106434 S IDA pe ci Co ResearchDigest al u Is ntr World Bank su ie e s on VOLUME 10 NUMBER 4 SUMMER 2016 Changes in Poverty and Female-Headed Households in Africa In Africa poverty is falling faster for percentage point increase in the share IN THIS ISSUE households headed by women than of the population living in female- for those headed by men headed households. A one-year rise Changes in Poverty and Female-Headed in women’s average age at first mar- Households in Africa … page 1 T riage leads to a 2.5 percentage point oday in Africa, a very sizable mi- increase in the population share living Growth and Capital Inflows nority of households are headed in female-headed households—an in Africa … page 2 by women—26 percent according effect almost as strong as that of an to data from the latest Demographic extra year of schooling. Higher overall Are Public Works Working in Malawi? … page 3 and Health Surveys. While there are life expectancy also has a positive ef- differences across countries, both fect—equal to a 0.5 percentage point Vulnerability to Climate Change the share of the population living in boost per extra year. This presumably in Coastal Bangladesh … page 4 female-headed households and the reflects the natural survival advantage share of households headed by wom- of women that is revealed with higher Improving Agricultural Data en have been rising over time. The life expectancy and the resulting preva- for Better Policies … page 5 data show quite clearly that the prob- lence of widow-headed households. ability that a woman age 15 or above Conflict and HIV also raise a country’s Enhancing Transparency of Large-Scale heads a household (controlling for her share of population in female-headed Land Acquisition … page 6 age) has been increasing over time in households. all regions and across the entire age Thus the prevalence of female- Explaining the Gender Gap distribution. headed households has been rising in Agricultural Productivity … page 7 What explains this? Milazzo and while poverty has been falling. Past van de Walle investigate this question literature has been generally sugges- Changing Patterns of Growth and Poverty Reduction in India … page 8 in a recent paper using the full series tive that female-headed households of Demographic and Health Surveys tend to be poorer. But does this imply Land Market Restrictions and fielded in Africa over the past 25 years that female-headed households have Rural Labor Markets … page 9 and covering 89 percent of Africa’s been left behind by recent improve- population. Their results suggest that ments in living standards? Female Deforestation Prevention Programs and economic growth brings lower female heads of household are a diverse Community-Managed Forestry … page 10 headship—presumably in part because group. Some—such as married women of lower work-related migration by with a nonresident (polygynous or mi- Recent Policy Research Working Papers men, associated with a growing local grant) husband, or educated women on IDA Countries … page 12 economy. Yet as the data show, female who may choose, and can socially and headship has been rising during a pe- economically afford, not to be mar- riod of growth. This seeming paradox ried or remarried—can be expected to is resolved by the fact that other things be relatively well-off. Others—war or are changing across Africa. AIDS widows, separated or abandoned Changes in demographic and popu- women, and single mothers who have lation characteristics, social norms, not “chosen” headship but simply have education, and the nature of the family no options—are frequently found to all appear to be encouraging female head disadvantaged households. headship. An extra year of school- ing (for men or women) produces a 3 (continued on page 11) 2 World Bank ResearchDigest Growth and Capital Inflows in Africa Questions about the relationship The second endogeneity bias comes literature that FDI inflows are generally between growth and capital inflows from the omitted variables problem: good for growth. have become increasingly important we do not know whether growth and By contrast, sovereign borrowing capital inflows are both driven by a shows no impact on growth. This find- for Sub-Saharan Africa third factor, such as global growth or ing is interesting, because developing D global interest rates. It is reasonable countries are often praised for being oes economic growth in devel- to believe that global growth provides able to raise money from the interna- oping countries attract more capital to a developing country and at tional financial markets. Doing so is capital inflows? And do capital the same time helps boost the coun- considered a new way for countries to inflows in turn foster further economic try’s growth through other channels help finance their investment needs growth? These are important ques- (such as exporting). The literature on without depending on aid. But the tions for many developing countries. aid effectiveness is well aware of the authors’ finding casts doubt on this If the answer to both is yes, economic endogeneity problem and has tried to belief. It suggests that sovereign bor- growth and capital inflows could form address it, with mixed results. rowing might not be deployed most ef- a positive feedback cycle in which one In a recent paper Calderón and fectively. Unlike aid, which comes with reinforces the other. Nguyen revisit the causal relationship clear purposes and stricter monitoring, This issue is becoming important between economic growth and three capital raised through sovereign bor- for Sub-Saharan Africa in light of the types of capital inflows—aid, FDI, and rowing is usually more freely spent. region’s strong economic growth and sovereign borrowing—using a two-step The authors speculate that sovereign increasing capital inflows in recent approach to address reverse causality debt inflows could be more prone to years. Foreign direct investment (FDI) and omitted variables issues. corruption, vested interests, and irre- into the region has grown from less In the first step, using rainfall as an sponsible investment decisions. than 1 percent of GDP in the 1990s to instrument for growth, the authors find Overall, the findings provide evi- as much as 4 percent of GDP in recent that for a set of 38 developing coun- dence to support advocating for aid years. Meanwhile, sovereign debt in- tries in Sub-Saharan Africa, for the pe- to poor countries such as those in flows have grown from almost nothing riod 1979–2012, economic growth does Sub-Saharan Africa. Aid could flow in the early 2000s to about 1 percent not attract aid, FDI, or sovereign lend- directly toward poverty reduction, so- of GDP. ing. This result suggests that capital cial programs such as education and It is commonly believed that flows to Sub-Saharan Africa are driven health care, and infrastructure—all of longer-term inflows, such as FDI, aid, either by noneconomic factors (such as which benefit growth. By contrast, the or sovereign debt, have positive ef- political factors) in the recipient coun- finding on sovereign borrowing sug- fects on the receiving economy. Theory tries or by “push” factors in the invest- gests that poor countries might want suggests that this could be the case ing countries. In other words, the flows to scale back their access to private because long-term capital inflows help could be driven by economic condi- international financial markets. Low countries ease their capital constraints tions in the investing countries and institutional and investment manage- and also provide capital for invest- not by those in the recipient countries. ment quality in these countries could ment. FDI or aid inflows could also In the second step the authors find make greater access to private finance bring foreign know-how and technolo- that aid and, to a lesser extent, FDI counterproductive. gies or encourage better governance have a positive effect on economic and support human capital develop- growth. This is after controlling for ment. In addition, long-term capital global factors, commodity prices, and inflows could insulate countries from growth of Sub-Saharan African coun- the inherent volatility associated with tries’ main trade partners. short-term capital flows. Aid seems to matter most to Empirically, however, the relation- growth: a 1 percent increase in inflows ship between capital flows and growth of official development assistance is less clear because of the endoge- (ODA) increases growth by 0.025 per- neity issues. The first endogeneity centage points. This finding helps con- bias is reverse causality: we cannot firm the role of aid in assisting poor tell whether a positive relationship countries. between capital inflows and growth FDI inflows also seem to support César Calderón and Ha Nguyen. 2015. “Do means that capital inflows cause a country’s growth, though by a much Capital Inflows Boost Growth in Developing Coun- growth, that growth attracts capital smaller magnitude (about one-tenth as tries? Evidence from Sub-Saharan Africa.” Policy inflows, or that both phenomena are much as aid). This result is neverthe- Research Working Paper 7298, World Bank, simply happening at the same time. less consistent with the finding of the Washington, DC. World Bank Research Digest 3 Are Public Works Working in Malawi? Malawi’s public works program increase its effect on food security, through reduced labor supply or in- aims to increase food security potentially at the cost of investment in creased reservation wages. and the use of fertilizer. Evidence fertilizer. While the study can rule out The program offers households the significant improvements in the two suggests that it is falling short opportunity to earn about $22 at plant- outcomes specifically targeted by the L ing season and another $22 later in program—food security and the use abor-intensive public works the year (in a country with a per capita of fertilizer—there may have been programs are common social income of only $320). But results show small, diffuse increases in these or protection tools in low-income that it does not have a measurable other outcomes that are too small to settings. These programs require that short-term effect on lean season food detect. For example, changes in weekly beneficiaries work in order to receive security for participating (treated) spending (the interval captured in the a cash payment or in-kind transfer. households. Nor does rescheduling survey period) might be too small to Among the most widely known are the second work cycle from the harvest detect if households spread consump- the National Rural Employment season to the lean season generate tion across the program period (four Guarantee Act in India and the measurable improvements in food or eight months, depending on treat- Productive Safety Net Project in security. ment group) or saved for even longer Ethiopia. Such programs are wide- The failure of Malawi’s public works durations—especially since extra spread in Sub-Saharan Africa, though program to improve food security in spending may have been spread across on a smaller scale: 39 of 48 countries either the short run (through con- many different categories of goods. in the region have government- sumption support) or the longer run This interpretation allows for the pos- supported public works programs. In (through greater use of fertilizer) is sibility that the program was welfare- Malawi the program doubled in size troubling because it is the largest so- improving for households that chose in 2012 to cover about 500,000 house- cial protection scheme in one of the to participate and thus is consistent holds annually. world’s poorest countries. Public works with their revealed preferences, while Several studies of cash-for-work programs in other countries differ in still ineffective in achieving its main programs have focused on their poten- some elements of their structure—for policy objectives. tial effects in crowding out other work example, Ethiopia’s program offers The indirect effects of the public (with people working less on their own more income—and have been more works program are small and, sur- farms, for example) or on the extent effective. That 24 extra days of work prisingly, negative. In Northern and of self-targeting (which people choose during the lean season in Malawi do Central Malawi food security of un- to work at the wage rate offered). Yet not significantly improve food security treated households in participating there is limited evidence about how may be because of the low daily wage villages is not only lower than food such programs affect consumption in the public works program. Longer security among their treated neigh- and food security. This will depend on duration and more flexible schedules bors, it is also lower than food security such design features as the size of the are avenues for future investigation. in nonparticipating (control) villages. transfer, the season in which the pro- Malawi’s public works program This is in contrast to expectations and gram operates (an aspect associated is timed to coincide with the plant- to the effects of other large-scale trans- with opportunity costs especially in ing season to promote take-up of the fer programs (such as Oportunidades rural areas, where it may compete with country’s fertilizer subsidy scheme. But in Mexico). An explanation for this smallholder farming), and the mode the study’s results do not support the unexpected finding has proven to be of payment (whether a lump sum pay- hypothesis that the two programs are elusive. ment or multiple payments). complementary. While households par- Using a randomized controlled tri- ticipating in the public works program al, a recent study by Beegle, Galasso, are more likely to receive fertilizer and Goldberg evaluated Malawi’s coupons (consistent with the policy public works program. The study in- of linking this participation with the troduced two variants relative to the fertilizer subsidy) and thus pay less for standard model of the program as the fertilizer they use, they do not use implemented in 2012/13: the timing more fertilizer. and the payment schedule (lump sum The study found no evidence that or split payments). There were two lev- the program increased the ownership Kathleen Beegle, Emanuela Galasso, and Jessica els of randomization: across villages of durable goods. Nor did it find evi- Goldberg. 2015. “Direct and Indirect Effects of and across households in participating dence that the program affected prices Malawi’s Public Works Program on Food Secu- villages. The study examined whether by injecting cash into the economy, or rity.” Policy Research Working Paper 7505, World changes to the timing of the program that it led to labor market tightening Bank, Washington, DC. 4 World Bank ResearchDigest Vulnerability to Climate Change in Coastal Bangladesh In coastal regions of Bangladesh In densely populated, land-scarce studied species can be purchased in rising salinity from climate change Bangladesh poor households face dis- affordable quantities by the rural poor poses serious threats to the health advantages in accessing land and are and shared more equitably among forced to settle in low-lying regions household members, including women and livelihoods of poor people close to the coast. A poverty map de- and children. I veloped by the Bangladesh Bureau of Health risks from drinking water n a changing climate Bangladesh Statistics, the World Food Programme, salinity are also significant. Earlier re- is exceptionally vulnerable to sea- and the World Bank identifies a high search in coastal Bangladesh indicated level rise—because two-thirds of its incidence of poverty in the southwest links between drinking water salinity territory is less than 5 meters above coastal region, home to 9.9 million and maternal hypertension during sea level. Sea-level rise is virtually cer- poor people in 2010. Alarmingly, this pregnancy. In a third paper Dasgupta, tain to continue beyond 2100, even in figure includes 5.9 million extremely Huq, and Wheeler extended the analy- the most optimistic emissions scenar- poor individuals who cannot afford sis by studying the postnatal impact ios, so Bangladesh needs to anticipate even the basic-needs food expen- of prenatal salinity exposure in coastal potential impacts and plan for adapta- diture. Living conditions are harsh; Bangladesh. After controlling for many tion. Most research has focused on the the area is prone to tidal surges and other determinants, they found that long-run effects of progressive inun- cyclones; and residents have already saline water ingestion during the final dation from sea-level rise and losses experienced widespread inundation month of pregnancy significantly af- from heightened cyclone-induced and salinization of soil and water. GIS fects mortality risk for infants less than surges. Less attention has been paid overlays of poverty and salinity maps two months old. The estimated impact to salinity from saltwater intrusion, show that salinization is already affect- is comparable in magnitude to the along with its impact on livelihoods ing 2.5 million poor residents, includ- estimated effects of traditionally cited and adaptation alternatives, though ing 1.4 million who are extremely poor. determinants of infant mortality, such soil and water salinity monitors indi- By 2050 these numbers rise to 2.9 mil- as maternal age and education, gen- cate increasing salinization in many lion and 1.7 million in the best-case der of the household head, household coastal areas. A set of recent papers climate scenario, and 5.2 million and wealth, toilet facilities, drinking water explored issues relating to saltwater 3.2 million in the worst-case scenario. sources, and cooking fuels. intrusion in coastal Bangladesh, fo- Empirical research on the implica- Worldwide, about 600 million cusing on the effects on poor people. tions for livelihoods is progressing people inhabit low-elevation coastal In one paper Dasgupta, Kamal, as better data become available. In a zones that will be affected by progres- Khan, Choudhury, and Nishat assessed second paper Dasgupta, Huq, Mustafa, sive salinization. Recent research sug- the vulnerability of coastal Bangladesh Sobhan, and Wheeler studied the im- gests that the sea level may rise by to saltwater intrusion by 2050, using pact of salinization on the habitats of one meter or more in the 21st century, information from monitoring stations, fish species that are important sources which would increase the vulnerable hydrological models, and forecasts of of protein and market revenue for tens population to about one billion by rainfall, temperature, and sea-level of thousands of poor people in the 2050. Experience with salinization in rise for alternative climate change coastal region. Progressive saliniza- Bangladesh therefore offers a caution- scenarios. Their analysis focused on tion will affect freshwater fish species ary case from a global perspective. the southwest coastal region, which in several ways, including reproductive Many countries still lack systematic re- already has a salinization problem that cycles, reproductive capacities, spawn- search on the impacts of saltwater in- will only worsen as climate change ing areas, and migration patterns. The trusion on livelihoods and adaptation continues. Scientists have not reached study focused on 83 species consumed options. The estimated magnitudes of consensus on the timing and spatial by households in the region. these impacts in Bangladesh suggest impacts of climate change, so the Using species-specific salinity that such research is urgently needed study used alternative climate sce- tolerance ranges and poverty maps, in other poor countries with low-lying narios to predict low- and high-salinity the study estimated that salinization- coastal areas. trends from a March 2012 baseline. It induced species losses will outnumber Susmita Dasgupta, Mainul Huq, Md. Golam concluded that by 2050 salinization gains by a factor of six in high-poverty Mustafa, Md. Istiak Sobhan, and David Wheeler. of river water in coastal Bangladesh areas. In Bangladesh chronic and acute 2016. “The Impact of Aquatic Salinization on Fish will pose major risks to drinking wa- malnutrition levels are already higher Habitats and Poor Communities in a Changing ter quality, irrigation water for dry- than the World Health Organization’s Climate: Evidence from Southwest Coastal season agriculture, and coastal aquatic thresholds for public health emergen- Bangladesh.” Policy Research Working Paper ecosystems. cies. This finding therefore has po- 7593, World Bank, Washington, DC. Rising salinity has particularly se- tentially serious implications for food rious implications for poor people. security and malnutrition, since the (continued on page 7) World Bank Research Digest 5 Improving Agricultural Data for Better Policies By supporting better policies, better of agriculture, sometimes in collabora- The LSMS-ISA initiative offers a num- agricultural data could improve the tion with the national statistics office. ber of suggestions. Countries could lives of poor people, especially in Sample surveys such as farm surveys take a tremendous step forward if are a third source of agricultural data. they were to regularly collect reliable, Sub-Saharan Africa These surveys, the backbone of agri- nationally representative agricultural I cultural statistics in Africa, vary greatly data through a multitopic, multisec- n Sub-Saharan Africa, where most in content, frequency, and quality. In tor household survey instrument that people in extreme poverty live in many countries farm surveys are of- accounts for differences across indi- rural areas and are engaged largely ten complemented by other types of viduals within households. Another in agricultural activities, the devel- household surveys that capture some important improvement would be opment of agriculture, particularly data on agriculture, usually conducted the construction of comprehensive smallholder agriculture, remains a by the national statistics office or by libraries for all possible nonstandard critical driver of poverty reduction. Yet private firms. units in each specific region of a coun- despite the key role of agriculture, se- What are some of the current prob- try—something that should be car- rious weaknesses persist in the mea- lems with agricultural data? Lack of ried out to the extent possible for all surement of agricultural outcomes institutional links is a common issue; countries in Sub-Saharan Africa. And and in our understanding of the fac- in many countries the data collected as GPS technology becomes more ac- tors hampering growth in smallholder by the ministry of agriculture are curate, affordable, and user-friendly, agriculture. underutilized and are not analyzed GPS-based area measurement can Past investments and technical in conjunction with data available provide a practical alternative to self- assistance in the area of agricultural from the national statistics office or reporting by farmers. This approach is statistics have failed to produce sus- from other line ministries. In addi- being increasingly applied in surveys tainable systems, and existing sta- tion, the data often have limited policy worldwide. tistics continue to suffer from poor relevance because of their failure to Knowledge gaps in agricultural sta- quality, lack of relevance, and limited capture a comprehensive set of infor- tistics remain prevalent, and the chal- use in national policy dialogues. mation on rural households—informa- lenges ahead are daunting. Given the Further compounding the problem is tion that is crucial for understanding importance of agriculture in promoting that the poorest countries often have the links between agriculture and such growth and poverty reduction, improv- the poorest data. These countries, for aspects of well-being as health, nutri- ing the quality, availability, and policy which agriculture can be most criti- tion, and food security. relevance of agricultural data is of cal as a source of livelihood, are least Methodological standards for mea- paramount importance—particularly able to direct their limited resources suring agricultural productivity are par- for African countries, which lack funda- to improving the quality of agricultural ticularly weak. There are challenges in mental information to inform the de- statistics. quantifying both agricultural produc- sign of effective policies. Doing so will In a recent paper Carletto, Jolliffe, tion and land area. Those in quantify- require a concerted effort by individual and Banerjee address the challenge of ing agricultural production include dif- countries and stakeholders to develop improving agricultural statistics that ficulties in recall for extended-harvest and implement global standards and come from household surveys. Taking crops, use of nonstandard units, lack best practices in agricultural statis- a targeted and selective approach, of information on the state of the crop, tics. A strategy centered on improving they investigate a number of issues difficulties in valuing own production, methodologies, building capacity, and that are the focus of a recent World lack of adequate price data, and the strengthening institutions could lead Bank initiative, the Living Standards practice of intercropping, to name a to better-informed agricultural policies Measurement Study–Integrated few. Challenges in quantifying land with the potential to improve the lives Surveys on Agriculture (LSMS-ISA). area include known inaccuracies of of the millions of people involved in Where do agricultural data come self-reporting by farmers, current limi- the agricultural sector worldwide. from? They generally come from dif- tations of satellite imagery, and use ferent institutional sources, a situa- of nonstandard units. Another is the tion that typically results in conflict- time-consuming and costly nature of ing estimates. Virtually all countries techniques such as traversing (based have routine data systems based on on the use of compass and tape), often resident or local extension officers considered the gold standard for mea- Gero Carletto, Dean Jolliffe, and Raka Banerjee. employed by the ministry of agricul- suring land area. 2015. “From Tragedy to Renaissance: Improv- ture. A second source of agricultural How can we improve the ways that ing Agricultural Data for Better Policies.” Policy data is the agricultural census, also we collect agricultural data, particu- Research Working Paper 7150, World Bank, usually implemented by the ministry larly data on agricultural productivity? Washington, DC. 6 World Bank ResearchDigest Enhancing Transparency of Large-Scale Land Acquisition Reliable data on commercial farms Policy decisions on these as well as Beyond providing analytical in- could support policy actions aimed private investment will require regu- sights that are of interest in their at improving performance and lar access to up-to-date information. own right, having reliable data on Yet few countries have developed the outcomes by commercial farms can investment levels in the sector systems to regularly provide the data generate feedback loops for policy T needed. Even much of the empirical formulation. In Ethiopia these include he 2007/08 commodity price literature is still based on case studies a continued impetus to improve the boom triggered a global “land whose representativeness is difficult survey instruments used, efforts to rush,” with investors seeking to establish, often focusing on pro- better link to existing data (such as on to acquire agricultural land in Sub- cess rather than on quantification of investment licenses and survey plans) Saharan Africa. Nearly a decade outcomes. Lack of reliable and regular that are issued and maintained by later, there appears to be agreement information will make it difficult to the Investment Agency and regional on some stylized facts about this manage investment risks or country land authorities, and exploration of phenomenon. One is that weak or risks, enforce laws, and document techniques (such as yield monitoring fragmented institutions compromised compliance with global standards so based on remotely sensed data) that countries’ ability to channel this as to attract capable and responsible could help complement, cross-check, demand toward areas where it would (institutional) investors. and eventually even partially replace yield the highest returns or to reject In a recent paper Ali, Deininger, and survey information. nonviable proposals from inexperi- Harris explore ways of satisfying such While the descriptive evidence enced investors. As a consequence, information needs sustainably and ef- reported in the study only scratches the benefits have been less than were fectively. In doing so, they draw on a the surface, it provides an interesting expected and a sizable share of inves- nationally representative large-farm basis for future research. For example, tors either went out of business or survey in Ethiopia conducted between the commercial-farm survey could be failed to fully utilize all the land that 2010/11 and 2013/14. Ethiopia is an linked more directly to ongoing data was allocated to them. interesting case because it has a long collection efforts in the smallholder Another of these stylized facts is tradition of collecting systematic data sector to look at the interaction be- that while the demand for land has on the performance of large (state) tween large farms and neighboring retreated from the 2008 levels, it is farms—data that, because of poor smallholders—particularly the poten- expected to continue, though at levels quality, were often not reported or tial spillover effects and the channels much lower than those observed at whose collection was even intermit- through which they materialize. Also the height of the land rush. Indeed, if tently stopped. interesting would be to investigate guided by a coherent and enforceable Data collected with an improved the determinants of entry and exit for policy, responsible agricultural invest- survey instrument allow the authors to commercial farms and, more generally, ment could provide countries whose draw interesting conclusions on com- the dynamics of firm performance over economy depends on agriculture with mercial farming in the country. First, time. opportunities to add value and gener- even at the peak of the land rush the ate local benefits. vast majority of land transferred to To realize these benefits will re- investors in at least partly operational quire policy decisions on a number of commercial farms went to Ethiopians issues, including the following: First, rather than to foreigners. Second, where in the value chain—upstream about 55 percent of the land trans- in agroprocessing, in mixed nucleus ferred remains unutilized, with the estate models with outgrowers, or fully main constraints to expansion relating own production—investment would be to technology and labor. Third, with most desirable and what complemen- one permanent job per 20 hectares tary public inputs may encourage such plus some temporary jobs, commercial investment. Second, how well land farms fail to generate much employ- that has been transferred to investors ment. Finally, for most crops the yields is utilized and, if poorly, what reme- of commercial farms are roughly twice dial action (for example, canceling li- those of smallholders; however, the censes) may be needed or appropriate. highest yields are normally obtained Daniel Ali, Klaus Deininger, and Anthony Har- And third, how performance compares by those in the 10- to 20-hectare cat- ris. 2015. “Using National Statistics to Increase between local producers and outside egory, which often also managed to Transparency of Large Land Acquisition: Evidence investors and what this implies for expand their cultivated area quite from Ethiopia.” Policy Research Working Paper regulating the sector. rapidly. 7342, World Bank, Washington, DC. World Bank Research Digest 7 Explaining the Gender Gap in Agricultural Productivity Women play a vital part in Ugandan of this unexplained gap is attributed Programs designed to encour- agriculture, yet have much lower to differential returns to the child age women to plant high-value cash productivity than men. What could dependency ratio and one-fifth to dif- crops could also be explored. But such ferential returns to transport access. programs should carefully assess the help close this gap? This implies that women’s greater risk that male farmers might claim a W child-care responsibilities and their stake in plots on which cash crops are omen make up about half greater difficulty accessing input and planted, resulting in heavy leakage of the agricultural labor force output markets from areas without benefits from female to male farmers. in Sub-Saharan Africa. But transport are the largest drivers of the Such programs also should carefully evidence shows that plots managed by gap. Others include differences in the ensure access to adequate output mar- women are 20–30 percent less produc- uptake of cash crops and differences in kets before the switch to a new crop tive on average than those managed the uptake of and returns to improved and provide complementary inputs by men. This agricultural productivity seeds and pesticides. and extension services critical to its gap contributes to income inequality These findings have far-reaching success. between women and men. In some policy implications. One implication is Female farmers’ limited use of hired cases the productivity gap is also that low-cost interventions designed to labor provides weak evidence that accompanied by an inefficient over- ease child-care constraints on female they face constraints in hiring workers. allocation of inputs to male-managed plot managers have the potential to Evaluating projects that provide agents plots, resulting in large aggregate substantially improve equality in the to help women farmers find, supervise, productivity losses for the agricultural distribution of resources between male and finance payment of hired labor sector. and female managers within Ugandan could offer useful insight. Estimates from the Food and households. Community-based child- Agriculture Organization of the United care interventions are one possibility, Nations have shown that closing the though there is little existing evidence Daniel Ali, Derick Bowen, Klaus Deininger, and agricultural productivity gap globally of their efficacy. Marguerite Duponchel. 2015. “Investigating the could increase agricultural output in Given the disproportionate travel Gender Gap in Agricultural Productivity: Evidence lower-income countries by 2.5–4 per- costs faced by female farmers, inter- from Uganda.” Policy Research Working Paper cent—enough to reduce the number ventions that bring extension services 7262, World Bank, Washington, DC. of people who are undernourished by closer to dwellings, provide access to 12–17 percent, or by 100–150 million. market information through mobile Thus the size of a country’s gender gap phones, or provide better access to in agricultural productivity, and the transport may help close the produc- factors that drive it, are of great inter- tivity gap. Another possibility would est to policy makers. be to promote use of existing women’s Using national data from the groups for commercial purposes, such Uganda National Panel Survey for as collective access to input and out- 2009/10 and 2010, a recent paper by put markets. Ali, Bowen, Deininger, and Duponchel The use of inputs (improved seeds, estimates the gender gap in agricultur- chemical fertilizer, and pesticides) is (continued from page 4) al productivity in Uganda. The analysis not widespread enough to contribute also unpacks this gap into portions substantially to the aggregate pro- Vulnerability to Climate Change accounted for by differences in house- ductivity gap. But the extremely large hold, plot-manager, and parcel charac- difference between male and female in Coastal Bangladesh teristics and in returns to these charac- farmers in the use of physical inputs, teristics. Overall, the results show that and the statistically significant con- Susmita Dasgupta, Mainul Huq, and David output per unit of land is 17.5 percent tribution of such inputs to the pro- Wheeler. 2015. “Drinking Water Salinity and lower on female-cultivated than male- ductivity gap, suggest that programs Infant Mortality in Coastal Bangladesh.” Policy cultivated plots. to encourage their adoption should Research Working Paper 7200, World Bank, Although men have greater access incorporate elements to ensure equal Washington, DC. to inputs, input use is so low and in- adoption by male and female farm- Susmita Dasgupta, Farhana Akhter Kamal, verse returns to plot size so strong in ers. For example, programs could be Zahirul Huque Khan, Sharifuzzaman Choudhury, Uganda that smaller female-managed explored that finance the purchase of and Ainun Nishat. 2014. “River Salinity plots have a net endowment advantage inputs through such mechanisms as and Climate Change: Evidence from Coastal of 12 percent, revealing a larger unex- vouchers, loans, or transfers that are Bangladesh.” Policy Research Working Paper plained gap of 29.5 percent. Two-fifths aligned with farmers’ cash flow cycle. 6817, World Bank, Washington, DC. 8 World Bank ResearchDigest Changing Patterns of Growth and Poverty Reduction in India As the structure of India’s economy if it is measured using the national overall contribution to the decline in has changed, urban growth has accounts. poverty has also dwindled. Whereas become the new engine of poverty Did the changing pattern of India’s before 1991 the primary sector ac- growth affect the pace of poverty de- counted for about two-fifths of the reduction in the country cline—and if so, in what way? The total decline in poverty, after 1991 its I authors’ research shows that during contribution fell to less than 10 per- ndia’s economic take-off during the the past two decades the poor gained cent of the total—and larger—decline 1990s and early 2000s is now part more from urban than from rural in poverty. of the country’s economic folklore. growth, marking a change in the earlier At the same time, the contribution After 1991 per capita income grew relationship between the pattern of of the other sectors to poverty reduc- nearly two and a half times as fast in growth and poverty reduction. Before tion has risen substantially. While real terms as in the preceding three 1991 rural growth largely determined growth in the secondary sector (mainly and a half decades. During this time poverty reduction in the country be- manufacturing and construction) ac- poverty also fell more quickly. But did cause traditionally the vast majority counted for about 25 percent of the the faster pace of poverty reduction of poor Indians lived in rural areas fall in poverty after 1991, the tertiary after 1991 simply reflect faster growth, and depended on the rural economy sector alone (mainly services) con- or did poverty become more respon- for sustenance. While urban growth tributed more than 60 percent of the sive to growth? And did the changing reduced urban poverty, it contributed decline. Since 2000 India’s construc- pattern of India’s growth matter? little to national poverty reduction. tion boom—which has made intensive Datt, Ravallion, and Murgai investi- Since the early 1990s, however, use of low-skilled labor—has helped gate these questions in a recent paper. this pattern has undergone a strik- secondary sector growth become more The authors begin with the question ing change. Urban growth has now pro-poor. of whether India’s growth has become emerged as a major driver of national In short, the Indian economy is more pro-poor. One measure of the poverty reduction. Since 1991 urban changing, and so is the relationship extent to which growth is pro-poor is growth has been responsible for about between economic growth and poverty the elasticity of poverty reduction to 80 percent of the total decline in pov- reduction in the country. The process growth—in other words, by how much erty. This happened directly, through of structural transformation of the poverty declines for every 1 percent urban growth having a larger impact economy has intensified, displacing increase in per capita income or con- on urban poverty—but even more im- the traditional sources of both eco- sumption. Thus the question of wheth- portant, it also happened indirectly, nomic growth and poverty reduction. er poverty in India has become more through urban growth having a sub- As this process continues, the country responsive to growth can be rephrased stantial impact on rural poverty. This can be expected to increasingly turn as whether the elasticity of poverty re- indicates that the growth of urban to growth in its urban and nonagri- duction to growth has increased. areas, which have both bigger popula- cultural sector to drive future poverty It turns out that the answer to this tions and higher productivity, has been reduction. question depends to some extent on good for poverty reduction as a whole how growth is measured. If it is mea- in India. sured by mean changes in per capita The authors next investigate which consumption derived from household sectors have emerged as the primary surveys, there is strong evidence that drivers of India’s growth. Before 1991 growth in the post-1991 period was rural growth, especially in the farm not only faster but also more pro-poor: sector, mattered most for poverty the elasticity of the headcount index reduction. But in recent times it has to growth increased from 1.5 (before become more difficult to attribute pov- 1991) to 2.7 (after 1991). If growth is erty reduction to any specific sector. measured by per capita income or con- Since 1991 all sectors have contrib- sumption from the national accounts, uted to reducing poverty. Indeed, with the evidence still points to a higher the greater integration of the Indian elasticity for the headcount index after economy, growth in one sector has 1991. For poverty measures that take begun to transmit gains elsewhere to the depth or severity of poverty into a greater extent than before, and the account, the evidence is mixed: higher imbalance in the growth process has Gaurav Datt, Martin Ravallion, and Rinku Mur- elasticity of poverty reduction after ceased to matter. gai. 2016. “Growth, Urbanization, and Poverty 1991 holds only if growth is measured Because the primary sector (mainly Reduction in India.” Policy Research Working using the household surveys, but not agriculture) has declined in size, its Paper 7568, World Bank, Washington, DC. World Bank Research Digest 9 Land Market Restrictions and Rural Labor Markets Restricting land sales can slow the costs of migration out of villages channels. If labor is mobile, rural the shift of labor out of agriculture, due to inability to sell the land. The wages will remain largely unaffected by reduce rural wages, and increase increase in migration costs results land market restrictions if either col- from the fact that a household loses its lateral value or insecurity of property women’s labor force participation income stream from the land when it rights is the main channel of causa- M decides to leave the village. The papers tion. In contrast, wages will be lower in any developing countries develop theoretical models to investi- areas with a higher proportion of land impose restrictions on sales gate the effects of increasing migration under restrictions if migration costs and rentals of agricultural costs arising from land market restric- are the primary channel through which land. One rationale for doing so is to tions on structural shifts from agricul- the restrictions affect the rural labor promote own farming: self-farming is ture to nonagriculture and on female market and the structural transforma- considered conducive to increasing labor force participation and wages. tion of employment. agricultural productivity, since the In the case of the structural trans- To provide empirical evidence on owner of a farm has a self-interest in formation of rural employment, the the effects of land market restrictions, conserving the quality of the land and theoretical model finds that land Emran and Shilpi take advantage of investing in its improvement. Another market restrictions reduce rural-urban a historical natural experiment in Sri rationale is to guard against landless- migration and increase the employ- Lanka, where malaria played a unique ness—reflecting a concern that in ment share of agriculture—under the role in land policy during precolonial the face of income and consumption assumptions that agriculture is more times. After people fled malaria- shocks, poorer farmers may be forced labor intensive than nonagriculture afflicted areas, the government took to sell their land at unfavorable prices, and that demand for locally produced ownership of abandoned land under depriving them of their most valuable nonfarm goods and services is income the Crown Lands Ordinance of 1840. asset. inelastic. Under these two assump- After Sri Lanka’s independence in Emran and Shilpi examine the tions, sales restrictions in land mar- 1948, these lands were distributed longer-term consequences of land kets may result in a “reverse structural among households under settlement market restrictions on rural labor mar- change” in the rural economy, with the programs following the imposition kets in two recent papers. The authors labor share in agriculture increasing in of sales and rental restrictions. The address two main questions: Do land response to the restrictions. identification strategy relies on the market restrictions affect the struc- In the case of female labor force fact that malaria was eradicated from tural transformation from agriculture participation, the theoretical model Sri Lanka by 1951 and that eradication to nonagriculture (manufacturing and shows that women face higher migra- was made possible by an exogenous services) in rural areas? And how do tion costs than men because of their breakthrough in technology (the in- these restrictions affect rural labor higher productivity in the provision of troduction of the pesticide DDT). Both market outcomes for women? Both home goods (child and family care). As papers thus rely on the incidence of issues have important bearings on a consequence, female participation in malaria more than half a century ago economic welfare and development. the village labor force can increase as a to identify the effects of land market Recent studies have found that labor result of land market restrictions. restrictions. The strength of the identi- productivity is significantly higher in Considering other channels through fication scheme also derives from the nonagricultural activities than in agri- which land market restrictions work, fact that, unlike China and Vietnam, culture and concluded that a structural the authors argue that the “reverse Sri Lanka imposes no formal restric- transformation from agriculture to structural change” can also occur if the tion on the geographic mobility of nonagriculture is desirable for eco- restrictions destroy the collateral value households. nomic growth and poverty reduction. of land and thus act as impediments The empirical results show that If land market restrictions tie the rural to entrepreneurship in nonagricultural land market restrictions significantly labor force, particularly women, to low- activities. How lower collateral value reduce the probability of participation paid agricultural jobs, they may lead to affects female labor force participa- in nonagricultural activities: starting adverse poverty outcomes—contrary tion depends on the substitutability from its mean value, a one percent- to their policy objectives. of labor and capital. If land market re- age point increase in the share of Many economic studies show that strictions lead to insecurity of property land under restrictions reduces this land market restrictions can affect eco- rights, they can decrease female labor probability by 1.38 percent. At a more nomic outcomes through their effects force participation, though the effects disaggregate level, the adverse effects on the insecurity of property rights on employment transformation are are more pronounced for employment and the collateral value of land. The ambiguous. papers by Emran and Shilpi focus on Both theoretical models indicate a different mechanism: an increase in a way to discriminate among different (continued on page 11) 10World Bank ResearchDigest Deforestation Prevention Programs and Community-Managed Forestry Evidence from Nepal suggests that how such contracts should be struc- example, CF households that believe community-managed forests can be tured. For example, would community they have equitable access to their an effective part of deforestation members prefer that REDD+ payments CF community forest funds are more go to forest user groups for commu- likely to accept REDD+ contracts. But prevention programs like REDD+ nity projects or to households? What respondents are less likely to accept M direct-use ecosystem services are most REDD+ contracts if they believe that ost deforestation today is costly for villagers to give up? And how village authorities engage in more for- occurring in developing coun- does community-level enforcement af- est monitoring (in contrast to outside tries. But the consequences fect acceptance of REDD+ contracts? enforcement) and if they think that are global, with forest degradation In a recent study Dissanayake those authorities do not appropri- accounting for 11–20 percent of and nine coauthors attempted to ately enforce CF rules and regulations. annual greenhouse gas emissions. fill the gaps in knowledge through a These results suggest that good gover- Curbing deforestation and forest choice experiment carried out in rural nance—including ensuring equitable degradation is believed to be a highly Nepalese communities. The goal of access to CF funds, preventing the cost-effective way to address climate this experiment was to understand misuse of funds, and ensuring proper change and also support adaptation. respondents’ preferences on the insti- monitoring of forests—can support the One approach aimed at reducing tutional structure of REDD+ contracts. REDD+ program by reducing costs. deforestation and forest degradation Specific objectives included identifying Using the results, the study cal- in developing countries is to provide household preferences on the attri- culated the opportunity costs of the payments for ecosystem services— butes of REDD+ contracts, estimating REDD+ contract obligations. It esti- such as through the REDD+ (Reducing the opportunity costs faced by house- mated the costs per ton of carbon at Emissions from Deforestation and holds when participating in REDD+, $24.20 for CF communities and $17.90 Forest Degradation) program created and evaluating the impact of belong- for non-CF communities. These esti- under the United Nations Framework ing to legally sanctioned community- mates are consistent with findings in Convention on Climate Change. In managed forest groups on these the current literature that the oppor- addition to sequestering carbon, preferences and opportunity costs. To tunity cost of carbon sequestration in REDD+ may bring important benefits address the last objective, the study community forests is low compared to developing country households split the sample into households that with other abatement options. But the and communities through poverty are members of the legally recognized estimates are higher than many of the reduction. Nepal Community Forestry Program other estimates for the avoided costs About 18 percent of forests world- (CF) and those that are not part of that of deforestation, which are typically wide, and 25 percent of those in de- program (non-CF). based on engineering or land use mod- veloping countries, are controlled by In both CF and non-CF communi- els and do not account for community- communities either by law or in prac- ties the study found that households and household-level opportunity costs. tice, and this share is rapidly increas- prefer higher REDD+ payments and The study shows that community- ing with ongoing forest decentraliza- would rather not take on REDD+ obli- controlled forests, at least in Nepal, tion. The effectiveness of programs like gations or reduce firewood collection can be an effective part of the REDD+ REDD+ in preventing deforestation and grazing without adequate com- program. When implementing the con- is linked to their adoption by com- pensation. On the key policy question tracts, however, it is vital to account for munities with community-controlled of how to divide up REDD+ payments, proper enforcement and governance. forests. For these communities, the the study found that respondents Most important, given the relatively decision to accept REDD+ contracts prefer that more of the payments go high opportunity costs, REDD+ deals depends on incentives, benefit-sharing to communities rather than to house- may need to be more remunerative arrangements, the opportunity costs of holds. This result indicates a high de- than previously envisioned. carbon sequestration, the allocation of gree of trust in forest user group com- forest management decision-making munities, mirroring the study’s focus rights, and community interactions. group findings. As REDD+ contract Yet there is limited evidence-based requirements become more stringent, Sahan T. M. Dissanayake, Prakash Jha, Bhim knowledge about community and respondents would like to be sure Adhikari, Rajesh Bista, Randall Bluffstone, Har- isharan Luintel, Peter Martinsson, Naya Sharma household preferences and the oppor- that their households receive more Paudel, E. Somanathan, and Michael Toman. tunity costs of carbon sequestration in direct benefits in exchange for those 2015. “Community Managed Forest Groups and developing countries in the context of sacrifices. Preferences for REDD+ Contract Attributes: A community-managed forests for initia- The study found that community Choice Experiment Survey of Communities in Ne- tives such as the REDD+ program. Very dynamics play a very important part in pal.” Policy Research Working Paper 7326, World basic knowledge is needed even about the adoption of REDD+ contracts. For Bank, Washington, DC. World Bank Research Digest11 (continued from page 1) Changes in Poverty and Female-Headed Households in Africa So what has happened to the living a male adult household member, decline in poverty despite their smaller standards of female-headed house- and the same for nonmarried heads. overall share in the population. holds in the aggregate? The authors And the finding that poverty is falling But why has poverty fallen faster for examine this question by calculating faster for female-headed households female-headed households? Perhaps country-specific changes in the head- is robust to allowing for the generally poor female-headed households face count index of poverty based on real smaller size of such households and relatively high economic returns to household per capita consumption economies of scale in consumption. the new opportunities unleashed by expenditures, in 2005 dollars adjusted The living standards of the different growth. Perhaps they have benefited for purchasing power parity, separately types of female-headed households disproportionately from the expan- for male- and female-headed house- followed dissimilar paths across coun- sion of social protection in Africa. Or holds. Spells of comparable survey tries and periods with no one type perhaps the group of people living in pairs for the same country allow this to consistently outperforming the oth- female-headed households is funda- be done for 27 spells and 24 countries ers, yet with at least one type usually mentally changing over time. A super- that account for about 80 percent of outperforming male-headed house- ficial examination does not support Africa’s current population. holds. There is little discernible pat- any of these explanations, but this new The results show that poverty de- tern across countries. One category of stylized fact about poverty in Africa clined for both groups of households, female-headed households does well warrants a closer look. but in most countries it fell faster for in one country or period while another female-headed households. This is category does best elsewhere. also true when one allows for the di- A decomposition of the change versity among female-headed house- in poverty indicates that rather than Annamaria Milazzo and Dominique van de Walle. holds—for example, comparing house- putting a brake on poverty reduction, 2015. “Women Left Behind? Poverty and Head- holds with widowed and nonwidowed female-headed households are con- ship in Africa.” Policy Research Working Paper heads, married heads with and without tributing appreciably to the overall 7331, World Bank, Washington, DC. (continued from page 9) Land Market Restrictions and Rural Labor Markets in manufacturing and services, with a greater burden of land market restric- costs in rural areas—both through a particularly large negative effect for tions on older women and a much slower transformation from agricul- services. An increase in the land under smaller effect on men. tural to nonagricultural activities and restrictions has a positive effect on the Emran and Shilpi also estimate through lower wages and lower per probability of being employed in ag- the effect of land market restrictions capita consumption. The burden of the ricultural wage work, but no effect on on overall wages (for both men and restrictions falls disproportionately on the probability of being self-employed women) and per capita household older women who take up the work left in nonagricultural activities. Land consumption. The results show signifi- behind by men and younger women. market restrictions thus hold back the cant negative effects on both. A one These longer-term costs should be structural change from agricultural to percentage point increase in the share weighed against possible benefits nonagricultural employment in a rural of land under restrictions reduces per before imposing restrictions on land economy. capita consumption in a village by sales and rentals. In investigating the effects on labor 0.084 percent. Overall, the evidence on market outcomes for women, Emran wages in both papers indicates that and Shilpi find that their labor force the effects of land market restrictions participation increases in response to work primarily through the migration M. Shahe Emran and Forhad Shilpi. 2015. “Do land market restrictions: a one per- cost channel. Land Market Restrictions Hinder Structural centage point increase in the share of While concerns about agricultural Change in a Rural Economy? Evidence from Sri land under restrictions leads to about productivity and poverty among farm- Lanka.” Policy Research Working Paper 7525, a 2.3 percent increase in women’s la- ers often underlie the restrictions im- World Bank, Washington, DC. bor force participation. But this same posed on land sales and rentals, the ———. 2015. “Land Market Restrictions, Wom- increase in the land under restrictions studies by Emran and Shilpi show that en’s Labor Force Participation, and Wages in a leads to a 1.7 percent decrease in the by increasing migration costs, these Rural Economy.” Policy Research Working Paper female wage. The results also suggest restrictions can impose significant 7524, World Bank, Washington, DC. 12World Bank ResearchDigest Recent Policy Research Working Papers on IDA Countries 7164 Promoting Handwashing and Sanitation: 7296 Preferences for REDD+ Contract Attributes in 7416 Oral Democracy and Women’s Oratory Evidence from a Large-Scale Randomized Low-Income Countries: A Choice Experiment Competency in Indian Village Assemblies: A Trial in Rural Tanzania in Ethiopia Qualitative Analysis Bertha Briceño, Aidan Coville, and Sebastian Sahan T. M. Dissanayake, Abebe Damte Paromita Sanyal, Vijayendra Rao, and Umang Martinez Beyene, Randall Bluffstone, Zenebe Prabhakar 7183 Short-Term Impacts of Formalization Gebreegziabher, Peter Martinsson, Alemu 7419 Pronatal Property Rights over Land and Assistance and a Bank Information Session on Mekonnen, Michael Toman, and Ferdinand Fertility Outcomes: Evidence from a Natural Business Registration and Access to Finance M. Vieider Experiment in Ethiopia in Malawi 7305 Firms’ Locational Choice and Infrastructure Daniel Ayalew Ali, Klaus Deininger, and Francisco Campos, Markus Goldstein, and Development in Tanzania: Instrumental Niels Kemper David McKenzie Variable Spatial Autoregressive Model 7426 Changing Wage Structure in India in the 7219 Advanced-Country Policies and Emerging- Atsushi Iimi, Richard Martin Humphreys, and Post-Reform Era: 1993–2011 Market Currencies: The Impact of U.S. Sevara Melibaeva Hanan G. Jacoby and Basab Dasgupta Tapering on India’s Rupee 7306 Crop Choice and Infrastructure Accessibility 7435 Formalizing Rural Land Rights in West Africa: Yuki Ikeda, Denis Medvedev, and Martin in Tanzania: Subsistence Crops or Export Early Evidence from a Randomized Impact Rama Crops? Evaluation in Benin 7241 Are Women Less Productive Farmers? How Atsushi Iimi, Richard Martin Humphreys, and Markus Goldstein, Kenneth Houngbedji, Markets and Risk Affect Fertilizer Use, Sevara Melibaeva Florence Kondylis, Michael O’Sullivan, and Productivity, and Measured Gender Effects in 7334 Quality and Accountability in Healthcare Harris Selod Uganda Delivery: Audit Evidence from Primary Care 7440 The Impact of Violence on Individual Risk Donald F. Larson, Sara Savastano, Siobhan Providers in India Preferences: Evidence from a Natural Murray, and Amparo Palacios-López Jishnu Das, Alaka Holla, Aakash Mohpal, Experiment 7244 Household Responses to Shocks in Rural and Karthik Muralidharan Pamela Jakiela and Owen Ozier Ethiopia: Livestock as a Buffer Stock 7347 Gold Mining and Proto-Urbanization: Recent 7466 Quantifying Spillover Effects from Large Farm Daniel Ayalew Ali Evidence from Ghana Establishments: The Case of Mozambique 7278 Firm Productivity and Infrastructure Costs in Marcel Fafchamps, Michael Koelle, and Klaus Deininger, Fang Xia, Aurelio Mate, and East Africa Forhad Shilpi Ellen Payongayong Atsushi Iimi, Richard Martin Humphreys, and 7370 Promoting Democracy in Fragile States: 7473 Households or Locations? Cities, Catchment Sevara Melibaeva Insights from a Field Experiment in Liberia Areas and Prosperity in India 7279 Firms’ Locational Choice and Infrastructure Eric Mvukiyehe and Cyrus Samii Yue Li and Martin Rama Development in Rwanda 7384 The Impact of Secondary Schooling in Kenya: 7551 Managing Food Price Volatility in a Large Atsushi Iimi, Richard Martin Humphreys, and A Regression Discontinuity Analysis Open Country: The Case of Wheat in India Sevara Melibaeva Owen Ozier Christophe Gouel, Madhur Gautam, and Will 7280 Firm Inventory Behavior in East Africa 7394 Can Improved Biomass Cookstoves J. Martin Atsushi Iimi, Richard Martin Humphreys, and Contribute to REDD+ in Low-Income 7557 The Distribution of Consumption Expenditure Sevara Melibaeva Countries? Evidence from a Controlled in Sub-Saharan Africa: The Inequality among 7281 Agriculture Production and Transport Cooking Test Trial with Randomized All Africans Infrastructure in East Africa: An Application Behavioral Treatments La-Bhus Fah Jirasavetakul and Christoph of Spatial Autoregression Abebe D. Beyene, Randall Bluffstone, Sahan Lakner Atsushi Iimi, Liangzhi You, Ulrike Wood- Dissanayake, Zenebe Gebreegziabher, Peter 7646 Decomposing Response Errors in Food Sichra, and Richard Martin Humphreys Martinsson, Alemu Mekonnen, and Michael Consumption Measurement: Implications for 7285 Smallholders’ Land Ownership and Access in Toman Survey Design from a Survey Experiment in Sub-Saharan Africa: A New Landscape? 7411 Recasting Culture to Undo Gender: A Tanzania Klaus Deininger, Fang Xia, and Sara Sociological Analysis of Jeevika in Rural Jed Friedman, Kathleen Beegle, Joachim De Savastano Bihar, India Weerdt, and John Gibson 7290 Costs and Benefits of Land Fragmentation: Paromita Sanyal, Vijayendra Rao, and Shruti 7665 Short-Term Effects of India’s Employment Evidence from Rwanda Majumdar Guarantee Program on Labor Markets and Daniel Ayalew Ali, Klaus Deininger, and 7412 Job Opportunities along the Rural-Urban Agricultural Productivity Loraine Ronchi Gradation and Female Labor Force Klaus Deininger, Hari K. Nagarajan, and 7295 The Economic Viability of Jatropha Biodiesel Participation in India Sudhir K. Singh in Nepal Urmila Chatterjee, Rinku Murgai, and Martin Govinda R. Timilsina and Ujjal Tiwari Rama Working Papers can be downloaded at http://econ.worldbank.org To download the World Bank Research E-Newsletter, go to http://econ.worldbank.org/research_newsletter The World Bank Research Digest is a quarterly publica- The Research Digest is financed by the Bank’s Editorial Committee: Indermit S. Gill (managing editor), tion disseminating findings of World Bank research. Research Committee and managed by DECDP, the Aslı Demirgüç-Kunt, and Shiva S. Makki. Editor: Alison The views and interpretations in the articles are those research support unit of the Development Economics Strong; production: Roula Yazigi. For information or of the authors and do not necessarily represent the Senior Vice Presidency (DEC). The Research Digest is free subscriptions, send email to research@worldbank. views of the World Bank, its Executive Directors, or the not copyrighted and may be reproduced with appropri- org or visit http://econ.worldbank.org/research_digest. countries they represent. ate source attribution. The World Bank 1818 H Street, NW Washington, DC 20433, USA Printed on Recycled Paper