the amount of calories produced increased by 20 percent. Second, we see that the agricultural production is heavily concentrated on the production of grains, roots, and tubers that accounted for 85 percent of all calories produced in the region in 2015. 7 Legumes and nuts account for 12 percent of the total calorie production. Strikingly, the remaining six food groups account only for about 3 percent of the total calories produced in the region. The production of Vitamin A-rich fruits and vegetables is particularly low in Amhara. The region's agricultural production would only provide 0.6 calories per day from this food group. Using the same data, Baye et al. (2019) calculate that this level of production would mean that nearly 33 percent of the population in Amhara suffered from Vitamin A deficiency; see Table 4.2. The implicit assumption here is that the availability of energy (and nutrients) is constant across seasons. This is unlikely to be the case: agricultural production in Ethiopia is largely rain-fed (Taffesse, Dorosh, & Gemessa, 2012), and therefore highly seasonal. Hirvonen, Taffesse, and Worku (2016) show how household energy intakes vary across seasons. 7 It is instructive to compare this production share to international dietary recommendations. For example, the recently developed diet recommendation by the EAT-Lancet commission recommends that only one-third of the total daily calories come from grains, roots, and tubers (Willett et al., 2019).In this section, we assess the availability of nutritious foods in rural markets. As shown by Sibhatu and Qaim (2017), nutritious foods, such as fruits, vegetables, meat, and eggs are largely sourced from markets in rural Ethiopia. The exception is dairy for which the markets are missing or poorly functioning (Hoddinott, Headey, & Dereje, 2015).Together, the PSNP and FtF survey teams visited 108 rural communities in 36 woredas in Amhara. In each community, the team visited the local food market to assess food availability and prices. A novel feature of these market surveys was that the enumerators recorded whether the food item is available in the market.8 Both surveys fielded a similar market questionnaire.We group this data using the 10-food group categorization used to analyze women's dietary diversity in Section 3.2, with the exception that we do not include the starchy staple food group in the analysis.9 Table 5.1 shows the market availability in the beans and peas food group. We see that market availability is not a constraint for this food group; more than 80 percent of the markets have chick peas, horse beans, and lentils. Source: Own calculation from BMGF-PSNP (March-2019) and FtF (September-October 2018) surveys.Table 5.3 shows the market availability for the items in the dairy food group. Here we see considerable differences across the two surveys. More than 30 percent of the markets visited in the FtF survey had fresh milk, whereas only 19 percent of the markets visited in the PSNP surveys had fresh milk. Large differences in the availability of powdered milk and yoghurt are also observed. Still, it becomes clear that dairy is poorly available in the Amhara markets. Source: Own calculation from BMGF-PSNP (March-2019) and FtF (September-October 2018) surveys.Table 5.4 shows the market availability for meat products (flesh foods). We see that more than 75 percent of the FtF markets had beef, while only 26 percent of the markets in PSNP areas had beef. Other cut meat (goat, lamb, chicken, camel) was rarely available in these markets. If the consumer wanted chicken, he or she had to buy a live chicken and slaughter it at home as chicken meat is not available in these markets. Eggs are available in almost all markets (Table 5.5) but the availability of Vitamin A-rich dark green leafy vegetables varies (Table 5.6). Ethiopian kale (gommen) is available in less than half of the markets in the PSNP areas and in more than 70 percent of the markets in FtF areas.Spinach was available in 26 percent of the markets in PSNP areas and in nearly half of the markets in FtF areas. Source: Own calculation from BMGF-PSNP (March-2019) and FtF (September-October 2018) surveys.Table 5.7 shows the availability of foods under the other Vitamin A-rich vegetable and fruit group. The availability of fruits varies by season and may explain some of the differences between two data sets. In the PSNP survey, mangoes were found in 42 percent of the markets and papayas in 40 percent of the markets. About 15 percent of the markets in PSNP areas had pumpkin. Carrots were available in 59 percent of the markets in the PSNP areas. In the FtF survey, mangoes were found in 38 percent of the markets and papayas in 24 percent of the markets. Carrots were found in 62 percent of the FtF markets and pumpkin in 45 percent of the FtF markets. Source: Own calculation from BMGF-PSNP (March-2019) and FtF (September-October 2018) surveys.Other types of vegetables are widely available (Table 5.8). Onions, tomatoes, and green pepper were on sale in most markets. Lettuce was on sale in about 30 and 40 percent of the PSNP and FtF markets, respectively. Cauliflower and mushrooms were rarely available. Lettuce was available in 26 percent of the markets in PSNP areas and 41 percent in the FtF areas. Source: Own calculation from BMGF-PSNP (March-2019) and FtF (September-October 2018) surveys.As for fruits that are not rich in Vitamin A, the market availability is high for bananas, oranges, and lemons (Table 5.9). Avocados were found from 17 and 43 percent of the markets in PSNP and FtF localities, respectively. Meanwhile melons and cactus fruits were rarely found at the time the enumerators visited these markets. In this section, we assess the affordability of nutritious foods in Amhara. We use the seven-food group categorization used to assess children's diets in Section 2.1, except we omit the first food group: grains, roots, and tubers. We define affordability as the share of total income needed to consume the recommended daily amount of the food group. 10Ethiopia has not yet developed itsown nutritional guidelines or eating recommendations.Therefore, we have to use international nutritional guidelines to determine the recommended intake for each food group. Here, we use the recently developed recommendations by the EAT-Lancet Commission on Food, Planet and Health (Willett et al., 2019). The EAT diet recommendations attempt to maximize health benefits while minimizing the diet's negative impact on environment. Hence, the diet recommendation puts more emphasis on plant-based foods (legumes, nuts, vegetables, and fruits), and less on meat and eggs. 11 Table 6.1 maps these recommendations to the six food groups.10 These calculations do not account for refuse factors and therefore these affordability estimates should be considered as lower bound.11 As noted by Willett et al (2019), the recommended amount of animal-source foods may be sub-optimal for many sub-Saharan African countries. More specifically, the authors write (p.10) that \"[b]ecause many regions, such as sub-Saharan Africa, still face severe burdens of undernutrition and malnutrition, and growing children often do not obtain adequate quantities of nutrients from plant-source foods alone, the role of animal-source foods should be examined carefully\".  (Willett et al., 2019). The purpose of the affordability analysis carried out here is not to calculate the cost of an optimal diet for children or women. Rather, we aim to provide a sense of the price of these foods relative to the household income levels in the region.For income, we rely on CSA's Household Consumption Expenditure (HCE) survey report for the region based on data collected between July 2015 and June 2016 (CSA, 2018b). The HCE survey does not record incomes, and, therefore, we proxy household incomes using CSA's estimates of household consumption-expenditures. The mean annual per capita consumption expenditure in the region was estimated as 8,094 birr (CSA, 2018b). Table 6.2 shows the mean annual per capita consumption expenditure estimates for each quintile. Table 6.4 provides the results.12 Following the EAT diet would mean that the average resident in Amhara would spend 5 percent of their income on legumes and nuts, 15 percent on dairy, 38 percent on meat, 4 percent on eggs, and 2 percent on Vitamin A-rich fruits and vegetables, and 7percent of other fruits and vegetables. These percentages are much higher for the poorest households for which many of these recommendations are out of reach. Therefore, the poorest households may require special attention when nutrition-sensitive agriculture and other policies are planned. 13Despite the moderate emphasis on meat products in the EAT diet, consuming the recommended intake (71 grams) would be very expensive in Amhara. For the average resident in the region, this would mean spending 38 percent of their budget just on meat products. For the poorest households, a daily ration of 71 grams of meat would take 78 percent of their total budget.The EAT-Lancet commission recommends consuming 13 grams of eggs every day. Considering that the local egg weighs about 35 grams, this means consuming two to three eggs per week. For the average resident in Amhara, this would mean allocating 4 percent of the budget on eggs. For the poorest, the corresponding share would be 9 percent.A large part of the daily calories in the EAT diet comes from the consumption of fruits and vegetables. Here, we have divided these into items that are rich in Vitamin A and other items.Consuming 200 grams of fruits and vegetables that are rich in Vitamin A would mean allocating a mere 2 percent of the budget on these items at the mean-income level. For the poorest, the corresponding share is 4 percent. Following the recommended intake for other fruits and vegetables would take 7 percent of the total budget at the mean-income level and 15 percent at the income level of the poorest households. Finally, pumpkin and cabbage were identified as the cheapest items in their food groups (Table 6.3). These foods are not widely consumed in Ethiopia, perhaps because they can only be bought in large quantities (in terms of weight) and in absence of preservation technologies, are not actually affordable. This prompts us to explore the sensitivity of these estimates considering the second cheapest item in the fruit and vegetable food groups (see Appendix B). These results are presented in Table 6.5. As expected, the percent of income needed to meet the recommended intake somewhat increases. For the average resident, following the EAT recommendation for Vitamin A-rich fruit and vegetables would cost 6 percent of the total budget. The corresponding share among the poorest households is 12 percent. As for other fruits and vegetables, 13 percent is needed at the mean-income level and 27 percent at the income level of the poorest household. This report has analyzed the consumption, production, market availability, and affordability of nutritious foods in the Amhara region of Ethiopia. In this section, we summarize these findings.Children's and women's diets in the region are extremely monotonous. Only 3 percent of the children 6-23 months meet the WHO recommended dietary diversity (four food groups out of seven). Less than 3 percent of the women meet the recommended dietary diversity for adult women (five food groups out of 10). Further analyses suggest that one reason for poor dietary diversity among infants and young children is delayed initiation of complementary foods. Other possible explanations include misperceptions around when to introduce certain foods to young children as well as poor affordability of certain foods, such as ASFs.Nearly 45 percent of children would need to add one or two additional food groups to their diet to meet the four-food group recommendation. Should this happen, the share of children achieving the minimum dietary diversity would increase to 48 percent. With this in mind, the following passages discuss the non-starchy staple food groups and their potential of reaching the most households in the region to improve dietary diversity.Legumes and nuts are widely available in the markets and their production is at relatively reasonable level, supplying more than 500 calories per person per day. They are also consumed by more than half of the children and nearly 80 percent of adult women. Therefore, and considering the dietary gaps with respect to other food groups, legumes and nuts have a limited potential for improving dietary diversity in Amhara.Dairy products are not widely consumed among children and women in Amhara. Their production is relatively low, supplying only about 50 calories per day per person. Furthermore, their market availability and affordability are poor. While dairy is an important source of many nutrients and highly beneficial input for child growth (Headey, Hirvonen, & Hoddinott, 2018), their commercial value chains are extremely complex requiring refrigeration or industrial processing (e.g. powder or ultra-high temperature processing) to maintain food safety. Given this, increasing dairy consumption further in Amhara would require considerable investments.Very few children and adult women consume meat products in Amhara. Their production is relatively low, supplying only about four calories per day per person. Furthermore, while their market availability is relatively good, meat products are not affordable for most households.High adherence to Orthodox fasting practices that prohibit consumption of and contact with animal-source foods for a large part of the year further complicate matters. We return to the Orthodox fasting issue below. All things considered, improving the availability and affordability of meat products is challenging in the Amhara region and will require innovative solutions.Dried shredded or powdered meat products hold promise as they simultaneously solve several issues regarding perishability, affordability (can be sold in small portions), and acceptability (does not require home slaughtering during fasting).Egg consumption is not widespread in Amhara. Also, their production levels are relatively low, supplying only five calories per person per day. However, their market availability is good with most rural markets stocking eggs. They are also relatively affordable, and have high nutritional content and recent evidence suggests eggs have a high potential for reducing stunting rates in poor countries (Headey et al., 2018;Iannotti, Lutter, Bunn, & Stewart, 2014;Iannotti et al., 2017). Therefore, eggs could have a high potential for improving dietary diversity in Amhara.Despite their relatively good availability and affordability, Vitamin A-rich fruits and vegetables are not frequently consumed among children and adult women. Therefore, Vitamin Arich fruits and vegetables could have a high potential for improving dietary diversity in Amhara. However, high perishability and the seasonal availability of these products should be taken into account.The consumption of other types of fruits and vegetables is infrequent among children. More than half of the women consume other vegetables (most likely onion). Their availability in the region is good with most markets stocking onions, tomatoes and bananas. They are also reasonably affordable. Therefore, other fruits and vegetables could have a high potential for improving dietary diversity in Amhara. However, high perishability and the seasonal availability of these products should be taken into account.One key challenge in increasing ASF consumption in Ethiopia relates to religious fasting practices. This is particularly the case for the Orthodox Christians who comprise more than 80 Young children and pregnant and lactating women are exempted from fasting. To underline this, in January 2016, the Ethiopian Orthodox Church together with USAID issued a press release instructing women and children to abstain from fasting during the first 1,000-day period. 14Still, children are perceived to be influenced by fasting practices because adults in the households do not want to prepare foods containing ASF. A qualitative formative study conducted during the Lent fasting period in the Amhara region highlights this but also points out that eggs and milk were considered more acceptable to be fed to children as they required less preparation -and perhaps more importantly, did not require slaughtering (Kim et al., 2018).To address these issues, behavior change interventions will need to carefully tailor the messaging during the fasting periods. Given the caregiver acceptability and market availability, it may be strategic to focus the behavior change messages more on eggs and dairy instead of meat products. An alternative strategy would be to promote lightly processed and preserved meat products (e.g. dried meat) so that no slaughtering is required. Finally, the Alive & Thrive project has successfully involved religious leaders to improve feeding practices in the Amhara region. 15This seems like a good model to follow.This study has limitations. First, although we use geographically widespread surveys to assess women's dietary diversity, in the absence of regionally representative data, we cannot be sure","tokenCount":"5510","images":["192175356_1_1.png","192175356_2_1.png","192175356_8_1.png","192175356_12_1.png","192175356_13_1.png","192175356_14_1.png","192175356_15_1.png","192175356_16_1.png","192175356_17_1.png","192175356_18_1.png","192175356_19_1.png","192175356_39_1.png","192175356_40_1.png"],"tables":["192175356_1_1.json","192175356_2_1.json","192175356_3_1.json","192175356_4_1.json","192175356_5_1.json","192175356_6_1.json","192175356_7_1.json","192175356_8_1.json","192175356_9_1.json","192175356_10_1.json","192175356_11_1.json","192175356_12_1.json","192175356_13_1.json","192175356_14_1.json","192175356_15_1.json","192175356_16_1.json","192175356_17_1.json","192175356_18_1.json","192175356_19_1.json","192175356_20_1.json","192175356_21_1.json","192175356_22_1.json","192175356_23_1.json","192175356_24_1.json","192175356_25_1.json","192175356_26_1.json","192175356_27_1.json","192175356_28_1.json","192175356_29_1.json","192175356_30_1.json","192175356_31_1.json","192175356_32_1.json","192175356_33_1.json","192175356_34_1.json","192175356_35_1.json","192175356_36_1.json","192175356_37_1.json","192175356_38_1.json","192175356_39_1.json","192175356_40_1.json","192175356_41_1.json"]}

data/part_2/0286791193.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"664805415258c7e888d4ccbbf5a4d2ac","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/5e1d9b2c-050c-4cba-be56-31cdcc1171fb/retrieve","description":"After a decade of stagnation during the 1990s, investments and human resource capacity in public agricultural research and development (R&D) averaged more than 20 percent growth in Sub-Saharan Africa (SSA) during 2001-2008. In 2008, the region spent $1.7 billion on agricultural R&D (in 2005 purchasing power parity dollars)—or $0.8 billion (in 2005 constant US dollars)—and employed more than 12,000 full-time equivalent (FTE) agricultural researchers. Most of this growth, however, occurred in only a handful of countries and was largely the result of increased government commitments to augment incommensurately low salary levels and to rehabilitate neglected infrastructure, often after years of underinvestment. Many countries—particularly those in francophone West Africa, which are threatened by extremely fragile funding systems—face fundamental capacity and investment challenges. National investment levels in such countries have fallen so low as to leave them dangerously dependent on often volatile, external funding sources. Despite the overall capacity growth recorded, average qualification levels have deteriorated in a number of countries. Some reported large influxes of BSc-qualified scientists, often in response to prolonged recruitment restrictions, further straining already inadequate training opportunities and far exceeding the capacity for appropriate oversight and mentorship by senior researchers, given years of nonreplacement of retiring and departing scientists. Notwithstanding the challenges facing many countries, renewed commitment to agricultural R&D by governments and donors indicates improved prospects for agricultural R&D for a number of African countries. Regional initiatives are also a key factor in increasing research coordination and collaboration and ensuring the prioritization and efficiency of research. Increased and sustained investment from national governments, regional and international organizations, and large donors will go a long way toward stabilizing investment and capacity levels and enabling real progress for agricultural R&D in the region. Building on the strategic recommendations of various highly influential reports and meetings, and taking into account the various investment and capacity challenges outlined in this report, four key areas with strong implications for policy must be addressed by governments, donors, and other stakeholders: (1) decades of underinvestment in agricultural R&D; (2) excessive volatility in yearly investment levels; (3) existing and imminent challenges in human resource capacity; and (4) the need to maximize regional and subregional cooperation in agricultural R&D.","id":"1772539163"},"keywords":[],"sieverID":"9f58d0ea-eb50-4156-83e0-926759075b37","pagecount":"44","content":"The International Food Policy Research Institute (IFPRI) was established in 1975. IFPRI is one of 15 agricultural research centers that receives its principal funding from governments, private foundations, and international and regional organizations, most of which are members of the Consultative Group on International Agricultural Research.T his report was developed with funding from the Bill & Melinda Gates Foundation and additional support from the International Food Policy Research Institute. The authors thank their country collaborators and the more than 370 agricultural research agencies that participated in national survey rounds; without their commitment, this report would not have been possible. The authors also thank Kathleen Flaherty and Michael Rahija for their excellent help in preparing this publication and the two anonymous reviewers for reviewing and commenting on a draft version.IFPRI gratefully acknowledges the generous unrestricted funding from Australia, Canada, China, Denmark, Finland, France, Germany, India, Ireland, Italy, Japan, the Netherlands, Norway, the Philippines, South Africa, Sweden, Switzerland, the United Kingdom, the United States, and the World Bank.A fter a decade of stagnation during the 1990s, investments and human resource capacity in public agricultural research and development (R&D) averaged more than 20 percent growth in Sub-Saharan Africa (SSA) during 2001-2008. In 2008, the region spent $1.7 billion on agricultural R&D (in 2005 purchasing power parity dollars)-or $0.8 billion (in 2005 constant US dollars)-and employed more than 12,000 full-time equivalent (FTE) agricultural researchers. Most of this growth, however, occurred in only a handful of countries and was largely the result of increased government commitments to augment incommensurately low salary levels and to rehabilitate neglected infrastructure, often after years of underinvestment. Many countries-particularly those in francophone West Africa, which are threatened by extremely fragile funding systems-face fundamental capacity and investment challenges. National investment levels in such countries have fallen so low as to leave them dangerously dependent on often volatile, external funding sources. Despite the overall capacity growth recorded, average qualification levels have deteriorated in a number of countries. Some reported large influxes of BSc-qualified scientists, often in response to prolonged recruitment restrictions, further straining already inadequate training opportunities and far exceeding the capacity for appropriate oversight and mentorship by senior researchers, given years of nonreplacement of retiring and departing scientists.Notwithstanding the challenges facing many countries, renewed commitment to agricultural R&D by governments and donors indicates improved prospects for agricultural R&D for a number of African countries. Regional initiatives are also a key factor in increasing research coordination and collaboration and ensuring the prioritization and efficiency of research. Increased and sustained investment from national governments, regional and international organizations, and large donors will go a long way toward stabilizing investment and capacity levels and enabling real progress for agricultural R&D in the region.Building on the strategic recommendations of various highly influential reports and meetings, and taking into account the various investment and capacity challenges outlined in this report, four key areas with strong implications for policy must be addressed by governments, donors, and other stakeholders:(1) decades of underinvestment in agricultural R&D;(2) excessive volatility in yearly investment levels;(3) existing and imminent challenges in human resource capacity; and (4) the need to maximize regional and subregional cooperation in agricultural R&D.The 2003 Maputo Declaration directed all member countries of the African Union (AU) to increase agricultural investments to at least 10 percent of their national budgets. To gauge progress toward this target, the Comprehensive Africa Agriculture Development Programme (CAADP) under the AU's New Partnership for Africa's Development (NEPAD) agreed to monitor agricultural expenditures, setting a 6 percent yearly target for growth in agricultural gross domestic product (AgGDP) in countries where agriculture plays a dominant economic role. During 2000-2008, the region's GDP grew by an average of more than 5 percent per year-more than twice as fast as in the two preceding decades-but AgGDP grew by only 3 percent per year on average (Montpellier Panel 2010).One of CAADP's four foundational pillars focuses on increasing investments in agricultural research, extension, education, and training as a means of promoting growth in agricultural productivity (NEPAD-CAADP 2010). Moreover, NEPAD's African Ministerial Council on Science and Technology established and adopted a Consolidated Plan of Action for developing regional science and technology (S&T). This plan calls for substantial increases in national R&D budgets, with each country taking concrete measures to allocate at least 1 percent of its GDP to R&D (NEPAD 2006). In order to measure, monitor, and benchmark the inputs, outputs, and performance of agricultural S&T systems at the national and regional levels and to assess progress toward the successful implementation of CAADP targets related to S&T, quantitative data are essential. S&T indicators are an indispensable tool when assessing the contribution of agricultural S&T to agricultural growth and, more generally, to economic growth. They assist research managers and policymakers in formulating policy and making decisions about strategic planning, priority setting, monitoring, and evaluation. They also provide information to governments and others involved in the public debate on the state of agricultural S&T at national, regional, and international levels.This report assesses long-term trends in investments and human resource capacity in public agricultural R&D in SSA, particularly focusing on developments during 2000-2008. The analysis uses information from a set of country notes prepared by the Agricultural Science and Technology Indicators (ASTI) initiative of the International Food Policy Research Institute (IFPRI), using comprehensive datasets derived from primary surveys conducted during 2009-2010. The sample includes 32 countries that contribute more than 90 percent of the region's agricultural GDP. 1 These datasets have been linked with investment and human resources data collected in the region by ASTI during 2002-2003, as well as with ASTI's global datasets, to provide a wider context for agricultural R&D investment trends in SSA over time and in contrast to other regions. 2 The analysis in this report concludes with suggested future directions needed to address the financial and human capacity challenges that many countries currently face.The institutional structure of most of the region's agricultural research has remained constant since 2000, 3 but Mozambique and Tanzania are important exceptions. The Agricultural Research Institute of Mozambique was established in 2005 with the objective of centralizing agronomic, veterinary, animal production, and forestry research (Flaherty, Mazuze, and Mahanzule 2010). In contrast, Tanzania reversed its earlier decision to consolidate national research activities under the Department of Research and Development (DRD) and instead created a dedicated agency for R&D activities related to livestock (Flaherty  and Lwezaura 2010). R&D systems in some other countries also underwent important structural changes. The National Agricultural Research Organisation (NARO) in Uganda, for instance, was transformed from an agency to a consortium in efforts to improve its response to client needs and its ability to oversee and guide agricultural R&D service provision (Flaherty, Kitone, and Beintema 2010). In Nigeria, the Agricultural Research Council of Nigeria (ARCN) was created to improve coordination and linkages across research agencies and between research providers and clients, and to redress overlaps in mandates within the institutes (Flaherty et al. 2010a).In most of the smaller countries, agricultural research is undertaken by a handful of government agencies and university faculties; systems in the large countries like Ghana, Kenya, Nigeria, South Africa, and Sudan are, understandably, far more complex. Nevertheless, the majority of SSA countries have a single national agricultural research agency that accounts for the bulk of agricultural R&D capacity and investments. Examples include the National Agricultural Research Institute of Niger (INRAN), the Togolese Agricultural Research Institute, and the National Agricultural Research Institute in Eritrea. In some countries, an umbrella organization like Ghana's Council for Scientific and Industrial Research (CSIR) or South Africa's Agricultural Research Council (ARC) oversees and coordinates the R&D activities of a large number of commodity or thematic centers, whereas in a country like Mauritania, the national crop, livestock, and fisheries research agencies operate independently of each other without a coordinating body. Overall, the government sector still dominates agricultural research in the region, but its relative share has declined over time. In 1991, government agencies employed 82 percent of full-time equivalent (FTE) public agricultural R&D staff in SSA on average, but this share had fallen to 73 percent in 2008 (see Box 1 for an explanation of FTEs).The absolute number of FTE researchers in the higher education sector for the 32 countries included in this study-hereafter referred to as \"the ASTI countries\"-more than doubled during 1991-2008, mainly as a result of the establishment of new higher education units involved in agricultural research. Most of these new agencies were located in Nigeria and Sudan. Despite the high and increasing number of higher education agencies conducting agricultural research in SSA, the individual capacity of most of them, in terms of FTE researcher numbers, remains small. During 1991-2008, the higher education sector's share of public agricultural research staff (in FTEs) grew from 15 to 24 percent. Although the amount of time staff spend on research has gradually risen over time, in 2008 it still amounted to less than 25 percent of their time on average.In contrast, the nonprofit sector's share increased marginally during this period, from 2 to 3 percent. The sector's overall growth has been slow compared with the government and higher education sectors. Most nonprofit institutions in SSA are linked to producer organizations and receive most of their funding via levies on production or exports. Although other forms of nonprofit institutions exist in a number of countries, including Benin, Madagascar, and Togo, they play a limited role in agricultural research.Little information could be obtained on capacity or expenditure trends in private agricultural R&D in SSA. 4 Most private for-profit companies still outsource their research to government agencies or universities, or they import technologies from abroad. Only a limited number of private companies operate their own research programs, and the companies that do so often employ only a handful of researchers  5 have all made considerable progress in coordinating agricultural research activities in their member countries through the establishment of various research networks. These networks have proved to be a successful method of collaboration and information sharing. They allow specialization of particular national agricultural research systems in certain fields and have proved to be particularly beneficial for small countries lacking a critical mass of agricultural R&DComparing R&D data is a highly complex process due to important differences in price levels across countries. The largest components of a country's agricultural R&D expenditures are staff salaries and local operating costs, rather than capital investments that are traded internationally. For example, the wages of a field laborer or lab assistant at a research facility are much lower in Kenya than in any European country; locally made office furniture in Senegal is considerably cheaper than a similar set of furniture bought in the United States.Standard market exchange rates are the logical choice for conversions when measuring financial flows across countries. However, they are far from perfect currency converters for comparing economic data. At present, the preferred conversion method for calculating the relative size of economies or other economic data, such as agricultural R&D spending, is the purchasing power parity index. PPPs measure the relative purchasing power of currencies across countries by eliminating national differences in pricing levels for a wide range of goods and services. They are also used to convert current GDP prices in individual countries to a common currency. In addition, PPPs are relatively stable over time, whereas exchange rates fluctuate considerably (for example, the fluctuations in the US dollar-euro rates of recent years).ASTI bases its calculations of human resource and financial data on full-time equivalent staffing, or FTEs, which take into account the proportion of time researchers spend on R&D activities. University staff members, for example, spend the bulk of their time on nonresearch-related activities, such as teaching, administration, and student supervision, which need to be excluded from research-related resource calculations. As a result, four faculty members estimated to spend 25 percent of their time on research would individually represent 0.25 FTEs and collectively be counted as one FTE.Sources: Beintema and Stads (2008a and forthcoming) and ASTI's website (www.asti.cgiar.org/methodology).staff. Nonetheless, reaching agreement on regional priorities has often been difficult as countries continue to pursue self-sufficiency in fields of agricultural R&D in which they are weak (IAC 2004). Aside from regional networks, most of the centers under the Consultative Group on International Agricultural Research (CGIAR) have offices in Africa, often with considerable research facilities and staffing. These centers are a key source of agricultural innovation for many countries, providing new crop varieties that are subsequently tested by national agricultural R&D agencies under local conditions. Several other international and regional organizations have a presence and conduct agricultural research in SSA, including the Center for International Cooperation and Agricultural Research for Development (France), the Institute for Research for Development (France), and the World Vegetable Center. Staff and expenditure levels at these international centers are excluded from the analysis of this report because its focus is on national level investments and capacities.A bsolute levels of public agricultural R&D spending and staffing varied considerably across the 32 ASTI countries (Table 1 ratios of 3.9 and 5.2 percent, respectively. For Mauritius this reflects the high level of investment in sugarcane research. In contrast, a large number of countries recorded intensity ratios of 0.5 percent or lower. Gauging researcher numbers or spending levels against total population or economically active agricultural population also facilitates cross-country comparisons (Figure 1b). 6 In 2008, SSA employed 68 FTE researchers per million economically active agricultural population. Again, wide variation occurred across the ASTI countries. Botswana, Gabon, and Namibia employed more than 200 FTE researchers per million economically active agricultural population. A particularly high ratio in Mauritius (3,103 FTEs) again reflects the leading role the country plays in sugarcane research. Gabon, Nigeria, and Sudan had relatively high ratios of research staff compared with their spending intensities.Although intensity ratios provide useful insights into relative investment and capacity levels across countries, they take into account neither the policy and institutional environment within which agricultural research occurs nor the broader size and structure of a country's agricultural sector and economy. For example, small countries require more human resource and capital investments, because they are unable to benefit from the economies of scale available to larger countries. More important, a high intensity can actually reflect reduced agricultural output rather than higher investment, as is noted for Botswana. Detailed analysis is needed to ensure a clear understanding of the implications of intensity ratios across countries.In the late-twentieth century, greater instability was evident in agricultural R&D in SSA compared with other world regions, mainly due to political unrest, social and economic hardship, and institutional vulnerability. Spending levels fluctuated in many countries, and overall growth slowed over time (Beintema and Stads 2006). This trend appears to have reversed, at least in the aggregate for the 2001-08 period. In 2008, public agricultural R&D investments for SSA as a whole-based on data for the 32 ASTI countries and estimates for 14 other, often small countries-totaled $1.7 billion in inflation-adjusted PPP dollars-or $0.8 billion in 2005 constant US dollars. This was almost 20 percent higher than the $1.4 billion (in 2005 PPP dollars)or $0.6 billion (in 2005 US dollars)-recorded in 2001 and marks a considerable shift away from the slow 1.0 percent yearly growth in agricultural R&D investments recorded during the 1990s (Figure 2 top graph). 7 Overall, 2008 investment levels in SSA were comparable to those in individual countries like Brazil and India with high investment levels (Box 2).Growth in agricultural R&D capacity was strong in the 1970s and 1980s, at 5.4 and 3.8 percent per year, respectively, but during the 1990s it slowed to a mere 1.3 percent per year. Since the turn of the millennium, growth in researcher numbers has once again accelerated. In some countries, renewed growth was due to the cessation of long-term recruitment bans, whereas in other countries it stemmed from increased involvement in agricultural research by the higher education sector. In 2008, SSA employed 12,120 FTE researchers, compared with 9,824 FTEs in 2001.The relative growth, in terms of the intensity ratio, has also increased since the turn of the millennium. The aforementioned 2008 investment of $0.61 for every $100 of agricultural output was considerably higher than the average of $0.49 during the late-1990s (Figure 2 bottom data). This was mainly due to the aforementioned low growth in agricultural R&D spending during that decade, along with higher increases in AgGDP. Although SSA's intensity ratio has increased since 2000, it is still below the levels of the late 1980s and early 1990s.Country-level data reveal that the regionwide spending and capacity increases of roughly 20 percent during 2001-2008 were largely driven by only a handful of countries. More than one-third of the growth in public agricultural R&D spending during this period is attributable to a $110 million increase in spending in Nigeria. Ghana, Sudan, Tanzania, and Uganda also experienced relatively high increases in total spending of between $25 million and $56 million each (in 2005 PPP dollars) (Figure 3a). In contrast, Ethiopia and South Africa experienced notable declines ($28 million and $12 million, respectively). Nigeria was also the main driver of the regional growth in researcher numbers, accounting for 724 of the region's 2,285 increase in FTE researcher numbers during 2001-2008 (Figure 3b). Ethiopia, Kenya, and Sudan reported significant R&D staff increases as well.South Africa experienced the largest decline in public agricultural researcher numbers (140 FTEs), whereas changes in investment and capacity levels in the remaining ASTI countries were less severe in absolute terms during 2001-2008.   In 2000 (the latest year for which global data were available), SSA contributed 5 percent of the $25 billion spent on public agricultural R&D globally (in 2005 PPP prices), compared with 7 percent in 1981. This decline resulted in part from relatively low yearly spending growth during the 1990s, combined with a very strong increase in public agricultural R&D spending in the Asia-Pacific region, specifically in China and India (Beintema and Stads 2010). No recent information on worldwide public agricultural R&D spending is available, but investments in Brazil, China, and India have continued to rise, so SSA's overall share is unlikely to have increased despite gains since 2000 (Figure 2).Increases in Brazil, China, and India were mostly the result of renewed government commitment to public agricultural R&D rather than increased donor funding, which is low compared with levels in many SSA countries. By way of magnitude, Brazil and India both spent slightly less on public agricultural R&D than SSA as a whole.   U nsurprisingly, the countries identified as the main drivers of regional growth in agricultural R&D spending and capacity were those with the largest absolute spending and capacity, as indicated in Table 1. These \"Big Eight\" 8 -Nigeria, South Africa, Kenya, Ghana, Uganda, Tanzania, Ethiopia, and Sudan-accounted for 70 percent of regional public agricultural R&D spending and 64 percent of all researchers in 2008 (Table 2). This is a considerable increase above 1991 levels, when the corresponding shares were 53 and 55 percent, respectively. Because of the relative size of these countries, it is not surprising that developments in capacity or spending, whether positive or negative, have such a significant impact on regional trends, so these countries warrant individual attention.The Drivers of Regional Growth Africa's \"Big Eight\" As established, the overall increase in agricultural R&D investments in SSA is driven by a handful of countries, but the underlying cost-category breakdown reveals that different factors drove growth. The rapid increase in Ghanaian agricultural R&D spending, for instance, was driven almost entirely by increased salary expenditure at CSIR rather than expanded research activities or greater investment in equipment or infrastructure (Figure 4A). The unprecedented increase in expenditure on salaries needs to be understood in the context of years of underfunding, during which salary levels became increasingly incommensurate and uncompetitive (Flaherty, Essegbey, and Asare 2010).At DRD in Tanzania, on the other hand, spending on salaries has remained relatively stable over time (Figure 4B). Prior to 2005, DRD was highly dependent on World Bank funding and capital investments were high, but spending plummeted once the World Bank project ended. Thereafter, the Tanzanian government increased its commitment to agriculture and agricultural research over time, thereby allowing greater expenditure on research activities and equipment and infrastructure (Flaherty and Lwezaura 2010). Similarly,  the government of Nigeria significantly increased its funding to the national agricultural research institutes (NARIs) and other government agencies from the late 1990s, enabling the purchase of equipment and the rehabilitation of facilities (Figure 4C). Nonetheless, despite these remarkable increases, investment levels remain below those required to restore facilities to prior levels and to sustain the country's agricultural research needs (Flaherty et   4D). In 2005, the Ugandan government approved a much-needed 100 percent salary increase and thereafter a yearly increase of 10 percent.Understandably, this had a significant impact on overall spending levels (Flaherty, Kitone, and Beintema 2010).Though increases and decreases in the absolute levels of agricultural R&D spending and capacity of the Big Eight overshadow those of many of the smaller countries in SSA, a closer look at relative shifts in investment and capacity levels over time reveals some interesting cross-country differences and challenges. Some of the region's smallest countries have such low and declining levels of investment and human resource capacity that the effectiveness of their national agricultural R&D could be questioned. This also highlights the need for regional initiatives to address the unique needs of small countries and to take advantage of collaborative synergies. Considerable differences were reported not only in absolute investment levels among the ASTI countries but also-more importantly-in the magnitude of growth over time. During 2001-2008, 13 of 29 countries (excluding Rwanda, Mozambique, and Zimbabwe) experienced negative yearly growth in public agricultural R&D spending, ranging from -1.6 to -12.4 percent per year (Figure 5A), which is sizeable given that spending in SSA as a whole actually increased throughout this period. Of these 13 countries, 7 are francophone countries located in West and Central Africa. With the exception of Gabon and Mali, these countries also experienced negative growth during the 1990s, which is a major area of concern. Falling investments in agricultural R&D in these countries resulted mainly from the completion of large donor-funded projects, often financed through World Bank loans (Burkina Faso, Guinea, Senegal, and Togo). Comparing the 2001-2008 growth rates with those of the 1990s clearly illustrates the volatility of agricultural spending levels for many of the region's countries. Eritrea and Ethiopia, for example, experienced negative growth during 2001-2008 (-12.4 and -4.5 percent per year, respectively) following a decade of particularly high positive growth (32.7 and 11.0 percent, respectively), which is indicative of high dependency on donor funding (see the next section for details).In contrast, agricultural R&D investments in a number of countries increased substantially after 2000. Eight countries recorded yearly growth rates of more than 6 percent, including four of the Big Eight countries (Ghana, Sudan, Tanzania, and Uganda). Spending in Nigeria, which accounted for more than one-third of the absolute growth in spending during 2001-2008, grew at a comparatively moderate average rate of 3.2 percent per year. For some countries, growth reflected the re-establishment of agricultural R&D systems after periods of political unrest, whereas in others-such as Nigeria, Sudan, Tanzania, and Uganda-growth stemmed from increased national government commitment to agriculture in general and to agricultural R&D in particular.Growth in agricultural research staffing varied less across countries compared with total spending (Figure 5B). In line with reduced spending levels, a number of francophone countries in West and Central Africa also reported declining capacity during 2001-2008. Gabon, Niger, the Republic of Congo, and Togo recorded yearly declines of -1.8 percent or more. This is a worrisome trend because capacity in these countries was low to begin with. Even more challenging, the pool of well-qualified and experienced researchers in many countries is aging; many will be lost to retirement in the next decade (see \"Staffing Trends\" for details). Nevertheless, research staffing increased or remained fairly constant in most countries during 2001-2008. Despite strong losses in spending levels in Eritrea and Ethiopia throughout this period, researcher numbers increased by 6.6 and 6.0 percent per year, respectively. In Ethiopia this growth was driven by the development of the RARIs and the higher education sector. Notes: The figure excludes Mozambique, Rwanda, and Sierra Leone (for both spending and capacity) and Malawi and Zimbabwe (for spending only) because time-series data did not date back to 2001. Growth rates are missing for Eritrea, Mauritania, and Namibia (for spending and capacity) and Tanzania and Uganda (for spending only) due to lack of time-series data for the full period of 1991-2001. Compound yearly growth rates are calculated using the least-squares regression method.Unsurprisingly, variation is significant across countries and agencies (Figure 6). The government funds the bulk of agricultural R&D activities in some countries, whereas other countries are extremely dependent on external funding. A number of R&D agencies generate significant revenues by selling goods and services, whereas for others the proceeds of sales are channeled to the national treasury, thereby eliminating the incentive to pursue such endeavors.In addition to differing across countries at a specific time, sources of funding differ substantially across time. KARI, for example, is a relatively wellfunded institute, receiving constant support from the Kenyan government, attracting large sums of donor funding, and generating its own revenues (Figure 7A; Flaherty et al. 2010b). In contrast, the Mauritius Sugar Industry Research Institute (MSIRI), a nonprofit agency, is almost entirely funded through a tax on sugar production (Figure 7B). Given falling world market prices of sugarcane in recent years, overall funding to MSIRI has declined (Rahija, Ramkissoon, and Stads 2010). Côte d'Ivoire's National Agricultural Research Center (CNRA) is funded largely by the private sector (mostly coffee, cocoa, rubber, and oil palm producers) supplemented by limited funding from the national government (Figure 7C). Foreign donors stopped supporting the center shortly after the civil war broke out in 2002, and they have not returned since (Stads and Doumbia 2010). Niger's INRAN is an example of an institute that was extremely dependent on donor and development bank funding during the 1990s, but with the completion of a large World Bank-financed project in 1998, the institute fell into severe financial crisis, and the situation remains precarious (Figure 7D; Stads, Issoufou, and Massou 2010). Like INRAN, many other African institutes have extremely fragile and donor-dependent funding systems.The higher education sector is excluded from this section because agricultural R&D funding data for this sector are extremely difficult to obtain. Given that teaching is the core business of agricultural faculties, dedicated R&D budgets are rare or ad hoc. Many universities fund R&D activities through public grants, student tuition fees, and internally generated resources. Like the government sector, donor funding plays an important role in many countries. African universities often maintain close linkages with universities in developed countries and benefit from funding as part of joint research projects. Although the role of higher education in agricultural R&D in SSA could increase if it were to receive sustainable research funding, it is unlikely that funding would increase to levels seen in countries like India, Mexico, or the United States. Agricultural higher education agencies in SSA are often fragmented and fall under universities with a broader (nonagricultural) mandate; independent agricultural universities or colleges are scarce. Further, many higher education agencies in SSA are underfunded and understaffed and have therefore limited time available for activities other than education (World Bank 2007c).F unding for African agricultural R&D is derived from a variety of sources, including national governments; donors, development banks, and (sub)regional organizations; producer organizations; the private sector; and internally generated revenues. Given that funding data were not available for all 32 ASTI countries (including some important Big Eight countries), it is not possible to present regionwide funding trends. Evidence presented above, however, indicates that growth in spending in Ghana, Nigeria, Sudan, Tanzania, and Uganda-the main drivers of regional growth-was largely the result of significant injections of government funding.Who's Footing the Bill? Notes: Gabon, Ghana, Malawi, Nigeria, the Republic of Congo, and Zimbabwe were excluded due to lack of complete data. SROs indicates subregional organizations; \"producer organizations\" include contributions through export or production levies; \"own income\" includes sales of goods and services and contractual research performed for public and private agencies. Funding shares for some research agencies fluctuated over time.  2009). Shares of funding in this category are generally much higher in SSA (Figure 6), although they are negligible in many middle-income countries in southern Africa or in countries afflicted by political unrest. Donor funding is provided by:• multilateral bodies, such as the European Union, CGIAR, and the United Nations;• bilateral donors, such as foreign governments and private foundations; • (sub)regional organizations such as FARA, ASARECA, and CORAF/WECARD, which, in turn, are also recipients of donor resources (largely through multidonor trust funds); and • development banks, such as the World Bank and the African Development Bank, which provide loans and grants.Contributions from donors, development banks, and (sub)regional organizations have declined for a large number of countries, which can be attributed largely to the overall decline in World Bank-funded projects since the late-1990s and early 2000s. Such projects variously focused on agricultural and economic development, and in terms of agricultural research generally aimed to reshape national agricultural research systems, provide much-needed Notes: SROs indicates subregional organizations; \"producer organizations\" include contributions through export or production levies; \"own income\" includes sales of goods and services and contractual research performed for public and private agencies.CNRA's funding structure constitutes a unique and exemplary regional case study. The second National Agricultural Services Support Project (PNASA II), which was launched in 1998 and administered by the World Bank, stipulated that CNRA would be structured as a public-private entity, with 40 percent of its funding being contributed by the government and 60 percent derived from the private sector. To this end, the Inter-Professional Fund for Agricultural Research and Extension (FIRCA) was established in 2002. FIRCA relies on financial contributions not only from the government but also from the country's producers, who pay membership subscription dues through commodity-specific producer organizations. At least 75 percent of the subscription fees raised through agricultural production in a given subsector are allocated to programs serving the needs of that subsector. The remaining funds are allocated to a solidarity fund, and a marginal share underwrites FIRCA's operating costs. The purpose of the solidarity fund is to finance programs designed to serve production sectors (mostly food crops) unable to raise sufficient funding through their own subscription fees or that have difficulty doing so because of the way they are structured. In 2008, the amounts raised and contributed by the coffee, cocoa, rubber, and oil palm producer organizations represented 91 percent of total subscription dues raised by all the producer organizations combined. Private Sector. Commercializing research outputs is often achieved through partnerships with the private sector. 9 In a few of the region's countries, private investment in agricultural R&D is increasing and creating an income stream for agricultural research agencies (Echeverría and Beintema 2009). In Senegal, for instance, large companies like SODEFITEX (cotton producers) and SUNEOR (groundnut producers) fund research activities conducted by the Senegalese Agricultural Research Institute (ISRA), the country's main government agricultural R&D agency (Stads and Sène 2010). These activities are often ad hoc, such as research on cash crops at ISRA in Senegal, but in many countries they are formalized, for example at CNRA in Côte d'Ivoire (Box 3). Although private-sector funding offers valuable potential support in developing financial and human resource capacity in agricultural R&D, it can be implemented only in countries with the necessary enabling policy environment, including strong intellectual property legislation, minimal barriers to importing and testing new technologies, and tax exemptions on research expenditures and venture capital (Alston, Pardey, and Piggott 2006). In many countries in SSA, the necessary policy environment is extremely weak or nonexistent. Furthermore, in many countries internally generated income is channeled back to the treasury, eliminating any incentive for research agencies to explore contract-based research for the private sector. It should be borne in mind that the proliferation of contracts to carry out research on behalf of agrobusinesses entails the risk of excessively skewing the research agenda away from basic research in favor of applied research and seed multiplication. Attention to the balance between these types of research is therefore needed when private contract-based funding is expanded.Research can also be funded through levies on agricultural production or exports. Commodity levies are important in several countries in SSA, including Kenya (coffee and tea), Malawi (tea and tobacco), Mauritius (sugar), South Africa (sugar), Tanzania (tea and coffee), and Zambia (cotton). Research levies have often been established in countries experiencing long-term funding instability that have an identifiable group of beneficiaries able to contribute to the cost of research. In other countries, levies date back to the colonial times. Levies are mostly applied to cash or export crops because they pass through a limited number of collection points. They are a less efficient mechanism for commodities and countries where most of the output is consumed on-farm or are traded in local markets because collection costs would be too high. Levies have the benefit of involving farmers in the research agenda and providing relevant research outputs. They may provide additional funding for agricultural R&D, but there is always the risk that they will simply replace rather than augment government funding.Levies do, however, have a number of potentially negative impacts. These include the risks of promoting price disincentives and suboptimal funding levels due to spillovers and delayed research outputs. One way to counter these potential risks is for governments to provide funding to match the commodity levies (Kangasniemi 2002; Echeverría and Beintema 2009).Competitive funding mechanisms have gained ground but are limited in Africa compared with other developing regions of the world. These funds typically finance R&D through grants allocated on the basis of scientific merit and congruence with broadly defined agricultural R&D priorities. Competitive funds are believed to attract research resources while lowering execution costs, encouraging demand-driven activities, and promoting research partnerships (Echeverría and Beintema 2009). A main concern, however, is long-term sustainability, given that many mechanisms are dependent on external funding. In the late 1990s and early 2000s, various competitive funds were established as components of World Bank projects in countries like Kenya, Mali, Senegal, and Tanzania. Though many funds have built-in sustainability mechanisms, overall funding will decline once initial endowments have been exhausted. Other countries, including Nigeria, Uganda, and Zambia, have attempted to establish competitive grant schemes, but without substantial initial injections of funding, these schemes have generally faltered. 10  Women's participation in agricultural research has increased around the world over the past several decades, but it remains low in many countries, especially in the developing world. In 2008, 22 percent of FTE researchers in the ASTI countries were female, compared with 18 percent in 2001. Shares of female scientists in East and southern Africa are generally higher than those in West Africa. More than 30 percent of the agricultural researchers in Botswana, Eritrea, Mauritius, South Africa, and Sudan in 2008 were female. In contrast, corresponding shares in Ethiopia, Guinea, Mauritania, Niger, Senegal, and Sierra Leone were less than 10 percent. Overall, women are more represented in junior roles requiring only a BSc-level qualification. Female researchers also tend to be more prevalent in the higher education sector. Although the share of professional women employed in agriculture is increasing, the vast majority are entry-level staff and students (that is, BSc graduates or students). The need for greater representation by women in agricultural research in SSA is urgent. Women in senior positions as scientists, research managers, lecturers, and professors can provide valuable insights and perspectives to assist research agencies in addressing the unique and pressing challenges facing African farmers, many of whom are women. An increasing number of support staff (technicians, research assistants, laboratory assistants) have BSc, MSc, and occasionally PhD qualifications, but they are not classified as researchers (Figure 10). In Senegal, for example, the minimum requirement for a researcher is an MSc degree, so the 105 scientists employed at ISRA with BSc qualifications are classified as technicians (Stads and Sène 2010). Half of the technicians and other research support staff at NARO in Uganda held MSc or BSc degrees, and most attained these qualifications without official NARO backing. Although the number of research positions at NARO increased in recent years, promotional opportunities remain limited because applicants must meet specific minimum requirements, including having an MSc degree (Flaherty, Kitone, and Beintema 2010). Unlike the situations in Senegal and Uganda, support staff at Tanzania's DRD are promoted to researcher status upon obtaining their BSc degrees (Flaherty and Lwezaura 2010). The pool of degree-qualified support staff is sizeable in some countries. In Senegal, 43 percent of all degree-qualified research staff are technicians, and in Mauritius, Nigeria (government agencies only), and Uganda this share is about 25 percent (Figure 10). It is important to capture quantitative information on research technicians, whogiven proper training and promotional opportunitieswill be a valuable resource in the future development of African agricultural R&D.In many of the region's countries, salary and retirement packages and conditions of service are poor. In addition, many agencies have outdated infrastructure and insufficient operating budgets to conduct research. Even with the aforementioned increase in training opportunities in a number of countries, research agencies have difficulty retaining staff once they attain higher degrees and can attract offers of better remuneration and conditions, either in the higher education or private sectors or abroad (FARA 2006;  World Bank 2007c). A major concern in many countries, particularly in West and Central Africa, is a rapidly aging pool of scientists, many of whom will approach retirement within the next decade. In Cameroon and the Republic of Congo, for instance, agricultural researchers at the main agencies are well over 50 years old on average. In 2007, 27 percent of researchers in Ghana's CSIR were 51 years or older, and half were between 41 and 50 years old (Figure 11). 12 That year, 36 percent of KARI's researchers in Kenya were 51 years or older. With 59 percent of its researchers averaging well over 50 years in 2007, Senegal has one of the oldest pools of scientists in West Africa. In 2003, ISRA in Senegal employed 70 PhD-qualified scientists compared with just 54 in 2008. Some of these scientists left ISRA to take advantage of opportunities in the higher education and private sectors, where salaries are reported to be up to three times higher than in the public sector; many of the more senior researchers also retired (Stads and Sène 2010).A number of agencies are instituting staff retention strategies. KARI in Kenya, for example, has introduced regular staff performance evaluations, which form the basis for promotion. The institute is also working on other incentives, such as better medical benefits. KARI also requires that staff offered training commit to working for the agency for a set period of time. In the mid-2000s, KARI and other institutes convinced the government to increase the retirement age from 55 to 65 years, not only to address the shortage of senior staff, but also to offset the time it takes for staff to qualify for and then undertake MSc and PhD training. It made sense to extend the productivity of these researchers once they became fully qualified. The higher retirement age also provided an incentive for junior staff, including technicians with diplomas, to pursue higher training, even through self-sponsorship (Flaherty et al. 2010b).Attracting and retaining staff is an even more seriously problem in countries with small research capacities. The National Agricultural Research Institute in The Gambia is a case in point. During 2003-2009, the institute lost seven PhD-qualified researchers through retirement, departure, or death, and many of its remaining staff members lack advanced training or experience (Stads and Manneh 2010). This lack of a critical mass of well-qualified researchers in small countries also highlights the need for regional initiatives focusing on the needs and vulnerabilities of such countries. N ew quantitative evidence presented in this report shows that, following a period of Well-developed national agricultural research systems and adequate levels of investment are important prerequisites for agricultural development, food security, and poverty reduction. In recent years, governments have exhibited renewed interest in supporting agricultural development in SSA. CAADP, the G8 L'Aquila summit, the UN High-Level Task Force on the Global Food Security Crisis, as well as international efforts to re-engage in climate change mitigation and natural resources management all contributed to returning agriculture and agricultural R&D to the political agenda. But this political support must be translated into a set of specific directives by governments, donors, and other R&D stakeholders if the many challenges facing agricultural R&D systems are to be addressed. Countries that have increased their expenditures have directed most of the funds toward salary increases and the rehabilitation of infrastructure and equipment. Nevertheless, these important investments have to be complemented with additional allocations to increase the number, variety, and intensity of actual research activities. National governments should urgently address underinvestment in agricultural R&D. They will need to make more funds available to support research activities if investments are to translate into improved agricultural productivity. Increased government support to agricultural R&D should also include funding to allow universities to establish and maintain basic research programs, which to date have been limited. Moreover, governments, donors, and regional and international organizations must cooperate more closely and increase their commitments to agricultural R&D if SSA countries are to meet CAADP's 6 percent yearly target for AgGDP growth or the poverty and hunger targets of the Millennium Development Goals. Finally, diversification of funding sources is needed, for example, through the sale of goods and services and increased participation in and funding of research by the private sector. As stated, this, in turn, requires that national governments provide a more enabling policy environment.Investment Levels. The time-series data in this report reveal that agricultural R&D funding in SSA has been highly volatile. Many countries continue to be extremely dependent on unstable inflows of donor funding and development bank loans, and in many instances the completion of large donor-funded projects has precipitated severe financial crises, seriously undermining any progress made. In this way, very often the gains achieved through donorfunded projects are quickly eroded in the absence of viable mechanisms to sustain them. Donor funding is typically short term and ad hoc, calling into question the long-term effectiveness and efficiency of this type of funding. In addition, research by nature involves inherent time lags between investment in R&D and the attainment of returns to those investments in the form of tangible benefits (Alston, Pardey, and Piggott  2006); this further highlights the need for long-term, stable funding. Volatility in year-to-year spending levels can be halted only through stable and sustainable levels of government funding. Governments have to clearly identify their long-term national R&D priorities and design relevant, focused, and coherent R&D programs accordingly. Donor funding needs to be better aligned with national priorities, and consistency and complementarities between donor programs need to be ensured. Real progress can be achieved only with sustained, long-term backing from national governments, donors, and regional and international organizations.Addressing Existing and Imminent Challenges in Human Resource Capacity. Growing concern exists regarding the lack of human resource capacity in agricultural R&D to enable satisfactory responses to emerging global challenges. National governments and donor organizations must expand their investments in agricultural higher education to allow universities to increase the number and size of their MSc and PhD programs and to improve the curricula of existing programs. The regional community has an important role to play in this regard, particularly when it comes to small countries with limited or nonexistent MSc or PhD training opportunities. In recent years, various regional capacity-building initiatives have begun, but these will have to be further expanded in order to address some of the capacity challenges evidenced in this report, including aging pools of scientists and increasing shares of junior research staff in a large number of countries. As a result of prolonged recruitment freezes, many countries lack middle-level staff needed both to take on seniority as older scientists retire and to train and mentor the younger researchers coming up behind them. In addition to university training programs, agricultural research agencies will need to establish mentoring programs to facilitate on-the-job training for junior scientists. National governments must also promote (agricultural) science as a valuable career path for young people, which should include strengthening primary-and secondary-level education in the sciences. Moreover, many countries with serious capacity gaps will have to increase the civil servant retirement age or institute flexible working arrangements to ensure that retired researchers can contribute to much-needed training and mentorship initiatives.Cooperation in Agricultural R&D. Because of the high fixed costs inherent in research, small countries generally lack the required critical mass of agricultural R&D capacity and hence face enormous challenges in producing or accessing relevant, highquality research outputs (World Bank 2007a). Very often, the only viable-and efficient-solution is regional collaboration. Through regional initiatives, technological innovation in one country can quickly have an impact in other countries with similar agroclimatic conditions, creating what is known as a leapfrog effect. Creative efforts to build and enhance strong subregional linkages need to be further strengthened in order to maximize these synergistic opportunities. Because many of the regional efforts have a network approach, the CGIAR will continue to act as a critical provider of agricultural technologies in most SSA countries, as well as supporting capacity building efforts.Monitoring the performance, inputs, and outcomes of agricultural S&T systems is fundamental to assessing progress toward CAADP's targets and the strategic recommendations and policy directives espoused in the various influential reports and meetings described above. Up-to-date information is critical to accurate interpretations of the current status and direction of national agricultural research systems in SSA countries. Regular collection of data on agricultural S&T capacity and investments, as undertaken by ASTI, is therefore essential.","tokenCount":"7616","images":["1772539163_1_1.png","1772539163_18_1.png","1772539163_18_2.png","1772539163_18_3.png","1772539163_18_4.png","1772539163_18_5.png","1772539163_18_6.png","1772539163_18_7.png","1772539163_18_8.png","1772539163_18_9.png","1772539163_18_10.png","1772539163_18_11.png","1772539163_18_12.png","1772539163_18_13.png","1772539163_18_14.png","1772539163_18_15.png","1772539163_18_16.png","1772539163_18_17.png","1772539163_18_18.png","1772539163_18_19.png","1772539163_18_20.png","1772539163_18_21.png","1772539163_18_22.png","1772539163_18_23.png","1772539163_18_24.png","1772539163_18_25.png","1772539163_18_26.png","1772539163_19_1.png","1772539163_19_2.png","1772539163_19_3.png","1772539163_33_1.png","1772539163_33_2.png","1772539163_44_1.png"],"tables":["1772539163_1_1.json","1772539163_2_1.json","1772539163_3_1.json","1772539163_4_1.json","1772539163_5_1.json","1772539163_6_1.json","1772539163_7_1.json","1772539163_8_1.json","1772539163_9_1.json","1772539163_10_1.json","1772539163_11_1.json","1772539163_12_1.json","1772539163_13_1.json","1772539163_14_1.json","1772539163_15_1.json","1772539163_16_1.json","1772539163_17_1.json","1772539163_18_1.json","1772539163_19_1.json","1772539163_20_1.json","1772539163_21_1.json","1772539163_22_1.json","1772539163_23_1.json","1772539163_24_1.json","1772539163_25_1.json","1772539163_26_1.json","1772539163_27_1.json","1772539163_28_1.json","1772539163_29_1.json","1772539163_30_1.json","1772539163_31_1.json","1772539163_32_1.json","1772539163_33_1.json","1772539163_34_1.json","1772539163_35_1.json","1772539163_36_1.json","1772539163_37_1.json","1772539163_38_1.json","1772539163_39_1.json","1772539163_40_1.json","1772539163_41_1.json","1772539163_42_1.json","1772539163_43_1.json","1772539163_44_1.json"]}

data/part_2/0299713515.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"3a99ac746cfc53d6ac7455cd4955faed","source":"gardian_index","url":"https://newpathwaysagric.files.wordpress.com/2021/05/new-pathways.pdf","description":"The creation of the WTO in 1994 marked a new departure in multilateral trade relations, for several reasons, among them the successful inclusion of a specific agreement on agriculture, putting an end to the de facto exclusion of the sector from multilateral disciplines which had persisted since the founding of the GATT in 1948. By accepting the Uruguay Round Agreement on Agriculture, Members of the fledgling WTO acknowledged the disarray which had been created by unbridled subsidies and trade restrictions in the sector and set about rectifying this situation. Moreover, in agreeing Article 20 on the continuation of the reform process, Members also implicitly acknowledged that the Agreement on Agriculture (henceforth AoA) was merely a beginning. Explicitly it was agreed that negotiations for continuing the reform process would begin one year before completion of implementation, that is in the year 2000. These negotiations were then made an element of the new round of multilateral trade negotiations launched in 2001 and known as the Doha Development Agenda (DDA).","id":"-426642802"},"keywords":[],"sieverID":"d18a3f8f-257f-481e-a3d1-9cd6a2f73031","pagecount":"39","content":"All the members of the \"pathways\" group, contributed to the work. All subscribe to the objective to re-invigorate and re-energise discussion on agricultural trade at and beyond the WTO. The proposals and suggestions made are broadly supported by each contributor, but this does not mean that there is full consensus on every aspect and detail of the proposal.This paper has been prepared by an informal network of individuals including former senior officials in governments and international organisations, academics, trade negotiators and analysts from think tanks and institutes, all of whom have been, and continue to be, actively engaged in reflections on agricultural policy and agricultural trade. Each participant has contributed under his or her own responsibility and authority and no support has been sought or received from any government or institution. The full list of the members of this pathways group with a brief description of former and current affiliations is included in Appendix 3.The creation of the WTO in 1994 marked a new departure in multilateral trade relations, for several reasons, among them the successful inclusion of a specific agreement on agriculture, putting an end to the de facto exclusion of the sector from multilateral disciplines which had persisted since the founding of the GATT in 1948. By accepting the Uruguay Round Agreement on Agriculture, Members of the fledgling WTO acknowledged the disarray which had been created by unbridled subsidies and trade restrictions in the sector and set about rectifying this situation. Moreover, in agreeing Article 20 on the continuation of the reform process, Members also implicitly acknowledged that the Agreement on Agriculture (henceforth AoA) was merely a beginning. Explicitly it was agreed that negotiations for continuing the reform process would begin one year before completion of implementation, that is in the year 2000. These negotiations were then made an element of the new round of multilateral trade negotiations launched in 2001 and known as the Doha Development Agenda (DDA).It is not the intention here to provide an exhaustive account of what has since transpired. Suffice it to say that it has so far not been possible to deliver the next tranche of a comprehensive reform process in agriculture. This is notwithstanding some notable achievements such as the Trade Facilitation Agreement negotiated in 2013, and the highly significant decision made at the 10 th Ministerial Conference in Nairobi in December 2015 to completely eliminate export subsidies in agriculture. That same Ministerial Conference acknowledged that WTO members have different views on how to address the negotiations, some reaffirming full commitment to conclude the DDA while others believe new approaches are necessary. However, all members made it clear that they were strongly committed to advance negotiations on the remaining Doha issues including all three pillars in agriculture. (WTO 2015) Despite this commitment, agriculture has continued to be a difficult and sensitive topic, and while the failure to reach agreement on agricultural matters cannot be held entirely responsible for the uncertain situation concerning the DDA, nor for the serious systemic problems currently facing the WTO, there is no doubt that the intractability of the agricultural negotiations has contributed to the paralysis. Arguably, agriculture also serves to illustrate the broader, more systemic difficulties experienced by the WTO, in trying to deliver a significant multilateral agreement while preserving founding principles such as consensus and nondiscrimination, as WTO membership expanded and geopolitical as well as economic balances in the world were undergoing radical change.It is against this background that our informal network of experts has been brought together with a view to exploring pathways for progress to be made in agriculture. This paper attempts to bring together in a single proposal a host of ideas, some new and some less so, developed by contributors to the network and by others. At the centre of the paper are suggestions specifically addressing conceptual and technical aspects of the agricultural negotiations, ranging from \"fixes\" to the original agreement, to more radical, conceptually different approaches. Alongside the explicit ambition to offer something that could assist governments and the WTO to see a way out of the deadlock in agriculture, is a broader ambition. That broader ambition seeks to position this effort in a world where geo-political factors have shifted dramatically, where challenges such as resource depletion and climate change have intensified, and where the multilateral system itself is under existential threat.The late 20 th and early 21 st centuries have been marked by some dramatic shifts in economic weight as reflected in GDP and trade statistics. According to the IMF, emerging and developing economies now account for more than 58% of global GDP measured in purchasing power parity terms. By this measure the United States ranks second worldwide at almost sixteen percent, after China with nineteen and the European Union ranks third at almost fifteen (IMF 2021, DataMapper). According to the latest WTO trade data, developing economies now account for 43% of global merchandise trade, and 21% of world merchandise trade occurs among developing countries. China ranks first in the world in terms of merchandise exports with a world share of almost 13%. Economic integration has not been limited to merchandise trade, developing countries providing 25% of world services exports and accounting for 34% of imports (WTO 2019a,b, UNCTAD 2020).These trends are mirrored in global developments of agricultural and food production, consumption and trade. The past three decades have seen East Asia (i.e. mainly China) increase its share of global agricultural production from 25% in 1993-95 to 33% in 2013/15, that of South and Central America from 9% to 11%, while Europe's share has fallen from 26% to 14% and that of North America from 13% to 10% (OECD 2019). Moreover, projected growth in production will continue to be stronger in regions outside the traditional agricultural powerhouses of the US and Europe (OECD/FAO 2019). Patterns in agricultural trade have also evolved significantly, with agricultural trade tripling between 2000 and 2014, while the share of developing countries has risen sharply, from 26 to 39% of imports since the midnineties, and from 31 to 40% of exports over the same period (Glauber 2019).Significant policy shifts have also been occurring. OECD's annual monitoring and evaluation exercise provides an up-to date estimate of the levels of support and protection of the sector in its own members and in a group of emerging economies, which together account for close to three-quarters of global value added in agriculture (OECD 2019b). The composition and amounts of different kinds of domestic support as defined at the WTO have also changed significantly (Brink and Orden 2020). The data reveal persistent, vast disparities among countries in the level of their interventions in the sector, ranging according to the OECD's PSE measure, expressing government support as a share of gross farm receipts, from negative support to a maximum around 60%. Trends also diverge significantly, with the OECD group of countries mostly following a declining or stable trend , with average PSE levels currently (that is the period 2017-19) at 17.6%, while a group of emerging and developing countries who began the 21 st century with very low levels (of positive support) now provides support on average in the region of 13%. Another group of countries, among them India and Argentina, deploy trade and other policy measures that implicitly or explicitly tax their agricultural sectors.Clearly, as weight has shifted towards the emerging and developing countries in terms of production, consumption, trade and farm support, they have become important actors in the policy sphere, deploying the full range of domestic and trade interventions, and exerting considerable influence on the functioning of international markets for agricultural and food products. Agriculture in many of the countries in this group is dominated by large numbers of smallholders with relatively low productivity, as illustrated by the disparity between agriculture's share of employment and GDP. Many of these smallholders are poor, many are net buyers of foodstuffs from the markets, and have limited or no capacity to absorb shocks. Protection leading to higher prices damages net buyers, while low prices harm small farmers and may push them out of business. Governments attempt to reconcile these divergent impacts. This and other features of emerging and developing country agriculture complicate the political economy of agricultural policy making and significantly impact their capacity to engage in multilateral negotiations if these are seen as limiting their policy choices in the domestic arena. On the other hand, all countries have an interest in participating in negotiations to reduce distortions and the resulting damage to their interests caused by the behaviour of third countries. Our proposals have been attentive to these challenges and try to encourage policy adaptations that facilitate important social priorities in minimally distorting ways.The Marrakesh Agreement creating the WTO clearly acknowledged that natural resources, environment and sustainability are essential elements to be taken into account in the pursuit of growth and development through improved trade (WTO 1995). Several decades later, these issues have taken on even greater prominence and urgency. The scientific evidence concerning humankind's contribution to climate change is incontrovertible as are the conclusions that the world is not on track to prevent it from becoming a serious threat to the livelihoods and lives of a significant share of humanity. Agriculture and food production are projected to be one of the sectors most severely affected (IPCC 2018). Concerns are also growing about biodiversity loss and the threat it poses, inter alia, to the resilience of the food system. Agriculture is contributing to the problem, but it is also potentially an important part of the solution. Similarly, there are concerns worldwide about depletion of natural resources such as soil and water, where agriculture is strongly implicated. Agricultural water risk hot spots have been identified where a continuing failure to address over-exploitation could have implications for global food security (OECD 2017). Governments around the world need, as a matter of urgency, to move away from policies that encourage intensification of production of specific crops and the use of methods and inputs that are contributing to climate change and are harmful to sustainability. A renewed effort to address agricultural market distortions could provide an important impetus here, as well as nudging governments to redeploy scarce fiscal resources to more effective, targeted measures. While significant improvements have been made in food security as measured by the percentage of the world's population experiencing hunger, actual numbers of undernourished have increased recently, giving rise to worries that the Sustainable Development Goal to eliminate hunger by 2030 may not be reached. In recent years nutrition dimensions of food security have come into stronger focus as micro-nutrient deficiencies, and rates of overweight and obesity have increased to alarming proportions Trade has an important role to play in this respect both as a vector for growth and poverty alleviation, and as one element in a comprehensive approach to food security encompassing production, imports, exports and stockholding. Consideration also needs to be given as to how the multilateral trading system can better support, not just greater access to needed volumes of foodstuffs, but also better nutritional outcomes.Volatility, broadly defined as large and unanticipated fluctuations in agricultural and food prices, has long been a concern for many governments. Behind issues of price volatility often lie concerns about price levels, specifically the impact of high prices on the food security of vulnerable households and countries, and the impact of low prices on vulnerable producers. Agricultural markets are more subject to volatility than other markets due to the dependence of the sector on weather, climate and other natural phenomena, to the low price elasticity of demand and to the lagged nature of supply responses. Developing countries, in particular, worry about the transmission of price volatility from international markets and the impact on their poorer consumers and producers, although volatility stemming from domestic market failures or supply disruptions is likely to be more problematical (G20, 2011).Following the international price peaks experienced in 2007-8 and again in 2011, many analysts concluded that the world was entering a period of increased volatility. While prices have been relatively stable since, and may have returned to the previous pattern of slow decline in real terms, greater volatility cannot be ruled out in the future (OECD/FAO 2019). Climate change will force changes in production patterns which themselves may be disruptive, while the predicted increase in the incidence of extreme weather events may significantly impact production. Specific countries or regions may face supply shortfalls. In an increasingly globalised world, transmission of contagious animal diseases could also cause price volatility, locally, regionally or even globally.An open, robust and predictable global trading system for agricultural and food products will be key to mitigating the impacts of increased volatility, alongside appropriate risk management instruments for producers and safety net programmes for vulnerable consumers. On the contrary, countries wishing to maintain high barriers to trade or to rely solely on their own production will be at risk of even greater volatility. The larger the number of countries participating in international trade, governed by agreed and predictable behaviours, the greater the likelihood that an episode of volatility can be contained, and that supplies can continue to flow to regions experiencing a supply shortfall. A renewed effort to achieve a more transparent, robust and well-functioning multilateral trading system for food and agriculture will be needed to ensure that vulnerable producers and consumers are able to cope with the expected disruptions.The coronavirus outbreak of early 2020 spread rapidly and quickly acquired pandemic proportions. Some disruptions to food production and distribution logistics initially gave cause for concern, but the system proved to be resilient, adjusting relatively quickly to the evolving sanitary situation and the associated changes in consumption behaviour. Governments stepped in to support the supply side and also quickly deployed help to vulnerable consumers. Some governments implemented trade liberalising measures, other took trade restrictive steps, but many of the latter were quickly removed. Overall the global food system continued to function well. Global agricultural commodity and food prices have increased since mid 2020 due to a combination of weather and market factors, and while overall, market fundamentals remain comfortable, the situation requires careful monitoring. Panic buying and ill-advised trade restrictions implemented by governments have not been on the scale seen during the food price crisis when they actually caused or exacerbated the problems in some sectors. Nevertheless many of the most vulnerable households have experienced an increase in food insecurity although the channel through which this has occurred has mainly been the loss of employment and income, rather than disruptions to the food system itself. Overall, the resilience shown by the sector demonstrates the importance of transparency, (e.g. through the Agricultural Market Information System, AMIS), and of keeping markets open, the effectiveness of reforms that have already occurred and the importance of consolidating and continuing the process (FAO/AMIS 2021, FAO 2020, OECD 2020a).The world's population is predicted to continue to grow, reaching according to various estimates between 9.5 and 10 billion people by 2050, up from the current 7.3 billion. Analysts concur that with continuing growth and related improvements in incomes, demand for food will grow more than commensurately, particularly demand for animal protein. The most often quoted projection is for an increase in food demand of 50% (from a 2012 base) by 2050 (FAO 2017). The literature also concurs in saying that much of that increased demand will have to be met through trade. This is because population growth will outstrip production capacity in some regions, because of societal demand for a more nutritious, varied and less seasonal diet, and because of the changes that resource depletion and climate change will provoke in the patterns of production around the world (Gurria and Graziano da Silva 2018;Diaz-Bonilla and Hepburn 2016). To ensure food security worldwide, it will be crucial therefore that the international community delivers a more open trading system in which both importing and exporting countries can place their trust.The importance of trade has been explicitly acknowledged in the formulation of the Sustainable Development Goals adopted by the United Nations in September 2015 which include as Goal 2 \"End hunger, achieve food security and improved nutrition and promote sustainable agriculture\". In support of Goal 2, Target 2.B states the need to \"correct and prevent trade restrictions and distortions in world agricultural markets, including through the parallel elimination of all forms of agricultural export subsidies and all export measures with equivalent effect, in accordance with the mandate of the Doha Development Round\", while Target 2.C calls on governments to \"adopt measures to ensure the proper functioning of food commodity markets and their derivatives and facilitate timely access to market information, including on food reserves, in order to help limit extreme food price volatility\". These globally agreed declarations testify to the clarity and unanimity of UN members concerning the role of trade and of well-functioning international markets in the achievement of food security and sustainable agriculture and food systems. The vital importance of trade for the world's agricultural and food economy, and for global food security, was again emphasized by Ministers of Agriculture from 71 nations meeting in January 2020 at the 12th Berlin Agriculture Ministers' Conference on the occasion of the Global Forum for Food and Agriculture (GFFA).Finally, most recently, the coronavirus pandemic has brought sharply into focus the importance of transparent and well functioning global markets for agriculture and food products. In this context, the global community, armed with the lessons learned from the 2007/08 food price crisis, predominantly acted to keep markets open while providing direct assistance to affected producers and consumers, so that major disruptions were largely avoided (see Appendix 1).Research continues to find that there are significant economic benefits from deepening market integration in terms of growth and development. Benefits accrue first and foremost in the countries undertaking reforms and tend to be commensurate with the starting levels of distorting interventions. Benefits accrue inter alia because of the opportunities provided to access new technologies and innovations. Benefits are greater when market integration is both broad and deep, covering the entire economy and the largest possible number of countries. Gains arise in large part from increased participation in global (but also domestic) value chains, because as import barriers come down economies can access cheaper, high quality inputs (Greenville et al, 2019, OECD 2016). The growing integration among developing countries means that much of the benefits that potentially accrue to them arise from reform efforts in other developing countries. There are therefore compelling reasons to continue to seek progress in market opening, and to prefer multilateral solutions to bilateral or preferential agreements.The AoA was based on clear economic concepts relating to the harmfulness of policy interventions in terms of production and trade distortions. It also benefitted from a body of empirical evidence that enabled the negotiating process to be reasonably well informed. With respect to the domestic support pillar, measurements of domestic support deemed to be distorting of production and trade became the benchmark from which reductions were required. Market price support was captured by a price gap methodology which compared applied administered prices to external reference prices. The latter were fixed as the average of the 1986-88 period and have remained unchanged since. As a result, the data on which the AoA and the commitments accepted under it were based are now totally outdated. Moreover, many countries have fallen behind in their notifications, and some of them are now seriously out of date.An honest and open dialogue towards progress would be facilitated by a clear reiteration that the fundamental objective is to reduce obstacles to and distortions of trade in a multilateral setting. In other words, it should be clear that there is no intent to seek to provide countries with greater scope for border protection or distorting domestic measures than is currently permitted.Another prerequisite for open and honest dialague is an up-to-date and accurate picture of the policies currently pursued by WTO members and their possible impacts on other members. Data gathering and analysis undertaken by the OECD in the 1980s provided valuable background to the Uruguy Round negotiations. Similar information, but for a much larger group of countries, is available from work currently undertaken regularly by OECD, FAO (MAFAP2 ) and in the Consortium for Measuring the Policy Environment for Agriculture hosted by IFPRI to which the IDB and the World Bank also contribute (FAO/MAFAP, IFPRI 2020, OECD 2020b). While not all WTO members are covered, all the systemically important countries are, as well as a large number of others.Moving the basis of the discussions from the parameters laid out in the AoA to a contemporary picture of the world presents some risk. It is essential that technical or negotiated solutions be found to ensure that what has already been achieved in the course of the Uruguay Round AoA is not undermined. In reality, many countries are well below their permitted levels of certain types of domestic support, or apply tariffs well below bound levels. For these countries new disciplines could largely involve \"eliminating water\". Many countries have also already made additional concessions in regional or preferential agreements and have made the political and economic case for them to their populations. For both of these reasons it should be possible to find a way forward in the multilateral arena of the WTO.Trade in general, (and agricultural and food trade specifically) is not an end in itself, neither is it the panacea for all the problems facing countries at different levels of development, nor is it the cause of all ills. What therefore is the appropriate scope of multilateral deliberations on agricultural trade (and by extension of this paper)? The intent is to encroach to the minimum extent possible on domestic prerogatives while also acknowledging that there are many important, broad, linked issues where the multilateral dimension is important and where the international community needs to forge agreements that are consistent and mutually supportive or reinforcing.First and foremost the intent is to support the World Trade Organization in the fulfillment of its fundamental objective as described in the preamble to the Marrakesh Agreement as \"raising standards of living, ensuring full employment and a large and steadily growing volume of real income and effective demand, and expanding the production of and trade in goods and services, while allowing for the optimal use of the world's resources in accordance with the objective of sustainable development\" including by \"entering into reciprocal and mutually advantageous arrangements directed to the substantial reduction of tariffs and other barriers to trade and to the elimination of discriminatory treatment in international trade relations\" (WTO 1995).Generating sustainable growth and development is of course a key priority for governments globally. But it is not their only preoccupation. Trade is a vehicle whereby growth can be generated and societal problems can be solved, but it may also inadvertently aggravate some problems. The societal issues which may be impacted by trade include food security and nutrition, poverty alleviation, employment, biosecurity risks, climate change mitigation, and environmental and natural resource pressures. These issues also potentially have important cross border and multilateral dimensions. Any proposals concerning agricultural and food trade therefore need to also take these broader issues into account. In this context, the following may be helpful in structuring reflections and delimiting the appropriate scope of a multilateral trade agreement on agriculture under the auspices of the WTO.First, some measures, generally considered as unambiguously production and trade distorting, also generate negative externalities or public \"bads\", some with a global dimension. Reforming such measures would constitute \"win-wins\" and should be prioritized as such. Disciplining policies that simultaneously distort trade and aggravate environmental and resource problems or contribute to climate change therefore come clearly under the remit of multilateral agricultural trade negotiations and should be given high priority.Second, as was already the intention in the AoA, international agreements can be designed to nudge governments towards beneficial interventions, for example to overcome under-provision of certain public goods, and away from interventions that are harmful to growth and development. Trade-offs can be made explicit and the specific characteristics of effective policy measures that are minimally or non-production and trade distorting can be defined. In this light a thorough review of the scope of the different elements (in particular the domestic support categories) of the AoA and of the specific provisions governing them is fully warranted. The purpose is to ensure that general and specific provisions are accommodating enough to enable a wide range of non-trade objectives to be met while minimizing production and trade distorting effects.Research shows that there is a very strong concordance between policy effectiveness in many areas and diminishing levels of trade distortion. This is particularly the case in relation to climate, environmental services and risk management. International agreements can be designed to encourage governments to find the optimal trade-offs, prohibiting or otherwise disciplining measures which both distort trade and exacerbate environmental or climate problems, while defining the broad contours of exemptible policies that are less or nondistorting but also more effective and efficient in relation to other objectives.Is there a role for trade policy itself in governments' efforts to meet specific objectives such as addressing climate change which has a clear global dimension, or in preserving or improving the nutrition status of populations? Analysts point to enormous technical and measurement difficulties in designing such measures, and in forging international agreement concerning them, particularly where the agriculture sector is involved. The risks of disguised protectionism, arbitrariness or double taxation are substantial. Border tax adjustments or equivalent instruments would need to comply with the WTO's founding principle of nondiscrimination. The Paris Agreement on climate change enshrines a bottom up approach allowing countries freedom to decide how they will achieve their targets, which seems difficult to reconcile with border measures designed to prevent emissions leakage from specific sectors or products. In addition, there are important unresolved issues around which international agreement should take precedence if contradictions arise. Similarly, even if part of a broad policy effort including nutrition education and social safety nets, designing trade measures to specifically support better nutrition outcomes is challenging.The issues raised above go way beyond the agriculture sector and way beyond the remit of the WTO. Considerable political will is needed across a broad spectrum of countries and multilateral institutions, including the WTO, to find innovative and consensual solutions. For these reasons and notwithstanding the huge importance of the underlying issues, this paper stops short of considering the use of trade policy instruments in pursuit of global or domestic objectives that are not currently within the remit of the WTO or in the frame of the AoA. The primary intention is to support multilateral efforts to contribute to sustainable growth and development through trade, specifically through agriculture and food trade. A fundamental underlying hypothesis is that there is a very large overlap between disciplines which can contribute to that, while also being beneficial in pursuit of other objectives. This is particularly significant in relation to environment, resources and climate, all of which suffer from the kinds of policy interventions that lead to more intensive use of natural resources and of fossil fuel-based inputs in agriculture. Therefore, particular attention is paid to a) aiming for the strongest discipline on interventions that are simultaneously harmful to trade and to other societal objectives, b) identifying policies where there are beneficial overlaps between trade and other outcomes and c) nudging governments towards those more effective and equitable interventions by continuing to define a comprehensive list of exemptible measures in the form of revisions to Annex 2 (Green Box) of the AoA.At the time of completion of the AoA tensions in agricultural trade were an issue mainly for the OECD countries, although many developing countries suffered from the price depressing effects of policies in OECD countries and from the increased market volatility they were causing. The solutions adopted, in many respects, reflected that situation. The starting point was empirically established levels of domestic support, import barriers and export subsidies. The effect was to impose discipline and reduction commitments, but at the same time, to legalise (or legitimise) the residual, sometimes very high levels of support and protection.The failure to continue the reform process in any comprehensive way in the intervening years has led to the emergence of a keen sense of grievance on the part of some developing countries, who, rightly or wrongly, feel that their policy choices are unfairly constrained by the terms of the AoA, while the original protagonists are perceived as having wide latitude in their policy choices, including in relation to decoupled income supports and other direct payments which can be excluded from discipline under the terms of Annex 2 of the AoA. Issues around cotton subsidies, public stockholding for food security purposes, and the adequacy or not of de minimis provisions have also attracted much attention in this respect. Any proposal which fails to address these issues will fall short of expectations.It follows that any new agreement will need to move towards more equal treatment in terms of policy flexibility across the whole spectrum of levels of development. This implies efforts by WTO members commensurate with their starting levels of support and protection and an agreed long-term objective to move towards convergence. In parallel, WTO members need reassurance that they will be able to implement policies appropriate to their actual level of development, associated developmental and social needs, and capacities. Specific solutions will be needed for some of the most contentious and sensitive issues. The updating exercise which is proposed to establish the current situation of countries will be an important step in the process, shedding light on the extent to which the AoA provisions may or may not have effectively constrained countries in their policy choices to date.The AoA was innovative in many respects, but it was also onerous in terms of the technical and data capacity required to complete the schedules and fulfil the ongoing notification requirements. To the extent possible it is desirable to simplify the disciplines which in turn would reduce the notification burden. At the same time a significant technical effort is required to clarify terms used, close loopholes and ensure consistency across the different pillars.While the period for preparing and submitting schedules under the AoA was relatively short, a strong case can be made for proceeding slowly once agreement has been reached on the broad terms and modalities of a new agreement. Countries will need time to build technical capacity and gather and verify the data which will subsequently be used to determine compliance with commitments made. The capacity of international organisations such as the OECD, the FAO and IFPRI could be harnessed to assist in the work of clarifying, simplifying and ensuring consistency. Different modalities could be envisaged including the creation of a consultative, technical group under the auspices of the WTO's Committee on Agriculture. A deliberative function could be created within the Committee on Agriculture itself to discuss and clarify technical and interpretation issues as they arise. In the context of wider WTO reforms a more proactive role could also be envisaged for the WTO Secretariat in the verification of notifications or in the provision of updated data where countries themselves fail to respect their notification commitments.Behind the innovations contained in the AoA (three pillars, empirical measures of interventions as the starting point) was the idea that protection at the border, distorting domestic support and direct or indirect export subsidies constitute an interlinked system. Disciplines needed to be coherent and mutually reinforcing. The ability of governments to intervene in domestic markets, for example, to raise prices to farmers, depends crucially on their ability to simultaneously control imports and, in many cases, to dispose of surplus production through export competition measures. It was therefore well understood that each element would need to be tightened. This remains the case today, although the Nairobi agreement to eliminate export subsidies has to a very large degree closed off one of the \"valves\" that governments previously had recourse to in the event that domestic production stimuli provided behind high tariff walls overshot the needs of the domestic market.Nevertheless, any revised or new agreement will still need to take these interconnections into account. This means, among others, that continuing attention will have to be paid to more indirect forms of distorting export competition such as state trading, sales from government stocks destined for export, and any other measures that could constitute export subsidisation. Similarly, in any simplification of domestic support disciplines the strength of the remaining market access barriers will be a crucial determinant of the potential to ease and simplify domestic support disciplines. This cannot be an exact science but will be reflected qualitatively in the proposals to follow,The following sections are organized largely along the same lines as the ongoing discussions at the WTO, covering seven themes. These reflect the original three pillars of the AOA, additional topics or sub-topics which have come to prominence in the intervening years as experience accumulated in implementation of the AoA, and new topics which emerged in light of market and policy developments. Under several of the themes, options are presented, rather than one, single proposal. These options involve increasing levels of ambition and commensurately require increasing willingness on the part of governments to contribute to the process. Even the least demanding of the options presented requires all countries, without exception, to move away from long held positions, to be open both to conceptual changes and technical fixes, and to be prepared to engage in a holistic process that will, by definition, require trade-offs and compromises across the full range of issues on the table . \n\nConsiderable thought has been given as to how to incorporate in the proposals the essential characteristics of any potential new deal, as outlined in the previous section. The starting point is a current picture of the world. The proposals if implemented would constitute progress in terms of removing obstacles to trade, they allow sufficient scope for governments to pursue a wide range of economic, social, environmental and climate objectives, they address grievances and imbalances arising from the AoA, they require contributions from governments commensurate with their actual situation in development, market and policy terms, and they are mutually consistent and reinforcing across the three pillars. A simultaneous implementation of the most ambitious options would lead to a significant narrowing of the difference in treatment between agriculture and other goods sectors at the WTO, opening up the possibility that in the long term, trade in agriculture need no longer be disciplined by a specific, complex agreement and instead could be subject to general market access provisions and the general provisions of a revised SCM Agreement.In the proposals presented here, only the broad framework for an agreement is specified, specific modalities -that is the scale and speed of implementation of required concessionsbeing a matter for negotiation among countries.The proposals presented here cover tariff simplification, tariff reductions, the AoA's special safeguard (SSG), a new special safeguard measure (SSM) and tariff rate quotas (TRQs). Exemptions and carve-outs are kept to a minimum in order to ensure that any agreement actually results in improved market access. The options presented vary from technical amendments to existing disciplines to approaches that are conceptually different.Tariff simplification WTO members agree not to create additional compound or other complex tariffs, nor to transform existing tariffs into more complex forms.Tariffs are reduced over an agreed period of years in a tiered fashion, with required reductions higher and reduction periods longer the higher the tier of the base tariff. Base tariffs from which reductions are to be made are currently bound tariffs.The number of tariff lines eligible for the SSG is reduced to zero over a six year period.Developing countries 3 may invoke a price-based SSM. The SSM can be invoked if the cif price declines to less than an agreed per cent of its recent average. The additional duty must not exceed an agreed per cent of the price drop. The additional duty cannot be charged for more than an agreed, limited number of months.Tariff quota volumes remain unchanged. TRQ fill rates are to be notified, along with abovequota imports of the same product.In addition to the ban on creation of new non-ad-valorem tariffs, all bound and applied non-ad valorem tariffs are converted to ad valorem tariffs, based on AVEs calculated for the most recent available data base (either WTO data or any other agreed data base).The Swiss Formula 4 is used for calculating final bound tariffs for all tariff lines, starting from base tariffs. Higher bound tariffs are thus reduced by more than lower tariffs. Tariffs are 3 By whatever definition is applied across the WTO 4 Under the Swiss Formula, used in the Tokyo Round for reducing non-agricultural tariffs, the final tariff TF (in per cent) is calculated from the base tariff TB (in per cent) and an agreed parameter A as follows: TF=A*TB/(A+TB). Under this formula, tariffs with TB=A are reduced by one half, while tariffs above (below) A are reduced by more (less) than one half.reduced to that final level over a period of x years. For tariffs that have to be reduced by more than one half, the reduction period can be longer.The number of tariff lines eligible for the SSG is reduced to zero over a three year period.As in Option 1, but with the additional provision that the additional duty must not raise the total duty charged to more than the pre-agreement bound tariff.As in Option 1As in Option 2 with the addition of the following provisions covering the Special Safeguard Mechanism:For a transition period of five years, countries may invoke a price-based SSM for a product whose final bound tariff level is below y percent. The SSM can be invoked if the cif price declines to less than an agreed per cent of its recent average. The additional duty must not exceed an agreed per cent of the price drop or an agreed per cent of the difference between the base tariff and the final bound tariff, whatever is lowest. The additional duty cannot be charged for more than an agreed, limited number of months.LDCs may implement these market access disciplines on a best endeavours basis. Other developing countries may be accorded longer time periods to implement the disciplines resulting under the various provisions and formulae. The proposed SSM is available only to developing countries. (The determination of the basis on which countries are defined as developing is a systemic issue going beyond any agriculture agreement and therefore beyond the scope of this paper).This section deals with domestic support as covered by Article 6 of the AoA. Proposals covering Annex 2 (Green Box) measures are covered in the next section. Three options are proposed in respect of Article 6, (\"domestic support measures in favour of agricultural producers\"), which includes AMS support, blue box support and Article 6.2 support) and are presented in a graduated sequence. The first option is based on the framework adopted in the Agreement on Agriculture (AoA) concluded in the Uruguay Round, the two subsequent options are derived from that first option but propose an increasingly simplified approach.Where possible generic provisions and conditions are proposed rather than attempting to define specific parameters and design features of policies. Under all of these three options the possibility to exempt measures considered non or only minimally production and trade distorting under Annex 2 (Green Box) is retained.This option retains major elements of the conceptual framework developed for the AoA, updates all parameters, clarifies technical and legal issues that have arisen to date and closes loopholes. It proposes that new limits be designed from a revised base.A prerequisite for Option 1 (and all others) is for all domestic support notifications (as specified in the Committee on Agriculture's document G/AG/2 of 19955 ) to be submitted for all years including the notification year (calendar, marketing, crop year etc.) that ended in the calendar year preceding the agreement. Moreover, all notifications are also to be verified in the Committee on Agriculture.The second step is to replicate the notifications for the notification years ending in three recent years to be agreed (e.g. 2018 to 2020), using parameters reflecting fully the new rules defined below. This means, among others, that members operating market price support programmes involving an applied administered price will use updated external reference prices.The three-year average of each year's data as required under the new rules then forms the basis for the new commitments (as did the \"offers\" in the verification part of the process of establishing Schedules in the UR). This means that future commitments on domestic support start from a 'clean sheet', much like was the case in the UR.The AoA Total AMS (support subject to discipline and reduction commitments) is redefined so as not to exempt previous Blue Box payments (Article 6.5 in AoA) and the input subsidies component of the Development Box (Article 6.2 of the AoA). The aggregate of this redefined Total AMS and all de minimis AMSs would be denoted \"Domestic Support\" (DS) and expressed as a % of the member's value of total agricultural production, yielding the \"%DS\".Values of total agricultural production used in these calculations are to be included in notifications. Development Box (i.e., Article 6.2) support other than input subsidies would continue to be exempted for developing countries.Instead of the Bound Total AMS under the AoA, new disciplines would be agreed requiring tiered reductions in the %DS, progressing over an agreed number of years, such that a significant degree of convergence is achieved across the WTO membership. Tiered refers to identifying groups of members whose initial %DS are within a particular range, or tier, and applying the same relative reduction for all members in a given tier, with high-range tiers requiring larger reduction than lower tiers. Alternatively a Swiss formula approach, as used for tariff reductions in the Tokyo Round, could be agreed (see the proposal on Market Access).A benchmark level of %DS would be agreed (akin to the present de minimis allowance). Members already below this level would agree not to exceed it. As a confidence building measure, developed countries in this group (i.e. with a %DS below the benchmark on average in three recent years to be agreed) would be asked to observe a standstill, that is not to take measures to increase their applied %DS, or not to increase it by more than a defined margin.The exact level of the benchmark can only be determined when the full up-to-date information on countries' current levels is available, The value it would take should be ambitious and incentivise some further reform, particularly for members providing the highest levels of trade distorting support.DS would include both product-specific and non-product specific support. To avoid unbalanced profiles of support for individual products, an anti-concentration provision would be added, based on measurement of product-specific %DS. The provision could take the form of product-specific limits expressed as a share of value of production of that product, steeper cuts for more highly supported products, or limits on the share of the permitted %DS that could be granted to any one product. In order to make this provision effective, agreement needs to be reached on how 'products' (or product groups) are precisely defined.Under this option, external reference prices would be agreed to be either a) a moving average, or b) a moving Olympic average, of the most recent five years. To make sure that price data are actually available, a one-year lag could be introduced. This means that in any year t, for which the %DS commitment has to be honoured, the external reference prices on which calculation of the current (i.e., applied) %DS is based are calculated from actual prices prevailing in years t-2 to t-6. In other words, the external reference prices to be considered in setting policies for year t are already known when policy decisions are made. The external reference prices used in calculating the %DS, along with data sources used, are to be included in notifications.Loopholes, from an economic perspective, which emerged in implementation of the AoA need to be closed. A new agreement would clarify that market price support (MPS) calculations are to be undertaken for all products that are subject to an administered or procurement price underpinned by government purchases. Where a government enforces price controls by requiring private agents to buy at government-determined prices, the respective controlled prices are used as administered prices. For products with administered or procurement prices, measured price gaps are to be applied to the total production and not procured or pre-announced quantities only. Non-exempt direct payments based on a price gap between an administered price and a reference price are accounted for using budgetary outlays.An additional important clarification is that if the administered price is set below the comparable international reference price, MPS is not calculated, i.e. it is set to zero.Any other technical aspects which have created loopholes or errors from an economic or practical perspective in the application of the AoA would need to be clarified in defining the new rules.It is proposed to strengthen the role of the WTO Committee on Agriculture in reviewing members' progress in the implementation of commitments by laying down more formal procedures for discussing and clarifying issues of interpretation along the model of specific trade concerns in the SPS and TBT Committees.This variant is derived from Option 1 by proposing a much simplified and stylized approach to capturing price support, no longer needing international reference prices. Price support and budgetary outlays would be captured in two different indicators which are not comparable and therefore not additive, though each indicator would be comparable across members.A new indicator of price support is proposed to capture the support associated with administered prices underpinned by government purchases. It consists of multiplying the amount actually purchased by governments (or agencies acting on their behalf, or by private agents buying at government-determined prices) by the relevant administered or procurement price (rather than by the difference between the administerd price and a reference price), summing across all commodities covered by such schemes. This new variant of measuring price support could be called \"Procurement Expenditure\" or PE. The indicator would be kept current, with both administered prices and value of production updated as and when the data are available. While this indicator is, strictly speaking, less economically meaningful than those proposed under Option 1, this approach is motivated by an effort to avoid, in most cases, the difficulties of defining appropriate reference prices, and by the argument advanced by governments, with a certain justification, that they cannot enter into legally binding commitments under which compliance depends on a parameter (fluctuating world market prices) they cannot control.Limits would then apply to the percentage share of PE in the total value of agricultural production covered by such measures, \"%PE\". A new de minimis percentage for %PE would be defined. Progressive, tiered cuts would be required by members above this de minimis percentage. Developed countries with a base period %PE below the benchmark would be asked to observe a standstill, that is not to take measures to increase their applied %PE, or not to increase it by more than a defined margin.As in Option 1 no accounting would be necessary if the administered price is below the pertinent international reference price although governments using this provision would be required to provide the data supporting its use.Alongside this new indicator of price support, a separate indicator of trade distorting budgetary outlays would be established, including all non-price support covered above in Option 1: the AoA AMS support (net of market price support) including Blue Box payments and the input subsidies component of Article 6.2. The total of these budgetary outlays could be dubbed \" Support through Budgets\" (SB). Expressed as a % of value of total agricultural production it yields the \"%SB\". New disciplines are agreed with progressive tiered reductions in the %SB, such that a significant degree of convergence is achieved across the WTO membership. Members below a de minimis or benchmark level (to be defined) would not be required to make any reductions. Development Box (i.e., Article 6.2) support other than input subsidies would continue to be exempted for developing countries. Developed countries with a base period %SB below the benchmark would be asked to observe a standstill, that is not to take measures to increase their applied %SB, or not to increase it by more than a defined margin.As under Option 1, anti-concentration provisions would be added. Under Option 2 they would apply separately to the product-specific %PE and the product-specific %SBThis option is a further step towards simplification and is proposed only if significant progress has been made in the market access and export competition pillars. It would entail giving up on AoA-type disciplines regarding administered prices and thus not requiring the measurement of MPS. Limits would apply only to budgetary outlays. Stringent rules regarding the types of measures considered as generating harmful trade effects would underpin this option. Discussions are ongoing at the WTO about how to better capture and discipline the full range of direct and indirect subsidies currently affecting markets for steel, aluminium, semi-conductors and other products (OECD 2019c(OECD , 2019d)). In this context, a revised ASCM also involving quantification of subsidies and a system of permitted \"green light\" measures would allow for significant convergence between subsidy rules applying generally and those applying only to agriculture.Under this option, a new indicator of budgetary expenditure would be calculated, consisting of two components. Its first component is the SB as suggested under Option 2. Its second component is budgetary outlays involved in government procurement (or subsidized private intervention buying) to underpin administered prices (this latter type of budgetary outlays is not included in the AMS under the AoA). This second component would, for example, include budgetary outlays which arise if the price at which the government (or an agent) buys the produce from agricultural producers is above the price at which the government (or an agent) sells that produce.6 This more comprehensive indicator of budgetary outlays could be denoted \" Support Expenditure\" (SE). Expressed as % of the value of total agricultural production it yields the \"%SE\".Budgetary outlays arising from domestic food aid (to be defined as under the current Annex 2 (Green Box), paragraph 4) would not however be included. That is, where food is sold to clearly defined sections of the population in need, below the prevailing domestic market price, the outlay resulting from the gap between the prevailing domestic market price and the price charged to those sections of the population would not be included in the %SE.The %SE would be subject to progressive and tiered reductions over an agreed period until an agreed benchmark level is reached.The rationale behind this option is that administered prices cannot, in the absence of budgetary outlays, provide longer-term effective support to domestic production in a situation where export subsidies are eliminated and border protection is constrained through appropriately low tariff bindings. Thus, Option 3 is relevant only in a situation where significant progress has been made on market access and with full compliance with the provisions prohibiting export subsidies and disciplining other export competition measures. Countries with low market access barriers (to be defined) and fully in compliance with all export competition provisions could be allowed to opt for this approach immediately.Regarding Annex 2, two options are proposed with different levels of ambition. A first option envisages some adjustments to the current provisions while keeping the basic framework. A second more radical proposal involves a conceptually and practically different approach.In formulating these options the idea is retained that the Green Box seeks to be a list of policy measures under which support is not subject to limit (in other words any policies not meeting the criteria defined in the Green Box are subject to the limits on domestic support set elsewhere in the agreement, as it is proposed that the Blue Box disappears). The sole exception to this concerns a constraint on decoupled income support under Option 2. In addition, conditions or provisions which cannot reasonably be measured or verified in a multilateral setting are avoided. As for article 6 domestic support, generic provisions and conditions are proposed rather than attempting to define specific parameters and design features of policies.This option retains the original framework but seeks to alter and clarify specific provisions.The overarching conditions for claiming exemption from reduction commitments would be expanded to include ………\"shall not have properties that incentivise the production of any individual agricultural product or group of products.\"Paragraph 2 (f), General Services -marketing and promotion services.It is proposed to narrow the range of this provision to cover market information only or to delete it in its totality.It is proposed to clarify that the sales from food security stocks referred to in paragraph 3 relate to sales other than those to targeted recipients of domestic food aid -urban and rural poor in developing countries as in the footnote. This could be done by merging the text and footnote 6. [Like other elements of the Green Box, budgetary outlays for domestic food aid are exempted from constraints on domestic support.] It would also be specified that, as in Options 1 and 2 on domestic support, there is no requirement to measure market price support if applied administered prices are below international reference prices as defined under rules for measuring market price support.The following text amendments are proposed sub-para (d) to read: \"The amount of such payments in any given year shall not be related to, or based on, the quantity or use of the factors of production employed in any year after the base period.\" sub-para (e) to read: \"Eligibility for such payments shall not be determined by a requirement to produce any agricultural product nor by a requirement not to produce any given agricultural product.\"The remaining paragraphs of Annex 2 would also benefit from a number of specific amendments. An illustrative example of what could be done is provided in Appendix 2.Under this option distinctions would be drawn among different types of measures, essentially between direct payments on the one hand and social, developmental and public good programmes on the other, with different levels of specificity concerning programme design features, and differentiated notification and scrutiny provisions. Additional provisions and quantitative constraints are added to decoupled direct payments to underscore their desired transitional characteristic, based on the results of multiple empirical studies showing that large direct payments, even if 'decoupled' in the sense of paragraph 6, are not entirely production-neutral.Part A of the Green Box under this option would cover all direct payments currently exemptible under paragraphs 5 and 7 to 13 (with the exception of paragraph 12), and as amended above under Option 1.Part B of the Green Box would include a wide range of social, environmental, climate and developmental measures as described in paragraphs 2 to 4 and paragraph 12, as amended in Option 1. Notification of expenditures would be waived for these measures (for all members) and replaced by a periodic reporting and review in a process within the CoAg to be defined. Alternatively, Green Box notification requirements for the proposed Part B could be waived for LDCs.Part C would cover decoupled income support as currently governed by paragraph 6 of the Green Box. This Part C could be considered as a separate, intermediate category between the strictly limited domestic support (DS) and the Green Box. It would be governed by the following rules..Pre-existing programmes meeting the general and specific criteria applicable to this category will continue to be exemptible, provided they are capped at their percentage of the total value of production in the most recent period and scheduled to be reduced. The rate of reduction from the base period could be the same as that decided for the %DS of the country in question, but the time period for implementation longer. Alternatively it could be decided to continue the annual reductions of exempted decoupled income support until it reaches zero.New schemes under this heading are exempted on condition that they meet the general and specific criteria applicable and are designed as compensatory payments in the context of reforms resulting in an elimination or reduction of other programmes previously classified as non-green box support. The level of these new compensatory payments must not exceed the reduction in other forms of support resulting from the policy reform.The percentage share of such new compensatory payments in total value of production introduced in year t may be denoted %NDIS(t) (new decoupled income support). The actual level of %DS to be notified in year t is denoted %DS(t). The maximum level of new compensatory payments that can be exempted is then defined by %NDIS(t) = %DS(t-1) -%DS(t).8 As required for pre-existing schemes, compensatory payments under new schemes are required to be reduced. The rate of reduction from the level at which the new payments are introduced and the time period for implementation would be the same as those applying to pre-existing programmes.In addition to Table DS:2 notifications (\"new or modified\") as presently required under G/AG/2, more complete details of direct payment programmes falling into categories A and C, both characteristics of programme implementation and expenditures, are to be notified annually through the usual notification processes by all members claiming exemption for payments under such programmes.Proposals made above on Domestic Support and the Green Box have implications for public stockholding for food security purposes and provide several elements of a long-term solution to this issue.In particular, the use of recent reference prices for the calculation of AMS (instead of the fixed external reference price, average 1986-88, or later periods used by countries acceding to the WTO) and the explicit provision that no price gap is calculated when the domestic administered price is lower than the pertinent external reference price virtually eliminates the probability that public stockholding programmes for food security give rise to market price support which would be included as AMS. A country operating a programme with procurement prices above international reference prices would nevertheless be obliged to calculate the related AMS (as implied in option 1 for domestic support above) and to ensure that the new domestic support provisions are respected.In addition, a clarification is proposed whereby domestic food aid in the form of cash transfers to needy households or programmes providing vouchers or coupons to the same group of needy households are not considered as relevant to the measurement of support to agricultural producers and need not be notified, under the condition that they do not impose any specific obligations on recipient households to purchase domestically produced food or any other restrictions other than for health purposes (such as no alcohol, soda etc). It would also be made clear that the constraint under paragraph 3 of the green Box whereby sales from food security stocks cannot not be made at less than current domestic market prices does not apply to domestic food aid in the sense of paragraph 4.Moreover it is proposed to specify that public stockholding only for food security purposes is included under these provisions, which excludes public stocks used to \"buffer\" prices in the domestic market. Additionally, a total ban on sales from food security stocks at below market prices to any actor other than those administering domestic food aid should be made explicit.The provisions governing public stockholding and domestic food aid are available to all WTO members whether they already operate such schemes or not.The Nairobi Ministerial decisions eliminated export subsidies and exhorted members not to take measures that would circumvent that decision. They also contained new disciplines in relation to export financing, food aid and exporting agricultural state trading enterprises. Consequently, more progress has been made in the export competition pillar than in either of the other two pillars of the original AoA.To strengthen and consolidate achievements on export competition a first step would be for all countries that have not yet done so to submit updated schedules of commitments to the WTO, reflecting the implementation of the Nairobi Ministerial decision. A second step would be for all members deploying any export competition measures to agree to comply in a timely and comprehensive manner with the reporting requirements already in place in principle since Bali, but largely not respected in practice. Improved monitoring in the CoAg on the basis of thorough and up-to-date information would enable members to judge the effectiveness of the current disciplines and to consider whether further tightening would be beneficial.Another useful step would be to agree additional commitments on the monetisation of food aid and on export credits and export credit guarantees.New disciplines proposed elsewhere in this document, notably on domestic support and on public stockholding, highlight the need to ensure that market operations carried out by governments in the name of of food security do not lead to indirect export subsidisation. It is proposed that reporting requirements on government agencies involved in the management of public stocks be sufficiently detailed and explicit to allow the detection of sales at below market prices for subsequent export, and that specific provision be made for discussion of these issues during meetings of the CoAg.Enhanced monitoring of the activities of state trading enterprises would also be important in this respect. In addition, agricultural state trading enterprises should be subject to whatever enhanced disciplines and transparency provisions that might be agreed for such entities in general at the level of the WTO.It should be remembered that all export competition provisions are potentially subject to challenge under the Agreement on Subsidies and Countervailing Measures.Existing WTO rules forbid quantitative export restrictions except where a critical food shortage threatens. Export taxes are generally permitted and are not bound. During the 2007-2009 period, export restrictions caused or exacerbated problems in global food markets (G20, 2011).It is therefore proposed that, in a first step, WTO members agree immediately to exempt purchases for humanitarian purposes by recognised international agencies such as the WFP from any export restrictions whether in the form of export taxes or quantitative restrictions.It is also proposed that Members agree to bind export taxes (e.g. at the highest level imposed in the most recent five year period) and to embark on a process akin to tariffication by calculating the ad valorem equivalent (AVE) of quantitative restrictions and to begin a process of gradual reduction, leading over an agreed number of years to an elimination of all export taxes. Tiered reductions as for trade distorting domestic support could be used, or alternatively a Swiss formula, as is proposed for import tariffs under option 2 in the section on market access. A safeguard measure could be devised that makes more explicit the terms under which exceptions or exemptions from the agreed reduction of export taxes would be allowed.Members agree to notify other members likely to be affected by the imposition of an export restriction in as timely a manner as possible, on a best endeavours basis. The rationale for this approach is that governments may be reluctant to provide prior notification in relation to foodstuffs because of the risk of provoking the shortages they are hoping to avoid.There is a longstanding ministerial commitment to address \"ambitiously, expeditiously and specifically\" the issue of trade distorting domestic support for cotton. Reflecting the importance granted to this issue, progress has already been made, including in the form of strengthened monitoring processes piloted by the WTO Secretariat, the granting of duty free and quota free market access by developed countries (and developing countries on a best endeavours basis) to least developed countries, and the Nairobi agreement for the immediate elimination of export subsidies, compared to the more gradual elimination for other products. These measures together with domestic reforms in some key countries have already contributed to improving the international market environment for cotton.General proposals made in the context of the current project under the headings of domestic support, green box and market access would contribute to a further significant improvement of the situation regarding cotton. To deepen and accelerate this process, it is proposed that under a new agreement on agriculture, all limited domestic support which is specific to cotton, should be reduced at the rate applicable to the highest tier, irrespective of the observed level of cotton specific support (%DSCotton) at the outset. Agreed reductions on cotton could be applied immediately or at an accelerated pace relative to the generally agreed reductions.In addition, it is proposed that remaining import tariffs (applying to other than LDCs) also be subject to larger reductions than tariffs on other agricultural products and that these reductions be applied immediately or at an accelerated pace. If a new agreement on market access foresees tiered reductions of tariffs, then import tariffs for cotton and cotton products are reduced by the highest level of reduction. If the new market access agreement is based on the Swiss formula, a smaller value of the parameter A in that formula than applied to other agricultural products is used to calculate final bound tariffs on cotton and cotton products (such as half the value of A used for other agricultural products). In addition, it is proposed that all tariff rate quotas be abolished and the within-quota tariffs are applied.Earlier sections have drawn attention to the extent to which economic and trade weights and patterns have been altered fundamentally by the sustained growth achieved during the past 3 to 4 decades, and the contribution that deeper economic integration, facilitated to a large degree by the WTO itself, has made to that process. As a result, interdependencies among countries are deeper and wider than before, and policy actions taken in one country may spill over to a large number of other countries. While linkages between the most advanced and the emerging economies have deepened, so also have those between and among emerging and developing countries, with so-called south-south trade increasing dramatically in importance. These developments are mirrored in the agriculture and food sector.The effort required from countries in the context of multilateral trade negotiations needs to reflect their economic heft and the attendant responsibilities. In the specific context of agriculture and food, this translates into a guiding principle which has permeated this document, namely that participation in the effort to reduce obstacles to trade in agriculture should be commensurate with a country's role in global agricultural markets and with the extent to which it engages in market and trade distorting polices with the potential to disrupt markets and impose costs on other WTO members.No attempt is made here to propose a classification of members as an alternative to the current system of self-nomination. Thus, when reference is made to specific provisions for developing countries in the preceding texts, the intention is neither to reflect the status quo, nor to propose any specific alternative. This is a difficult and sensitive issue which can only be resolved at the level of the WTO more generally and not in the specific context of the sectoral discussion on agriculture. Nevertheless, and pending agreement among the WTO membership, a new agriculture agreement could incorporate helpful advances. Systemically important countries in the global agriculture and food system which currently enjoy developing country status could voluntarily opt not to claim SDT, recognising in this way that they are important players whose actions can drive world market developments. Along the lines of the respective decisions at the Hong Kong and Bali Ministerials of the WTO, LDCs should be offered duty-free and quota-free access to all other members' markets, while being invited to contribute to market opening and domestic support reforms in their own countries on a best endeavours basis. LDCs could also be offered technical assistance to design programmes that are Green Box compatible and to avoid programmes that are not. In addition, longer time periods could be allowed for members required to implement the most onerous levels of disciplines, irrespective of their development status now or resulting from a new agreement on that issue.Recent years have brought recognition that the functioning and effectiveness of the WTO needs to be improved, and members are actively engaged in driving the process. Issues have been raised concerning all three major functions of the organisation: negotiating new trade agreements, monitoring existing agreements, and the dispute settlement processes. The Doha Round, launched in 2001, has not been completed, and countries disagree on its current status (while remaining committed to addressing the outstanding issues). The WTO has therefore not been able to reach closure on a number of traditional issues, among them agriculture. Neither has it been able to advance on important \"modern\" issues such as ecommerce, or investment facilitation. Criticisms have been levelled also that the organisation is not able to deal with non-market-oriented policies and practices, that definitions of subsidies incorporated in current agreements are inadequate, that self-designation of developing country status is inappropriate, and that the activities of state -owned enterprises cannot be adequately disciplined. Civil society and some governments call for regime coherence to ensure that sustainable development goals and climate change are properly taken into account.Myriad solutions have been put forward. Specific proposals aim to speed up the dispute settlement system and to define more carefully the appropriate scope of its activities. There are ongoing discussions about strengthening the deliberative and monitoring functions of WTO Committees and bodies, as a way to avoid litigation. There are calls for an evidencebased method of identifying the level of development of member countries, in place of the current self-determined method. The need for much improved transparency, including through full compliance with notification requirements is acknowledged. Faced with the immense difficulty (or even impossibility) of achieving consensus among the WTO's 164 members, some are exploring the possibility of forging plurilateral agreements.Clearly these issues are closely intertwined with the agriculture negotiations, but they also go much beyond them. Undoubtedly, the difficulties besetting the agriculture negotiations have contributed to the general paralysis. It is also clear that the implications of resolving the issues mentioned here go way beyond the agriculture and food sector. In the preceding text therefore, to the extent possible, we have refrained from encroaching on these broader issues. Nevertheless, the pathways group understands that the proposals made here are unlikely to achieve any traction other than in a situation where there is a political willingness among the member countries to advance on the systemic issues currently besetting the WTO, as well as a political willingness to advance on agriculture. It will also be important for agreement to be reached quickly on a functioning Dispute Settlement body after a long hiatus. Members would be more inclined to embrace reform commitments in food and agriculture and other areas if they have confidence that enforcement provisions are effective.Implementation of the market access provisions of the AoA was completed in six years for developed countries and ten years for developing countries. Nevertheless, further market opening and tariff reductions have continued, partly as a result of unilateral initiatives, partly in the context of bilateral or other preferential agreements. As a result, applied tariffs across a wide range of agricultural products and a large number of countries are lower than at the time of completion of implementation of the AoA (although domestic support has almost always been omitted from preferential agreements). Preferential agreements more broadly may show the way to resolution of some of the issues that are slowing down the multilateral process. Countries participating in deep preferential arrangements have already experienced the adjustments required as well as the benefits of deeper integration. An indication of their willingness to implement similar concessions multilaterally (on a voluntary and possibly timelimited basis) would also help towards rebuilding trust and confidence in the multilateral system.The Uruguay Round AoA was, to a significant degree, made possible by the existence of a body of policy data which enabled the main protagonists to have a clear understanding of where they stood relative to their trading partners. The data currently being generated by the monitoring processes built into the agreement no longer, for reasons explained above, provide that snapshot of current patterns of trade distorting domestic support. It will be difficult to make progress without a comparable, accurate and up-to-date quantification of current policy settings. Several international organisations which are active in gathering data and generating relevant indicators have recently come together with a view to harmonising definitions and methods 9 . This work could inform early-stage discussions, and give countries a more accurate and economically meaningful view of where they are situated relative to their negotiating partners. The data and analysis needed to do this are already in the public domain (IFPRI 2020, FAO-MAFAP, OECD 2020b)The ideas and options presented here constitute a \"whole\". They have been designed to point the way to achieving significant improvement in the functioning of global markets for agriculture and food products, while allowing ample scope for win-win policy solutions supportive of social, developmental, environmental, and climate goals as well. The overall package is balanced, allowing grievances to be addressed, but also aiming to persuade countries to concede on long held positions on some issues in order to gain significantly from the overall package. A holistic approach going forward on all the issues under negotiation is therefore to be preferred over piecemeal efforts. significant risks, including the risk of hoarding and inappropriate policy measures, but, overall, the food system has shown strong adaptability and resilience.As countries across the globe moved to confine their populations in the attempt to contain the spread of the virus, there were immediate impacts on logistics and transport. Most heavily affected was air freight, generally used for high value, perishable and seasonal products. Air freight of food products is closely linked to passenger traffic, which essentially collapsed and which has, so far, recovered to only a small fraction of its previous volume. Road transport was also affected to some degree, although less so for agricultural and food products than for other types of merchandise. Movement of bulk commodities, which generally occurs by ship, was the least affected, although there were some disruptions.On the production side, the most immediate effects have been felt in terms of shortages of seasonal, often migrant labour, which is important in many sectors but notably in fruit and vegetables. Issues also arose in abattoirs and meat processing plants, also often staffed by migrant workers and where the conditions are particularly favourable to the propagation of the virus. Significant falls in production of meat have been reported in some countries, although they seem to have been shortlived. Fears of shortages of inputs such as fertilizers, agricultural chemicals and seeds seem to have abated and again, disruptions have been relatively short in duration when they occurred.The coronavirus pandemic has led to very significant shifts in consumption patterns around the world. Confinement orders led to some initial hoarding and panic buying on the part of consumers, heavily favouring staples, and frozen or other long-life foodstuffs. Driven by cautious behaviour or by closure orders, the volume of food consumed away from home fell significantly, leaving large amounts of fresh produce -meat, fish, fruits and vegetableswhich could not immediately be redirected to retail outlets. However, commentators generally agree that the system adjusted rather rapidly to the new circumstances, notwithstanding disruptions leading to losses and waste in some areas.Coronavirus is having dramatic impacts on employment and incomes, particularly among populations already suffering from poverty and precarity. This is the channel whereby the most dramatic effects have been seen, although rising prices which began to emerge later are adding to the problem. Across the world, charities and food banks report large increases in the number of people needing food assistance. The ranks of the unemployed have swelled alarmingly. The degree of suffering has varied according to the willingness and capacity of governments to deploy emergency assistance whether in kind or in the form of augmented cash transfers or unemployment benefits. The World Food Program estimated that 270 million people would experience severe food stress as a result of the pandemic by the end of 2020, an 82% increase compared to the pre-pandemic situation. While not restricted to emerging and developing countries, these developments have hit this group of countries particularly hard, in part because of the high degree of informality in the labour markets. Also, as outlined by IFPRI, countries where the food system is transitional in nature have experienced particular difficulties because of the large number of relatively small actors in the chain, whose activities have been restricted. These transitional food systems have fared less well than subsistence models where production and consumption are largely local, and less well than large integrated modern systems which have been able to continue operations without being overly disrupted. Overall, impacts on access to food related to loss of employment and income are likely much more severe in developing countries. The WFP has warned that \"The COVID-19 pandemic and related containment measures are expected to continue exacerbating economic crises and acute food insecurity, particularly in fragile economies. … . Even if COVID-19 can be contained in some parts of the world, slow vaccine roll-out in countries with poor health services could prolong restrictions, dimming the prospects for swift economic recovery.\" (WFP 2021).Because governments moved quickly to classify agricultural and food production, processing and distribution as essential activities, agricultural and food trade seems to have fared better than many other sectors, exports actually increasing through the early months of 2020, before declining in May. (Later data are not yet available). As communicated regularly to governments and to the general public by AMIS, production and stocks of the major food staples it covers are at or near all-time highs and market fundamentals are considered comfortable. But international prices of most agricultural commodities have risen sharply and food inflation is observed in some countries. These developments will require careful monitoring. AMIS underlines the importance of an effective and well-functioning trade environment ir order to safeguard market stability and prevent excessive levels of pirce volatility (AMIS Market Monitor, May 2021).Therefore, as long as continuity in trade, trade policy, transport and logistics is assured, there is no fundamental, underlying issue in relation to basic food availability. This international effort, launched by the G20 in 2011 to provide real-time information and transparency about the state of world markets for key staple food commodities in the wake of the 2007-08 food price crisis, has proved its worth.Governments did nevertheless undertake a certain number of policy initiatives, some trade restrictive, and some trade enabling. A number of domestic measures were also taken aimed at ensuring that the food system could continue to function or providing assistance to those impacted by virus related restrictive measures or by the associated economic downturn.New SPS measures were initially restrictive as governments sought to ensure that food trade was not a vector for transmission of the virus. Subsequently measures to facilitate trade such as permitting electronic certification became more common. Other trade enabling measures included tariff reductions, postponements of payment of VAT or import duties, and increases in tariff rate quotas.Similarly to 2007/08 a number of governments put export restrictive measures in place with a view to ensuring supplies to their own populations and avoiding price hikes domestically. Measures included export bans, quotas or licensing arrangements and affected mainly but not exclusively food staples or related products. While about 20 countries had implemented such measures some have already expired or been eased after some time. Among the WTO membership, a relatively small number of countries notified the measures as required by the Agreement on Agriculture. Overall however, there have been fewer measures and their impacts have been less damaging than during the 2007-08 crisis when the proliferation of measures considerably exacerbated the evolving crisis.In addition to activating safety nets, many governments have taken measures directly impacting the agriculture or food sectors. Some stepped up food purchasing and distribution schemes with the dual intent of supporting farmers and poor consumers. Others stepped in to help farmers by easing credit etc.The coronavirus pandemic is ongoing with a strong resurgence of the disease in many parts of the world. Nevertheless, where newly developed vaccines are being deployed there is some hope of containing the situation and of a return to normality in the not too distant future. Global value chains and international trade have demonstrated considerable resilience as governments moved to ensure that the sector receive priority support as an essential activity. Governments have also, for the most part, not hampered the operation of international trade, and in some cases have taken liberalising measures. These outcomes bear witness to the key role played by transparency (e.g. in the form of AMIS) and to the importance of multilateral institutions and legal frameworks within which governments have largely operated, and which have enabled continuity in the smooth operation of international trade in food. As the pandemic persists and provided that governments continue to exercise restraint, trade will continue to be an important element in ensuring food security across the world, including the food security of the most vulnerable groups. In the context of this paper, recent events are a reminder of the importance of a well functioning, transparent and predictable world trading system in agriculture and food products and underline the urgency of further consolidating and building on reforms already undertaken. (FAO/AMIS 2021, FAO 2020, OECD 2020a).","tokenCount":"13532","images":[],"tables":["-426642802_1_1.json","-426642802_2_1.json","-426642802_3_1.json","-426642802_4_1.json","-426642802_5_1.json","-426642802_6_1.json","-426642802_7_1.json","-426642802_8_1.json","-426642802_9_1.json","-426642802_10_1.json","-426642802_11_1.json","-426642802_12_1.json","-426642802_13_1.json","-426642802_14_1.json","-426642802_15_1.json","-426642802_16_1.json","-426642802_17_1.json","-426642802_18_1.json","-426642802_19_1.json","-426642802_20_1.json","-426642802_21_1.json","-426642802_22_1.json","-426642802_23_1.json","-426642802_24_1.json","-426642802_25_1.json","-426642802_26_1.json","-426642802_27_1.json","-426642802_28_1.json","-426642802_29_1.json","-426642802_30_1.json","-426642802_31_1.json","-426642802_32_1.json","-426642802_33_1.json","-426642802_34_1.json","-426642802_35_1.json","-426642802_36_1.json","-426642802_37_1.json","-426642802_38_1.json","-426642802_39_1.json"]}

data/part_2/0337134352.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"335cfaf1e506c872164f300e6456a385","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/07e992d2-0f03-40bd-88d8-728890a2dd14/retrieve","description":"Using the SAM multiplier model for Egypt, we simulate the individual and combined effects of a collapse in the tourism sector and reductions in Suez Canal revenues and in foreign remittances under more and less pessimistic scenarios. SAM multiplier models are well-suited to measuring short-term direct and indirect impacts of unanticipated, rapid-onset demand- or supply-side economic shocks such as those caused by the COVID-19 pandemic. We model the demand shocks as the anticipated reductions in tourism, Suez Canal, and remittances revenues.","id":"603758662"},"keywords":[],"sieverID":"aaac9829-d213-4cd2-b6ae-97879959c6a1","pagecount":"4","content":"The economic impacts of the COVID-19 crisis are increasingly hitting low-and middle-income countries and the poor. International travel restrictions and the full or partial closure of businesses and industries in Asia, Europe, and North America have led to a collapse in global travel and are expected to reduce the flows of remittances. Tourism and remittances are important sources of employment and incomes for the poor. This post assesses the potential impacts of the expected reductions in these income flows by using Egypt as a case study.The pandemic is likely to have a significant economic toll. For each month that the COVID19 crisis persists, our simulations using IFPRI's social accounting matrix (SAM) multiplier model for Egypt suggest national GDP could fall by between 0.7% and 0.8% (EGP 36-41 billion or US$2.3-$2.6 billion). Household incomes are likely to fall, particularly among the poor.Egypt is a rising star among emerging economies. Even though several reforms remain to be completed, the reform program launched in 2016 has started to bear fruit: Egypt has achieved economic growth of over 5% in the last two years. The tourism sector recorded its highest revenues in 2018-19, another sign of increased stability. Continued efforts aimed at improving Egypt's business climate were expected to lead to even stronger private sector growth and economic diversification in 2020 and beyond. This progress will almost certainly be interrupted by the COVID19 pandemic. While the government is taking actions to contain the spread of the virus -including the suspension of commercial international passenger flights, school and sports clubs closures, and a nationwide nighttime curfew -and the number of reported infections in Egypt is currently low compared to that of many other countries, the global economic slowdown is expected to have major knock-on effects for Egypt. International travel restrictions are already curtailing tourism to the country. The global slowdown is likely reducing payments received from the Suez Canal and remittances from Egyptians working abroad. These three sources together account for 14.5% of Egypt's GDP. Thus, any disruptions to these foreign income sources will have far-reaching implications for Egypt's economy and population.Using the SAM multiplier model for Egypt, we simulate the individual and combined effects of a collapse in the tourism sector and reductions in Suez Canal revenues and in foreign remittances under more and less pessimistic scenarios. SAM multiplier models are well-suited to measuring short-term direct and indirect impacts of unanticipated, rapid-onset demand-or supply-side economic shocks such as those caused by the COVID-19 pandemic. We model the demand shocks as the anticipated reductions in tourism, Suez Canal, and remittances revenues.Our results show the potential significant impact on the economy and people for each month that the COVID-19 crisis persists. If the dynamic effects of the COVID-19 shock on the Egyptian economy are different than those simulated, our results could be either under-or over-estimations of the aggregate economic impact of the crisis. Also, effects from other channels may reinforce the effects of the pandemic.These expectations also assume that there is no change in the current government policies in place to combat the crisis. This is important to note, as the government is taking aggressive action to contain the disease and support the economy and people. As such, the model scenarios do not consider the impacts of specific government economic policies, but are intended to support government decision-makers in determining the scale of their support to the economy and to Egyptian households.Figure 1 breaks down estimated losses in GDP, which may hit 0.8% per month in the more pessimistic scenario. Lower tourist spending will affect not only hotels, restaurants, taxi enterprises, and tourist guides, but also food processing and agriculture. Lower public revenues from Suez Canal fees are likely to affect the government budget. Lower remittances income will likely reduce household consumption of consumer goods and hit sectors producing intermediate goods. We estimate that the absence of tourists alone may cause monthly losses of EGP 26.3 billion, or $1.5 billion. Thus, the loss in tourism revenues accounts for about two-thirds of the total estimated impact.Household incomes are estimated to decline by between EGP 153 or $9.70 (less pessimistic scenario) and EGP 180 or $11.40 (more pessimistic scenario), per person per month for each month that the crisis continues (between 9.0% and 10.6% of household income). The expected reduction in tourism has the strongest effect on all households, making up more than half the economic impact for all household types in the model (Figure 2). Households are also affected directly and indirectly by lower remittances from abroad. The less pessimistic combination scenario assumes a 10% reduction in Suez Canal revenues and in remittances. The more pessimistic scenario assumes a 15% reduction in these payments. Both combination scenarios assume a complete absence of international tourists. GDP = gross somestic product.-0.8 -0.6 -0.4 -0.2 -0.1 -0.3 -0.5 -0.7 -0.9 0.0 While all households are hurt by lower tourist expenditures, it is poor households -and especially those in rural areas -that suffer the most from lower remittances. Due principally to the relatively greater decline in remittances that they experience, rural poor households are estimated to lose in total between EGP 104 and 130 ($6.60-$8.20) per person per month, or between 11.5% and 14.4% of their average income, while urban poor households will see their incomes decline somewhat less, between EGP 80 and 94 ($5-$6) per person a month, or between 9.7% and 11.5% of their average income.If the crisis persists for at least three to six months, as many now believe likely, the cumulative loss from these three external shocks alone could amount to between 2.1% and 4.8% of GDP in 2020. Importantly, our simulations measure only the effects that might result from specific impact channels, namely, foreign sources of remittances and revenues. Domestically, restrictions on movement of people and goods within the country and on certain productive activities may also have adverse economic impacts. On the other hand, some sectors may benefit, such as information and communications technologies, food delivery, or the health-related goods and services sectors.The authorities have begun a course of decisive action to curb the virus outbreak by allocating EGP 100 billion ($6.3 billion) and have enacted tax breaks for industrial and tourism businesses, reducing the cost of electricity and natural gas to industries, and cutting interest rates. They are also considering providing grants to seasonal workers. Additional measures may also be in the works, such as increasing cash transfer payments to poor households, increasing unemployment benefits, and providing targeted support to specific sectors.While the country's focus currently is rightly on fighting the health crisis and mitigating its immediate impacts, planning on how to re-open the economy should start now. To emerge stronger after the COVID-19 crisis, both the public and private sectors should continue to strengthen their Unless governments around the world take decisive action, the case of Egypt suggests that poverty is likely to increase in countries where tourism and remittances play a large role. It is also a strong reminder of the interconnectedness of the world and the importance of global cooperation to end this crisis and to be better prepared for the future.","tokenCount":"1181","images":[],"tables":["603758662_1_1.json","603758662_2_1.json","603758662_3_1.json","603758662_4_1.json"]}

data/part_2/0347683295.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"218a625862776ac06665ef3a00e5f981","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/98474670-7d78-4d6c-90e3-baae5ed82615/retrieve","description":"Farmer hiring of agricultural machinery services is common in South Asia. Informal fee-for-service arrangements have positioned farmers so they can access use of machinery to conduct critical, timesensitive agricultural tasks like land preparation, seeding, irrigation, harvesting and post- harvesting operations. However, both the provision and rental of machinery services are currently dominated by men, and by most measures, it appears that women have comparatively limited roles in this market and may receive fewer benefits. Despite the prevailing perception in rural Bangladesh that women do not participate in agricultural entrepreneurship, women do not necessarily lack a desire to be involved. Using a mixed methods approach involving literature review, secondary data collection, focus groups and key informant interviews, and a telephone survey, we studied the gendered differences in women’s and men’s involvement in emerging markets for rice and wheat reaper-harvester machinery services in Bangladesh. We find that women benefit from managing and sometimes owning machinery services, as well as from the direct and indirect consequences of hiring such services to harvest their crops. However, a number of technical, economic, and cultural barriers appear to constrain female participation in both reaper service business ownership and in hiring services as a client. In addition, women provided suggestions for how to overcome barriers constraining their entry into rural machinery services as an entrepreneur. Men also reflected on the conditions they would consider supporting women to become business owners. Our findings have implications for addressing social norms in support of women’s rural entrepreneurship and technology adoption in South Asia’s smallholder dominated rural economies.","id":"-684228055"},"keywords":["gender","cultural norms","agricultural technology","scale-appropriate mechanization","rural machinery service provision","Bangladesh v"],"sieverID":"acf6e5c0-13a2-4c9c-b778-da37705c4d1b","pagecount":"48","content":"established in 1975, provides research-based policy solutions to sustainably reduce poverty and end hunger and malnutrition. IFPRI's strategic research aims to foster a climate-resilient and sustainable food supply; promote healthy diets and nutrition for all; build inclusive and efficient markets, trade systems, and food industries; transform agricultural and rural economies; and strengthen institutions and governance. Gender is integrated in all the Institute's work. Partnerships, communications, capacity strengthening, and data and knowledge management are essential components to translate IFPRI's research from action to impact. The Institute's regional and country programs play a critical role in responding to demand for food policy research and in delivering holistic support for country-led development. IFPRI collaborates with partners around the world.Farmer hiring of agricultural machinery services is common in South Asia. Fee-for-service arrangements have made agricultural mechanization substantially more accessible to smallholder and resourceconstrained farmers who no longer need to purchase capital intensive machinery to make use of equipment to conduct critical, time-sensitive agricultural tasks like land preparation, seeding, irrigation and harvesting. Appropriate machinery options can help farmers to reduce their own labor and drudgery, as well as expenditures on hired labor. Conversely, farmers who own machinery can benefit as rural entrepreneurs who offer machinery to farmers on an affordable fee-for-service basis (Baudron et al. 2015;Mottaleb et al. 2016).However, like many new agricultural technologies and market opportunities in South Asia, it appears that the early adopters, both the farmer-renters and machinery service providers, are predominantly men. This is especially the case in rural Bangladesh, where women face strong social and cultural constraints to participating in paid work outside the home, including agricultural work in the field (Sraboni et al. 2014). While closing the gender gap in women's access to agricultural technology is considered a key strategy for rural women's economic empowerment (FAO 2011), the case of fee-forservice markets in Bangladesh illustrates how there are conflicting explanations in the literature on gender and technology adoption for why this gender gap exists and what should be done about it.One view in the literature is that there is a gender gap in adopting new technology primarily because women face greater constraints to accessing capital, credit, and information than men do (Doss 2001;Quisumbing and Pandolfelli 2010;Fletschner and Kenney 2014;Peterman, Behrman, and Quisumbing 2014). In addition, women's limited access to complementary inputs, including land and the ability to marshal family or paid labor, can reduce women's willingness to adopt technology (Doss and Morris 2001). Development approaches informed by this view often focus on supporting women's access to the complementary resources needed to adopt technology, including subsidies and direct transfers of technologies. Women's formal ownership is often a goal.A second perspective in the literature suggests that \"ownership\" is not necessarily an accurate representation of who benefits or bears the costs from a given technology, especially in households with multiple decision makers and income generation sources (Theis et al. 2018). This strand of literature examines how the benefits of a technology are distributed as people negotiate for control over a technology and the resources associated with its use, both within adopting households (Agarwal 1994;Doss 2001; Njuki et al. 2014) as well as within communities (Rahman and Routray 1998;Kaaria and Ashby 2001;Beuchelt 2016). In this case, subsidies or direct transfers of technologies targeting women are not seen as a guarantee of women's control over the technology, especially for women in male-headed households. At the same time, even if men are considered the formal owner a given technology, women are not completely excluded from rights to make use of or benefit from the technology. Importantly, some technologies might save labor for men but increase time use for women. For example, while irrigation technology can facilitate intensified double and even triple cropping within the same calendar year, it can also increase women's time use and drudgery for weeding multiple crops (Chancellor and O'Neill 2000).Finally, a third perspective questions whether agricultural technologies available in the market are relevant and useful for women-if they are not, why would women adopt them? This approach emphasizes that women have needs and preferences that can be distinct from men, which are often inadequately taken into account in research and development processes (Doss 2001;Rathgeber 2011;Ragasa 2014). Instead of working to promote women's uptake of technologies they may not want or need, there is increasing evidence that research for development programs can improve the rate of farmer adoption by assessing whether existing technologies are compatible with women's roles and responsibilities (Meinzen-Dick et al. 2014). Willingness to pay for a specific technology can differ for men and women within the same household, as they prioritize reducing drudgery and raising household income differently (Padmaja and Bantilan 2008;Khan et al. 2016). When men hold primary control over household finances and purchasing decisions, they conversely may not necessarily prioritize women's preferences for certain technologies (Rathgeber 2011).These three views provide different explanations for why women do not use new technologies.Each explanation suggests different interventions to improve women's equity in access, use and benefits of technology. Policymakers and donors often choose one explanation to guide their policy and programming actions without adequate assessment of the others. But the interpretations of each narrative are not mutually exclusive. Each can be seen as interlocking pre-conditions for women's technology adoption: a woman must have awareness and access (narrative 1) to desirable technologies she prefers (narrative 3) and be able to effectively negotiate the benefits and costs associated with technology use (narrative 2). Rather than choosing a single narrative on an a priori basis, if each narrative is investigated in a given setting, development actors will be better able to describe gender dynamics around agricultural technology and choose strategic, multi-faceted actions to advance gender equity.In this paper, we assess the relevance of these different perspectives in explaining women's apparent limited inclusion in fee-for-service machineries in Bangladesh. In so doing, they can help us articulate what gender equity with respect to agricultural machinery might look like in this context. How women themselves assess their current roles in agricultural technology markets, and in what ways, if any, they wish to expand their participation, is of central relevance when considering the implications of this market on gender dynamics, and what approach would be most effective in advancing gender equity.This research was conducted in three phases. The first phase entailed literature review and secondary data analysis to explore gendered patterns of agricultural machinery use in Bangladesh using a large, nationally representative survey. The second phase consisted of qualitative field research to provide detailed insights as to the gendered aspects of reaper-harvester use in southwestern Bangladesh. During the third phase, a telephone survey was conducted to explore the implications of harvesting machinery on women's daily activities and labor allocation. Ethical approval for the second research phase was obtained from the Institutional Review Board of the International Food Policy Research Institute.Informed consent was obtained from all participants in the second and third phases of research.We focused on machinery services for multi-crop reaper-harvesters, which enable farmers to rapidly cut rice and wheat, although sesame, jute, and other crops can be harvested. We chose this technology because relative to other agricultural technologies such as power tillers and seeders, reapers only formally entered the Bangladesh market in the last decade and thus present an opportunity to study emerging gender dynamics and opportunities associated with a newly introduced technology. Two types of reapers are common in Bangladesh: the self-propelled 'walk behind' model and the two-wheel tractor-attachable and rideable models (Figure 1A and 1B).Both technologies are designed to reduce drudgery and accelerate the rate at which crops can be cut for drying and/or carrying from the field at maturity. Service providers running reaper businesses often hire skilled machine operators to harvest farmers' fields. Studies indicate that reapers can reduce the time and costs of harvesting by 80% and 60%, respectively, while enabling the rapid clearing of land so that farmers can sow the next crop by the recommended planting date (Theis et al. 2018). In the initial phase of research, we conducted a literature review, which aimed to provide insights into women's roles in agriculture in Bangladesh and, more broadly, to assess existing evidence on gender and agricultural technology, with specific attention paid to research exploring how agricultural technology benefits women. Additional insights on gender and rural mechanization in Bangladesh were gleaned from analysis of data from the 2015 Bangladesh Integrated Household Survey (BIHS), a nationally representative, multi-topic household survey in Bangladesh covering 6500 households (IFPRI 2016). Specifically, we conducted a descriptive analysis of agricultural machinery usage among farms of different sizes and of differences in men's and women's time use across a range of activities. Findings from both of these exercises informed the design of the qualitative protocols used in the second research phase.Field study sites were chosen in late 2017 in coordination with staff from the Cereal Systems Initiative for South Asia (CSISA) project. These included four villages from Jheneidah and Faridpur districts, with anadditional four villages (two in Faridpur, one in Jheneidah, and one in Magura) selected to reach women machinery service providers who were not available in the primarily selected villages. Site selection was based on CSISA field staff's experience and prioritized a high density of machinery service providers and high uptake rates of reaper technologies, while maintaining comparability across sites in terms of agricultural and economic livelihoods and infrastructure. Further selection criteria included whether people in the villages tended to hold more progressive or more conservative gender norms, with sites representing each set of views selected to provide a basis for comparison.Three categories of respondents were targeted: reaper service providers, farming households using reaper services, and farming households not using reaper services. Reaper service providers were purposively selected with the assistance of CSISA field staff in the district offices. The other categories were selected as advised by local researchers and through snowball sampling while speaking to respondents. Male and female reaper service providers were targeted, as were women and men registered as renting machinery services in databases maintained by CSISA. Husbands and wives (sometimes women and sons-in-law) involved in the service provider business or rental of machinery services were both interviewed, separately. Women in non-mechanized farming households, who had never rented any machinery services, were also interviewed. Along with these three major categories, key informants (e.g., community leaders, CSISA staff, machinery dealers, community leaders, etc.) were also interviewed.Data were collected through semi-structured interviews, focus group discussions, and key informant interviews. The first round of data collection was conducted from October to November 2017 in Jheneidah and Faridpur districts. The first round of data collection interviewed husband and wife machine service providers (18 women, 18 men), men and women who hire these machines for their farms (25 women, 17 men), women farmers in non-mechanized farming households (12 women), CSISA staff (8 men), and community leaders (1 woman, 1 man) (Table 1). Semi-structured interview guides were developed and translated into Bangla. The interview guides were pre-tested for cultural relevance in communities in Jheneidah. A second round of data collection was conducted from October to November 2018, mainly in Jheneidah district, and focused on clarifying questions with previously selected respondents, as well as additional key informants from CSISA staff and local women's groups. For the second round of data collection, new interview guides were translated into Bangla, pre-tested, and refined to explicitly probe responses provided in the first round. The main topics covered in both rounds included perceived impacts and gender roles in reaper and machine service provision and rental.All interviews were transcribed verbatim in Bangla, translated into English, and uploaded to NVivo (Ver. 12, QSR International, Doncaster, Victoria, Australia). Interviews were conducted or supervised by one of the authors, a trained Bangladeshi sociologist, and a team of trained graduate-level facilitators fluent in Bangla and English. Interviews were mostly held in a private household compound or at an office (mainly for CSISA staff). Focus group discussions were conducted with men and women separately. They provided contextual information and normative perspectives on how women and men perceive the impact of reaper rental on their lives. Key informant interviews focused on men and women's perspectives on gender roles in the service provider business and in renting reaper services.The transcripts were coded following principles of thematic analysis (Braun and Clarke 2006;Nowell et al. 2017). Data were coded using both pre-set and emergent themes. During data analysis, themes were compared across the transcripts to assess men and women's perspectives on women's current and potential and/or desired roles in local service provider (LSP) businesses and in machine rental. A second round of data collection further probed the priorities and key barriers identified in the first round by respondents and gathered feedback on options to increase equity in machinery services.In the third phase of research, a telephone survey of farm households that hire rice and wheat reaping services was conducted to examine the implications of mechanized harvesting on women's time and labor allocation during the harvesting period, both before and after households adopted reaping services. The CSISA project maintains a database of 6,674 farmers making regular use of machinery services. A random sample farm households was drawn for reaping of rice (n = 254) and wheat (n = 255). All households had hired reaping services for at least two seasons (rice or wheat) over the last two years.Eight enumerators, five of whom were women, were trained to telephone households in March of 2019.Households were called and recalled until contact was made and enumerators were permitted to speak with the primary woman of the household. After being read a standard confidentiality agreement, respondents were asked to estimate the proportion (in hours) of their activities (inclusive of child care, cooking for family members or farm laborers, livestock, poultry or aquaculture activities, harvesting and post-harvest, farm laborer supervision, leisure and other income generating activities and sleep) within a typical 24-hour day during aman rice and wheat harvesting prior to and after the household's adoption of mechanized reaper harvesting. Statistical differences in data for before and after reaper adoption were analyzed for each crop and activity individually by employing two-tailed paired T-tests in JMP (Ver. 14) software.Research results are provided below starting with the first phase that included literature review and exploration of the BIHS survey results. Results from the second phase involving field research are organized into two broad categories: women's roles in providing machine services and women's roles in renting machinery.The growing fee-for-service model for farm mechanization challenges conventional understandings of what constitutes gender-equitable technology adoption. Policy and development interventions tend to focus on reducing the gender gap in agricultural technology adoption by promoting women's ownership of technologies (Peterman et al. 2014). Conceptually, individual ownership of machinery, rather than rented access or use rights, can increase the value of women's assets and can help strengthen bargaining power within relationships. However, when considering machinery rental or hiring markets, where smalland medium-sized farmers access machines on a fee-for-service basis, ownership is not necessary. This model of agricultural 'services provision' however permits machinery owners to recoup capital investment costs by serving neighboring farmers as business clients (Mottaleb et al. 2016). Owning agricultural machinery is therefore not only a significant financial investment but may also represent a commitment to starting and operating a business. For this reason, promoting women's ownership requires not only the ability to make the initial purchase but also implies the capacity and financial skills to initiate and successfully run a profit-making business.Conservative cultural norms in Bangladesh however result in a situation in which it relatively uncommon for women to be directly involved in owning/renting-out or renting-in use of machinery services. This situation appears to reinforce barriers to entry for women in either role. Conversely, increasing the number of women as service providers may help to reduce social restrictions for women renters, and vice versa. Mechanized land preparation in Bangladesh using two-wheel tractors (2WTs) has expanded significantly since import tariffs on 2WTs were removed in 1995. Presently, more than 550,000 two-wheeled power tillers, most of which are made in China under the Sifeng and Dongfeng brands (Alam et al., 2017;Krupnik et al. 2013), are used in Bangladesh. Demand for 2WTs has remained high in Bangladesh, not only because power tillers or reapers can be attached and used for land preparation or harvesting, but also because these implements can be removed and substituted with trailers and used for hauling outside the farm, underscoring the importance of modular equipment and ways for farmermachinery owners to bundle services and generate year-round income. 2WTs have been used in Bangladesh since the 1980s (Mottaleb et al. 2016) and now have an annual import market estimated at over $50 million per year. Harvesting equipment is conversely newer, with a market value of $1.2 million per year (Alam et al. 2017).Who accesses and uses agricultural machinery in Bangladesh? Early critique and debate on rural mechanization centered on the themes of labor displacement and uneven access. Use of tractors in South Asia, for example, was reported as favoring mainly wealthier farmers who could afford to purchase imported machinery rather than smaller landholders (Pearse 1980). More recently, data from the 2015 BIHS show that 80.6 percent of farms in Bangladesh used 2WT-operated power tiller services, supplied mainly through fee-for-service arrangements (Figure 2). The BIHS however does not capture data on emerging reaper-harvester service provision markets. The majority (44 percent) of Bangladesh's farmers are small (farming 0.2-0.6 ha), while 36 percent are marginal (cultivating less than 0.2 ha). Conversely, ownership of 2WT is greatest amongst medium (0.6-1 ha) and large-scale farmers (greater than 1 ha; 3.7 and 13.0 percent ownership, respectfully), while only 2.4 and 0.4 percent of small and marginal farms, respectively, own 2WT (Figure 2). Nonetheless, 84 and 76 percent of small and marginal farms, respectively, rented power tillers during the past year (compared to 84 and 79 percent of medium-and large-scale farms, respectively) (Figure 2). In comparison, only 15 percent of farms used draft animals (with little variation by farm size), indicating a clear preference among farmers for 2WT-operated power tiller services.In terms of gender, the primary users of hired operated power tiller services are male. Of the 82.3 percent of farms that use 2WT-operated power tiller services, 82.5 percent of service provision businesses are exclusively male-managed. A similar gender disparity exists in terms of the providers of custom-hire machine services in Bangladesh. Of the 257 2WT power tiller service providers identified in the 2015 BIHS, 98 percent are male. While additional support may be needed to strengthen access to these services by marginal farmers, mechanized land preparation is nearly equally distributed among small, medium, and large-scale farmers in Bangladesh, with little difference between these groups owing largely to the pervasive system of custom hiring service provision (cf. Mottaleb et al. 2016).This system is however generally inaccessible to women both as users of custom-hire services or as providers, a result of generally conservative cultural norms in Bangladesh that restrict women's ability to engage in farm production and management. When assessing how agricultural technology can serve women's strategic interests in Bangladesh, it is important to recognize that women play important and growing, but less visible, roles in agriculture than men (Zaman 1995;Mahmud, Shah, and Becker 2012;Amin 1997;Bose, Ahmad, and Hossain 2009). Definitions of economic activities often fail to capture women's contributions to household production, and much of women's work is not visible to outsiders (Zaman 1995).In rural Bangladesh, the practice of purdah, often translated as female seclusion, excludes women from public spaces and restricts women's interaction with men outside the family, with important implications for women's mobility, economic activity, and well-being (Feldman and McCarthy 2006;Kabeer 1990;Amin 1997;Bose, Ahmad, and Hossain 2009;Mahmud, Shah, and Becker 2012;Ahmed and Sen 2018). Amin (1997) argues for viewing purdah as a broader code of conduct for female morality, with observance granting prestige, and shame and loss of status accompanying violations of the practice.The cultural and religious norms behind purdah reinforce women's economic dependence on men by segregating women's labor to the private sphere while men handle public affairs of the household (Kandiyoti 1988;Zaman 1995;Kabeer 1994). In general, women need to seek permission or at least inform the husband or in-laws when leaving the homestead (Mahmud, Shah, and Becker 2012).Observing purdah often concentrates rural women's work on home-based or near-home activities referred to as 'bari' (home-based productive activities with exchange and/or market value, for example including homestead gardening, postharvest processing, livestock and poultry rearing), and a range of other domestic activities. Figure 3 shows a breakdown of the average number of minutes per day devoted to different categories of market (paid) and non-market (unpaid) work activities, for men and women, ages 15-64 years old, in Bangladesh. While men and women both spend roughly 8 hours working each day, the vast majority of men's time is spent in market activities, such as farming, wage/salary employment, or own business work, whereas women's time is mostly spent in non-market work activities, such as cooking, cleaning, and caring for children and the sick and elderly. Considering rice farming activities, women's main responsibilities include pulling seedlings, transplanting, weeding and harvesting (although even these activities may be restricted in parts of the country), while men engage in land preparation, harvesting, and threshing, which are already largely mechanized (Pandey et al. 2010).Women in many areas assist with manual harvesting and bundling and carrying crops to threshing centers.They also manage most post-harvest rice processing, including drying, parboiling, cleaning, grading and storage of rice. In contrast, men manage field-based agricultural work, market all household products, including those produced by women, and shoulder the socially-assigned responsibility to support the family economically (Zaman 1995;Begum 1985). Women's economic contributions are cost-saving but may not generate cash, and these activities blend with their other tasks within the compound (Amin 1997). According to recent national statistics, the majority of rural women (71 percent in [2005][2006] work as unpaid family workers, compared to 12 percent of rural men (BBS 2015). own farms, and landless women often work as wage laborers on the farms of others (Zaman 1995).Between 1999-2000 and 2005-2006, men's employment in the agricultural sector declined by six percent, while female labor force participation in agriculture more than doubled from 3.8 million to 7.7 million.As a result, the proportion of women in the agricultural labor force increased from less than 20 percent to 33.6 percent (Jaim and Hossain 2011). These dynamics are also changing as gains from female employment outweigh the potential cost to social identity (Ahmed and Sen 2018). As out-migration of men from rural areas reduces agricultural labor supply, women are beginning to fill roles previously assumed by men (Pandey et al. 2010). Between 2005Between -2006, about 17 percent of rural men and three percent of rural women worked as agricultural day laborers (BBS 2015). Female wage laborers work at lower wages compared to their male counterparts, and can be paid irregularly and/or in food or in kind (Amin 1997;Rahman and Routray 1998;Kelkar 2009). Women may also assume managerial roles in crop production in the absence of their husbands, but the extent of women's involvement depends on the length of time that the husband is away, as temporary and seasonal migrants will often return to their farms for key activities in the crop production cycle (Jaim and Hossain 2011).Previous studies have argued for greater nuance around the concept of women's empowerment in Bangladesh. Mahmud et al. (2012) describe how empowerment is more visible in women's selfperceptions and relational aspects of their lives than in their personal autonomy. Despite women's low mobility and control over income, women report relatively high levels of involvement in decision-making and self-esteem (Mahmud et al. 2012). This may represent a strategic \"patriarchal bargain\" for women (Kandiyoti 1988) and a reflection of the social risks associated with the pursuit of autonomy in personal or economic decisions, as other scholars in South Asia have noted (Kabeer 2011;Rashid 2013).Starting business as an agricultural machinery service provider can be a significant investment.Considering reaper-harvesters, costs can vary between USD 500 to 2,000, depending on the make and model. While women have some access to finance through village savings and loans and microcredit offerings, these mechanisms are typically insufficient large capital investments. Men finance the machine through a combination of cash, trade in crops, taking loans from NGOs such as BRAC, renting-out land, and receiving loans from machinery dealers. In focus groups, when women were asked how they would prefer to finance such a purchase, most women indicated that personal savings or selling vegetables or livestock were the best options. Both men and women appreciated potential earnings from reaper services provision, which can exceed USD 1,250 per year from harvesting monsoon 'aman' season rice in October-December, in addition to wheat in March-April. These crops are typically grown in sequence, enabling reaper service providers to capture at least two harvesting business opportunities within one calendar year.Three of the women interviewed independently managed reaper service businesses. Their husbands had left agricultural activities to them due to their husbands' physical disabilities or employment as a machinery dealer. A third owned a reaper as a member of a women's group and has an adult son who assists in the business. One woman respondent explained, \"I take all the decisions. Starting from appointing the driver, I take care of which field, at what price would be worked upon in addition to the amount of oil required and what to do if the machine gets damaged.\" However, she notes that, \"Even though I do all the work, at the end of the day I have to explain all the expenditures, profits and income to my husband.\" Though she knows people criticize her behind her back for her involvement, she brushes it off, saying \"the business brings me my income. No one else is going to give me this money without reason.\"While over 90% of reaper service providers registered by CSISA are men, their wives often assist in running the business, and men acknowledged their wives' contributions. A key informant estimated that there are \"30 percent [women] who actively participate in the service provision business though officially they are not recognized.\" Despite women's ownership of businesses, men typically supervise machine operators or drive reapers themselves, although women frequently help in advertising the reaper to other farmers through their social networks. Women also bring fuel for the machine and food and water for the reaper operator while in the field (especially if the field is next to the house). They also engage by cleaning and maintaining the machine when it is in storage. Some men taught their wives how to fix and repair reapers, and others expressed interest in their wives receiving this type of training to assist the business. Some wives also assist with accounting and keeping track of dues that are yet to be paid off by farmers.Many women also actively communicated with prospective clients when their husband is not at home. One man explained, \"If anyone comes to my house to rent the machine, but I am not present, they could easily go to another service provider, however, my wife makes sure that does not happen. She talks to them, which is why no renters go empty handed and…I get saved from facing any losses. This is extremely helpful for me as well.\"Husbands indicated that wives could further contribute to the business if they could collect money from clients and drive the reaper machines but these tasks are not socially acceptable. \"Getting payments from people in the market is also something they cannot do,\" explained one male service provider. Another explained, \"It would have been really good if she could drive the machine because then we would not have to hire drivers. However, since she is a woman, she cannot do it.\" A wife of a reaper service provider said, \"I wish I could drive, though, especially when my husband is resting.\"Men often noted women's inability to drive machines as a reason for why they were not suited to work in this business. In general, men clarified that they were not personally opposed to their wives driving the machine, but the reactions of others upon seeing a woman driving the machine would be humiliating. A male service provider noted:I do not have any problems if my wife or any other women drives machines in the field. But I think it would be better if they could be more focused on the accounting or taking care of the machines. This is mainly because if the women drive the machines, people would demean and mock them… people react in this way because tader chokhe bhalo lage na (society does not like seeing a woman driving a machine). However, if they drive the surrounding plots of land beside the house then there would not be any problem.Another male service provider explained:I do not think that a woman driving a machine is something negative. Women do not drive in our area (western part of the country) and there is no norm for them to do so. Women also expressed openness to operating reapers. The wife of a service provider shared, \"I would try, in order to expand our business more. If that period comes and the times change, then I would learn to drive, if I am taught in order to run the machine. However, in Bangladesh women do not drive machines.\"Effective supervision of machinery operators is a challenge for men and women service providers. Drivers need to be skilled to properly operate the reaper and avoid damaging the farmer's crop or the machine itself. Furthermore, when the operator interacts directly with the farmer, they must be trusted to collect and return the payment to the business owner. Supervising a machinery operator is however not socially unacceptable for women, at least compared to driving the machines: \"If I supervise the machines in the plot of land, the people in my village would not judge and demean me, but they would if I drive the machines\" (aspiring female service provider). However, some male reaper operators indicated that they would be unlikely to work for women machine owners, though both men and women are concerned that unsupervised operators will try to exploit them. A woman reaper service provider described, \"The operator does not give the earned money back properly. This mainly happens because I am a woman.\"Yet women perceived that their particular social relationships could create exceptions. A female service provider explained she \"might not get as many renters as a man. They might not value the female service providers as much as a male service provider. The people who would help me such as the driver might think that since I am a woman, they could do my work later.\" In contrast, she pointed out that:Operators in my village would be more understanding and they would say omuk apa bhabi fupur kaj kore di agey karon amar nijer gramer manush, onnoder gram e emon hobe na (the relationship between the people in the village, allows the LSPs to think of me as an aunt and therefore they would give importance to my work.) If it is another village, they would not value me as an LSP as much as in my village.Interview participants recognized that the problem of operating reapers not the only limitation to women's involvement in the service provider business. Men generally have closer relationships to those knowledgeable about agricultural machinery, such as NGOs, extension, research institute staff and machine dealers, and can ask them questions about the machine, participate in field demonstrations, and discuss the machine in person at the machine dealer shops or tea shops. Men in a focus group comprised entirely of machinery service providers in Jheneidah shared how they first heard about the machine by calling a number on an advertising billboard, watching a video at the tea stall, and a television show ('Krishi Dibanishi') that features agricultural activities.In contrast, women are restricted from spending time with strangers and in public places, thus reducing their chance to learn about new technologies and business opportunities. Women's mobility restrictions were often the first explanation men and women provided for why there are no female service providers. Nearly all respondents acknowledged that it is considered culturally shameful for women to interact in public. A wife of a reaper service in a slightly more conservative area acknowledged, \"There is a huge difference between a village and a town. If I get out of my house or go to the market, the people tend to point me out and say that, 'look at that woman, she is going outside'.\" One woman shared, \"I wanted to be involved in buying machineries since the very beginning. The women would be able to run a business if they want to but they would not be able to run the machine directly in the farm or go for work in other areas. How can a woman go to work in another place?\"Asked to identify the most important factors in learning about machinery, most men indicated interaction with extension agents, NGO representatives, or participating in field demonstrations. Most women conversely explained that they learned about machinery through their husband. In addition, interview respondents indicated that extension agents and NGO workers were more comfortable interacting with men than women, and hence target their assistance to men at the expense of women. One woman explained, \"The agricultural officers never talk about buying a machine under my name. They always called my husband to buy machines. If there is any meeting, only he is called. If the agriculture officers told my husband to buy a machine under my name, then my husband would have agreed.\"In addition, men's social networks of other male farmers allow them to publicize the availability of reaper services through word of mouth and in public spaces, including by displaying their machines at markets and providing field demonstrations in their own and neighboring villages. One man uploaded the reaper machine onto his Facebook page to advertise and sent SMS messages through his mobile phone to his contacts. Sometimes they will operate the machine for free on influential people's farms to increase public awareness and interest. One man noted, \"The way men can communicate, a woman cannot. It is not possible for the women to inform about their machine in public places, like a man does.\"Women and men recognize social judgment about women's contact with the public as a major constraint to running a business that requires significant interaction with strangers, often on an ad hoc basis in response to sudden demand from clients or urgent repair needs. A male respondent commented, \"if the woman suddenly needs something, she cannot run over to the market, such as for buying oil or spare-parts\". Another explained, \"It is not like the woman can go to the market to purchase oil. They would be heavily criticized by society if they are out and about in the market.\" A third respondent explained \"If someone wants to hire a woman's services, it is not possible for her to go immediately…but when it comes to men this issue does not arise.\" Women respondents also indicated this as a key constraint to taking on more responsibility or starting their own business.In addition to concerns about being judged for contact with the public, women regarded their lack of experience interacting with strangers not only as an obstacle to doing business but also as a factor limiting their confidence and knowledge. One woman shared her uncertainty about her own competence, tied to her limited exposure to the public. She explained that she and her husband have a mostly cooperative relationship where they \"discuss on all household issues.\" She is empowered to make some decisions for small issues, but \"important issues must be decided by my husband if we have a different opinion…such as marrying off my daughter…because as a woman I understand much less. My husband goes out, unlike me, and meets many people, therefore knows more about their attitudes, way of life and their characters, making him more sensible, with better instincts about people.\" Similarly, a woman described that because her son-in-law goes to the field and the market, \"…he has a good idea and experience about the world. According to me, visually seeing is the reality, but hearing is not reliable. In addition to the fact that he understands better.\" Without experiencing much of this world first hand, she felt uncertain about her judgment and ability to interact in this world. Similarly, another female respondent explained that while her husband listens to her while taking decisions, \"…he is more experienced than me and I might or might not understand certain decisions, yet he discusses it with me from a sense of responsibility, therefore usually he does what he thinks is right.\"Another woman expressed insecurity about how to act in public. Considering what it would be like if \"the business is in my name\", she responded that she would have to go to different places for activities like trainings, but \"it is difficult for me to go where crowds of people have formed. I usually do not go in these kinds of places but if I have to go, I am concerned about how to do it, how to talk, whether or not I can actually do this. However, I think once I learn, it will get better.\" \"Nonetheless,\" another woman noted, \"if I get started running the business, I would have to meet and work with new people, which would increase my knowledge about the outside world more.\"Women's expressions of insecurity about their competence may in part also be a strategy to avoid potential criticism about their arrogance. Women were very concerned not to appear to be displacing their husband in his role the primary income provider for the family. One man explained that \"society would criticize her for wanting to own a business, especially if the husband is present. The society would criticize her for thinking she understands more than her husband.\"When asked if women could be involved in the business in the future, a male service provider acknowledged that working in a business would be an important opportunity for women to develop new skills. He noted that lack of knowledge is not innate to being female, but rather a result of not having had exposure to these ideas. He explained:I do think women will come in future. Women involving them in the business can be good. They should not idly sit about. Involving themselves would allow them to increase their knowledge. If they practice, they would slowly understand. If we do not tell them or make them understand the issues, they obviously would not be able to do it. Now they can do it by themselves. If the women are given the knowledge, then they would also be able to do what a man does. If the women are given opportunity from the family and community, they would come.In most cases men made the decision to invest in a reaper, sometimes only informing their wives afterwards. But when asked to reflect on who is the owner of the machine, most men said that they were the joint owner of the machine with their wife, one male service provider explained that \"I consider us both to be the owners of the machines. This is because we are both part of this family and therefore the machines cannot just belong to one individual. Apart from this, we both take decisions regarding the income-expenditure and any other factors related to our family and business.\"Respondents were asked to reflect on whether women could be joint or individual owners of the machine and run the business. Many men thought their wives capable of running the business, but only after men had established initial relationships with farmers and operators. Most men and women did not expect that women's ownership would affect husband and wife's respective roles in the business (\"whether it is in her name or mine, it is the same\"). Most also did not perceive this would affect their day-to-day relationships. However, some were concerned with the effect ownership could have on inheritance. One male service provider explained:Personally, I do not see any advantages or disadvantages in buying a machine in my wife's name, but I do not think there is any risk if I end up doing so. We already have two children, so I do not worry anymore. Many men can be anxious about the fact that if there are properties in their wife's name, then their daughters would get more share than their sons. Maybe this is one of the reasons that the men who do not want to buy machines in their wife's name, especially those men who just have daughters. They worry that all their properties would be inherited by their daughters, when they are not there anymore. However, if there is one son and one daughter, these men might feel differently.Because this respondent has both a son and a daughter, he is comfortable with the ownership of the machine being in his wife's name. According to Islamic law, sons inherit more property than daughters if the property belongs to the father, but daughters inherit more if it is in the mother's name. If he only had a daughter, there would be a risk of shifting the property to another location and family, since after marriage girls usually go to a husband's home, often in a different village. Since he has a son, most of the assets and property will remain in this man's family, so they do not need to worry about property in his wife's name.One man articulated the value to his relationship of increasing his wife's ownership: \"Moreover, my wife also has certain wishes and wants. She might think, ami eto kichu kori, tumi amar jonno ki korcho (contemplate on the fact that she does so much for me, but exactly what do I do for her).Therefore, her being a legal owner might be a way to show that she is valuable as well. The mutual understanding between us really good.\" Another respondent explained that now that he and his wife are middle aged, there is no risk if the machine is bought in his wife's name. Earlier in their marriage, there could have been a concern of his wife leaving him and taking this property with her. However, after many years of marriage and having children together, they have developed a mutual trust and cooperative relationship reducing fears of abandonment.Women were often interested in buying the machine jointly with their husband, parents-in-law, or other household member. Respondents noted how this would diffuse risk associated with the investment and help them run the business, especially in the most socially unacceptable tasks for women of operating the machine and collecting money from farmers. In addition, women suggested that their husband would not support them in a solo business and may even mock them for proposing to start her own business.One woman explained, \"Even in the future I am not interested in buying a machine in my name, it should be under my husband's. This is because if I buy it under my name, my husband will say, 'Since the machine is in your name, then you should give the money for the machine and you earn,'\" which she knew she would have difficulties doing completely on her own. Similarly, another woman in Jheneidah district worried that her husband would say, \"You own it so you should run it.\" Though men might be the owners and expect women's support, women sensed that if they proposed independent ownership their husbands would challenge them to do it on their own, rather than help them, perhaps to emphasize the futility of the idea.In a focus group discussion in Jheneidah with women in households that hire reaper services, women described how they and their families would face scorn if they were individual owners of the machine: \"…people will say 'the wife has been keeping her husband under her feet'\" and \"People would laugh and say 'See! The machine is owned by the wife!'\"In contrast, another woman affirmed, \"I do not listen to what other people say, however as long as my husband does not say it, everything is fine. If I have money, anyone who talks to me would think before they speak. My value would also increase in the society if I buy a machine in my name and start a business as a woman LSP.\" This woman described her and her husband as having a \"good understanding\" by saying:We never use loud voices against each other. He is concerned about me just the way I am concerned about him…When he comes home working the whole day and sees that I have not finished all my work, he never questions why I have not been able to finish, even though it is quite late, and tends to ask if I have eaten or not. I have an important and valuable position in the family and it will stay like this in the future and will not change. In fact, if I start the business, it would increase more.In addition to concerns about women's roles in reaper service business, men and women had broader reservations about women's involvement in economic activities. Men and women often pointed to fear of social scorn. In a focus group in Jheneidah, men said, \"We fear about what people will say about us. Our fear is, if the women start running a business their family status will decrease. People will not respect us.\"In general, men were cynical about women's ability to participate in economic activity, let alone lead a business venture. In a focus group discussion with male farmers who rented machinery, they shared comments such as: \"My wife would remain under my authority, I would not be under hers\"; \"If my wife starts earning money she would be acting like she is the leader;\" \"It is a rule that in every sector, women should be behind the men;\" and \"However, the women should not be too far behind, but definitely behind.They can never come leading.\"Disparagement against working women often focused on how her economic activity reflects on male household members' ability to provide for the family. A female respondent explained:The society would criticize my husband saying…, 'Why you are living for? Your wife took over the responsibility and started operating machine in the field, it will be bringing huge income now, you should die!' It means, according to them, my husband is not man enough to handle his work, so his wife is taking over, who would earn more, and he should die out of shame. All these sarcastic judgmental comments would be hurtful. People would treat me like a laughing stock and ask my husband that maybe I should work in their farms as well to mock him.A female respondent from a mechanized farming household explained that her mother-in-law criticizes her poultry business, saying \"Men should earn and women should spend what earning comes from the men, women should not think about this income issue, it's men's job not women's.\" Sons may also feel ashamed by their mother working. A wife in Jheneidah explained that while she used to work in the field when her sons were young, she does not work in the farm anymore because her sons forbid her. She explained, \"If I go, then people would criticize me by saying 'What a shame that the mother is working in the farm, while she has two adult sons!'\" This same woman has a homestead poultry farm but gives her earnings to her husband, who she said tells her, \"You give all the money to me. If you need it, I will give you. It is not good for women to keep money with them.\" She explained that \"he is very kaachal parra lok (unnecessarily argumentative), therefore I give all the money to him in order to avoid any kind of dispute.\" At the same time, she was secretly saving money, presumably from these earnings, in an Akti Bari Akti Khamar (One House One Farm) group, despite her husband saying it is \"forbidden\" to generate interest.Some men and women suggested that familial support could override social contempt towards women's work. Women expressed that if they have their husband's permission and earn money, social norms such as purdah would not be an obstacle, and the increased income would raise their status and earn them respect. Many men agreed. A male service provider in Faridpur explained that women could say, \"This is my work, so I have to do it. No matter what others say. Those people, who talk, do not give me food to eat. If the husband supports and helps their wives, what others say should not matter.\" Men in a focus group discussion in Jheneidah pointed out that women will not become service providers because \"they have no interest\" and \"family does not support the woman to do this business. Family and society is in fact the main barrier for women to become women service providers or start the business.\" Wives of service providers in Faridpur explained, \"The reason for no women service providers is the husband would not let them conduct any business. They [men] believe that mohilara porda poribeshe khabar dibe (a women should remain veiled and serve food only\"). It the husband's job to provide food for them. In these cases, even if the family is supportive, however the husband does not give his consent, then it is his wishes which gets priority.\"Another wife of a service provider proposed that in the future, more education in the village would \"encourage more women\" to see that \"it is not wrong to work to provide food and shelter for the family. …Even if their husband is present, they would be able to do everything (including business).\" Her quote illustrates the perception that the husband decides whether she can start an enterprise or not. It also reflects a common view held by women that women without husbands or male family members, or from families where the men are incapacitated, can exercise relative autonomy in their economic activities (\"those women can work\").One woman, whose husband remarried and lives with the other family, shows how separated or single women can have more autonomy and might even be admired for starting a business. She explained:It would increase my value in the family and the society as a woman, if I could start this business.The people would in fact say, 'See, being a woman, she has bought a machine!' In other words, people of the society would be at awe and be surprised that being a woman, I have bought a machine. In addition, my sons do not argue or speak over in terms of decision-making. In fact, for the service provider business… my sons would be dancing with joy if I bought a machine for the service provision business.Other women echoed this idea when they suggested that even in the future, women without husbands were most likely to be able to become business owners. One wife of a service providers shared:I think the women who do not take any orders from their husband, and live according to their wishes, independently, they can do anything in life, and those women are the ones who can run their own business…In the future, if the environment in the village gets better…there will be a group of women would be able to run a business, who have no husband, since if they have a husband, he would not let his wife work.Most families hiring reaper services appear to be switching from employing wage laborers to harvest wheat or rice paddy to machinery services. Respondents describe that a shortage of wage laborers and increased rural labor costs during the peak season prompted them to hire the reaper services as an alternative to manual harvesting . Identifying, negotiating with, contracting, and supervising hired laborers is a source of stress and uncertainty for farmers, given the time-sensitive nature of the harvest.Moreover, harvesting time for wheat and aman rice corresponds with a period of unpredictable weather and heavy rains, so the ability to complete the reaping in a shorter period of time (for example, one day versus five days) with the machine is another advantage over laborers. Farmers also indicated that the speed of reaping with also allows farmers to proceed more quickly in preparing fields for the next crop grown in rotation. In addition, reaping with a machine is considerably less costly than hiring wage laborers, and both men and women appreciate greater flexibility in paying for the reaper services after the market day (hat bazars) or in installments instead of the immediate pay required by wage laborers. As with the service provider business, most of the farmers hiring reaper services are men. However, women in these households note that they experience some specific positive changes from mechanization, particularly in their workload and their involvement in decision-making. Men also recognize these benefits for women.Both men and women pointed out that one of the main benefits for women from hiring a reaper is that it frees them from the responsibility of providing meals and accommodation to the laborers over the days or weeks that hired harvesting laborers work. Women describe how they have to make food that laborers prefer, provide cigarettes and betel leaf, and \"keep them happy,\" or else \"endure their anger\" if the food does not taste good. Further exploration of these issues in focus groups indicated that because of reapers, women feel that they do not need to go to the field to bring the laborers food, which in turn increases their feeling of social status. Women also indicated they were pleased that their husbands do not have to work alongside the laborers in the field. Some women can even coordinate with the reaper operator, freeing the husband to take part in other economic activities during this time. If women in poorer families are able to afford machinery services, they strongly value the time, labor, and cost savings from mechanization.These women appreciated the respite from heavy physical labor, and say they spend time resting, with family, participating in religious activities, or pursuing other economic activities.These findings were backed by telephone survey results that indicated significant reductions in women's time spent cooking for harvesting laborers (a mean reduction of 2.45 and 1.63 hours/day during rice and wheat harvesting periods, respectively) harvesting, bundling, carrying, threshing and bagging activities (reduction of 3.04 and 3.30 hours/day for rice and wheat), and also supervision of laborers (reduction of 0.35 (rice) and 0.33 (wheat) hours/day) (Figure 3). Reductions these activities resulted in significant increases in time allocated to child care (a mean increase of 1.13 and 1.10 hours/day during rice and wheat harvesting periods, respectively) cooking for family members, (increase of 0.38 (rice) and 0.37 (wheat) hours/day), livestock and poultry arearing associated with income generation (increase of 0.64 (rice) and 0.60 (wheat) hours/day), leisure time (increase of 2.29 (rice) and 1.66 (wheat) hours/day), and sleeping (increase of 1.30 (rice) and 1.00 (wheat) hours/day). Other income generating activities were significantly different yet minor for wheat only (increase of 0.25 hours/day).Women in focus groups echoed these findings and also reported reductions in post-harvest labor from using machine services for the harvest. Both men and women in focus groups noted that harvested wheat is 'cleaner' when reaped with a machine than with laborers; this saves women the chore of cleaning the wheat of soil and other particles before mechanical threshing. In addition, the reaper can harvest the entire plot in a day, so that women can \"..thresh and beat all of them [paddy crop] together, therefore finishing all the work together at once\" rather than over the five days that laborers would take to complete the harvest. Some women, however, still need to provide labor for bundling the reaped crops and carry them to the house if they are not able to hire a wage laborer to do so.Some women also used to work alongside the laborers to harvest, bundle and carry rice or wheat from the field to the compound for threshing. Given strong cultural opprobrium toward women working on the farm, women who labor in the field are usually from lower income families, though not as poor as women who work as laborers on others' plots of land. To hide the fact that they needed to work in the field, two women in focus groups whose families lease and sharecrop land to farm, said that they used to help with the harvest by cover of night, \"…since if I cut them in the morning the society would heavily criticize me by saying that being a mother, a daughter and a wife, how I could cut the crops in the farm.\"Other women pointed out it is acceptable for them to work in the field if the land is next to the house, but not if it is far away. Women were strongly aware of the scorn towards women who work in the field, but it was rare that women internalized this notion, and instead characterized it as a rule that they had to comply with, whether they agreed or not, for the sake of their families' reputation. Many women echoed the words of a woman in Jheneidah, who said, \"I personally think, there is nothing wrong working in the field, but my thought cannot change anything until the society thinks it is acceptable.\"In focus groups, several men and women mentioned that after hiring reaper services, women's involvement in intrahousehold decisions had increased. One man pointed out that his wife had become more involved in agricultural decisions. Previously, \"she had no idea about how many wage laborers to appoint or when to appoint,\" but now she \"communicates with the service provider and keeps information about this. Even though she has not had any direct role in terms of selling, but the usage of the money … is part of her role now as well.\" Similarly, several women noted that the savings generated from switching to the reaper from laborers provided an opening for greater joint decision-making within the household.One woman explained:If I say something to my husband he listens to me now. Previously he would not, since in poor families there is no time to listen... The crops that were produced were used for our food, therefore leaving no crops to sell, so why would he listen to me? The money that gets saved[now] can now be used for other expenses such as my children's education or the household and that is my decision. I do not have to tell my husband.A man from a different mechanized farming household echoed this pattern, saying \"After starting to use the reaper, I and my wife started to take decisions in terms of the family income usage. I never used to get time before to discuss anything properly, however using reaper saves time and therefore I can discuss with my family where and when money needs to be spent.\" Changes in decision making are far from an inevitable outcome from mechanization, and many other variables are likely at play. However, it is clear that for some families, the time and cost savings from shifting to the machine reduces stress in the family and raises household status. In addition, women take on a new role with machine services, helping to supervise the reaper machine rather than serving the laborers.Most service providers interviewed mentioned that they do have a few female clients. They pointed out that these women are all from families where male family members are absent-husbands have migrated, passed away, are sick or incapacitated, or are occupied with non-farm employment. These women are already more directly involved in agricultural activities and often earn lower incomes. Wives of service providers in an focus group conducted in Faridpur explained, \"All those women whose husbands do not really understand [agriculture], those women do not really have any option, other than come as customers.\" As such, most service providers did not consider it likely that more women would hire machines in the future. They explained that there is only a small group of women who are both involved in agriculture and able to afford machine services. Women who have agricultural land and whose husbands are absent often lease or sharecrop their land to other farmers rather than work it themselves.Poorer women with little or no landholdings, on the other hand, either harvest themselves without hiring laborers or machines or, if they have no land, work as laborers on others' farms.It seems that for the women who do hire machine services, they are already non-compliant with dominant gender norms, and thus are comfortable speaking to men or women service providers. Wives of service providers in Faridpur explained that women who seek reaper services will speak to men or women, noting, \"This cannot be considered as something which is bad, since the work needs to be done first.\" If they keep purdah they would only talk to women in the service provider household, but this is rare as it is nearly impossible to keep strict purdah if a woman needs to provide for her family.Whether women farmers are more comfortable renting from women service providers remains to be seen, since it is still rare for women to take on either role. But there is some indication that women farmers are more comfortable seeking out advice from women service providers. In one case, a prominent woman service provider described how women, including very poor women, come to consult with her.\"Six to seven women came in, saying that they wanted to cultivate wheat, and asked my opinion on what they should do. That is when I rented the seeder and tiller machine both out to them.\" About a quarter of her clients are women, higher than for other male service providers. She said that these women had learned about the reaper service from their involvement in a Boithok (training session) called Krishok Math (farmers field). She also indicated that at times she provides services for free to poor women.Most of the women interviewed had husbands, sons, or sons-in-law who facilitated renting the reaper, but in a few cases, women were interviewed who had sought out machine services themselves because they were widowed or had a husband who \"seems to live in another world.\" A widow with a small son who leases land first heard about the reaper from her neighbor who works as a service provider and as an agricultural officer. Sometimes she pays in cash directly, but she also works as a wage laborer in his field as payment. For her, the time savings of using the reaper is valuable so that \"other works [economic activities] do not get hampered,\" and she pursues other jobs like weeding, tilling, road construction, and working in other people's houses.Though there is a readily apparent gender gap in the formal ownership of reaper service businesses and hiring of reaper machines in study locations within Bangladesh, women play key roles in both ownership and hiring dynamics, and they relay generally positive impacts of harvesting technologies on their lives and livelihoods. Many women in male-headed households contribute to activities essential for running service provider business and completing harvests, and they perceive this for the most part as within the bounds of social acceptability. When women are involved in either service provision business activities or hiring of machine services, it appears to increase their self-esteem and their husband's respect for them.Women's involvement with machinery, only visible upon closer examination, problematizes the idea that they are completely excluded from machinery service markets.In reaper service provider business, women are involved, in some way, in many activities. The major exception is in operating the machine, which is widely considered unacceptable for women given cultural norms in Bangladesh. However, many male service providers supervise machine operators, rather than drive the machines themselves, so if women do not drive the machines it is not a disqualifying factor from their involvement. Indeed, it is a challenge for both men and women to effectively supervise the driver and ensure he operates the machine properly. Many male service providers, having witnessed the contributions of their wives, believe that with proper training and support, women are capable of running the business as well as men. Men who had worked as service providers, and thus understood the work and had benefited from their wives' involvement, were more enthusiastic about opportunities for women in this business, compared to men who had rented machines but never run the business.For women who hire reaping services, they strongly value the time and cost savings. The former frees time for increased child and family care, income generating activities, leisure, and sleep during harvesting periods. In addition, they imply an increased sense of social status when women switch from serving wage laborers and meeting all their needs by cooking and serving them during harvest to supervising and assisting a machine operator. Some women and men noted there are new opportunities for women to participate in agricultural and income decisions with their husbands now that they have slightly more time and money.At the same time, it is clear that women do not have the same opportunities to participate in this market as men do. Above all, they are constrained by strong social disapproval of women's mobility and involvement in economic activities. Women's restricted mobility not only reduces women's chances of learning about new technologies and business opportunities-men were much more likely to initially hear about reapers through spending time in the market and through their contacts with extension officers, machine dealers, and other farmers-it makes it difficult to run a business that relies on interaction with male strangers to advertise and coordinate machine services. Women also strongly noted that their lack of exposure to the public -common given cultural hegemony and purdah in Bangladesh -undermines their confidence in general knowledge of the world and their ability to make sound decisions. This was commonly cited by women as a reason they would not begin a reaper service provision business. While some women and men believed that women's work is unacceptable on religious grounds, others were not personally opposed, but noted that women's economic activity tends to reflect negatively on male family members -husbands, fathers, fathers-in-law, and even sons within a village. Women across social classes weigh to what extent to limit or hide their economic involvement to avoid judgment.While few women run a service provider businesses on their own, some women from poorer households, who take on agricultural duties in the absence of male family members, are hiring reaper services. If women do not have male family members or are very poor, they work in agriculture out of necessity. Still, some try to hide their labor by working at night or only working if the field is near the house. This market segment of women, severely time and resource-constrained, is perhaps most directly benefited by this technology, as compared to women who are less involved in agriculture and come from wealthier or male-headed households. It is also more socially acceptable for women to rent machines if they do not have male family members, as these women are usually already parting with social norms given the necessity of their involvement in agriculture and economic activities. They do not seem concerned by social disapproval in renting the machines as the advantages are too great, but others acknowledge that it is the \"helpless\" women who tend to rent machines. Most of these women managed to learn about reaper technologies when they saw them being used in their village, rather than through their contacts with NGOs or extension workers or interaction in the marketplace. In the absence of male family members, these women have fewer mobility restrictions than those in male-headed households, but they are still not as likely to be aware of new agricultural technologies as men. In households with adult males, many men and women relay that they jointly own the machinery, and some men were open to the idea of women's formal ownership of a machine, as long as men continue doing most of the work. Some women perceive that their husbands will not be supportive if they start a business individually, implying that they would rather see them fail. For these reasons, most women preferred to have joint ownership over the machine and business, rather than individual in their name.Despite their lack of individual ownership over the machines, and less visible roles in providing and renting machine services, women manage to benefit in important ways through the current arrangements. Women are cautious about increasing their involvement through individual ownership or management of the business, because of the broader constraints they face to mobility and interacting with strangers and social disapproval towards women in business. Against the current landscape of social norms, many women do not perceive that the possible rewards of business involvement are worth the social risks, although some husbands are supportive of ownership in their wife's name as long as the work is still jointly managed. Women point out that even if their expanded role is somewhat socially nonconforming, they are willing to do it as long as their husbands are supportive. When they believe that their husband would not support them, they are not however interested in attempting business on their own. On the machine rental side, focus group results suggested that the women who benefit the most from mechanization are also likely to be those who are least likely to have access to the machines, given fewer financial resources and more restricted exposure to new information about the machines. Some pay for the services with their own labor, indicating that they may still value saving this labor if only as a way to substitute it for more remunerative work.The results presented in this paper complicate conventional wisdom about how to leverage new agricultural technologies for women's empowerment. Machinery ownership is not necessary for farmers to benefit from machinery use in their own fields. This is clearly evident given Bangladesh's and South Asia's large fee-for-service machinery market. There are, however, opportunities on the service provider side to expand women's ownership over machines, especially as service providers consider purchasing additional machines and have experience working with their wives running the business. Though women's ownership in this context rarely conveys women's control over the asset or even the earnings, focus group respondents clarified that legal ownership of the asset in the wife's name can provide her with a safety net in case of separation or death of a husband. Importantly, some women say they would be proud to have a machine in their name. Our results also highlight the scope to build women's skills and confidence in the business, especially in advertising, accounting, and mechanics/maintenance roles.However, as with other agricultural technologies, women still lag behind men in learning about the technology, especially when they have mobility restrictions, more gender-segregated social networks, and less contact with machine dealers or extension officers. This restricts women's access both to renting the machines and to starting or participating more confidently in their husband's business. Joint learning opportunities for husbands and wives, as well as for women without adult male family members, could encourage rental of reaper machinery services. Reaching women through their networks-including savings groups and NGOs-would help equalize their access to information, which is currently skewed towards men. These groups could also be used to collectively hire reaper services for use in fields partially or wholly managed by women. In addition, donors or government programs could assist with reducing the direct and transactions costs associated with the rental of machinery services for femaleheaded households, who benefit substantially from these services, but note that the cost is high and sometimes even pay for these services with their own labor.The common narrative that women would adopt technology if constraints to access (financial, information, complementary resources) were alleviated appears to apply only to a limited extent to poor women renting machinery services. However, it is less relevant to women in male-headed households renting machines, or to women engaged in service provider businesses. Though men nearly always tend to be the owners and renters of machines, this does not mean that their wives do not contribute or benefit.There are therefore important opportunities for development initiatives to build on the women's roles that are currently socially acceptable, as initial entry points to expand respect for women's competence, strengthen their asset ownership, and widen the bounds of socially acceptable behavior and engagement in agriculture and remunerative activities. Distinguishing between ownership and the opportunities provided women's roles is an important place to start, and should be recognized in the design of appropriate public policy and rural development initiatives.","tokenCount":"12115","images":["-684228055_1_1.png","-684228055_1_2.png","-684228055_1_3.png","-684228055_10_1.png","-684228055_10_2.png","-684228055_34_1.png"],"tables":["-684228055_1_1.json","-684228055_2_1.json","-684228055_3_1.json","-684228055_4_1.json","-684228055_5_1.json","-684228055_6_1.json","-684228055_7_1.json","-684228055_8_1.json","-684228055_9_1.json","-684228055_10_1.json","-684228055_11_1.json","-684228055_12_1.json","-684228055_13_1.json","-684228055_14_1.json","-684228055_15_1.json","-684228055_16_1.json","-684228055_17_1.json","-684228055_18_1.json","-684228055_19_1.json","-684228055_20_1.json","-684228055_21_1.json","-684228055_22_1.json","-684228055_23_1.json","-684228055_24_1.json","-684228055_25_1.json","-684228055_26_1.json","-684228055_27_1.json","-684228055_28_1.json","-684228055_29_1.json","-684228055_30_1.json","-684228055_31_1.json","-684228055_32_1.json","-684228055_33_1.json","-684228055_34_1.json","-684228055_35_1.json","-684228055_36_1.json","-684228055_37_1.json","-684228055_38_1.json","-684228055_39_1.json","-684228055_40_1.json","-684228055_41_1.json","-684228055_42_1.json","-684228055_43_1.json","-684228055_44_1.json","-684228055_45_1.json","-684228055_46_1.json","-684228055_47_1.json","-684228055_48_1.json"]}

data/part_2/0355655600.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"32956562903f3b286a8799b850b34c89","source":"gardian_index","url":"https://dataverse.harvard.edu/api/access/datafile/:persistentId/?persistentId=doi:10.7910/DVN/28540/XZWBFG","description":"The Tanzania Social Accounting Matrix (SAM) for the year 2009 is a detailed representation of Tanzania's economy. It separates 58 activities and commodities; labor by different level of education levels; and households by rural/urban areas as well expenditure quintiles. Labor and household information is drawn from the most recent Tanzania Household Budget Survey. Finally, the SAM identifies government, investment and foreign accounts. It is therefore an ideal tool for economywide impact assessments, including SAM-based multiplier analysis and computable general equilibrium (CGE) modeling.","id":"228550853"},"keywords":[],"sieverID":"3b0bb4bd-8380-464e-be1d-1a2e612cd551","pagecount":"26","content":"This paper documents a Tanzania Social Accounting Matrix (SAM) for the year 2009. The national SAM is based on newly estimated supply-use tables, national accounts, state budgets, and balance of payments. The SAM reconciles these data using cross-entropy estimation techniques. The final SAM is a detailed representation of Tanzania's economy. It separates 58 activities and commodities; labor by different education levels; and households by rural/urban areas as well expenditure quintiles. Labor and household information is drawn from the most recent Tanzania Household Budget Survey. Finally, the SAM identifies government, investment and foreign accounts. It is therefore an ideal tool for economywide impact assessments, including SAM-based multiplier analysis and computable general equilibrium (CGE) modeling.This paper outlines the construction of a 2009 social accounting matrix (SAM) for Tanzania. A SAM is a consistent data framework that captures the information contained in the national income and product accounts and the supply-use table (SUT), as well as the monetary flows between institutions. A SAM is an ex-post accounting framework since, within its square matrix, total receipts must equal total payments for each account contained within the SAM. Since the required data is not drawn from a single source, information from various sources must be compiled and made consistent. This process is valuable since it helps identify inconsistencies among Tanzania's statistical sources. For example, there are invariably differences between the incomes and expenditures reported by households in Living Standards Surveys. SAMs are economy-wide databases which are used in conjunction with analytical techniques to strengthen the evidence underlying policy decisions.Section 2 reviews the general structure of SAMs and Section 3 presents the key features of the Tanzania SAM. The first step in constructing a SAM is compiling information from various sources into a SAM format or framework known as the 'prior SAM'. The construction of the prior SAM takes place in two stages. A 'macro SAM' is first constructed using aggregate information from national accounts and other macroeconomic databases. This SAM is then disaggregated across sectors, factors and households to derive a more detailed 'micro SAM'. Given the diversity of its data sources, the prior SAM is invariably inconsistent (i.e., there are inequalities between receipts and payments). Section 4 describes the data sources used to construct the prior SAM. Finally, Section 5 outlines the basic cross-entropy estimation approach used to reconcile the imbalances in the prior SAM.The second step in constructing a SAM is reconciling receipts and payments so that row and column totals are equal (i.e., 'balancing' the SAM). This is also done in two stages. The reliability of the various data sources is first assessed based on the observed inequalities between row and column accounts. The SAM is then balanced using cross-entropy econometrics. The crossentropy approach is described in Section 5 together with a description of the constraints imposed during the estimation procedure. The final section summarizes the details of the new Tanzania SAM.One way of depicting the economy is the circular flow diagram shown in Figure 1, which captures all transfers and real transactions between sectors and institutions. Production activities purchase land, labor and capital inputs from the factor markets, and intermediate inputs from commodity markets, and use these to produce goods and services. These are supplemented by imports (M) and then sold through commodity markets to households (C), the government (G), investors (I) and foreigners (E). In the circular flow diagram, each institution's expenditure becomes another institution's income. For example, household and government purchases of commodities provide the incomes producers need to continue the production process. Additional inter-institutional transfers, such as taxes and savings, ensure that the circular flow of incomes is closed. In other words, all income and expenditure flows are accounted for and there are no leakages from the system.A SAM is an economy-wide data framework usually representing the real economy of a country, as depicted in Figure 1.1 More technically, a SAM is a square matrix in which each account is represented by a row and column. Each cell shows the payment from the account of its column to the account of its row -the incomes of an account appear along its row, its expenditures along its column. The underlying principle of double-entry accounting requires that, for each account in the SAM, total revenue (row total) equals total expenditure (column total). Table 1 shows an aggregate SAM (with verbal explanations in place of numbers).The SAM distinguishes between 'activities' (the entities that carry out production) and 'commodities' (representing markets for goods and non-factor services). SAM flows are valued at producers' prices in the activity accounts and at market prices (including indirect commodity taxes and transactions costs) in the commodity accounts. The commodities are activity outputs, either exported or sold domestically, and imports. In the activity columns, payments are made to commodities (intermediate demand), and factors of production (value-added comprising of operating surplus and compensation of employees). In the commodity columns, payments are made to domestic activities, the rest of the world, and various tax accounts (for domestic and import taxes). This treatment provides the data needed to model imports as perfect or imperfect substitutes vis-a-vis domestic production.   The government is disaggregated into a core government account and different tax collection accounts, one for each tax type. This disaggregation is necessary since otherwise the economic interpretation of some payments is often ambiguous. In the SAM, direct payments between the government and households are reserved for transfers. Finally, payments from the government to factors (for the labor services provided by public sector employees) are captured in the government services activity. Government consumption demand is a purchase of the output from the government services activity, which in turn, pays labor.The SAM contains a number of factors of production, which earn incomes from their use in the production process, and then pay their incomes to households, government and the rest of the world. Indirect capital earnings are taxed according to average corporate tax rates and some profits may be repatriated abroad. The remaining capital earnings, together with other factors' earnings (e.g. land and labor) are paid to households. Households use their incomes to pay taxes, save, and consume domestically produced and imported commodities.The national SAM of Tanzania identifies 58 sectors, of which 26 are in agriculture (see Table 2). Agricultural production is divided into crop agriculture (21 subsectors), livestock (3), fishery and forestry. Another main sectors identified in the SAM are in industry, which is separated into mining, manufacturing (18), utilities (2) and construction. Finally, the SAM also contains information on 10 different service sectors, including private services (7 subsectors) and public or government services (3).Table 3 shows the structure sectoral of gross domestic product (GDP). Agriculture accounts for 27.7 percent of total GDP in Tanzania, most of which is generated by crop agriculture, particularly maize and rice. The country exports some of its agriculture commodities, where coffee, cashewnuts, cotton and tobacco are the major ones. All these commodities comprise almost half of total agriculture exports. Industry in Tanzania is almost as big as agriculture sector (24 percent), where manufacturing and construction give the highest contribution. Half of total industry's exports come from mining, while the rest are mainly generated form manufacturing commodities, where process foods contributes the most. The services sectors hold major role in the economic activity, where it accounts for 48.2 percent of total GDP in the country. This includes services provided by both private as well as the public sector. Trade services accounts for almost half of total services provided by private sector, while most of services provided by public sector comes from government administration.  As mentioned above, one of the reasons for aggregating the sectors in the SAM was to expand the detail on factors' incomes and expenditures. Table 4 shows factor income shares within different sectors. The SAM differentiates between different kinds of factors, including labor, capital, agricultural land, and livestock stocks. Labor is further disaggregated by workers' education levels. \"Primary\" refers to workers with some primary schooling (grades 1-4); \"secondary\" includes workers with some secondary schooling (grades 5-11); and \"tertiary\" includes workers who have completed secondary school or higher education (12 or higher).A larger share of labor income in agriculture is earned by primary-educated workers compared to the national average. By contrast, very little labor value-added in agriculture comes from tertiaryeducated labor (only 0.2 percent). This reflects the general lower-skilled intensity of agriculture.The remaining agricultural value-added is earned by capital (19.8 percent) and agricultural land and livestock (44.7 percent).Industry, by contrast, is far more intensive in its use of higher skilled labor. For example, manufacturing labor value-added is mainly generated by secondary-and tertiary-educated workers, although the former dominates overall. However, while labor is still an important factor input into the manufacturing sector, it is capital that is responsible for most value-added generated in the industrial sectors. For example, capital in the mining sector accounts for 98.3 percent of total value-added. This reflects the higher capital-intensity typically associated with industrial production.Finally, services are the most intensive user of higher-skilled labor, with tertiary-educated workers generating 30.9 percent of total sectoral GDP. Moreover, labor is also a more important overall than capital. This is particularly true for the government sector (including health and education) where value-added is overwhelmingly generated by high-skilled workers. Identifying different factors, especially labor, is critical to capturing the effects of policy changes and external shocks on the distribution of household incomes. Table 5 summarizes how different households generate their incomes. For example, the table indicates that households as a whole in Tanzania earn 30 percent of their incomes from capital returns (i.e., mostly nonfarm enterprise profits). Most of their incomes come from labor wages and salaries, particularly from secondary and tertiary-educated labor. Land and livestock returns are also important, accounting for about 12.7 percent of total household incomes. However, it is the disaggregation of households into representative groups that is the main feature of the SAM. The new 2009 SAM separates households in the 2007 Tanzania Household Budget Survey across rural/urban areas and further disaggregates farm/nonfarm specifically for rural households (NBS 2011). Households are also disaggregated based on national per capita expenditure quintiles.Table 5 shows that lower-income households rely heavily on lower-skilled labor incomes. Capital is also less important for lower-income households. For example, while households in the top expenditure quintile receive more than one third of their income from capital, this accounts for only 2.3 percent of incomes for households in the lowest quintile. Farm households are amongst the poorer sections of Tanzania's population and this is reflected in the greater similarities between their income patterns and those of the lower quintile households. Finally, the SAM is an economywide data framework and so it captures not just how households earn their incomes, but also how they spend them. This completes the flow of incomes from production to incomes (via factor markets) to consumption (i.e., both the demand and supply of goods). Here the detailed information captured in the household breakdown reveals differences that can be important for assessing policies and shocks. For example, a much larger share of lowerincome households' consumption spending goes on agricultural goods and processed foods (73 percent) compared to higher-income households (46.3 percent). While the table shows only food versus non-food spending, the SAM contains detailed information on household spending on all 58 commodity groups. This information is crucial for assessing such external shocks as changes in world commodity prices.ion This section has provided some information on the key features of the SAM. It is focused on sector, factor and household disaggregation. However, the 2009 Tanzania SAM also contains detailed information on how government earn and spend their revenues. Similarly, the SAM maintains a consistent accounting of savings and their use to finance investment and changes in inventories or stocks. Compiling this comprehensive database necessitates drawing on a wide range of data sources, which are identified and discussed in the next section. The initial task in building a SAM involves compiling data from various sources into the SAM framework. For Tanzania, this information was drawn from national accounts, government or state budgets and balance of payments. This information often uses (1) different disaggregation of sectors, production factors, and socio-economic household groups, (2) different years and/or base-year prices, and (3) different data collection and compilation techniques. Consequently, the initial or prior SAM inevitably includes imbalances between row and column account totals.The macro SAM shown in Table 7 is an aggregation of the more detailed micro SAM. This section explains how each macro SAM entry is derived and disaggregated to arrive at the prior micro SAM. Each entry in the SAM is discussed below. The notation for SAM entries is (row, column) and the values are in billions of 2009 Tanzanian Shillings. The final disaggregated SAM is quite large and is included in the accompanying spreadsheet file.Total value-added or GDP at factor cost (Factors,Activities)... 25,503 This is the value of gross domestic product (GDP) at factor cost or alternatively, total value-added generated by labor, capital and land. Sectoral GDP for 19 sectors is drawn from national accounts (NBS 2011) and is further disaggregated across 58 sectors using GDP shares from the Economic Survey (MFEA 2010) for industry and services, and production quantity and price data for agriculture (NBS 2012). Value-added is then further divided into the returns to labor; capital; land and livestock capital using technical coefficients from the Input-output table.Labor income is split across educational groups: \"uneducated\" includes all workers without any formal school education; \"primary\" refers to workers with some primary schooling (grades 1-4); \"secondary\" includes workers with some secondary schooling (grades 5-11); and \"tertiary\" includes workers who have completed secondary school or higher education (12 or higher). Workers' incomes from wage and non-farm enterprises are drawn from the 2007 Tanzania Household Budget Survey. Capital is disaggregated into agricultural capital, livestock capital, land and nonagricultural capital. ii.This is the value of intermediate inputs used in the production process (i.e., the \"use\" matrix). The technical coefficients are based on the input-output table (NBS 1999). These coefficients are the share of inputs used per value unit of output.iii.This is the value of total marketed output (i.e., the \"supply\" matrix). Since all output is assumed to be supplied to markets, this value is equivalent to gross output, where gross output is the sum of intermediate demand and GDP at factor cost. Production of individual commodities was \"backed-out\" assuming the same technologies for all commodities produced by the same industry. This produces a diagonal supply matrix.This is the cost of trading and transporting goods from the farm or factory to domestic markets or to the border (in the case of exports), and vice versa (in the case of imports). These margins were taken from the earlier 2001 SAM (Thurlow and Wobst 2003).v.While the macro SAM shows only a single row and column for taxes, this account actually consists of a number of distinct tax accounts, including specific accounts for direct, indirect and trade taxes. The commodity tax entry can therefore be disaggregated to include indirect sales taxes (932) and import tariffs (1,770). These aggregate values of individual taxes were taken from government accounts (MFEA 2009). Aggregate tax revenues were disaggregated across commodities using information from earlier SAM (Thurlow and Wobst 2003), which then proportionally scaled so that the sum of individual tax collections matches the aggregate values reported in government accounts.vi.The value of total imports of goods and services was initially taken from national accounts (NBS, 2011). Goods and service imports were disaggregated using information from FAO data set (FAOSTAT 2011) and earlier SAM (Thurlow and Wobst 2003).vii.The payment from households to commodities is equal to household consumption of marketed production. The total level of private consumption of each commodity is based on national accounts (NBS 2011). However, we split this consumption value into consumption of marketed production and own production (see below). Households in the SAM are disaggregated by rural and urban areas, where rural is further disaggregated into farm and nonfarm households; and national per capita expenditure quintiles. All data disaggregation was based on information from the 2007 HBS (NBS 2011b). Consumption shares for each commodity were used to disaggregate consumption spending across the various household groups.viii. Own consumption demand (Activities Households)… 2,713 Some households who live in rural area produce commodities for their own consumption. The total value of this consumption is estimated based on information from the 2007 HBS (NBS 2011b) and national accounts (NBS 2011).ix.The total value of government consumption spending is taken from national accounts (NBS 2011). All of government spending is for the purchase of the government services commodity. In this way the government is treated as both a sector producing government services, and a demander of these services.x.The aggregate value of investment demand is taken from national accounts (NBS 2011) and disaggregated across commodities using information from the earlier SAM (Thurlow and Wobst 2003). Note that this aggregate value includes both public and private investments.x.The value of total exports of goods and services was taken from national accounts (NBS 2011). Goods and service exports were disaggregated using information from FAO data set (FAOSTAT 2011) and earlier SAM (Thurlow and Wobst 2003).xi.Households receive factors includes directly from labor, livestock and agricultural land. The total value of these receipts depends largely on the sectoral value-added composition. They are distributed to different representative household groups based on incomes reported in the 2007 HBS. Labor income distribution is based on reported wage receipts and half of reported farm/nonfarm enterprise earnings (assuming that the remaining half is returns to capital). Earnings from crop land and livestock capital are distributed based on the reported incomes from these separate farm enterprises in the HBS.xii.This is remittances paid to domestic households as reported in the balance of payments (MFEA 2009). This aggregate value was disaggregated across representative household groups using information on total household survey in the HBS (NBS 2011b).xx.Personal income taxes are paid by households to the government. The value of these taxes is taken from the government budget (MFEA 2009). This was disaggregated across representative household groups using information reported by households in 2007 HBS (NBS 2011b).xxi. Household private savings (Savings, Households)… 5,859In the absence of supporting data, household savings is treated as a residual balancing item after accounting for all incomes and expenditures.xxii. Foreign transfers to the government (Government, Rest of world)… 913Government income from the rest of the world is the value of \"foreign grants\" as reported in the balance of payment (MFEA 2009).xxiii. Public savings or recurrent fiscal balance (Savings, Government)… 148Government savings includes public investment and is treated as a residual balancing after accounting item for all government revenues and payments (MFEA 2009).This is the current account balance or the total value of foreign savings. It is treated as a residual balancing item after accounting for all foreign receipts and payments. The result is lower to the current account balance reported in the balance of payments (TSH 3,083 million).The range of datasets used to construct the prior micro SAM implies that there will inevitably be imbalances (i.e., row and column totals are unequal). Cross-entropy econometrics is used to reconcile SAM accounts (see Robinson et al., 2001). This approach begins with the construction of the prior SAM, which as explained in the previous section, used a variety of data from a number of sources of varying quality. This prior SAM provided the initial 'best guess' for the estimation procedure. Additional information is then brought to bear, including knowledge about aggregate values from national accounts and technology coefficients. A balanced SAM was then estimated by minimizing the entropy 'distance' measure between the final SAM and the initial unbalanced prior SAM, taking into account of all additional information.Balancing procedure for the SAMThe balancing procedure takes places in two stages. First, a very detailed national SAM was constructed using the supply-use table, national accounts, state budgets and balance of payments. At this stage, the SAM contains aggregate entries for factors and households. This aggregate national SAM was then balanced using cross-entropy.After balancing the national SAM, it was then disaggregated across factors and households. Since the aggregate national SAM is balanced, this results in imbalances for the household accounts only. These household accounts were again balanced using cross-entropy, but holding all other non-household-related entries of the national SAM constant. Given the imbalances in the household survey between incomes and expenditures, the target household income/expenditure total for the final balanced SAM was the expenditure totals in the unbalanced prior SAM.Table 8 presents the equations defining the SAM estimation procedure. Starting from an initial estimate of the SAM, additional information is imposed in the form of constraints on the estimation. Equation 1 specifies that row sums and corresponding column sums must be equal, which is the defining characteristic for a consistent set of SAM accounts. Equation 2 specifies that sub-accounts of the SAM must equal control totals, and that these totals are assumed to be measured with error (Equation 3). An example would be the estimate of GDP provided by national accounts, which is the total value of the Factor-Activity matrix in the prior SAM. The matrix G is an aggregator matrix, with entries equal to 0 or 1. The index k is general and can include individual cells, column/row sums, and any combination of cells such as macro aggregates. Equation 4 allows for the imposition of information about column coefficients in the SAM rather than cell values, also allowing for error (Equation 5). Error on each constraint k e AError on each cell coefficient W, W � Weights and prior on error term for each constraint k or cell coefficient i,jV � Error support set indexed over w for each constraint k or cell coefficient i,j Equations The error specification in Equations 2 and 3 describes the errors as a weighted sum of a specified 'support set' (the V parameters). The weights (W) are probabilities to be estimated, starting from a prior on the standard error of measurement of aggregates of flows (Equation 8) or coefficients (Equation 9). The number of elements in the error support set (w) determines how many moments of the error distribution are estimated. The probability weights must be non-negative and sum to one (Equations 8 and 9). The objective function is the cross-entropy distance between the estimated probability weights and their prior for the errors in both coefficients and aggregates of SAM flows. It can be shown that this minimand is uniquely appropriate, and that using any other minimand introduces unwarranted assumptions (or information) about the errors.Various constraints were imposed on the model according to the perceived reliability of the data.Certain values that appeared in the supply-use table and national accounts were maintained in order to remain consistent with the overall macro structure of the economy. The macroeconomic aggregates that were maintained in the micro-SAM include: total labor value-added; total capital value-added; household final demand; government spending; investment demand; exports; imports; government borrowing/saving; current account balance; sales taxes; import tariffs; direct taxes on enterprises; government transfers to enterprises; enterprise transfers to the rest of the world; enterprise transfers to government; household transfers to government; government transfers to the rest of the world; and household foreign transfers received. The same standard errors were applied to all representative household groups.","tokenCount":"3858","images":["228550853_5_1.png"],"tables":["228550853_1_1.json","228550853_2_1.json","228550853_3_1.json","228550853_4_1.json","228550853_5_1.json","228550853_6_1.json","228550853_7_1.json","228550853_8_1.json","228550853_9_1.json","228550853_10_1.json","228550853_11_1.json","228550853_12_1.json","228550853_13_1.json","228550853_14_1.json","228550853_15_1.json","228550853_16_1.json","228550853_17_1.json","228550853_18_1.json","228550853_19_1.json","228550853_20_1.json","228550853_21_1.json","228550853_22_1.json","228550853_23_1.json","228550853_24_1.json","228550853_25_1.json","228550853_26_1.json"]}

data/part_2/0356891213.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"2659090bf40e8dec5ee4d3015808c10d","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/f7eb483a-61d1-47fb-94cf-b73a46e2a8e0/retrieve","description":"The authors examine the lessons that rural financial institutions such as banks and cooperatives can learn from the informal lending sector. They also considers the roles government should play in the provision of financial services. The report’s findings are gleaned from a series of detailed household surveys conducted in nine countries of Asia and Africa: Bangladesh, Cameroon, China, Egypt, Ghana, Madagascar, Malawi, Nepal, and Pakistan. Most of the poor in these countries could benefit from credit, savings, and insurance services, but what is available varies greatly from country to country.","id":"419954002"},"keywords":[],"sieverID":"24d94598-b82c-476f-aabd-e3aee6cec1b7","pagecount":"2","content":"INT. FG0v POUCY n many developing countries, poor rural households face se~&~• cf>d~itJiliifQ)ten they seek credit from formal lending institutions. Formal financial services such as those offered by banks are often not available to those below the poverty line because of restrictions requiring that loans be backed by collateral. Nor do banks welcome the small amounts the poor want to save. As a result, the poor usually turn first to informal sources such as friends, relatives, or moneylenders, who loan small amounts for short periods, or to informal, indigenous institutions such as savings clubs and lending networks to borrow enough to purchase food and other basic necessities. These informal networks are frequently successful in tiding the poor over difficult times, such as a bad harvest, and they enable poor households to build up savings for investments that can help lift them out of poverty.A recent Food Policy Report, Rural Finance and Poverty Alleviation, by Manfred Zeller and Manohar Sharma, examines the lessons that rural financial institutions such as banks and cooperatives can learn from the informal lending sector. It also considers the roles government should play in the provision of financial services. The report's findings are gleaned from a series of detailed household surveys conducted in nine countries of Asia and Africa:n looking at the lessons that can be learned from studying the relationship between informal lenders and their poor clients, the report finds the following: ( 1) A credible longterm relationship is the key to enforcing loan repayment: the borrower will repay the loan if he or she expects to be able to borrow again in the future. (2) Financial services should be tailored to the demand patterns of the borrowers. For example, farm loans that can only be used for seeds or fertilizer reduce the flexibility of the household to make the best use of the loan. (3) Decisionmaking on loans granted should be made at the local level. ( 4) Institutions ought to have clear plans for loan recovery before lending begins. (5) Group-based transactions hold promise, but more research is needed to compare group lending and saving activities with other member-based institutions such as credit unions and village banks. (6) Saving services should be provided. (7) Incentives for managers of rural financial programs should be built into the programs.n recent years, micro finance institutions designed to serve the poor, such as the Gramecn Bank in Bangladesh, have received wide attention, but these institutions depend on subsidies from national governments and international donors. Zeller and Sharma argue that these subsidies represent good investments of public funds on two counts: they enable services to be offered that the marketplace is not willing to provide on its own, and they have been proven to alleviate poverty.Although excessive government interference and rigid regulations have suppressed innovation in financial services, liberalization of financial markets alone has not been able to trigger the kinds of innovation that reduce transaction costs for the poor. Rural financial markets in developing countries have inherent problems that make investments risky and costly: clients are too scattered, rural clients all want to borrow at the same time (in the preharvest season) and to save immediately after the harvest, and the poor own few assets to secure loans. Private-sector financial institutions are reluctant to take on these risks. In the long run, however, innovations that improve the usefulness of these institutions to the rural poor will also improve the efficiency and sustainability of rural financial programs.illce the market, by itself, has not U been able to stimulate much institutional research and experimentation in rural areas, public support in the development phase is critical. \"Once viable prototypes are identified, they will eventually be adopted by the private sector,\" the report says. \"Well-directed support to promising microfmance institutions is likely to have payoffs in both services to the poor and reduced costs of services in the long run.\"Access to credit or participation in a credit program positively affected household income in four out of five countries assessed, the report finds. Households with improved access to credit were also better able to adopt technology; they spent more on food and, in some cases, had higher calorie intakes. Access to financial services improves the incomes of and opportunities for the rural poor, and provides support to tide families over difficult times. And poor households strive to repay loans so that they will be able to borrow another time.But, for the poorest of the poor, the report indicates that financia l services must be offered in combination with other programs such as training in basic literacy, enterprise management, and education in nutrition, health, and family planning.The report makes a strong case for strengthening rural financial markets through appropriate public intervention. In the long run, public investment in institutional innovations will pay off in efficient microfinance institutions that offer full-fledged savings and credit services to the rural poor.__________________ ._,_,__ ___________ _ Please send the Food Policy Report Rural Finance and Poverty Alleviation, by Manfred Zeller and Manohar Sharma.Name --------------------------------------------------------------------------------Organization•---------------------------------------------------------------------------Address -------------------------------------------------------------------------------The report will be sent free of charge by surface airlift. Please allow 3-4 weeks for delivery.","tokenCount":"852","images":["419954002_1_1.png","419954002_2_1.png"],"tables":["419954002_1_1.json","419954002_2_1.json"]}

data/part_2/0371745114.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"227f9c15e4f94c9b9a4c6fe9e2b29147","source":"gardian_index","url":"https://dataverse.harvard.edu/api/access/datafile/:persistentId/?persistentId=doi:10.7910/DVN/LY2OFY/Y1MPYC","description":"This dataset is the result of the frontline health workers survey that was conducted to gather data at baseline within the context of an overall evaluation of the franchise model for Alive & Thrive (A&T) in Viet Nam. The overall aims of the evaluation were to assess the impact of the franchise model on (1) age-appropriate IYCF practices among children <2 years of age and (2) stunting among children 2-5 years of age.</p>\r\n\r\nA&T is an initiative funded by the Bill & Melinda Gates Foundation to reduce undernutrition and death caused by suboptimal IYCF practices in three countries (Viet Nam, Bangladesh, and Ethiopia) over a period of six years (2009-2014). The goal of A&T is to reduce avoidable death and disability due to suboptimal IYCF in the developing world by increasing exclusive breastfeeding (EBF) until 6 months of age and reducing the stunting of children under two years of age. A&T applied principles of social franchising within the government health system to deliver the interventions.</p>\r\n\r\nA&T’s Viet Nam strategy is designed to support improvements in infant and young child feeding (IYCF) in three key ways: (1) improving policy and regulatory environments; (2) shaping IYCF demand and practice; and (3) increasing supply, demand, and use of fortified complementary foods. In order to achieve this, the A&T Viet Nam program has been divided into three main focus areas namely advocacy, community, and the private sector. In addition, a communications component is integrated into each of these focus areas to support their activities.</p>\r\n\r\nAmong several activities, the franchise model is a core initiative of the community model to provide quality nutrition counseling to women and families at health facilities at all levels. Implemented in cooperation with the Vietnamese government and select private clinics, franchises will deliver a package of focused IYCF counseling services to pregnant women, lactating mothers, and their families, based on a franchise service package. Focused training and capacity building for healthcare workers will be undertaken to enable the health system to provide franchise services. Individualized services will be supported through mass media campaigns aimed at generating demand for franchise services and promoting optimal IYCF practices.</p>\r\n\r\nThe baseline survey was conducted in 40 communes across four provinces, Thai Nguyen, Thanh Hoa, Quang Ngai, and Vinh Long, between June and August 2010 by the IFPRI team in collaboration with the Institute of Social and Medicine Studies (ISMS). The survey included four components—(i) household survey, (ii) community survey, (iii) frontline health workers survey, and (iv) health facility assessments survey. \r\n\r\nThe commune health center staff survey data provide information on the staff knowledge and attitudes related to IYCF practices, the training they had previously received on nutrition, their IYCF-related activities and time commitment, and their job motivation and satisfaction.","id":"784317380"},"keywords":[],"sieverID":"b03ce4c1-8bf2-43b9-83e2-fafd5f0887d5","pagecount":"20","content":"We are from the Institute of Social and Medicine Studies (ISMS), and we are collaborating with Institute of Food and Policy Research Institute (IFPRI) to conduct an evaluation of nutrition services for young children in your community. This evaluation is being carried out in 40 communes and some Reproductive Health Center and Provincial Hospital, and your commune was one of those selected. The focus of this evaluation is to have general understanding about the commune as well as health services at the communes.We are inviting you to be a participant in this study. If you decide to participate, we will ask you a series of questions about knowledge, attitude and your actual work about child breastfeeding and complementary feeding status. The answers you give will help provide better information to policy-makers, practitioners and program managers so that they can plan for better services that will respond to your needs and your community.The information you will provide during the interview is strictly confidential, will only be available to the project investigators, and will not be provided to anyone else. You will only be identified through code numbers. Your identity will not be stored with other information we collect about you. Your responses will be assigned a code number, and the list connecting your name with this number will be kept in a locked room and will be destroyed once all the data has been collected and analyzed. We will use approximately 30-45 minutes of your time to collect all the information. All questions in this survey are supported by Ministry of Health and National Institute of Nutrition.There will be no cost to you other than your time. Your participation in this research is completely voluntary. You are free to withdraw your consent and discontinue participation in this study at any time. You also have the right to refuse to answer specific questions. There will be no risk as a result of your participating in the study. The researcher read to me orally the consent form and explained to me it meant and agree to take part in this research. I understand that I am free to discontinue participation at any time if I so choose, and that the investigator will gladly answer any question that arise during the course of the research. We would first like to talk to you to know about your perceptions related to feeding babies and young children. If you feel that a statement here does not apply to your job description, please tell us so.(Interviewer: IF the health staff says this does not apply, CIRCLE \"0\"). ","tokenCount":"430","images":[],"tables":["784317380_1_1.json","784317380_2_1.json","784317380_3_1.json","784317380_4_1.json","784317380_5_1.json","784317380_6_1.json","784317380_7_1.json","784317380_8_1.json","784317380_9_1.json","784317380_10_1.json","784317380_11_1.json","784317380_12_1.json","784317380_13_1.json","784317380_14_1.json","784317380_15_1.json","784317380_16_1.json","784317380_17_1.json","784317380_18_1.json","784317380_19_1.json","784317380_20_1.json"]}

data/part_2/0380121055.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"e6518d6f96c680d79bb9eecb6e32fb13","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/9cb8392a-d41e-4c69-8bbc-00cf863f747d/retrieve","description":"This study aims to understand the implications of stricter food safety regulations and certification systems to the food industry and to find ways to manage risks and costs associated with these regulations and systems. This paper empirically examines the timing of initial decisions to adopt food safety systems and subsequent decisions to maintain the certification. Survival models are used to evaluate firm-level decisions among seafood processors in the Philippines. Whereas initial certification decisions were influenced mainly by easily obtainable a priori indicators such as output price, scale of production, and association membership, decisions to continue certification were influenced by a larger number of less-visible factors including price differentials across markets and cost structures. Managerial hubris may have played a role in initial certification decisions, but decertification decisions were more informed by realized cost–benefit comparisons.","id":"-2109299831"},"keywords":["food safety","survival analysis","seafood industry JEL Classification: Q18","Q13","D22","L6"],"sieverID":"b95dea30-62cc-4cc7-b9b7-b04291d6fc8b","pagecount":"28","content":"established in 1975, provides evidence-based policy solutions to sustainably end hunger and malnutrition and reduce poverty. The Institute conducts research, communicates results, optimizes partnerships, and builds capacity to ensure sustainable food production, promote healthy food systems, improve markets and trade, transform agriculture, build resilience, and strengthen institutions and governance. Gender is considered in all of the Institute's work. IFPRI collaborates with partners around the world, including development implementers, public institutions, the private sector, and farmers' organizations, to ensure that local, national, regional, and global food policies are based on evidence. IFPRI is a member of the CGIAR Consortium.The importance of achieving and maintaining food safety has never been more apparent. High-profile outbreaks and rising consumer concerns increase pressure on public and private decision makers to identify and resolve systemic problems. Numerous studies have focused on the adoption of certification and management systems (privately or publicly managed) as a way to improve food safety performance (Hammoudi, Hoffmann, and Surry 2009;Menard and Valceschini 2005;Henson and Caswell 1999). A particular focus has been the Hazard Analysis and Critical Control Points (HACCP) system-one of the more widely used methods of food safety management, a process standard recommended by the Codex Alimentarius Commission of the Food and Agriculture Organization of the United Nations and the World Health Organization, and a required system in the European Union (EU).Prior literature has identified a number of capacity-and incentive-related factors that affect initial adoption of HACCP, including length of time to develop and implement the program, technical expertise and support, availability of human resources, production technology and design, company size, and level of institutional support (Fotopoulos and Kafetzopoulos 2011). In addition, institutional theorists have identified social acceptability, credibility, and legitimacy as important drivers of firms' choices to adopt HACCP (Meyer and Rowan 1977;Scott 2001;Bansal and Clelland 2004).Some studies have suggested that food producers and processors do not always embrace HACCP with the anticipated enthusiasm (Panisello, Quantick, and Knowles 1999;Taylor 2001;Panisello and Quantick 2001;Taylor and Taylor 2004;Ehiri, Morris, and McEwen 1995;Herath and Henson 2006). Further, there are indications of a lack of sustainability in the system; for example, more than 60 percent of certified firms in the seafood industry in the Philippines have recently discontinued EU HACCP certification. The lack of understanding about such decertifications is a major knowledge gap in food safety policy (Ragasa, Thornsbury, and Bernsten 2011).After initial certification, a firm's decision to continue with, or discontinue, the certification at any point in time will be influenced by the same set of incentive-and capacity-related factors, but now with additional information on realized costs and outcomes. 1 Firms with the necessary capacity to implement the program successfully and subsequently realize the expected net benefits from HACCP certification will continue with their programs, but firms that lack capacity or fail to realize the projected net benefits may seek decertification. Failure to realize anticipated benefits may be the result of ex ante unrealistically optimistic projections (that is, managerial hubris) or subsequent changes in markets and external conditions. Literature in strategic management identifies managerial hubris as a major cause of adverse firm performance, especially in explaining the failure of major strategic moves, such as mergers and acquisitions, or decreases in firm profitability (Roll 1986;Jiang et al. 2011). Other studies suggest that firms often misestimate costs of regulations including those for food safety (Gray and Shadbegian 1993;Joshi, Krishnan, and Lave 2001;Morgenstern, Pizer, and Shih 2001;Ragasa, Thornsbury, and Joshi 2011).Prior to certification, the actual costs and benefits to a firm are not known. Uncertainty can be reduced through ex ante information collection and assessment but cannot be entirely eliminated. Once certification is undertaken and outcomes are realized, firms have additional information and investments that will influence subsequent decisions about sustainability. The difference between anticipated and realized outcomes is known only over time and will vary across firms.Survival analysis is used to assess longevity or rate of survival in relation to an event of interest where there are subjects or observations who do and do not experience that event of interest at alternative points in time (Menard 2008;Singer andWillet 1993, 2003). Survival (or duration) models have been widely used in epidemiology and medical research to explain occurrence of and survival from a disease (for example, Kurian, Sigal, and Plevritis 2009;Song and Lawson 2009;Spitale et al. 2009;Madan et al. 2008). Application to economic problems is less frequent but targeted to issues where duration is the focus-that is, exit or survival of firms (Olmos 2010;Tiller, Feleke, and Starnes 2009;Dimara et al. 2008;Soderbom, Teal, and Harding 2006); rate of technology adoption (Abdulai and Huffman 2005); rate of contract termination (Olmos 2010); length of visitor stay (Barros and Machado 2010;Barros, Butler, and Correia 2010;Gokovali, Bahar, and Kozak 2007); timing of loan default (Roszbach 2004); infrastructure failure (Debon, Carrion, and Solano 2010); and employee retention (Mattox II and Jinkerson 2005).In the case of food safety, a subset of firms initially adopt EU HACCP systems, some of which maintain their certification and some of which subsequently decertify, which makes survival analysis an appropriate tool to evaluate the time-path of food safety certification. Earlier studies do not explicitly address the effect of factors on the time-path of adoption or certification (Abdulai and Huffman 2005). Although some prior studies have analyzed factors affecting food safety controls, including HACCP, using discrete choice models such as logit and probit (Herath, Hassan, and Henson 2007;Henson and Holt 2000), to the best of our knowledge, this study is the first to apply survival or hazard functions to food safety certification decisions. We hypothesize that the time duration before a firm gets certified depends on the strength of incentive (that is, expected net benefits) and capacity factors (that is, firm capacity for implementation or early adoption). We hypothesize that conditional on initial certification, the period for which a firm remains certified (that is, does not get decertified) is positively associated with the strength of incentive and capacity factors. We expect the ex post decisions over decertification to be more strongly associated with incentive and capacity than the ex ante decisions about initial certification, because decertification decisions are likely based on known information in contrast to projections and managerial hubris.Firm-level data were collected from 59 seafood processors located in the Luzon, Visayas, and Mindanao areas in the Philippines.2 Face-to-face interviews using a semistructured questionnaire were conducted in September to December 2005 with plant managers to collect data covering the period 1998 to 2005 of their operations. 3 The sample includes a variety of firm sizes (15 cottage scale, 14 very small, 6 small, 15 medium, and 9 large) and product categories (17 frozen tuna processors, 10 milkfish, 9 shrimp, 8 canned tuna, and 15 other products). 4 Respondents included 41 firms that initially received EU HACCP certification, 15 of which were subsequently decertified. 5 The remaining 18 firms in the sample never received EU HACCP certification. 6 The timing of initial adoption and the duration of certification varied across the sample firms (Table 2.1). Table 2.2 shows incentive-and capacity-related explanatory variables. Incentives include expected or perceived net benefits, both financial and nonfinancial, from adopting HACCP standards. Financial variables include average price received for products sold (output price level, or Price) and difference in export price between EU and United States (Pdeuus), and between EU and other markets (Pdeuoth). Estimated prices for labor (PL), materials (PM), and capital (PK) are used as indicators of firm-level input cost structure. Since higher revenues increase the likelihood of higher benefits from certification, we hypothesize that both the output price level and output price differential variables will be positively associated with initial certification but negatively associated with the decertification decision. As HACCP certification involves additional capital and operational investments and higher input prices adversely affect incremental benefits from certification, we expect input prices (PL, PM, and PK) to be negatively associated with initial certification decisions but positively associated with subsequent decertification decisions.  Difference in export price between the EU and the United States as received by the processors. Price measured as value of exports divided by the volume (US$/kg).Difference in export price between the EU and other markets as received by the processors. Other markets include the local Philippine markets and other importing countries (aside from European Union and the United States). Price measured as value of exports and sales divided by the volume (US$/kg).These are location-specific wage levels. This represents three-fourths of the minimum wage rate by region (US$/day). The adjusted minimum wage reflects the variability of costs of goods across regions. Location can reflect the differences in the cost of factors of production including labor and other inputs needed for HACCP. aWeighted average of the prices of fish/seafood raw materials based on firmlevel interviews and financial and income statements from the Securities and Exchange Commission (US$/kg).Weighted average of the tax-adjusted interest and dividend (as a ratio of the long-term debt and equity) plus the weighted economic depreciation rate of fixed assets (US$/unit of capital). b Market diversification nmarket Number of markets, in terms of country of product destination, the firm has.Number of distinct product forms in the production line. For instance, frozen shrimp is distinct from frozen octopus or frozen deboned milkfish.Access to capital credit (=0,1)Dummies for reported difficulty in accessing credit by firms (1 = difficult; 0 = not difficult).Number of years in business, which represents length of experience in terms of dealing with buyers and negotiating with input suppliers; also represents the period to develop trust and long-term relationship with buyers.member Number of producer, processor, or exporter associations of which the respondents are a member.active (0,1)Dummy representing active membership of respondents in associations. \"Being active\" means actively joining the activities of the associations and/or actively seeking advice, assistance, or support from such associations.Age of plant facility.Size of production Y Volume of production (in metric tons).Source: Authors' compilation from various sources. Notes a Minimum wages are not actually followed in the Philippines, as in most developing countries based on the personal communication with J. Price Gittinger, November 15, 2006, World Bank, Washington, DC. The minimum wage rate is from the Philippine Department of Labor and Employment and the National Statistics Office. b Rates of economic depreciation per asset class adopted from Hulten and Wykoff (1981).Nonfinancial incentives include diversification, measured by both the number of markets in which the firm sells in terms of country of product destination (nmarket) and the number of distinct product forms that the firm processes in its production line (nform), and volume of output (Y) as a measure of size. Greater diversification, both in terms of geographical area and the number of products, gives the firm flexibility to continue business even without HACCP certification, thereby reducing its incentives to seek HACCP certification. Hence, we hypothesize that both market diversification (nmarket) and product diversification (Nproduct) will be negatively associated with initial certification but positively associated with decertification decisions. Because of significant fixed costs associated with HACCP certification, we expect economies of scale in realizing benefits from certification and hypothesize a positive association between volume of production and initial certification, and a negative association with subsequent decertification decisions.Capacity variables focus on the financial, human, and institutional resources to set up and manage HACCP systems.7 A dummy variable credit represents difficulty in accessing credit in addition to the more direct measure of price of capital (PK). Number of years in business (yrbus) is a proxy for operational experience and human capital. We hypothesize that a firm with higher difficulty in accessing capital, higher price of capital, and fewer years in business (that is, low operational capacity) has a lower probability of achieving initial HACCP certification and a higher probability of being subsequently decertified. We include age of the plant (age) in our estimations, and as older plants are likely to need higher investments in upgrading to meet HACCP standards, we hypothesize a negative association with the initial certification decision.Variables used as measures of institutional support include the number of association memberships (member) and whether the firm was actively involved in such associations (active). These variables can be viewed either as proxy measures for the degree of external institutional pressures faced by the firm or as additional firm capacity because active membership provides the firm with access to resources, know-how, and potentially experienced manpower. We hypothesize a positive relationship between these variables and the likelihood of a firm receiving initial HACCP certification, and a negative association with its subsequent decertification.Survival analysis (also called duration or hazard analysis) is used to explain the intertemporal nature of firm action (Hosmer and Stanley 1999;Clarke 2003;Wooldridge 2003). We use simple means comparisons between certified and decertified firms, and logit model results, as benchmarks for comparison with survival model results. The first logit model analyzes the likelihood that a firm initially adopts EU HACCP at any point between 1998 and 2005. A second logit model analyzes the likelihood that a certified firm is decertified at any point during the period. These two models assume that the probability of an event occurring in period t is independent of time before the event occurs (constanthazard-rate assumption), but one can hypothesize that the longer the time period over which a firm has operated without HACCP, the higher the likelihood that it will not adopt HACCP in the next period. Similarly, it is likely that the longer a firm remains certified, the greater the likelihood that this firm will not be decertified in the next period.For initial adoption, the survival time is the number of years before a firm adopts HACCP after the year the EU markets imposed the system (that is, 1999) (Table 2.1[a]). For decertification decisions, survival time is the number of years between initial certification and the year that a firm is decertified (Table 2.1[b]). The distribution of survival time can be characterized by three equivalent functions: the density function, the survivor function, and the hazard function. The density function iswhere T is a nonnegative random variable that takes values t to measure the time spent in a particular state. The survivor function is defined as the probability that the event of interest has not occurred by duration t-that is, the random variable T exceeds (or at least equals) the specified time t. For the discrete case, it is given bywhere j denotes a failure time. S(t) is a nonincreasing function with a value of one at the time origin and a value of zero as t goes to infinity. The opposite of the survivor function, the hazard function, is the relationship between the hazard rate and the time already spent in that state and is presented asThe hazard function represents the conditional probability of an event occurring at the next time point, given that the event has not occurred up to a previous point. Estimation in this article employs the hazard function.Hazard functions can be estimated using either semiparametric or parametric methods. 8 Semiparametric models are more appealing as they allow greater flexibility; in semiparametric models the baseline hazard λ 0 (t) is left unspecified. 9 Hazard is a function of a vector of explanatory variables X with coefficients ß that can be estimated. The baseline hazard λ 0 is given as 8 There are several parametric models to be used based on the distribution of the hazard/risk: exponential, Weibull, Gompertz, log-logistic, log-normal, and generalized gamma models. The common approach in selecting among such approaches is to use the Akaike information criterion (AIC). Akaike (1974) proposed penalizing each log likelihood to reflect the number of parameters being estimated in a particular model and then comparing them. AIC is defined as, where c is the number of explanatory variables and p is the number of model-specific ancillary parameters, that is, two ancillary parameters for generalized gamma and one for the other models. The preferred model has the smallest AIC value.9 This is often preferred especially in applications to social sciences, in which there is not enough theory to make a strongwhere  is a positive function of X and ß, and h 0 (t) characterizes how hazard changes as a function of time. The baseline hazard depends on t, but not on X, meaning that it captures individual heterogeneity unexplained by the explanatory variables. Alternatively, the baseline hazard can be interpreted as the probability of an event occurring if the explanatory variables are set equal to zero.We adopt the semiparametric procedure developed by Cox (1972), where (X,β) is equated with exponentiated (X,β) given as(5)To estimate the hazard function, observed time periods of the event occurring are ordered by length from smallest to largest, t 1 <…<t n . The conditional probability that the event happens in the first observation at time t 1 , given that any of the n observations could have been failed at t 1 , isor the contribution of the shortest observation to the partial likelihood. More generally, the contribution of the jth shortest observation to the partial likelihood is given byFor HACCP adoption, the numerator is the conditional probability of the jth firm that initially adopts (or decertifies) at time t j , whereas the denominator is the sum of the partial likelihood for all other firms that have not initially adopted (or that have not been decertified) just prior to time t j . The likelihood is formed as the product of these contributions and may be written asThe likelihood function depends only on the unknown coefficient vector ß and can be maximized using standard methods. The functional form of the hazard rate does not need to be specified. The Cox log-likelihood function is then as follows:where ) ,. In terms of interpreting the coefficients, the hazard ratio (HR), which is exp(), is the change in the rate of the event occurring for every unit change in the explanatory variable case to assume a constant, increasing, or decreasing baseline hazard.X i .10 For a dichotomous explanatory variable, the difference in the rate of X i = 1 and X i = 0 is equal to HR. An HR greater than one (HR > 1) increases the rate of the occurrence of an outcome by ) 1 ( * 100  HR percent; an HR that is less than one (HR < 1) decreases the rate of an event occurring by / 1 / * 100  HR percent.Durbin-Wu-Hausman test results indicate that explanatory variable output (Y) is endogenous in all models. Size of production can affect a firm's decisions about whether to adopt and when to adopt. At the same time, decisions about HACCP adoption may also affect the size of production through market access. To correct for endogeneity, we use two-stage regression and first estimate a simple production function (equation 10) including product form (shrimp, milkfish, tunacan, tunafroz), capital availability (corp), and years of experience (yrbus). Test results reject the null hypothesis of no joint significance among the above explanatory variables, and thus pm, corp, shrimp, milkfish, tunacan, and tunafroz can be used as instrumental variables in the Y model. 11 Predicted Y (yhat) is an explanatory variable in the second-stage survival estimation to generate unbiased and robust estimates.The Durbin-Wu-Hausman test was also performed for a possible endogeneity problem associated with association membership in the HACCP initial participation model. For example, the decision to join an industry organization could be influenced by the participation in HACCP. However, the test indicated no endogeneity issue.The baseline results in Table 4.1, without consideration of the timing of decisions, indicate that output price and quantity are positively associated with the initial decision by seafood processors to adopt HACCP. A higher output price received provides more incentive to adopt, and estimates suggest that a US$112 increase in output price level is associated with a 32 percent increase in the likelihood of being certified at least once. A larger output (or firm size) allows the firm to capture some economies of scale and creates an incentive for certification. A means comparison indicates a significant difference between the average output of firms that initially adopt HACCP and those that do not. Regression estimates suggest that a 1,000-metric ton increase in annual output is associated with a 9 percent increase in the likelihood of being certified. None of the variables related to cost structure (that is, input price), diversification, or capacity displayed statistically significant associations with the initial certification decision even though means comparisons indicate differences between firms in all capacity-related variables.  c Due to a multicollinearity problem from the high correlation between output price level and price of materials, we were able to use only one of these variables at a time in the models. However, the two models showed similar results. The one presented here is the one using output price level. d Due to a multicollinearity problem from the high correlation between \"number of association memberships\" and \"being active in these associations,\" we were able to use only one of these variables at a time in the models. However, the two models showed similar results. The one presented here is the one using \"being active in these associations.\" *** Significant difference in averages at < 0.01 level of significance; ** at < 0.05 level of significance; * at < 0.10 level of significance.When the timing of the initial adoption decision is considered, survival analysis results show that the number of product forms and membership and active involvement in industry associations are significant in addition to output price level and output size.A $1 increase in the output price level is correlated with a 44 percent increase in the rate of initial adoption among firms, implying high sensitivity of the rate of initial adoption to changes in output price. A 1,000-metric ton increase in annual output is associated with a 1 percent increase in the rate of initial adoption, suggesting that the larger the firm size, the larger the capacity of a firm to accommodate cost increases. Adding an additional product form in the production line is correlated with a 26 percent decrease in the rate of initial adoption, perhaps because of costly processes to avoid cross-contamination. A firm active in trade or processor associations is twice as likely to adopt compared with an inactive firm, suggesting that associations may be sources of institutional pressure yet still provide needed capacity, information, technical support, and resources.Overall results indicate that initial certification decisions are more strongly influenced by easily obtainable a priori indicators-namely, output prices, firm output, and to some extent association membership and product diversification. Less easily observable factors such as input cost structure and financial and operational capacity appear to have no influence on the initial adoption decision.Explanatory variable mean values are statistically similar for both certified and decertified firms (Table 5.1) with the exception of the average EU-US price differential. Firms that remained certified (until 2006) reported no difference in prices received in the US and EU markets, but decertified firms (in 2006) reported receiving significantly lower prices from EU buyers compared with US buyers (by $0.94/kilogram on average) with the constant-hazard-rate assumption. c Due to a multicollinearity problem from the high correlation between output price level and price of materials, we were able to use only one of these variables at a time in the models. However, the two models showed similar results. The one presented here is the one using output price level. d Due to a multicollinearity problem from the high correlation between \"number of association memberships\" and \"being active in these associations,\" we were able to use only one of these variables at a time in the models. However, the two models showed similar results. The one presented here is the one using \"being active in these associations.\" *** Significant difference in averages at < 0.01 level of significance; ** at < 0.05 level of significance; * at < 0.10 level of significance.Logit model results indicate that price differentials across markets (especially between the EU and the United States) and access to credit are significant factors in the sustainability of HACCP certification among Philippine seafood processors. A $1 per kilogram increase in EU prices, compared with US prices, is estimated to result in a 94 percent decrease in the likelihood of being decertified, 13 confirming high sensitivity of EU HACCP decertification to price differentials between the EU and US markets. Firms that received higher prices in the US market and sold a significant share of their output in the United States had no incentive to continue with EU HACCP certification. Logit model results also suggest that relatively lower EU prices versus prices in other markets (Japan, other non-US, and domestic) were negatively associated with continued certification. Cost structure was not significant in explaining decertification decisions when timing was not considered.With respect to capacity, only access to credit had a statistically significant effect on continued certification decisions in the logit model: firms that reported difficulty accessing credit had a higher likelihood (42 percent) of being decertified than firms reporting no difficulty. Capital requirements of food safety systems increase the demand for scarce funds, and public-sector credit support may be needed to jump-start and sustain food safety systems.When the constant-hazard-rate assumption is relaxed (that is, survival analysis), the rate of decertification is significantly affected not only by EU-US price differentials but also by other incentiveand capacity-related factors such as output price level, input prices, number of markets, number of product forms, access to credit, and membership and active involvement in associations. Reported hazard ratio estimates suggest that a $1 increase in the difference between EU and US prices is associated with a 99 percent decrease in the rate of decertification. With an average EU-US price differential in 2004-2008 of -$0.25/kilogram, a dollar increase will make EU prices four times more attractive than US prices and will likely eliminate almost all EU HACCP decertifications by firms. Results suggest that firms choose their certification strategies depending on their target market. In our sample, firms that exported mainly to the EU were less likely to decertify from EU HACCP, whereas firms that exported mainly to the United States and other markets were more likely to decertify. Hence HACCP capacity building appears to be tailored to requirements of the importing nations.Similarly a $1 increase in the output price level is associated with a 70 percent reduction in the rate of decertification, suggesting very high sensitivity of the rate of decertification to changes in price level. An increase in the number of product forms by one is associated with a 2.5-times increase in the rate of decertification (that is, firms handling multiple product forms were highly likely to decertify). In contrast, an additional market destination is associated with a 53 percent decrease in the rate of decertification, indicating that reputation achieved through EU HACCP certification was important in continued certification for firms selling in a number of markets.Cost variables were also significant when time was considered; a $1 increase in the price of labor is associated with a 63 percent decrease in the rate of decertification. Although this result appears to be inconsistent with the initial hypothesis that higher wage rates will create incentives for decertification because of the additional labor involved with maintaining safe practices, discussions with respondents suggest that higher wage rates created incentives for the firms to seek higher prices through HACCP certification as a competitive strategy because higher local wage rates made them uncompetitive in domestic markets. Firms that reported difficulty in accessing credit had a rate of decertification three times higher than those that reported no difficulty. These results suggest that easy access to credit is a critical factor driving both initial and sustained certification.A firm that is active in an association has a 70 percent lower rate of decertification than a firm that is not active. Institutional pressures imposed through association membership and the potential preferential access to resources because of active participation play important roles in influencing continued certification. Building institutional pressures and collective capacity through industry associations can help promote early adoption as well as continued certification. Such associations can provide a venue to mobilize resources and share expertise and relevant information about food safety systems and market opportunities. Shared information through industry associations can reduce erroneous decisions arising from managerial hubris.Scale economies did not appear to play a significant role in the decertification decision, likely because firms without adequate scale economies were filtered out in the initial certification decision. Even though improving food safety is paramount across all firm types and firm sizes, some firms may need different forms of incentives and capacity building on a more sustained basis.Lack of knowledge about the sustainability of food safety certification systems constitutes a major gap in our understanding of food policy and regulation. Relaxing the constant-hazard-rate assumption using survival analysis techniques reveals the significant influence of additional factors on the conditional decisions. Differences between survival and logit model estimates for both initial certification and subsequent decertification decisions suggest that assuming constant hazard rates may lead to potentially misleading results and misestimation of the influence of other relevant drivers.The initial decision to certify appears to be influenced significantly by scale economies and easily observable information such as output prices, access to credit, and institutional pressures from association membership. Other incentive-and capacity-related factors do not appear to have a significant influence.After certification, a firm's decision to continue or discontinue certification at a point in time will be influenced by the same set of incentive-and capacity-related factors, but with the availability of additional information concerning realized costs and benefits. As a result decertification decisions appear to be significantly affected by a larger number of revenue, cost, and nonfinancial factors (that is, output price differentials, product and market diversification, input costs, and institutional factors). Although logit model results suggest that price differences between the EU, US, and other markets and credit availability are the main drivers of the decertification decision, estimates using survival analysis techniques reveal the significant influence of additional factors, including the extent of market and product diversification, institutional pressures, and labor costs.Managerial hubris, likely in the absence of adequate information, may have played a significant role in initial certification decisions, but decertification decisions were based on more-informed cost and benefit information. The results support our hypothesis that compared with initial certification, decertification decisions were better-informed, calculated business decisions. The results suggest that an increased emphasis on policy measures aimed at discouraging decertification, versus a focus limited to capacity building to achieve initial certification, may be necessary to sustain food safety initiatives such as HACCP.","tokenCount":"5061","images":["-2109299831_1_1.png"],"tables":["-2109299831_1_1.json","-2109299831_2_1.json","-2109299831_3_1.json","-2109299831_4_1.json","-2109299831_5_1.json","-2109299831_6_1.json","-2109299831_7_1.json","-2109299831_8_1.json","-2109299831_9_1.json","-2109299831_10_1.json","-2109299831_11_1.json","-2109299831_12_1.json","-2109299831_13_1.json","-2109299831_14_1.json","-2109299831_15_1.json","-2109299831_16_1.json","-2109299831_17_1.json","-2109299831_18_1.json","-2109299831_19_1.json","-2109299831_20_1.json","-2109299831_21_1.json","-2109299831_22_1.json","-2109299831_23_1.json","-2109299831_24_1.json","-2109299831_25_1.json","-2109299831_26_1.json","-2109299831_27_1.json","-2109299831_28_1.json"]}

data/part_2/0386182555.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"17d4c42f6d03d734f7535ed344719381","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/547bf8c2-f563-4c90-872a-ab84ae4d729b/retrieve","description":"After three decades of fluctuating but overall diminishing public agricultural research and development (R&D) spending in Zambia, the downward trend of investment accelerated during 2001-08. In 2008, Zambia spent 20 billion kwacha or 8 million PPP dollars on public agricultural R&D, both in 2005 constant prices, compared with 24 billion kwacha or about 10 million PPP dollars in 2001, and 89 billion kwacha or 37 million PPP in 1991. Unless otherwise stated, all dollar values in this note are expressed in purchasing power parity (PPP) prices. PPPs reflect the purchasing power of currencies more effectively than do standard exchange rates because they compare the prices of a broader range of local-as opposed to internationally traded-goods and services. Public agricultural R&D capacity increased in the 1980s, then experienced a sharp decline in the early 1990s. A period of growth in the mid-1990s was then followed by another period of contraction, up through 2006, primarily due to a government-sector hiring freeze imposed from 2002 until 2007. When recruitment resumed, research capacity grew quickly and returned to mid-1990s levels by 2008, with a total of 209 full-time equivalent (FTE) research staff employed that year.","id":"119340371"},"keywords":[],"sieverID":"65c7eea0-276e-4910-bdc8-137afdb174a0","pagecount":"8","content":"Zambia's historical trend of declining public agricultural research and development (R&D) investments continued during 2001-08 due to weakened government and donor support.• The country's agricultural research capacity also deteriorated during 2001-06, both in terms of numbers of full-time equivalent researchers and levels of educational qualiications. This can largely be attributed to a government-sector hiring freeze during 2002-07, after which staf numbers once again began to rise, but predominantly in the category of junior (BSc-qualiied) staf.• The most signiicant institutional change in Zambia's public agricultural research system was the 2005 upgrade of the Soil and Crops Research Branch (SCRB) to a ministerial department under the name Zambia Agricultural Research Institute (ZARI).• Funding for agricultural research in Zambia is primarily derived from the national government, supplemented by limited support from foreign donors and development bank loans. Government funding, however, is largely allocated to salaries and overhead, making it crucial that agencies secure donor funding for operating and capital costs related to research.A fter three decades of luctuating but overall diminishing public agricultural research and development (R&D) spending in Zambia, the downward trend of investment accelerated during 2001-08. In 2008, Zambia spent 20 billion kwacha or 8 million PPP dollars on public agricultural R&D, both in 2005 constant prices (Figure 1, Table 1), compared with 24 billion kwacha or about 10 million PPP dollars in 2001, and 89 billion kwacha or 37 million PPP in 1991. Unless otherwise stated, all dollar values in this note are expressed in purchasing power parity (PPP) prices. 1 PPPs relect the purchasing power of currencies more efectively than do standard exchange rates because they compare the prices of a broader range of localas opposed to internationally traded-goods and services. Public agricultural R&D capacity increased in the 1980s, then experienced a sharp decline in the early 1990s (Figure 2). A period of growth in the mid-1990s was then followed by another period of contraction, up through 2006, primarily due to a government-sector hiring freeze imposed from 2002 until 2007. When recruitment resumed, research capacity grew quickly and returned to mid-1990s levels by 2008, with a total of 209 full-time equivalent (FTE) research staf employed that year. 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 Nonprofit  1981 1984 1987 1990 1993 1996 1999 2002 2005 2008   In a 2008 sample of nine public agricultural research agencies in Zambia, almost a quarter of all FTE researchers were female-a signiicant increase over the 9 percent share reported in 2000 (ASTI 2009;Beintema et al. 2004). ZARI was largely responsible for this shift, increasing its share of female researchers from 5 percent in 2000 to 20 percent in 2008. Shares of female researchers varied signiicantly at the other government and higher education agencies.The ratio of support staf to researchers decreased on average from 5.6 in 2001 to 3.9 in 2008 (ASTI 2009)  Underlying datasets can be downloaded using ASTI's data tool at www.asti.cgiar.org/data. This brief presents aggregated data; additional graphs with more detailed data are available at asti.cgiar.org/zambia/datatrends.1.2 technicians, 0.7 administrative staf, and 2.0 other support staf. The nonproit agencies reported high ratios of support staf to researchers (12.0 at GART and 8.0 at CDT), whereas the higher education sector employed only 1 to 2 support staf per researcher on average in 2008-a consistent and understandable inding given that research is not their primary mandate.Intensity ratios are commonly used to compare agricultural R&D spending and capacities across countries. One indicator of public research intensity is total agricultural R&D spending as a percentage of agricultural output (AgGDP). Zambia's agricultural intensity ratio continued a sharp decline from the 1990s. In 2008, for every $100 of agricultural output the country only invested $0.29 in agricultural R&D, down from $0.47 in 2001 (Figure 3). This decline was the combined result of increasing AgGDP and declining agricultural R&D spending. The ratio of agricultural researchers to farmers also declined to a low of 48 FTE researchers for every million farmers in 2005, before improving again in 2007, and inally reaching 67 FTE researchers per million farmers in 2008 based on the recommencement of staf recruitment in the government sector.The main change in the institutional structure of public agricultural R&D in Zambia since 2000 is the transformation of the Soil and Crops Research Branch (SCRB) under the Department of Research and Specialist Services (DRSS), which was administered by the Ministry of Agriculture, Food, and Fisheries (MAFF). In 2005, SCRB became ZARI, a department under the renamed Ministry of Agriculture and Cooperatives (MACO). ZARI comprises four technical divisions: Crop Improvement and Agronomy, Soil and Water Management, Plant Protection Quarantine, and Farming Systems and Social Sciences. The upgrade included an expanded management team from a single deputy director, to a director and two deputies-one to oversee the four technical divisions, and the other to centrally coordinate research services and zonal development programs, while maintaining a focus on Zambia's three agroecological regions. ZARI operates nine agricultural research stations with outreach programs to focus on local constraints. It was originally intended that ZARI would be given semiautonomous status (Elliott and Perrault 2006) to increase its ability to charge for services and directly attract donor funding; as it stands, however, ZARI continues to operate as a department under MACO.Other signiicant changes since 2000 involved the restructuring of MACO, which resulted in the creation of a separate ministry for livestock and isheries that now administers both CFRI and CVRI. Previously, forestry research was conducted by the Forestry Research Branch (FRB) under the Ministry of Environment and Natural Resources. In 1995, CBU was requested to run FRB as a research and training entity and in 2004, it became part of CBU.The National Institute for Scientiic and Industrial Research (NISIR) manages SUUGR, LPRC, and PHFPN. Some of its programs and units underwent reorganization since 2000. SUUGR was formed from the Tree Improvement Research Centre, while PHFPN had previously been known as the Food Technology Research Unit. The Water Resources Research unit merged with the unit dealing with energy and the environment. Responsibility for NISIR falls under the Ministry of Science, Technology, and Vocational Training (MSTVT). The Science and Technology Policy of 1996 that established NISIR also established the National Science and Technology Council (NSTC) under MSTVT. The legislation enacting the policy allowed for the centralization of all agricultural research activities under the management of MSTVT and under the coordination of NSTC, but MAFF was reluctant to relinquish its research mandate (Elliott and Perrault 2006). In 2009, MSTVT revised the 1996 policy. The new policy ASTI Website Interaction www.asti.cgiar.org/zambia  A list of the 6 government agencies, 2 nonproit, 6 higher education, and 3 private agencies included in this brief is available at asti.cgiar.org/zambia/agencies.  Detailed deinitions of PPPs, FTEs, and other methodologies employed by ASTI are available at asti.cgiar.org/methodology.  The data in this brief are predominantly derived from surveys. Some data are from secondary sources or were estimated. More information on data coverage is available at asti.cgiar.org/zambia/datacoverage. More relevant resources on agricultural R&D in Zambia are available at asti.cgiar.org/ zambia.  1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 Figure Forestry research is now oicially administered by CBU, as previously mentioned. Further, CBU's role in agricultural research may increase in the years to come because the recent expansion of its School of Natural Resources to include degrees in agroforestry, wildlife management, and isheries and aquaculture.Collaboration among agencies at national, regional, and international levels is integral to Zambian agricultural research. Numerous projects on a range of commodities and themes are implemented jointly with centers of the Consultative Group on International Agricultural Research (CGIAR). ZARI participates in a number of South African Development Community (SADC) regional networks, such as the SADC Plant Genetic Resources Center (SPGRC) and its Food, Agriculture, and Natural Resources (FANR) Directorate, as well as the Southern Africa Root and Tuber Research Network (SARRNET), and the African Center for Fertilizer Development (ACFD). UNZA belongs to the Regional Universities Forum for Capacity Building in Agriculture (RUFORUM) and the SADC Bean Research Network (Haazele 2008).Half of all the public agricultural researchers employed in Zambia in 2008 were trained to the postgraduate level, a signiicant shift from 2001 when 70 percent of researchers held postgraduate degrees (Figure 4). The increasing share of BScqualiied staf at ZARI and other government agencies stems from the aforementioned government-sector hiring freeze; lack of appropriate training opportunities; and the concurrent reduction in the number of senior researchers based on losses to other agencies, retirement, or death. The loss of PhD-qualiied researchers particularly afected ZARI, where the number of PhD-qualiied staf fell from 15 FTEs in 2001 to 9 in 2008. The number of female researchers with PhD degrees changed little, although the share shifted signiicantly given that the majority of newly hired female researchers were BSc-qualiied (ASTI 2009; Beintema et al. 2004). A recent ZARI report noted a number of constraints to hiring and retaining qualiied staf, including inadequate compensation and beneits, low morale, a slow recruitment process, and a lack of an established staf training program (ZARI 2009). Public agricultural research institutions in many developing countries are facing similar challenges because agencies in the higher education sector, the private sector, and abroad are able to ofer more lucrative packages, often under more attractive conditions.The general increase in the number of BSc-qualiied researchers was even more apparent at the other government agencies and GART. Overall, the share of BSc-qualiied staf at the other government agencies increased from 20 percent in 2001 to 58 percent in 2008. At GART, the increase in researchers with BSc degrees equalized shares of all three types of degree qualiications.As is the case in most universities in Africa and other regions of the world, a greater share of staf in the higher education sector in Zambia have postgraduate degrees compared with government agencies. The number of staf with PhDs remained relatively stable during 2001-08, but an increase in the number of MSc-qualiied staf caused a shift in the share of staf with PhD degrees from 57 percent in 2001 to 43 percent in 2008. MSc-qualiied faculty staf now represent more than half of the FTE research staf at the agricultural higher education agencies. Again, this structural shift stems from the retirement of many PhD holders while newly recruited staf have had limited training opportunities. In addition, the high proportion of PhD-qualiied staf reported during the late-1990s and early 2000s resulted from training programs supported by the U.S. Agency for International Development (USAID), the World Bank, and the Swedish International Development Cooperation Agency (Sida), but no programs have been ofered in more recent years.The majority of technicians in the government, nonproit, and higher education sectors held some type of diploma or degree, but relatively few held BSc or higher degrees. At ZARI, for example, no technicians held BSc or higher degrees in 2008, whereas 195 technicians had some other type of diploma or degree (ASTI 2009). The allocation of research budgets across salaries, operating costs, and capital investments afects the eiciency of agricultural R&D, so detailed cost category data were collected from the government agencies as part of this study. On average, at least half of all expenditures at ZARI and the other government agencies during 2001-08 were allocated to salaries (Figure 5). In 2008, expenditures at ZARI were evenly split between salary and nonsalary costs. 4 The 4 billion kwacha (in constant 2005 prices) spent on operating and capital costs by ZARI represented a signiicant increase over the 2006 levels of only 1 billion kwacha.The government began investing in capital improvements at ZARI in 2008, but at a fraction of the levels recorded in the late-1990s (Beintema et al. 2004). In a recent report by ZARI, lack of infrastructure was noted as a major constraint to research, particularly in terms of lack of laboratory equipment and vehicles, and inadequate buildings, staf housing, irrigation, and communication facilities (ZARI 2009). Delays in and shortfalls from budgeted funding were also cited as signiicant constraints, both from the government and from foreign donors. Given the size of the country, transportation costs are also a continual challenge for ZARI and other agencies.Agricultural R&D in Zambia was primarily funded by the government, supplemented by foreign donors and minor contributions on the part of the agencies through the sale of goods and services. In 2008, 96 percent of funding for ZARI was supplied by the national government, with donors contributing 4 percent (Figure 6). Donor funding to ZARI that year represented a third of the amount contributed in 2001. Other government agencies received higher shares of funding from donor agencies and through the sale of goods and services. In general, government funding mainly supports salaries and overhead, making donor funding crucial to the support of operating and capital costs associated with research.A recent public expenditure review notes that it is diicult to accurately identify the levels of donor funding to MACO because inancial reports often omit such data (Orlowski et al. 2010). The review does estimate that 78 percent of donor funding to MACO in 2009 was provided by three major donors: the African Development Bank, the World Bank, and the European Union. Other donors to agricultural research agencies in Zambia include the International Fund for Agricultural Development (IFAD), the Food and Agriculture Organization of the United Nations (FAO), the Norwegian Agency for Development (NORAD), the governments of France and Finland, Sida, USAID, the Japan International Cooperation Agency (JICA), the Department for International Development (DFID), the International Atomic Energy Agency (IAEA), the International Maize and Wheat Improvement Center (CIMMYT), the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), the International Institute of Tropical Agriculture (IITA), the International Livestock Research Institute (ILRI), the International Center for Agricultural Research in the Dry Areas (ICARDA), and the World Resources Institute.In the nonproit sector, the government and foreign donors contributed most of CDT's funding, but a small amount was derived from cotton producer organizations such as the Zambia Cotton Ginners Association. In contrast, GART only received a small share of its funding from the government, given that its major source of funding is the sale of goods and services, including commercial farming and contract research. GART also received donor funding from the governments of Norway and Sweden and the UN Common Fund for Commodities (GART 2008).The World Bank was a major source of funding to agricultural research in Zambia during the 1990s and early 2000s, providing loans through projects co-inanced by the government and other donors. The irst project, the Zambia Agricultural Research and Extension Project (ZAREP), began in 1987 and provided US$40 million in funding for infrastructure, staf training, and improved institutional management and collaboration (Beintema et al. 2004). The Agricultural Sector Investment Program (ASIP), which took a broader sectoral approach, followed from 1996 until 2001 (World Bank 1995). The agricultural research component of this project, which totaled US$35 million, focused on supporting 2 0 0 1 2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 2 0 0 6 2 0 0 7 2 0 0 8 2 0 0 1 2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 2 0 0 6 2 0 0 7 ZARI Other government (4) 2 0 0 8 privatization through the creation of GART, funding staf training, and re-equipping and rehabilitating the country's research stations.There were no large-scale donor-funded projects from 2002 to 2006, so funding levels declined signiicantly. In 2006, the World Bank-funded Agricultural Development Support Project (ADSP) began, which is expected to run until 2012 at a total project cost of US$37 million (World Bank 2006). ADSP focuses on market-oriented development, supporting feeder roads and agro-industry. A small amount of funding is allocated to agricultural research and extension. ZARI and CDT were named as beneiciaries of the institutional development component of the project, with budgets of US$0.8 and US$1.5 million, respectively.A number of competitive grants are available for agricultural R&D in Zambia. The Agricultural Innovation Fund (AGRIFU) is one such source of funding, for CDT and other agencies. Details on the size and scope of this fund were unavailable, however. There are two competitive funds for research managed by NSTC on behalf of MSTVT: the Strategic Research Fund (SRF) and the Youth Innovators Fund (YIF) (NSTC 2010). SRF grants aim to support research projects on an institutional basis, whereas YIF supports individuals or groups of researchers between the ages of 15 and 35 years. Another fund-the Science and Technology Development Fund (STDF)-was enacted in 1997, but it is not yet operational. It is intended that STDF grants will target individual researchers or agencies, and could fund postgraduate research studies, capacity building, research-related travel, and various research projects. In contrast with AGRIFU, agriculture is one of several priority areas of these three funds.Funding for agricultural research at universities is derived from diverse sources. As a public institution, UNZA, for example, solicits research funding through competitive grants at both the institutional and individual levels. At the university level, the Directorate of Research and Postgraduate Studies oversees institutional-level eforts to raise funding, whereas individuals can source funding by responding to (local or international) public announcements or by securing research on a contract basis.Given that the allocation of resources across various lines of research is a signiicant policy decision, detailed information was collected on the number of researchers working in speciic commodity and thematic areas (in FTEs).The predominant focus of agricultural research in Zambia was crops. In 2008, 59 percent of researchers were involved in crop research, while 15 percent focused on livestock, 6 percent focused on forestry, 4 percent focused on isheries, and 3 percent focused on natural resources (Figure 7). These shares shifted somewhat from 2000, when crop research accounted for half the country's FTE researchers, and natural resources was the focus of 11 percent of researchers (Beintema et al. 2004).Taking a closer look at crop and livestock research, maize was the most heavily researched crop, accounting for 18 percent of the crop researchers at ZARI and 20 percent of crop and livestock researchers at GART (Table 2). Other important crops included sorghum, cassava, fruit, and vegetables. At CVRI and LPRC-the government agencies with livestock research mandates-dairy and beef were the major areas of research. Researchers at GART and the higher education agencies also spent signiicant time on livestock research, with poultry being their primary focus.  Crop research dominates in terms of thematic focus at ZARI and in the nonproit sector. In 2008, crop genetic improvement and crop pest and disease control each accounted for 7-8 percent of FTE researchers (Table 3). Natural resources themes were also strong at ZARI and in the higher education sector, resulting in a 9 percent share of FTE researchers focusing on soil issues, 7 percent focusing on water issues, and 9 percent focusing on other issues related to natural resource research (including forestry). The combined efects of a government-sector hiring freeze and lack of training opportunities resulted in signiicant erosion of research staf capacity. Although staing levels increased to 209 FTE researchers in 2008, the composition shifted towards junior rather than senior researchers, meaning those holding BSc rather than PhD degrees. ZARI was particularly afected by a reduction in the number of PhD-qualiied researchers.In addition to training and capacity limitations, ZARI and the other government research agencies have also faced challenges in supporting the operating and capital costs associated with research. A number of needs have been identiied at ZARI, such as infrastructure, laboratory equipment, communication facilities, and vehicles. Delays and reductions in the disbursement of budgeted funding from both the national government and foreign donors continue to constrain the eicient management of research funding.Although GART has been successful in generating funding through the sales of goods and services, as well as attracting donor funding and strengthening linkages with UNZA, other trusts within the nonproit sector have not fared as well. They were originally created for the purpose of increasing the lexibility and eiciency of research funding and management, in addition to promoting public-private partnerships. They, however, still depend on national government funding and have yet to meet the expectations of their mandate.Although the recent rise in the number of agricultural researchers is positive-as is the upgrade of ZARI to a ministerial department, and increased investment under ADSP since 2007-Zambia's agricultural R&D agencies are still contending with the efects of long-term underinvestment and continue to struggle with funding issues that hinder their ability to contribute more efectively to the country's agricultural and economic development.1 Financial data are also available in current local currencies or constant 2005 US dollars in the ASTI Data Tool, www.asti.cgiar.org/data.2 Financial data for these private companies were unavailable; for more detailed information on the private sector in Zambia, see Mwala and Gisselquist (2010, forthcoming).3 Of note, operations by both these development trusts have been severely constrained by lack of funding, so they have not performed as expected.4 Donor funding allocated collectively to operating and capital costs could not be disaggregated; hence, only salary and nonsalary costs could be identiied. ","tokenCount":"3497","images":[],"tables":["119340371_1_1.json","119340371_2_1.json","119340371_3_1.json","119340371_4_1.json","119340371_5_1.json","119340371_6_1.json","119340371_7_1.json","119340371_8_1.json"]}

data/part_2/0386276309.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"765e3301be4e949235eb09f8d2020794","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/26d8dc7b-17f3-479d-bd54-16504cd359af/retrieve","description":"Angeline Munzara IFPRI-FAO conference side event, \"Accelerating the End of Hunger and Malnutrition\" November 28–30, 2018 Bangkok, Thailand","id":"-193005220"},"keywords":[],"sieverID":"94dce884-1765-4ce7-94d9-407d8d12237f","pagecount":"7","content":"Saving Group: Uganda• Conflict and climate change in fragile settings have led to protracted crises, forcing populations into refugee and internally displaced people's camps and disrupting local economies.• Resulting in high rates of hunger & malnutrition as capacity of parents and caregivers to provide well for their children is reduced.Solution is Holistic: SDG 2 calls for multi-sectorial responses to addressing hunger and malnutrition. Not only is humanitarian assistance is sufficient to address the challenge. Solutions must bridge the humanitarian and development divide to build resilient and productive societies and end intergenerational cycles of poverty.SDG 2:3 \"By 2030, double the agricultural productivity and incomes of smallscale food producers, in particular women, indigenous peoples, family farmers, pastoralists and fishers, including through secure and equal access to land, other productive resources and inputs, knowledge, financial services, markets and opportunities for value addition and non-farm employment . \"• Doubling agricultural productivity and incomes ensures that all elements of food security are met (availability; accessibility; affordability and cultural compatibility) • Financial inclusion has a role to play to support productivity and increasing incomes. • 1.7 billion adults have no access to banking services (World Bank, Global Findex Index 2017: Measuring Financial Inclusion and the Fintech Revolution)Saving Group: Ethiopia • Can easily be applied in fragile contexts to build resilient and productive livelihoods of the most vulnerable communities. • Offers a bridge between the humanitarian and development divide to start the process of emergency recovery for long lasting food and nutrition security. • S4T groups when integrated as an underlying pillar in nutrition sensitive value chains, they help to boost agricultural productivity of small holder farmers when they have access to agricultural inputs on time and when they are taught on financial literacy and the importance of saving when they sell their produce. • Provide a platform for education on nutrition, health, water and sanitation, peace building etc to members of the group.","tokenCount":"317","images":["-193005220_1_1.png","-193005220_1_2.png","-193005220_1_3.png","-193005220_2_1.png","-193005220_2_2.png","-193005220_2_3.png","-193005220_3_1.png","-193005220_3_2.png","-193005220_3_3.png","-193005220_4_1.png","-193005220_4_2.png","-193005220_4_3.png","-193005220_5_1.png","-193005220_5_2.png","-193005220_6_1.png","-193005220_6_2.png","-193005220_7_1.png","-193005220_7_2.png"],"tables":["-193005220_1_1.json","-193005220_2_1.json","-193005220_3_1.json","-193005220_4_1.json","-193005220_5_1.json","-193005220_6_1.json","-193005220_7_1.json"]}

data/part_2/0387890218.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"014c0e2880ea8a70e7119dd2030e4b7d","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/ed36b5d7-2e2a-4696-9fbd-19ebe1754dca/retrieve","description":"Agriculture, natural resources, and the nutrition landscape in South Asia (SA) are unique. As in other Asian regions, SA’s agriculture is largely smallholder based and continues to employ a large share of the workforce, compared with Latin America. Compared with Africa south of the Sahara (SSA), agricultural modernization and intensification in SA have progressed more substantially. However, compared with East Asia (that is, Northeast and Southeast Asia), the speed of agricultural transformation has remained slower in terms of the exit of labor from farming, despite comparable intensification levels. The region is still home to almost 300 million of the poor, a majority of whom live in rural areas, engaging in agriculture. SA is also one of the regions with the scarcest natural resource endowments per capita, including water resources.1 Finally, the multiple burden of malnutrition persists, as the region remains one of the largest contributors to global undernutrition, while simultaneously overnutrition continues to emerge (Meenaksh 2016). Therefore, in SA, understanding the evolution and implications of agricultural development is important particularly in the context of agricultural transformation, natural resource management, poverty, and food/nutrition security improvement.","id":"-1476896660"},"keywords":[],"sieverID":"1e8a0bd0-c9a3-48fd-af1e-ea43eab920e4","pagecount":"42","content":"the nutrition landscape in South Asia (SA) are unique. As in other Asian regions, SA's agriculture is largely smallholder based and continues to employ a large share of the workforce, compared with Latin America. Compared with Africa south of the Sahara (SSA), agricultural modernization and intensification in SA have progressed more substantially. However, compared with East Asia (that is, Northeast and Southeast Asia), the speed of agricultural transformation has remained slower in terms of the exit of labor from farming, despite comparable intensification levels. The region is still home to almost 300 million of the poor, a majority of whom live in rural areas, engaging in agriculture. SA is also one of the regions with the scarcest natural resource endowments per capita, including water resources.1 Finally, the multiple burden of malnutrition persists, as the region remains one of the largest contributors to global undernutrition, while simultaneously overnutrition continues to emerge (Meenaksh 2016). Therefore, in SA, understanding the evolution and implications of agricultural development is important particularly in the context of agricultural transformation, natural resource management, poverty, and food/nutrition security improvement.Research in the agricultural economics field has shed important light on various aspects of agricultural development patterns in the last few decades in the region, ranging from productivity growth, changes in resource endowments, intensification/modernization, urbanization and nutritional transition, and their drivers as well as inhibitors. The contribution of this chapter is to provide, through the review of productivity growth patterns and seminal papers in the literature, narratives of how these aspects may be interconnected in terms of their collective implications for the challenges and opportunities of future agricultural development in SA.1 Renewable internal freshwater per capita in SA is greater than that of only the Middle East and North Africa and is a fraction of the resources of East Asia and the Pacific, SSA, or Latin America (World Bank 2018).Chapter 4 TAbLE 4.1 Estimates of land and labor productivity and endowments in agriculture, South Asian countries, 1960s-2010sThe second section of this chapter highlights that labor productivity (which is closely related to returns to labor) seems inherently low in SA and that agriculture seems to have contributed to income growth through land productivity growth. But the region may be approaching a point where further growth must come from greater labor productivity growth rather than land productivity growth. We explore this issue by examining the changes in labor and land productivity growth, as well as the land-labor ratio. The third section highlights that, in SA, because of low-income status and limited resources for food imports, domestic production remains an important source of food for consumers. Demand for diversification and improved nutritional intake might have been and may continue to be met by diversification in agricultural production in the region, while growing exportable high-value agricultural products may induce some specialization.The fourth section then discusses key technological and institutional innovations, emerging rural employment as well as challenges, and growing inefficiency of small farms, with potential implications for the future labor productivity increase in SA. The section highlights innovations in technologies and institutions that have emerged, and related policies. Growing evidence of the inefficiency of small farms is reviewed. Emerging rural employment growth patterns, where the agricultural sector continues to be dominant but the nonfarm sector, such as the garment sector, is gradually expanding as a source of employment, are also discussed for their implications for labor productivity. It also touches on some inequality issues that are associated.The last section summarizes the main messages and discusses some forward-looking emerging issues.This section reviews the key patterns of agricultural productivity growth in SA, important aspects of the growth patterns of total factor productivity (TFP), and changes in natural resource endowments. The section aims to show that the past labor productivity growth in SA has been associated largely with the growth in land productivity, although TFP growth, particularly since the 1980s, has contributed to it as well. The section also shows that natural resource endowments in SA (water, soil, land) have been gradually eroded, potentially limiting substantial future land productivity growth in the region.This section extends analyses by Hayami and Ruttan (1985, Chapter 5) of factor endowments and partial productivity growth, focusing on key SA countries. Table 4.1 summarizes the changes in partial productivity of land and labor, as well as land-labor ratios, in seven SA countries from the 1960s to 2010s. This table corresponds to Table 5-1 in Hayami and Ruttan (1985). Figure 4.1 then plots the labor and land productivity for the seven countries presented in Table 4.1, corresponding to Figure 5-1 in Hayami and Ruttan (1985). To highlight the differences between SA and East Asia, which were analyzed in Chapter 3, Figure 4.1 also plots trajectories for East Asian countries (countries shown without labels). Table 4.1 and Figure 4.1 can be interpreted in the following way: many countries from the 1960s to the 2010s exhibited slopes that were steeper than the uni-A/L lines (dotted lines in Figure 4.1 along which agricultural output per agricultural land ([A]) and agricultural output per worker ([L]) change at the same proportions), indicating that increases in labor productivity were smaller than those in land productivity. Afghanistan and Bhutan even had absolute declines in labor productivity during these periods. While Bangladesh has been an exception, labor productivity growth has been only marginally higher than land productivity growth. Even in Pakistan, where land per labor has been historically high, the recent trend has exhibited greater land productivity growth than labor productivity growth. These patterns suggest that increases in land productivity also contributed to the increase in labor productivity between the 1960s and 2010s. Few of these countries have experienced growth paths that were more toward greater labor-productivity improvement than land-productivity improvement, which many higher-income countries have experienced. However, shifts from paths steeper than the uni-A/L line to paths that are parallel to it, which Bangladesh has experienced, suggest that those countries have managed to move out of the conditions prevalent in earlier periods, when high growth rates of the labor force in agriculture largely eroded the gains in labor productivity growth arising from land productivity growth.Comparisons between SA countries and East Asian countries suggest that the trajectories of many SA countries are similar to those of East Asia, with SA lagging behind rather than following a different path. The differences between SA and East Asia might have been due to factors that have led to slower growth in human capital and land quality in SA, such as education and irrigation, as described in more detail in the later section. Table 4.2 summarizes the growth rates of labor productivity and land productivity, and the land-labor ratio of the 1960s, 1980s, and 2010s. This table corresponds to Table 5-2 of Hayami and Ruttan (1985). There is some intraregion diversity in productivity growth and the changes in factor endowments. Land productivity (Y/L) has exhibited some convergence, in which countries with low initial levels (India and Pakistan) have experienced faster growth compared with countries with higher initial levels (Bangladesh, Sri Lanka). On the other hand, signs of convergence are much weaker for labor productivity. Although Pakistan, with the highest labor productivity, experienced lower growth rates than other countries in the 2000s and 2010s, its growth rate had been high up to the 1990s. In contrast, Nepal and Sri Lanka, whose labor productivity was lower in the 1960s, have seen slower growth rates up to the 2010s. While land-labor ratios have remained relatively constant in Bangladesh, Nepal, and Sri Lanka from the 1960s to the 2010s, those in India and Pakistan have declined more sharply. Because of this pattern, labor productivity growth in India and Pakistan has been much lower than the land productivity growth during this period (1.4 and 1.2 percent as opposed to 2.1 and 2.7 percent). The absorption of the growing rural labor force in nonfarm sectors in India and Pakistan has been much slower than in the other three countries in SA. Sri Lanka has experienced particularly slow labor and land productivity growth rates, with labor and land productivity being caught up by Bangladesh (labor productivity) and India or Nepal (land productivity) by the 2010s. The slow productivity growth in Sri Lanka in the last three decades has been partly due to the slow output growth rates of nonrice commodities that have accounted for about two-thirds of total outputs, including vegetables and fruits (plantains, etc.), tea, coconuts, and rubber (these are discussed more at the beginning of the third section).Figure 4.2 and Figure 4.3 illustrate the changing relationship between the land-labor ratio and tractor horsepower (HP) per worker, and between land productivity and fertilizer input per hectare, corresponding to Figures 5-3 and 5-4 in Hayami and Ruttan (1985). These figures illustrate the agricultural growth process that involves the substitution of man-made inputs for labor and land. Fertilizer and tractor HP are used as proxies for the factors that substitute for land and labor, respectively.Most SA countries have seen substantial increases in tractor HP per worker, without increases in land-labor ratios in the past five decades (Figure 4.2). In some countries, like India and Pakistan, the land-labor ratios have declined alongside the increase in tractor HP per worker. These patterns are, however, still consistent with Hayami and Ruttan (1985). In Hayami and Ruttan (1985, Figure 5-3), at a lower level of tractor HP per male worker, land-labor ratios are generally flat. Once it exceeds a threshold, the increase in tractor HP per male worker becomes increasingly more positively associated with the land-labor ratio. In the past five decades, while tractor HP per worker has increased considerably in SA countries, its level still has been too low to translate into significant increases in land-labor ratios. However, by the 2010s, many countries had reached the threshold of tractor HP per worker. Therefore, these countries are expected to see growth of the land-labor ratio as tractor HP per worker further increases in the future. As shown in the fourth section, tractor usage for land preparation has reached almost 100 percent in most SA countries (except Nepal), primarily through tractor rentals. Therefore, further increases in tractor HP per worker may accompany substantial reductions in the absolute number of agricultural workers. Figure 4.3 illustrates the process where constraints in agricultural production due to land resource endowments have been mitigated through the increased use of fertilizer (and other yield-enhancing inputs) per unit of land. Figure 4.3 suggests that many SA countries have seen clearer trends in this aspect of land-constraint mitigation as indicated by strongly positive Source: authors' calculations and Fao (2020a). Note: Cv (chevaux vapeur), used in Fao (2020a), is a unit of metric horsepower (hp), approximately equivalent to mechanical horsepower. Cv is assumed to be 10 and 40 on average for two-wheel tractors (2wt) and four-wheel tractors (4wt), respectively, unless otherwise stated. importantly, however, average tractor Cv has generally increased over time, and ignoring that trend could underestimate the overall trend of tractor Cv growth. For Bangladesh and india, interpolation used a constant growth rate (not linear). For nepal, Fao figures seem to closely match with other sources and therefore FaoStat figures were used for 4wt. For 2wt, interpolation used a constant growth rate (not linear); there were also around 1,000 minitillers in 2010, but they were excluded. For pakistan, several studies indicate Cv to be slightly higher than 40. also, information about 2wt in pakistan has been very difficult to obtain. therefore, the tractor Cvs used for pakistan are more conservative estimates. For Sri lanka, information about 4wt has been relatively limited. therefore, the growth rate between 1980 and 2002 for which information is available was applied to extrapolate the growth from 2002.slopes. These patterns seem to have become reinforced since the 1980s. It is interesting to note that curves are convex rather than concave, which indicates that decreasing marginal productivity of fertilizer is more than offset by fertilizer-using technological changes. This is similar to the pattern observed in East Asia.Altogether, the investigations of factor endowments, land and labor productivity growths, and the trends in the land-and labor-substituting input uses have revealed that SA still follows the agricultural growth trajectory hypothesized for land-scarce countries by Hayami and Ruttan (1985). Pre-1980s had been characterized by modest land productivity growth and fairly small labor productivity growth due to the growth of the number of agricultural workers relative to land. However, since the 1980s, the land-labor ratio has become relatively constant and the continuous growth of land productivity seems to have largely translated into the increase in labor productivity. The land productivity growth might have been primarily led by land-saving inputs like fertilizer and other yield-enhancing inputs. The use of power and machinery has continued growing, and SA seems to have reached the point where further increases in power and machinery use per worker may be associated with a significant increase in the land-labor ratio. Key evidence on agricultural TFP growth in SA countries is mostly provided in the literature as part of TFP analyses for a larger set of countries. For SA as a whole, TFP growth had been lower between the 1960s and mid-1980s and turned upward since the mid-1980s, although it has been generally lower than in most other regions throughout the periods since the 1960s, largely due to the slow efficiency growth (despite comparable technical improvements), including the post-1980s period (Nin-Pratt and Yu 2010). TFP growth between 1965 and 1996 was also lower than in East Asia, due to slower efficiency growth (Suhariyanto and Thirtle 2001). Both studies, however, also suggest that since the 1980s, technical changes in SA have been faster than in other parts of Asia, while slower growths of TFP in SA during the last several decades have been largely due to slower technological progress. Coelli and Rao (2005) suggest substantial variations in TFP growth rates within SA between 1980 and 2000. During this period, Bangladesh and Pakistan observed the highest TFP growth, India and Nepal saw medium TFP growth, and Sri Lanka saw the lowest growth.More detailed country-level investigations reveal spatial, regional, and temporal variations. In Indian Punjab, TFP growth would have been considerably higher, about 3 percent per year during the Green Revolution era (1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975) compared with 1 to 2 percent per year in the post-Green Revolution era (1986)(1987)(1988)(1989)(1990)(1991)(1992)(1993)(1994)  (Murgai 2001). TFP growth in the Indo-Gangetic Plain (IGP) region of India was 1.2 percent per year from 1981 to 1996, with more impressive TFP growth during the 1980s than in the 1990s (Kumar, Kumar, and Mittal 2004). TFP growth seems to have been particularly high in early periods. This is potentially because the shift from traditional varieties to modern high-yielding varieties (HYVs), compared with subsequent shifts from the original Green Revolution varieties to the second-and third-generation varieties, might have been much more drastic in terms of the technological progress. Lower TFP growth in the post-Green Revolution era might have been partly due to land quality, including depletion/pollution of groundwater resources, a buildup of soil salinity and waterlogging, nutrient mining, micronutrient deficiencies, deteriorating water quality, the formation of subsoil compaction, and increased pest buildup (Kumar, Kumar, and Mittal 2004). The development of roads and investment in irrigation infrastructure had positive causal effects on TFP growth in rural India between 1971 and 1994 (Zhang and Fan 2004).In India, the TFP level has remained diverse across states (Mukherjee and Kuroda 2003). The TFP growth rates have also varied across regions and over time (Rada 2016), reflecting important dynamics. Since the 1980s, the northern, eastern, central, and northeastern regions have still grown the fastest, due to the remaining effects of the aforementioned Green Revolution technologies. The western and southern regions experienced more rapid growth since the 2000s-partly contributing to their catching up. Rada (2016) speculates that these patterns have been associated with shifts away from grains (commonly grown in the north) to higher-value crops including horticulture and livestock products. Meanwhile, the northern and eastern regions also saw recovery in TFP growth in the 2000s. This had been partly through reduced water use intensity enabled by investments in efficient irrigation methods, including drip irrigation (Rada 2016), which partly mitigated the overexploitation of groundwater in these regions, as described in a later subsection. In the northeastern region, deficiencies in transportation and market infrastructure, among others, relatively suppressed TFP growth even during this period (Rada 2016).Similarly, in Bangladesh the TFP growth between 1948 and 2008 exhibited heterogeneity across regions (Rahman and Salim 2013). Growth had been the fastest in Chittagong (largely due to technical-efficiency gains and some scale-efficiency change). It had been slow in the Chittagong Hill Tracts, largely due to inefficiency in exploiting economies of scope in production achievable through diversifying into the production of crops suitable for hilly environments. According to Rahman and Salim (2013), regional variations in the efficiency in exploiting economies of scope explain significant variations in TFP growth across Bangladesh.Changes in Natural Resources (Water, Soil) The productivity gains in SA agriculture might have been achieved at the cost of key natural resource endowments, including soil (land degradation) and water. Such a decline in natural resource endowments suggests that future land productivity growth in SA faces considerable challenges. FAO (1994) suggests that by the 1990s, land degradation in SA had intensified primarily through erosion of soils by water and wind, the decline of soil fertility, and deterioration of soil quality through waterlogging or salinization, among others. These conditions are likely to have remained since then. FAO (1994) estimates that by the early 1990s, 25 percent of areas under crop and pasture in SA had been affected by water erosion (Table 4.3).2 More than half of the land in Bangladesh and Sri Lanka had suffered soil fertility decline. More than one-third of the areas were affected by different types of land degradation in different countries. Land degradation has become costly in SA. The estimated economic losses due to land degradation in SA by the 1990s were US$5.4 billion from water erosion, $1.8 billion from wind erosion, $0.6-1.2 billion from soil fertility decline, $0.5 billion from waterlogging, and $1.5 billion from salinization-the total being equivalent to 7 percent of agricultural gross domestic product (GDP) then (FAO 1994).SA, which is the most water-scarce region in the world, has also suffered from a declining groundwater table lately. Since 1960, SA has used the most groundwater in the world, and this pattern has accelerated since the 1990s (Shah et al. 2009). Combined with the rising pumping costs due to rising diesel prices, the lowering of the groundwater table has further raised irrigation costs, particularly in the IGP. This has lately led farmers to shift from more water-intensive crops to less water-intensive crops or to rainfed farming systems (Shah et al. 2009).This section describes the key characteristics of agricultural growth, patterns of urbanization, and nutritional transitions in SA, with the aim of highlighting how they may be interlinked with each other, creating unique challenges and opportunities for further agricultural development and food and nutrition security improvements in the region. Source: Data from Fao (1994).agriCultural Development anD moDernization in South aSia 121Unique agricultural growth patterns in SA can be highlighted by the analyses of the growth patterns of value-added, modernization, linkages with domestic and international markets, and diversifications.Agricultural value-added in SA has increased steadily in the past several decades, and the analysis of its growth patterns in comparison with other regions like East Asia or SSA reveals important differences in underlying conditions. The growth rates of agricultural value-added in SA have been 3.2 percent in the past three decades starting in 1985, higher than the rate of 2.4 percent between 1961 and 1985 (Table 4.4). Such acceleration in growth is consistent with the TFP growth patterns described (Nin-Pratt and Yu 2010) in the previous section. However, the growth rate may have been somewhat slower than in East Asia during both periods. This relatively low growth rate of value-added in SA is consistent with the slow labor productivity growth discussed earlier. This is related to persistently high shares of the workforce being employed in the agricultural sector in many SA countries, except in Sri Lanka. While the share has constantly declined in SA in the past three decades (from 62 percent in 1991 to 44 percent in 2017), the speed had been slower than the pace in East Asia, where the share declined from 55 percent in 1991 to 29 percent in 2017 (Table 4.4). While the agricultural sector in SA has undergone substantial modernization in the past few decades, as described in the next section, such modernization might have remained labor-intensive. In other words, SA seems to have been slow in transforming the economic structure away from agriculture.Modernization of agriculture in SA is pronounced in terms of increased adoptions of modern varieties and increased modern inputs use, including chemical fertilizer, machines, and agrochemicals, as well as more systematic uses of water through irrigation (Figure 4.4), although, as is discussed later, use of irrigation on major crops like rice in SA has lagged behind East Asia. Chemical fertilizer use intensity grew substantially up to the 1980s, and it kept growing afterward. By the 2000s, most countries exceeded the intensity of 100 kilograms of nutrients per hectare. The shares of cultivated areas prepared by tractors and irrigated areas have varied more widely across the countries than fertilizer use. Pakistan, Sri Lanka, and India led the growth of tractor uses,  1985 1985 1961-1970 1971-1980 1981-1990 1991-2000 2001-2010 2011 1991  Contrary to the progress in production of staple crops, the modernization of the livestock sector has remained somewhat limited. The dairy sector in SA is still challenged by the limited supply of improved indigenous cattle breeds, limited artificial insemination to provide crossbred cattle, and the generally low quality of crop residue used for most feed resources (McDermott et al. 2010).Modernization at the postharvest stage has been more heterogeneous, with areas closer to major urban markets experiencing faster modernization. The adoption of modern processing facilities has grown gradually around major urban areas. For example, Bangladesh has seen the emergence of large (automatic) rice mills displacing the traditional small huller mills extensively in parts of the Chittagong Division (Reardon et al. 2012). In the country as a whole, however, the majority of paddy is still processed by traditional mills. Recently, demand has shifted from cheap, coarse rice to higher-value (medium and fine) rice, and there are increasing price premiums for the latter in major urban centers like Dhaka, Bangladesh, accompanied by the growing sales by the modern retail sector (Minten, Murshid, and Reardon 2013). The shift for these modern agrifood systems, however, kept modernization to the postharvest stage, and there was relatively little effect on modernization at the farm level (Minten, Murshid, and Reardon 2013). Similarly, branding by the private sector is still often plagued with incompleteness or misinformation (Minten, Singh, and Sutradhar 2013).Domestic production and consumption seem to continue to be dominant sources of supply and demand for agricultural commodities in SA, despite growing external trade. Table 4.5 shows the shares of the net exported quantities as a percentage of the total quantities domestically produced for each group of food commodities (negative values indicate net imports). For example, 12 indicates that the net exports were equivalent to 12 percent of the quantities domestically produced.Table 4.5 indicates that the quantities of cereals, starchy roots, vegetables, fruits, meat, eggs, milk, and fish products that have been exported or imported have remained fairly small compared with the domestic production of these commodities (generally accounting for 20 percent or less). Only pulses and tree nuts, and other food items that may be relatively minor nutrient sources, have seen substantial increases in the contributions of trade. Generally, domestic productions of key agricultural commodities have remained important sources of domestic food and nutrition, while domestic markets have remained the major drivers of agricultural growth, in SA. This pattern is in contrast to some developing countries in SSA, where the reliance on cereal imports has increased substantially in the past few decades.DIVERSIFICATION SA, as a whole, has seen a gradual decline of the share of cereals, and increasing shares for vegetables, and milk, in net production values (Table 4.6). The trends in India, given its dominant size, largely represent the overall patterns in SA. However, there have been considerable variations and divergence over time across countries outside India (Figure 4.5). In Bangladesh, diversification has been relatively limited, with the share of cereals remaining around 60 percent throughout the past five decades, while the shares of higher-value crops have not changed significantly.3 Sri Lanka has actually experienced an increased share of cereals during the last five decades, accompanied by decreasing shares of oil crops, fruits, stimulants (tea in particular), and tobacco/rubber. In contrast, Nepal has seen a rapid shift from cereals and milk to vegetables and fruits and, to a lesser extent, spices. Pakistan, with historically low shares of cereals and higher shares of milk, meat, and fiber crops, has largely maintained the same level of agricultural diversification.This diversification pattern might have been driven by both the changes in demand and deliberate efforts on the supply side. For example, Sri Lanka had been a major rice importer in the region in the 1960s but has gradually reduced import reliance, partly through the government's effort in achieving rice self-sufficiency. Another contributing factor might have been the declining real price of rice and other grains, which is likely to have induced a shift of the production toward high-value crops.Aquaculture production has grown in the last few decades in many SA countries, often 10-fold between the 1980s and the 2010s in gross production value. While the production value is still small compared with the entire agricultural production, it has grown to the equivalent of 20 percent of gross agricultural production in countries like Bangladesh (Table 4.6). The sector has contributed to poverty reduction through reduced price of fish products and their increased consumption by low-income consumers (Toufique and Belton 2014), through improved economic activities of women (Kawarazuka and Béné 2010), and through employment (Paul and Vogl 2011). Its growth has, however, also caused environmental damage, including mangrove  1961-1970 1971-1980 1981-1990 1991-2000 2001-2010  degradation, saltwater intrusion, sedimentation, pollution, and disease outbreaks (Paul and Vogl 2011).The diversification patterns have also been associated with cross-country variations in productivity growth, particularly the slower growth in Sri Lanka than in other SA countries described earlier. Table 4.7 summarizes the output growth between the 1980s and the 2010s by commodity groups in Sri Lanka and other SA countries. Output growth for vegetables and fruits has been particularly slow in Sri Lanka (0.3 percent per year) compared with SA as a whole (3.6 percent per year). In other SA countries like India, Nepal, and Pakistan, while rice output growth rates have been similar to that of Sri Lanka, nonrice commodities have grown by 3 percent or more (for example, milk in India and Pakistan; meat and cotton in Pakistan; meat, vegetables, and fruits in Nepal). In Bangladesh, rice output growth has been faster, at 2.8 percent per year. 1961-1970 1971-1980 1981-1990 1991-2000 2001-2010  agriCultural Development anD moDernization in South aSia 127In relative terms, the pace of urbanization in SA seems to have been slow compared with both growth in the East Asia and Pacific region and the historical experiences of developed countries. Urbanization levels are the lowest in Nepal (19 percent) and Sri Lanka (18 percent), while the Maldives (47 percent), Bhutan (39 percent), and Pakistan (39 percent) are the most urbanized countries in the region (Table 4.8). The urban population accounted for 35 percent in Bangladesh and 33 percent in India. The agricultural growth and modernizations in SA described in the previous section, therefore, seem to have largely taken place in the rural sector. However, in absolute terms, SA's urban population growth has been and is likely to be significant. The urban population increased from about   4.8). Urbanization still has been one of the drivers of dietary diversifications in SA in recent years (Joshi, Gulati, and Cummings 2007).Urbanization might have also posed certain challenges to food security. The 2007-2008 food crisis demonstrated the vulnerability of urban Proportion of gross aquaculture production to gross agricultural production in values (agricultural production value = 100)Source: authors' calculations based on Fao (2020a), Fao (2019) for aquaculture production values, and Fao (2020a) for commodity groups.Note: gross aquaculture production value does not include capture fishery. the \"proportion\" simply compares the sizes and does not mean the share of aquaculture production to agricultural production, since the latter figure does not include the former. -= data not available.populations, especially slum dwellers, to shocks in agricultural markets (OECD-FAO 2010). Large urban settlements in SA are marked by widespread slums. The share of the urban population living in slums in SA is high (except in Bhutan and Sri Lanka), ranging from 17 percent in India to 89 percent in Afghanistan (Ellis and Roberts 2016). At least 130 million people were living in informal urban settlements in SA in 2009 (ADB 2014). Slum populations often do not have access to water and sanitation facilities, making them more likely to suffer from disease and malnutrition. At the same time, where it has occurred, urbanization might have been associated with the dramatic increase in the proportion of dietary energy derived from oils and fats and considerable increases in the consumption of fruits, vegetables, and dairy and other animal products. Food consumption patterns have been changing across regions, with consumption of noncereal crops growing in both rural and urban areas.Rural and urban food consumption patterns are not uniform across SA, but some common trends emerge. Food accounts for a smaller share of consumption expenditure in urban areas than in rural areas, and urban households have more diverse diets than rural households (Joshi, Parappurathu, and Kumar 2016).SA seems to be going through a nutrition transition (Chapter 10)-defined as the process of changes in the food environment, physical activity, and lifestyle that result in declining levels of undernutrition and increasing levels of overnutrition over time (Popkin 1993). Reduced staple crop prices in countries like India and Bangladesh, realized through investments in public research and development, have led to significant reductions in poverty, and increases in disposable incomes have led to increased dietary diversification through consumption of higher-value nonstaple commodities (Pingali 2012). Considerable progress has also been made on improving nutritional outcomes. For instance, Nepal has recorded the fastest reduction in child stunting in the world recently, from 56 percent in 2001 to 36 percent in 2016 (Nepal, Ministry of Health 2016). Despite such transitions and progress, challenges remain. SA remains the global hub of nutrition insecurity. In 2016, two out of every five of the world's stunted children lived in SA, and more than 15 percent of children under age 5 were wasted. In fact, stunting levels in SA (38 percent) were comparable to those in SSA (37 percent) and over three times higher than in East Asia and the Pacific (12 percent) or Latin America (11 percent).Investments into staple crops are also said to have crowded out the required investments for traditional crops that were important sources of critical nutrients, such as legumes and pulses (Pingali 2012). Consequent increases in their prices in India are considered to have reduced the consumption of these commodities (Pingali 2012).South Asia has also seen a rapid emergence of overnutrition (Figure 4.6). Obesity reached 20 percent in India and 13 percent in Nepal by 2013. In India, overnutrition and associated noncommunicable diseases have also gradually spread from urban to rural areas lately (Meenakshi 2016). In SA, about one in four adults are currently overweight.This section highlights key technological and institutional innovations and key policies that might have partly enabled the agricultural growth in SA described in the previous sections. The section also discusses that these innovations have so far been associated with labor-intensive agricultural transformation in SA. Despite the gradual shift in the comparative advantages from smallholders to larger farms, such effects might not have been strong enough  , 2013;Bangladesh, 2014;Bhutan, 2010;india, 2015;maldives, 2009;nepal, 2016;pakistan, 2012;Sri lanka, 2012;and South asia, 2015. poverty head count ratio and undernourishment are given as percentage of population; stunting figures pertain to children under 5 years of age. the prevalence of overweight data are all from 2013.to overcome constraints for farm size expansions. Similarly, despite the growing wage gap between the agricultural sector and the nonagricultural sector, the former has remained one of the primary sources of rural employment.SA has seen considerable progress in technological and institutional innovations.For varietal technologies, public-sector-led conventional plant breeding has continued producing newer generations of improved varieties (Hossain et al. 2003). Genetic improvement efforts had been significant, not only for widely documented breeding for staple crops, but also for vegetables-including, for example, application of molecular-based methods for breeding virus-resistant tomatoes in southern India (Weinberger and Lumpkin 2007).Over the years in the region, the public sector has accumulated plant-breeding technologies and knowledge, as well as a supply of superior germ plasms, which have raised the potential of developing hybrid varieties. At the same time, there has been increased participation of the private sector in agricultural research and development over the past few decades. Recently, hybrid varieties have been developed for maize, rice, and wheat (Matuschke, Mishra, and Qaim 2007), as well as pearl millet (Matuschke and Qaim 2008). The adoption rates of hybrid maize and hybrid rice have grown considerably. Furthermore, the increased applicability of biotechnologies (including genetic engineering) has induced private-sector participation, which has resulted in the successful development of Bt cotton and vegetable seed developments (see, for example, Krishna and Qaim 2007). Strengthening intellectual property rights is also likely to have accelerated private-sector participation (Pray and Nagarajan 2014). In countries like India, this progress has coincided with the deregulation of the seed sector that started in the 1990s (Pray and Nagarajan 2014).Mechanical tools have spread widely in SA over the past few decades, but possibly in ways not displacing substantial labor force in agriculture, which is appropriate in raising labor productivity in the labor-abundant environment. This has been achieved by the spread of various types of small-scale machines, including mechanical water-lifting tools, tens of millions of which are in SA (Mandal, Biggs, and Justice 2017). Since the 1990s, the use of smallscale shallow tube wells and motorized pumps has increased the extraction of groundwater and surface water in Bangladesh, where importation of shallow tube wells was liberalized in the 1990s (Ahmed 1995), as well as Nepal and the IGP of India (Shah et al. 2009). Power tillers have spread as well for land preparation and local transportation. At the same time, four-wheeled, riding tractors that have been more common in India and Pakistan might have been more suitable in rice-nonrice crop rotation common in parts of SA like India, and mechanization might have led to significant area expansions, sustaining the overall demand for labor (Pingali 2007). The growth of the domestic manufacturing industry, such as in India, facilitated the development of small four-wheeled tractors that are more suitable on smallholder-dominated upland in the region (Pingali 2007). These mechanization patterns may explain why the land-labor ratio did not increase with the increases in tractor HP per worker shown in Figure 4.2.SA has seen various institutional innovations in the past few decades. The growth of market institutions has been observed in the growing participation of new actors in processing, trading, and wholesale for both staple crops like rice and potato (Reardon et al. 2012) and nonstaple crops (Fafchamps, Hill, and Minten 2008). Vertical integration has also been strengthened for crops like sugarcane in India (Patlolla, Goodhue, and Sexton 2015). However, constraints remain for food sanitation along the value chain for vegetables/fruits and dairy value chains in India (Fafchamps, Hill, and Minten 2008;Kumar, Wright, and Singh 2011). Markets for minor crops have also expanded. Increased market participation has had positive effects on on-farm varietal diversity for millet in India (Takeshima and Nagarajan 2012). Marketing institutions including contract farming and cooperatives have become increasingly important for production and marketing of high-value vegetables and fruits, poultry, and higher-quality milk, which has led to the adoption of improved food safety practices (Joshi, Gulati, and Cummings 2007;Kumar et al. 2018), as will be discussed in Chapter 11 in depth.Market institutions have also developed regionally. The South Asian Free Trade Area (SAFTA) agreement in 2004 potentially increased market integration across SA countries. Trade liberalization in the early 1990s might have improved the resilience of countries like Bangladesh against food crises caused by natural disasters like floods (del Ninno and Dorosh 2001).Institutional innovations have also been widespread for common resource management. Transfer of forest use rights to local communities reduced forestry resource extraction in Nepal (Edmonds 2002). In India, increased demand for wood products has led to increased forestry cover, reversing the trend of deforestation (Foster and Rosenzweig 2003). Community irrigation systems have also grown (Bardhan 2000). Environmental efficiency has also been raised by new production technologies, such as Bt cotton that has reduced chemical pesticide use in Pakistan (Kouser and Qaim 2015).The increased stocks of physical infrastructure (for roads, electricity, information and communications technology, etc.) as well as institutional infrastructure have been important drivers of agricultural diversification away from traditional low-value crops to higher-value crops (Joshi, Gulati, and Cummings 2007). Better road infrastructure has often reduced the transaction costs for marketing perishable commodities like vegetables and fruits (Joshi, Gulati, and Cummings 2007). Infrastructure for veterinary institutions and artificial insemination centers for livestock was likely to have encouraged diversification in the production of livestock products (Joshi, Gulati, and Cummings 2007).Many SA countries, including Bangladesh, India, and Pakistan, have expanded key infrastructure during the past several decades (Table 4.9). Paved roads expanded by more than five times between the 1970s and 2000s in all three countries. The number of ground line telephones multiplied by almost 5 in Bangladesh and almost 10 in India and Pakistan during the same periods. The public expenditures on infrastructure during these periods have been substantial.The series of innovations are also likely to have been partly induced by the economic and macro policy reforms introduced in many SA countries in the past few decades. Many SA countries, including India, Pakistan, and Sri Lanka, have gradually reduced the distortions of their exchange rates, which had typically been fixed and overvalued (Sahoo, Nataraj, and Dash 2013). India since the 1980s and Pakistan throughout the 1980s saw significant devaluations that helped raise the competitiveness of some of the export crops. Unlike some other developing countries (particularly those in SSA), these SA countries have managed to stabilize the exchange rates even after the shift to the floating exchange rate system. For example, the real exchange rate in Pakistan remained largely unchanged from the late 1980s to the 2010s (Spielman et al. 2017).Investment policies have also been reformed in many SA countries. For example, in 1997, Pakistan allowed foreign direct investment (FDI) in the agricultural sector, though it had restricted FDI to the manufacturing sector in the previous period (Sahoo, Nataraj, and Dash 2013). Similarly, in Nepal, the share of FDI flowing into the agricultural sector started increasingly, from almost zero in 2000 to 4 percent in 2010 (Sahoo, Nataraj, and Dash 2013).However, the extent of the effectiveness of macro policies in the agricultural sector remains uncertain. While trade liberalization policies have significantly reduced the tariff rates for nonagricultural goods, average tariffs for agricultural goods have remained higher, around 40 percent in India and 30 percent in Sri Lanka (Sahoo, Nataraj, and Dash 2013). While SA countries have seen increased shares of tax revenues originating from income tax and value-added tax, rather than seigniorage and tariff (Aizenman and Jinjarak ), the taxation of agricultural incomes appears to have remained incomplete, especially in India and Pakistan.SA countries have also supported provisions of institutional credit, insurance, and agricultural inputs. For example, India has made efforts to allocate credit to the agricultural sector by increasing the overall volume of credit that goes to the sector, by waiving institutional debt for small farmers and allowing onetime settlement, and through interest-subvention schemes that reward prompt repayment by borrowers (Narayanan 2016). Consequently, in India the flow of formal-sector agricultural credit has been consistently increasing, with the ratio of agricultural credit to agricultural GDP rising from 10 percent in 1999-2000 to about 38 percent in 2012-2013 (Kumar et al. 2017), and the share of credit-financed portion out of the spending on agricultural inputs increasing from 21 percent in 1995/1996-2003/2004to 69 percent in 2004/2005-2011/2012(Narayanan 2016)). Institutional credit has been positively associated with agricultural development in various ways in countries like India (see, for example, Narayanan 2016). Between the mid-1990s and early 2010s, greater credit flows had been associated with the increased use of chemical fertilizer, use of pesticides, or investments into agricultural capital like tractors. At the same time, the effects of credit on agricultural GDP, efficiency, and productivity have been relatively limited (Narayanan 2016). These patterns suggest that institutional credit might have been more associated with the increase in agricultural productivity where capital inputs (purchased by credit) offered higher productivity than noncapital inputs (see Chapter 16 for further discussion). In Bangladesh, formal credit for machinery purchase has been increasingly provided by the importers that have comparative advantages in monitoring and loan payment collections, rather than directly through the bank. Khandker (2007) found that microfinance institutions in Bangladesh enhanced flood-affected households' access to finance and thereby played a central role in enhancing their coping ability after devastating floods.The uptake of agricultural insurance has been rising, albeit slowly, in India. By 2016, 57 million farmers in India had been covered by some type of insurance (Gulati, Terway, and Hussain 2018). The cost-effectiveness of these subsidized insurance programs remains ambiguous with limited empirical evidence (see Chapter 17). In parts of India, indexed insurance increased investments into potentially higher-return cash crops that are more sensitive to rainfall (Cole et al. 2013). In parts of Bangladesh, indexed insurance significantly enhanced risk-management effects and expanded cultivated area and spending on agricultural inputs (Hill et al. 2019). However, further studies are needed on the effects of agricultural insurance on agricultural productivity, profitability, and farm incomes, for which evidence remains thin.Among input subsidies, fertilizer subsidies often accounted for 2-3 percent of total public expenditures and often more than 30 percent of agricultural expenditures in Bangladesh, India, and Pakistan between 1980and 2010(Rashid et al. 2013). These fertilizer subsidies were effective in reducing fertilizer prices during this period-as much as 30 percent in India, 9 percent in Bangladesh, and 3 percent in Pakistan (Rashid et al. 2013). However, it is also important to note that these subsidies also led to significant distortions of fertilizer markets. Combined with exchange rate distortion, fertilizer prices often deviated substantially from international prices (Rashid et al. 2013). The subsidies have also been rather regressive; that is, larger farms generally benefited more than smallholders. Furthermore, fertilizer subsidies have also been considered partly responsible for various environmental damages caused by their overuse, including groundwater pollution in Bangladesh (Alauddin and Quiggin 2008) and India (Rashid et al. 2013), as well as air pollution in India (Aneja et al. 2012).The average farm size in SA has been declining over the past several decades. This trend has been consistent across many SA countries, including Bangladesh and Nepal with very small farm sizes (less than 1 hectare on average), India with slightly larger farm sizes (around 1 hectare on average), and Pakistan with modestly larger farm sizes (around 3 hectares on average) (Table 4.10).Declining farm sizes and the process of land fragmentation in SA countries have been partly influenced by government policies in these countries, including the law of inheritance of paternal property, lack of progressive tax on inherited land, heterogeneous land quality, and an underdeveloped land market (Niroula and Thapa 2005). Pakistan has imposed a progressive land tax, where larger farms are taxed at a higher rate than smaller farms (Adamopoulos and Restuccia 2014). In India and Nepal, a land reform program called the land-to-the-tiller program was implemented, albeit with a limited implementation capacity, whereby the land was transferred from large landholders to tenants who actually cultivate the land, and ceilings were set for the land ownership.Evidence suggests potentially negative effects of declining farm size, as well as fragmentation, on productivity and efficiency. Adamopoulos and Restuccia (2014) estimated that the aforementioned tax policies in Pakistan reduced average farm size and agricultural productivity by 3 percent. The aforementioned policies on the land-to-the-tiller program in India and Nepal discouraged optimal land concentration that would achieve economies of scale in production (Otsuka, Liu, and Yamauchi 2016).The negative productivity and efficiency effects of declining farm size have also been aggravated by accelerating land fragmentation. Land fragmentation was found to increase the cost of cultivation in India (Deininger 2017)  and had negative effects on rice productivity and efficiency in Bangladesh (Rahman and Rahman 2009). Niroula and Thapa (2005) conclude, based on reviews of the literature in SA, that fragmentation of small landholdings and tiny land parcels have been detrimental to land conservation and economic gain and that they have discouraged farmers' adoption of agricultural innovations. Meanwhile, relations between farm size and productivity have been changed by growing adoptions of mechanization in SA in the last few decades. Even though the machine rental markets providing hiring services have been relatively efficient and have managed to enable not only larger farmers but also smallholders to adopt mechanization, mechanization, particularly recent large-scale mechanization, has shifted the comparative advantage to larger farms. In Nepal, the adoptions of mechanical technologies (tractors, in particular) have directly raised the returns to scale in agriculture, shifting the comparative advantage from smallholders to larger farms (Takeshima 2017a). Aside from mechanization, Deininger et al. (2018) argue, the labor market imperfections in India have dissipated over time, and this too has weakened the inverse relationship between farm size and output per area. According to them, previously family labor was overused for farming because they were rationed out of off-farm labor markets, which were less functioning, and the extent of overuse was greater among smaller farms with greater family labor endowments. In the last few decades, however, an increasingly functioning labor market has started absorbing more family labor into the off-farm market, while reducing on-farm family labor use to a more optimal level.With labor market transformation, increased infusion of capital into the agricultural sector in the form of mechanization, and the resulting shift in comparative advantages toward larger farms, the institutional failures in developing efficient land markets are now becoming increasingly important bottlenecks in SA.In SA, the agricultural sector has remained the major source of rural employment in the past few decades, despite the gradual expansion of the nonfarm sector. The persistently high contribution of the agricultural sector as the source of rural employment is consistent with the presentations in earlier sections, where labor use per hectare has remained unchanged, even though labor productivity in agriculture has risen substantially in many SA countries. In India, the agricultural sector has long competed with the industrial and service sectors for workers. Where labor mobility has been low, the industrial sector has often moved to areas with lower agricultural productivity and thus lower wages (Foster and Rosenzweig 2004). However, rural labor mobility was not always low. In Nepal and Pakistan, the adoption of modern varieties in favorable areas benefited marginal areas through increased labor demand and migration (Upadhyaya, Otsuka, and David 1990;Renkow 1993).Rural employment in the agricultural sector has remained high and, in some cases, has increased in absolute terms. The crop sector rather than the livestock sector has often been the major source of agricultural employment, as in, for example, Pakistan (Spielman et al. 2017). New agricultural technologies in SA, particularly in India, have been relatively more skill intensive, raising the returns to primary schooling (Foster and Rosenzweig 1996) and thereby raising the demand for moderately educated labor. Output growth has increased demand for harvesting labor, even for labor-saving varieties like Bt cotton in India (Subramanian and Qaim 2009). Such growth for harvesting labor is, however, also a reflection of the fact that wages are still low, preventing the spread of mechanical harvesters that has been widely observed in East Asia lately. Once the wage level rises to the critical threshold, rural employment in SA is likely to shift considerably to the nonfarm sector.In SA, the share of rural employment in the nonfarm sector has remained low and has just started rising gradually. Table 4.11 compares the shares of rural employment among different economic activities in the five SA countries in recent decades. Direct comparisons across countries or over time should be made with caution, due to the differences in the classifications of activities, measurements, and periods. However, these figures collectively provide useful insights.The construction sector has been one of the fastest-growing sources of rural employment among the nonfarm sector, particularly in India and Nepal. To a lesser extent, rural employment in the manufacturing of farm products has grown in parallel with the overall farm output growth. The rice-milling sector, which is still dominated by small-scale mills in India and Bangladesh, and potato cold storage facilities in Bangladesh have provided significant sources of employment despite relatively high capital intensity (Reardon et al. 2012).4 Employment growth in the manufacturing of nonfarm products has been driven by textile industries. For example, the garment industry has grown considerably in Bangladesh, and by 2007-2008, 2.5 million workers had been employed in the sector (Mottaleb and Sonobe 2011). The growth of the Bangladesh garment industry has been enabled primarily by international transfer of production technologies and skills in management and international marketing, coupled with enhanced education levels (Mottaleb and Sonobe 2011). Furthermore, Bangladesh succeeded in creating an enabling environment for garment-sector growth by maintaining stability and predictability of garment-industry policies, rather than focusing on heavier interventions that may be more substantial in scope but can also bring significant unpredictability from the private sector's perspectives (Ahmed, Greenleaf,  Deininger, Jin, Sur (2007). Note: For Bangladesh, \"transport\" includes \"commerce\"; \"other services\" includes \"personal, financial, and community services\"; \"construction\" includes \"construction and utilities mining.\" transport includes \"transport and commerce.\" For india, \"construction\" includes \"mining\"; \"trade\" includes \"trade, hotel, and restaurant.\" \"other services\" includes \"electricity, gas, and water supply; finance; real estate, etc.; and public administration and other services.\" For nepal, figures for the urban and rural sectors, as well as male and female, and across agroecological belts in takeshima (2017b) are combined using the population weights. and Sacks 2014). Where these conditions exist, export-oriented manufacturing may grow substantially even in the rural areas in SA (see Chapter 11).Other major sources of rural employment include government services. In the 1990s, government jobs provided about 25 percent of rural nonfarm earnings in rural Pakistan and nearly 20 percent of rural nonfarm employment in rural India (Haggblade, Hazell, and Reardon 2010).The expansion of public works social protection policies has had certain effects as well on rural employment in SA. A notable example is India's Rural Employment Guarantee program, which has offered employment on local public works to any rural adult residents demanding work for up to 100 days. This program has been one of the largest public works schemes and had covered about 30 percent of rural households in India from 2009 to 2015, generating more than 2 billion person-days of work per year from 2008 to 2014 (Drèze and Khera 2017). Although there are still issues with these programs, including corruption and the lack of accountability in program implementation, the program has been found to significantly reduce lean-season poverty, to increase consumption particularly among scheduled castes (Bose 2017;Drèze and Khera 2017), and to increase wages (Merfeld 2019).Despite the agricultural and rural development described in SA in this chapter, the region has still lagged behind Northeast Asia (such as the Republic of Korea) and some countries in Southeast Asia in terms of structural transformation. This is despite the fact that these countries had started with similar factor endowments and level of economic development in the 1950s and had similar access to the Green Revolution technologies. While it is difficult to pinpoint exact factors that have led to such divergence, two potential factors are irrigation and education.The difference in the use of irrigation for Green Revolution technologies between SA and East Asia might have been responsible for the differences in productivity growth between these regions. In Bangladesh and India, which account for the majority of the rice area in SA, the share of the total rice area that was irrigated remained lower than in many East Asian countries, and this has been consistent with differential rice yields in these regions (Figure 4.7). Often, irrigation had been the most important factor affecting the adoption of improved rice varieties in Asia during the Green Revolution (Estudillo and Otsuka 2012). This was because Green Revolution technologies had been developed more for irrigated areas, while the development of varieties for rainfed areas came much later. Importantly, however, SA countries have also been catching up with East Asia. In the first 30 years of the country's independence, Bangladesh invested US$5 billion (in real terms) in large-scale irrigation projects, in part through the financial loans from the development partner (Rashid et al. 2013). Similarly, India invested more than $9 billion in public irrigation projects between 1971 and 1995 (Rashid et al. 2013). These investments may help narrow the transformation gap between SA and East Asia.Slower labor movement out of the agricultural sector in SA might have been partly associated with a historically lower human capital level in the region. Average education levels among adults have remained lower than those in East Asia over the last half century, despite the gradual increases during this period (Figure 4.8). These differences are likely to have affected the overall economic growth in these regions. Siddiqui and Rehman (2017) suggest that educational differences largely affected the growth difference between East Asia and SA. In India and China, education has had similar effects on growth, and thus the gap in education between these two countries has accounted for the gap in economic transformation (Bosworth and Collins 2008). SA has one of the highest labor-land ratios in the world. However, the region seems to have followed the Asian-type agricultural growth path characterized by Hayami and Ruttan (1985). Since the 1980s, even the most land-scarce countries, like Bangladesh and Nepal, seem to have started reversing the previous trend of declining labor productivity up to the 1970s. The region appears to be finally shifting toward the beginning of the transformation process with greater increases in labor productivity than land productivity, which many advanced countries have experienced in the past. The underlying technology landscape seems to have shifted as well; mechanization growth in the past three decades has been considerable and seems to have reached the stage where further increases in power and machinery use are associated with a significant increase in the land-labor ratio. This is consistent with the growing evidence in the region that smaller farms are becoming inefficient relative to the larger farms. Institutional failures in developing land markets can become increasingly important bottlenecks. Land productivity growth in SA has followed diverse patterns, with some countries shifting from cereals to higher-value vegetables and fruits, some Other countries in East Asia countries shifting from traditional export crops to cereals, and some countries shifting from traditional cereals to higher-value cereals. Yet, there has been a convergence of agricultural output per unit of land within SA. Technological innovations, ranging from the adoption of machines and chemicals to improved varieties, have been led by both the public and the private sector in SA. Public support and institutional innovations seem to be associated with expanded provisions of agricultural finance and insurance, as well as collective actions in milk production, improved food safety practices, and management of common resources including irrigation facilities and forests. Liberalization and regional integration might have facilitated the trade growths of agricultural commodities within the region, and they have potentially improved national resilience against food crises. SA, however, still faces various emerging challenges. Natural resources, including water resources, continue to degrade. Nutritional improvements still seem to lag behind agricultural growth. The varying agricultural growth might also have important implications for inequality, between urban or peri-urban areas experiencing fast modernization and rural areas where the low-paying agricultural sector remains the dominant source of employment. Raising the labor productivity in agriculture is likely to be critical in SA, and it will require, among other things, careful balancing of support for rural nonfarm employment, facilitating labor movement from farm to the nonfarm sector, and continued support for the farm sector in ways that lead to productivity enhancement without resource degradation.","tokenCount":"9399","images":[],"tables":["-1476896660_1_1.json","-1476896660_2_1.json","-1476896660_3_1.json","-1476896660_4_1.json","-1476896660_5_1.json","-1476896660_6_1.json","-1476896660_7_1.json","-1476896660_8_1.json","-1476896660_9_1.json","-1476896660_10_1.json","-1476896660_11_1.json","-1476896660_12_1.json","-1476896660_13_1.json","-1476896660_14_1.json","-1476896660_15_1.json","-1476896660_16_1.json","-1476896660_17_1.json","-1476896660_18_1.json","-1476896660_19_1.json","-1476896660_20_1.json","-1476896660_21_1.json","-1476896660_22_1.json","-1476896660_23_1.json","-1476896660_24_1.json","-1476896660_25_1.json","-1476896660_26_1.json","-1476896660_27_1.json","-1476896660_28_1.json","-1476896660_29_1.json","-1476896660_30_1.json","-1476896660_31_1.json","-1476896660_32_1.json","-1476896660_33_1.json","-1476896660_34_1.json","-1476896660_35_1.json","-1476896660_36_1.json","-1476896660_37_1.json","-1476896660_38_1.json","-1476896660_39_1.json","-1476896660_40_1.json","-1476896660_41_1.json","-1476896660_42_1.json"]}

data/part_2/0393777159.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"f9957dcd8b28429de93b445e0f80442a","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/cb39344f-d2e0-4320-a377-365e5558d996/retrieve","description":"","id":"-1808362931"},"keywords":["Climate change","adaptation","Southern Africa"],"sieverID":"c117492d-8e28-404f-9c5d-96143047a27e","pagecount":"40","content":"2007, the Discussion Paper series within each division and the Director General's Office of IFPRI were merged into one IFPRI-wide Discussion Paper series. The new series begins with number 00689, reflecting the prior publication of 688 discussion papers within the dispersed series. The earlier series are available on IFPRI's website at www.ifpri.org/pubs/otherpubs.htm#dp.2 IFPRI Discussion Papers contain preliminary material and research results. They have not been subject to formal external reviews managed by IFPRI's Publications Review Committee but have been reviewed by at least one internal and/or external reviewer. They are circulated in order to stimulate discussion and critical comment.The International Food Policy Research Institute (IFPRI) was established in 1975. IFPRI is one of 15 agricultural research centers that receive principal funding from governments, private foundations, and international and regional organizations, most of which are members of the Consultative Group on International Agricultural Research (CGIAR).IFPRI's research, capacity strengthening, and communications work is made possible by its financial contributors and partners. IFPRI gratefully acknowledges generous unrestricted funding from Australia, Canada, China, Denmark, Finland, France, Germany, India, Ireland, Italy, Japan, the Netherlands, Norway, the Philippines, Sweden, Switzerland, the United Kingdom, the United States, and the World Bank.The climate of southern Africa is highly variable and unpredictable and the region is prone to extreme weather conditions, including droughts and floods (DFID 2004;Kinuthia, 1997). Climate change with expected long-term changes in rainfall patterns and shifting temperature zones are expected to have significant negative effects on agriculture, food and water security and economic growth in Africa; and increased frequency and intensity of droughts and floods is expected to negatively affect agricultural production and food security (DFID 2004;Kinuthia, 1997). According to DFID ( 2004) climate change will result in northern and southern latitudes getting drier while the tropics are expected to become wetter.Moreover, climate variability is expected to increase with increased frequency and intensity of extreme weather conditions in Africa. The implications for southern Africa are that the region would generally get drier and experience more extreme weather conditions, particularly droughts and floods, although there would be variations within the region with some countries experiencing wetter than average climate.Agricultural production remains the main source of livelihoods for most rural communities in developing countries and sub-Saharan Africa in particular. Here, agriculture provides a source of employment for more than 60 percent of the population and contributes about 30 percent of Gross Domestic Product (GDP) (Kandlinkar and Risbey 2000). Climate change will have greater negative impacts on poorer farm households as they have the lowest capacity to adapt to changes in climatic conditions. Adaptation measures are therefore important to help these communities to better face extreme weather conditions and associated climatic variations (Adger et al. 2003). Adaptation has the potential to significantly contribute to reductions in negative impacts from changes in climatic conditions as well as other changing socioeconomic conditions, such as volatile short-term changes in local and international markets (Kandlinkar and Risbey 2000). Therefore, an analysis of adaptation options and constraints to adaptation is important for the agricultural communities of southern Africa.A better understanding of farmer perceptions regarding long-term climatic changes, current adaptation measures and their determinants will be important to inform policy for future successful adaptation of the agricultural sector. This paper provides insights on farmer perceptions regarding changes in climate, adaptation options and their determinants as well as barriers to adaptation.A number of economic impact assessment studies in southern Africa use the Ricardian crosssection approach for measuring impacts of climate change on agriculture, including Mano and Nhemachena (2006) for Zimbabwe; Gbetibouo and Hassan (2005) and Benhin (2006) for South Africa; Jain (2006) for Zambia. The advantage of using this approach is that it incorporates adaptation in the analysis of impacts of climate change. The cross-sectional Ricardian model implicitly assumes that farmers are rational and adapt to changes in climatic conditions in their decision making process. The limitation of this approach in analyzing adaptation is that the underlying assumptions that \"historical choices made in the market implicitly map agricultural (and other sectoral) outputs to climate variables\" fails to explicitly model adaptation in the agricultural sector, (Kandlinkar and Risbey 2000). This study addresses this limitation by using a multivariate discrete choice response model to analyze adaptation in three countries of southern Africa: South Africa, Zambia and Zimbabwe.To our knowledge, no studies published to date investigated the determinants of farm-level adaptation options to climate change in the context of southern Africa. Understanding the determinants of household choice of adaptation options can provide policy insights for identifying target variables to enhance the use of adaptation measures in agriculture. Maddison (2006), using the data set also used in this study, did not distinguish the determinants underlying each individual, potential adaptation option.Instead, he aggregated adaptation measures into two options of whether a farmer adapts or not. The decision of not adapting was then used in a sample selection Heckman model to analyze the determinants of not adapting to changes in climatic conditions. Other studies that analyzed adaptation using the same GEF/WB/CEEPA data set considered single adaptation options focusing mainly on climate related factors (Kurukulasuriya and Mendelsohn 2006a;2006b andSeo andMendelsohn 2006).This study adds to these analyses by distinguishing household and other socioeconomic factors affecting propensity of use of each of the main adaptation measures available to farmers. The distinguishing feature is that it uses a multivariate discrete choice econometric model to simultaneously examine the relationships between each adaptation option and a common set of explanatory variables.The advantage of using this approach as opposed to univariate (single-equation) technique is that it explicitly recognizes and controls for potential correlation among adaptation options and therefore provides more accurate estimates of relationships between each adaptation option and its explanatory variables. The univariate technique on the other hand is prone to biases due to common factors in situations where there are unobserved and unmeasured common factors affecting the different adaptation options.The primary objective of this study is to develop and apply empirical methods to assess farmers' adaptation in southern Africa. The specific objectives are 1) to identify farmers' perceptions towards climate change adaptation measures taken; 2) to identify the determinants of farm-level adaptation strategies to changing climatic conditions; and 3) to identify alternative adaptation measures that countries in southern Africa can employ to stabilize national and regional food security in the face of anticipated changes in climatic conditions.The next section presents a brief review of the literature on adaptation to climate change in agriculture. Section 3 reviews some empirical adaptation studies and the model and data sources are presented in section 4. The empirical results and discussion are presented in section 5 and the last section presents conclusions and implications for policy.Adaptations are adjustments or interventions, which take place in order to manage the losses or take advantage of the opportunities presented by a changing climate (IPCC 2001). Adaptation is the process of improving society's ability to cope with changes in climatic conditions across time scales, from short term (e.g. seasonal to annual) to the long term (e.g. decades to centuries). The IPCC (2001) defines adaptive capacity as the ability of a system to adjust to climate change (including climate variability and extremes), to moderate potential damages, to take advantage of opportunities, or to cope with the consequences. The goal of an adaptation measure should be to increase the capacity of a system to survive external shocks or change. The assessment of farm-level adoption of adaptation strategies is important to provide information that can be used to formulate policies that enhance adaptation as a tool for managing a variety of risks associated with climate change in agriculture.Important adaptation options in the agricultural sector include: crop diversification, mixed croplivestock farming systems, using different crop varieties, changing planting and harvesting dates, and mixing less productive, drought-resistant varieties and high-yield water sensitive crops (Bradshaw et al. 2004). Agricultural adaptation involves two types of modifications in production systems. The first is increased diversification that involves engaging in production activities that are drought tolerant and or resistant to temperature stresses as well as activities that make efficient use and take full advantage of the prevailing water and temperature conditions, among other factors. Crop diversification can serve as insurance against rainfall variability as different crops are affected differently by climate events (Orindi and Eriksen 2005;Adger et al. 2003). The second strategy focuses on crop management practices geared towards ensuring that critical crop growth stages do not coincide with very harsh climatic conditions such as mid-season droughts. Crop management practices that can be used include modifying the length of the growing period and changing planting and harvesting dates (Orindi and Eriksen 2005).Use of irrigation has the potential to improve agricultural productivity through supplementing rainwater during dry spells and lengthening the growing season (Baethgen et al. 2003;Orindi and Eriksen 2005). It is important to note that irrigation water is also subject to impacts from climate change. Use of irrigation technologies need to be accompanied by other crop management practices such as use of crops that can use water more efficiently. Important management practices that can be used include: efficient management of irrigation systems, growing crops that require less water, and optimizing of irrigation scheduling and other management techniques that help reduce wastage (Loё et al. 2001).Adaptation occurs at two main scales: (a) the farm-level that focuses on micro-analysis of farmer decision making and (b) the national level or macro-level that is concerned about agricultural production at the national and regional scales and its relationships with domestic and international policy (Bradshaw et al. 2004;Kandlinkar and Risbey 2000). Micro-level analysis of adaptation focuses on tactical decisions farmers make in response to seasonal variations in climatic, economic, and other factors. These tactical decisions are influenced by a number of socioeconomic factors that include household characteristics, household resource endowments, access to information (seasonal and long-term climate changes and agricultural production) and availability of formal institutions (input and output markets) for smoothening consumption, (Figures 1 and2). Farm-level decision making occurs over a very short time period usually influenced by seasonal climatic variations, local agricultural cycle, and other socio-economic factors.Macro-level analysis on the other end focuses on strategic national decisions and policies on local to regional scales taking into account long term changes in climatic, market and other conditions over longtime periods (Bradshaw et al. 2004;Kandlinkar and Risbey 2000). The level of analysis for this study is the local farm-level where micro-analysis of adaptation will be analyzed to find potential ways of improving agricultural production at the farm level.  Adapted from Jawahar and Msangi (2006) The figure shows critical adaptation points highlighted in light grey color and key constraints (dark grey color) affecting successful crop and or livestock production in the face of changing climatic conditions. Adaptation measures can be supply-side measures (such as providing more water), demandside measures (such as reuse of water) and combinations of both (such as changing crop varieties). While some measures may be taken at the individual or farm level, others require collective action (rainwater harvesting) or investments at the agency or government level (for example, building dams, releasing new cultivars that are more water efficient) (Jawahar and Msangi 2006).Resource limitations and poor infrastructure limit the ability of most rural farmers to take up adaptation measures in response to changes in climatic conditions. With resource limitations, farmers fail to meet transaction costs necessary to acquire adaptation measures and at times farmers cannot make beneficial use of the available information they might have (Kandlinkar and Risbey 2000). Labor availability is considered an important input constraint. The expectation is that farm households with more labor are better able to take on various adaptation management practices in response to changes in climatic conditions compared to those with limited labor. Lack of market access can also limit the potential for farm-level adaptation. Farmers with access to both input and output markets have more chances to implement adaptation measures. Input markets allow farmers to acquire the necessary inputs they might need for their farming operations such as different seed varieties, fertilizers, and irrigation technologies. On the other end, access to output markets provide farmers with positive incentives to produce cash crops that can help improve their resource base and hence their ability to respond to changes in climatic conditions (Mano et al. 2003).Information concerning climate change forecasting, adaptation options, and other agricultural production activities remains an important factor affecting use of various adaptation measures for most farmers. Lack of and or limitations in information (seasonal and long-term climate changes and agricultural production) increases high downside risks from failure associated with uptake of new technologies and adaptation measures (Jones 2003;Kandlinkar and Risbey 2000). Availability of better climate and agricultural information helps farmers make comparative decisions among alternative crop management practices and this allows them to better choose strategies that make them cope well with changes in climatic conditions (Baethgen et al. 2003, see also Figure 2). Failure to implement adaptation options and poor agricultural performances by many African farmers has been blamed on lack of information and resources (Archer et al. 2005). Southern Africa for example, has early warning units and meteorological departments, but the information does not reach all intended users (Archer et al. 2005). Adaptation policy measures need to consider how information concerning adaptive measures, forecasts, and production cycles can best reach farmers to help them respond to changes in climate. Climate change policy measures regarding information need to put in place information pathways that ensure that important climate change information is timely disseminated to the farmers.Improving the adaptive capacity of disadvantaged communities requires ensuring access to resources, income generation activities, greater equity between genders and social groups, and an increase in the capacity of the poor to participate in local politics and actions (IISD 2006). Thus, furthering adaptive capacity is in line with general sustainable development and policies that help reduce pressure on resources reduce environmental risks, and increase the welfare of the poorest members of the society.The empirical estimation of the determinants of adaptation strategies takes into account the various issues and factors raised in the discussion above. Some of these factors are considered as explanatory variables in the model to help assess their impact on the propensity of adoption of various adaptation strategies. Examples of factors considered include farmer education level, access to markets and information (extension services) and other household characteristics that are discussed in the empirical estimation section below. Early impact assessment studies ignored adaptation (Tol et al. 1998cited in IPCC 2001) giving rise to the so-called \"dumb farmer\" scenarios. The \"dumb farmer\" scenarios represent any agent that is assumed to continue to act as if nothing has happened as he or she does not anticipate or respond to changes in climate (Rosenberg, 1992;Easterling et al. 1993;Smit et al. 1996cited in IPCC 2001). The implication of ignoring autonomous and planned adaptations in impact assessment models is that they fail to make a distinction between potential and residual net impacts. As a result, their usefulness in assessing Rosenzweig and Parry (1994) showed that there is great potential to increase food production under climate change in many regions of the world if adaptation is taken into consideration. In another study, Downing (1991) showed that adaptation has the potential to reduce food deficits in Africa from 50 adaptations to changes in climate. Kurukulasuriya and Mendelsohn (2006a) and Mendelsohn and Dinar (2003), explored the importance of water availability in the Ricardian model by estimating the role of irrigation as an adaptation measure against unfavorable climatic conditions. This was a significant step in addressing the shortcomings of past Ricardian studies of agriculture (Mendelsohn et al. 1994;1996) that were criticized for failing to take into account the effects of irrigation and other water supplies (Cline 1996;Darwin 1999). The studies showed that irrigation is an important adaptation measure that can significantly help reduce the negative impacts associated with changes in climate. Kurukulasuriya and Mendelsohn (2006b) and Seo and Mendelsohn (2006) both used multinomial logit models to analyze crop and livestock choice as adaptation options, respectively. The study on crop choice showed that crop choice is climate sensitive and farmers adapt to changes in climate by switching crops. The results from choice models from the livestock study showed that farmers in warmer temperatures tend to choose goats and sheep as opposed to beef cattle and chicken. Goats and sheep can do better in dry and harsher conditions than beef cattle. Maddison (2006) reports that perception results on climate change showed that a significant number of farmers believe that temperature has already increased and that precipitation has declined for eleven African countries. Farmers with the greatest farming experience were more likely to notice changes in climatic conditions which according to the study are consistent with farmers engaging in Bayesian-updating of their prior beliefs. The study also reported that farmer experience, access to free extension services and markets are important determinants of adaptation.Better understanding of the demand for adaptation measures requires farm household characteristics to be matched with use of adaptation measures. By identifying the important determinants of adoption of the various adaptation measures important policy information on supporting policies for farm-level adaptation strategies can be obtained.The study identified seven common adaptation measures: using different varieties, planting different crops, crop diversification, different planting dates (given the high number of statements that the timing of rains is changing), diversifying from farm to non-farm activities, increased use of irrigation, and increased use of water and soil conservation techniques. The statistical model for assessing determinants of adaptation options assumes that use of each adaptation option is related to a number of socioeconomic factors, and farmer perceptions about changes in climatic variables.Descriptive statistics (means) were used to characterize farmer perceptions on changes in long-term temperature and precipitation changes as well as various adaptation measures being used by farmers and barriers to adaptation. The multivariate probit technique is used to analyze the determinants of adaptation measures. The multivariate probit model simultaneously models the influence of the set of explanatory variables on each of the different adaptation measure while allowing the unobserved and unmeasured factors (error terms) to be freely correlated (Lin et al. 2005;Green 2003;Golob and Regan 2002).Complementarities (positive correlation) and substitutabilities (negative correlation) between different options may be the source of the correlations between error terms (Belderbos et al. 2004). Another source of positive correlation is the existence of unobservable household-specific factors that affect choice of several adaptation options but are not easily measurable such as indigenous knowledge. The correlations are taken into account in the multivariate probit model.Another approach would be to use a univariate technique such as probit analysis for discrete choice dependent variables to model each of the adaptation measures individually as functions of the common set of explanatory variables. The shortfall of this approach is that it is prone to biases caused by ignoring common factors that might be unobserved and unmeasured and affect the different adaptation measures. In addition, independent estimation of individual discrete choice models fails to take into account the relationships between adoptions of different adaptation measures. Farmers might consider some combinations of adaptation measures as complementary and others as competing. By neglecting these common factors the univariate technique ignores potential correlations among the unobserved disturbances in adaptation measures, and this may lead to statistical bias and inefficiency in the estimates (Lin et al. 2005;Belderbos et al. 2004;Golob and Regan 2002).A multinomial discrete choice model is another alternative to the multivariate model with seven endogenous, discrete choice variables. In the multinomial discrete choice model the choice set is made up of all combinations of adaptation measures or 2 7 = 128 available alternatives. With a problem of this size (128 alternatives and 19 explanatory variables) estimating a multinomial logit (MNL) model is possible.The shortfall of this technique is that interpretation of the influence of the explanatory variables on choices of each of the seven original separate adaptation measures is very difficult. The usefulness of a MNL is limited by the property of independence of irrelevant alternatives (IIA). In such situations estimation of multinomial probit (MNP) and \"mixed\" or random-coefficients MNL are more appropriate and both Bayesian and non-Bayesian simulation methods can be used to estimate parameters of large MNP and mixed logit models (Golob and Regan 2002). The shortfall of this technique is that all multinomial replications of a multivariate choice system have problems in interpreting the influence of explanatory variables on the original separate adaptation measures.This study uses a multivariate probit econometric technique to overcome the shortfalls of using the univariate and multinomial discrete choice techniques. Following Lin et al. (2005), the multivariate probit econometric approach used for this study is characterized by a set of n binary dependent variables y i (with observation subscripts suppressed), such that:where x is a vector of explanatory variables, The maximum likelihood estimation maximizes the sample likelihood function, which is a product of probabilities (2) across sample observations. Computation of the maximum likelihood function using multivariate normal distribution requires multidimensional integration, and a number of simulation methods have been put forward to approximate such a function with the Geweke-Hajivassiliou-Keane (GHK) simulator (Geweke et al. 1997;Hajvassilion et al., 1996) being widely used, (Belderbos et al. 2004). This study follows the GHK simulator approach that uses STATA routines based on Cappellari and Jenkins (2003) to estimate the model1 .The marginal effects of explanatory variables on the propensity to adopt each of the different adaptation measure are calculated as:where i P is the probability (or likelihood) of event i (that is increased use of each adaptation measure), ) (⋅ φ is the standard univariate normal cumulative density distribution function, x and β are vectors of regressors and model parameters respectively (Hassan 1996).Econometric analysis with cross-sectional data is usually associated with problems of heteroskedasticity and multicollinearity and the effect of outliers in the variables. Multicollinearity among explanatory variables can lead to imprecise parameter estimates. To explore potential multicollinearity among the explanatory variables, we calculated the Variance Inflation Factor (VIF) for each of the explanatory variables. The VIFs ranges from 1.07 to 1.53 which does not reach convectional thresholds of 10 or higher used in regression diagnosis (Lin et al. 2005). In the analysis multicollinearity does not appear to be a problem. To address the possibilities of heteroskedasticity in the model, we estimated a robust model that computes a robust variance estimator based on a variable list of equation-level scores and a covariance matrix.Responses to questions on farmer perceptions were coded as binary variables. Responses to the question on whether farmers had witnessed changes in temperature were classified as falling into one or more of six different categories: 'warmer,' 'cooler,' 'more extreme,' 'other,' 'no change,' and 'don't know.' The question on whether the farmer had witnessed changes in precipitation was classified as falling into one of seven different categories. No less than 25 different categories were identified for adaptations to climate change and 12 different barriers to climate change were identified for the eleven African countries in the study (Maddison 2006).Temperature and precipitation data came from the Africa Rainfall and Temperature Evaluation System (ARTES) (World Bank 2003). This dataset, created by the National Oceanic and Atmospheric Association's Climate Prediction Center, is based on ground station measurements of precipitation.Seven dummy variables are the dependent variables for the model: using different varieties; planting different crops; crop diversification; different planting; diversifying from farm to non-farm activities; increased use of irrigation; and increased use of water and soil conservation techniques). Summary statistics of the identified main adaptation measures are presented in Table 1.  Farmer perceptions regarding long-term changes in temperature and precipitation are presented in Figures 3 and4, respectively. 2 Perceptions on long-term temperature and precipitation changes were divided into six and seven categories respectively as can be seen in the Figures. The results indicate that most farmers perceive that long-term temperatures are increasing. On the other hand, the overall perception on longterm changes in precipitation is that the region is getting drier and that there are pronounced changes in the timing of rains and frequency of droughts.2 Farmers were asked whether they have noticed changes in long term temperature and precipitation and to explain the change. They were also asked follow up questions on the adjustments they made in response to the changes in temperature and precipitation. Percentage of respondets (%)Table 3 presents various adaptation strategies being used by farmers in response to changing climatic and other socioeconomic based on the survey observations. The adaptation strategies are grouped into adaptations by country, and farmer perceptions regarding temperature and precipitation. As indicated in the results, less than 40 percent of the respondents are not adopting any adaptation strategies. As has been described above, these adaptation options can be classified into two main modifications in the production systems including increased diversification and escaping sensitive growth stages through crop management practices that ensure that critical crop growth stages do not coincide with very harsh climatic conditions in the season such as mid-season droughts. Increased diversification through engaging in production activities that are more drought-tolerant and or resistant to temperature stresses as well as activities that make efficient use and take full advantage of the prevailing water serve as an important form of insurance against rainfall variability. Growing a number of different crops in the same plot or in different plots reduces the risk of complete crop failure as different crops are affected differently by climate events. Farmers are using crop management practices that include use of irrigation, water and soil conservation techniques and varying planting and harvesting dates to ensure that critical, sensitive growth stages do not coincide with very harsh climatic conditions in the season. These strategies can also be used to modify length of the growing season; for instance irrigation and water conservation techniques are an important source of additional water that can be used to lengthen the growing period of crops. It is important to note that these adaptation measures should not be taken as independent strategies but should be used in a complementary way. For instance the use of irrigation technologies needs to be accompanied by other good crop management practices such as use of crops with better use of water; use of efficient irrigation systems, growing crops that require less water and using improved irrigation water use practices.Although farmers reported to be using these adaptation measures in response to changes in climate, we note that these actions might be profit-driven rather than responses to changes in climate.However, for the purpose of this study we assume that farmers are using these measures as a response to climate change. This assumption is based on questions about farmer perceptions on climate change and the actions they are taking to reduce the impacts of climate change on agricultural production. We however, acknowledge that to properly answer the question of whether farmers are minimizing losses due to climate change or are maximizing profits subject to markets and other socioeconomic constraints, a structural model would be required. This is not the scope of this paper and is an area that can further be explored. The study also assessed farmer perceived barriers to using various adaptation measures. The results presented in this section only provide a conjecture of barriers and do not go on to really measure such barriers. Measuring the barriers requires an ex ante simulation of a structural model or perhaps some experimentation framework in which behavioral responses can be elicited. This is however, beyond the focus of this paper and is thus given as an area that can be explored further.Results on barriers to taking up adaptation options in southern Africa are presented in Figure 5 below. The results indicate that lack of credit and information concerning climate change forecasting (both short term variations (stv) and long-term climate change (ltcc) and information concerning adaptation options and other agricultural production activities; rationing of inputs, and lack of seed inputs are important constraints for most farmers. Lack of credit, rationing of inputs, and lack of seed limit the ability of farmers to get the necessary resources and technologies they might want in order to adapt their activities to changing climatic conditions. Since most smallholder farmers are operating under resource limitations, lack of credit facilities and other inputs compound the limitations of resource availability and the implications are that farmers fail to meet transaction costs necessary to acquire the adaptation measures they might want to and at times farmers cannot make beneficial use of the available information they might have (Kandlinkar and Risbey 2000).The study estimated a multivariate probit model and for comparison a univariate probit model for each of the seven adaptation options. Results from the multivariate probit model of determinants of adaptation measures are presented in Table 4. The results of the correlation coefficients of the error terms are significant (based on the t-test statistic) for any pairs of equations indicating that they are correlated. The results on correlation coefficients of the error terms indicate that there are complementarities (positive correlation) between different adaptation options being used by farmers. The results supports the assumption of interdependence between the different adaptation options which may be due to complementarity in the different adaptation options and also from omitted household-specific and other factors that affect uptake of all the adaptation options. Another important point to note from the results is that there are substantial differences in the estimated coefficients across equations that support the appropriateness of differentiating between adaptation options. Percentage of respondents (%)The univariate probit models can be viewed as a restrictive version of the multivariate probit model with all off-diagonal error correlations set to zero (i.e. 0 = ij ρ for j i > ), (Lin et al. 2005;Belderbos et al. 2004) ; probability > 2 χ = 0.0000 justifying estimation of the multivariate probit that considers different adaptation options as opposed to separate univariate probit models and consequently the unsuitability of aggregating them into one adaptation or no adaptation variable as was the case by Maddison (2006). The following summarizes results from the multivariate probit analysis:  Noticing climate change increases the probability of uptake of adaptation measures. Farmers who are aware of changes in climatic conditions have higher chances of taking adaptive measures in response to observed changes. It is an important precondition for farmers to take up adaptation measures (Madison 2006). Raising awareness of changes in climatic conditions among farmers would have greater impact in increasing adaptation to changes in climatic conditions. It is therefore important for governments, meteorological departments, and ministries of agriculture to raise awareness of the changes in climatic conditions through appropriate communication pathways that are available to farmers such as extension services, farmer groups, input and output dealers, radio and televisions among others. This needs to be accompanied by the various crop and livestock management practices that farmers could take up in response to forecasted changes in climatic conditions such as varying planting dates, using irrigation, or growing crop varieties suitable to the predicted climatic conditions.Access to free extension services significantly increases the probability of taking up adaptation options except moving from faming to non-farming. Extension services provide an important source of information on climate change as well as agricultural production and management practices. Farmers who have significant extension contacts have better chances to be aware of changing climatic conditions and also of the various management practices that they can use to adapt to changes in climatic conditions.Improving access to extension services for farmers has the potential to significantly increase farmer awareness of changing climatic conditions as well as adaptation measures in response to climatic changes.Farmers with access to credit and markets have higher chances of adapting to changing climatic conditions. Access to affordable credit increases financial resources of farmers and their ability to meet transaction costs associated with the various adaptation options they might want to take. With more financial and other resources at their disposal farmers are able to change their management practices in response to changing climatic and other factors and are better able to make use of all the available information they might have on changing conditions both climatic and other socioeconomic factors. For instance, with financial resources and access to markets farmers are able to buy new crop varieties, new irrigation technologies, and other important inputs they may need to change their practices to suit the forecasted and prevailing climatic conditions.Increasing mean annual temperature increases the probability of farmers to respond to changes in terms of changing management practices. Increasing warming is associated with decreases in water resources (surface and ground), and high evapotranspiration rates. Resulting water shortages leads to a variety of farmer responses, including changes in crop and livestock management practices. For instance farmers change to drought-resistant crops or varieties; vary planting dates so that critical crop growth stages do not coincide with peak temperature periods; diversify crop and non-farm income options; use water and soil conservation techniques to conserve the little rain that is received; and use irrigation technologies to supplement rainwater and increase the crop growing period.Increasing mean annual precipitation increases the probability of farmers changing their management practices, in particular, growing crop varieties that suit the prevailing and forecasted precipitation. Less precipitation increases the probability of farmer to efficiently use water resources for food production and other uses. Use of water conservation techniques increases with decreasing precipitation because farmers have learnt from drought experiences to conserve rainwater in times of good rains so that it is available for future use in dry periods. Increasing knowledge and empowering communities to use water conservation techniques such as water harvesting can significantly help farmers cope with changing rainfall and temperature regimes.Private property increases uptake of adaptation measures. Farmers who own their farm have a higher propensity to invest in adaptation options compared to no ownership. The implication of this finding is that it is important for governments to ensure that even in the communal systems that characterize most of the smallholder farming systems in the region, tenure arrangements are secure to facilitate investments in long-term adaptation options by farmers. Ownership of land act as a positive incentive in facilitating farmer investments on their farms that include investments in adaptation and good crop and livestock management practices. Conservation technologies have a higher chance of uptake when farmers feel secure about land ownership.Mixed crop and livestock farmers are associated with positive and significant adaptation to changes in climatic conditions compared to specialized crop and or livestock farmers. The results imply that mixed farming systems are better able to cope with changes to climatic conditions through undertaking various changes in management practices. Country fixed effects were also included and the results for Zambia are shown in Table 4.Including either South Africa or Zimbabwe resulted in each being dropped due to multicollinearity. The country effects from Zambia have significant effects on adaptation indicating the importance of national policies concerning adaptation to climate change.This study was based on micro-level analysis of adaptation that focuses on tactical decisions farmers make in response to seasonal variations in climatic, economic, and other factors. These tactical decisions are influenced by a number of socioeconomic factors that include household characteristics, household resource endowments, access to information (seasonal and long-term climate changes and agricultural production) and availability of formal institutions (input and output markets) for smoothening consumption. Farm-level decision making occurs over a very short time period, usually influenced by seasonal climatic variations, the local agricultural cycle, and other factors. Adaptation is important for farmers to achieve their farming objectives such as food and livelihood security.Descriptive statistics (means) were used to characterize farmer perceptions on changes in longterm temperature and precipitation changes. Perception results indicate that farmers are aware that the region is getting warmer and drier with increased frequency of droughts and changes in the timing of rains. Observed trends of temperature and precipitation support farmer perceptions. The implication is that farmers need to adjust their management practices to ensure that they make efficient use of the limited rainfall and water resources for food production and other needs. Farmers identified lack of credit and information concerning climate change forecasting (both short-term variations and long-term climate change and information concerning adaptation options and other agricultural production activities); rationing of inputs and lack of seed resources as important constraints. Addressing these issues can significantly help farmers tailor their management practices to warmer and drier conditions.Important adaptation options being used by farmers include crop diversification, using different crop varieties, changing planting and harvesting dates, increased use of irrigation, increased use of water and soil conservation techniques, and diversifying from farm to non-farm activities. The adaptation options being used by farmers can be classified into two main modifications in the production systems (a) increased diversification and (b) escaping sensitive growth stages through crop management practices that ensure that critical crop growth stages do not coincide with very harsh climatic conditions in the season such as mid-season droughts. Increased diversification through engaging in production activities that are drought tolerant and or resistant to temperature stresses as well as activities that make efficient use and take full advantage of the prevailing water and temperature conditions, among other factors, serves as an important form of insurance against rainfall variability. Growing a number of different crops in the same plot or in different plots reduces the risk of complete crop failure as different crops are affected differently by climate events. It is important to note that these adaptation measures should not be taken as independent strategies but should be used in a complementary way. For instance use of irrigation technologies need to be accompanied by other crop management practices. Supporting farmers in increasing these adaptation measures through providing the necessary resources such as credit, information and training can significantly help farmers increase and sustain high productivity levels even under changing climatic conditions. This paper explored the determinants of household use of different adaptation measures using a multivariate probit model. The model allows for the simultaneous identification of the determinants of all adaptation options, thus limiting potential problems of correlation between the error terms. Correlation results between error terms of different equations were significant (positive) indicating that various adaptation options tend to be used by households in a complementary fashion, although this could also be due to unobserved household socioeconomic and other factors.Multivariate probit results confirm that access to credit, free extension services, farming experience, mixed crop and livestock farms, private property and perception of climate change are some of the important determinants of farm-level adaptation options. Use of different adaptation measures significantly increase for household with more access to these factors. Designing policies that aim to improve these factors for smallholder farming systems have great potential to improve farmer adaptation to changes in climate. For example, more access to credit facilities, information (climatic and agronomic) as well as access to markets (input and output) can significantly increase farm-level adaptation.Government policies need to support research and development that develops and diffuses the appropriate technologies to help farmers adapt to changes in climatic conditions. Government responsibilities are usually through conscious policy measures to enhance the adaptive capacity of agricultural systems.Examples of these policy measures include drought resistant crop technologies, improving climate information forecasting and dissemination, or promoting farm-level adaptation measures, such as the use of irrigation technologies. Accessibility to key agricultural production information like these water and soil conservation techniques as well as the other adaptation options identified above is essential in promoting farmer adaptation to changes in climate.To properly answer the question of whether farmers are minimizing loses due to climate change or maximizing profits subject to markets and other socioeconomic constraints, there is a need to develop a structural behavioral model. This is not the scope of this paper and is an area that should be further explored.","tokenCount":"6595","images":["-1808362931_1_1.png","-1808362931_1_2.png","-1808362931_1_3.png","-1808362931_1_4.png","-1808362931_1_5.png","-1808362931_1_6.png","-1808362931_1_7.png","-1808362931_3_1.png","-1808362931_3_2.png","-1808362931_3_3.png","-1808362931_3_4.png","-1808362931_3_5.png","-1808362931_3_6.png","-1808362931_3_7.png","-1808362931_15_1.png","-1808362931_15_2.png","-1808362931_15_3.png","-1808362931_15_4.png","-1808362931_15_5.png","-1808362931_15_6.png"],"tables":["-1808362931_1_1.json","-1808362931_2_1.json","-1808362931_3_1.json","-1808362931_4_1.json","-1808362931_5_1.json","-1808362931_6_1.json","-1808362931_7_1.json","-1808362931_8_1.json","-1808362931_9_1.json","-1808362931_10_1.json","-1808362931_11_1.json","-1808362931_12_1.json","-1808362931_13_1.json","-1808362931_14_1.json","-1808362931_15_1.json","-1808362931_16_1.json","-1808362931_17_1.json","-1808362931_18_1.json","-1808362931_19_1.json","-1808362931_20_1.json","-1808362931_21_1.json","-1808362931_22_1.json","-1808362931_23_1.json","-1808362931_24_1.json","-1808362931_25_1.json","-1808362931_26_1.json","-1808362931_27_1.json","-1808362931_28_1.json","-1808362931_29_1.json","-1808362931_30_1.json","-1808362931_31_1.json","-1808362931_32_1.json","-1808362931_33_1.json","-1808362931_34_1.json","-1808362931_35_1.json","-1808362931_36_1.json","-1808362931_37_1.json","-1808362931_38_1.json","-1808362931_39_1.json","-1808362931_40_1.json"]}

data/part_2/0396878911.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"adbcb2d36b0bbf37c0383bd2de45dc8b","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/e06cb750-7d98-4a9a-8ae3-e28aa581cd46/retrieve","description":"There is a low number of studies on the impact of policy oriented research (PORIA) and an even lower number of those that undertake a quantitative assessment of impacts. In comparison, there are numerous quantitative impact assessment studies of technology research, thus creating an imbalance of evidence for decisionmakers interested in investing in agricultural research. There are many reasons for this, but one has been the challenges in PORIA, notably in methods to assess attribution of policy outcomes to research and the measurement of impacts of policy outcomes. To respond to this, a workshop was convened from November 12-14, 2014 at IFPRI headquarters in Washington, DC. It was cosponsored by IFPRI; the CGIAR Research Program on Policies, Institutions, and Markets (PIM); and the Standing Panel on Impact Assessment (SPIA) of the CGIAR’s Independent Science and Partnership Council (ISPC).","id":"381323935"},"keywords":[],"sieverID":"63da4014-ad74-457b-928d-17d168e2d42f","pagecount":"3","content":"seeks sustainable solutions for ending hunger and poverty. IFPRI is a member of the CGIAR Consortium. (www.ifpri.org and www.cgiar.org).The Independent Impact Assessment Reports are the product of externally conducted impact assessment studies of IFPRI's research. These studies are organized and overseen by an external impact assessment coordinator commissioned by IFPRI who arranges for external experts to conduct the studies and who oversees a peer review of the draft reports by at least one internal and one external reviewer. Any opinions expressed are those of the author(s) and do not necessarily represent the opinions of IFPRI.There is a low number of studies on the impact of policy oriented research (PORIA) and an even lower number of those that undertake a quantitative assessment of impacts. In comparison, there are numerous quantitative impact assessment studies of technology research, thus creating an imbalance of evidence for decisionmakers interested in investing in agricultural research. There are many reasons for this, but one has been the challenges in PORIA, notably in methods to assess attribution of policy outcomes to research and the measurement of impacts of policy outcomes.To respond to this, a workshop was convened from November 12-14, 2014 at IFPRI headquarters in Washington, DC. It was cosponsored by IFPRI; the CGIAR Research Program on Policies, Institutions, and Markets (PIM); and the Standing Panel on Impact Assessment (SPIA) of the CGIAR's Independent Science and Partnership Council (ISPC).The workshop brought together practitioners and clients for such impact assessments in order to discuss approaches and methods for PORIA that would meet the needs of the research and funding community and overcome some of the challenges currently faced in PORIA studies.This brief summarizes the workshop discussions and the paper that was developed from it.The usual practice in assessing impacts from policy-oriented research (POR) is to start with the POR and work through to its impact. Key steps of an ex post impact assessment are as follows:1. Assess the POR outputs for relevance, timeliness, and quality. 2. Assess the influence of the research outputs on any policy outcome (or change), in the context of other providers of POR and influencers of policy. 3. Assess the economic, social and environmental impacts of the policy change, and compare these to the impacts that would have occurred with a relevant counterfactual. 4. Attribute a share of the gain in (3) to the POR. 5. Given a quantitative evaluation of the impact of the policy change in (3), an attribution share from (4), and an estimate of the costs of the POR, calculate a cost-benefit ratio.Each step presents its own challenges, and the workshop discussions of these challenges and best practice solutions are summarized below for each of the five steps.• Outputs should include research publications and their quality and use, outreach and communications events, encounters with decisionmakers (policymakers, NGOs, private sector, etc.).• Identifying the outputs is relatively straightforward if a research team has kept adequate records, but this has been rare in the past. Moreover, overlapping projects and researcher responsibilities can make it hard to assign outputs to specific POR investments.• Bibliometric analyses provides an assessment of the quality of the research, but it still remains a judgment on how important this research is among other research in the thematic area.• This generally takes the form of a narrative, and should be undertaken or at least validated by a skilled external assessor.• In principle, one should try to assess influence over a policy change that the POR team had targeted in its project design, but sometimes serendipity rears its head and should not be ignored. Moreover, one should account for spillover effects of POR to other policy decision processes. • There is need to decide whether one is assessing the influence of the evidence generated by POR, or the institution(s) that undertook and communicated the POR. This is done at least implicitly with a counterfactual in mind-i.e., what would have been the policy change/outcome in the absence of the evidence, or the institution? This is important because research is but one contribution to policymaking and it may be a relatively minor one. This not easy and a common assumption is that the research was not essential to the policy change, but only to the speed of policy change, by providing the decisionmaker with information that helped to justify the decision. • The workshop noted that it may not be important to rigorously quantify the level of attribution and this will be very difficult to do in many cases. For some purposes, it may suffice to demonstrate that the POR made a contribution to the policy decisionmaking process. In this respect, there was agreement that evaluators should adopt legal standards of causality such as 'probable cause', 'preponderance of evidence' or 'beyond reasonable doubt' and not be expected to prove the influence. To attempt to claim attribution or influence may moreover not be well taken by policymakers who will rightly claim ownership of the decision. • The complexity of tracing influence increases as one moves from POR targeted at a micro (project), country, regional, or global problems. This is not only more difficult conceptually, but would be more costly to do as well. • For many purposes, an impact assessment ends with the assessment of a POR's influence over a policy change, since researchers cannot reasonably be held accountable for the implementation of a policy change or for other factors that may affect its impact.• Quantitative methods (e.g., randomized trials, econometrics, and simulation models) can be useful for assessing the impacts of some kinds of policy changes, but one of the main challenges for PORIA is that many types of policy changes cannot be evaluated in this way, and evaluators have to rely on qualitative methods instead.• Most assessments use an existing (old) policy as the counterfactual, but if the policy was going to change anyway, then the correct counterfactual is the policy that would have been adopted without the benefit of the POR. This can be hard to estimate unless the POR is intended to modify a planned policy change (e.g., help refine the design of a conditional cash transfer program that is already going to be implemented). • Sometimes the counterfactual is the new policy but brought forward in time as a result of the POR. If POR stops a bad policy from being implemented, then the policy change is not implementing it and the counterfactual is the bad policy. In some cases, the policy implemented partly as a result of POR may prove to have negative impacts as compared to a counterfactual (e.g., if the POR did not anticipate certain dynamics in the economy).In practice, there can be a significant separation in time between steps 2 and 3. That means that the ideal timing of a study focusing on influence may be too early to detect policy impacts. A PORIA that is implemented after policy impacts runs into difficulties of assessing influence which may have taken place many years earlier.Furthermore, assessment of influence requires different skills than estimation of quantitative impacts. Existing PORIA studies rarely assemble the required breadth of team members and include information collected at different times along the impact pathway.• The method of attribution depends on the type of counterfactual used. If the counterfactual is the old policy then the impact gain from the new policy is due to many things, including the POR, and an attribution share needs to be allocated to the POR. • If the counterfactual is what the policy change would have been without the POR, then if correctly estimated all the gain can be attributed to POR. This, however, may be difficult to identify in practice. • There can be sensitivities among policymakers and partner organizations in trying to assess the influence of specific research organizations, so many prefer to assess contribution instead, i.e., that the \"researchers were part of the conversation\"). • Unfortunately, one does need something firmer and more causal-an attribution-for any cost-benefit analysis. For many purposes, it is better to undertake the analysis for the output or knowledge generated by the POR rather than any particular organization that worked on it. • Even when quantitative methods can be successfully used to evaluate impacts in step 3, still the evaluator typically has to rely on qualitative approaches in step 4.• If the POR being evaluated is a research program, then the total costs should be included, not just the costs of any successful projects (cherries) that may have been evaluated.• There have been very few cost-benefit estimates for POR, and most of the ones that have been done are for micro-oriented POR studies within single countries where it is relatively easy to track influence and impacts. These studies mostly yield good cost-benefit ratios.We expect much higher cost-benefit ratios for POR that leads to cross-country spillovers or global international public goods, but sadly there have been hardly any attempts to quantify these.When evaluating an entire research program or institution, it is rarely possible to evaluate in depth all the component projects. A pragmatic compromise is to assess the relevance, outputs and influence of a high proportion of projects, and then select a few case studies for in-depth analysis. In these cases, a particularly important decision is the choice of which projects to select for impact assessment. If this is not an a priori decision, but rather ex post (after some observations can be made on possible influence and impact), then one choice is between purposefully selecting possible successes (i.e., cherry picking) versus a more neutral approach.Reasons to cherry pick:• The successes may be enough to justify an entire research program or institution; as with oil drilling where the odd gusher justifies the cost of all the trial wells. • There is really no need to assess the impact of failed projects because there would not be anything to measure. However, it is important to include the costs of all the research projects in a portfolio, including the failures, when calculating a cost-benefit ratio or rate of return.Reasons not to cherry pick:• POR not like oil drilling because there should be less chance involved in choosing successful POR than wells. Good ex-ante impact analyses can help ensure that more POR projects become cherries. • While it is reasonable to expect some POR to fail, one would also like to know why failures occurred (was it due to poor design, poor management or changed circumstances). Either way, picking a representative sample can make good sense. While cherry picking may be adequate to justify overall investment in POR it will not likely be able to inform the decision of which types of POR to invest in, i.e., which are more likely to lead to policy outcomes and impacts.A common problem in impact studies of POR is a lack of credible evidence available for ex-post evaluations. Very few research teams compile a systematic evidence trail about the impact of their work, leaving it to evaluators to try and create such evidence in an ex-post setting. Sadly, much valuable information relevant to the conduct and influence of research is lost as memories fade or individuals with knowledge of specific research activities relocate. Regular and well-designed monitoring and evaluation may be expected to pay high dividends on a number of fronts: by sustaining long-run support from developing country governments and donors alike, by providing learning and feedback loops, and by providing solid empirical data on outputs and, with time, outcomes. Establishing such a system involves clarification of intended outcomes and theories of change, improved tracking of outputs and outreach activities and timely coordination of studies of influence or impact.The table below summarizes these key steps. ","tokenCount":"1936","images":["381323935_1_1.png","381323935_1_2.png","381323935_1_3.png"],"tables":["381323935_1_1.json","381323935_2_1.json","381323935_3_1.json"]}

data/part_2/0409997010.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"03ae6d0c39309be82e38914fefd8a7ad","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/fbcd9792-98c1-4e11-bffd-0813e6b580c3/retrieve","description":"Claudia Ringler SEMINAR Irrigation Investment Policy: Does Scale Matter? Co-organized by United States Agency for International Development (USAID) and IFPRI MAY 24, 2022 - 9:30 TO 11:30AM EDT","id":"269620819"},"keywords":[],"sieverID":"b251d8b1-55c1-4b00-a8f7-0aa32c90ffe4","pagecount":"15","content":"The end of irrigation? Some of the world's largest freshwater lakes have dramatically shrank due to irrigation (Aral Sea / Lake Chad)  Some irrigation depends largely on snowpack/glacier meltwater--, f.ex. Pakistan. With climate change, these resources will be exhausted in the near future  Little movement to conserve water in some of the most arid / water scarce areas","tokenCount":"60","images":["269620819_1_1.png","269620819_1_2.png","269620819_2_1.png","269620819_2_2.png","269620819_2_3.png","269620819_2_4.png","269620819_2_5.png","269620819_3_1.png","269620819_3_2.png","269620819_4_1.png","269620819_5_1.png","269620819_6_1.png","269620819_7_1.png","269620819_8_1.png","269620819_8_2.png","269620819_9_1.png","269620819_9_2.png","269620819_9_3.png","269620819_9_4.png","269620819_9_5.png","269620819_10_1.png","269620819_10_2.png","269620819_11_1.png","269620819_11_2.png","269620819_12_1.png","269620819_12_2.png","269620819_13_1.png","269620819_13_2.png","269620819_13_3.png","269620819_13_4.png","269620819_13_5.png","269620819_14_1.png","269620819_15_1.png","269620819_15_2.png"],"tables":["269620819_1_1.json","269620819_2_1.json","269620819_3_1.json","269620819_4_1.json","269620819_5_1.json","269620819_6_1.json","269620819_7_1.json","269620819_8_1.json","269620819_9_1.json","269620819_10_1.json","269620819_11_1.json","269620819_12_1.json","269620819_13_1.json","269620819_14_1.json","269620819_15_1.json"]}

data/part_2/0430408156.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"5db9134d190f004dc8a9f5dbb93da673","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/8f104aaa-eb3e-420b-bf80-1bbdad7bc59b/retrieve","description":"Prabhu Pingali IFPRI-FAO conference, \"Accelerating the End of Hunger and Malnutrition\" November 28–30, 2018 Bangkok, Thailand","id":"-760191676"},"keywords":[],"sieverID":"bcdd7f32-1205-4665-80b4-3eb309a23627","pagecount":"9","content":"a specific technological intervention, policy change, or institutional reform, often a combination thereof, that leads to transformative change at scale in the sustainable reduction of hunger and malnutrition.Transformative change in food systems requires a transition from a focus on quantity to an emphasis on quality, diversity & safety.Inter-sectoral synergies are essential for sustainable reductions in malnutrition in all its forms.","tokenCount":"60","images":["-760191676_1_1.png","-760191676_6_1.png","-760191676_6_2.png","-760191676_8_1.png","-760191676_8_2.png","-760191676_8_3.png","-760191676_8_4.png","-760191676_8_5.png","-760191676_8_6.png","-760191676_8_7.png"],"tables":["-760191676_1_1.json","-760191676_2_1.json","-760191676_3_1.json","-760191676_4_1.json","-760191676_5_1.json","-760191676_6_1.json","-760191676_7_1.json","-760191676_8_1.json","-760191676_9_1.json"]}

data/part_2/0433069231.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"1e0141765e552e26c47c08abdaa03da3","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/51a2e621-f7ad-4c7f-a1f1-e4ce58c2a41e/retrieve","description":"Which social safety-net progammes reach poor households? How cost efficient are they? This research shows that public works programmes have great potential for targeting poor households. However there is a great variability in their performance. Human capital subsidies provide a promising approach for addressing poverty and have been shown to have a substantial impact on nutrition, health, and education outcomes.","id":"30084425"},"keywords":[],"sieverID":"453d8be6-5438-41d4-a5c8-11ee92db6644","pagecount":"2","content":"T here growing recognition that public social safety-net systems can play a crucial role both in protecting households from poverty and in promoting long-term development. Indeed, for many of the world's poor, public safetynet programmes are their only hope for a life free from chronic poverty, malnutrition, and disease. However, their performance has been variable, reflecting a number of shortcomings that undermine their effectiveness. Frequently, a large portion of safety net budgets is eaten up by administrative costs, corruption, and operational inefficiency. Or the transfers (in the form of food, other in-kind transfers, or cash) themselves fail to reach the most vulnerable groups.Typically three broad intervention types absorb the bulk of governments' safety net budgets: public food subsidies (including both universal and targeted subsidies), public works, and humancapital subsidies in education and health sectors. Empirical evidence clearly shows that universal public food subsidies are rarely a cost-effective way of getting resources to the poor, reflecting both high leakages to the non-poor and economic inefficiencies resulting from distorted consumer and producer prices. Leakages of subsidies to the non-poor due to poor targeting increase the budget cost of transferring subsidies to the targeted poor population. Due to this leakage, it costs governments $3.3 to transfer $1 to the poor through universal food subsidies. For this reason, this modality for food distribution to the poor is often viewed as a stopgap policy until more effective policy instruments can be developed. Although the targeting of public food subsidies (e.g., through ration shops) can improve cost effectiveness, their performance has not always been satisfactory, often reflecting the high administrative costs, corruption, and leakages to the non-poor. Again, focusing only on leakages (i.e., ignoring administrative costs), on average it costs $2.6 to transfer $1 to the poor.Traditionally, one of the most popular programmes has been public works, which employ the poor on projects that maintain or create a physical asset-a road, an irrigation system. More recently, human capital subsidies in the form of transfers conditioned on poor children attending school or health clinics are increasing in popularity. A recent IFPRI/World Bank study has reviewed the available evidence on the design and performance of social safetynet programmes from 47 countries across Africa, Asia, Eastern Europe, and Latin America. The findings show that such programmes are generally successful at getting a high proportion of transfers to the poor, with the poor receiving, on average, around 25 percent more than they would without targeting. This increases to 35 percent when universal food subsidies are not included. However, both the large variation in performance (within targeting methods, programme types, and regions) and the large number of poorly targeted programmes found in practice (a staggering one quarter were benefiting the nonpoor) highlight the need to pay sufficient attention to detailed programme design and implementation issues.Research indicates that public works and human capital development hold strong promise for improving the livelihoods of the poor. But research also tells us that much more can be done to design these programmes to better transfer resources and pull households out of destitution.Public works programmes that transfer resources in the form of cash or food Because these design features are often not in place, studies have shown that participants often lose income from other sources in order to participate in public works projects. These lost wages can constitute anywhere from 25-50 percent of what they could earn from public works. Using a 25 percent wage loss, the cost of transferring income to poor households through public works is calculated to increase from $1.3 presented above to $1.7 for every dollar transferred to the poor. The choice and quality of project output also matter and community participation in the selection and implementation of projects have been shown to have high returns. However, although the emphasis on community asset maintenance or creation gives these programmes greater benefits in terms of the community as a whole, it also means that these programmes can be a very costly way to transfer income to poor households. For example, when materials, management, and equipment account for 30 percent of total programme costs, the total costs of transferring resources to the poor through public works raises to $2.4 to transfer $1.Human Capital Subsidies, which are transfers of food or cash conditioned on households investing in their children's nutrition, health, and education status, provide a promising approach for addressing poverty. Invariably, the poorest households are not only poor in terms of income and consumption levels, but also in terms of their nutrition, health, and education levels. By increasing human capital in poor households, these types of programmes can contribute significantly to breaking the inter-generational transmission of poverty.Human capital subsidies are generally very well targeted, using a combination of geographic, demographic, proxy-means and community targeting methods. Because, on average, 68 percent of the benefits accrue to poor households, it costs $1.4 per $1 transferred to the poor, i.e., slightly lower than the average leakage cost of public works. In addition, administrative costs appear to be relatively low, on average, accounting for around 20 percent of the budget. Combining this with the targeting efficiency, human capital subsidies cost in total $1.7 to transfer $1 to the poor and out perform the best public works programmes discussed above.Human capital subsidies also have been shown, through rigorous evaluations, to have a substantial impact on nutrition, health, and education outcomes. For example, in Bangladesh's Food for Education Programme, it has been estimated that by giving food assistance linked to school attendance, primary school enrolment increased from 9 percent to 17 percent. Other education programmes, particularly in Latin America, have proven even more successful. In Nicaragua, one of the lowest income countries in Latin America, human capital subsidies resulted in an increase in primary school enrolment rates by 22 percentage points (from 69 to 91 percent). The impacts of these programmes on nutrition and health are equally impressive. In Mexico's programme, there was a significant increase in child growth and a reduction in the probability of stunting for children in the critical age range of 12 to 36 months. These results are consistent with a 16 percent increase in mean child growth per year. Source: See reference below.Leakages plus other costs Public works $1.6 ($1.3-$4) $3.0 ($2.4-$7.6) Human capital subsidies $1.4 ($1.25-$2.3) $1.7 ($1.6-$2.9) Notes: Numbers refer to average cost of transfer (with the range given in brackets). For public works, other costs include lost earnings (25 percent of project wage) and management and material costs (30 percent of programme budget). For human capital subsidies, other costs include administrative costs associated with targeting, implementing, and monitoring the programme (20 percent of programme budget).","tokenCount":"1103","images":["30084425_1_1.png","30084425_1_2.png"],"tables":["30084425_1_1.json","30084425_2_1.json"]}

data/part_2/0433610191.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"16b377a4dc3c496c08181d07314af2c5","source":"gardian_index","url":"https://dataverse.harvard.edu/api/access/datafile/:persistentId/?persistentId=doi:10.7910/DVN/ZZRV3J/LGWNGX","description":"The project-level Women’s Empowerment in Agriculture Index (pro-WEAI) is a survey-based index designed for use by agriculture development projects to measure project impacts on women’s empowerment, diagnose areas of women’s disempowerment, and inform strategies to address deficiencies in specific settings. Pro-WEAI was developed jointly by the International Food Policy Research Institute (IFPRI), the Oxford Poverty and Human Development Initiative (OPHI), and 13 agricultural development projects in Africa and South Asia as part of the Gender, Agriculture, and Assets Project, Phase 2 (GAAP2). Pro-WEAI is an adaptation of the Women’s Empowerment in Agriculture Index (WEAI), originally developed in 2012 by IFPRI, the United States Agency for International Development (USAID), and OPHI. This data package contains baseline data from five GAAP2 projects and Stata .do files necessary for replicating the results of Malapit et al. (2019). The data include all items needed to calculate pro-WEAI, as well as the basic demographic information (gender and household type) used in the analysis.","id":"1610208548"},"keywords":[],"sieverID":"49d04605-bc0e-46e3-b28a-ec9e5b19e0aa","pagecount":"2","content":"In the process of preparing the replication results, small corrections were made to the code responsible for calculating the Gender Parity Index. The results produced below are not qualitatively different from those published in the original paper and are presented here.To replicate all the results five do-files and two data sets are needed, namely: To replicate the results all these files must be in the same folder. Download all seven files and save them into one folder on your computer. We will call this folder your working directory. Once you have successfully completed the replication, all replication results will be saved to this folder.1. Save all seven (five .do files and two .dta files) files in the same folder on your computer. 2. Change the working directory in the Development of pro-WEAI -Master do file.do to the folder where you have saved all the files. This is found on line 12 of the do file. Type or copy and paste your folder address here. 3. Run Development of pro-WEAI -Master do file.do 4. Open your working directory, all the results should now be saved here.","tokenCount":"185","images":[],"tables":["1610208548_1_1.json","1610208548_2_1.json"]}

data/part_2/0434426212.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"5401bbd833d2314d6b11029d9cc0df2c","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/3188c291-b8cb-4393-a2ac-0dcfb23b8be0/retrieve","description":"","id":"-2081571970"},"keywords":[],"sieverID":"c175fdbb-779c-4d0c-ae42-9f3dd99e4ead","pagecount":"2","content":"P epsiCo is a global business operating in more than 200 countries and territories and rooted in creating and delivering iconic, great tasting foods and beverages. A critical aspect of its operations is the ability to take successes in one part of the business and scale them elsewhere. This is increasingly common in agriculture supply chains as participants replicate and adapt to ensure a reliable supply of raw materials that meet cost and quality standards. An important focus of the company's scaling practices is that sustainability issues be factored in at the start of supply chain development.The ability to scale makes an especially significant impact when the company expands to new markets and creates new products that demand the development of a sustainable agricultural supply chain to provide raw material ingredients. Emerging markets present great opportunities, but they also present significant challenges. The latter includes an insufficient number of farmers growing targeted crops, gaps in yield and crop reliability, minimal access to capital for purchasing inputs or technology, inability to meet quality criteria or properly store crops, and inadequate infrastructure to transport materials and finished goods through the value chain to market. The ability to scale up, replicate, and adapt business models is crucial to success.Company agronomists, procurement specialists, and business development associates working in the field develop and execute business models that expand agricultural supply chains to meet market demands. Associates often contract directly with the growers, training them on agronomic best practices, quality criteria, and storage practices that will help increase yields, productivity, and economic returns.The PepsiCo model for scaling up agricultural supply chains, technology transfer, and agronomic education is used in similar fashion across countries and regions. In each case, the process is adapted to fit local culture, agricultural maturity, politics, and market demands. Following a description of the general scaling-up model, this brief examines two examples.The model has seven steps:1. Develop a plan for new market entry or demand for new crop procurement. The market plan includes clear direction on the commodity needed, the delivery schedule, product specifications, and the cost and quality needed for product manufacture in order to make the business model work for that market.2. Conduct sourcing survey(s). The agriculture procurement team identifies local sourcing opportunities for existing crops as well as growing parameters, such as climate zone and soil type, needed for crop expansion No pilots were necessary as the growers had been growing corn their entire lives and PepsiCo Mexico already knew how to increase yields. The infrastructure needed to make this work included training, extension, and a secure market for the farmers' crops, which presented a lower risk for financial institutions. PepsiCo provided the market security in the form of guaranteed contracts and FUNDAR facilitated training and extension. Through this partnership, PepsiCo scaled up the corn supply chain in Mexico through technology transfer and the sharing of practices already in use elsewhere in the country. At the same time, the project reduced freight cost by sourcing 40 percent of supply closer to the Guadalajara plant, and growers saw yields and incomes increasing more than 100 percent.The second example involves the demonstration and deployment of various technologies that significantly increase yields and overall productivity in India. PepsiCo began developing a potato supply chain in India in 1994 and gradually transitioned from working with aggregators to direct contracting with growers as government policy permitted. This change in policy allowed PepsiCo to work more closely with individual farmers, resulting in more efficient grower training, new technology deployment, and, thus, scaling up of the agricultural supply chain. In India, developing a market plan and the sourcing survey were carried out in parallel. Agronomists looked for climate conditions and soils suitable for potato production. After focusing on areas that fit the crop needs, PepsiCo sought to understand current production practices and opportunities to influence these to deliver the required quality and volume that would benefit both the farmers and the company. As in Mexico, the company sought out partner institutions that could help to gain access to farmers and provide necessary inputs. It found key partners in the Central Potato Research Institute and the National Bank of India.One limitation to scaling up sufficient potato production was water. About 40 percent of potato farming in India is in waterscarce or drought-prone areas. Through pilots, the company confirmed that the introduction of drip irrigation, while not a new technology globally, had the potential to save significant amounts of water while increasing yields and tuber quality. To fill this infrastructure gap in technology and improvement of grower yields, PepsiCo helped to deploy drip irrigation in Maharashtra and Haryana states and currently has trials in Gujarat and West Bengal in India, as well as in some areas of China and the UK. In this case, the scale up has been both in expanding the potato supply chain in India and in transferring technology in areas where the company saw clear opportunity. As a result of this program, farmers found price stability, consistently higher returns, and training and technology transfer leading to an increase in productivity. In West Bengal alone, farmers gained access to technology, expertise, and the enabling environment that came from the company's partnership with the Central Potato Research Institute and the International Potato Centre for processing grade seed potato, with chemical companies for agrochemicals at subsidized prices, with loans from the state bank at an 8 percent annual interest rate, with crop and weather insurance companies, and with a cold chain company leading to new cold storage for 10,000 tons of potatoes.In these examples, the barometer of success is that while PepsiCo's business in India and Mexico is expanding and has been established for the long term, the company has simultaneously mobilized the drivers-increased yields and successful technologies-for farmers to increase productivity and economic returns through successful scaling-up efforts in agriculture supply chains. These efforts entail three key lessons learned. First, it is imperative to ensure a market for the supply chains. Second, partnerships can help ensure access to a reliable supply that meets company standards and is mutually profitable to both grower and buyer. Third, overall costs are reduced when sustainability is part of the business plan from the start. Beth Sauerhaft ([email protected]) is director of global environmental sustainability with PepsiCo in New York. Ian Hope-Johnstone ([email protected]) is director of sustainable agriculture with PepsiCo in the United Kingdom.","tokenCount":"1062","images":[],"tables":["-2081571970_1_1.json","-2081571970_2_1.json"]}

data/part_2/0437080268.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"0de499eb0074add7832f90e7f992dda6","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/8ecf6490-32bf-4ca4-9f53-b4a302751dd7/retrieve","description":"This country factsheet presents key agricultural R&D indicators in a highly accessible visual display. The publication also features a more in-depth analysis of some of the key challenges that the country’s agricultural R&D system is facing, and the policy options to address these challenges.","id":"970831164"},"keywords":[],"sieverID":"cb564990-1ddd-45fe-b749-3011e367bf93","pagecount":"4","content":"Lesotho's agricultural research spending fell by one-third during 2009, and thereafter remained fairly constant, in inflation-adjusted terms. Spending contracted further in 2014 due to reduced government support to DAR-the country's main agricultural research agency-and the cessation of research at MFLR. The 2015 spike in spending was due to increased government investment in DAR's infrastructure.In 2016, Lesotho invested 0.94 percent of its AgGDP in agricultural research, which aligns with the 1 percent minimum level recommended by the African Union and the United Nations. This result, however, is more a reflection of Lesotho's small size than its commitment to agricultural R&D. The country's total number of agricultural researchers fell during 2009-2016 due to staff resignations, transfers, and training leave at DAR, and a decline in time allocated to research at NUL-FA.DAR is almost entirely funded by the government, but allocations have only been sufficient to cover the cost of salaries. The only exception to this was 2015, when the government allocated funding for upgrades to infrastructure. Research funding has been restricted to small research grants and alliances with regional and international agencies.Lesotho has insufficient agricultural research capacity. As of 2016, only 14 percent of DAR's researchers held PhD degrees; its pool of researchers is young and inexperienced; and opportunities for training, mentoring, and career growth are limited. In addition, staff morale and retention is low due to uncompetitive remuneration packages and insufficient resources for the conduct of effective research.Increased allocations of funding for infrastructure stemmed from the government's commitment to commercializing agriculture, both in the areas of technology generation and dissemination, and through numerous specialized services, such as seed and soil testing, as well as pest and disease diagnosis.Resolving constraints to researcher capacity is a priority for DAR. Improvements have been made in recent years, but more efforts are needed. One solution, used effectively in other countries, could be hiring recently retired researchers as consultants to train and mentor junior scientists over a transition period.Attaining (semi)autonomous status, another trend in many countries, could help DAR to diversify and increase its funding base, and support the development of critical mass of competent researchers. During 2015-2016, DAR and NUL-FA published an average of 0.5 and 13.5 journal articles, respectively. NUL-FA also published books and book chapters. Publications per researcher averaged 0.48 across the two-year period. ","tokenCount":"381","images":["970831164_1_4.png","970831164_4_2.png","970831164_4_3.png","970831164_4_4.png","970831164_4_5.png","970831164_4_6.png"],"tables":["970831164_1_1.json","970831164_2_1.json","970831164_3_1.json","970831164_4_1.json"]}

data/part_2/0443853547.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"4d0fd850241022bbf550ea2cd0f6a8cc","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/98052a42-0e94-4c2d-84a0-3739282a910d/retrieve","description":"This study is undertaken to quantify the benefits of contract farming (CF) on farmers’ income in a case where new market opportunities are emerging for smallholder farmers in Nepal. CF is emerging as an important form of vertical coordination in the agrifood supply chain. The prospect for CF in a country like Nepal with accessibility issues, underdeveloped markets, and a lack of amenities remains ambiguous. Contractors find it difficult to build links in these cases, particularly when final consumers have quality and safety requirements. However, a lack of other market opportunities makes the contracts more sustainable. The latter happens if there are product-specific quality advantages because of agroecology and, more important, lack of side-selling opportunities. Concerns remain about monoposonistic powers of the buyers when small farmers do not have outside options. Results of this study show that CF is significantly more profitable (81 percent greater net income) than independent production, the main pathway being higher yield and price realization. The positive impact of CF on farmers’ profits can help Nepal in harnessing the growing demand for pulses, especially in neighboring international markets, like India.","id":"124487805"},"keywords":["contract farming","lentil","income","small farmers","Nepal JEL classification: Q12","Q13","Q17","Q18"],"sieverID":"025d6040-6b65-451a-9d7d-0c5302d71dee","pagecount":"32","content":"established in 1975, provides evidence-based policy solutions to sustainably end hunger and malnutrition and reduce poverty. The Institute conducts research, communicates results, optimizes partnerships, and builds capacity to ensure sustainable food production, promote healthy food systems, improve markets and trade, transform agriculture, build resilience, and strengthen institutions and governance. Gender is considered in all of the Institute's work. IFPRI collaborates with partners around the world, including development implementers, public institutions, the private sector, and farmers' organizations, to ensure that local, national, regional, and global food policies are based on evidence. IFPRI is a member of the CGIAR Consortium.Lentil is Nepal's number one agricultural export commodity. 1 In Nepal it is the most significant pulse in terms of both area and production, and it constitutes more than 60 percent of the total pulses produced in the country. Lentil has a share of about 13 percent of total agricultural exports from Nepal. However, a vast majority of the 0.7 million farmers engaged in lentil cultivation are smallholders. Small farmerdominated lentil production in Nepal has historically been characterized by traditional technologies and postharvest practices. More importantly, small farmers have been forced to sell to local traders, who enjoy significant market power, implying a low share in value for the farmers.Because of these factors, profitability has traditionally been quite low for lentil growers in Nepal. In recent years, as global prices of pulses have been persistently high, including in India, the high value element in lentil has been elevated further by the ban on exports from India. This has opened up new avenues for Nepalese exports. With trade potentially expanding on both intensive and extensive margins, new opportunities for value addition and premium returns for higher quality of lentils has emerged for Nepali growers as processors, millers, and traders aim to expand production and improve quality and safety of the produce.As the demand emerged for a greater volume of lentils with better quality and food safety, there was a need for vertical and horizontal coordination. Farmers formed cooperatives and have tried to link directly with firms or larger traders through contract farming (CF). By doing so, bypassing the traditional buyers, that is, traders, there has been a push towards commercialization of lentil production in Nepal. With coordination arrangements such as CF, yields in lentils seem to have improved with adoption of technology. Moreover, price realization of the farmers has been lifted with the compression of the chain and greater bargaining power of the farmers. This is probably due to the changed scenario wherein farmer groups instead of individual farmers form forward linkages with firms and traders. Higher prices for growers also could be due to improved product quality involving cleaning, grading, and sorting of lentils.According to the Nepal Economic Agriculture and Trade (NEAT) report, due to interventions by NEAT, between 2011 and 2013, lentil yield increased by more than 50 percent and farmer sales by more than US$4.5 million 2 (United States Agency for International Development [USAID]/Nepal 2013). The share of farmers using improved seed increased from 4.4 percent to 92 percent, and the area under lentil cultivation increased by 25 percent over the baseline. The nature of contracting has been such that farmer groups sell in bulk at a premium to processors or traders-whoever pays a higher price. The CF arrangement means that farmers now sell beyond the village trader's guild.Elsewhere, a series of studies shows that market liberalization is transforming agricultural production patterns in developing countries and driving the emergence of several innovative models for linking farmers with markets (Simmons, Winters, and Patrick 2005). CF has emerged as an integral part of this agricultural transformation process, often facilitating direct firm-farm linkages. Though the potential benefits of CF in principle are significant for both contractors and contracted, particularly for the products where quality and safety are important, its role and possible impacts in the developing countries are often controversial. One of the most contentious issues in CF is the threat of exclusion of smallholders, particularly when the higher transaction cost, along with stringent demand for quality and safety, may preclude smallholder farmers from participating in CF (Pingali 2006).With this in the backdrop, CF in lentil in Nepal is an ideal case to study from the point of view of small farmers' outcomes and delivery of product attributes such as food safety in an exportable crop. This is so because on the one hand a vast majority of lentil farmers in Nepal are small but on the other hand opportunities for exports are continuously emerging. These opportunities, however, require fundamental changes in the demand for quality and safety of the product. Overall the marginal and small famers have little marketable surplus, low education, inefficient production technologies, and underdeveloped 1 Globally, the most traded pulse is lentils. 2 All dollars are US dollars. infrastructure (for example, transportation, cold storage, and information channels), constraining their access to remunerative markets.CF has indeed been shown to be a remedial institutional mechanism with the potential to increase productivity, reduce transaction costs, minimize risks for farmers, and enhance safety and quality of the produce for consumers (Minot 2011;Birthal, Joshi, and Gulati 2005;Ramaswami, Birthal, and Joshi 2006;Kutlu 2012;Jia and Bijman 2014;Kumar et al. 2016). How far this has been the case in lentil production in Nepal remains unanswered, the suitability of the case for such an investigation notwithstanding.This study is aimed at identifying the factors that motivate farmers' participation in CF in an overwhelmingly smallholder-dominated context, that is, lentil production in Nepal. Next, it also assesses the impact on farmers' economic welfare, that is, on yield, cost of production, and profit. As discussed above, lentil is one of the high-value cash crops in Nepal with high export potential. If the farmers can link up with export markets and get a fair share in the value, given the large number of small farmers, it can be a significant contributor to poverty reduction. This study thus provides an empirical analysis of the impact of CF in lentil on outcomes such as farm profits, efficiency, yields, and production costs of smallholders in Nepal.The precise research questions addressed in this paper are the following: What factors motivate farmers to participate in CF in lentil? Does CF raise profits, increase yield per hectare (ha), and reduce total cost of production?These questions are quite important because they relate to the prospects of small farms. Greater efficiency and profitability in potentially high-value crops supported by coordination arrangements seem to be sine qua non for the survivability of a large number of small farms, as is the case in lentil production in Nepal. In Nepal, the average farm size is less than 0.7 ha, which is much smaller than in Thailand (3.4 ha), India (1.15 ha), and South Korea (1.5 ha).On the policy side also the present study is quite pertinent. Nepal is one of the countries where policy makers are still uncertain about the promotion of CF. CF in Nepal is in its infancy, and given the pertinence of the case of lentil, an empirical analysis is useful to assess the role and impact of CF in Nepal. Since the government is still seeking evidence on CF while formulating a national policy on it, a proper and credible analysis can be really useful in crafting the appropriate agribusiness development policies in Nepal.The paper is organized as follows. In Section 2, we provide a brief background on the lentil subsector in Nepal. In Section 3, we describe the survey data and the methodological approach. The estimation results are presented and discussed in Section 4, and Section 5 concludes and provides some policy implications.Lentil is the most important pulse and a key cash crop in Nepal. During triennium ending (TE) year 2013, Nepal, with its 4.6 percent share, was the sixth largest producer of lentil in the world, after Canada (35.1 percent), India (22.2 percent), Turkey (8.9 percent), Australia (8.3 percent), and the United States (4.8 percent). In TE 2013, Nepal produced 214,000 tons of lentil in an area of 207,300 ha with an average yield of 1,033 kilograms (kg)/ha. 3 Between TE 1981 and TE 2013, the share of lentil in the gross cropped area of Nepal increased from 3.4 percent in TE 1981 to 4.2 percent in TE 2013. The contribution of lentil to the agricultural value of production also has risen marginally from 2.4 percent in TE 1981 to 2.9 percent in TE 2013. Further, lentil emerged as the most valuable export commodity of Nepal with its 11.4 percent share in agricultural exports in TE 2013 (Table 2.1).  (FAO 2015). Note: GCA = gross cropped area; TE = triennium ending; VOP = value of production. Between TE 1981 andTE 2013, there was a significant increase in the area, yield, and production of lentil in Nepal. The area under lentil cultivation more than doubled from 97,000 ha to 207,000 ha, with an average annual growth of 2.5 percent (Table 2.2). The yield of lentil increased from 497 kg/ha to 1,033 kg/ha in the same time period. A more than twofold increase in the area as well as the yield of lentil has resulted in a rise in production by more than four times, from 48.7 tons to 214.0 tons. Over time the farmers adopted new technologies and took advantage of the remunerative export market (Shrestha, Neupane, and Adhikari 2011). Though the yield of lentil in Nepal is almost at par with the global average (1,097 kg/ha), it is far behind the yields of some major producers such as New Zealand (2,580 kg/ha), China (2,294 kg/ha), Australia (2,045 kg/ha), Turkey (1,719 kg/ha), Canada (1,674 kg/ha), and France (1,656 kg/ha).Lentil is cultivated across all development regions in Nepal (Table 2.3). The share of the central region in production is the highest (32.0 percent), followed by the mid-western (28.0 percent), eastern (17.4 percent), far-western (13.1 percent), and western (9.6 percent) regions. In terms of agroclimatic zones, the cultivation of lentil is concentrated in the terai region with a share of greater than 95 percent in total production. The contribution of the hills region is around 4 percent, and that of the mountains region is meagre (0.4 percent). The yield is also highest in the terai region (1,039 kg/ha), followed by the hills (939 kg/ha) and mountains (694 kg/ha) regions. The highly favorable agroclimatic conditions, suitable soil, and easy access to main highway routes are considered the major drivers for concentration of commercial cultivation of lentil in the terai region (USAID 2011). Lentil, a winter-season crop, is cultivated after harvesting paddy. The lentil plant uses residual moisture in the soil to meet its water requirement. The major lentil-producing districts in Nepal are Dang Deukhuri, Rautahat, Kailali, Bardiya, Bara, Siraha, Banke, Nawalparasi, Parsa, Rupandehi, Saptari, Sunsari, Kapilvastu, Chitwan, Kanchanpur, and Jhapa-all terai districts (Figure 2.1).  , 2011, 2012, 2013(Nepal, MoAD 2013).Note: ha = hectare.Source: Authors' work based on statistical information on Nepalese agriculture (Nepal, MoAD 2013).Lentil accounts for 90 percent of the total export of pulses from Nepal. Overall, there has been an increasing trend in lentil exports. The export of lentil from Nepal increased from 11,383 tons in 1991 to 22,890 tons in 2013, that is, an annual growth rate of 2.1 percent. In value terms at constant prices, the export of lentil has grown at 2.6 percent per year during the same period (Table 2.4, Figure 2.2). The unit value (in constant dollars) of lentil exports has hovered between $616 per ton and $646 per ton between 1991 and 2013. This means that the Nepalese lentil has been fetching a stable price in the international market in the long run. Despite being a major producer and exporter, Nepal imports lentil to meet domestic demand, especially during off-seasons. In recent years, the import of lentil has significantly increased from a mere 4 tons in TE 1991 to 280 tons in TE 2001 and further to 6,744 tons in TE 2013. Despite a rise in Nepal's import of lentil in recent years, its trade surplus has increased from $7.3 million in 1991 to $10.5 million in 2013. Moreover, unit values in exports have been higher than in imports in most years as the lentil from Nepal is perceived as better in quality (USAID 2011).  We selected these districts due to a high concentration of lentil contract farmers and the presence of a number of firms and cooperatives that procure lentil for processing. The firms and cooperatives that establish vertical and horizontal coordination with farmers include Durali Cooperative, Hare Krishna Cooperative, Komal Cooperative, Janmukhi Cooperative, and Sidharth Mills in the Bardiya district; Krishak Upkaar Cooperative and Banke Daal Factory in the Banke district; and Shri Ram Farmer Cooperative in the Chitwan district.We surveyed 602 lentil farmers comprising 300 contract farmers and 302 noncontract farmers, chosen randomly from 27 wards under seven village development councils from the three sample districts. The share of sample size allocated to each sample district was in proportion to the number of contract farmers. Hence, the number of farmers identified for survey from Bardiya, Banke, and Chitwan districts were 300, 201, and 101, respectively. The village development councils within districts were also selected based on the presence of contract farmers specializing in lentil production. The distribution of sample households from each district is given in Table A.1.The data collected through field surveys in the three terai districts were used for empirical analysis. One of the main objectives of the study is to estimate and compare profits for contract and independent lentil growers. Profits are used as a proxy for farmer's welfare. Measurement of farm profits, however, is complex because of incomplete markets and unobserved transaction costs that make it difficult to properly price inputs and outputs (Barrett 1997). We try to build up information about profits by collecting disaggregated information about elements of cost through presurvey interactions to try to minimize measurement errors.Specifically, the components of costs of lentil production comprise labor (own and hired), rental value of land, seed (including seed treatment), fertilizer and manure (if used), pesticide (if applied), and rental for machinery (if employed). Tax on land also was considered part of the costs. In addition, there are postharvest costs related mainly to transport of the produce to market. In the broad costing group, information was collected at a detailed level to get as accurate an estimate of costs as possible. For example, labor costs were obtained for different activities such as land preparation, farm yard manure application, mulch collection, plantation/sowing, irrigation, weeding, spraying, harvesting, and cleaning. Information about different inputs used and their prices was obtained from the respondents separately to estimate the cost of cultivation of lentil. Profits of the farmers were then calculated as the difference between total revenue and total costs. Partial budget analysis was carried out to estimate the costs and returns for both contract and independent growers.As part of the implications of CF on the well-being of farmers, we attempt to answer two specific research questions through econometric methods. The first question is about identifying characteristics of farmer households that are associated with whether a farmer is part of a CF arrangement, that is, the issue of participation. Note that we put the issue forward as one of participation and not selection since several of the characteristics of the farmers that we observe now would be different at the time of selection. The second question pertains to assessing the impact of CF on the farmers' economic welfare (profits).One of the big problems in the CF literature is the identification of the causal impact of participation in CF on farm profits. It is straightforward to see that several of the observed and unobserved characteristics that result in positive or negative selection into participation in CF are also likely to have an effect on farm profits (such as skills in farming or social connectedness). Alternatively, participation in CF is usually not random but based on specific characteristics including location. To the extent that we include all three districts from the terai areas, we mitigate the problem of the location effect in contracting to some degree. Yet with the possibility of omitted variables, this implies that simple linear estimates of the effect of contracting on profits can be biased. To try to address the issue of the nonrandom nature of participation in CF, several papers have used a two-step procedure (for example, Bellemare 2012; Ito, Bao, and Sun 2012; Katchova and Miranda 2004;Miyata, Minot, and Hu 2009;Simmons, Winters, and Patrick 2005;Wang, Zhang, and Wu 2011;Gupta and Roy 2012) in assessing the impact of CF on farm returns. We also involve a two-step procedure using instrumental variables (IVs) to address the issue of endogeneity of the contracting variable.In the first stage, the dependent variable is binary (farmers' participation in contracts = 1, otherwise = 0), and the independent variables are a mix of qualitative and quantitative variables; we use a logit model to examine the role of factors associated with a farmer's being in contract or being independent. Specifically, the logistic regression is given by the following:) where p represents the probability that the farmer participates in CF and βj are regression coefficients estimated by the maximum likelihood method. Xj represents the vector of characteristics of farmer\uD835\uDC57\uD835\uDC57. These include several socioeconomic and demographic characteristics of the farm households. The details of the variables are given in Appendix Table A.2.In the second stage, to assess the impact of CF on the farmers' profits, the profit function can be represented aswhere, \uD835\uDF0B\uD835\uDF0B \uD835\uDC56\uD835\uDC56 is net profit per kg received by a farm household from cultivation of lentil, \uD835\uDC51\uD835\uDC51 \uD835\uDC56\uD835\uDC56 is a dummy variable (= 1 if farmer is in contract, 0 otherwise), \uD835\uDC4B\uD835\uDC4B \uD835\uDC56\uD835\uDC56 is a vector of observable farm and operator characteristics, and \uD835\uDF00\uD835\uDF00 \uD835\uDC56\uD835\uDC56 is an error term.As discussed above, estimation of equation 2 using a simple ordinary least squares (OLS) regression may result in biased estimates of the impact of contracting on farm profits. This is because farmers are not randomly chosen in contracts. Farmers either are selected for contract by contractors or decide to participate in contracts of their own accord. Both of these possibilities signify nonrandom selection. Hence, the unobserved factors could be guiding farmers' decisions to enter into contracts. Thus, di, the variable representing participation of farmers in contracts, is likely endogenous and could be correlated with the error term εi.Without the benefit of a randomized assignment of lentil farmers into contracts, we rely on IV techniques given that unobserved characteristics such as hidden entrepreneurial ability of a farmer can play a role in the decision to participate in a contract. Therefore, we use the IV technique to minimize bias in the estimates of the impact of CF on farm profits. An ideal IV should not correlate with the dependent variable in equation 1; however, it should be correlated with di, the variable representing participation in CF. In addition, the variable should not be from the vector of farm and operator characteristics, Xi. It is indeed hard to find an ideal instrument in this setting.We identify three IVs, namely, (1) the proportion of contract farmers in each ward for all households, (2) the proportion of contract farmers by caste group in each ward for all households, and (3) a categorical variable representing farmers producing organically certified produce, in the survey database. These three variables are network variables. We hypothesize that given a location, as a greater proportion of the farmers in the geographical and social neighborhood who are contracting increases, it would increase the likelihood of a particular farmer's getting into contract him-or herself. We take care to define the neighborhood as minutely as possible to avoid the relationship of the instrument to the dependent variable through alternative channels.Consider, for example, networks defined at a broader level: say, the district level. Here, owing to the size, the network measure could lead to effects on profits through channels such as prices of inputs and outputs faced by a farmer. Hence, our network variables based solely on geographical proximity or augmented with social proximity are defined at the ward level, that is, a subdistrict region. Defined so minutely, we can argue that profits per kg from lentil cultivation would likely be independent of the geography-or geography-plus-social-identity-based proportions of the farmers contracting. Further, we believe that social proximity based on caste is quite important in rural settings of Nepal. It is possible that households from the same village might not mingle with each other if they have different castes while farmers from different villages could interact if they belonged to the same caste. In the context of rural Nepal, homophily based on caste is likely to be important, which motivates us to create network measures by going beyond geographical proximity per se.To get the share of contract farmers in each ward, we take the number of contract farmers in a particular ward (while excluding from it the respective farmer for whom the network measure is being created) and divide it by the total number of farmers in that ward. The share of contract farmers is determined for all the households under respective wards whether or not that particular farmer is contracting. Similarly, the share of contract farmers by caste group in each ward is equal to the number of contract farmers of a caste group in a particular ward divided by the total number of farmers in that ward. We conducted Hausman's test for endogeneity and found CF to be endogenous, which indicates nonrandomness in the selection of farmers for contracting.We check for the strength of these instruments in the first stage by including them in the regression of participation in CF on its determinants. If the network variables as constructed above are strongly correlated with d i, that is, if participation in CF and our argument of their not being systematically related with per unit profit in lentil cultivation holds, the required conditions for an instrument would be met.In the first stage, the binary variable participation in CF is regressed on characteristics and the IVs share of contract farmers in wards, share of contract farmers by caste group in wards, and share of organically certified producers. The second stage estimates the contribution of participation in CF on profits instrumented from the first-stage regression.In this section we try to find whether general characteristics of contract farmers of lentil in Nepal differ from those of noncontract farmers. A simple look at the data in Table 4.1 reveals significant differences in some characteristics and small differences in others. For example, these two groups of farmers differ significantly from each other in terms of operational holding size, gross cultivated area, cropping intensity, and household size. The average household size of CF households is 6.7 as compared to 6.2 of non-CF households. The average size of operational land holding of contract farmers (1.0 ha) is significantly higher than that of noncontract farmers (0.6 ha). Similarly, contract farmers have a higher average cultivated area (2.0 ha) than independent farmers (1.3 ha). On the other hand, cropping intensity was higher among independent farmers than among contract farmers. Also, the proportion of households headed by females is significantly higher among independent farming households than among CF households. On the other hand, CF and non-CF households did not differ significantly in terms of age, education, occupation, household size, irrigation, experience in farming, incidence of migration, and monthly remittances. One of the apprehensions related to CF in developing countries stems from the threat of exclusion of smallholders. The critics of CF argue that to reduce their transaction costs, firms can prefer to tie up with a few large farmers instead of dealing with a large number of smallholders. The distribution of sample households contracting in lentil indicates a reasonable presence of smallholders in the contractual arrangements. In the sample of contract farmers, around 60 percent are marginal and small with less than 1 ha of land, while this share for noncontract farmers is 85 percent (Table 4.2). The distribution of sample households based on caste also does not reveal any bias against lower-caste farmers. Tribals, Other Backward Castes, and Dalit castes constitute 78 percent of contract farmers and 76 percent of noncontract farmers. A similar pattern is evident in the case of education. Thus, the distribution of households does not reveal existence of any systemic bias against farmers on the basis of farm size, social caste group, or education. This section assesses the impact of CF on yield, production cost, output prices, and profits of lentil cultivators. The average yield of lentil is higher for contract growers (11.4 quintals/ha) than for noncontract producers (10.1 quintals/ha), and it differed significantly at the 1 percent level (Table 4.3). In addition, the average price realized by the contract farmers (NPR 8,886/quintal) was significantly higher in relation to independent farmers (NPR 7,528/quintal). However, there is not much difference in the cost of cultivation between contract and independent farmers.Source: Authors' calculations based on 2015 field survey. Note: Price represents average weighted price received by farmer by selling produce to various marketing channels or avenues. ha = hectare; NPR = Nepalese rupees; Q = quintal. *Significant at the 10 percent level. ***Significant at the 1 percent level. On average, contract farmers realize around 80 percent higher profits than do independent lentil growers. Moreover, the higher profit realization holds for all categories of farmers except large farmers (Table A.3). Important to note, marginal farmers seem to derive the greatest benefit from CF. The per unit profit for marginal contract farmers (NPR 4,440/quintal) is more than two times that of marginal independent farmers (NPR 2,059/quintal). In fact, per unit profit from participation in CF depicts an inverse relationship with farm size. Several studies examine the effects of CF. In general, studies report substantial positive impact on gross margins, crop income, or total household income (Wainaina, Okello, and Nzuma 2012;Kalamkar 2012;Ramaswami, Birthal, and Joshi 2006;Tripathi, Singh, and Singh 2005;Birthal, Joshi, and Gulati 2005;Singh 2002;Warning and Key 2002;Leung, Sethboonsarng, and Stefan 2008;Bellemare 2012;Michelson 2013;Miyata, Minot, and Hu 2009;Xu and Wang 2009;Zhu 2007;Simmons, Winters, and Patrick 2005). Our findings are consistent with the findings of the cited studies.There can be several factors behind the observed increase in farmers' income from CF, such as better quality inputs, choice of appropriate technologies, and better realized price. CF in our case seems to make a significant difference in yield, price realization of the produce, and reduced cost of production. Consequently, contract farmers-especially smallholders-have improved profitability from lentil cultivation.Note that the composition of cost of cultivation is similar for both contract and independent farmers. Labor expenses account for more than 24 percent of the total costs of lentil cultivation for contract farmers and 27 percent for noncontract farmers (Table 4.4). Harvesting and threshing together constitute more than three-fourths of the labor costs for both categories of farmers. The input costs (predominantly seed cost) have a share of more than 10 percent in total costs. However, contract farmers have a relatively higher share of fertilizers (2.74 percent) in overall expenses than do noncontract farmers (1.89 percent). Also, there is more expenditure on using machinery and equipment among contract farmers (14.8 percent) than among their independent counterparts (12.1 percent). The details of costs by farm size for both contract and independent farmers are given in Table A.4. This section identifies the determinants of lentil farmers' participation in CF using a logit model for this purpose. The dependent variable is the binary variable participation in CF, and explanatory variables include a variety of sociodemographic and economic characteristics such as age, gender, household size, education, caste, migration, access to mobile phone connectivity, and location (district) of the village. The choice of the explanatory variables was guided by previous empirical literature on the subject (for example, Bellemare 2012; Roy and Thorat 2008;Kumar, Shinoj, and Shivjee 2013;Fischer and Qaim 2012;Miyata, Minot, and Hu 2007). Table 4.5 reports estimates of the regression model. The results reveal that household size, farm size, caste, location of village, and mobile phone connectivity are significantly associated with farmers' participation in CF. Households with small family sizes have a higher propensity to participate in CF. The relationship between land size and participation in CF is positive. Important to note, households with large farm sizes are significantly more likely to engage in CF. Other Backward Castes households are more likely to participate in CF than are Dalit households. Education did not show any significant association with participation in CF. Many papers posit a positive relationship between education and CF (Zhu and Wang 2007;Arumugam et al. 2011;Hu 2012); a number of studies show a negative or insignificant relationship (Guo, Jolly, and Zhu 2005;Ramaswami, Birthal, and Joshi 2006;Miyata, Minot, and Hu 2009;Wang, Zhang, and Wu 2011;Bellemare 2012;Ito, Bao, and Sun 2012;Wainaina, Okello, and Nzuma 2012;Wang, Yu, and Li 2013). Farmers who have access to mobile phones have a better chance of participating in CF. Age, gender, incidence of migration, education, main occupation, and remittance do not have any effect on households' participation in CF.This section presents the results of the estimation of the impact of CF on net profits of lentil farmers in Nepal. Table 4.6 reports parameter estimates for both IV and OLS regressions. The first column reports parameter estimates of the first stage, similar to the coefficients reported in Table 4.5 except for the inclusion of the IVs as regressors. All the regressions include district fixed effects, and standard errors are clustered at the district level. The Hausman test shows endogeneity when IVs are proportion of contract farmers in a ward and proportion of contract farmers by caste (Table A.5). The second column of each specification in Table 4.6 shows that the contract has a significant positive impact on the unit profit of lentil farmers.Contract farmers earn a higher profit ranging from NPR 13.27 per kg to NPR 35.72 per kg of production. Both the OLS and the two-stage least squares regression models provide evidence for significantly higher profits for contract farmers. While the OLS estimate shows an average increase of NPR 17.01 per kg of lentil in operating profits, the IV estimates show an increase ranging from NPR 13.27 per kg to NPR 35.73 per kg for different specifications. These figures come from our specifications, including district fixed effects, to account for observed and unobserved characteristics of locations.Specification 1 makes use of proportion of contract farmers by caste in a ward as the IV. Specification 2 involves a categorical variable representing farmers producing organically certified produce as the IV. Specification 3 uses proportion of contract farmers in a ward as the IV. Specification 4 has two IVs: proportion of contract farmers in a ward and proportion of contract farmers by caste in a ward. Specification 5 uses two IVs: proportion of contract farmers by caste in a ward and a categorical variable representing farmers producing organically certified produce (please refer to equations 1 and 2 in the Methodology subsection).  Using data collected in 2015, this study quantifies the impacts of participation in CF in lentil on farmers' returns, yields, cost of production, and adoption of food safety measures at the farm level. Our results show that farm size and access to mobile phones are significant determinants of participation in CF.Family size and caste attributes also are related to participation in CF. Though an overwhelming majority of farmers in lentil cultivation are small farmers, there still is stratification, with land size positively associated; comparatively large farmers have a greater chance of participating in CF. Conditional on participation, contract farmers earn significantly higher profits, realize higher yields, and register lower costs of production. These findings potentially have significant policy implications. One of the strongest criticisms of CF in developing countries stems from the perception that small farmers will be exploited by the \"big\" integrators (Gupta and Roy 2012). In fact, there has been an intense debate in the formal literature. Some researchers and policy makers perceive that CF is close to bonded labor, while the other group perceives that CF is the way out for promoting agricultural commercialization (Bellemare 2012). The (ADS 2014) has adequately emphasized promotion of agricultural commercialization and CF.In this context, these findings suggest that CF can increase households' income substantially with minimum stratification among small farmers in terms of participation. Further, Nepal has the opportunity to exploit the huge Indian pulse market, which imports 2 to 3 million metric tons of pulses annually to meet its domestic demand. CF through a collective mechanism could be one of the most promising vehicles to increase lentil production and enhance quality to harness the potential of the neighboring international market. The Agribusiness Promotion Act, which is in the offing, can further give a boost to promote CF to harness this potential. The legal system for export licensing is in place, but adequate attention needs to be given to ensure the quality and safety of the produce. Policy makers in Nepal should devise appropriate strategies and mechanisms to promote CF in such commodities, which can contribute to enhancing farmers' welfare and mitigating poverty.   ","tokenCount":"5491","images":["124487805_1_1.png","124487805_10_1.png","124487805_12_1.png"],"tables":["124487805_1_1.json","124487805_2_1.json","124487805_3_1.json","124487805_4_1.json","124487805_5_1.json","124487805_6_1.json","124487805_7_1.json","124487805_8_1.json","124487805_9_1.json","124487805_10_1.json","124487805_11_1.json","124487805_12_1.json","124487805_13_1.json","124487805_14_1.json","124487805_15_1.json","124487805_16_1.json","124487805_17_1.json","124487805_18_1.json","124487805_19_1.json","124487805_20_1.json","124487805_21_1.json","124487805_22_1.json","124487805_23_1.json","124487805_24_1.json","124487805_25_1.json","124487805_26_1.json","124487805_27_1.json","124487805_28_1.json","124487805_29_1.json","124487805_30_1.json","124487805_31_1.json","124487805_32_1.json"]}

data/part_2/0451660960.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"51ec6a34169afcc1425ded2ad9d94c0f","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/916bb0ae-7c74-4cf7-b8fd-f75779804c22/retrieve","description":"The fifth round of the Myanmar Household Welfare Survey (MHWS), a nationally and regionally representative phone survey, was implemented between March and June 2023. It follows from four rounds that were carried out quarterly beginning in December 2021. This report discusses the findings from the fifth round related to livelihoods, shocks, asset and income poverty, and coping strategies.","id":"1339102442"},"keywords":[],"sieverID":"414a4613-0a78-4dc7-b7df-be0a257f74ce","pagecount":"58","content":"The fifth round of the Myanmar Household Welfare Survey (MHWS), a nationally and regionally representative phone survey, was implemented between March and June 2023. It follows from four rounds that were carried out quarterly beginning in December 2021. This report discusses the findings from the fifth round related to livelihoods, shocks, asset and income poverty, and coping strategies.The security situation in Myanmar continued to deteriorate during the fifth-round recall period, which spanned from January to June 2023. Households felt insecure in their communities, as reported by 21 percent of households in both rural and urban areas, an increase compared to the previous year. This is because crime and violence continued to increase, affecting 18 and 10 percent of communities, respectively. Further, 7 percent of households were directly affected by violence, either through violence against a household member, robbery, or appropriation and/or destruction of their assets.Households faced multiple shocks besides insecurity. Disruptions to the internet and electricity also negatively affected household wellbeing and livelihoods. Further, households struggled to receive medical services. Finally, while school attendance recovered compared to the previous year, it declined compared to the last quarter of 2022 and was still under 70 percent in some states/regions. In R5 income-based poverty increased by 9 percent compared to R2 (+5 percentage points) and declined by 6 percent compared to R4 (-4 percentage points). Sixty-one percent of the population was income poor. The fall in income poverty between R4 and R5 was largely attributable to rising income outpacing a relatively low rate of food inflation at the beginning of 2023 -8 percent increase in food prices between R4 and R5. Casual wage-earning households, both farm and non-farm, had the highest levels of income poverty. Remittances, assistance from family and/or friends, and salaried income were the primary factors inversely associated with households' probability of being income-poor, whereas remittances, non-farm business income and larger household sizes were inversely associated with asset poverty.Seventy-one percent of households used at least one coping strategy to meet daily needs during the month prior to the fifth-round survey. The three most common coping strategies used were spending savings, reducing non-food expenditure, and reducing food expenditure. This has been consistent across rounds. Further, some households exhausted some or all their coping strategies.Compared to the other states/regions, households in Kayah and Chin were the most vulnerable, though our survey struggled to capture some of the most conflict-affected areas, especially in Sagaing. Households in Kayah and Chin were more likely to be impacted by conflict, have income loss, and be income poor. Despite reporting comparatively less conflict, households in Rakhine were also vulnerable; 72 percent of households in Rakhine were income poor and many were mortgaging/selling assets to cope. Further, because most households in Rakhine were surveyed in early May, the welfare indicators for Rakhine do not capture the disastrous effects of cyclone Mocha.From January through June of 2023, households continued to be impacted by security, climatic, and economic shocks. During the recall period for the survey, fighting was ongoing in the states/regions of Kayah, Chin, Sagaing, Kachin, Kayin, Mon, eastern Bago, and Tanintharyi (OCHA 2023a). Along with the endless devastating conflict, in May 2023, Cyclone Mocha made landfall in Rakhine State. The cyclone devastated households across Rakhine State including farmers who lost their harvest and much of their livestock. The cyclone then travelled north bringing rain, flooding, and strong winds to Chin, Sagaing, Magway and Kachin (OCHA 2023b). Households' agricultural production was impacted by this drought and flooding, as well as by other climatic shocks such as irregular temperatures and rainfall that occurred across the country. Further, despite inflation cooling in early 2023, households continued to be affected by economic shocks including high food and fuel prices and job loss. Disruptions to the internet and electricity worsened over the survey period, with most households having regular blackouts. All these factors continued to negatively impact household welfare.This paper provides an overview of the vulnerability and welfare of households across Myanmar for the fifth round (R5) of the Myanmar Household Welfare Survey (MHWS). The MHWS is a representative phone survey at the national, urban/rural, and state/region levels. The fifth round (R5) of the Myanmar Household Welfare Survey (MHWS) was carried out between March and June 2023.1 For most indicators, there was a recall period of three months, so therefore most indicators report on the period spanning January through June of 2023. This recall period includes the end of the dry season, which stretches from November to April in the largest part of Myanmar as well as monsoon planting, which began at the end of the survey collection period in May.In this paper, we provide an update of households' livelihoods and economic status. Thereafter, we examine the security, climatic, health, service, and economic shocks that Myanmar households face. Third, we analyze changes in asset and income poverty for Myanmar's households. Fourth, we study the coping strategies households employ to meet their daily needs. Finally, we explore the association of shocks and household characteristics with asset and income poverty.The paper is organized as follows: Section two describes the data and methodology. Section three provides an overview of the livelihood and economic status of households. Section four is a description of shocks that have negatively affected Myanmar's people including security, climatic, service, and economic shocks. Section five provides an update on asset and income poverty. Section six is an overview of the coping strategies that households employ. Section seven explores characteristics associated with income and asset poverty. Section eight concludes.The analysis presented in this paper relies on data from the fifth round of the MHWS. The fifth round of the MHWS was collected through phone survey interviews between March and June 2023 and has 12,953 respondents. Because most questions were asked for a three-month recall period, the data covers the time spanning from January to June. The survey intends to monitor household and individual welfare through a range of different indicators including wealth, livelihoods, food insecurity, diet quality, health shocks, and coping strategies. A novel sampling strategy in combination with the development of household and population weights allows for estimates that are nationally, regionally, and urban/rural representative (MAPSA 2022a; MAPSA 2022b).The analysis is mainly descriptive and employs straightforward indicators, although the construction of indicators related to shocks and poverty requires more detail. The shock indicators include only self-reported shocks. In the MHWS, respondents were asked about different shocks that their households or their communities experienced in the past three months. Depending on the date the household was interviewed, the past three months includes January-March 2023, February-April 2023, March-May 2023, or April-June 2023. Because of the difficulty in surveying conflict affected areas, it is likely that these MHWS estimates of shocks underrepresent the extent of insecurity in the country.The poverty line is the minimum welfare level for an individual not to be considered deprived. In previous in-person nationally representative surveys (the Myanmar Poverty and Living Conditions Survey (MPLCS) of 2014-15 and the Myanmar Living Conditions Survey (MLCS) of 2017), the share of poor was calculated using a consumption expenditure aggregate. Unfortunately, in a phone survey, collecting detailed expenditure information is not feasible. Therefore, we use an income-based poverty measure to determine the number of households that fall below the poverty line. Our income-based poverty measure is a comparison of total household income with the national poverty line. Total household income is the sum of income from 13 different economic activities plus remittances and other transfer income received in the past month. It is adjusted for household size using standard adult equivalency scales. Separately, the national food poverty line from the first quarter of 2017 -which was 1,037 kyat (MoPF et al., 2019) -was updated first with the official food CPI until mid-2020, and then with a temporal MAPSA food price index from a national survey of food vendors (MAPSA, 2022c). Then, a spatial deflator was applied to adjust food prices for rural and urban areas within each state/region based on price information from the MAPSA food vendor survey. The nonfood poverty line is calculated using the ratio of the food to the nonfood poverty lines in 2017 and the total poverty lines are the sum of the food and nonfood poverty lines. The income-based poverty measure is found to be highly correlated with the MLCS 2017 consumption-based poverty measure at the state/region level (MAPSA, 2022c).We compare our different indicators of vulnerability and welfare by the households' main source of income and asset class. We divide households into five groups by their main source of income: non-farm business, non-farm salary, non-farm wage, farm wage/salary, and own farming. Households were categorized into three asset-class groups based on the number of assets they own: asset-poor (0-3 assets), asset-low (4-6 assets) and asset-rich (7-10 assets). This categorization is based on a count of 10 assets including: improved housing (semi-pucca, bungalow/brick, apartment/condominium), flush toilet, improved water source (piped into house or bottled water), grid-based electricity (not solar), rice cooker, fridge, TV, wardrobe, car/motorcycle/tuk-tuk, and a working computer/laptop/iPad. Finally, we employ regression analysis to identify factors associated with household income poverty and asset poverty. We use random effects panel linear regressions to estimate the association between specific types of shock and the likelihood of being asset and/or income poor. We include three types of shocks in our analysis: security, climatic, and migration shocks. All three shock indicators are self-reported measures pertaining to the three months prior to the survey round. The security shock indicator is a measure of community insecurity. The climatic shock indicator measures whether the household was negatively impacted by natural or climatic shocks. We also include an indicator of the respondent's inability to work due to a lack of employment or seasonal, safety, and movement restrictions. In our analysis, we control for the main household income source, other sources of income and other household and respondent characteristics. State/region dummies and round dummies are also included in the models. It is important to note that our estimates are only associations between our independent and dependent variables.In rural areas, household farming is the most important income generating activity with 56.0 percent of households engaged in household farming and 36.4 percent of rural households identifying it has their primary livelihood in the three months prior to the R5 interview (Table 1). Compared to R2, in R5 there was a 3.2 percentage point drop in the share of rural households engaged in household farming but no significant change in the share reporting household farming as their main livelihood (Table A. 1 Table A. 2). Casual wage employment is the second most common livelihood for rural households with 15.9 percent of households primarily engaged in non-farm wage employment and 14.0 in farm wage employment. The share of rural households with farm wage employment declined by 9.9 percentage points between R4 and R5 and the share who consider it their primary livelihood declined by 3.7 percentage points. This is likely due to seasonality between the survey rounds; March through mid-June is low season for farming. However, between R2 and R5, the share of rural households with farm and non-farm wage livelihoods increased by 3 percentage points overall. Finally, in R5, 18.8 percent of rural households depended primarily on non-farm business income, which declined by 3.6 percentage points compared to R2.In urban areas, the most important source of income in the three months prior to the R5 interview was the operation of household businesses, both in terms of the share of households operating a businesses (48.6 percent) and the share reporting it as their primary livelihood (36.5) (Table 1). However, the share of urban households operating a business dropped by 9.8 percentage points since R2 and the share of households reporting it as their main livelihood fell by 4.4 percentage points (Table A. 3 Table A. 4). Non-farm salary employment was the second most common livelihood in R5 (28.4 percent of urban households) followed by non-farm wage employment (22.1 percent).In the past year, there has been an overall reduction in household engagement in income earning activities. The share of households engaged in each income generating activity either declined or increased by a small, statistically insignificant amount. Furthermore, there was a statistically significant decline in the number of household income sources, in both rural and urban areas. The number of sources was the lowest in R5 compared to all other rounds. Other sources of income, including remittances, were important for both rural and urban households, and were the main source of income for 7.8 percent of households nationally. Most of this other income was from remittances or transfers from family or friends. Very little of the other income was from local and international relief organizations or from government transfers. Table 2 presents different transfers into the household. Between the first half of 2022 and the first half of 2023, apart from remittances, most transfers into households declined. Unemployment benefits are not common in Myanmar; around 0.1 percent of households in R5 received unemployment. Pensions are more common. Around 2.7 percent of rural households and 6.9 percent of urban households received pensions in R5. Support from relief organizations was infrequent. Local relief organization provided support to around 0.8 percent of households in R5, a decrease from the previous year. International relief organizations provided support to about 1.3 percent of households during the same months, again a decline from the previous year and from the previous quarter. A more common form of support was food, non-food items and cash given by friends or family. In R5, 7.0 percent of households received money from this source, but again this was a decline from the previous year and from the previous quarter. Finally, the most important source of support was remittances, 16.4 percent of households received income from remittances in R5, which remained constant from 2022 to 2023. Median real household income increased by 8.1 percent between R4 and R5 indicating that nominal household income outpaced food inflation and household purchasing power increased -though income dynamics varied considerably between household groups. Overall, this is the first round to round increase in median real income since the survey began in late 2021 (Table 3). Real income is the value of income after adjusting for inflation. Real income in Tables 3 and4 is presented in terms of the value of the kyat in R5 and provides a measure of the purchasing power of income over time. Median real income rose in rural areas by 10.9 percent but essentially stagnated in urban areas. Table A. 5 in the appendix presents average real household income. Average income is more prone to the influence of very high values compared to median income but is useful for testing the statistical significance of changes over time. Between R4 and R5, average real income rose by 27.4 percent with a rise in both rural and urban income (35.8 and 7.8 percent, respectively). Table 3 andTable A. 5 both indicate that households whose main livelihood is farming are driving the overall increase (33.5 increase in median real income). Household groups reliant on non-farm wage income achieved no gains in median real income (-1.7 percent) and changes in both average farm and non-farm wage income are statistically insignificant (Table A . 5).In contrast, between R2 and R5, median real total income deteriorated in all household groups except own farm households, with a 10.6 percent decline in rural areas and a 10.0 percent decline in urban areas (Table 3). Farm earnings are inherently linked to food prices which contributed to real farm income outpacing food inflation. In contrast, though, median nominal income rose in the last year for all livelihood groups, but income in household groups other than own farm did not rise by enough to compensate for inflation, thus non-farming households had declining real income. In the past year, median real income earned from own farm activities, remittances, and other sources, (i.e., rent, pensions, unemployment, and support from friends/family and organizations) increased, while farm wage and non-farm salary income declined, and nonfarm wage and non-farm business income declined only slightly. Table 4 presents median real income for households who earned income from each source in the 30 days prior to the interview. Because households may have income from multiple sources or may not even earn income in their primary livelihood during the 30-day recall period, measuring income by livelihood group does not depict actual earnings in each income category. It is important to note that wage income presented in Table 4 reflects both wage rates and hours of work and thus does not depict wage rates over time as hours worked also fluctuates. In January through June of 2023, 40.2 percent of households reported lower income compared with last year, with 24.5 percent facing a significant reduction in income (greater than 20 percent) and 15.7 percent experiencing a small reduction in income (1-20 percent). In the MHWS, households were asked to reflect on their own change in income over the past year. The 40.2 percent of households who reported lower income compared with last year builds on the 55.4 percent of households who earned less income in January through June of 2022 compared to 2021 (Table 5). For 24.9 percent of households interviewed in R2 and R5 this is two subsequent years of income reduction. At the same time, 34.6 percent of households did not see a change in their total household income compared with last year. Further, 25.2 percent of households saw their income increase compared to last year. In R5, self-employed non-farm workers along with casual non-farm and farm wage earning households were the most likely to report income loss compared to the previous year. Forty-five percent of self-employed non-farm workers, 45.5 percent of casual non-farm earning households and 45.4 percent of farm wage and/or salary earning households reported lower income (the percent of households who reported a large and or small reduction) this year compared to the last (Table 5). This is significantly lower than in R2 where 59.6 percent of nonfarm wage households and 63.5 percent of farm wage households reported lower income compared to the previous year. But again, for many households this is two subsequent years of perceived income loss.Compared to households earning money from other income streams, households employed in non-farm salaried work, both farm and non-farm, were the least likely to see an income reduction compared to the previous year. Further, 29.6 percent of non-farm salaried households saw an increase in income, compared to the previous year. Farming is another income source, where many households felt that they are doing better compared to last year. While 36.3 percent of farmers reported less income compared to last year, 32.5 percent reported more income, with 11.6 percent reporting a large increase in income.  Kayah, Chin, and Rakhine suffered from the highest shares of households with selfreported reduced income; 64.6, 54.8 and 52.2 percent of households, respectively. Table A. 6 shows the share of households who felt they had lower income this year compared to last in each state/region of the country, and Figure 2 shows the share of households who are economically affected. We classify households as economically affected if they reported a large or small reduction in income compared to last year or if they had no income at all in the past three months (Figure 2). While households in Kayah, Chin, and Rakhine were the most vulnerable, more than 45 percent of households in Kayin, Sagaing, and Tanintharyi were economically vulnerable as well. Rakhine and Tanintharyi are the only states/regions where there were no improvements in the number of households with lower income this year compared to last. While households in Kayah still fared the worst in the country, they were less economically vulnerable compared to the previous round. On the other hand, in Ayeyawady, there was the largest increase in households with higher income this year compared to last, followed by Mandalay and Kachin.Note: Households are classified as economically affected if they experienced a large or small reduction in income or if they had no income at all in the past three months. Source: Author's calculations based on MHWS data.In January through June 2023, 21.0 percent of households in Myanmar felt that their community was very or somewhat insecure (Table 6). The number of households who feel insecure has increased steadily since the same period last year (R2), in which 19.6 percent of households felt unsafe in their community. Households' trust in their community has also continued to erode, with 23.2 percent of households having no or low trust in their community. Further, violence has also increased since the same time last year. In R5, 9.6 percent of households reported that there was violence in their community (Table 6). This is an increase from 7.0 percent in R2. The percentage of urban households that felt insecure in their community decreased considerably compared to the previous round, from 27.1 percent in R4 to 21.0 percent in R5. Rural insecurity remained high at 20.9 percent of rural households. At the same time, more urban households felt a low level of trust in their communities compared to rural households, and significantly more than at the same time last year (R2), and six months ago (R4) (MAPSA 2023a). Further, households reporting violence also increased in urban communities, from 9.1 percent in R2 to 11.8 percent in R5. In rural communities, there was also an increase in reported violence over the year.Lawlessness is on the rise in Myanmar. In R5, 19.7 percent of households reported a lot or some gambling in their community, 18.3 percent reported a high risk of burglary, theft, or robbery in their community, and 15.0 percent reported drug use. 2 These issues were more prominent in urban areas, compared to rural areas (Figure 3). Petty crime was particularly widespread in urban areas in the first half of 2023, with 29.1 percent of urban dwellers reporting a risk of being robbed. Another crucial challenge is that 14.4 percent of respondents felt that it was dangerous for them to move around and do everyday tasks in the first half of 2023. Again, this impacted more households in urban areas than rural areas. Further, 10.2 percent of urban households and 6.4 percent of rural households reported that it was common for them to pay bribes to authorities. Finally, 2.1 percent of respondents revealed that there was a risk of kidnapping in their community. The three state/regions where households felt the most insecure between January and June 2023 were Kayah (52.3 percent of households felt insecure), Chin (42.9 percent), and Sagaing (41.6 percent) (Table A. 7 for national Table A. 8 for urban Table A. 9 for rural). The number of households feeling insecure increased in Kayah and Sagaing from the end of 2022 to the first half of 2023, while insecurity fell in Chin and Kachin. Compared to the first half of 2022, during the first half of 2023, households felt more insecure or equally insecure in all states/regions except for Yangon, Chin, and Kachin. But of course, insecurity in Chin and Kachin fell from very high-levels and those states are still the second and fifth most insecure states, respectively.Respondents in Chin (36.2 percent) and Kachin (33.2 percent) had the lowest levels of trust in their community (Table A. 7). Though, levels of trust were similarly low in Kayah (32.7) and Kayin (32.8). Further, trust significantly declined between the first half of 2022 and the first half of 2023 in Sagaing, Kachin, Ayeyawady, Tanintharyi, Bago, Shan, and Rakhine. In the remaining states/regions, levels of social trust remained low. Community insecurity and lack of social trust may be a result of an uptick in crime or violence in the community.The three state/regions where households reported the most violence were Kayah (19.7 percent of households), Chin (18.2 percent of households), and Tanintharyi (17.3 percent of households. Violence significantly increased between the first half of 2022 and the first half of 2023. The largest increases over the period occurred in Magway, Kayin, and Shan. But reported violence also increased significantly in Bago, Mandalay, Tanintharyi, Rakhine, and Nay Pyi Taw. In Kachin, Kayah, Chin, and Sagaing, violence remained the same, and the highest over the  Along with the highest incidence of violence, 31.8 percent of households in Kayah reported not being able to move around to complete everyday tasks, 28.9 percent of households reported a risk of petty crime, 5.1 percent reported a risk of being kidnapped, and 28.3 reported high drug use in their community (Table A. 11 and Figure 5). In Chin, where violence was equally widespread, 32.9 percent households reported not being able to move around to complete everyday tasks, and a concerning 5.8 percent of households reported there was a risk of kidnapping. In Tanintharyi, while mobility was higher comparatively, risk of kidnapping, drug use, and gambling were the second highest, respectively, of all the states/regions. Further, in Tanintharyi the largest number of households feared having to pay a bribe. Households in Kayin reported the highest levels of gambling. Finally, 54.3 percent of respondents in Kachin reported some or a lot of drug use in their communities. Seven percent of respondents were directly negatively impacted by violence and/or crime against their household, including 0.9 percent of households who had a member assaulted or detained, 1.7 percent of households who suffered the destruction or appropriation of an asset, 3.0 percent of households who were impacted by theft or robbery, and 1.1 percent of households who were forced to give bribes or payments (Table 7). The incidence of households or household members being victims of theft/burglary was much higher in urban areas, 4.5 percent versus rural areas 2.4 percent. Theft/burglary of interviewed households decreased compared with the same period last year (R2) and compared with the last quarter of 2022 (R4). This is because crime rates dropped significantly in urban Yangon, Nay Pyi Taw, and Mandalay (Table A . 8). While fewer households had members robbed or their household burgled, more households had an asset destroyed or appropriated and more households had to pay bribes. In R5, households in Kayah state continued to suffer from high levels of violence and crime against them. In Kayah, 9.3 percent of households suffered damage to an asset or had an asset appropriated, and 9.4 percent of households endured theft/robbery (Figure 5 and Table A. 7). While in Kayah, numbers fell slightly compared to the end of 2022, in Chin state, destruction/appropriation of assets increased to 6.4 percent and theft and robbery to 5.1 percent.Source: Author's calculations based on MHWS data. 9 While the lowest levels of reported insecurity continued to be in Nay Pyi Taw (7.9 percent), Bago (11.4 percent), and Ayeyarwady (12.5 percent), these regions are still confronting much of the same risks as experienced across the country (Table A. 7). In Bago and Ayeyawady, a similar percentage of households were burgled or had their members robbed. Further, more than 15 percent of households reported high gambling where they lived, and more than 10 percent feared petty crime.In R5, 14.0 percent of farm households reported being negatively impacted by at least one climatic shock over the past three months. The recall period for R5, January through June, begins in the pre/post monsoon season and continues into the beginning of the monsoon. The number of households experiencing climatic shocks was identical to that one year prior. At the same time, the climatic shocks reported were slightly different. The two largest climatic shocks reported were strong winds (6.4 percent of households) and irregular temperature and rainfall (4.0 percent of households) (Figure 7). The incidence of flooding was more prevalent in the same period last year. It should be noted that in May 2023 cyclone Mocha hit Myanmar destroying households in Rakhine, Chin, Sagaing, Magway and Kachin. Most households in these states and regions were interviewed prior to this disaster, so the full extent of this shock is not captured in our dataset.3 At the regional level, intense winds were a danger to households in Chin and Rakhine, negatively impacting 19.3 and 17.6 percent of households, respectively (Table A. 11). Drought was the most prevalent in Chin, with 6.9 percent of households negatively impacted. Flooding was also an issue in Chin with 6.9 percent of households negatively impacted there. Further, in Sagaing, 5.6 percent of households were negatively impacted by flooding. Finally, irregular temperatures or rainfall were important issues in Kayin and Bago. Banking difficulties improved between January through June of 2022 and January through June of 2023, but 2.7 percent of households still paid agent fees to obtain cash while 1.5 percent of the households reported that they faced other financial issues. In the MHWS, households were asked if they had significant difficulties obtaining cash from banks or other financial institutions (Figure 8). At the beginning of 2022, 10.7 percent of households had to pay agent fees to obtain cash, 5.8 percent of household could not take out cash because the bank was either closed or had no cash, and 3.9 percent of households could only withdraw a limited amount of cash. In February through June of 2022, banking difficulties declined significantly, but 7.2 percent of households still needed an agent to obtain cash in each round. In February through June of 2023, these numbers declined to 2.7 percent of households. Further, other issues such as ATMs not working, banks closed, and use of the banking system declined to nearly zero. Between the first half of 2022 and 2023, there was a decline in the number households that accessed power from the government or national grid from 65.2 percent to 63.7 percent (Table A. 12). While most urban residents accessed power from the national grid, 90.3 percent, only about half of rural residents, 53.3 percent, did. Instead, 31.1 percent of rural residents accessed electricity from solar home systems, the usage of which increased between 2022 and 2023. The use of rechargeable battery systems also increased to 7.4 percent in rural areas while the use of community and household generators fell.For residents that accessed electricity from the national power grid, 84.3 percent of households had a power cut of at least one hour from 8:00 am to 8:00 pm for all seven days of the week prior to the interview. In Mandalay and Yangon, 93.6 and 93.0 percent of respondents reported having at least a one-hour power cut per day for seven straight days. In Nay Pyi Taw, on the other hand, 17.0 percent of respondents reported no days of daily power cuts. In Kayah and Tanintharyi, respondents recorded the fewest power cuts during the day, with 33.7 and 23.3 percent of households in the two states experiencing no power cuts during the day, respectively. At the same time, however, most of these townships, instead, faced power cuts at night. Twenty-nine percent of households reported that they were negatively affected by this loss of electricity. The loss of electricity was particularly detrimental to urban residents with 44.3 percent of urban households reporting that they were negatively impacted by this loss (Table 8). In Yangon and Mandalay, the greatest number of households reported that they were negatively affected by the loss of electricity (Table A. 13,Table A. 14,Table A. 15). Between January and June of 2023, almost half of the households (48.8 percent) did not have access to the internet regularly. During the same period in the year prior, 54.2 percent of households could not access the internet or could only access it a few times per month (Figure 10). In R5, 21.3 percent of households could not access the internet at all in the month prior to the survey, compared to 22.6 percent in R2, which shows that there has been no improvement in access to the internet over the course of mid-2022 to mid-2023. Internet access was especially difficult in Chin and Sagaing where 75.1 and 68.9 percent of households could not access the internet at all in the month prior to the survey. In Chin, only 2.7 percent of respondents could access the internet anytime they wanted to.The lack of internet access was a result of internet service disruptions, as reported by 28.4 percent of households. Households also reported not being able to afford to pay for the internet because of high fees (9.3 percent), a limited budget, or no working mobile phone (26.0 percent). But between January to June of 2023 and 2022, these issues declined by 6.0, 3.7, and 1.6 percentage points, respectively. On the other hand, 31.6 percent of households reported that they could not access the internet because they had no electricity or there were service problems. This is compared to 7.1 percent in the previous year. Internet service disruptions were the primary reason for the lack of internet in Sagaing, (82.4 percent of households reported that this is why they had no internet access), Chin (83.9), and Kachin (46.3 percent). In Yangon, Kayin, Shan, Bago, Rakhine, and Kayah, electricity access was the most cited hindrance for accessing the internet. Of the households who needed medical services, 6.4 percent of households in the month prior to the survey could not access medical services and 14.9 percent of households could only access medical services once or twice. Among households that needed medical services, access to them has increased since February-June of 2022. But in some states/regions medical access continues to be limited. In Chin, 26.9 percent of households could not access medical services in the last month. This is in addition to the 36.4 percent of Access to schooling improved tremendously from January through June of 2022 to the same period in 2023, from 52.5 percent of children 5 to 14 enrolled to 76.7 percent. At the same time, enrollment declined by 3.5 percentage points compared with R4 (July-December) of 2022. In February through June of 2022, only 52.5 percent of children 5 to 14 years were attending school, 40.8 percent in urban areas and 56.5 percent in rural areas. In the third quarter of 2022, this number jumped to 76.8 percent nationally, 74.1 percent in urban areas and 77.7 percent in rural regions. While in the fourth quarter of 2022, there were small increases in enrolment to 79.3 percent of children attending school, 80.7 percent of rural children and 75.2 percent of urban children, this declined again in the first half of 2023. Similar to enrollment levels in the third quarter of 2022, in the first half of 2023, 76.1 percent of children were enrolled in urban areas and 76.9 percent in rural regions.Compared to Q4 of 2022, there were significant declines in enrollment in Yangon, Chin, Magway, Rakhine, and Shan. In many states/regions enrollment is alarmingly low. In Sagaing enrolment was only 44.6 percent of students showing no change from Q4 of 2022. In Chin, Tanintharyi, and Kayah enrollment was 43.2 percent, 62.5 percent, and 57.6 percent, respectively, with a statistically significant decline in enrolment in Chin. In no state/region did enrolment increase between the end of 2022 and the first half of 2023 (Figure 11). In the MHWS, we asked about domestic and child labor. Nearly eight percent of households had a child under fifteen who worked a paid work week in the three months prior to the interview. Rural children were more likely to be engaged in paid work compared to urban ones, 7.9 percent of rural households had a child work for wages compared to 6.5 percent of urban households. Children also worked as domestic laborers in their household or in another family's home. Two percent of households had a helper other than their own children for domestic work and about 2.2 percent of households sent their children to work somewhere else for an income and/or for accommodation before they were 18 years old. The share of rural households which sent their children was double that of urban households, 2.6 percent versus 1.3 percent, respectively.Source: Author's calculations based on MHWS data.The rate of food inflation slowed to 8.1 percent between R4 and R5, an average monthly rate of 1.5 percent, which is the lowest average monthly increase in food prices between the MHWS survey rounds, considerably lower than the monthly average of 3.6 percent between R2 and R4 (equivalent to a 42 percent annual increase). Inflation is measured by the changing cost of a fixed basket of food items over time. In the past year, the rate of food inflation has been higher in rural compared to urban areas (45.2 and 35.6 percent respectively between R2 and R5) but the cost of the food inflation basket has narrowed considerably since the first survey round (Figure 12).Rice is the largest contributor to rising food costs in the past year, with long grain prices increasing by 87.6 percent between R2 and R5, while vegetable prices increased by 61.0 percent, pulses by 43.1 percent, animal source foods by 27.5 percent, and bananas by 26.3 percent. Though edible oil prices were a significant factor in food inflation in the first part of 2022, rising by 60 percent between R1 and R3, they increased by merely 1.6 percent in the past year. Falling onion prices drove the lowered inflation rate between R4 and R5 with most other food groups increasing but at slower average monthly rate as compared with R2 to R5. However, the reprieve from high food inflation appears to be short lived. Evidence from MAPSA's most recent rounds of food vendor surveys indicate a 19 percent rise in food prices between R5 of the MHWS to late July/early August, an average monthly increase of 3.2 percent. During this period, rice prices increased by 26 percent (see MAPSA (2023) for a description of previous food vendor surveys). In January through June of 2023, 30.8 percent of households were negatively impacted by higher food prices (Table 8). This is much lower compared to the last quarter of 2022, where 61.5 percent of households were negatively impacted by higher food prices. Although food prices rose slightly more in rural areas compared to urban (8.5 percent versus 7.3 percent), fewer rural households were negatively impacted by higher food prices, possibly because they were able to supplement from their own-production, and or as farmers, benefited from higher prices. In January through June of 2023 the number of households impacted by high fuel prices decreased considerably to 28.6 percent of households from 57.5 percent of households in the previous quarter. While fuel prices jumped considerably between Q1 and Q2 of 2022, they have leveled off at the higher price ever since. In Kayah households were still overwhelmingly negatively impacted by price shocks; 59.5 percent of households were negatively impacted by high food prices, while 60.4 percent of households were negatively impacted by higher fuel prices (Table A Twenty percent of households were negatively impacted by a loss of employment, which was an improvement from 37.3 percent in July-December of 2022. But in Kayah, 57.1 percent of households reported a loss of employment in R5, while 38.7 percent of households reported a loss of employment in Chin. This was particularly an issue in the urban areas of Kayah and Chin. At the same time, there was a statistically significant decline in the number of income sources between R2-R5 and R4-R5 in both rural and urban areas. In addition to losing income streams, households continued to face numerous challenges with earning income including reduced working hours and higher prices of farm and non-farm business inputs.Seventeen percent of salaried/wage workers reported reduced working hours or less work as their main challenge from January through June of 2023, compared to 21.8 percent a year earlier (Table A. 16). In the MHWS, households reported the main challenge they faced in the last three months, based on their principal source of income. Reduced working hours was the largest challenge faced by salaried/wage workers. This was a bigger issue in rural areas, 18.9 percent of wage/salaried workers versus 12.8 percent of wage/salaried workers in urban areas. Further, 5.8 percent of wage/salaried workers reported low/reduced wages as their principal challenge, which is a gradual improvement from 10.8 percent a year prior. In fact, 65.9 percent of wage workers stated they faced no difficulty in R5, compared to 53.5 percent of workers a year earlier. On the other hand, 1.6 percent of workers reported that their wages were not paid or were paid late, which is greater than a year prior. While nationally, 3.7 percent of wage/salary workers reported it was unsafe to travel to their work location, in Kayah, 13.5 percent of wage/salary workers reported this issue, and in Kayin 8.7 percent. Further, in Chin and Kayah, 26.4 and 24.3 percent of wage-earning households reported less work and reduced working hours as their most important challenge.The main challenges that farmers faced in R5 were high input prices or mechanization services (15.1 percent) and weather (13.2 percent). Compared to the last quarter of 2022, fewer households reported high input prices and high fuel prices as the most important issue they faced (Table A. 17). But high input prices were still a considerable issue in Kayah, faced by 34.3 percent of farmers, Rakhine, 25.3 percent of farmers, and Shan, 21.5 percent of farmers. The high price of fuel was mainly an issue in Kachin; it was the main challenge faced by 7.7 percent of farmers there. Issues with pests/diseases (6.7 percent) declined as well but were quite high among peri-urban farmers in Yangon, 17.6 percent of farmers. It is important to note that while nationally, 3.8 percent of farmers faced issues hiring workers, in Kayin, 9.2 percent of farmers faced this issue. This was a consistent issue in Kayin in 2022 as well. Finally, weather negatively impacted crop production most in Magway, 21.4 percent of farmers, and Kachin, 17.4 percent of farmers.The main issues farmers faced in terms of selling their crops were low prices for crops (9.2), though this reduced significantly from R2 (21.9 percent), and difficulty reaching traders (4.5 percent) (Table A. 18). Low prices for crops continued to be a significant issue in Kachin, Rakhine, Tanintharyi, and Shan where at least 15 percent of farmers struggled with low prices. In Kayah, 12.9 percent of farmers reported that there were not many traders with whom to sell their crops. In Chin and Sagaing, 20.4 percent and 11.4 percent of farmers stated that buyers or traders could not reach their farm because of conflict, which was a considerable increase from the previous quarter.For non-farm enterprises, 10.4 percent reported high prices of raw materials as their main challenge in R5 (Table A. 19). Increasing fuel prices declined as a prominent issue in R5, with 3.8 percent of non-farm enterprises reporting high fuel prices as the main issue they faced compared to 10.8 in the previous year. Twelve percent of non-farm business owners reported that their greatest challenge was that no customers bought their products. This was particularly an issue in Kayin and Chin. This is likely due to the low purchasing power of households across the country. A growing issue that non-farm enterprises are facing is that people are not paying off their debts, and more people are buying on credit. This increased from 0.4 percent in the first half of 2022 to 3.3 percent in the first half of 2023. In Rakhine 7.7 percent of non-farm enterprises faced this issue. Difficulties hiring workers and electricity supply problems have also become more prominent challenges in 2023. Finally, 5.9 percent of non-farm businesses stated that customers could not reach their business, which has not declined since 2022. This was an important challenge in Kayah, Chin, and Sagaing.Between December-February 2021/2022 and March-June of 2023 the percentage of the population defined as asset poor (0 -3 assets) increased statistically significantly from 33.8 percent to 37.1 percent (Figure 13). While the percentage of the asset low population (4 -6 assets) did not change, the percentage of those who are asset rich declined from 26.5 percent to 23.3 percent. Except for computer/laptops/tablets, the ownership of rice cookers, fridges, TVs, wardrobes, car/motorcycle/tuk-tuk declined (Table A. 20). Further, the use of flush toilets declined from 6.6 percent to 6.1 percent. The use of improved water sources also declined. While bottled water use increased slightly to 26.5 percent of households, water piped into the household declined from 4.0 to 2.9 percent. The use of public tap water also decreased. At the same time, the use of protected wells, springs, or ponds increased as did the use of surface water. Adjusted in accordance with food inflation, the poverty line increased by 42.3 percent between R2 and R5 and 8.1 percent between R4 and R5. The poverty line represents the cost of acquiring a basic bundle of food and nonfood needs. The cost of a bundle is estimated in a base year (2015 in Myanmar) and then in subsequent periods adjusted for food inflation to estimate its current cost. Thus, a non-poor household falls into income-based poverty when their income does not keep pace with the rising costs of the poverty line. To measure poverty in the MHWS we update the poverty line in each round using a food price index (Figure 12). We do not collect sufficient information on nonfood items to separately adjust the food-and nonfood-poverty lines.Rising real income between R4 and R5 led to the first round to round reduction in income-based poverty since the MHWS survey began in late 2021 with R5 poverty falling to 61.0 percent of the population, which was a 5.6 percent decline compared to R4 (Table 9). The 6.4 percent decline in rural poverty is statistically significant but the -3.1 percent decline in urban poverty is not. Reductions in income-based poverty between R4 and R5 are largest in households whose primary livelihood is own farming (-12.5) while reductions are smaller and/or statistically insignificant in other livelihood groups. These results are largely consistent with changes in median (Table 3) and average real income by livelihood group (Table A . 5).Despite reductions in income-based poverty in the first half of 2023, poverty increased by 8.5 percent of the population compared to the same time last year (R2), an increase which was significant in rural (7.5 percent) and urban areas (11.9 percent) (Table 9). Compared to other livelihood groups, non-farm salary households saw a considerable rise in poverty between R2 and R5 resulting in a 22.4 percent increase in poverty since last year. Once again, changes in poverty between R2 and R5 closely mirror reductions in real median income (Table 3).Income-based poverty is negatively associated with asset ownership; in R5, poverty reached 73.8, 59.5, and 42.0 of the population for households classified to be asset poor, asset low, and asset rich, respectively. Between R2 and R5 income-base poverty in assetpoor households increased by only 3.6 percent compared to 7.6 and 9.2 percent in asset-low and asset-rich households. In every state/region, income-based poverty peaked in a round prior to R5 (Figure 14). In most areas poverty was highest in R3 or R4 with the exceptions of Kachin and Kayin where poverty was highest in R1 and R2, respectively, and declined since (-14.9 and -9.9 percent between R2 and R5, respectively). Over the past year poverty increased in all other state/regions with the largest rise in Bago, 30.9 percent. Despite declining overall poverty headcounts between R4 and R5, poverty headcounts remained unchanged in Kayah and increased in three states during this period: Chin (3.3 percent), Shan (6.7 percent), and Tanintharyi (3.8 percent).Source: Author's calculations based on MHWS data.Households whose main source of income is farm wages are consistently the most vulnerable livelihood group, followed by non-farm wage earners. In R5, 88.8 and 75.2 percent of the population living in farm and non-farm wage households were income-poor, respectively. (Figure 15). In both R2 and R4, the poverty rate in farm wage households was about 45 percent higher than the national average and 23 percent higher in non-farm wage households. Any increase in poverty for wage earning households is dire, particularly as those who were already poor are likely becoming even poorer.Households reliant on other forms of income, particularly remittances, are the most resilient livelihood group with poverty rates only rising by 2.8 percentage points since the beginning of 2022 and 2.2 percentage points between R2 and R5. A large decline in median income in salaried households in the last year (-14.3 percent, Table 3) resulted in a 22.4 percent increase in their poverty rate. This large increase and the relatively stable poverty level of the \"other income\" livelihood group, has put the two livelihood groups nearly on par, with poverty rates of about 50 percent -the lowest of all livelihoods.Income poverty in farm households follows a more complex pattern than other livelihood groups. As farm income is highly seasonal, so is the income poverty status of households whose primary livelihood is own farming (Figure 15). Poverty in farm households was highest in R3, which corresponded to a period considered the lean season, prior to the harvest and sale of major crops. The R2 and R5 surveys were implemented over similar periods in 2022 and 2023, eliminating seasonal influences. Farm earnings are inherently linked to food prices which contributed to farm income outpacing food inflation (Table 3). Thus, unlike other livelihood groups in which poverty increased significantly between R2 and R5, poverty in farm households was essentially unchanged.Note: R5 poverty rates are labeled. Source: Author's calculations based on MHWS data.Overall, 71.3 percent of households used at least one coping mechanism to deal with lack of food or money in the past 30 days, 71.7 percent of rural residents and 70.2 percent of urban residents (Table 10). Shocks can be particularly damaging to household well-being, when either the household cannot deploy a coping mechanism to ensure the same living standard or, the household is forced to use a coping mechanism that results in permanent loss of assets, income, or safety. In the MHWS, households identified all the coping strategies they used in the past 30 days to cope with lack of food or money. On average, households reported using 2.1 different coping mechanisms over the 30 days prior to the R5 interview. This marks a significant decline in both the percentage of households using a coping strategy and the average number of coping strategies used in January through June of 2023, compared to the last quarter of 2022, and compared to the same period last year.  Notes: 1 Household assets include radio, furniture, television, jewelry, etc. 2 Non-agric productive assets include sewing machine, wheelbarrow, bicycle, car, etc. Asterisks show significance differences between rounds; * p < 0.10, ** p < 0.05, *** p < 0.01. Source: Author's calculations based on MHWS data.Overall, the most common coping strategies were spending savings (57.5 percent), reducing non-food expenditures (45.0 percent), and reducing food expenditures (44.8 percent). Lower food inflation between R4 and R5 may have reduced the need for households to spend more savings and cut back on food and non-food expenditure. Overall, fewer households reported using these coping strategies in February through June of 2023 compared to the previous quarter and to the same period last year. But thirty-one percent of households reduced both their food expenditure and their non-food expenditure, while 22.6 percent of households had to reduce food and non-food expenditure and spend their savings. 4 Further, 17.3 percent of panel households reduced their non-food expenditure in all five periods, 9.6 percent of households spent some of their savings in all five periods, while14.7 percent of households reduced their food expenditure in all five periods. Finally, households who reduced their food expenditure did so mainly by decreasing their spending on meat (88.1 percent), fish (79.8 percent), and oils, fats, and butter (72.7 percent) (Table A. 21). Rural households decreased their expenditures on those food groups more than urban households. Between R2 and R5, though, the largest decreases in expenditure were reported in sugary products, dairy, and grain.To meet daily needs, 15.0 percent of households mortgaged household assets and 10.6 percent sold household assets. Mortgaging assets was more common in rural areas while selling assets was more common in urban areas. Household assets include gold, jewelry, furniture, technology, and appliances. The most common asset sold and/or mortgaged was gold and/or jewelry. Among panel households, 13.2 percent of households sold assets in more than two periods, while 22.1 percent mortgaged assets in more than two periods. Further, 2.1 percent of panel households mortgaged or sold assets in all five periods. Two percent of households sold non-agricultural productive assets including sewing machines, wheelbarrows, bicycles, cars, and other means of transportation, and less than one percent mortgaged these assets. Some households also mortgaged or sold critical assets such as their dwelling (1.2 percent) or agricultural land (0.3 percent). Further, 2.4 percent of agricultural households mortgaged or sold agricultural productive assets, which is higher than the previous period and year. Given the recall period of 30 days, the number of households that have mortgaged and/or sold assets continues to be concerning.The number of households who borrowed money, 27.7 percent, decreased significantly from the previous round, and the previous year. At the same time, however, 47.7 percent of households continued to be in debt. Households also pursued risky activities to meet their daily needs. This includes 3.5 percent of households that engaged in income-generating activities that they themselves considered risky, and 7.5 percent of households where children worked to complement household income. Finally, 1.4 percent of families migrated with their entire household to deal with the dire economic situation in the month before the survey round. There was no decrease in these three coping strategies over the course of the year.Among households who used only one coping strategy, the most common coping strategy was spending savings (67.8 percent) (Table A. 22). When households used two coping strategies, most households spent their savings (73.8 percent), and additionally, households also began to reduce their food and non-food expenses, around 53 percent, respectively. Among households that used three coping strategies, nearly all of them spent savings and reduced food and non-food expenditure. Further, around 35.6 percent borrowed money and 51.0 percent reduced their expenditure on health. When households used four coping strategies, they began to increasingly mortgage and sell households assets. Finally, households that used six or more coping strategies, began to sell non-agri-productive assets (10.8 percent), sell agricultural assets (9.9 percent), engage in high-risk activities (25.8 percent), and migrate with their household (11.0 percent).The situation of households is dire in Kayah and Chin and continues to decline in Rakhine and Kayin as shown by the number of coping strategies used. At the same time, no state/region has been spared from the conflict and economic downturn, and in Mandalay and Nay Pyi Taw, where coping strategy use is lowest, still 66 percent of households in each region used at least one coping strategy. Figure 16 andTable A. 23 show coping strategies in each State/Region of the country. In Kayah, 88.1 percent of households used at least one coping mechanism in the past 30 days, and households used on average 3.0 different coping mechanisms. Further, compared to other states/regions, more households in Kayah spent their savings (75.4 percent), reduced their non-food and food expenditure (71.6 and 72.8 percent), sold household assets (20 .6 percent), and sold non-agri productive assets (5.8 percent). Compared to other states/regions, more households in Chin borrowed money (39.1 percent), reduced expenditure on health (52.7 percent), and mortgaged/sold agricultural productive assets (4.7 percent). In Rakhine State, 79.9 percent of households applied at least one coping mechanism, while using 2.9 mechanisms on average. Rakhine had the greatest number of households mortgage and sell household assets, 27.3 and 20.8 percent, respectively. Also alarming, is the percent of households who engaged in high-risk activities to meet daily needs, including 11.5 percent in Chin, 10.5 percent in Kayah, and 9.7 percent in Kachin. Further, in Kachin 11.9 percent of households had children working while in Kayah and Sagaing, 9.7 percent had children working. Further, approximately 7.2 percent of households in Kayah and 6.0 percent in Chin migrated from these states.Asset poor households were more likely to use coping strategies than asset low and asset rich households. Figure 16 shows different coping strategies used by asset class for January through June of 2023. During that period, 62.1 percent of asset poor households reduced their non-food expenditure, 65.3 percent reduced their food expenditure, and 72.3 percent spent their savings. Particularly striking is the difference between asset poor and asset rich households in terms of buying food using credit and borrowing money. Fifty-four percent of asset poor households bought food using credit compared to 15.1 percent of asset rich households. Further, 50.1 percent of asset poor households borrowed money compared to 19.5 percent of asset rich households. Finally, asset poor households were most likely to sell and mortgage assets. In this section, we explore how shocks and household characteristics are associated with vulnerability. More specifically, we explore the extent to which household characteristics and different shocks are associated with whether households are poor in terms of their income poverty or asset poverty status. Households are considered income poor if their per adult equivalent daily income is less than the poverty line and households are considered asset poor if they own fewer than four out of ten key assets.The results show that households facing security and climatic shocks experience increased income and asset poverty (Figures 17 and18). On the other hand, high levels of migration into the community and recent migration by the household are negatively associated with income poverty. Households' income and livelihood profiles matter. Households whose main source of income is from farm wages have a 22.4 percentage point higher probability of being income poor compared to own farm households and a 26.8 percentage point higher probability of being asset poor (Appendix Tables A.24 and A.25). Similarly, non-farm casual wage households are more likely to be income and asset poor than farm households by a magnitude of 12.6 and 13.6 percentage points, respectively. Households earning money from salaried labor are less likely to be income poor than farm households, whereas households with nonfarm business income are less likely to be asset poor. Households where the primary respondent is not able to find work are 14.0 percent more likely to be income poor. Being income poor increases a household's probability of being asset poor by 9.1 percentage points. Assistance helps to avert income poverty. Households who received remittances are 16.5 percentage points less likely to be income poor and households who received assistance from family and friends are 10.8 percentages points less likely to be income poor.Households in which the respondent has completed only primary education are more likely to be income poor and asset poor by 9.0 and 11.5 percentage points, respectively. Households with more dependents are more likely to be both income and asset poor. However, larger household sizes are associated with a higher probability of income poverty (15.2 percentage points) but a reduced probability of asset poverty (3.8 percentage points). Finally, rural households are more likely to be both income and asset poor, but the probability is much higher for asset poverty compared to income poverty (16.0 vs 2.5 percentage points, respectively).Note: The dependent variable is income-based poverty. Households are defined as income poor if they have income per adult equivalent per day less than the poverty line. The model also controls for state/region, survey rounds, the sex of the respondent, remoteness, and a township-level indicator for violent shocks based on secondary information from the ACLED dataset. Source: Author's calculations based on MHWS data.Note: The dependent variable is asset poverty. Household are asset poor if they have fewer than 4 assets. The model also controls for state/region, survey rounds, the sex of the respondent, remoteness, and a township-level indicator for violent shocks based on secondary information from the ACLED dataset. Regression is limited to R1 and R5, which are the rounds when information is collected on asset ownership. Source: Author's calculations based on MHWS data.Vulnerability is increasing in Myanmar. The MHWS survey data for R5, which spans the period of January to June 2023, reveals an increasing frequency of shocks encountered by households, and associated negative consequences for household welfare. The security situation continued to deteriorate, and 21 percent of households felt insecure in their communities, an increase compared to the previous year. This is because crime and violence continued to increase, affecting 18 and 10 percent of communities, respectively. Further, 7 percent of households were directly affected by violence, either through violence against a household member, robbery, or appropriation and/or destruction of their assets. In R5, climatic shocks were equally prevalent compared to the same time last year, though the most common types of shocks (strong wind and irregular temperature and rainfall) differed. Due to the timing of the survey, the full extent of the impacts of cyclone Mocha could not be captured.Disruptions to the internet and electricity also negatively affected household wellbeing and livelihoods. Further, households struggled to receive medical services. Finally, while school attendance recovered compared to the previous year, it declined compared to the last quarter of 2022 and was still under 70 percent in some states/regions. Sixty-one percent of the population was income poor in R5. Income-based poverty increased by 9 percent compared to the same time last year but declined by 6 percent compared to compared to the last quarter of 2022. This decline was largely attributable to rising income outpacing a relatively low rate of food inflation (8 percent) in the first half of 2023. Over the past year, poverty increased in all state/regions with the exception of Kayin and Kachin where poverty was highest in the first half of 2022 and continues to decline. Despite an overall decline in poverty since the end of 2022, poverty increased in three states: Chin, Shan, and Tanintharyi.Households relied on coping strategies to meet their daily needs. Seventy-one percent of households employed at least one coping strategy to meet their daily needs during the month prior to the survey round. The three most common coping strategies used were spending savings, reducing non-food expenditure, and reducing food expenditure. This has been consistent across rounds. Further, some households exhausted some or all of their coping strategies.Myanmar's households may be more vulnerable than described in this report. Because most households in Rakhine were surveyed in early May, the welfare indicators for Rakhine do not capture the disastrous effects of cyclone Mocha. Further, our survey struggled to capture some of the most conflict-affected areas, especially in Sagaing. Finally, since internally displaced persons or other households in particularly precarious situations have limited access to phones, they are under sampled.Regression analysis reveals associations between shocks and the probability of being income or asset poor, though these associations are relatively small. Our descriptive statistics and regression analysis reveal that agricultural/non-farm causal wage-earning households are among the most vulnerable. They use the greatest number of coping strategies and are more likely to be economically affected and income poor. Remittance income was the only factor we found that reduces a household's probability of being both income and asset poor.              Note: There was no option for difficulty and other listed options were multi-select responses so the total sum of percent will be greater than 100 in Round 1.Source: Author's calculations based on MHWS data.     ","tokenCount":"10535","images":["1339102442_1_1.png","1339102442_1_2.png","1339102442_14_1.png","1339102442_17_1.png","1339102442_17_2.png","1339102442_18_1.png","1339102442_18_2.png","1339102442_19_1.png","1339102442_19_2.png","1339102442_58_1.png"],"tables":["1339102442_1_1.json","1339102442_2_1.json","1339102442_3_1.json","1339102442_4_1.json","1339102442_5_1.json","1339102442_6_1.json","1339102442_7_1.json","1339102442_8_1.json","1339102442_9_1.json","1339102442_10_1.json","1339102442_11_1.json","1339102442_12_1.json","1339102442_13_1.json","1339102442_14_1.json","1339102442_15_1.json","1339102442_16_1.json","1339102442_17_1.json","1339102442_18_1.json","1339102442_19_1.json","1339102442_20_1.json","1339102442_21_1.json","1339102442_22_1.json","1339102442_23_1.json","1339102442_24_1.json","1339102442_25_1.json","1339102442_26_1.json","1339102442_27_1.json","1339102442_28_1.json","1339102442_29_1.json","1339102442_30_1.json","1339102442_31_1.json","1339102442_32_1.json","1339102442_33_1.json","1339102442_34_1.json","1339102442_35_1.json","1339102442_36_1.json","1339102442_37_1.json","1339102442_38_1.json","1339102442_39_1.json","1339102442_40_1.json","1339102442_41_1.json","1339102442_42_1.json","1339102442_43_1.json","1339102442_44_1.json","1339102442_45_1.json","1339102442_46_1.json","1339102442_47_1.json","1339102442_48_1.json","1339102442_49_1.json","1339102442_50_1.json","1339102442_51_1.json","1339102442_52_1.json","1339102442_53_1.json","1339102442_54_1.json","1339102442_55_1.json","1339102442_56_1.json","1339102442_57_1.json","1339102442_58_1.json"]}

data/part_2/0458332581.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"bbf68b368d30fafda089f4c615e4b954","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/3ce325ab-2b48-40b8-a0c2-7fd96fcd19ab/retrieve","description":"Researchers and policymakers have long understood the benefits of crop insurance but have been consistently disappointed by the poor performance of these programs. Rarely have programs seen sizeable take-up rates without support through large government subsidies, and in many countries, demand has been meager even at prices well below fair-market rates. Experiences from India have largely followed this trend, despite a number of large policy initiatives. Limited demand stems from low perceived value, arguably because the existing insurance products are unsuited to farmers’ needs. The present study fills an important gap in rural development by improving upon existing insurance policy design by incorporating product characteristics better suited to farmers’ preferences. To do so, we conducted a discrete choice experiment with agricultural households in four states in India. While farmers seem to like several of the features of policies offered under existing programs, our results suggest they would generally be willing to pay more than the highly subsidized rate they currently pay and are also clearly dissatisfied with delayed and uncertain indemnity payments and would be willing to pay a significant premium for more assured and timely payment delivery.","id":"-1119488321"},"keywords":["Q10","Q11","Q18 Crop insurance","discrete choice experiments","willingness-to-pay","India"],"sieverID":"766578d3-7e00-4868-a572-b6b4bd5aa9e7","pagecount":"43","content":"established in 1975, provides research-based policy solutions to sustainably reduce poverty and end hunger and malnutrition. IFPRI's strategic research aims to foster a climate-resilient and sustainable food supply; promote healthy diets and nutrition for all; build inclusive and efficient markets, trade systems, and food industries; transform agricultural and rural economies; and strengthen institutions and governance. Gender is integrated in all the Institute's work. Partnerships, communications, capacity strengthening, and data and knowledge management are essential components to translate IFPRI's research from action to impact. The Institute's regional and country programs play a critical role in responding to demand for food policy research and in delivering holistic support for country-led development. IFPRI collaborates with partners around the world.Indian agriculture presently is marred by climate risks. Evidence seems to suggest that Indian agriculture may be in the midst of a transition to a new monsoon normal: in five of the six years between 2009-2015, monsoon rains have been weak and unevenly distributed over both time and space, and three of the seven years from 2009 through 2015 have been officially designated as all-India drought years. Total rice production has suffered as a result of these vagaries in monsoon rainfall, both as a result of decreases in harvested area as well as through reductions in rice yields.The erratic pattern of rainfall in the last few years imposes several limitations on the cultivators' decision-making behavior with regards to investments in their crops. Their perceptions of multifarious risks such as droughts, floods, market prices etc. may serve as decisive factors in their input decisions. The effects of these can be seen in their reduction of farm investments in higherrisk higher return activities, such as higher-yielding seed varieties, investing in irrigation or new machinery. It is in scenarios such as this, that crop insurance becomes relevant. By providing protection against unforeseen weather-related circumstances, insurance serves as a risk transferal mechanism, thereby encouraging farmers to undertake investments in high-value cultivation. It should therefore be expected that in setups where crop insurance is available, there is an increasing demand for the same. However, India's experience with crop insurance has not always followed an upward trend.Agricultural insurance, specifically insurance against crop loss, has been around for many years in India. Pilot crop insurance programs implemented since 1972-73 led to the first major government crop insurance program in 1985-86, the Comprehensive Crop Insurance Scheme (CCIS) that was subsequently replaced by the improved National Agricultural Insurance Scheme (NAIS) in 1999-2000 (Nair, 2010a;Sinha, 2004). These programs used an ''area approach\", whereby insurance payouts are made to all farmers in an area where average yields fall below the guaranteed yield (Nair, 2010a). Despite having several national-level programs to promote insurance, however, only about 20 percent of gross cropped area was covered under various insurance schemes as recently as 2014. The Indian Prime Minister at the time, Narendra Modi, launched a new policy in 2016 -Pradhan Mantri Fasal Bima Yojana (PMFBY) -which replaced the NAIS. 1 The existing agricultural insurance scheme was revamped by aggregating all existing variants. Since its launch, approximately 55 million farmers have been insured, exceeding enrollment from the previous scheme by approximately 40 percent. On an average in the first two years of operation, 55 million hectares of cultivable land were insured. Opening up for participation from private insurance companies and heavily subsidizing the cost of insurance are perceived to be the key factors affecting this rise in uptake. Under this scheme, farmers pay a premium of maximum 2 percent of the sum insured during monsoon season (also known as kharif) sowing, 1.5 percent of the sum insured during the winter season (also known as rabi) sowing for food and oilseed crops, and a maximum of 5 percent of the sum insured for commercial crops, regardless of season. 2 The difference between actuarial premium rates and the farmer shares is shared equally between the union and the federal state governments. As a result, in the year 2017-18, the total premiums collected by all insurers together was INR 232 billion (USD 327 billion) of which farmers' share was INR 39 billion (USD 55 billion) and a government subsidy of INR 193 billion (USD 272 billion). By paying a premium of INR 700 (USD 10) a farmer can insure up to INR 35,000 (USD 500) of losses per hectare, assuming effective execution. Loss assessment and indemnity payments are calculated based on crop-cut experiments at the panchayat level (i.e., a collection of nearby villages). Yet only 25 percent of insured farmers purchased insurance of their own volition; the remaining 75 percent were insured as part of compulsory default coverage under the scheme where any farmer who has applied for seasonal agricultural credit is mandated to purchase insurance coverage, often without their explicit knowledge. Experiences with crop insurance thus vary widely between credit-recipients (loanees) and those not availing credit (non-loanees).This scenario raises a few interesting questions. Why is voluntary participation still very low when premiums for farmers are quite affordable for most? Is it the quality of insurance product and not the WTP that is a deterrent? Are certain characteristics of the insurance scheme more problematic than the others? What features should comprise an optimal agri-insurance scheme in India, and are there possibilities for trade-offs between the various components of the policy? This study attempts to answer some of these important and highly policy-relevant questions.At present, no assessment has been done in India to determine farmers' true WTP for comprehensive multi-peril insurance policies. We try to fill this gap by conducting discrete choice experiments with farmers across four geographically variant states in India. The methodology is designed to gauge their valuation for not only the totality of a multi-peril insurance product, but also the specific characteristics, such as the coverage period, method of yield loss assessment, total sum insured, levels of actuarially fair premium rates and timing of insurance payouts. While our choice sets are agnostic to any specific insurance scheme, they include all the important attributes that are present in the current large-scale new insurance program in India. Our results, therefore, not only contribute to the broader literature on WTP for multi-peril crop insurance, but perhaps are especially valuable for actual policy makers to optimize insurance design.While insurance is relevant as one of the many tools to manage and cope with all forms of agricultural risks, insurance demand behavior gets complicated by compounding of multiple factors, such as adverse selection and ambiguity aversion behavior (Elabed and Carter, 2015), thereby reducing the WTP for insurance. These traditional indemnity-based insurance programs are subject to a myriad of well-documented challenges, including information asymmetries in the form of moral hazard and adverse selection (Hazell, 1992;Morduch, 2006;Barnett et al., 2008;Miranda and Farrin, 2012). Traditional indemnity-based crop insurance programs are also prone to other challenges, including high administrative costs (in particular, the cost of assessing losses), and the covariance of insured farmers' risks that increases the insurers' risk of insolvency or, at the least, increases their costs of reinsurance. All of these challenges are perhaps most pronounced in developing countries, where information asymmetries, knowledge gaps, and other structural and operational issues are even more widespread. Moreover, despite a great deal of research, there remains relatively scant evidence to suggest that traditional crop insurance positively affects farmer welfare in either developed or developing countries (Hazell, 1992;Skees et al., 1999;Smith and Watts, 2009). Most crop insurance programs in the developed world have been propped up by large government subsidies, and many developing countries exploring such programs are following suit. Given this, and the scant empirical evidence of high WTPs for multi-peril crop insurance in developing countries, it is difficult to predict the commercial viability of such programs, particularly since even evidence from developed countries suggests that risk aversion among farmers is not high enough to pay for purely private actuarial premiums (Goodwin, 2001, Smith andGlauber, 2012).In an assessment of Australian wheat farmers, Patrick (1988) found almost negligible willingness to pay full costs of offering insurance above the actuarially fair premium, and no buyers in the instance of loading factor exceeding 20 percent. In another assessment in Australia, farmers were not willing to pay higher than 5 percent of the actuarially fair premium (Bardsley et al., 1984). Smith and Goodwin (1996;2010) assessed crop insurance in the US and found that farmer's willingness-to-pay (WTP) for multi-peril risk protection was not higher than the costs of bearing an insurance program. This, they argue, is not indicative of farmer's risk aversion, but instead a reflection of farmers' alternative risk management mechanisms such as diversification, off-farm employment or self-insurance. As a result, there is hardly any multi-peril crop insurance scheme that is not highly subsidized. In the US and Canada the average subsidy rates have been around 60 percent in recent times, in Spain and Portugal nearly 70 percent and in Japan roughly 50 percent (Du et al., 2016, Mahul andStutley, 2010).In developing countries, crop insurance is one of the many tools governments use to smooth farm incomes such as quotas, minimum price support systems, input subsidies and low interest agricultural loans, among others (Mahul and Stutley, 2010). In the presence of these, it is difficult to determine the real demand for insurance. Some options may also promote moral hazard where a combination of high input subsidy, low interest loans and insurance lead to poor management practices in a low investment -assured return setting (Hazell and Hess, 2010). In Burkina Faso, Sakurai and Reardon (1997) find that expectation of public food aid reduced the demand for drought insurance. This is known as a 'Samaritan's Dilemma' (Coate, 1995). This is especially relevant where area-based approaches are prevalent for both crop insurance and disaster payments, such as loan waivers. In such a context, low risk farmers who are indemnified in an insured area may simply want to wait for a low probability disaster payment rather than investing in crop insurance. Self-insurance (grain storage, livestock sales or social networks) is another factor that conflates with demand for formal insurance in the developing countries (Kazianga andUdry, 2006, Ambrus et al., 2014). Where insurance is compulsory, that is, bundled together with crop loans, low risk farmers that have not applied for credit may not want to buy insurance knowing it crosssubsidizes high risk farmers (Report GoB, 2009). In South Asia, a peculiar interaction further complicates the understanding of insurance demand: forgiveness of agricultural credit. This hampers repayment culture and solvency of banks (Report GoB, 2009) while at the same time not translating into higher agricultural investments or productivity (Kanz, 2016). In some recent interactions with farmers in the Indian state of Karnataka, the authors discovered that indebted farmers do not visit rural banks (that are in-charge of dishing out insurance). This is out of, both, fear of having to repay and hope that there will be a political intervention near to an electoral event when outstanding loans would be forgiven (IIMA, 2018).This implies that estimating insurance demand through observed prices (in this case, premium rates) may not yield reliable results. Therefore, in recent times, there have been some, though very limited, efforts to estimate demand or the WTP for insurance using direct valuation methods, such as contingent valuation (CVM) or discrete choice experiments (DCE). Liesivaara and Myyrä (2017) conducted a split sample DCE to include disaster aid as a constant variable in estimating WTP for different attributes of a crop insurance product in Finland. They found that expectations of disaster relief meant farmers would be less worried about crop losses. In such a situation, premiums would have to be highly subsidized for insurance take-up implying expansive use of taxpayers' money for very low marginal benefits, which has implications for policy. Fahad and Jing (2018) use a CVM to estimate the possible premium range that farmers would be willing to pay to insure themselves from risks of flooding in a high flood prone region of Pakistan. Those who said 'yes' to participation in an insurance product, were given six starting bid levels in the range of 0.07 to 0.71 USD as options for the monthly premiums. There were lower and upper bounds to the dichotomous choice bids and for each choice of premium, the reasons for rejecting a higher bid were asked. This helped reveal that access to credit, irrigation, exposure to path adverse weather events and other socio-economic constraints affected insurance demand of a farmer. Arshad et al. (2016)  The Liesivaara and Myyrä (2017) study evaluates attributes of an insurance product, but the relevance of the insights are largely limited to the EU context. Moreover, the focus has been on interaction with co-risk mitigation options. The studies in Pakistan by Arshad et al., 2016 andFahad andJing, 2018, on the other hand offer meaningful insights for insurance demand in a developing country context, but have two limitations: first, the assessment is for two named perils, floods and droughts, thus limiting insights on multi-peril insurance products; second, they adopt a holistic CVM approach which can only speak generally of WTP for insurance. It gives no insights on how farmers value the various attributes (such as coverage period, timeliness of indemnity payments or yield loss assessment accuracy) within an insurance product, and thereby, does not help in optimizing insurance design. An assessment of WTP for multi-peril insurance in India, that also evaluates the preferences for attributes, is therefore, novel and helps to understand insurance demand behavior in developing countries more comprehensively. This is especially true in a context where one of the largest government subsidized multi-peril insurance programs in the world is currently operational. It provides an opportunity to validate outcomes and experiment with optimized insurance design.The present study uses discrete choice experiments to better understand Indian farmers' preferences for various elements of crop insurance. Discrete choice experiments allow researchers to analyze stated preferences for products or services, but beyond that they allow researchers a means for parsing out preferences for specific characteristics or attributes of a good or service.This is particularly useful if the researcher believes, as Lancaster (1966) suggested, it is not the good or service that is the object of utility, but rather it is from the underlying characteristics of the good or service from which utility is derived. In a discrete choice experiment, preferences are elicited through survey participants' responses to a series of hypothetical choice scenarios. These survey-based exercises are referred to as experiments because the researcher controls the combination of product characteristics to which the survey participant is exposed.We assume that observed choices arise from a process of utility maximization (McFadden, 1974).Specifically, within the context of a discrete choice experiment, it is assumed that the observed (stated) choice that an individual makes within a particular choice scenario is the choice that, on average, maximizes her utility among the set of potential alternatives. Utility can be conceived of as consisting of both a systematic, deterministic component, and a stochastic component. The deterministic component reflects individual tastes and preferences that map the expression of product characteristics directly into utility, while the stochastic component reflects, among other things, random variations in tastes and preferences and errors in optimization. We can write our random utility model as:where \uD835\uDC62\uD835\uDC62 \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 is the observed indirect utility (i.e., the utility of the utility maximizing option \uD835\uDC57\uD835\uDC57)obtained by individual \uD835\uDC56\uD835\uDC56 during choice scenario \uD835\uDC61\uD835\uDC61; \uD835\uDC5D\uD835\uDC5D \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 is the price of option \uD835\uDC57\uD835\uDC57 faced by individual \uD835\uDC56\uD835\uDC56 during choice scenario \uD835\uDC61\uD835\uDC61; \uD835\uDC65\uD835\uDC65 \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 is a vector of (non-price) insurance policy characteristics or attributes; \uD835\uDEFC\uD835\uDEFC \uD835\uDC56\uD835\uDC56 represents individual \uD835\uDC56\uD835\uDC56's preferences for policy price; \uD835\uDEFD\uD835\uDEFD \uD835\uDC56\uD835\uDC56 is a vector of preference weights for the corresponding elements of \uD835\uDC65\uD835\uDC65 \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 ; and \uD835\uDF00\uD835\uDF00 \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 is a Gumbel (Extreme Value Type I) distributed error term with farmer-specific variance \uD835\uDC49\uD835\uDC49\uD835\uDC49\uD835\uDC49\uD835\uDC49\uD835\uDC49(\uD835\uDF00\uD835\uDF00 \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 ) = \uD835\uDF0E\uD835\uDF0E \uD835\uDC56\uD835\uDC56 2 (\uD835\uDF0B\uD835\uDF0B 2 /6), where \uD835\uDF0E\uD835\uDF0E \uD835\uDC56\uD835\uDC56 is a farmerspecific scale parameter. In many applications, it is assumed that there is no heterogeneity in this scale parameter, and furthermore the scale parameter is simply normalized to 1 for ease of computation (i.e., \uD835\uDF0E\uD835\uDF0E \uD835\uDC56\uD835\uDC56 = \uD835\uDF0E\uD835\uDF0E = 1). Taking partial derivatives of \uD835\uDC62\uD835\uDC62 \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 with respect to the attributes provides estimates for the change in utility associated with incremental changes in the expression of the attributes; in other words, the \uD835\uDEFD\uD835\uDEFD \uD835\uDC56\uD835\uDC56 terms can be directly interpreted as marginal utilities. The ratio of two marginal utilities is directly interpretable as the marginal rate of substitution between the two attributes (i.e., the rate at which an individual would be willing to give up a unit of the attribute in the denominator to acquire an increment of the attribute in the numerator). If one of the marginal utilities is the marginal utility of income, then the marginal rate of substitution with respect to income is an estimate of WTP. We are rarely able to directly observe the marginal utility, but this can be proxied by the marginal disutility of product cost. Since cost is almost always deemed to be one of the important features driving purchase decisions, it is almost universally included as an attribute in a DCE. An estimate for the WTP for a specific attribute would therefore just be the ratio of the marginal utility of the attribute to the marginal disutility of product price.If one assumes that, in addition to homogeneity in the scale parameter, preferences are fixed in the population, then estimating marginal utilities and arriving at estimates for WTP is relatively straightforward using conditional logit estimation. The assumption of fixed (or constant)preferences in the population is quite restrictive, however, and imposes some potentially unrealistic assumptions on, among other things, the substitution patterns that are permitted by the model. A common approach to incorporating preference heterogeneity is to estimate the choice model using a mixed logit (also known as a random parameter logit) model. Under this approach, the researcher assumes a distribution for the preference parameters, and derives an estimate for WTP as the ratio of the random parameters. This approach, however, can lead to distributions for WTP that have undefined moments (e.g., the ratio of two normally distributed random variables takes a Cauchy distribution, for which neither the mean nor the variance are defined).Even if we permit preference heterogeneity, there is still the potential violation of scale homogeneity. If we permit scale heterogeneity, then we cannot simply proceed with a conventional mixed logit estimator. Note that, since utility is ordinal, we can divide equation ( 1) by the scale parameter to obtain a scale-free equivalent (Scarpa et al., 2008):where, now, \uD835\uDF08\uD835\uDF08 \uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56\uD835\uDC56 is an i.i.d error term with constant variance \uD835\uDF0B\uD835\uDF0B 2 6 ⁄ . We can re-write the re-scaled utility coefficients asImportantly, note that if \uD835\uDF0E\uD835\uDF0E \uD835\uDC56\uD835\uDC56 varies randomly in (2), the utility coefficients in (3) will be correlated, since \uD835\uDF0E\uD835\uDF0E \uD835\uDC56\uD835\uDC56 enters into the denominator of each of the re-scaled utility coefficients. Even if \uD835\uDF0E\uD835\uDF0E \uD835\uDC56\uD835\uDC56 does not vary, the utility coefficients could still be correlated simply due to correlations among tastes for various attributes (Scarpa et al, 2008). Since the WTP for a given attribute is the ratio of the marginal utility of that attribute to the marginal (dis-)utility of the policy price, we can write \uD835\uDEFE\uD835\uDEFE \uD835\uDC56\uD835\uDC56 = \uD835\uDF13\uD835\uDF13 \uD835\uDC56\uD835\uDC56 /\uD835\uDF06\uD835\uDF06 \uD835\uDC56\uD835\uDC56 , and can re-write (3) aswhich re-parameterizes utility in WTP space rather than preference space (Train & Weeks, 2005;Scarpa et al., 2008). Now, rather than assuming the distributions for the marginal utilities (the \uD835\uDEFD\uD835\uDEFD \uD835\uDC56\uD835\uDC56 terms), the researcher can directly specify the distribution for (individually-scaled) WTP (the \uD835\uDEFE\uD835\uDEFE \uD835\uDC56\uD835\uDC56 terms) without having to worry about ratio distributions with undesirable properties.Consequently, researchers have much more direct control over the distributional features of marginal WTP in the underlying population under this specification than they would otherwise (Thiene & Scarpa, 2009). Furthermore, Train & Weeks (2005) and Hensher and Greene (2011) have found that this transformed model generally produces more reasonable estimates of WTP than when WTP is calculated as the ratio of utility parameters. This model can then be estimated by appealing to the generalized multinomial logit (GMNL) model developed by Fiebig et al. (2010), which Hensher and Greene (2010) have demonstrated is a generalization of choice models estimated in both preference space as well as WTP space.To allow for even greater flexibility in estimation, we consider the possibility that the randomly distributed WTPs for the different insurance product attributes could be correlated. As was previously mentioned, if there is scale heterogeneity, then WTPs will be correlated by definition, and even if there is no scale heterogeneity, there is the possibility for correlated WTPs simply due to correlation among preferences for different attributes. Hensher, Rose, and Greene (2015) have further noted that, in virtually all data sets, there are likely unobserved effects that are correlated among alternatives in a given choice situation and allowing for WTP parameters to be correlated is one way to account for this. Failing to control for this can lead to imprecise estimates of WTP, which has obvious implications for the reliability of the policy implications that be derived from these estimates (Mariel & Meyerhoff, 2018).While there are potentially innumerable different dimensions with which to characterize and differentiate insurance products, to maintain tractability we are necessarily limited in the scope of attributes over which we can attempt to elicit preferences. As such, we narrowed the field of potential attributes to those which we assumed would be particularly salient in farmers' minds when they evaluate risk management alternatives. In particular, we were interested in estimating farmers' preferences for the insurance coverage period, the method of loss assessment, the delivery of insurance payments, the coverage amount (referred to in the Indian context as the insured sum), and the cost of insurance. For the coverage period, there are several potential alternatives that insurance providers could consider. For example, under PMFBY, insurance covers the entire period from pre-sowing until after harvest. Other alternatives could include only the period from sowing until harvest, or merely pre-sowing or post-harvest.For the method of loss assessment, we consider not only the crop-cutting experiments at the village or panchayat level that are currently being utilized under PMFBY, but also loss assessments from remote (e.g., satellite-based) sensors, as well as rainfall-based indices, in which insurance payments could be issued if rainfall at the district level falls below 75 percent of long-run historical averages. While crop-cutting experiments may provide loss assessments that are highly correlated with a particular farmer's on-field experiences, they are very costly to administer and highly susceptible to accidental and purposeful measurement errors. Other methods for assessing losses, such as using remote sensors, are generally quite inexpensive and may eliminate problems of moral hazard and adverse selection, but typically either lack a high degree of transparency (in the case of remote sensing technologies) or exhibit relatively low correlations with actual on-farm performance (in the case of rainfall-based weather indexes), especially in irrigated agricultural systems.The timing of insurance payments -in particular the long delays that farmers endure -has been often identified as a problematic feature of crop insurance in India, including under PMFBY. We were interested in seeing whether farmers would be willing to pay a premium for an insurance policy that would provide assured, timely payments if farmers experienced a loss during the coverage period. In the discrete choice experiment, we allowed for insurance policies to provide indemnities within six weeks of losses being assessed, with 100 percent certainty, or for a 50 percent chance that payments will arrive within six weeks of losses being assessed, and a 50 percent chance that payments will be delayed for more than 6 months.Perhaps of greatest interest to both insurance providers and their clients is the cost of insurance.Controlling for policy price is important because it allows for direct estimation of a monetary welfare metric -namely, WTP. Recall that the preference parameters \uD835\uDEFD\uD835\uDEFD can be interpreted as marginal utilities. The ratio of any two marginal utilities can be interpreted as the marginal rates of substitution. The ratio of the marginal utility of a policy attribute with respect to the marginal utility of income (or the marginal disutility of policy cost) provides a direct estimate of the amount of money an individual would be willing to give up (or would demand) in exchange for an incremental increase in the expression of the policy attribute. While farmers are currently only asked to pay a mere 1.5-2.5 percent of the insured sum as premium under PMFBY, the very low and seemingly arbitrary figures might not truly reflect the value that farmers derive from crop insurance. In our discrete choice experiment, we allowed for the policies to have three different premium rates, including 2.5 percent, 4 percent, and 10 percent.We also included in our experiment a variable capturing the insured sum of the hypothetical insurance policies. This is not because we were especially interested in preferences for larger policies versus smaller policies (we would assume a priori that larger payouts would be preferable to smaller payouts), but more so because we needed for there to be a baseline against which the study participants could evaluate the policy premium and the other insurance policy characteristics.In our experiment, we allowed for the insured sum to take three possible levels, specifically INR 20,000, INR 30,000, or INR 40,000 per hectare. Table 1 summarizes the various attributes of an insurance policy and their various level included in this experiment.Expanding equation ( 4) based on the above discussion of product attributes, our base utility function accounting only for main effects can be written as:where \uD835\uDC36\uD835\uDC36\uD835\uDC36\uD835\uDC36\uD835\uDC36\uD835\uDC36 1 , \uD835\uDC36\uD835\uDC36\uD835\uDC36\uD835\uDC36\uD835\uDC36\uD835\uDC36 2 , and \uD835\uDC36\uD835\uDC36\uD835\uDC36\uD835\uDC36\uD835\uDC36\uD835\uDC36 3 are binary variables corresponding to insurance coverage from sowing to planting, coverage during pre-sowing, and coverage during post-harvest, respectively, with the coverage period extending from pre-sowing to post-harvest serving as the reference category. Similarly, \uD835\uDC3F\uD835\uDC3F\uD835\uDC3F\uD835\uDC3F 1 and \uD835\uDC3F\uD835\uDC3F\uD835\uDC3F\uD835\uDC3F 2 are binary variables corresponding to loss assessments from remote sensors and rainfall-based indices, respectively, with loss assessments from crop-cutting experiments at the village or panchayat level serving as the reference category. \uD835\uDC47\uD835\uDC47\uD835\uDC56\uD835\uDC56\uD835\uDC43\uD835\uDC43\uD835\uDC56\uD835\uDC56\uD835\uDC47\uD835\uDC47\uD835\uDC47\uD835\uDC47 is a binary variable equal to one if the insurance payment is guaranteed to be delivered within six weeks of the loss assessment, and zero otherwise. The \uD835\uDC43\uD835\uDC43\uD835\uDC49\uD835\uDC49\uD835\uDC43\uD835\uDC43\uD835\uDC43\uD835\uDC43\uD835\uDC56\uD835\uDC56\uD835\uDC62\uD835\uDC62\uD835\uDC43\uD835\uDC43 and \uD835\uDC46\uD835\uDC46\uD835\uDC62\uD835\uDC62\uD835\uDC43\uD835\uDC43 terms are continuous variables capturing the monetary cost farmers are required to pay insurance and the insured sum, respectively. While the premium attribute was previously discussed percentage rate of the insured sum, this rate was converted into a monetary figure when participants completed choice tasks by multiplying the premium rate by the insured sum for each choice alternative.Because a full factorial experimental design -consisting of all possible combinations of insurance policy attributes across the competing alternatives -would be intractable in any real-world research setting, we set out to create a fractional factorial design that satisfied some wellestablished design criterion. In our particular case, we specified an orthogonal experimental design with three hypothetical alternatives in each choice set, and underlying utility functions consisting of all main effects and first-order interaction effects. 3 The experimental design was generated using Ngene 1.1.2, a software package specially-designed for generating discrete choice experiments (Ngene, 2014). In sum, the experimental design resulted in a total of 36 unique choice sets, each consisting of three hypothetical alternative insurance policies. The 36 choice sets were blocked into six groups of six choice sets each. Each household in the sample was then randomly allocated to one of the six groups, and then would be expected to respond to the 6 choice sets assigned to that specific choice set group.The data used in the present study come from a household survey conducted across four Indian states (Gujarat, Himachal Pradesh, Karnataka, and Uttar Pradesh; see Figure 1). While not intended to be nationally representative, the diversity of state coverage allows for some heterogeneity in agro-ecological and social conditions. The survey and discrete choice experiment were conducted 3 Orthogonal experimental designs have the properties of attribute balance and independent estimability of all parameters. In practice, this implies that each attribute column in the design matrix is uncorrelated.each of the four states, two districts were sampled based on their share of primary crop cultivation in the state. Within each district, two blocks were randomly selected, with three villages subsequently selected at random from each block. Within each village, 12 households were randomly selected from village lists, resulting in an initial sample of 576 households. Respondents were administered a set of survey questions that sought information on their demographic characteristics, cultivation practices, household asset ownership, income sources and experience with insurance policies, translated in their local languages. The discrete choice experiments were also administered using standardized, scripted protocols. Both the survey questions and the discrete choice experiment were administered in the primary language in each state (Gujarati in Gujarat, Kannada in Karnataka, and Hindi in both Uttar Pradesh and Himachal Pradesh). Table 2 summarizes the characteristics of households in the sample on both a pooled and statewise basis.Table 3 presents results our base estimates of WTP for insurance product features under two specifications. Column (1) reports results from a generalized mixed logit regression assuming that WTP parameters are uncorrelated, while column (2) reports results permitting free correlation of the WTP parameters. Under both specifications, we permit both preference and scale heterogeneity, with the WTPs assumed to be normally distributed. The upper panel in Table 3 reports the estimates of the mean WTP for the corresponding attributes, while the lower panel reports the corresponding distribution parameters (standard deviations).Although there is not a sizeable difference in model fit between the two specifications, nor are there dramatic changes in the WTP coefficient estimates, the model permitting free correlation in WTP parameters is superior to the more restrictive specification assuming that the WTP parameters are uncorrelated based on a likelihood ratio test. Additionally, when we permit correlations, all of the means and standard deviations for the WTPs are statistically significant, indicating that not only are the mean WTPs significantly different from zero, but also that there is a significant amount of heterogeneity in insurance policy preferences within the population.Indeed, when we allow for correlated WTP parameters, the standard deviations of the WTP distributions are all greater than when we preclude correlations. For the discussion that follows, we will rely upon the estimates assuming the WTP parameters are correlated. Figure 2 illustrates the empirical distributions of farmers' WTP for the various insurance policy attributes considered in the present study.Interestingly, the estimates suggest that farmers would be willing to pay significantly higher premiums than they currently are asked to pay. The coefficient of sum insured is approximately 0.1, meaning that if the sum insured rises by one unit (here, 1 unit=INR 1,000), they would be willing to pay roughly INR 100, or about 10 percent of the sum insured. While that is close to the actuarially-fair cost of insurance (Joint Group, 2005), it is still quite low relative to what would likely be needed for insurance to be commercially viable without large government subsidies, but the premium is about 5 times higher than what farmers are required to pay for crop insurance under PMFBY. For example, for a base policy with an insured sum of INR 30,000 per hectare, these results suggest that farmers would be willing to up to INR 3,000 per hectare. Under PMFBY, where farmers are typically only asked to pay at most a premium of 2 percent of the sum insured, they cost to farmers would only be INR 600 for a comparable policy.For insurance policy attributes that enter the utility function as binary variables (coverage level, loss assessment, and timing of indemnity payments), the omitted category is always the status quo under the existing policy regime. Consequently, many of the estimated WTPs reported in Table 3 can be interpreted either as premia that individuals would be willing to pay (in the case of guaranteed indemnity payments within 6 weeks of loss assessment) or discounts that would be demanded (in the case of coverage period or loss assessment methods) for alterations of the status quo insurance policies. For example, farmers in our sample would require discounts for policies with shorter coverage periods (i.e., anything other than pre-sowing to post-harvest). These required discounts are quite sizeable for policies that cover only the pre-sowing period (INR 5,000) or only the post-harvest period (INR 4,800). Both of these required discounts exceed the INR 3,000 that we might expect farmers to be willing to pay for a hypothetical base policy with an INR 30,000 insured sum. Even in the choice experiment itself, the highest possible cost that farmers faced was INR 4,000 (a maximum 10 percent premium on a maximum insured sum of INR 40,000 per hectare). The fact that these implicit discount requirements are so large speaks volumes about farmers' preferences and clear dissatisfaction with the limited coverage offered by these policies.Other things equal, therefore, we would not expect farmers to purchase any policy that covers only these tail ends of the agricultural season. There is a smaller (though still nontrivial) discount requirement for policies covering cropping from sowing to harvest (INR 1,000). In sum, though it seems farmers would not be interested in policies that protect against risks either leading up to or following the monsoon season, they also clearly perceive some risk of crop loss due to sources apart from just rainfall (whether deficiencies or excesses) which presumably would be covered by a policy covering the sowing to harvest period.Farmers would demand a discount for insurance that assessed losses by remote sensing or rainfall indices, though the discount that would be demanded is not huge (roughly INR 400 for each of the two alternative methods). Farmers evidently prefer to have losses assessed by crop-cutting experiments conducted at the panchayat level, despite the fact that the costs for crop-cutting experiments make the insurance policies more expensive for the insured. While it is difficult to estimate the cost of conducting a crop-cutting experiment, and the costs will likely vary widely from state to state, we can roughly estimate the cost based on historical data. For example, we know that a crop-cutting experiment cost about INR 300 in 2004(World Bank, 2007). Based on general inflation levels over the ensuing 14 years, this corresponds to a 2018 equivalent price of about INR 500 per crop-cutting experiment. This assumes that, for example, wages for the agricultural workers responsible for the crop cuts just keep up with the cost of living. For remotely sensed or index-based loss assessments, the variable costs are essentially nil. Consequently, the administrative loads on these types of policies would be considerably less, thereby lowering the cost of insurance -something that would clearly be attractive for farmers. Whether the reduction in farmers' price would exceed the discount they would demand due to a preference for cropcutting experiments remains to be seen.Farmers would be willing to pay substantially more for insurance if they could believe that payments would be timely. On average, farmers would be willing to pay a premium of over INR 1,000 if indemnity payments would be guaranteed within 6 weeks of the loss assessments. This is quite important, and has implications for the ultimate design of improved crop insurance policies.One of the primary concerns that has arisen with regards to assessing losses by crop-cutting experiments is that these can take a long time and lead to delayed indemnification. Other methods for assessing losses (e.g., by remote sensing or based on parametric weather indices) can facilitate much more rapid payment deliveries.As previously discussed, column (2) in Table 3 reports WTPs allowing for preferences for the different policy attributes to be correlated.Table 4 reports the covariance, correlation, and Cholesky (lower) decomposition matrices that reflect these correlations. In particular, the correlation matrix provides details into how preferences co-move. For example, preferences for certain and timely indemnity payments are negatively correlated with preferences for all the other product features, except loss assessments via remote sensing. This correlation is not especially strong (0.127), but it is nonetheless interesting that increasing valuations for guaranteed, timely indemnity payments are positively correlated with insurance policy designs that increase the likelihood of timely payment delivery.The Cholesky decomposition matrix provides information about the degree of variation directly attributable to the different attributes. The first element is simply the standard deviation for the random WTP coefficient associated with the sowing to harvesting coverage level. Subsequent diagonal elements represent the amount of variance attributable to random WTP coefficients once the correlations with the other coefficients have been removed. The off-diagonal elements represent the amount of cross-coefficient correlation that was previously confounded with standard deviations for models not controlling for these correlations (Hensher et al., 2005). For example, the amount of variance directly attributable to the indemnity payment timing random WTP coefficient is not 0.767, as would be suggested based on just examining the standard deviation of the WTP distribution (from Table 3), but is really 0.645: there are negative portions due to negative correlations with the various coverage periods, loss assessments from weather indices, and sum insured that would otherwise be confounded within the standard deviation estimate if the correlation were not accounted for.In addition to controlling for correlated preferences, it may also be useful to consider the salience of the different attributes when respondents are evaluating choice scenarios. To evaluate this, we consider inferred attribute non-attendance using the method proposed by Hess & Hensher (2010).We use the individual-level estimates of the WTP distributions for the various insurance policy attributes, and compare the variation in individual WTP estimates relative to the expected WTP level. Specifically, we compute a noise-to-signal ratio by dividing the standard deviation of the WTP distribution by the mean WTP for each individual and for each policy attribute.Table 5 reports the proportion of respondents in the sample who were deemed to have ignored the different insurance policy attributes based on this procedure. Most of the insurance policy features appear to be quite strongly attended to. Fewer than 5 percent of respondents appear to have ignored the timing of indemnity payments, and virtually no one appears to have ignored the sum insuredobviously an important feature of any insurance policy. Interestingly, there are mixed results when it comes to the coverage period. Nearly 25 percent of respondents appear to have ignored the coverage period if it covered the sowing to harvest period, but almost no one appears to have ignored the coverage period if it was either only the pre-sowing period or only the post-harvest period. For those that ignore the sowing to harvest coverage period, we can infer that they essentially view the policy as indistinguishable from one that covered the full pre-sowing to postharvest period (like those policies offered under PMFBY) and would therefore not demand a discount on the purchase of such a policy. A similar phenomenon arises for the alternative methods of loss assessment. Nearly 25 percent of respondents ignored each of the two alternative loss assessment methods. It might be tempting to suspect that these respondents are just not paying attention to the method with which losses are assessed, but this is not entirely accurate. There is essentially no correlation between respondents' behavior when it comes to attending to or ignoring these two loss assessment methods. In other words, rarely are those that ignored loss assessments via remote sensing the same individuals who ignored loss assessments via weather-based indices.Regardless, if we also infer that individuals view these loss assessment methods as indistinguishable from crop-cutting experiments, then this might further strengthen arguments for moving from expensive and time-consuming -not to mention rather subjective and opaque -crop-cutting experiments to other methods of loss assessment that are more economical and can facilitate more rapid indemnity payments.Using the individual-level conditional estimates of WTP, we next aim to isolate any systematic correlates of farmers' WTP for the difference insurance policy attributes. In so doing, we assume that attribute non-attendance (as defined above) indicates that the farmer would neither be willing to pay a premium nor demand a discount for the policy feature vis-à-vis the status quo policy.Table 6 gives the estimates of the OLS regressions of individual WTP for various attributes as a function of various household characteristics. Each column pertains to the regression for the WTP of the different insurance policy attributes. By and large, the results do not suggest much in the way of systematic determinants of WTP for the insurance policy features considered in the present study, perhaps confirming the old adage de gustibus non est disputandum. 4 Where there are some interesting and statistically significant effects that emerge are in regards to WTP for insurance policies with alternative methods of loss assessments. In particular, there are interesting results that emerge based on farmers primary crop. We find that farmers who primarily cultivate rice during the monsoon season have a significantly higher WTP for insurance policies with loss assessments based on remote sensing (vis-à-vis farmers who primarily cultivate non-cereals during the monsoon season). This is a promising result, given that there have already been researchers and development practitioners working on the ground in India and other countries (largely in southeastern Asia) piloting remote sensing for rice yield prediction to eventually inform crop insurance programs. 5 The predictive accuracy of the remote sensing technologies has been rather encouraging, with predictive accuracy ranging between 85 percent and 96 percent across three sites in Tamil Nadu, India when predictions were made at the block (sub-district administrative unit), compared with accuracy of 87 percent when predictions were made at the district level (Pazhanivelan et al., 2015). Other recent research has found remote sensing yield prediction accuracy in rice as high as 95 percent in China (Huang et al., 2013). The high spatial resolution and high -and increasing -predictive accuracy of these remote sensing technologies is a promising development, and the preeminence of rice cultivation across much of India would suggest a nascent market for such products.We also found that farmers who primarily cultivate maize during the monsoon season have a significantly lower WTP for insurance policies with loss assessments based on weather indices (vis-à-vis farmers who primarily cultivate non-cereals during the monsoon season). Interestingly, 369 out of the total 372 farmers who primarily cultivate maize during the monsoon season are from Himachal Pradesh, a state with very large climatic variations due to differences in altitude.These large variations in climate conditions may increase the likelihood that realized weather conditions on a farmer's field may not match the realized weather conditions at the location where the weather data comprising the index are collected, thus increasing the basis risk that insured farmers could be exposed to. This may be one of the primary reasons why maize farmers may dislike index-based loss assessments.Finally, we find evidence of a U-shaped relationship between cultivated area and farmers' WTP for crop insurance policies with loss assessments based on remote sensing technologies. Initially, WTP for crop insurance based on remote sensing is declining with increasing cultivated area, but begins to increase after farm sizes exceed nearly 43 acres. There are a couple of caveats that should be considered before placing too much emphasis on this result. First, the linear area effect is not statistically significant at conventional levels, so there may not be the initial negative relationship between area and WTP. Second, very few households in our sample -and, indeed throughout much of India -cultivate areas in excess of 43 acres. Consequently, the observed relationship may be a mere statistical aberration that should not likely have any bearing on actual agricultural policy.The results above would suggest that, other things equal, farmers would generally be interested in purchasing crop insurance similar to the products being offered under PMFBY, and furthermore at premium rates higher than they are currently being asked to pay. If that is indeed the case, then it is puzzling that crop insurance coverage remains so low in India, and seems to be declining.Recent reports suggest that gross cropped area covered by insurance policies under PMFBY fell by more than 20 percent, 59.55 million hectares in 2016-17 to 47.5 million hectares in 2017-18 (Business Standard, 2018). This remains less than 24 percent of gross cropped area in the country.At the same time, the number of insured farmers has also decreased by 14 percent, from 55 million to about 48 million. Related to the sluggish -and declining -enrollments, one of the major challenges that policymakers in India face regards the long time delay in delivering indemnity payments. There is also anecdotal evidence that insurance company representatives -who take part in the crop-cutting experiments -lower the threshold limit below which indemnities are issued, so that even farmers with substantial crop losses may not qualify for payment (Business Standard, 2018).Even before these recent declines, only about one third of farmers in India were insured. One obvious reason for the lack of coverage is that many farmers simply do not know about this scheme. From our data, nearly 35 percent of farmers across these four states had never heard of PMFBY. Furthermore, because holding crop insurance is typically compulsory for farmers accessing loans, there is evidence that insurance companies do not consider non-loanee farmers to be profitable (IIMA, 2018). There is also evidence that significant transaction costs hinder the broad uptake of crop insurance in India. These transaction costs arise not only in acquiring insurance, but also in filing claims. For example, farmers may have to travel several kilometers to reach the nearest financial institution. From our data, roughly 10 percent of farmers indicated that they would not likely purchase insurance if they had to travel far to submit the requisite paperwork for acquiring insurance. Farmers are also required to submit sensitive personal details, such as Aadhar (unique identifier) numbers, bank account details, or land records. Not only would such requirements exclude farmers who do not have, for example, land title (e.g., tenant or sharecropping farmers), but many farmers are evidently sensitive to sharing this information. From our data, roughly 8 percent of farmers would be unlikely to purchase insurance if they had to submit their Aadhar details; 19 percent of farmers would be unlikely to purchase insurance if they had to submit their bank account details; and 20 percent of farmers would be unlikely to purchase insurance if they had to submit a copy of their land records. Many farmers are also averse to having to file insurance claims in-person. Over 14 percent of farmers in our sample indicated that they would be unlikely to purchase insurance if they had to personally inform the insurer of losses.How can this ambitious government policy be amended to increase coverage rates and better meet the needs of Indian farmers. Clearly, the most pressing need is to expedite the delivery of indemnity payments. But it seems unrealistic to expect this to be implemented without concomitant changes How would insurance demand and farmer welfare change as the result of such a transition. Figure 3 plots empirical demand curves for two crop insurance products: a base policy similar to those being offered under PMFBY, and an alternative policy that has been modified so that losses are assessed via remote sensing, but also one that offers a guarantee that indemnity payments will be delivered within six weeks of the losses being detected. The horizontal axis depicts the percentage of cultivated area covered (or not) under crop insurance. Consequently, at any point at which the demand curve for the alternative policy is above the demand curve for the base policy, we would expect a higher proportion of cultivated area to be insured at a given price. For virtually all prices above INR 1,800 per hectare (representative of a 6 percent premium on an insured sum of INR 30,000 per hectare), demand for the alternative policy exceeds demand for the base policy. For example, at a price of INR 2,100 per hectare (representative of a 7 percent premium on an insured sum of INR 30,000 per hectare), we would expect roughly 46 percent of cultivated area to be insured under the alternative policy, but only 43 percent of cultivated area to be insured under the base policy. This may not seem like a large margin, but achieving 46 percent coverage of cultivated area is a marked improvement on the existing achievements under PMFBY.Perhaps it is more appropriate not to compare demand at the same price, but to consider demand at different cost structures, but for example, under comparable insurance company profit margins.While these data are not readily available, we can be quite certain that the farmers' cost for the alternative policy would be considerably less than their cost for the base policy in order to secure insurance companies the same profit margin, since the firms' administrative costs (most notably associated with loss assessments and reinsurance) would be considerably lower under the alternative policy. By extension, the price they would need to charge to maintain the same profit margins that they earn under the base policy would also be considerably lower. The welfare effects of this would be sizeable, reflected in the twofold impacts of an increased insured area and a reduction in the cost of insurance for those who would already otherwise be insured.This study explores Indian farmers' demand for crop insurance, particularly in light of the muchhyped and highly ambitious Pradhan Mantri Fasal Bima Yojana (PMFBY). While India has a long history with crop insurance, the level of farmers insured under various government programs has remained disappointingly low. Prime Minister Modi's PMFBY scheme is meant to improve upon some of the previous failed programs and aims to increase the area under crop insurance to as much as 50 percent of the gross cropped area in the country. Yet despite subsidies in excess of 75 percent, the level of insurance take-up has been slower than anticipated, and has actually declined in recent years. Arguably, one of the reasons why insurance demand has been so sluggish is that policies are rarely designed with the farmers in mind.Our study aims to fill this knowledge gap by assessing farmers' preferences for various crop insurance features, with the objective of optimizing the design of crop insurance to satisfy farmers' needs. To address this, we employed a discrete choice experiment with a sample of farmers from four Indian states (Gujarat, Himachal Pradesh, Karnataka, and Uttar Pradesh). By reparameterizing the random utility model that underlies farmers' decision making in the experiment, we are able to directly estimate farmers' WTP for various insurance product attributes allowing for preference and scale heterogeneity and without imposing unrealistic restrictions on farmers' sensitivity to insurance premiums.Our results suggest that farmers are generally willing to pay considerably more -nearly five times more -for crop insurance than they are asked to pay under PMFBY (a 10 percent premium on the sum insured versus a 2 percent premium on the sum insured). While farmers prefer loss assessments through crop cutting experiments, they also have a very strong preference for assurances that indemnities will be paid in a timely fashion (e.g., within 6 weeks of the loss assessment). Unfortunately, in this regard, farmers generally cannot have their cake and eat it too:the myriad challenges associated with the completion of 1.5 -2 million crop cutting experiments nationwide lead to significant delays in the delivery of payments. Other methods of loss assessment -such as relying on remote sensing technologies or basing payouts on a weather index -can facilitate much more rapid delivery of indemnities, and can reduce insurance companies' marginal costs to basically zero. Yet farmers are at least initially wary of these rather intangible method for loss assessments, and consequently would require a discount on their premiums to be enticed to purchase.When we consider the transition from the status quo variety of crop insurance to an alternative policy that incorporates assessment of losses via remote sensing with guaranteed payment of indemnities in a timely fashion, we find that there are significant welfare gains to farmers, especially if we consider the lower cost of insurance that could result from eliminating the need for crop cutting experiments. While we are not able to directly estimate the magnitude of these welfare effects on a national basis, we can easily qualify or characterize these impacts. First, even when we abstract from cost considerations, the excess value that is associated with the alternative policy vis-à-vis the base policy structure (i.e., total difference in area under the demand curves) is significant. When there is a change in the cost of insurance (which we assume because of the reduction in insurers' costs due to eliminating the need for crop cutting experiments), there are two effects: the increase in cropped area (which presumably translates into a more-or-less proportional increase in the number of insured farmers), and the increased surplus experienced by farmers that would already be insured now paying a lower price. Finally, to the extent that the higher WTP and lower cost of this alternative policy structure reduces the need for government subsidies, there is a reduction in the marginal excess tax burden needed to finance these subsidies.This study points to several avenues for future research. For starters, researchers or policymakers may wish to expand upon the choice experiment design used in the present study to consider other insurance product attributes. The attributes considered in the present study were thought to be the most salient in farmers decision-making, but admittedly any choice experiment design requires simplification and subjectivity. In addition, because there is no financial recourse for decisions made in the course of a choice experiment such as this, there is the potential for hypothetical bias to inflate the estimated WTP relative to what farmers actually would pay if they were to engage in actual insurance markets. This is a common criticism of choice experiments, though some authors have argued that the ability of such stated preference data to engender the estimation and prediction of real market behavior is comparable to those of revealed preference data (e.g., Louviere et al., 2000). Nevertheless, future research could consider other valuation elicitation methods that might be more immune to such potential for bias. The findings from this study and any future research could be very valuable for the design of alternative crop insurance products that could be piloted in an experimental setting. Piloting insurance programs with alternative crop insurance designs would provide more concrete insight into the potential for modified insurance products to increase the number of insured farmers and the total cultivated area insured.  Note: *** Significant at 1 percent level; ** Significant at 5 percent level; * Significant at 10 percent level.Standard errors in parentheses. Note: *** Significant at 1 percent level; ** Significant at 5 percent level; * Significant at 10 percent level. Standard errors in parentheses. Each regression controls for state level fixed effects.Dependent variable in each regression is the conditional mean (marginal) WTP for each of the insurance policy characteristics estimated by the generalized multinomial logit regression (see Table 3), adjusted for inferred attribute non-attendance (see Table 5). ","tokenCount":"9116","images":["-1119488321_1_1.png","-1119488321_1_2.png","-1119488321_40_1.png","-1119488321_41_1.png","-1119488321_42_1.png"],"tables":["-1119488321_1_1.json","-1119488321_2_1.json","-1119488321_3_1.json","-1119488321_4_1.json","-1119488321_5_1.json","-1119488321_6_1.json","-1119488321_7_1.json","-1119488321_8_1.json","-1119488321_9_1.json","-1119488321_10_1.json","-1119488321_11_1.json","-1119488321_12_1.json","-1119488321_13_1.json","-1119488321_14_1.json","-1119488321_15_1.json","-1119488321_16_1.json","-1119488321_17_1.json","-1119488321_18_1.json","-1119488321_19_1.json","-1119488321_20_1.json","-1119488321_21_1.json","-1119488321_22_1.json","-1119488321_23_1.json","-1119488321_24_1.json","-1119488321_25_1.json","-1119488321_26_1.json","-1119488321_27_1.json","-1119488321_28_1.json","-1119488321_29_1.json","-1119488321_30_1.json","-1119488321_31_1.json","-1119488321_32_1.json","-1119488321_33_1.json","-1119488321_34_1.json","-1119488321_35_1.json","-1119488321_36_1.json","-1119488321_37_1.json","-1119488321_38_1.json","-1119488321_39_1.json","-1119488321_40_1.json","-1119488321_41_1.json","-1119488321_42_1.json","-1119488321_43_1.json"]}

data/part_2/0471919304.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"0dda85e6a942f89bb04a5f97ec353dd4","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/9fbe933b-51c6-4892-b56f-f81bf1c480cf/retrieve","description":"Under the Plan for Accelerated and Sustained Development to End Poverty (PASDEP), implemented from 2005/06 to 2009/10, Ethiopia achieved rapid economic growth and laid a foundation for future growth by making substantial investments in infrastructure and human capital. The Growth and Transformation Plan (GTP) for 2010/11-2014/15, Ethiopia's new five year plan, sets even higher growth and investment targets. This paper analyzes these new GTP investment and growth targets using a Computer General Equilibrium (CGE) model of the Ethiopian economy to assess the implications of the plan on sectoral growth and household incomes. The analysis of the GTP investment plan indicates that achieving its high growth targets will require rapid increases in total factor productivity and large-scale mobilization of domestic and foreign savings. The 4.9 percent annual Total Factor Productivity (TFP) growth needed to reproduce the high Gross Domestic Product (GDP) growth under PASDEP (2005/06 to 2009/10) or to continue this growth into the future is very high in comparison to those that have been achieved in other fast-growing economies such as India, China, and Indonesia. Achieving the GTP target GDP growth rates requires even higher TFP growth: by 5.8 and 7.6 percent per year respectively in the medium and high growth GTP scenarios. To some extent, some of this productivity growth could be achieved through reduced underemployment. Nonetheless, these results suggest that the projected GDP growth outcomes are very optimistic.","id":"-12757023"},"keywords":["Research Officer","Ethiopia Strategy Support Program II","International Food Policy Research Institution (IFPRI) Postdoctoral Fellow","Development Strategy and Governance Division","International Food Policy Research Institution (IFPRI)"],"sieverID":"50612bab-6534-40eb-a650-e7439ede49e2","pagecount":"33","content":"The Ethiopia Strategy Support Program II (ESSP II) Working Papers contain preliminary material and research results from IFPRI and/or its partners in Ethiopia. The papers are not subject to a formal peer review. They are circulated in order to stimulate discussion and critical comment. The opinions are those of the authors and do not necessarily reflect those of their home institutions or supporting organizations.v   Appendix A.1. Specification of the CGE and micro-simulation model adapted from Dorosh and Thurlow (2009)  Under the Plan for Accelerated and Sustained Development to End Poverty (PASDEP), implemented from 2005/06 to 2009/10, Ethiopia achieved rapid economic growth and laid a foundation for future growth by making substantial investments in infrastructure and human capital. The Growth and Transformation Plan (GTP) for 2010/11-2014/15, Ethiopia's new five year plan, sets even higher growth and investment targets. This paper analyzes these new GTP investment and growth targets using a Computer General Equilibrium (CGE) model of the Ethiopian economy to assess the implications of the plan on sectoral growth and household incomes.The analysis of the GTP investment plan indicates that achieving its high growth targets will require rapid increases in total factor productivity and large-scale mobilization of domestic and foreign savings. The 4.9 percent annual Total Factor Productivity (TFP) growth needed to reproduce the high Gross Domestic Product (GDP) growth under PASDEP (2005/06 to 2009/10) or to continue this growth into the future is very high in comparison to those that have been achieved in other fast-growing economies such as India, China, and Indonesia. Achieving the GTP target GDP growth rates requires even higher TFP growth: by 5.8 and 7.6 percent per year respectively in the medium and high growth GTP scenarios. To some extent, some of this productivity growth could be achieved through reduced underemployment. Nonetheless, these results suggest that the projected GDP growth outcomes are very optimistic.Meeting the financing requirements for the GTP will also be challenging given the large amount that needs to be mobilized, as compared to the low historical savings in Ethiopia. In the model simulations where the additional finance is mobilized from domestic sources, household savings rates increase from an average of 5.5 percent of total household income in 2014/15 in the base case to 12.5 percent of total household income in 2014/15 in the increased household savings / medium case scenario. Achieving this high rate of domestic savings would also require changes in macroeconomic policy, such as increases in tax rates or increases in interest rates to encourage private savings. If the additional finance is mobilized through increased foreign borrowing or transfers, foreign savings would need to increase from 47.7 billion birr in 2014/15 of the base case to 96.8 billion birr in 2014/15 of the medium case scenario in real terms (i.e. from 7.5 to 14.8 percent of GDP).Regardless of the financing strategy, the high TFP and GDP growth rates under the GTP imply high average income growth for both poor and rich households, in both rural and urban areas. Because the GTP involves a greater concentration of investment in non-agricultural sectors than did PASDEP, growth of incomes of urban households is higher in the GTP than under PASDEP. Conversely, income growth of the rural poor is slightly lower under the medium growth scenario with domestic savings (10.0 percent) than under a continuation of PASDEP growth and investment (10.6 percent).Thus, this analysis shows that if the GTP investment and sectoral growth targets are achieved, real incomes of the poor in Ethiopia would rise substantially. The base simulations indicate that real incomes of the poor rose under PASDEP from 2005/06 to 2010/11. Under GTP, this real income growth would be accelerated, provided there is sufficient foreign savings or mobilization of domestic savings to achieve the targets. Nonetheless, the simulations also suggest that agricultural growth will still be crucial for raising incomes of Ethiopia's rural poor. Thus, investments that raise agricultural productivity will need to continue in order to ensure that the rural poor share in the substantial projected benefits that would result from achieving the high economic growth targets of the GTP.Countries devise and implement medium term development plans to articulate a development strategy that will provide the guiding framework for implementation of policies. Accordingly, the Ethiopian government has been designing and implementing strategies and plans to manage the overall development of the country and achieve its key objective of broad-based, accelerated, and sustained economic growth so as to eradicate poverty. During the last development plan, the Plan for Accelerated and Sustained Development to End Poverty (PASDEP): 2005/06-2009/10, the country achieved remarkable economic and social development.Based on its experience with implementation of PASDEP, the government has formulated the next five year plan, the Growth and Transformation Plan (GTP), for 2010/11-2014/15. The GTP is envisaged to sustain rapid and broad based economic growth and contribute to Ethiopia's goal of becoming a middle-income economy by 2020-2023. Economic growth and other development targets set in the GTP are even higher than those in PASDEP and earlier plans. Achieving these targets, however, will require very large productivity growth and huge amounts of financial resources anticipated to be sourced from increased domestic saving mobilization, foreign direct investment, and foreign borrowing.Despite the huge resource requirements of the GTP, the government's Medium Term macroeconomic fiscal Framework (MTF) envisages modest spending and revenue growth during the same period: 2010/11-2014/15. Much of the development spending is planned to be \"off-budget\" (i.e. not included in the general government spending) through domestic and foreign resource mobilization of public development enterprises. The participation of the private sector in savings mobilization and investment is acknowledged to be important in the GTP, but the private sector's role in financing is not explicitly stated or quantified.The basic objective of this study is to assess the implications of alternative financing options of the GTP. Using a computable general equilibrium (CGE) model of the Ethiopian economy, this study provides quantitative estimates of the impacts of planned GTP investments on various economic indicators, including overall GDP, real exchange rates, sectoral output, household income levels, and household expenditures. In particular, the study focuses on two major financing alternatives: increased domestic (household) savings and increased foreign savings (through foreign borrowing, foreign direct investment, or foreign grants).Section 2 presents a brief overview of the implementation of the previous five year plan (PASDEP), examining both macro-economic indicators as well as investments in infrastructure designed to build a foundation for further economic growth. It also briefly introduces the GTP and the broad economic targets. Section 3 describes the CGE model and the data base used for the analysis (with details of the model equations included in an Appendix). Simulation results are discussed in Section 4. Section 5 contains the summary and conclusions.Since the first Structural Adjustment Program (SAP) of Ethiopia (1993Ethiopia ( -1996)), the Ethiopian Government has implemented various development plans designed to promote broad-based and equitable economic growth and to eradicate poverty. The development plan for 2005/06-2009/10, the Plan for Accelerated and Sustained Development to End Poverty (PASDEP), was aimed at achieving all the Millennium Development Goal (MDG) targets and the government's vision for Ethiopia's development.In PASDEP, two alternative economic growth scenarios were considered. In the base case scenario, an average economic growth rate of 7 percent per annum was considered necessary to achieve the MDGs. For the high case scenario, which aimed beyond achieving MDG's targets, a 10 percent annual average economic growth target was set.During the time of PASDEP's implementation, substantial economic growth and significant progress on social and human development were achieved. Annual average GDP growth is estimated at 11 percent (MOFED 2010) which exceeded both the base case and high growth scenarios set in PASDEP (Table 2.1). Specifically, average annual growth rates in the agricultural and services sectors (8.4 percent and 14.6 percent, respectively) exceeded plan targets. However, growth in the industrial sector, though substantial (10.0 percent per year), fell short of both the base case (11.0 percent growth) and high case (18.0 percent growth) targets. The structure of aggregate demand changed very little during the PASDEP period: the share of consumption expenditure at the end of the PASDEP was the same as the level in 2004/05 (94 percent, Table 2.2). The share of gross capital formation (investment) in GDP declined slightly, from 23.8 to 22.3 percent. Most of this investment was financed from foreign sources (reflected in the resource balance, the difference between exports and imports of goods and services), which equaled 19.3 percent of GDP in 2009/10. The domestic savings rate also remained unchanged in the plan period in spite of a plan to increase the domestic saving 1 rate to 13.1 percent. At the end of the PASDEP period, the domestic savings rate was only 5.5 percent.1Unlike the UN System of National Accounts, which defines savings as income net of consumption (income -consumption), domestic savings in PASDEP and GTP is defined as domestic product net of consumption (GDP -consumption). As a net recipient of transfers from the rest of the world, Ethiopia's savings rate is therefore underestimated in PASDEP and GTP by the amount equivalent to the net income and current transfers residents receive from the rest of the world. The public and private investments during the five years of PASDEP resulted in substantial increases in physical infrastructure and human capital that have laid a foundation for further growth and poverty reduction. Road and telecommunications infrastructure have increased dramatically, greatly increasing the proportion of Ethiopia's population with access to urban markets and services.In 1984, only about 6 million people in Ethiopia (12 percent of the population) were within 3 hours travel time to an urban center (agglomeration) with a population of at least 50 thousand (Figure 2.1). By 1997 this figure had doubled to about 12 million people. In 2007, this figure had reached 38 million people (over half of Ethiopia's population) because of major investments in roads and to a lesser extent, increased rural-urban migration. These investments also reduced remoteness: the number of people who lived more than 10 hours from an urban center of more than 50,000 people fell from 40 percent in 1984 to only 12 percent in 2007 (Figure 2.2). Likewise, the percentage of people who lived more than 5 hours from a major urban center fell from 66 percent in 1994 to 38 percent in 2007, a level similar to that in Mozambique, though still about double the rate in densely populated (and more urbanized) Nigeria.    Finally, investments in education have resulted in major increases in primary school enrollment, with potential long term effects on household welfare and labor productivity. As shown in Table 2.4, in 1991 only 21.9 percent of Ethiopian primary school-aged children were enrolled in primary school (and only 18.8 percent of girls). By 2000, net primary school enrollment had almost doubled to 38.4 percent (32.5 percent for girls), and by 2007, these figures reached 71.4 percent for all children and 68.5 percent for girls. While net primary school enrollment in Ethiopia still lags behind that of most other eastern and southern African countries (an important exception is Sudan), the gap between Ethiopia and these countries has narrowed significantly. The government has formulated the five year Growth and Transformation Plan (GTP) (2010/11-2014/15) to carry forward the important strategic directions pursued in the PASDEP. In addition to achieving the MDGs in the social sector and establishing suitable conditions for sustainable nation building, the GTP has a major objective of maintaining at least an average real GDP growth rate of 11 percent. By sustaining economic growth throughout the five year plan period, the government aims to achieve the MDG targets by 2015 and its longer term vision of being a middle income country by 2020-2023.GTP is unique in comparison to past development plans of the country due to its high economic growth and other development targets. Following MOFED, two growth scenarios are considered in the GTP: medium growth and high growth. Under the medium growth case scenario, Ethiopia's economy is projected to grow at a rate attained during PASDEP (11.2 percent), as shown in Table 2.5. All MDG targets will be met under this scenario. Under the high growth scenario, an annual average GDP growth rate of 14.9 percent is targeted (see Table 2.6). The basic assumption for the high growth scenario is the doubling of agricultural value added, by scaling up the productivity of farmers to the productivity levels of existing best or model farmers (MOFED 2010). The GDP growth in the high growth scenario is thus significantly different from the growth in the medium growth scenario due to the large growth difference in the agriculture sector. The growth targets in the industry and service sectors in the high growth scenario are only slightly higher than in the medium growth case. According to the GTP, all the country's development policies and strategies are geared towards the main development agenda of poverty eradication. The GDP growth in the plan is expected to help create employment and raise income, contributing to poverty eradication. Poverty oriented government expenditure as a percentage of GDP are also set to grow from 12.3 percent in 2009/10 to 17.3 percent at the end of the plan period. Accordingly, the GTP aims to achieve a reduction of poverty from the estimated level of 29.2 percent in 2009/10 to 22.2 percent in 2014/052 .While sectoral and overall GDP growth outcomes as well as expenditure components of GDP in the GTP are projected in two scenarios, the rest of the social and infrastructural development outcomes and the government's fiscal operations are projected only in one (base-case) scenario. For achieving the objectives stated in the GTP, allocation of finances has been planned based on the Medium Term macroeconomic fiscal Framework (MTF).According to the MTF, domestic revenue mobilization is to be boosted with the full implementation of the on-going tax reform, improvement in efficiency of tax administration, and broadening the tax base in line with the economic growth. As a result, tax collection as a percent of GDP is expected to rise from 11.3 percent in 2009/10 to 15 percent in 2014/15, as shown in the following table. Government expenditure during the GTP is planned to prioritize the financing of ongoing projects and investments in pro-poor3 sectors, is expected to be based on the growth prospects of the country in the coming years, and is aimed at maintaining the deficit at a sustainable level. As a result, the average government deficit in the GTP is planned to be only 2.3 percent of GDP, of which one third is to be covered by external borrowing (Table 2.8). CGE modeling is clearly an appropriate methodology to approach this problem, as it allows capturing the most relevant interactions in the economic system and, in particular, reflecting the effects of the different savings mechanisms under study on different parts of the economic system. Also, by representing the workings of the economy in a structural way, CGE modeling allows for an explicit reflection on the main transmission channels at work in the repercussion of the effects of these saving mechanisms.The model used in this study is a recursive dynamic extension of the standard CGE model of the International Food Policy Research Institute (IFPRI), documented in Diao et al. (2011). This kind of dynamic model is based on the assumption that the behavior of economic agents (private and public) is characterized by adaptive expectations: economic agents make their decisions on the basis of past experiences and current conditions, with no role for forward-looking expectations about the future (Lofgren, Harris, and Robinson 2001). This is an alternative that captures developing countries' reality better than inter-temporal dynamic models that can be explained by economic agents who have forward looking (rational) expectations and make inter-temporally optimal decisions, in which everybody knows everything about the future, and they use that information in making decisions.The model is solved recursively-one period at a time. Since a recursive model is a series of static CGE models that are linked between periods, the equations in this model are separated into a within-period module, which presents the decisions in each time period, and a between-period module, which governs the dynamics of the model.This part defines a single period static CGE model. The model assumes that producers maximize profits subject to production functions. In this model a multi-stage production technology is adopted. At the top level, value added and intermediate inputs are combined via a Constant Elasticity of Substitution (CES) production technology to produce gross output. Factors of production are substitutable, via the use of constant elasticity of substitution (CES) functions. On the other hand, the intermediate inputs are aggregated in a fixed proportion-Leontief specification.The domestic output of each commodity is either domestically used or exported. Profit maximization drives producers to sell their products in domestic or foreign markets based on the potential returns. It is assumed that commodities sold domestically can only imperfectly be transformed into exportable commodities, via the use of constant elasticity of transformation (CET) functions. In an analogous way, the model incorporates imperfect substitutability between domestically produced and imported goods (i.e. Armington assumption).Domestic demand is the total sum of final consumption demands by households and government, and investment demand, intermediate consumption demands by activities and transaction services demand. Households' consumption demand is given by a linear expenditure system (LES), derived from maximization of a Stone Geary utility function. The model assumes households maximize utility subject to budget constraints.The common assumption here is that the economy observed is in general equilibrium. Equilibrium in the goods market is attained through the endogenous interaction of relative prices. In order to clear the factor market, skilled labor and capital are set to be fully employed and activity-specific, implying that sector-specific returns to these factors adjust to guarantee market clearing. Semi-skilled and un-skilled workers have a fixed wage, and their employment levels are determined by the producers' demand for them. Factor land is assumed to have a flexible rent with a certain distortion by sector, which makes it fully employed and mobile across sectors.The model includes three macroeconomic balances: the current account balance, the government balance, and the savings and investment balance. The current account balance is held constant by assuming flexible exchange rate at a fixed level of foreign savings (fixed in foreign currency). There is an implicit functional relationship between the real exchange rate and the trade balance.In the government account, the level of government expenditure, equal to consumption and transfers, is fixed in real terms while government revenue is determined by fixed direct and indirect tax rates. Government savings is determined residually as the gap between revenue and expenditure. This closure is chosen since it is assumed that changes in direct and indirect tax rates, as well as in government expenditure, are exogenously determined based on the economic policy.The macro closure applied for the saving-investment balance is the savings-driven investment closure in which the value of investment is determined by the value of savings, making the simulations possible. Fixed savings rates for all non-governmental institutions and flexible capital formation are specified so that all savings are channeled into investment.It should be noted that the model designed in this paper can only solve for the relative prices and the real variables of the economy. To achieve this and anchor the absolute price level, a normalization rule has been applied. The consumer price index (CPI) is chosen as the numéraire, so all changes in nominal prices and incomes in simulations are relative to the weighted unit price of households' initial consumption bundle (i.e a fixed CPI). The model is also homogenous of degree zero in prices. In macro terminology, the model displays neutrality of money.In every period the capital stock is updated with the total amount of new investment and depreciation. New capital is distributed among sectors based on each sector's initial share of aggregate capital income. Total labor supply is updated by the population growth rate, i.e. as population grows, the total labor supply increases at the same rate. The GTP has set an ambitious investment plan and proposed ways of financing it, though details of the amount of financing from residents and non-residents are not specified. In this study, two options are considered as a means to ensure sufficient finances for investment targets: increased domestic households saving rates and increased foreign savings (through foreign loans, transfers from non-residents, or foreign direct investment).The analysis consists of five simulations. In the base scenario (BASE), we assume a continuation of the historical growth trends of 2005/06-2009/10 for an additional five years, i.e. through 2014/15. The other four simulations model the medium and high growth scenarios of the GTP under the two different financing options: Medium Growth with Accelerated Domestic Savings (MAD), High Growth with Accelerated Domestic Savings (HAD), Medium Growth with Accelerated Foreign Savings (MAF) and High Growth with Accelerated Foreign Savings (HAF). In all the GTP related simulations, other parameters, apart from the ones used in the specific simulations, are growing at either a fixed trend rate or as indicated in the GTP. Thus, the GTP simulations differ from the BASE one in terms of the assumptions in growth rates of total factor productivity (TFP), the changes in production that cannot be explained in terms of changes in factor supplies4 , government expenditure, tax rates, and the simulation-specific household savings rates or foreign savings growth rate. For all simulations, annual labor supply is assumed to grow at an average rate of 2.3 percent per year while agricultural land area is set to grow at 2.8 percent per annum based on the recent trend in agricultural land expansion. The tax rate as a percent of GDP is assumed to grow at 3.0 percent in the BASE, and at 7.5 percent in the GTP scenarios. In each simulation, total factor productivity of each activity is adjusted to approximate the GTP overall targets for sectoral (agriculture, industry, and services) growth rates.5 If foreign savings continue growing at the historical trend rate, in order to finance the investment requirement of the GTP, the average household savings rate needs to grow by an average 21.8 percent and 31.1 percent in the medium and high case scenarios, respectively. This means households' marginal propensity to save need to reach 12.5 percent in 2014/15 in the medium case scenario from an average of 5.5 percent in 2014/15 in the base case, to generate sufficient household savings. This is an ambitious target since savings rates in excess of 10 percent are rare. Over the 1990-2008 period, Belgium and Russia achieved average savings rates of 12.5 percent and Italy's savings rate was 13.6 percent. More common are savings rates in the range of 4 to 6 percent, as in the Czech Republic, Slovak Republic, Korea, and the United States (OECD Factbook 2010). Currently, the real interest rate on domestic deposits is negative. To achieve the desired increase in savings, it would require a macroeconomic policy with proper incentives to encourage savings, such as allowing positive real interest rates on bank deposits. Alternatively, if there is no adjustment in households' propensity to save relative to past trends, foreign savings need to increase by 22 percent and 31 percent to meet the investment requirements in the medium and high scenarios of the GTP, respectively (Table 4.1). Over the five years of PASDEP, growth rates in value added (overall GDP) were very high, in spite of relatively modest increases in factor supplies. This implies a very high TFP growth rate-equal to 4.9 percent per year in the model simulation of 2005/06-2009/10 used to update the SAM. This TFP growth rate is in fact very high by historical standards, for example, exceeding the annual TFP growth rates of China (4.1 percent), India (2.8 percent) and Indonesia (1.6 percent) over the 1990 to 2008 period (Table 4.2). 6 To achieve the GDP growth rates under medium and high growth GTP scenarios requires even higher average TFP growth rates of 5.8 percent and 7.6 percent respectively (Table 4.1). These very high TFP growth rates suggest that it may be very difficult to achieve the plan's growth targets even if levels of financing and investment are forthcoming. However, to some extent, calculated TFP growth rates for Ethiopia may not be strictly comparable to those of developed countries due to methodological differences in the calculation. In essence, TFP growth is calculated as the difference between growth in value added minus the growth due to changes in labor, capital, and land inputs. In developed countries, growth in labor inputs is calculated using labor input measured in hours or days, and as hours or days worked per worker rise during periods of rapid growth, this increase in labor input is automatically taken into account in TFP calculations. However, for Ethiopia and most developing countries, economy-wide estimates of labor input are only available in person-years with no adjustment for changes in the amount of time a worker works per year.To the extent that Ethiopia's recent rapid economic growth is due to lower underemployment, actual hours and days worked would rise faster than the number of workers, and as a result, true TFP growth will be overstated.By construction of the simulations (setting total factor productivity growth to meet the medium and high growth scenario GDP growth targets), there is essentially no difference in real GDP growth according to the source of financing. Increasing household savings rates, however, results in significantly lower consumption growth-for example, 9.4 percent in the medium growth accelerated domestic savings scenario (MAD) as compared to 11.5 percent in the medium growth accelerated foreign savings (MAF) scenario (Table 4.3). Indeed, private consumption in the former is effectively no different from the baseline case. The other major macro-economic difference between simulations of increased foreign savings versus increased domestic savings rates, is that increased foreign savings leads to a real exchange rate appreciation as the increased financial resources make increased spending on nontradable goods and services (which results in a rise in their price) possible, while the increased spending on tradable goods and services leads to a reduction in exports and an increase in imports. In the medium growth accelerated foreign savings scenario (MAF), the real exchange rate appreciates by 3.5 percent, 2.1 percentage points more than in the medium growth accelerated domestic savings scenario (MAD). Likewise, export growth is considerably slowed: 16.9 percent with accelerated foreign savings (MAF), compared with 22.5 percent with accelerated domestic savings (MAD). Government savings are higher in the high growth scenario than in the medium growth scenario under both financing options. As expected, given a similar growth in government spending in both GTP scenarios, with slower productivity growth, income growth and government revenue growth are less and thus government savings fall over time; with faster productivity growth, however, tax revenues rise and government savings increase. Despite higher government savings in the high growth scenarios, the large investment demand still requires additional financing from household or foreign savings.Returns for all the factors increase significantly in all the scenarios, with generally higher returns when the investment is financed through foreign savings than through increased household savings (Table 4.4), since in the former scenarios the rise in the output price of non-tradable goods generally increases the incomes of labor, land, and capital in these sectors. In the base case, which broadly replicates the economic growth path achieved under PASDEP with relatively high agricultural growth, the average annual growth rates of income of agricultural labor (10.4 percent), land (12.3 percent), and livestock (14.3 percent) rise faster than incomes of skilled labor (9.2 percent) and capital (9.3 percent). Under the GTP scenarios involving greater investment and TFP growth in non-agricultural sectors, however, incomes of skilled labor and capital rise faster than agricultural or rural sectors (except livestock).  4.5). Overall, higher household income growth is achieved with increased foreign saving inflows (as compared to increased domestic savings) because these inflows increase total resources in the economy. It should be noted that household consumption also rises more with increased foreign savings than domestic savings. However, if the foreign savings are in the form of loans that are to be repaid, then the consumption growth will be temporary. Therefore, household income is chosen as the preferred indicator for the more permanent gains from increased foreign savings. Note that if the economy continues with its performance as in the PASDEP period, income of rural households would increase faster than that of urban households (Figure 4.1). In particular, in this scenario (BASE), income growth is lowest for both the urban poor and nonpoor and highest for the rural poor. Under the GTP, however, incomes of the non-poor increase faster than incomes of the poor under both financing options. This result is largely because of the high returns to capital and skilled labor under the GTP that accrue mainly to non-poor households. Comparing across financing options, rural domestic incomes rise less in the increased household savings scenario since household spending on agricultural products is reduced (relative to the increased foreign savings scenarios). On the other hand, urban household incomes rise less in the increased foreign savings scenario than in the increased household savings scenario because the real exchange rate appreciation in the former scenario makes domestic production of import substitute industrial products less profitable, thereby reducing earnings of skilled labor and capital.Household total expenditure increases significantly in all household groups, in all simulations and financing alternatives (Table 4.6). Total expenditures of the urban non-poor grows the fastest in the medium growth scenario of the GTP, as a result of faster growth in the industry sector compared to the other sectors. In the high growth scenario of the GTP on the other hand, expenditure of the rural non-poor grows more or less the fastest due to faster growth in agriculture. The rate of increase in household consumption significantly differs between the financing options. As expected, the rate of increase in total household expenditure under the domestic saving option is lower in all household groups than under the foreign financing alternative, as households are supposed to save more, at the expense of consumption. The poor are spending even less than they would be able to with the continuation of the economic performance during the PASDEP. Financing the investment through increased foreign savings allows each of the domestic household groups to increase their consumption level at a higher rate. This may seem to suggest that increasing foreign savings is a better alternative but, in reality, it simply reflects that the analysis that we are conducting ignores the accumulation of assets by the actors in the model: if foreign savings consist of foreign direct investment or foreign borrowing, they will tend to reduce the share of output that is available to Ethiopian residents.The analysis shows that regardless of the financing strategy, there will be high income growth for poor and rich alike under the GTP. Urban households will enjoy larger income increases than rural households. Further, unskilled labor and agricultural labor will not fare any better under the medium growth scenarios than under the baseline without the GTP (except for a slight improvement for unskilled labor under the MAF scenario). Consequently, the rural poor will have lower average income growth rates under MAD (10.0 percent) than under the baseline (10.6 percent).The Growth and Transformation Plan has set higher growth and investment targets than those of any of Ethiopia's earlier national plans and its implementation requires faster growth in total factor productivity and mobilization of huge savings both from domestic and foreign sources. The 4.9 percent annual TFP growth needed to reproduce the high GDP growth under PASDEP (2005/06 to 2009/10) or to continue this growth into the future is very high as compared to those that other fast-growing economies such as India, China or Indonesia have been able to achieve. Achieving the GTP target GDP growth rates requires even higher TFP growth: by 5.8 and 7.6 percent per year in the medium and high growth GTP scenarios, respectively. To some extent, some of this productivity growth could be achieved through reduced underemployment. Nonetheless, it is not clear how these high TFP growth rates can be achieved, suggesting that the projected GDP growth outcomes are very optimistic.Meeting the financing requirements for the GTP will also be challenging given the large amount that needs to be mobilized, as compared to the low historical savings in Ethiopia. In the model simulations where the additional finance is mobilized from domestic sources, households' savings rates increase from an average of 5.5 percent of total household income in 2014/15 in the base case to 12.5 percent of total household income in 2014/15 in the increased household savings/medium case scenario. Achieving this high rate of domestic savings would also require changes in macroeconomic policy, such as increases in tax rates or increases in interest rates to encourage private savings. If the additional finance is mobilized through increased foreign borrowing/transfers, foreign savings would need to increase from 47.7 billion birr in 2014/15 of the base case to 96.8 billion birr in 2014/15 of the medium case in real terms (i.e. from 7.5 percent to 14.8 percent of GDP).Results from the simulations on financing options also show that private consumption is lower when the GTP is financed through increases in household savings as compared to financing through increased foreign savings. However, although considerable increment in exports and imports were observed in both of the GTP related simulations, growth in exports is lower under the increased foreign savings scenarios due to an appreciation of the real exchange rate relative to the increased household savings scenarios.Regardless of the financing strategy, the high total factor productivity and GDP growth rates under the GTP imply high average income growth for both poor and rich households, in both rural and urban areas. Because the GTP involves a greater concentration of investment in non-agricultural sectors than did PASDEP, growth of incomes of urban households is higher in the GTP than under PASDEP. Conversely, income growth of the rural poor is slightly lower under the medium growth scenario with domestic savings (10.0 percent) than under a continuation of PASDEP growth and investment (10.6 percent).It should be noted that the results presented here do not take into consideration possible additional migration of labor from rural to urban areas arising from faster growing urban incomes, or possible positive agglomeration effects on total factor productivity from growth in urban centers. Including these factors would raise the incomes of urban households relative to rural households and raise total GDP, but would not reverse the general result that investments in agriculture reduce poverty faster than investments in non-agriculture (Dorosh and Thurlow 2011).Thus, this analysis shows that if the GTP investment and sectoral growth targets are achieved, real incomes of the poor in Ethiopia would rise substantially. The base simulations indicate that real incomes of the poor rose under PASDEP from 2005/06 to 2010/11 and that these incomes would also rise if PASDEP growth continued. Under GTP, this real income growth would be accelerated, provided there is sufficient foreign savings or mobilization of domestic savings to achieve the targets. Nonetheless, the simulations also suggest that agricultural growth will still be crucial for raising incomes of Ethiopia's rural poor. Thus, investments that raise agricultural productivity will need to continue in order to ensure that the rural poor share in the substantial projected benefits that would result from achieving the high economic growth targets of the GTP.households in the household survey. Households in the model receive income through the employment of their factors in both agricultural and nonagricultural production, and then pay taxes, save, and make transfers to other households. The disposable income of a representative household is allocated to commodity consumption derived from a Stone-Geary utility function (i.e. a linear expenditure system of demand).The model makes a number of assumptions about how the economy maintains macroeconomic balance. These closure rules concern the foreign or current account, the government or public sector account, and the savings-investment account. For the current account, a flexible exchange rate maintains a fixed level of foreign savings. This assumption implies that the country cannot simply increase foreign debt but has to generate export earnings in order to pay for imported goods and services. While this assumption realistically limits the degree of import competition in the domestic market, it also underlines the importance of the agricultural and industrial export sectors. For the government account, tax rates and real consumption expenditure are exogenously determined, leaving the fiscal deficit to adjust to ensure that public expenditures equal receipts. For the savings-investment account, real investment adjusts to changes in savings (i.e. savings-driven investment). These two assumptions allow the models to capture the effects of growth on the level of public investment and the crowding-out effect from changes in government revenues.Finally, the CGE model is recursive dynamic, which means that some exogenous stock variables in the models are updated each period based on inter-temporal behavior and results from previous periods. The model is run over the period 2005-2015, with each equilibrium period representing a single year. The model also exogenously captures demographic and technological change, including population, labor supply, human capital, and factor-specific productivity. Capital accumulation occurs through endogenous linkages with previous-period investment. Although the allocation of newly invested capital is influenced by each sector's initial share of gross operating surplus, the final allocation depends on depreciation and sector profit-rate differentials. Sectors with above-average returns in the previous period receive a larger share of the new capital stock in the current period.In summary, the CGE model incorporates distributional change by (1) disaggregating growth across sub-national regions and sectors;(2) capturing income-effects through factor markets and price-effects through commodity markets; and (3) translating these two effects onto each household in the survey according to its unique factor endowment and income and expenditure patterns. The structure of the growth-poverty relationship is therefore defined explicitly ex ante based on observed country-specific structures and behavior. This allows the model to capture the poverty and distributional changes associated with agricultural growth.       ","tokenCount":"6310","images":["-12757023_1_1.png","-12757023_1_2.png","-12757023_1_3.png","-12757023_10_1.png","-12757023_10_2.png","-12757023_11_1.png","-12757023_11_2.png","-12757023_21_1.png","-12757023_33_1.png","-12757023_33_2.png"],"tables":["-12757023_1_1.json","-12757023_2_1.json","-12757023_3_1.json","-12757023_4_1.json","-12757023_5_1.json","-12757023_6_1.json","-12757023_7_1.json","-12757023_8_1.json","-12757023_9_1.json","-12757023_10_1.json","-12757023_11_1.json","-12757023_12_1.json","-12757023_13_1.json","-12757023_14_1.json","-12757023_15_1.json","-12757023_16_1.json","-12757023_17_1.json","-12757023_18_1.json","-12757023_19_1.json","-12757023_20_1.json","-12757023_21_1.json","-12757023_22_1.json","-12757023_23_1.json","-12757023_24_1.json","-12757023_25_1.json","-12757023_26_1.json","-12757023_27_1.json","-12757023_28_1.json","-12757023_29_1.json","-12757023_30_1.json","-12757023_31_1.json","-12757023_32_1.json","-12757023_33_1.json"]}

data/part_2/0491860919.json

@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"33cfa2f2651899af36af8d2fe6d29f2e","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/40ab1ff0-6cdb-4c33-9ba5-47af6453247a/retrieve","description":"The patterns of Africa’s participation in fruit and vegetable value chains (FVVCs) clearly reflect the continent’s colonial past. The restructuring of African exports around a few commodities to serve European markets during the colonial period largely undermined the farming of local food crops, including indigenous fruits and vegetables. Postcolonial governments focused on cash crops as the main source of foreign exchange earnings, reinforcing the status quo. However, the mid-1980s witnessed a major shift in global demand away from traditional cash crops and toward high-value products, including fruits and vegetables. This shift was an opportunity for developing countries, including those in Africa, to diversify their exports and reduce their vulnerability to global commodity price fluctuations. Participation in FVVCs can also have positive impacts on employment creation, income mobility, and poverty reduction. Yet, Africa’s participation in FVVCs is undermined by a number of structural challenges, some of which are typical of FVVCs, and some related to long-standing issues facing African economies in general, and the agriculture sector in particular.","id":"1564303628"},"keywords":[],"sieverID":"ddb7e291-265b-47e3-bb71-68a374e8a59f","pagecount":"58","content":"1 There is no standard definition of the three levels of processing. Throughout this chapter, we define \"unprocessed\" fruits and vegetables as raw commodities, \"processed\" fruits and vegetables as products that are ready to consume, and \"semi-processed\" goods as goods that are neither raw nor ready to consume. 2 This methodology was used in previous editions of the AATM. In principle, the underlying logic is to choose a 5-year period of the most recent data and compare it with a 5-year period with a 10-year interval. This can help us track whether trade data reflect a longstanding pattern or whether there may have been disruptive changes during and after 2020. Vegetables, leguminous; chickpeas (garbanzos), dried Vegetables, leguminous; small red (adzuki) beans, shelled, dried Vegetables, leguminous; n.e.s., dried, shelled. Fruit, edible; fruit and nuts n.e.s. in heading no. 0812, provisionally preserved but unsuitable in that state for immediate consumption No RCA 9 14 Vegetables, frozen, n.e.s. in Chapter 7 Vegetables, mixtures of vegetables n.e.s. in heading no. 0712, whole, cut, sliced, broken, or in powder but not further prepared, dried Vegetables, leguminous; peas, dried Vegetables, leguminous; beans, dried, shelled Vegetables, leguminous; kidney beans, dried, shelled Vegetables, leguminous; lentils, shelled, dried Fruit, edible; strawberries, frozen Fruit, edible; raspberries and other berries, whether containing added sugar or other sweetening matter Fruit, edible; fruit and nuts n.e.s. in heading no. 0811, frozen Vegetables, potatoes, frozen Vegetables, leguminous; peas, frozen Vegetables, leguminous; beans, frozen Vegetables, leguminous (other than peas or beans), frozen Vegetables; spinach, frozen Vegetables; sweet corn, frozen Vegetable mixtures, frozen Vegetables, olives, provisionally preserved but unsuitable in that state for immediate consumption Vegetables, cucumbers, and gherkins, provisionally preserved but unsuitable in that state for immediate consumption Vegetables and mixed vegetables; n.e.s. in heading no. 0711, provisionally preserved but unsuitable in that state for immediate consumption Vegetables, onions, not further prepared, dried Vegetables, leguminous; broad beans, dried, shelled Vegetables, leguminous; n.e.s. in heading no. 0713, shelled, dried Fruit, edible; cherries, provisionally preserved but unsuitable in that state for immediate consumptionThe patterns of Africa's participation in fruit and vegetable value chains (FVVCs) clearly reflect the continent's colonial past. The restructuring of African exports around a few commodities to serve European markets during the colonial period largely undermined the farming of local food crops, including indigenous fruits and vegetables. Postcolonial governments focused on cash crops as the main source of foreign exchange earnings, reinforcing the status quo. However, the mid-1980s witnessed a major shift in global demand away from traditional cash crops and toward high-value products, including fruits and vegetables. This shift was an opportunity for developing countries, including those in Africa, to diversify their exports and reduce their vulnerability to global commodity price fluctuations. Participation in FVVCs can also have positive impacts on employment creation, income mobility, and poverty reduction. Yet, Africa's participation in FVVCs is undermined by a number of structural challenges, some of which are typical of FVVCs, and some related to long-standing issues facing African economies in general, and the agriculture sector in particular.From a theoretical perspective, it is important to understand the determinants of FVVCs in Africa before analyzing trade data. Three main theoretical frameworks can be evoked: the factor content theory, gravity models, and global value chain determinants. First, the factor content theory argues that countries export products that use their relatively abundant factors of production. Thus, if fruits and vegetables are intensive in land and water, they will be exported by African countries that are abundant in these factors. Second, the gravity model predicts that bilateral trade flows are based on the economic sizes and distance between two countries, which is reflected in trade costs. Trade costs include transport and storage infrastructure, such as cold storage facilities, as well as trade policies and trade barriers, and are affected by common borders, historical colonial links, and common languages. Trade policies that increase trade costs include tariffs and nontariff measures. Thus, gravity considerations play an important role explaining African trade patterns, given that African countries generally trade with countries characterized by large markets (the United States and China) or with countries with which they had colonial links (France, Portugal, Italy, and the United Kingdom). Third, the literature examines the determinants of upgrading-that is, participating in the downstream nodes-in a global value chain. These are mainly the skills of the labor force, trade policy at the origin and the destination, and technology transfer (Gereffi 2019). Clearly, in Africa, the lack of research and development (and thus innovation) in the agriculture sector and the presence of high tariffs on processed agrifood products help to explain the specialization of most countries in unprocessed products in the early stages of the value chain.Against this background, this chapter analyzes Africa's participation in FVVCs and discusses challenges and opportunities in this sector, including new prospects with the advent of the African Continental Free Trade Area (AfCFTA). The chapter is structured as follows. We begin with a brief overview of the importance of the fruit and vegetable sector for Africa and summarize the main findings on the benefits from participation in FVVCs, as well as the risks faced by African countries in this sector, with special attention to smallholders. The core of the chapter analyzes FVVCs at the global and African levels, including trends in exports and imports of fruits and vegetables, top exporters, and top export destinations. Throughout the analysis, we examine trade in fruits and vegetables at three levels of processing: unprocessed, semi-processed, and processed products. 1 We also compare two time periods 2 with a 10-year Chapter 4 -Fruit and Vegetable Value Chains in Africa interval: the first from 2008 to 2012, and the second from 2018 to 2022. Next, we highlight the different categories of fruits and vegetables that may present an opportunity for Africa, taking into account both supply and demand sides. Based on Africa's comparative advantage and global demand, we distinguish between the various processed, semi-processed, and unprocessed fruits and vegetables that Africa should develop and those it should not prioritize, both in the short and long term. The last part of the analysis focuses on the challenges affecting Africa's participation and upgrades in FVVCs. These range from production-specific issues to more general challenges related to poor infrastructure and restrictive trade policies. Finally, the chapter's conclusions provide some policy recommendations, focusing on opportunities for improved intra-African integration in FVVCs.African countries' current production and trade of fruits and vegetables reflect the focus of colonial powers on a few export commodities in each colony, beginning in the early 19th century. In the colonial period, African agriculture had to shift from the production of traditional food crops to export crops-primarily fiber (such as cotton), vegetable oils (such as palm oil and peanut oil), sugar, rubber, cocoa, coffee, and tea (Bjornlund, Bjornlund, and Van Rooyen 2020). The colonial export-oriented policies had major impacts on Africa's rich food system and food security. Following independence, African governments continued to implement the same policies, focusing on exporting one or two cash crops, to maintain the flow of foreign currency needed to fund their industrialization policies.Since the mid-1980s, international trade in fruits and vegetables has grown substantially, driven by rising incomes worldwide (Joosten et al. 2015) and by the rise of supermarkets in developing countries, which has further increased demand for high-quality food products (Swinnen, Colen, and Maertens 2013). The shift in global demand from traditional export crops to high-value products, including fruits and vegetables, has several implications for African countries. On the one hand, the shift creates new opportunities for African farmers to increase their participation in agrifood value chains. On the other, these developments entail potentially severe repercussions for smallholders, who constitute the majority of producers in Africa.To better understand these opportunities and challenges, it is important to understand the structure and governance of FVVCs. Rising trade in fruits and vegetables between developing and developed countries has shaped these value chains around structures that have left most African producers \"stuck\" in upstream, typically low value-added segments of FVVCs. First, growing international trade in fruits and vegetables was accompanied by rising flows of foreign direct investment (FDI) toward developing countries, including Africa. These investments are typically controlled by a small number of multinational companies. Second, the organization and governance of FVVCs is controlled by these large exporting companies, which adopt either a vertical integration structure 3 or rely on contract farming with smallholders (Van den Broeck and Maertens 2016). Third, exports of fruits and vegetables to developed countries require tighter food quality and safety standards, especially as the level of processing increases. This concentration of actors along the value chain, together with stringent standards and regulations, may explain in part why African exporting countries are largely positioned in upstream segments of the value chain, that is, where fruit and vegetable exports are mostly unprocessed, as we demonstrate in the next sections of this chapter.3 Vertical integration refers to a situation in which the whole supply chain is integrated and owned by one firm. ▪ Africa Agriculture Trade Monitor / 2024 Report Chapter 4 -Fruit and Vegetable Value Chains in Africa Nevertheless, participation and upgrade along FVVCs may have several positive implications for Africa. In addition to revenues from their traditional export crops, production of fruits and vegetables can help to diversify African countries' exports and reduce their vulnerability to global commodity price fluctuations. Moreover, compared with traditional cash crops, the value of fruits and vegetables per unit or per weight is higher (Swinnen, Colen, and Maertens 2013). As the demand for processed agrifood products grows, African countries can also benefit from upgrading along FVVCs to promote smallholder commercialization and rural development (Jenane, Ulimwengu, and Tadesse 2022). Ongoing shifts in global demand (especially in emerging markets) toward healthier diets including fruit and vegetable products present an opportunity for Africa to engage in processing activities along global and regional FVVCs.The horticulture sector is typically intensive in low-skilled labor, meaning that participation in FVVCs has potential to increase incomes and reduce poverty, especially for African smallholders. A recent study (Mossie et al. 2021) found that participation in apple and mango value chains in Ethiopia's Upper Blue Nile Basin is associated with 17 percent and 18.5 percent higher household consumption expenditures, respectively. Maertens et al. (2012), in a study in Madagascar, found that vegetables produced under contract farming systems with exporting companies accounted for 47 percent of the household income of involved farmers; and in Senegal, found that participation in bean and tomato value chains had important implications for female empowerment within rural households, due to the female labor intensity of these sectors. Moreover, Van den Broeck and Maertens (2016) suggest that the shift from smallholder contract farming to vertically integrated estate farming affects the labor intensity of FVVCs, as more workers are needed for postharvest activities. Unlike in contract farming, women provide most of the labor in these export companies. Consequently, FVVCs can improve income and food security outcomes not only for smallholders working in contract farming, but also for women through wage employment. Moreover, participation in FVVCs can improve African countries' foreign exchange earnings and trade balance, thus increasing their capacity to import food, among other vital products. Van den Broeck et al. (2018) found that participation in FVVCs increased food security in Senegal through the country's capacity to import food.Notwithstanding the positive outcomes of fruit and vegetable exports, Africa's participation in FVVCs is undermined by several issues. First, fruits and vegetables are seasonal, and their supply chain is characterized by high perishability and susceptibility to waste and loss. Loss can occur due to poor production and harvesting conditions, lack of adequate transportation or poor road conditions, improper packaging, and lack of appropriate storage and cooling. For example, losses in Kenyan production of mangoes are estimated to reach up to 60 percent, most of which occurs before or during harvesting (Ridolfi, Hoffmann, and Baral 2018). Given the agriculture sector's high labor intensity, important income fluctuations can result from yield variations, losses, and waste, which increase the vulnerability of African agricultural communities. In addition, seasonality and perishability are among the main obstacles preventing African firms engaged in processing activities from operating at full capacity year-round (Jenane, Ulimwengu, and Tadesse 2022).Finally, climate change is a major cause of yield variations and among the main challenges facing African agriculture and participation in global value chains. Africa is particularly vulnerable to climate change-related drought shocks, flooding, and extreme weather, which can have severe repercussions on the livelihoods of small farmers. In their study of two districts in Ghana, Williams et al. (2018) found that exposure to climate variability and low capacity to adapt to climate change are among the main factors increasing livelihood vulnerability among smallholder horticultural farmers. Similarly, results from a survey conducted in the Limpopo province of South Africa (Randela 2018) suggest that temperature variability has had a negative and significant impact on avocado yields.In sum, Africa's participation and upgrading along FVVCs entails several benefits and opportunities for different stakeholders, including African governments, investors, smallholders, and consumers. However, Africa's performance along the value chain is undermined by multiple factors, including the current governance structure of the value chains, in addition to other structural challenges, most notably, poor access to technology and know-how, inadequate infrastructure and logistics, restrictive trade policies (both in African countries and their main export destinations), stringent food safety and quality in destination countries, and intra-African sanitary and phytosanitary measures, among others. Following a thorough analysis of African trade and value chain participation and of potential in the fruit and vegetable sector, we offer a more detailed discussion of some of these challenges and limitations.This section provides an in-depth analysis of global and African trade along FVVCs. We begin with an overview of African exports and imports of fruits and vegetables by level of processing and then look at the major exporters and importers at the global level. We then apply this analytical framework to Africa, investigating the major actors at the continental level and the main export destinations for African fruit and vegetable products. We also explore the evolution of intra-African trade in fruits and vegetables and identify the major intracontinental exporters and importers. Throughout this analysis, we look at two periods with a 10-year time interval (2008-2012 and 2018-2022) and at processed, semi-processed, and unprocessed fruits and vegetables in order to identify any changes in the main trade trends over this time period.  processing, between 2003 and 2022. 4 Generally, Africa's exports of fruits and vegetables were dominated by unprocessed commodities over this 20-year period. Moreover, the value of unprocessed fruit exports exceeds that of unprocessed vegetables. In the case of fruit (Figure 4.1, panel A), the gap between the exports of unprocessed goods, on the one hand, and semiprocessed and processed products, on the other, is substantial: exports of unprocessed goods increased from US$2.752 billion in 2003 to $9.433 billion in 2022. 5 Over the same period, the value of semi-processed fruit exports increased from $90.9 million to $627 million and that of processed fruit exports from $359.3 million to $581 million. For vegetables (Figure 4.1,panel B), the value of unprocessed exports also remained above the values of semi-processed and processed products. In 2003, exports of unprocessed vegetables amounted to $759 million, while those of semi-processed and processed vegetables were $250 million and $513 million, respectively. By 2022, exports of unprocessed vegetables reached $3.419 billion, while those of semi-processed and processed products were $1.527 billion and $2.088 billion, respectively. 2 0 0 3 2 0 0 4 2 0 0 5 2 0 0 6 2 0 0 7 2 0 0 8 2 0 0 9 2 0 1 0 2 0 1 1 2 0 1 2 2 0 1 3 2 0 1 4 2 0 1 5 2 0 1 6 2 0 1 7 2 0 1 8 2 0 1 9 2 0 2 0 2 0 2 1 2 0 2 2 Unprocessed Semi-processed ProcessedSource: Authors' elaboration using the 2024 AATM database. However, imports of processed and semi-processed fruits increased much less. In 2022, imports of unprocessed fruits reached $1.904 billion, while the value of semi-processed fruit imports was only $28.6 million (up from $10.5 million in 2003), and that of processed fruit imports was $385 million (up from $150 million in 2003). The picture for vegetable imports is quite different. The value of processed vegetable imports has always exceeded the values of unprocessed and semi-processed vegetables, and the gap widened over the 20-year period. In 2022, imports of 6 The drop in 2017 may have been caused by adverse weather conditions in the world's main fruit-growing regions in 2016 and 2017, which disrupted global production of all major tropical fruits. Mango, papaya, and avocado production were affected by drought in parts of South America and Asia, while pineapple cultivation suffered from flooding in Central and South America. Moreover, tropical storms in the Caribbean in September and October 2017 affected fruit production in small island states (Altendorf 2017). Among the major challenges facing tropical fruit production is that these fruits are mostly grown by smallholders with little access to weather-resilient production systems. Aside from climate shocks, the drop in imports may also have been caused by changes in Egyptian imports. Data on top African importers show that Egypt was on average the top importer of unprocessed fruits in both periods and was likely to be among the top importers in 2016 and 2017 (this period is not covered). However, in 2016, Egypt raised tariffs on 53 lines of food and agricultural products, including fresh and processed fruits considered \"luxury\" products (USDA Foreign Agricultural Service 2016), and also devaluated its exchange rate, with impacts on imports. Source: Authors' elaboration using the 2024 AATM dataset.In sum, except for vegetable imports, Africa's participation in FVVCs is marked by a concentration of trade flows in unprocessed goods. The predominance of unprocessed exports may reflect the global shift in consumer preferences driven by an increasing awareness of the nutritional benefits of fresh fruit and vegetable consumption (especially tropical produce). In developed countries and also a number of developing countries, demand for fresh, high-quality fruits and vegetables is increasing. In the particular case of tropical fruit, the unit price is also typically higher for fresh than for processed items, which increases profit margins from unprocessed exports (Altendorf 2017). At the same time, the relatively modest value of semi-processed and processed fruit and vegetable exports may also reflect the multiple challenges facing African countries in upgrading along FVVCs, including the lack of processing capacities and necessary logistics (such as storage and transport) and the difficulty in meeting international standards for processed fruits and vegetables, or the escalation of tariffs in export destinations (Fukase and Martin 2018). A more detailed discussion of these challenges follows later in this chapter.Finally, the structure of imports may be largely attributed to several factors, including rising incomes, urbanization, and shifts in consumer preferences in Africa. On the one hand, imports of unprocessed fruits may reflect increasing awareness of the benefits of fresh fruit consumption or the growing demand for tropical (mostly imported) varieties, especially in the largest African economies. On the other hand, the predominance of processed vegetable imports could reflect shifts in consumer preferences toward new varieties that are not domestically available, in addition to increasing income and urbanization, which drive up demand for vegetable preparations. Finally, the structure of trade flows may also reflect the lack of domestic vegetable processing capacities.In the following section, we identify the world's major exporters and importers of fruits and vegetables by level of processing in order to investigate, first, whether African countries feature among top exporters or top importing markets, and second, how the top exporters and importers have changed over time.Figures 4.3 and 4.4 rank the top exporters of fruits and vegetables for two periods (2008-2012 and 2018-2022) and by level of processing. The figures suggest three key findings, including, above all, the absence of African countries among the top 10 exporters. Second, the group of top exporters has changed very little over time. Exports are largely dominated by Europe, the United States, and Canada, in addition to China and several Asian and Latin American countries (such as Brazil, Mexico, and Thailand). Third, the world's exports of fruits and vegetables tend to be highly concentrated in a small number of countries, especially for unprocessed exports. For example, the top three exporters of unprocessed fruits (Figure 4.3, panel B) together constituted more than 35 percent of the world's exports. For unprocessed vegetables (Figure 4.4, panel B), this share is as high as 47.3 percent.The presence of China among the top 10 exporters can be traced back to the country's efforts to raise its profile in the global market for agricultural commodities, as part of the agricultural reforms including (but not limited to) the liberalization of agricultural input and output markets in the 1990s and improvements in irrigation and agricultural technology (Guo 2020). Since the 2000s, Chinese exports of fruits and vegetables have been growing rapidly and marked by an increasing diversification of export markets (Mu and Jin 2020). Moreover, China's agriculture sector relies increasingly on intelligent agricultural production, networked agricultural operations, digital technology, big data, and artificial intelligence (Lyu 2020). In Brazil, the agribusiness sector represents 21 percent of the country's GDP (Barros 2020; Mu and Jin 2020). The important role of Brazilian agricultural exports dates to colonial times. Since the 1980s, however, Brazil's agribusiness sector has increased its focus on goods for which there is global demand, such as processed fruits, including orange juice and sugar. Chapter 4 -Fruit and Vegetable Value Chains in Africa are also concentrated among the top 10 (and sometimes, the top 3). For example, the top 10 countries accounted for more than 60 percent of the world's imports of unprocessed and processed fruits (Figure 4.5, panels B and F). In the case of semi-processed fruits (Figure 4.5, panel D), the top 10 importers accounted for as much as 70 percent of the world's imports. For unprocessed vegetables (Figure 4.6, panel B ), the United States and Germany together accounted for 32.5 percent of the world's imports, and in the case of processed vegetables (Figure 4.6, panel F), the top three importers accounted for 39.1 percent of the world's imports.China is also a prominent importer of fruits and vegetables at all levels of processing, a trend that reflects increasing incomes, especially in urban areas, and a growing domestic demand to try \"novelty products\" such as tropical fruits (Altendorf 2017). In the next section, we look at the top exporters and importers of fruits and vegetables among African countries.Since Africa is not among the world's top participants in FVVCs, this section provides a separate analysis of the top-performing African countries, regardless of their share in global trade in these categories. Figure 4.7 shows the top 10 African exporters of fruit by level of processing for both periods (2008-2012 and 2018-2022). Overall, the data suggest a concentration of top African exporters in a limited number of countries.For exports of unprocessed fruits (Figure 4.7, panels A and B), South Africa is by far the largest exporter, accounting for about half of the continent's exports, followed by Egypt and Morocco.Grapes and citrus fruits are among these countries' top exports. South Africa upgraded citrus fruit exports by planting high-quality varieties and responding to rising international standards (Chisoro and Roberts 2024). In the case of Egypt, the production of fruit (especially citrus and grapes) exceeds domestic consumption and is among the main sources of agricultural export revenues (Kassim et al. 2018). Similarly, Morocco is one of the main exporters of oranges and grapes, especially to the European Union (EU) (Santeramo and Lamonaca 2023). In the case of semi-processed fruit exports (Figure 4.7, panels C and D), the composition of the top 10 exporters is similar, but the ranking is different, with Morocco and Egypt together constituting more than 90 percent of these exports during the second period. Finally, South Africa exports nearly half of the continent's processed fruits (Figure 4.7, panels E and F), followed by Egypt (22.4 percent) and Kenya (16.7 percent). South Africa and Kenya, for example, account for 85 percent of Africa's exports of pineapple juice concentrate, while South Africa also accounts for more than half of the continent's orange juice exports, followed by Egypt (Schreinemachers et al. 2022).Figure 4.8 depicts the top 10 African exporters of vegetables by level of processing. Similar to our findings on fruit, exports of vegetables remain concentrated in two to three exporters whose shares are substantially higher than the rest. For unprocessed vegetable exports (Figure 4.8, panels A and B), the composition and ranking remained largely unchanged in both periods, with Morocco being the top exporter and accounting for 58.5 percent during the second period. Together, Morocco, Egypt, and Kenya export more than 90 percent of the continent's total exports of unprocessed vegetables. For example, Morocco and Egypt are the top exporters of fresh tomatoes (Schreinemachers et al. 2022). For semi-processed vegetables (Figure 4.8, panels C and D), Egypt is the top exporter, with a share of about 40 percent during both periods, followed by Ethiopia and Tanzania. Finally, for processed vegetables (Figure 4.8, panels E and F), Egypt is also the top exporter, accounting for 49.6 percent of total exports during the second period. Other top exporters include Morocco and South Africa.Chapter 4 -Fruit and Vegetable Value Chains in Africa Chapter 4 -Fruit and Vegetable Value Chains in AfricaIn the case of fruit imports (Figure 4.9), North African countries (such as Egypt and Morocco) are also among the top-ranked countries. For unprocessed fruits (Figure 4.9, panels A and B), Egypt, Algeria, Libya, and Morocco constitute 78.2 percent of African imports for the 2018-2022 period. For semi-processed fruit imports (Figure 4.9, panels C and D), South Africa was the top importer, with a share of 42.8 percent during the first period, followed by Egypt (29.9 percent). During the second period, Egypt and South Africa had comparable shares (about 29 percent), followed by Algeria and Morocco. The list of top importers of processed fruits (Figure 4.9, panels E and F) also reveals a strong presence of North African countries, with Egypt, Algeria, and Libya among the top importers in both periods. Together, the top 10 importers constitute 73.9 percent of the total African imports within this category for the second period.  Our analysis of the main African exporters and importers of fruits and vegetables highlights several interesting findings: on the one hand, North African countries (primarily Egypt and Morocco) play a major role in exports of fruits and vegetables, along with South Africa (in the fruit sector). The concentration of exports in the top 10 countries can be as high as 90 percent. On the other hand, imports of fruits and vegetables are more balanced. While North African countries and South Africa have a major presence, some countries in West Africa (Guinea, Nigeria, and Senegal) and in East Africa (Ethiopia and Kenya) are also among the top 10 importers.These findings have important implications for value chains. The concentration of fruit and vegetable exports in a relatively small number of countries-together with the predominance of unprocessed exports depicted in Figure 4.1-could be attributed to several factors. First, endowments play an important role, as trade in agriculture is primarily driven by factor endowments (land, climate conditions, and thus the ability to produce and export). Clearly, this explains why these countries have a high comparative advantage in agriculture exports.Second, income is one of the major determinants of the status of African participation in global value chains (GVCs). Our findings suggest that the income levels of the African exporters are the highest in the continent. South Africa, which has a heavy presence in FVVCs, has the highest GDP in the continent, followed by Egypt (which is also among the top African performers in agricultural GVCs). Income levels were also found to correlate with a higher per capita demand for fruit and vegetable products (Mensah et al. 2021), without these trade flows being necessarily related to other processing activities along GVCs (such as in the case of Algeria and Nigeria). This may explain why economies that are larger in terms of GDP per capita, such as South Africa, Egypt, Algeria, and Nigeria, are also among the top importers of vegetables across different levels of processing, although Africa as a whole has the lowest per capita production and consumption of fruits and vegetables in the world (Schreinemachers et al. 2022).Third, it is also important to note that a sufficient and predictable domestic demand for fruits and vegetables is essential for the development of competitive value chains in the first place. These, once developed, could realize economies of scale and later compete internationally. Integration in the global economy also contributes to technology transfers and efficiency gains. For example, countries engaged in FVVCs have higher levels of input and irrigation technology use, which are of particular importance for these value chains (Baumüller et al. 2020). South Africa, for example, has invested in planting high-quality varieties to meet rising international standards. Countries that export and import fruits and vegetables across different stages of processing are also more likely to be engaged in FVVCs, due to endowments in specific crops (such as pineapple in South Africa and oranges and tomatoes in South Africa, Egypt, and Morocco). Thanks to the availability of necessary capital and technology, these countries have better fruit and vegetable processing capacities and are therefore engaged in different processing stages along the value chain. However, for most African countries, innovation among agrifood processing firms is generally low, due to low investments in research and development and limited access to technology (Jenane, Ulimwengu, and Tadesse 2022).Most African countries also rely on small-scale fruit and vegetable production. Given the high perishability of most fruit and vegetable produce and the absence of processing capacities and reliable market outlets for smallholders, domestic and intraregional trade is usually more realistic than international trade (Schreinemachers et al. 2022).Fourth, infrastructure is a significant challenge for Africa's fruit and vegetable trade and for upgrading along value chains due to long distances between producers and consumers coupled with poor road conditions and a lack of refrigerated transportation. Finally, it is important to note that trade policy plays a limited role, as will be shown later in this chapter. On the one hand, the top exporting countries perform better than other African countries due to their endowments and comparative advantage rather than their trade policies.On the other hand, other African countries benefit from preferential access to export marketsthrough the African Growth and Opportunity Act,7 the Generalized System of Preferences, or the Everything but Arms8 initiative-yet are not strong exporters, as they face several nontariff measures that reduce their competitiveness.In this section, we explore the main destination markets for African exports of fruits and vegetables for the two time periods. Figure 4.11 shows the top 10 destinations for African fruit exports. Regardless of the level of processing, the top 10 importers are mostly European countries, the United States, and Canada. The imports of African unprocessed and semiprocessed fruits for both periods of the analysis (Figure 4.11, panels A-D) are concentrated in Europe and the United States and, to a lesser extent, Japan. Other countries, including China and Russia, are also among the top importers. For processed fruit exports (Figure 4.11, panels E and F), the second period shows a diversification of top importing countries, with four African countries among the top importers.The top 10 destinations for Africa's vegetable exports are shown in Figure 4.12. While the top importers of African unprocessed vegetables (Figure 4.12, panels A and B) are predominantly European countries, those of semi-processed and processed vegetables (Figure 4.12, panels C-F) are more diversified and include Arab, African, and Asian countries. Thus, as the level of processing increases and food safety and quality standards become more stringent, the top destinations are more diversified and reflect a stronger presence of Asian and Arab countries as the main importers.Overall, we find that African countries are positioned upstream along FVVCs (that is, more toward the origin of the value chain). The growth of exports of unprocessed fruits and vegetables is significantly greater than exports of semi-processed and processed products. Moreover, Africa's top importers of unprocessed fruits and vegetables are predominantly Europe and the United States, suggesting an upstream position of African countries in the value chain, with their specialization in raw, unprocessed commodities, which are later processed in developed countries and may even be re-exported to Africa for domestic consumption.As was mentioned, gravity considerations play an important role in explaining these trade patterns, given that African countries trade generally with countries characterized by large markets (the United States and China) or with countries with which they had colonial links (France, Portugal, Italy, and the United Kingdom). In fact, these trade patterns are largely in line with the historical role European countries played as the main destination for African exports of (especially unprocessed) fruits and vegetables. Despite the relative maturity and high degree of competition in European markets, these are likely to remain attractive to Africa's fruit and vegetable exports due to increased interest in plant-based diets and healthy foods, in addition ▪ Africa Agriculture Trade Monitor /Chapter 4 -Fruit and Vegetable Value Chains in Africato their relative profitability (COLEACP 2020). At the same time, trade with other regions has been growing rapidly, especially at the higher levels of processing. Between 2002 and 2017, Africa's exports of fruits and vegetables (processed and unprocessed) to East Asia grew at an average rate of 9.6 percent per year, compared with only 1.1 percent for exports to the EU (COLEACP 2020). For the 2018-2022 period, Asian and Arab countries are also among the top destinations for Africa's semi-processed and processed fruit and vegetable exports. This reflects the ability of African countries to cater to markets with less stringent standards and sanitary restrictions and highlights the potential for developing vegetable processing industries and serving geographically close markets in the Middle East or rapidly growing markets in Asia. Despite this potential to upgrade along FVVCs, it is important to note that tariff escalation contributes to the concentration of African exports in unprocessed products.Tariff escalation refers to situations in which lower tariffs are imposed on unprocessed products and higher tariffs on processed ones, which is common in Africa's main export destinations, including China, the EU, and the United States (Antimiani, Di Maio, and Rampa 2011).Our initial analysis also suggests the presence of regional value chains, with some African countries among the top importers of Africa's processed and semi-processed fruits and vegetables, which we discuss in the next section.Chapter 4 -Fruit and Vegetable Value Chains in Africa Source: Authors' own elaboration using the AATM 2024 database.Chapter 4 -Fruit and Vegetable Value Chains in AfricaIn this section, we examine intra-African trade patterns and compare the observations with the findings on global trade above in order to better understand the challenges facing intra-African trade in fruits and vegetables.Trade by level of processing Total intra-African fruit and vegetable trade amounted to $2.36 billion in 2022, of which $1.55 billion was trade in vegetable products and $812 million was trade in fruit products. This is an increase of $750 million, almost 50 percent, since 2012. Of total African exports of fruits and vegetables, intra-African trade accounts for only about 6 percent of trade in fruits and 17 percent of trade in vegetables. About 21 percent of fruit imports and about 11 percent of vegetable imports are sourced from another African country. 9 However, intra-African trade statistics are likely underestimating the total level of intra-African trade due to the high level of informal trade (Bouët, Cissé, and Traoré 2020). Over the 2018-2020 period, overall, about 40 percent of intra-African fruit and vegetable trade was in the form of unprocessed commodities, 50 percent in processed products, and about 10 percent in semi-processed products. This has not changed substantially over the past 20 years, with only a few variations in individual years.There are a few differences between intra-African trade in fruit and vegetable products (Figure 4.13). Among fruit products, there is virtually no trade in semi-processed products, while about 20 percent of intra-African vegetable trade is in semi-processed products. Furthermore, intra-African fruit trade is primarily unprocessed, and the share of unprocessed fruits even increased between the 2008-2012 and 2018-2022 periods. Conversely, about 60 percent of intra-African vegetable trade is in processed products (about the same in both periods). This is both a consequence of the definition of semi-processed products, which are fewer in number than unprocessed and fully processed products, and the fact that regional FVVCs in Africa are very limited. For instance, Odjo and Diallo (2022) discuss Africa's role in GVCs and show, despite an increasing trend, limited African participation. As a result, FVVC products are either fully processed or unprocessed. They argue that this is caused by small, narrow manufacturing sectors that require additional cross-border foreign direct investment. Limited cross-border infrastructure and complex trade regimes, including rules of origin regulations, also contribute to limited participation in regional value chains (Kornher and von Braun 2020).The implementation of the AfCFTA in coming years offers an opportunity to address these policy constraints and improve intra-African trade. The top 10 intra-African fruit and vegetable trade destinations are shown in Figures 4.14 and 4.15. Botswana, Kenya, and Mozambique are the main destinations of intra-African vegetable trade for processed, semi-processed, and unprocessed products. Several countries appear in the top 10 list for two levels of processing: Mozambique, Morocco, South Africa, Somalia, Libya, Botswana, and Namibia (Figure 4.15). Three of these countries-South Africa, Libya, and Namibia-are also in the top 10 of intra-African fruit destinations (Figure 4.14). These findings differ from those in the previous section on top African importers (both intra-and extra-African trade). Many of the top African exporters from North Africa are less relevant for intra-African trade. This hints at better trade integration of these countries within the Mediterranean region than with sub-Saharan African countries. Generally, all African regions, except Central Africa, are frequently ranked among the top 10 destinations. While the size and purchasing power of the import market may explain why several of the wealthier African economies are among the top fruit and vegetable import markets (for example, Morocco, Nigeria, and South Africa), it is remarkable that several small countries-Eswatini, Lesotho, and Djibouti-are also ranked among the top 10 intra-African import destinations. The low level of fruit and vegetable imports overall signifies demand-side constraints among African importers, evident from the low levels of per capita fruit and vegetable consumption across the continent. There is no clear clustering of top importers of unprocessed and semi-processed fruit and vegetable products, which again supports the supposition that regional FVVCs are not well developed. Therefore, enabling and promoting African regional FVVCs should be a priority in the AfCFTA implementation process.   Chapter 4 -Fruit and Vegetable Value Chains in AfricaIntra-African top exportersThe group of top 10 intra-African exporting countries is quite different and much more concentrated than the top 10 intra-Africa import destinations (Figures 4.16 and 4.17). For instance, Egypt and South Africa are among the top three intra-African exporters of fruits and vegetables at all levels of processing. It is not surprising to see these countries in the lead for intra-African fruit trade, as we have seen they are among the top African exporters (Figures 4.7 and 4.8). Hence, the top African fruit and vegetable exporters also lead in intra-African fruit and vegetable trade. In fact, these countries are global fruit exporters, specifically of citrus fruits, and therefore have developed internationally competitive FVVCs (Seleka and Obi 2018). At the same time, many other African countries lack the production capacity, such as irrigation and inputs, and value chain requirements, such as cold storage transport and facilities, to produce and trade fruit and vegetable products at large scale (Baumüller et al. 2021).Regarding  (COLEACP 2020). The development of this regional trade may have several causes: better infrastructure and improved market access of smallholders; more efficient and liberal trade policy frameworks in SADC and EAC; and comparative advantages in terms of geography as well as land and labor productivity. The first two will be directly addressed if the AfCFTA is successfully implemented. Generally, the concentration of intra-African exports is less pronounced than it was 20 years ago. Among the most traded vegetable products within Africa are food preparations and sauces as well as onions, beans, and tomatoes.  Chapter 4 -Fruit and Vegetable Value Chains in Africa In this section, we compare African supply to global demand for processed, semi-processed, and unprocessed fruits and vegetables in order to highlight those products for which there is potential for Africa to expand exports. We begin the analysis by categorizing all fruits and vegetables according to Africa's revealed comparative advantage (RCA)10 and global demand, as measured by world imports.11 This yields four categories: (1) All fruits and vegetables for which Africa has an RCA and global demand is high. This category constitutes a true opportunity for Africa, and countries should focus on these products and promote their export.(2) All fruits and vegetables for which Figure 4.18 shows the share of products categorized under each of these four groups in total African fruit and vegetable exports. One of the main findings is that, in both time periods, African countries did not export any processed or unprocessed fruits or vegetables for which they have an RCA and that enjoy high global demand. A modest share of semi-processed products (2 percent of all exported fruit and vegetables in the first period and 1 percent in the second period) satisfies both conditions. Fruits and vegetables for which Africa has an RCA, but for which global demand is low, also represent a relatively minor share in total exports of these goods. This share does not exceed 3 percent for any of the three levels of processing in either time period. While this category of exports may be useful in the short term, focusing on these products in the long term is not recommended, given the low global demand.A considerable share of fruits and vegetables exported by Africa do not have an RCA and face low global demand, meaning they should not be prioritized given the weak potential on the supply as well as on the demand side. Product shares in this category range from 11 percent for semi-processed products to 15 percent and 16 percent for unprocessed and processed products, respectively.Fruits and vegetables for which global demand is high but African countries do not have an RCA constitute a large share of all exported products. For both periods, this category represented 49 percent of all fruit and vegetable products, with a major share of unprocessed products (27 percent during the first period and 29 percent during the second period). This category of products has potential benefits for African countries, as they could be developed in the long term. However, two important factors may limit this potential: first, the ability of African countries to use their resources and endowments to develop these products and increase their exports; and second (and more importantly), the availability of water and suitable climate conditions to grow these crops. As mentioned, Africa is among the regions most exposed to extreme weather fluctuations, with severe consequences for agriculture, especially for smallholders.Water availability can also be problematic, as several African countries are characterized by either a high level of water stress (North Africa) or a low level of water productivity (sub-Saharan Africa). This affects products that are water intensive, such as oilseeds, nuts, rice, and oats.Chapter 4 -Fruit and Vegetable Value Chains in AfricaIn addition, the impact of climate change on agricultural markets may lead to a 0.26 percent reduction in total global GDP, with several African countries severely affected, according to Costinot et al. (2016). Similarly, Mahofa (2022) argues that, by the 2050s, climate change will affect production and thus increase African countries' cereal imports. In the same vein, Gouel and Laborde (2021) show that export shares for maize, wheat, and rice will decrease for Africa by 2080 due to declining yields. Chapter 5 of this report presents a thorough discussion on the impact of climate change on African comparative advantage. (2) High (low) demand refers to products whose world imports are greater (less) than the median world imports over the period of analysis. With (without) RCA refers to products whose revealed comparative advantage index is greater (less) than 1.Table 4.1 provides a more detailed classification of fruit and vegetable exports using these four categories. This detailed presentation can help to identify specific products for African countries to focus on in the long term. As mentioned, African countries do not have any processed or unprocessed fruit and vegetable exports for which their supply is competitive and global demand is high. Among semi-processed products, both mushrooms and truffles and specific bean species were the two product categories for which Africa had a comparative advantage and global demand was high during the 2008-2012 period. 12 During the 2018-2022 period, only semi-processed mushrooms and truffles were in this category.Some other products may be beneficial to export in the short term but should not be promoted in the long term, given relatively low global demand. These include specific processed roots and tubers (including arrowroot and Jerusalem artichokes, tapioca preparations) and some bean preparations, adzuki beans, and legumes (semi-processed) and unshelled hazelnuts, dried prunes, and dried apples (unprocessed).12 See Table A4 Another category of interest are products for which Africa could potentially build a comparative advantage in the long term to benefit from the high global demand. For processed fruits and vegetables, this category includes, among others, preserved vegetables, fruits, and nuts; preserved tomatoes; mixed frozen vegetable preparations; frozen potatoes; preserved olives; preserved sweet corn; fruit jams, purees, and pastes; preserved pineapples; orange juice; and some sauces (including tomato sauce and ketchup). Top African exporters of these products, such as Egypt and South Africa for orange juice and Egypt and Morocco for tomato products, could work on overcoming the challenges (related to resources or food safety, for example) to increase their share in the global market. The same applies to processed vegetables, for which the market is expanding both worldwide and in Africa with increasing urbanization and preference for easy-to-prepare meals. For semi-processed fruit and vegetables, potential products include frozen cooked or uncooked vegetables, dried mixtures of vegetables, dried peas, some types of dried beans (including kidney beans), dried lentils, berries (including frozen strawberries), and some types of frozen fruit and nuts. Finally, for unprocessed exports, this category includes numerous fruits (such as apples, oranges, mandarins, grapefruit, grapes, peaches, plums, strawberries, almonds, and shelled hazelnuts), some of the major tropical fruits (pineapples, mangoes, and guavas), and vegetables (such as potatoes, tomatoes, onions and shallots, garlic, broccoli and cauliflower, lettuce, carrots, cucumbers, and mushrooms).As mentioned, tropical fruits are typically more profitable in their fresh than processed state. Despite the multiple challenges facing the agriculture sector in Africa, some countries may seize this opportunity and work on scaling up their exports in this category.It is important to note that while we have assessed the potential evolution of supply and demand for fruits and vegetables in Africa in order to identify which sectors can generate higher benefits, the indicators we used do not account for other important factors, including the comparative advantage of competitors or possible changes in endowments or external conditions (such as climate change).Chapter 4 -Fruit and Vegetable Value Chains in AfricaThe description of past and current fruit and vegetable trade patterns of African countries and the market demand analysis clearly show the limited capacity of African countries to engage in global and regional value chains. The reasons for this are multifaceted and include sectoral, institutional, and structural issues. In this section, we discuss the challenges in detail, looking at production processes, post-production processes, and trade policy.Africa's agricultural sector is not performing at its full potential due to a variety of interrelated factors, including the lack of adoption and investment in production-enhancing technologies and inadequate institutional frameworks (Baumüller et al. 2020). As a result, African fruit and vegetable yields are far below yields in other regions (FAOSTAT 2024), which limits Africa's capacity to produce sufficient fruits and vegetables to meet consumption needs. A simulation using the International Food Policy Research Institute (IFPRI)'s IMPACT model shows that many countries around the world will need to increase fruit and vegetable production to achieve the World Health Organization's dietary recommendations, even if waste is reduced to zero (Mason-D'Croz et al. 2019). There are specific challenges to increasing FVVC productivity and production, which we discuss below.Access to seeds of high-quality improved varieties that are adapted to local agroecologies, pest risk, and farmer preferences is a key element to increasing fruit and vegetable yields in Africa. Current limitations reflect the lack of selection and breeding studies for fruits and vegetables that are suited for the region, and particularly for traditional African vegetables (Dinssa et al. 2016). Very few seed companies operate in Africa, and even fewer have invested in research and development to create locally adapted varieties. Instead, most of these seed companies have based their businesses on trading and distributing seeds (Afari-Sefa et al. 2012). Despite the arrival of multiple international seed corporations, there is still very little breeding of vegetables or other crops for the local market in sub-Saharan Africa (Access to Seeds Foundation 2019).The low adoption of agricultural inputs to increase soil fertility and of pesticides to control pests is another contributor to poor fruit and vegetable yields in Africa. Given the low levels of fertilizer use prevalent in Africa compared with other regions (Kirui, Kornher, and Bekchanov 2023), agricultural production is very responsive to increasing chemical fertilization (Kornher and von Braun 2024). For example, Rosegrant et al. (2014) show that the yield increases that could be achieved through nitrogen-efficient technologies in Africa are higher than in other regions. This is important for fruits and vegetables because they deplete soil nitrogen more than other products, and therefore low fertilization contributes to ongoing soil degradation in Africa (Elrys et al. 2020). Regarding pesticide use, African farmers applied the lowest levels in terms of cropland area, population, and the value of their agricultural production between 1990 and 2020 (FAO 2022). Among all pesticides, herbicides, fungicides, bactericides, and insecticides have relatively equal shares. Most of the pesticides applied in Africa are imported from outside Africa, with only about 10 percent of the pesticide trade occurring within Africa.The constraints to enhanced fertilizer and pesticide availability in Africa are similar to those limiting adapted seeds: Fertilizer and pesticide production is limited in Africa. However, unlike adapted seeds, fertilizers and pesticides are less context-specific. Therefore, decisions about FVVCs largely depend on frequent water applications in many parts of the world. Water is required in different stages of FVVCs, including growing, processing (produce washing, packhouse wash down, sanitation), and distribution (wash down). Many fruit and vegetable crops, such as tomatoes and cucumbers, have high water content, and their yields and quality deteriorate under water stress. Therefore, a secure and reliable water supply is important to ensure productivity and quality. In Africa, however, crop cultivation is primarily rainfed, and only about 5 percent of agricultural land is irrigated (FAOSTAT 2024). With progressing climate change and uncertain precipitation, fruit and vegetable yields, particularly in semi-arid areas, will be under stress, and the expansion of irrigated cropland will be essential to mitigate climate change effects on yields (Hess and Sutcliff 2018). For instance, North African countries, which have substantially more crop area under irrigation, have higher fruit and vegetable yields than countries in sub-Saharan Africa and export vegetable products to the European Union (ZEF and ICRIER 2021). This suggests that irrigation is essential to lift fruit and vegetable farming from the subsistence to commercial level.Irrigation expansion can be achieved through large-scale irrigation schemes that employ water diversion and dams, as well as through the adoption of small-scale irrigation systems in the form of local pumps with substantial impact. For example, the Bwanje Valley Irrigation Scheme in Malawi increases the agricultural incomes of participating farmers by 65 percent and their caloric intake by 10 percent (Nkhata 2014). For small-scale irrigation, studies have shown that the adoption of simple irrigation technologies in Burkina Faso contributed to an increase in national vegetable production of between 60,000 and 160,000 tons within nine years. Households in Tigray, Ethiopia, that use irrigation earn double the income of households that do not have access to irrigation, with overall income gains of around $150 per household per year (Malabo Montpellier Panel 2018).IMPACT model simulations run by IFPRI show that under current conditions, sub-Saharan Africa will require net imports of 36 million tons of vegetables in 2050. In contrast, if irrigation is expanded, the region could become a net vegetable exporter. 13 Besides the macroeconomic effects of irrigation through yield increase, the adoption of irrigation technologies is also found to have strong microeconomic effects. For instance, smallholder households that adopt irrigation can increase crop diversity and expand fruit and vegetable production. Smallholders, who frequently practice irrigation, grow more vegetables, fruit, and other micronutrient-rich crops, particularly during the dry season. This increases the consumption of nutritious foods and offers additional income opportunities when selling these products in the market (Ringler et al. 2022).Food loss and waste (FLW) is a major challenge to sustainable food systems and is estimated to reach about 20 percent of total production quantities in Africa, which is substantially higher than the global average (FAO 2024). Postharvest losses are caused by high perishability. ForChapter 4 -Fruit and Vegetable Value Chains in Africa instance, reported losses for FVVC products vary between 0 and 80 percent but are related to several other factors, such as weather and production and transportation mode (Santacoloma et al. 2021). In low-and middle-income countries, and Africa specifically, most FLW occurs in the field and postharvest and not during consumption. Causes include inadequate production methods, the incidence of diseases, poor on-farm storage after harvesting and before marketing, excessive temperatures and humidity during storage and handling, weather conditions, the type of packaging, and time delays and handling during transportation, such as delays caused by road harassment (Bouët, Sall, and Traoré 2023). Many of these factors are more relevant in Africa than in other regions, where storage facilities are better equipped, management is more professional, and infrastructure and energy are available to manage temperatures during transport and storage.Fruit and vegetable products are highly perishable, so losses are higher than for cereal and legume crops (Houngbo 2019). Houngbo (2019) estimates that FLW for fruits and vegetables in sub-Saharan Africa could be up to 50 percent. Specifically, they reach 55 percent for fruits and about 45 percent for vegetables (Santacoloma et al. 2021). These estimates are in line with a systematic literature analysis by Kitinoja and Kader (2015), who found that reported FWL globally is between 30 and 40 percent, with little change since 1970.Postharvest losses result in monetary losses at the production, processing, and wholesale level.In addition, perishability also leads to substantial quality deterioration in fruits and vegetables, which also contributes to monetary losses. Estimates suggest that monetary losses from both quality deterioration and losses range from 4.8 to 81 percent for tomatoes, amaranth leaves, okra, oranges, and mangoes that suffer damage, spoilage, or decay at the farm level; between 5.4 and 90 percent at the wholesale level; and between 7 and 79 percent at the retail level (Santacoloma et al. 2021). These losses, also high compared with those of other food crops, are a disincentive to the production and marketing of fruit and vegetable products in Africa.Post-production processes: Market access and infrastructureConstraints to agricultural production are only one element that hinders growth of FVVCs.Fruits and vegetables are produced worldwide, but not all producers and traders have equal access to markets because value chains are seldom organized efficiently. For instance, many small farmers cultivate different species and produce in small quantities, which makes formal vertical market linkages less likely. Small producers often sell through middlemen, and most of their transactions are informal and do not fulfill food safety or quality control (grading) requirements. However, in fruit and vegetable supply chains, we must distinguish between value chains intended for export and those intended for domestic markets.Agriculture in Africa remains largely at the subsistence level, although market participation has increased in recent years (Carletto, Paul, and Guelfi 2017). Formal and informal links to local and global value chains are important to encourage producers to allocate resources toward fruit and vegetable cultivation and invest in production technologies. Traditionally, vegetables are mainly grown for subsistence, often at a small scale in kitchen gardens close to the homestead (Issahaku et al. 2023), and therefore move along traditional value chains. However, many tropical fruit products are produced for international export, and the creation and expansion of GVCs has increased vertical market linkages for these products.There are several causes for the limited market participation of African vegetable farmers. First, transaction costs of trading with smallholder farmers, many of them located in remote Infrastructure quality is highly relevant for fruit and vegetable marketing, as these perishable products are particularly prone to damage during transportation. However, road and storage infrastructure in Africa is poor compared with other regions. Long transportation times related to poor road infrastructure and road blockages and trade bureaucracy are associated with high losses during transport of fruit and vegetables, though losses differ between products and value chains (Santacoloma et al. 2021). One particular culprit is the lack of frozen storage, which can reduce loss and damage, for many FVVCs in Africa.Numerous studies show that improved rural infrastructure increases agricultural output and/or revenue by lowering market transaction costs and increasing access to both input and output markets. Thus, improved infrastructure lowers the cost of inputs and increases the income farmers receive from their products but also enables smallholders in low-and middle-income countries to profit from more nonfarm opportunities (van Berkum 2021). Given the importance of transport and storage infrastructure for fruit and vegetable trading, improved infrastructure is clearly critical for development of FVVCs. For instance, Barrett et al. (2022) argue that the selection of areas for fruit and vegetable production in exporting countries is largely explained by infrastructure factors, such as road infrastructure and electricity.Fruits and vegetables are the most consumed nonstaple food crop in Africa. The EAT-Lancet Commission recommends consuming at least 240 to 300 grams of vegetables per capita per day. Low demand for fruits and vegetables is largely associated with relatively high prices.For instance, the relatively high prices for fruits and vegetable products, compared with other foods, in Africa and elsewhere contributes to low demand (Headey and Alderman 2019). In turn, limited demand in Africa for local fruit and vegetable products reduces the size of local markets and does not create incentives for local producers and value chains to increase production. However, when compared to other food products, the demand for fruits and vegetables in lowand middle-income countries is price-and income-elastic. Thus, even slight price adjustments for these goods can have a significant impact on demand. For instance, some studies indicate that a 10 percent increase in the cost of fruits and vegetables could result in an 8 to 10.5 percent decrease in consumption (Magrini, Balié, and Morales-Opazo 2017). Similarly, a small increase in income can lead to substantial increases in fruit and vegetable consumption.Chapter 4 -Fruit and Vegetable Value Chains in AfricaTrade policy and certification requirementsTrade policy issues are eminently relevant to Africa's FVVCs. Intra-African trade liberalization and international market access could create incentives to expand fruit and vegetable production and improve allocative efficiency. Currently, tariff escalation in high-income export markets promotes trade in raw products and hinders trade of processed fruit and vegetable products that could lead to increased value addition in Africa (van Berkum 2009). Nontariff measures (NTMs) and associated trade bureaucracy increase trade costs and may be used by importing countries to protect local producers. The trade costs associated with NTMs are exacerbated by the limited institutional export capacity, such as port efficiency, of African producers (Kornher, Sakyi, and Tannor 2024). In addition, ad hoc border closures, such as the situation between Uganda and Rwanda where the border has been closed for three years, disrupt trade flows and create uncertainty for traders.Globally, tariffs on agricultural products have been reduced substantially as a consequence of international trade liberalization, including World Trade Organization (WTO) reforms and the increasing number of regional and preferential trade agreements. Several African countries are granted market access to export markets under unilateral preferential trade agreements, such as the Everything but Arms agreement, and multilateral regional trade agreements, like the African Growth and Opportunity Act agreement with the United States and the Economic Partnership Agreements with the EU (Kornher and von Braun 2020). For the remaining trade partners, FVVCs are subject to tariff escalation, with tariffs on processed produce generally higher than on the raw commodities, which limits Africa's participation in value added from agricultural trade (Fusacchia, Balié, and Salvatici 2022). In addition, the EU's entry price system restricts fruit and vegetable imports from North Africa if prices fall below a set price threshold (Santeramo et al. 2023). In intra-African trade, FVVC trade within regional economic communities (RECs) benefits from reduced or suspended tariffs, but preferential access is not commonly expanded to countries outside an individual REC. However, these REC agreements will be subject to changes with the upcoming AfCFTA.Apart from tariffs, agricultural trade in general, and fruit and vegetable trade in particular, are subject to significant nontariff trade costs, primarily related to sanitary and phytosanitary standards (SPS). SPS is necessary due to the food safety issues and health risks associated with perishable products and is motivated by the precautionary principle in high-income importing countries, especially the EU (Otsuki, Wilson, and Sewadeh 2001). Between 1995 and 2000, nearly 270 SPS measures were imposed on imports of fresh fruit and vegetables worldwide (UNCTAD Trains). Often these standards and required certification deviate from the joint FAO/ WHO Codex Alimentarius Commission for food safety, which sets the international standards promoted by the WTO Agreement on Sanitary and Phytosanitary Standards. For instance, the EU's pesticide maximum residue levels are stricter than international standards in sectors where EU producers compete with African exporters (Kareem, Martínez-Zarzoso, and Brümmer 2018). Exporters must navigate numerous requirements and regulations, including soil quality checks and certification standards compliance, which increase operational costs. At the same time, agricultural exporters must comply with the quality standards of the private sector, such as the Global G.A.P. 14 Ensuring quality control and adherence to standards remains a challenge. Substandard farming inputs and poor awareness among farmers regarding approved chemicals and farming practices result in the use of banned or inappropriate inputs, which leads to rejected produce and compromises both local food safety and export potential. In the intra-African context, SPS and quality regulations are less strict, yet differences exist among and within RECs, posing challenges to harmonization and smooth cross-border trade in FVVC products within and between regions. Additionally, differences in other regional trade policies, such as food standards, grading systems, and border procedures, further complicate intraregional trade in fruit and vegetable products. The lack of uniformity in these policies creates barriers and inefficiencies in trade flows, hindering the sector's growth potential. These costs increase when countries are members of overlapping RECs that apply different standards.Some RECs, such as the Economic Community of West African States (ECOWAS), have advanced quality infrastructure-institutional and physical systems to ensure products are safe and of high quality-that supports the continental quality infrastructure. Within ECOWAS, regional quality infrastructure has been established to ensure compliance and safety of products, particularly in the fruit and vegetable sector. Under the auspices of the ECOWAS Scheme for Harmonization of Standards (ECOSHAM), more than 90 standards have been harmonized, covering various areas including agricultural and food products. ECOSHAM certification ensures broad acceptance of products in all ECOWAS member states, thereby facilitating access to other markets in ECOWAS. The ECOWAS SPS guide outlines comprehensive procedures for phytosanitary inspection, focusing on plants, plant products, and regulated articles in international traffic.In 2013, the EAC Partner States adopted the EAC SPS Protocol, with the primary aim of enforcing SPS measures and standards as well as promoting both intra-and interregional trade. As of 2021, all partner states had ratified the protocol, clearing the way for implementation and domestication of various SPS instruments (EAC 2024). To lay the groundwork for effective implementation and enforcement of the protocol, several key instruments were developed and adopted, including SPS measures and procedures for fish and fisheries, phytosanitary measures and procedures for plants, and food and feed safety measures. Additionally, a draft SPS bill that provides a legal framework for the enforcement of EAC SPS measures and instruments was adopted by the Sectoral Council on Agriculture and Food Security and is currently awaiting enactment by the East African Legislative Assembly. Moreover, harmonized SPS regulations and standard operating procedures necessary to facilitate the implementation of the SPS bill have been developed.The SADC Protocol on Trade (SADC Protocol) emphasizes the harmonization of SPS measures for agricultural and livestock production based on international standards, guidelines, and recommendations, with provisions for consultations to achieve agreement on the recognition of equivalent SPS measures (Article 16). This protocol also offers a framework for collaboration and cooperation on SPS issues, focused on facilitating the protection of human, animal, or plant life or health, enhancing the implementation of the WTO Agreement on the Application of SPS Measures, building technical capacity, providing a regional forum for addressing SPS matters, and resolving trade-related SPS issues (Annex VIII). The SADC SPS Coordination Committeetasked with addressing regional SPS issues, promoting transparency, and strengthening cooperation between national regulatory agencies responsible for SPS measures-plays a pivotal role. National Committees on SPS Measures are also established in each member state, responsible for their WTO SPS National Notification Authorities and Enquiry Points, with representatives appointed to serve on the SADC SPS Coordinating Committee. The SADC Protocol includes provisions such as Article 11 on Control, Inspection, and Approval Procedures, which improve on the WTO SPS Agreement by facilitating the acceptance of equivalent procedures and reviewing inspection, testing, certification, and approval systems to enhance access to traded products (CFTA 2017).Chapter 4 -Fruit and Vegetable Value Chains in AfricaThe Economic Community of Central African States (ECCAS) coordinates SPS activities in Central Africa through regional programs that support member states, with the assistance of technical and financial partners such as the Food and Agriculture Organization of the United Nations (FAO). While 9 of the 11 ECCAS member states are also members of the WTO and also implement the WTO SPS Agreement, the ECCAS SPS program is still in its early stages. Achievements include the development of a joint phytosanitary regulation project, capacitybuilding activities, establishment of National SPS Committees and Focal Points, launch of a regional pesticide registration system, establishment of an interstate Committee on Pesticides in Africa in Central Africa, operationalization of a Regional Animal Health Centre, adoption of coordinated approaches in monitoring cross-border diseases and zoonoses, implementation of regional programs on health safety during disease outbreaks and for vector-borne diseases, and operationalization of the Regional Food Safety Program in Central Africa (CFTA 2017).This chapter has analyzed Africa's participation in FVVCs, highlighting challenges and opportunities for increased fruit and vegetable trade and upgrading along these value chains. One of our main findings is that, over the past 20 years, African exports of fruits and vegetables have been consistently dominated by unprocessed goods. This is more pronounced for exported fruits than for vegetables. At the same time, imports are dominated by unprocessed fruits and processed vegetables. This means that Africa is in an upstream position along FVVCs. While this may be profitable in the case of tropical fruit exports, African countries may be still missing opportunities to secure a place in the expanding market for processed fruit and vegetable products.As far as the global market is concerned, African countries are entirely absent from the list of the top 10 fruit and vegetable exporting and importing countries, regardless of the level of processing. Exports are largely dominated by Europe, the United States, and Canada and, to a lesser extent, China and a few Asian and Latin American developing countries.Our analysis of African trade in fruits and vegetables suggests a strong presence of North African countries (primarily Egypt and Morocco), in addition to South Africa, as the top exporters and importers. These countries may be engaged in FVVCs due to resource endowments, suitable agricultural and export upgrade policies, and better processing capacities, logistics, and transport and storage conditions compared with other African countries. At the global level, Europe and the United States are the main importers of African unprocessed fruits and vegetables. As the level of processing increases, top importers are more diversified, with a larger presence of Asian and Arab countries as importers. This suggests the ability of African countries to meet demand for processed products in countries with less stringent standards and sanitary restrictions. At the intra-African level, fruit trade is primarily unprocessed, whereas nearly 60 percent of intra-African vegetable trade is in processed products, reflecting the growing urbanization and demand for easy-to-prepare meals. Egypt and South Africa are among the top fruit and vegetable intra-African exporters for all levels of processing, and some SADC and EAC countries are among the top intracontinental exporters.We combine data on global demand and African supply to single out specific fruit and vegetable products that Africa should focus on developing and exporting in the long term. Our findings suggest that Africa's exports of fruits and vegetables with an RCA and for which global demand is high are quite minimal. However, the structure of African exports reflects a strong presence of fruits and vegetables that could be expanded in the long term. These include a rich variety of fruits and their semi-processed and processed products (such as apples, citrus","tokenCount":"11637","images":["1564303628_1_3.png","1564303628_9_3.png","1564303628_17_3.png","1564303628_22_3.png"],"tables":["1564303628_1_1.json","1564303628_2_1.json","1564303628_3_1.json","1564303628_4_1.json","1564303628_5_1.json","1564303628_6_1.json","1564303628_7_1.json","1564303628_8_1.json","1564303628_9_1.json","1564303628_10_1.json","1564303628_11_1.json","1564303628_12_1.json","1564303628_13_1.json","1564303628_14_1.json","1564303628_15_1.json","1564303628_16_1.json","1564303628_17_1.json","1564303628_18_1.json","1564303628_19_1.json","1564303628_20_1.json","1564303628_21_1.json","1564303628_22_1.json","1564303628_23_1.json","1564303628_24_1.json","1564303628_25_1.json","1564303628_26_1.json","1564303628_27_1.json","1564303628_28_1.json","1564303628_29_1.json","1564303628_30_1.json","1564303628_31_1.json","1564303628_32_1.json","1564303628_33_1.json","1564303628_34_1.json","1564303628_35_1.json","1564303628_36_1.json","1564303628_37_1.json","1564303628_38_1.json","1564303628_39_1.json","1564303628_40_1.json","1564303628_41_1.json","1564303628_42_1.json","1564303628_43_1.json","1564303628_44_1.json","1564303628_45_1.json","1564303628_46_1.json","1564303628_47_1.json","1564303628_48_1.json","1564303628_49_1.json","1564303628_50_1.json","1564303628_51_1.json","1564303628_52_1.json","1564303628_53_1.json","1564303628_54_1.json","1564303628_55_1.json","1564303628_56_1.json","1564303628_57_1.json","1564303628_58_1.json"]}