Reducing Barriers to Saving in Malawi

On average, developing countries have fewer than 20 bank branches per 100,000 adults, and people deposit money at a rate one-third of that in developed countries.[i] This lack of formal financial services, along with many other factors, may inhibit farmers and other entrepreneurs, particularly in rural areas, from increasing savings and investments, and smoothing household consumption. Financial services could help farmers to accumulate funds to purchase tools such as fertilizer which are helpful for increasing production. If barriers to financial services are reduced or eliminated by offering enhanced savings products, what is the impact on the use of different agricultural inputs, farm output, and overall well-being in rural farming households?


Tobacco is one of Malawi’s primary exports, employing many of the country’s farmers. Income volatility influenced by macroeconomic forces can be particularly harmful to those farmers living near the poverty line, causing households to skip meals and forgo necessary healthcare expenses.

Opportunity International, an international NGO, opened the Opportunity Bank of Malawi (OBM) in 2002 with a license from the Central Bank of Malawi. OBM provides financial services to the rural poor and has partnered with researchers and two private agricultural buyers, Alliance One and Limbe Leaf, to offer enhanced savings products to tobacco farmers.

Description of Intervention:

The study assessed the impact of OBM’s savings programs on the behavior and well-being of local farmers. Farmers were organized in farmers clubs, with an average of 10-15 members, by one of the agricultural buyers. In exchange for group loans in the form of fertilizer and extension services, administered by OBM, the club allowed the commercial buyer to make the first offer on the national auction floor, essentially creating an exclusive relationship. Farmer clubs in this sample were randomly assigned to one of two savings account treatment groups or a comparison group. Clubs in the comparison group received information about the benefits of having a formal savings account. Clubs in the treatment groups received the same information about savings accounts and were also offered individual savings accounts into which proceeds, after loan repayment, would be directly deposited.

Farmers in the first treatment group were offered an “ordinary” savings account with an annual interest rate of 2.5%. Those in the second treatment group received the same individual savings account, in addition to a “commitment” savings account which allows farmers to specify an amount of money to be frozen until a specified date (e.g. immediately prior to the planting season, so that funds are preserved for farm input purchases).

To assess the impact of public information on financial behavior, farmer clubs in both treatment groups were randomly assigned to one of three raffle schemes providing information about club level savings. Raffle tickets to win a bicycle were distributed to participants on two occasions based on savings balances as of two pre-announced dates. One third of farmers received raffle tickets in private, one third received tickets in public when names and numbers of tickets were announced to the club, and one third was ineligible for the raffle.

Results and Policy Lessons:

Savings Behavior: Twenty-one percent of farmers who were offered a commitment savings product (no raffle), made transfers to their account, while 16% of farmers who were offered the ordinary savings product (no raffle) had their harvest proceeds directly deposited into their individual account, and no farmers in the comparison groups received funds directly in an OBM account. Overall, farmers in the six treatment groups deposited substantial amounts into their individual bank accounts; among farmers who were offered the commitment savings account, most of these deposits were made into the ordinary savings account.

Farmers in the commitment savings group had higher net savings during the pre-planting period, and the commitment savings treatment group overall withdrew more money during the planting season. This finding implies that these farmers were better able to save money and delay consumption until the lean season when food supplies from the last harvest were scarcer. Farmers in the ordinary savings group did not experience an increase in net savings during the pre-planting season, or an increase in withdrawals during the planting season, suggesting they were not able to smooth consumption as effectively.

Inputs, crop sales, and expenditures: In relation to those in the comparison group, farmers who were offered commitment savings accounts had more land cultivated, higher value of inputs, and greater value of harvest at a statistically significant level. These commitment savings farmers cultivated .33 more acres of land (compared to an average of 4.3 acres of land in the comparison group) and used 17.1% more inputs. This increase in land under cultivation and inputs used by the commitment savings group led to a 20.1% increase in value of crop output above the levels in the comparison group. Finally, farmers in the commitment treatment group increased total expenditures reported in the last 30 days by 13.5%. Overall, farmers in the ordinary savings group did not have outcomes that were different from those in the comparison group at a statistically significant level.

Evidence suggests that the positive results in the commitment savings group were not due to helping farmers solve self-control problems since most money accrued in ordinary savings accounts and actual commitment account balances were low. There was also no direct evidence that the results were derived from farmers keeping funds from their social networks. Psychological phenomena such as mental accounting may be behind the impact of the commitment accounts. However, the current study does not empirically test this hypothesis and psychological mechanisms are addressed in future research. Results from the public and private raffle treatments were inconclusive.


[i]Consultative Group to Assist the Poor/The World Bank, “Financial Access 2009:  Measuring Access to Financial Services around the World,” January 9, 2011).

For more details, see the Gates Foundation briefing note on this project.

Examining the Effects of Crop Price Insurance for Farmers in Ghana

Policy Issue:
Many small-scale farmers in the developing world face significant income uncertainty, and rural farmers who live from harvest to harvest don’t have much room for error. Variables beyond the farmers’ control, such as fluctuating crop prices, can make a significant difference in how much a family earns for the year.  Farmers may be unwilling to take on additional risks by borrowing and making long-term investments due this uncertainty. This reluctance is thought to contribute to the decision of many farmers not to invest in technologies such as hybrid seeds, fertilizer, or irrigation that could potentially improve crop yields. Many lenders are also extremely wary of extending credit to farmers, fearful that they will inherit the risks inherent to farming. Crop price insurance could help solve this problem, reducing the risk to farmers and providing them with encouragement to make investments in their farms. Lenders, too, may feel more confident in lending to farmers with greater income certainty, facilitating even more capital investments.
Context of the Evaluation:
In Ghana, 50 percent of the rural population lives in poverty. In the Eastern Region where Mumuadu Rural Bank (MRB) operates, an estimated 70 percent of households make a living in the agricultural sector, but agricultural loans make up only 2 percent of the bank’s loan portfolio. Focus groups with maize and eggplant farmers in the area revealed that farmers were hesitant to borrow for fear that fluctuations in crop prices could force them to default. Rainfall fluctuations, typically an important source of risk for farmers, are not a great concern in this part of Ghana. The prices offered for traded crops, however, do fluctuate greatly. Information gathered in baseline surveys suggested that there was a potential but untapped market for crop price insurance: farmers in the area served by MRB expressed that they would be willing to pay to guarantee a certain minimum crop price. Despite this encouraging baseline finding, banks and insurance providers face the challenge that insurance is not a commonly understood concept among farmers in the region.
Details of Intervention:
Researchers developed an agricultural loan product in coordination with MRB that had an insurance component that partially indemnified farmers against low crop prices. Specifically, if crop prices at harvest dropped below a set price floor (the 10th percentile of historical prices for eggplant and the 7th percentile for historical maize prices), the bank would forgive 50 percent of the loan and interest payments. Borrowers were not required to pay any premium for the insurance product. The goal of incorporating insurance into the loan product was to reduce farmers’ risk in borrowing to invest in agriculture inputs. The intervention targeted maize and eggplant farmers in particular because the crops are both commonly grown in the region and subject to volatile (but historically well documented) prices.
Standard Mumuadu procedure is to invite farmers to meet in a group with Mumuadu employees to discuss the bank’s financial services, and to encourage farmers to come to a branch to apply for a loan. The average loan size is approximately US$159, which represents a significant change in cash flow for the borrower. For this project, Mumuadu employees approached community leaders to obtain a list of all maize and eggplant farmers in the village. The same community leaders then invited farmers to attend one of the bank’s information sessions. Farmers on the list were randomly assigned to one of four groups, each of which received a variation on the Mumuadu marketing pitch. The four groups were:
  1. Farmers who were offered the standard Mumuadu loan product;
  2. Farmers who were offered the Mumuadu loan product with complimentary crop price insurance;
  3. Farmers who received financial literacy training, before being offered the standard Mumuadu loan product;
  4. Farmers who received financial literacy training, before being offered the Mumuadu loan product with complimentary crop price insurance.
Prior to the marketing of the loans, Mumuadu employees conducted a survey of the farmers, gathering information relating to their credit history, risk perception, financial management skills, and cognitive ability. An analysis of baseline data, bank administrative data, and a followup survey that focused on farmer investment decisions allowed researchers to draw conclusions on the effect of crop price insurance on borrower behavior and agricultural investment in Ghana.
Take up of loans among farmers was quite high, with 86 percent of farmers in the comparison groups choosing to borrow and 92 percent of farmers in the treatment groups taking out a loan.  This high take up across both treatment and control groups made an analysis of the features that predicted take up difficult.  In fact, the researchers found no systematic difference across the treatment and control groups when considering which features predicted borrowing. Overall, those who borrowed tended to be older, with higher scores on tests of cognitive ability.  They were also more likely to have a record of previous borrowing.  
Apart from predictors of borrowing, researchers were interested in whether crop price insurance changed farmers’ investment behavior. There is evidence that it did, but not overwhelmingly. The small sample and high take up across both groups may have played a role in this outcome. Farmers offered the insurance spent 17.9 percentage points more on agricultural chemicals (mostly fertilizer) than those who had not been offered the product. There was also a trend towards growing more eggplants and less maize among these farmers. Farmers offered the insurance were also between 15 and 25 percent more likely to bring their produce to markets rather than sell to brokers who come to pick up the crop. Anecdotally, it is believed that the so-called “farmgate” sellers offer guaranteed purchase contracts, but at lower prices locked in before harvest. Selling in the market, on the other hand, is a potentially more profitable but riskier option.  
There are a number of potential reasons why the researchers did not find large effects of the crop price insurance product on either or take up or investment, and further research in necessary to determine their roles. It is uncertain, for example, whether farmers truly understood the benefits of the insurance. Farmers may also have been reluctant to make long term investments changes before an insurance product demonstrates an established presence in the area. Alternatively, crop price uncertainty may not be as important of an indicator of investment decisions as previously thought. Further research, with a larger sample size, is needed to better understand the roles of risk, financial literacy, and product design in determining microinsurance impact.

Making Networks Work for Policy: Evidence from Agricultural Technology Adoption in Malawi

Agricultural productivity growth in Sub-Saharan Africa has lagged behind the rest of the world, largely due to low adoption of productivity-enhancing technologies. Farmers may be reluctant to adopt new technologies because they lack credible information about them. Researchers examined how social network-based information dissemination impacted agricultural technology adoption in Malawi. Results suggest that targeting seed farmers using a complex social network model was most effective in increasing adoption, but this approach was also expensive and difficult to implement.

Policy Issue:

Agricultural productivity growth in Sub-Saharan Africa has lagged behind the rest of the world, largely due to farmers’ low adoption of productivity-enhancing technologies. One reason farmers are reluctant to adopt new technologies may be that they lack credible information about them. If this is true, learning through social relationships could be a key way to disseminate important information. However, little research exists on how information spreads through social networks and how that impacts individuals’ behavior. To address this question, researchers examined how social network-based information dissemination impacted agricultural technology adoption in Malawi.

Context of the Evaluation:

Malawi is a predominantly rural country where agriculture is a main source of employment and growth. While overall poverty and malnutrition rates have improved in recent years, the country remains impoverished; 47.8 percent of children under five were stunted in 2010. Given that more than 60 percent of the population’s calorie consumption derives from maize, technology adoption and productivity improvements in the crop could have significant welfare impacts on the country.

Agricultural training in Malawi is disseminated through an extension system operated by the Ministry of Agriculture and Food Security. Agricultural Extension Development Officers (AEDOs) train two seed farmers of their choosing in villages under their jurisdiction. But because many AEDOs are responsible for thirty villages or more, direct contact with villagers is rare. In a 2007 national census, only 18 percent of farmers reported participating in any type of extension activity.

Details of the Intervention:

Following up on a similar intervention, Promoting Sustainable Farming Practices in Malawi, researchers examined the impact of targeting information based on social networks on technology adoption. In collaboration with the Malawi Ministry of Agriculture and Food Security, researchers modified the existing extension program to promote two technologies: pit planting, a more productive way to plant maize, and better crop residue management (CRM), which emphasizes retaining residue to improve soil quality.

In total, the program included approximately 5,600 households in 200 villages in three districts: Machinga, Mwanza, and Nkhotakota. In each village, AEDOs trained two chosen seed farmers—well-connected farmers who would subsequently promote the technologies among their communities. To identify potential seed farmers, researchers conducted a census in each village to identify social networks. Researchers then randomly assigned each village to one of the following four variations of the program, in which seed farmers were chosen through different selection methods. All other aspects of the program remained the same.

1.     Simple: Seed farmers were chosen based on a simple social network model that assumed that a farmer would be convinced to adopt a technology if he knows just one other farmer that has.  

2.     Complex: Seed farmers were chosen based on a complex social network model that assumed that a farmer must know multiple other farmers who have adopted the technology in order to be convinced to do so.

3.     Geo Treatment: Seed farmers were chosen based on the complex social network model, but individuals were considered socially linked based on geographic proximity rather than the social census. This geographic targeting was a cost-effective substitute for the more expensive and difficult social network modeling method.

4.     Comparison: Seed farmers were chosen at the discretion of the responsible AEDO in accordance with the existing extension system.

In addition to receiving training, seed farmers received a small in-kind gift if they adopted pit planting in the first year of the program. No incentive was offered for adopting CRM, or for convincing others farmers in the village to adopt.

Over the next three harvest seasons after randomization, researchers collected information from households on farming techniques, input use, yields, and other demographic data. Precipitation data was also collected to account for the impact of weather patterns.

Results and Policy Lessons:

The program convinced seed farmers to adopt pit planting, leading to increased yields for these farmers. While technology adoption by non-seed farmers was low across all groups, the Complex villages exhibited the largest increases in adoption relative to comparison villages. Results from the Geo Treatment group suggested similar but inconclusive increases in adoption.

The program increased the likelihood that seed farmers across all groups adopted the technologies. Compared to shadow seed farmers, or those farmers who would have been chosen as seed farmers under the selection methods used in the other villages, actual seed farmers were 25.1 percentage points more, or more than four times as likely, to adopt pit planting in the first year. Compared to shadow farmers, seed farmers who adopted pit planting experienced a 44.9 percent increase in maize yield.  Moreover, differences in yields were largest during low-rainfall conditions, an important resiliency feature of the technology.

Trained seed farmers were also 13 percentage points more likely to use CRM in the first year, but adoption dropped in the second year among both actual and shadow seed farmers. This drop-off indicates that the technology may not have been well suited for these farmers. Thus, only pit planting adoption is discussed below.

While there was no difference in pit planting adoption during the first season between the network-targeted villages and comparison villages, adoption was significantly higher in network-targeted villages in seasons two and three. In season two, adoption rates increased by 61.4 percent and 79.5 percent in Simple and Complex villages respectively, compared to comparison villages where the adoption rate was 4.4 percent.

The increase in adoption in network-targeted villages was largely driven by farmers who had links to both trained seed farmers, suggesting that the complex social network model was more effective than the simple model. Farmers who had connections to both were 31.7 percent more likely to have heard about pit planting in season one, and 88.6 percent  more likely to adopt it in season two, compared to farmers with no connections.

These results demonstrate that selecting seed farmers based on a complex social network model was most effective in increasing pit planting adoption. Further research is needed on other simple and inexpensive procedures that could replicate the results of the successful but data-intensive and expensive social network modeling approach.

Innovative Finance for Technology Adoption in Western Kenya

Adoption of agricultural technologies like fertilizer and improved seeds is low in African countries compared to other developing countries and evidence suggests that using these technologies can dramatically increase farm productivity and income. In an ongoing study, researchers are collaborating with One Acre Fund to examine the effect of credit and savings on the adoption of fertilizer and hybrid seeds, farm productivity, and farmer livelihoods.

Policy Issue:

Adoption of modern agricultural technologies—inorganic fertilizer and hybrid seeds in particular—can lead to large increases in productivity and profitability for many African smallholder farmers. Yet actual adoption rates for these technologies are low: fertilizer use in Sub-Saharan African countries is about twelve times lower than other developing countries and adoption of modern crop varieties is about five times lower than in Asian countries. One potential explanation is that low adoption is linked to poor access to credit or savings, meaning that farmers often have difficulty finding enough cash when it comes time to make these purchases. Improved access to credit, in the form of upfront fertilizer and seed loans which are paid back over time, could spur adoption. Yet credit is relatively expensive and inflexible compared to savings. Savings products could help farmers accumulate cash to make fertilizer or seed purchases, but it may be difficult for farmers to save due to competing uses for their money and demands from family or friends. While credit and savings may potentially benefit farmers, there is little evidence about the relative effectiveness of either tool for increasing use of agricultural technologies.

Context of the Evaluation:

About 79 percent of Kenya’s population lives in rural areas and relies on agriculture for most of its income. Smallholder subsistence agriculture produces about 75 percent of total agricultural output and the primary crop is maize. For many farming households, home maize storage is the primary source of savings and is typically much larger than other formal sources (e.g. bank accounts) or informal sources (e.g. community savings groups). Yet many households suffer from substantial post-harvest losses in maize stores, due to insects or rotting. Farmers also appear to use storage inefficiently: they sell their maize at low post-harvest prices and often buy it back later in the season at prices that are four to five times as high. As a result, farmers may lack the cash needed to purchase fertilizer or seed during planting season. Yet for farmers who need cash, access to credit remains low: according to a 2009 national survey, 60.4 percent of adult Kenyans reported no access to credit or loans, and less than 10% of farmers in the study region in Western Kenya report having access to formal credit. Based in Kenya, One Acre Fund (OAF) works with over 130,000 farm households in Kenya, Rwanda, and Burundi, focusing mainly on seed and fertilizer loan and delivery. When expanding into new areas, OAF holds initial community interest meetings and then organizes groups of farmers who all receive the same program or service.

Details of the Intervention:

In partnership with OAF, researchers are examining the effect of different credit and savings products on the adoption of fertilizer and hybrid seeds, farm productivity, and farmer spending on health, education, and food. In particular, they are currently evaluating whether well-timed access to credit allows farmers to make better use of storage and sell their output at higher prices, and they are studying how any additional profits are re-invested. Researchers first randomly divided 232 groups of farmers (5-7 per group) into one of the following groups: 

  1. Post-Harvest Loan: 77 groups (474 farmers) were randomly selected to be offered a loan directly following harvest in September or October. 
  2. Post-Harvest +3 Loan: 75 groups (480 farmers) were randomly selected to be offered a loan three months after Harvest in January. 
  3. Comparison Group: 80 groups (635 farmers) were offered no loan.

Stored maize served as loan collateral, and all farmers who received loans also received a laminated tag with OAF’s logo attached to farmers’ stored maize. Following random assignment of the loan treatments, a subset of individuals within each group was randomly selected to receive a simple cash lockbox to promote savings. Additionally, researchers randomly selected 159 farmers within the comparison group to receive laminated tags to test whether the tag alone allows farmers to credibly claim to friends and family that they cannot give away their stored maize. Stored maize is easily visible to family and visitors and local norms promote sharing of surplus maize, so the tag may allow farmers to justify not sharing.

Results and Policy Lessons:

Results forthcoming. 

Edward Miguel

Ex-combatant Reintegration in Liberia

For post-conflict societies, the challenges of reintegrating ex-combatants and war-affected youth are likely to far outlast and outsize the formal demobilization, disarmament and reintegration (DDR) of ex-combatants. These programs, conducted in war’s immediate aftermath, form an important part of a policymaker’s post-conflict toolkit. While ex-combatants receive special policy attention, poor and underemployed men are also widely considered a threat to political stability.

Find a more detailed policy brief here (PDF) and the full paper here.
Context of the Evaluation:

In Liberia, where the bulk of the population is young, poor, and underemployed, many rural youth continue to make their living through unlawful activities, including unlicensed mining, rubber tapping, or logging. Many of them are ex-combatants, and some remain in loose armed group structures, doing the bidding of their wartime commanders. While the security situation has steadily improved since 2003, the government, the UN, and NGOs fear that these youth are a possible source of instability, particularly in hotspot regions where mining, rubber tapping, or logging and the allure of “fast money” attract young men from around the country. These youth may also be recruited into regional conflicts as mercenaries. Agriculture is and will continue to be a major source of employment and income for rural Liberians. The international NGO Landmine Action (LMA, now known as Action on Armed Violence) runs an innovative and intensive agricultural training program, targeting ex-combatants and other high-risk youth in rural hotspots.

Description of the Intervention:

The LMA program is broader and more intensive than most ex-combatant reintegration programs, and is designed to rectify some of the main failings of prior demobilization programs: it is oriented towards agriculture (the largest source of employment in Liberia); it provides both human and physical capital; and it integrates economic with psychosocial assistance. It also targets youth at natural resource hotspots that presented the most immediate security concerns.

LMA took youth selected for the program to residential agricultural training campuses, where they received 3-4 months of coursework and practical training in agriculture, basic literacy and numeracy training, psychosocial counseling; along with meals, clothing, basic medical care, and personal items. After the training, counselors facilitated graduates' re-entry with access to land in any community of their choice.  Graduates received a package of agricultural tools and supplies, valued at approximately US$200. The program's total cost is approximately $1,250 per youth, excluding the cost of constructing the campuses. The program was designed to give youth a sustainable and legal alternative to illegal resource extraction, ease their reintegration into society, reduce the risk of re-recruitment into crime and insurrection in the future, and to improve security in hotspot communities.

LMA recruited twice as many youth as it had space for in its programs, and researchers randomly assigned half of the youth to treatment (receiving the program), and half to a comparison group (not receiving the program). By comparing these two groups 18 months after the program, researchers can see the effect of the intervention on agricultural livelihoods, shifts from illicit to legal employment, poverty, social integration, aggression, and potential for future instability.  Despite massive migration, 93% of the youth were found at the time of the endline survey. The qualitative study included observation and a series of interviews with 50 of the youth.

Results and Policy Lessons:

Engagement in agriculture: More than a year after completion of the program, program participants are at least a quarter more likely than the control group to be engaged in agriculture, and 37% more likely to have sold crops. Interest in and positive attitudes toward farming are also significantly higher among program participants. 

Illicit activities:The program had little impact on rates of participation in illicit activities like mining, but those who participated in the program do spend fewer hours engaged in illicit activities, as agricultural hours seem to substitute somewhat for hours spent in illicit activities.

Income, expenditures, and wealth:  There was a sizable increase in average wealth from the program, especially in household durable assets, but no change in current income (last week and last month), savings or spending for the average program participant. Overall, the evidence suggests that cash cropping provides periodic windfalls from sales, and that these are mainly invested in durable assets (and not necessarily in agricultural inputs or equipment).  Qualitative observations also suggest that access to markets may have been an important constraint on success.

Social engagement, citizenship, and stability:  There were small but positive improvements across most measures of social engagement, citizenship, and stability. While not all of the estimated impacts are large enough to be statistically significant, they nevertheless suggest a small but broad-based reduction in alienation and some gains in stability. The evidence on aggression and crime, however, does not point to a significant reduction in illegal or aggressive behaviors among program participants.

Interest and mobilization into the election violence in Cote d’Ivoire:Conflict broke out in Cote d’Ivoire shortly before the launch of the program evaluation.  Self reported rates of interest in the violence and mobilization were fairly low among the sample population, but they were especially low among program participants – they tended to report a third less interest in or links to recruiters and recruitment activities. Given the difficulty of shifting such behaviors, these impacts of the program are regarded as extremely promising.

More information can be found in the policy brief here (PDF) and full paper here.

The Socio-economic Impacts of Ebola on Households in Sierra Leone

The economic impacts of the Ebola virus must be monitored in real time for policymakers to estimate the short- and long-term costs of the epidemic and respond appropriately, yet information on the magnitude of the effects remains scarce. The aim of this monthly survey in Sierra Leone is to measure the economic and social impacts of the outbreak on households and provide timely updates on the survey's findings to policymakers.

Policy Issue:

The Ebola outbreak in West Africa may be disrupting local economies, and if so, the long-term impact could be considerable. Reports have circulated that the virus has increased prices, reduced the availability of essential goods, and impacted agricultural production, yet there is little to no quantitative data to back these claims and inform appropriate policy responses. There is a great need, therefore, to monitor economic impacts in real time and provide accurate date to governments and their development partners. This research in Sierra Leone aims to help fill the information gap and supply policymakers with timely updates to address the crisis in both the short- and long-term.

Note: This is not a randomized controlled trial.

Context of the Evaluation:

Since its initial appearance in March 2014 in rural Guinea, the Ebola virus has spread to three other West African countries.  As of November 12, 2014, Sierra Leone has had more than 4,900 confirmed cases and nearly 1,200 deaths.1 The situation has become even more challenging as the virus has now taken hold in the capital, Freetown. Two of the country’s fourteen districts have been quarantined for over two months, an additional three districts were quarantined in mid-September, and certain areas of the capital are also under isolation. 

In partnership with the World Bank, IPA is supporting Statistics Sierra Leone (SSL) with on-the-ground technical assistance and supervision for a monthly household survey that measures the economic impacts of the Ebola virus over time.

Details of the Intervention:

This multi-round survey is a tool for monitoring the socio-economic and service delivery impacts of the Ebola virus. IPA and Statistics Sierra Leone will jointly administer the cell phone survey to a subsample of households taken from a national survey. The focus of the data collection is on economic indicators – such as the functioning of labor markets, the availability of foodstuffs, agricultural production, other health outcomes.

We aim to conduct a 20-minute survey at monthly intervals, with some questions  administered every month and other questions rotated into the questionnaire based on the situation at the time. 

The questionnaire will incorporate baseline information from the nationally-representative Labor Force Survey (SLLFS), administered in 2014, to capture changes and, due to the brief nature of this survey, to target and streamline data collection.

Results and Policy Lessons:

Round 1: released January 12, 2015

  • In the first round of data collection, wage and non-farm self-employed workers saw the largest declines in employment in urban areas of Sierra Leone, with Ebola cited as one of the main reasons for not working.  An estimated 9,000 wage workers and 170,000 self-employed workers outside of agriculture are no longer working since the July/August 2014 baseline.  The percent of households engaged in a non-farm household enterprise that was no longer operating tripled from 4 to 12 percent and among those still operating these businesses, average revenue decreased by 40 percent.
  • These job losses have been caused mainly by the indirect effects of necessary preventive measures to restrict disease spread and by the general disruption to the economy caused by the outbreak. No differences were found in labor impacts between quarantined and non-quarantined districts, further highlighting the importance of economy-wide indirect effects. 
  • Food insecurity is high as the harvest continues.  There is no evidence thus far on negative impacts on agriculture due specifically to Ebola, but harvest activities are still ongoing and future rounds of data collection will track Ebola-related effects if and when they arise.
  • There is some evidence of a decrease in utilization of health services for non-Ebola conditions in Freetown.  In particular, a much lower proportion of women in the capital reported post-natal clinic visits than in 2013. In the rest of the country, however, there is little evidence of such a decline.

Round 2: released April 15, 2015

  • There are signs of improvement in Sierra Leone, but the economic situation remains uneven. While there have been overall improvements in employment since November – driven by urban areas, youth employment in Freetown has continuously declined and the percentage non-farm enterprises that are no longer operating has increased fourfold. 
  • Stability of earnings has depended on the employment sector. Wage workers are earning around the same as they did pre-crisis, while those operating non-farm household enterprises are seeing revenues around 54 percent lower than in July-August 2014. Women in particular are affected, mostly due to the fact that they are generally working in non-farm household enterprises, the sector most heavily impacted by Ebola.
  • Food insecurity, which was high in Sierra Leone even before the crisis, continues to be a concern. Nearly 70 percent of households taking at least one action to cope with food shortages in the week leading up to the survey. Coverage of social assistance thus far reflects the disease-specific targeting of the emergency response, suggesting efforts to reach the poorest will be key as the country moves toward recovery. 
  • Delivery of social services has generally improved. The utilization of maternal care services has increased significantly since November: the percentage of women who gave birth in a clinic up from 28 percent to 64 percent and the percentage who received at least one prenatal visit up from 56 percent to 71 percent. National educational radio programs, working to bridge gaps created by long-term school closings, have reached nearly 72 percent of households with school-aged children, who reported that at least some children listened to these programs.

A third round of mobile phone data collection in Sierra Leone is planned for April 2015, to continue to track and highlight the most pressing areas of attention for policy makers as they move toward the economic recovery phase.

[1] Source: CDC Ebola Outbreak Monitoring,

Disseminating Innovative Resources and Technologies to Smallholder Farmers in Ghana

Researchers are testing whether access to improved-yield agricultural inputs and agricultural extension advice—when provided individually or in combination—leads to more intensive land cultivation and increased earnings among farmers in northern Ghana who receive access to rainfall insurance. The evaluation of the program, Disseminating Innovative Resources and Technologies to Smallholders (DIRTS), is being conducted in partnership with the Ghana Agricultural Insurance Pool, Savanna Agricultural Research Institute, and the Ghanaian Ministry of Food and Agriculture.

Policy Issue:

In Ghana, as in other parts of Sub-Saharan Africa and across the developing world, investment in agricultural inputs such as fertilizer, high-yield seeds, and farm equipment is low among smallholder farmers. One potential reason farmers may underinvest is that they face uncertain rainfall. This may provide a disincentive to invest – for example, in the case of a drought, the greater a farmer’s investment the greater their losses. Another potential reason is that these inputs may not be widely available in local markets and farmers may not have the cash on hand to purchase them just prior to planting season. Farmers may also lack information on the benefits of these inputs and how to use them.

There are many potential solutions to these three barriers. Just as in other risky enterprises, farmers may invest more if they have access to insurance that protects them from catastrophic losses, in this case from abnormal weather events. By coupling free delivery with the option to buy fertilizer and other inputs just after harvest when they have cash on hand, farmers may find it easier to make these investments. Finally, agricultural extension could potentially help farmers learn about the benefits of these inputs and use them most effectively, particularly for inputs like improved seeds which have more complicated planting procedures. In the past, these interventions have been examined individually, however their effects when provided together are not well understood. The researchers wish to better understand how extension and improved input provision— in combination or alone—may augment the impact of rainfall insurance on farmers’ yields and profits.

Context of the Evaluation:

In Ghana, agriculture constitutes 23 percent of economic output and 42 percent of total employment, yet farmers in northern Ghana achieve just 30 percent of potential crop yields.1 The farmers participating in this study face chronically low harvests and cultivate primarily maize. On test plots of land, the partner organization Savanna Agricultural Research Institute found that farmers could likely earn a net profit of US$127 per acre using inorganic fertilizer alone; however, when using inorganic fertilizer as well as organic fertilizer as a top dressing, farmers could likely earn a return of US$190 per acre. A past pilot evaluation in northern Ghana showed a 37-93 percent take-up rate of rainfall insurance depending on the relative terms and price, suggesting that there is a high level of demand for the product. Furthermore, in the same evaluation, researchers found that farmers with insurance increased fertilizer use by 25 percent, cultivated area by eight percent and labor use by 13 percent.

Details of the Intervention:

Through a randomized evaluation of 3,240 households in 162 farming communities in northern Ghana, researchers are testing the impact of agricultural extension and improved input delivery on farm output among farmers who have access to agricultural insurance. Researchers are implementing the evaluation of Disseminating Innovative Resources and Technologies to Smallholders (DIRTS) in partnership with the Ghana Agricultural Insurance Pool, Savanna Agricultural Research Institute, the Ghanaian Ministry of Food and Agriculture, and Innovations for Poverty Action.

The rainfall index insurance and agro-inputs are sold at market prices to all interested farmers in the selected communities, whereas extension services are provided to ten randomly selected farmers in each of the communities in the extension arm. Specifically, for the first year of the intervention, the households are randomly assigned to four groups:

(1) Extension Advice: In 52 communities, farmers receive interactive, individual-level trainings on farming best practices in addition to access to insurance. Trainers are community-based and compensated for each farmer interaction. Twenty households per community are randomly selected to be surveyed as part of the evaluation. These 20 households are randomly assigned to four sub-treatment groups of five households each. These four groups are free insurance, free insurance plus extension, extension only, and a comparison group.

(2) Input Marketing and Delivery: In addition to access to the insurance product, farmers in these 31 communities receive the opportunity to buy commercial inorganic fertilizer, certified seeds and other agro-chemicals, and equipment at market price to be delivered before or right at the start of planting time. During the first year of the study, farmers receive this option at four points in time, including in harvesting time when they had cash on hand. Twenty randomly selected households per community are surveyed as part of the evaluation. Of these 20 households, ten households are randomly selected to receive the grant of free insurance and ten receive insurance. Both groups get access to inputs.

(3) Combined Intervention: In addition to access to the insurance product, farmers in 29 communities receive access to high-yield inputs with delivery as well as extension advice. Twenty households per community are randomly selected to be surveyed as part of the evaluation. These 20 households are randomly assigned to four sub-treatment groups. These four groups, composed of five households each, are: free insurance plus access to inputs, free insurance plus extension plus access to inputs, extension plus access to inputs, and access to inputs only.

(4) Insurance Only: In 50 communities, farmers just receive access to rainfall insurance. Farmers in these communities are given the opportunity to buy insurance at market price. Twenty households per community are randomly selected to be surveyed as a part of the evaluation. Of these 20 households, ten households are randomly selected to receive a grant of free insurance and ten serve as a comparison.

Results and Policy Lessons:

Results forthcoming.

Read our Community Extention Agent Updates:


[1] Fosu, Mathias. (2011). AGRA Soil Health Project 2010 Annual Report. 

The Impact of Ebola on Market Prices in Sierra Leone

Starting the summer of 2014, many risk factors pointed to a potential food crisis in areas of West Africa hit hardest by the Ebola outbreak. Innovations for Poverty Action, in partnership with researchers and the International Growth Center, began monitoring markets across Sierra Leone for changes in food prices and supply. Researchers are now providing rapid feedback to the government and other development partners on where food shortages are occurring. With our partners, we are continuing to monitor food markets across Sierra Leone.

Policy Issue:

The Government of Sierra Leone and its development partners must have valid, credible data and analysis to ensure that their policy responses to the Ebola outbreak are evidence-based and well targeted.While many reports are circulating about the outbreak’s effects on the economies of affected countries, hard data on what is happening on the ground is in short supply. Some communities have been quarantined under a “cordon sanitaire,” potentially affecting food prices and supply in these areas. A slowdown in agricultural production, local and international trade, and/or labor supply could also be impacting food security. By tracking food prices and the supply of staple foods in markets countrywide, this research aims to identify food insecurity before it happens and provide rapid feedback to the government.

Note: This is not a randomized controlled trial.

Context of the Evaluation:

To help contain the Ebola outbreak, the Government of Sierra Leone has restricted movement in and out of certain areas of Sierra Leone. The quarantine of certain areas under the cordon sanitaire was imposed around the towns of Kenema and Kailahun in the summer of 2014, and expanded to other areas in September. This cordon could be obstructing the movement and marketing of food and causing spikes in food prices.

Details of the Intervention:

Researchers are measuring the impact of the Ebola outbreak on market prices in Sierra Leone. IPA is gathering data from markets across the country, and researchers are measuring changes over time and comparing the information to similar data we gathered from the same markets in 2011 and 2012.

In mid-August we collected data from 153 markets on food availability, prices of key food stuffs (including imported and domestic rice, cassava, palm oil, and fish), and the number of traders operating in the market. We then conducted a second round of market surveys during mid-September in 157 markets.

IPA staff based in Freetown surveyed individuals by phone that manage markets across the country. (We gathered the phone numbers in 2011 and 2012 during a previous survey.)

Using data from previous surveys, we have been able to compare these prices to those at the same time of year in 2012 and, for most markets, in 2011. Because we have the GPS location of the markets it has also been possible to compare outcomes for areas badly hit with Ebola or subject to transport restrictions, such as closure of borders, with those that have been relatively Ebola-free and not subject to transport restrictions.

Collecting data every month during the course of the outbreak is allowing researchers to track shortages pre-harvest and potential surpluses post-harvest (which may accumulate if transport is disrupted and drive prices down in some areas).

The latest round of market surveys took place in early October shortly after all Sierra Leoneans were asked to stay at home for two days and after the introduction of new cordon restrictions in Port Loko, Moyamba, and Bombali.

Results and Policy Lessons:
The latest rounds of market surveys took place in late 2014, between November 19-26 and again between December 17-21.
  • Rain fall returned to more seasonal levels in November and December and the rice harvest was underway throughout the country by late November.
  • Prices for most basic foods remain at or below previous year’s prices.
  • With the rice harvest came a modest increase in the number of domestic rice traders, which has returned to levels similar to those in 2011.
  • Traders in markets for other staples like processed cassava (gari) and palm oil are still below those for previous years although the gap closed somewhat in December.
  • The number of markets reported to be closed in November is largely unchanged since October, and slightly lower in December.
  • Preliminary evidence suggests that international shipping in West Africa fell in September, but appears to be similar in October and November relative to 2013. These results are consistent with emerging results from other surveys (including surveys of households) which suggest informal activity has been depressed by the Ebola outbreak but that food prices have not been impacted. 

Earlier Findings:

Data collected in August and September revealed that prices were relatively stable. Evidence from October revealed prices of basic food commodities were not significantly higher than they were in recent years, but some findings suggested the situation was worsening:

The number of traders selling basic food items continued to fall in all districts. In Kailahun and Kenema (the first districts to be cordoned) there were 69 percent fewer domestic rice traders than in 2012 while the decline in newly cordoned areas is 29 percent. Other key findings from October 2014:

  • Prices of basic food commodities at markets were not significantly higher in October than they were at that time in previous years, nor were they higher on average in cordon areas.
  • There were outliers where prices were much higher and there were more of these outliers than in normal years.
  • There was an increasing number of markets that are closed. In most of these cases traders report they are selling food from their homes. However, it will be important to monitor food security at the household level to ensure that food (at reasonable prices) is reaching households especially in remote locations.
  • Very preliminary data suggested a new risk to food security, or at least a potential delay in the rice harvests. Rainfall in September was much higher than it usually is at this time of year, but it began to decrease in October. This could negatively impact the rice harvest or at the very least delay the rice harvests.
  • Food security is not just a function of food availability and price but also of income. The reduction in the number of traders suggested reductions in economic activity more generally.

Researchers presented data to the government after it was gathered.

Finding Missing Markets: An Agricultural Brokerage Intervention in Kenya

Policy Issue: 

In much of the developing world, farmers grow crops only for local or personal consumption, despite export options which are thought to be much more profitable. There are several plausible reasons why farmers might choose to grow crops for local markets, forgoing the opportunity to make more money through export crops. There may be information gaps about profitability of export crops, lack of access to the capital needed to make the switch to export crops, inadequate infrastructure to bring crops to urban centers, concern over risky export markets, or misinterpretation by researchers as to the true profit opportunities. 

Context of the Evaluation: 

Kenya’s horticultural sector, which includes fruit and vegetable production, has received a great deal of attention over the past decade due to the rapid and sustained growth of its exports to Europe. Although Europe’s appetite for Kenyan agriculture exports has been great, small farmers have largely failed to cash in on this opportunity. Many farmers instead receive below-market prices for their crops by selling them at the local market or to intermediaries, who then resell the produce at regional market centers or to export firms. In the study sample, about half of the household income came from agriculture, and most owned the land they cultivated, which was usually about one acre. Farmers grew subsistence crops (beans, maize, potatoes, and kale) 50 percent of the time and cash crops (coffee, bananas, and tomatoes) 34 percent of the time. Only 12 percent of farmers grew any export crops.  

Details of the Intervention: 

In collaboration with DrumNet, a Kenyan NGO, researchers evaluated whether a package of services could help small farmers adopt, finance, and market export crops, and thus make more income. DrumNet tried to link smallholder farmers to commercial banks, retail farm suppliers, transportation services, and exporters. To be a member of DrumNet, a farmer had to be a member of a registered self-help group (SHG), express interest in growing export crops marketed by DrumNet (i.e. French beans, baby corn, or passion fruit), and have irrigated land. 

In 2003, researchers randomly divided the 36 active self-help groups (SHGs) in the Gichugu area into three equal groups: the first treatment group received all DrumNet services; the second treatment group received all DrumNet services except for credit; and the third served as the comparison.

All individuals in the two treatment groups received a four-week orientation course, which explained the financing and selling process, and good agricultural practices. In additional, all treatment individuals opened a personal savings account with a local commercial bank to accommodate possible future business transactions. Individuals in the credit treatment group also contributed the equivalent of a week’s labor wages to an insurance fund, which would serve as partial collateral for a line of credit. After being organized into groups of five, which were jointly liable for individual loans taken out, individuals in the credit treatment group received an in-kind loan from a local agriculture supply store.

At harvest time, for individuals in both treatment groups, DrumNet negotiated prices with an exporter and arranged a produce pickup. Once the produce was delivered to the exporter, the exporter payed DrumNet who, after deducting any loan repayments, credited the remainder to the individual savings accounts that each farmer opened when they registered. 

Results and Policy Lessons: 

Impact of the DrumNet Program: One year after the program began, treatment individuals were 19.2 percentage points more likely to be growing an export crop, but there were no significant gains in income for the full sample. However, among first-time growers of export-oriented crops, program participation led to a 31.9 percent increase in income.

Out of the twelve SHGs in each treatment group, ten decided to take advantage of DrumNet services when credit was offered, compared to only five of twelve when it was not, implying that farmers perceived credit as an important factor for cultivating export crops. However, access to credit had no effect on income gains compared to no-credit SHG groups. 

Long-term Consequences: Unfortunately, one year after the evaluation ended, the exporter refused to continue buying from the DrumNet farmers since none of the SHGs had obtained EU export certifications. This led to DrumNet’s collapse as farmers’ export crops were left to rot and loans went into default. Farmers returned to growing for local markets, underscoring the original concerns over export market risk. 


Commitment Savings Accounts for Farmers in Rwanda

Many farmers in developing countries lack the capital necessary to invest in potentially profitable inputs, as these investments must be made months after the harvest when households lack cash. Commitment savings accounts, which have features to discourage withdrawals, have been shown to help the poor save, and could help farmers put aside money to invest in their farms. However, demand for these products is low. In an ongoing study, researchers are aiming to identify product features that increase demand for commitment savings accounts and evaluate the effect of the accounts on farm investments and yields.

Policy Issue:

Many farmers in Sub-Saharan Africa lack the capital necessary to invest in potentially profitable inputs, as these investments must be made months after the harvest when households lack cash. Saving may also be particularly hard due to demands from family and friends. Commitment savings accounts, which have features to discourage withdrawals, have been shown to help the poor save, and could help farmers put aside money to invest in their farms. However, demand for these products is low, and it remains unclear what causes individuals to save more. Savings products that allow customers to withdrawal money in the case of emergencies could make the accounts more popular, help more farmers invest in their farms, and in turn, increase farm profitability. To shed light on these questions, researchers are investigating how various product features impact demand for commitment savings products, and the how the saving accounts impact farm investment and yields.

Context of the Evaluation:

The majority of Rwandan farmers participate in two planting seasons. Farmers generally face a tight planting calendar, with harvest and planting times for the two main crop seasons nearly overlapping: Season A ends in January/February and Season B starts in February.

Many Season A crops (such as beans) require drying or other post-harvest processing before farmers can market and sell the crops. As a result, farmers rarely have income from Season A by the time they need to buy inputs for Season B. Many farmers in the region either use credit or reduce input use to deal with this shortfall.

Details of the Intervention:

Researchers are partnering with Rwanda’s Ministry of Agriculture and Animal Resources (MINAGRI) and local Savings and Credit Cooperatives (SACCOs) to evaluate the impact of input saving accounts on investment and yields. The commitment savings products are designed to allow farmers to exert self-control and save a portion of their season B income for the following seasons A and B. Researchers are working with 1,101 farmers from four cooperatives to determine whether pairing insurance with savings increases take-up or changes savings behavior. Farmers were randomly assigned to one of six groups:

·      Standard Commitment Savings Account (196 farmers): Farmers were offered an account that allows free withdrawals in only two periods corresponding to the planting seasons. All other withdrawals incur a withdrawal penalty.

·     Commitment Savings Account with Health Exemption (231 farmers): Farmers were offered the standard commitment savings account, with no withdrawal penalty if the account holder provides documentation of an adverse health event.

·     Commitment Savings Account with Agricultural Exemption (230 farmers): Farmers were offered the standard commitment saving account, with no withdrawal penalty if the local agronomist verified an adverse agricultural event.

·      Commitment Savings Account with Private Exemptions (217 farmers): Farmers were offered the standard commitment savings account, but individuals were privately informed that they were exempt from all withdrawal fees.

·      Comparison group (227 farmers): Farmers were offered financial literacy training and the opportunity to register for a standard savings account, but not a commitment savings account.

In addition, researchers are privately announcing to half of the farmers who register for a commitment savings account that they will be exempt from all withdrawal fees on the account – i.e. the account will in practice function like a standard savings account, but is labeled as a commitment savings account. By privately announcing that the commitment account will function like a standard account, researchers will be able to identify whether the commitment creates a mechanism for shielding money from others or from one’s own impulses.

Results and Policy Lessons:

Results forthcoming.


The Market for Local Agricultural Information in Western Kenya

Providing better information to farmers about local agricultural conditions could enable them to make more informed and locally appropriate agricultural decisions, with potentially positive consequences for their income, food production, and the environment. This study in Western Kenya will test for failures in the market for local agricultural information and measure the impact of disseminating local information on farmers’ decisions to invest in agricultural inputs.

Policy Issue:

Across Africa, soil and agro-climatic conditions vary dramatically from place to place, and some researchers argue this high variability has hindered the continent’s ability to reap the benefits of the Green Revolution like Asia and Latin America have. After all, the effectiveness and/or profitability of specific agricultural inputs can be highly dependent on local conditions such as soil type, micro-climate, and local prices. Providing better information to farmers about local agro-climatic conditions may help them make more informed and appropriate agricultural decisions, with potentially positive consequences for their income, food production, and the environment. However, information has traditionally been difficult to sell: since it can be easily shared, it is hard for a producer of information to recover its costs. Failures in the market for local information might warrant the development of policies that would support the creation and dissemination of local information.

Context of the Evaluation:

In Western Kenya, soil tests that can determine what type and amount of inputs are suitable for a given area are rarely used. Most small-scale farmers have never formally experimented with different fertilizers on their own land. While soil tests and experimental plots might be too expensive for every farmer to implement (especially since farms are small), testing soils for some farmers in the area and sharing these results with neighboring farmers could potentially be a cost-effective way to produce locally-relevant information about soil conditions and appropriate inputs.

Details of the Intervention:

This study in rural Western Kenya aims to measure farmers’ demand for local agricultural information, evaluate the impact and cost-effectiveness of different information interventions, and explore the potential for institutional mechanisms to address market failures.

First, researchers will conduct a randomized evaluation among a group of 1,069 smallholder farmers to better understand how two types of information affect their decisions to purchase fertilizer. Eight-hundred respondents will be randomly assigned to receive soil tests or both soil tests and test plots on their land. The remaining 269 respondents will serve as a comparison group. During an initial survey, soil samples will be collected and sent to a lab for testing, to measure the soil’s acidity and nutrient composition, and to determine the best type of inputs to improve the soil quality.

Farmers assigned to receive test plots will receive fertilizer and technical assistance to set up plots using different types of fertilizer. After harvesting with the farmers, researchers will analyze the test plot yields to inform the farmers of which inputs are most profitable for their land. All respondents will also receive fertilizer coupons to measure the effect information has on decisions to purchase fertilizer.

Second, researchers will measure demand for local agricultural information from a new group of farmers’ identified in the first evaluation. To do so, the research team will elicit farmers’ willingness to pay for test plot and soil test results from other farmers in their area.  

Third, researchers will also work with the second study sample to examine the relative effectiveness of two additional sources of agricultural information on farmers’ knowledge and investment decisions. In a partnership with the Kenya Agriculture and Livestock Research Organization, they will randomly divide the sample into three groups: one incentivized to attend open farmer field days, one to receive an SMS intervention, and a comparison group. The farmer field days will exhibit test plots and provide information about best farming practices. The SMS intervention will provide farmers with general tips about how to increase yields throughout the season.

Finally, once results from the endline surveys are in, researchers will compare all of the information inventions tested to identify the most cost-effective and scalable solutions.

Results and Policy Lessons:

Results forthcoming.


Evaluating the Integrated Agriculture Productivity Project in Bangladesh

Despite recent economic growth in Bangladesh, food insecurity remains widespread. Researchers evaluated the impact of an agricultural training program for farmer groups on technology adoption in rural Bangladesh, and investigated what drives adoption and who is affected by the training, both directly and indirectly.

Policy Issue:

In South Asia’s highly agrarian and populous countries, farm holdings are typically small and agriculture is characterized by intensive cultivation.1 Despite recent economic growth, 17 percent of the population was undernourished in 2011-2013, making the region home to the highest number of undernourished people in the world.2 Adoption of certain farming technologies, such as improved seeds, farm-yard manure, and integrated pest management, may enable farmers to produce more food and earn more income, yet many farmers have not adopted these technologies. This research contributes to our understanding of technology adoption in the region.

Context of the Evaluation:

Over the last two decades, Bangladesh has achieved impressive economic growth and poverty reduction. Its agricultural sector grew at a rate of 4.8 percent between 1990 and 2005. However, poverty-related food insecurity is widespread, bolstered by the soaring prices of key staples like rice, wheat, and legumes. The country has a poverty rate of over 30 percent and a highest incidence of malnutrition: in 2008, Bangladesh’s food insecure population was estimated at 65.3 million, or 45 percent of the population.3

The Government of Bangladesh is pushing for increased use of technology and more intensive agricultural practices to improve food security and sustain economic growth.  The government is implementing the Integrated Agriculture Productivity Project (IAPP) under the supervision of the World Bank and U.N. Food and Agriculture Organization, and with funding from the Global Agriculture and Food Security Program. The project aims to enhance the productivity of crops, livestock and fisheries. It is taking place among small-scale and marginal farmers in northern and southern districts of Bangladesh, which are characterized by high rates of poverty and food insecurity, and vulnerability to the negative effect of natural shocks, such as tidal surges in the south and flash floods and drought in the north.

Details of the Intervention:

Researchers are conducting a four-year randomized evaluation in 316 villages to study the impact of the IAPP’s livelihoods field schools (LFS) on technology adoption, and to investigate what factors drive adoption of new technologies. The LFS consists of groups of approximately 25 farmers who meet twice a month with extension agents. The aim of the school is to promote adoption of new seed varieties and improved production practices.

To evaluate the overall impact of IAPP activities on farmer livelihood and the effect of  LFS on technology adoption, in 2012 researchers randomly assigned 96 villages in six districts to either the treatment group, which participated in program activities immediately, or the comparison group, which received the program two years later. In treatment villages, one demonstration farmer was chosen for each type of technology introduced into the farmer group. This demonstration farmer received free seeds, fertilizer, and training. The selected farmer cultivated the promoted crop in the first year, and the rest of the group was expected to learn from his or her experience. In the second year, the rest of the farmers were encouraged to grow the crop. Farmers that adopted the technology in the second year received free seeds and some training sessions but no other inputs.

In addition, to test which demonstration plot approach produces the greatest increase in adoption within the farmers’ field school model, a randomized evaluation is being conducted in 220 villages in 2 districts. Villages have been randomly assigned to one of five treatment arms:

1.)   Regular demonstration plots (54 villages):These villages have the standard IAPP project activities beginning in the second year of the project  

2.)   Shared demonstration plots (56 villages):These villages also have the standard IAPP project activities beginning in the second year of the project. However, each demonstration package of seeds, fertilizer, and training have been shared by two to four group members having contiguous plots, as opposed to just one in the traditional/regular demonstration plot system.

3.)   Incentives for self-demonstration (54 villages):These villages also have the standard IAPP project activities beginning in the second year of the project. However, instead of demonstration plots, all farmer group members have been offered a chance to try out the new technology by themselves. The demonstration package was divided among all the farmers who were interested.

4.)   Short-term comparison (36 villages):These villages have the standard project activities beginning in the fourth year of the project. There was no project activity during the second and third year when the treatment villages were phased in for the “Demonstration Plot Evaluation” exercise.   

5.)   Long-term comparison (20 villages):These villages will have the standard project   activities in the fifth and final year of the project.

This study is evaluating the impact of the technology adoption program on the LFS members’ crop productivity, fisheries productivity, and livestock productivity. The study is also measuring the impact of IAPP on farmers’ overall income, income from agriculture, and food security in relation to farmers who did not receive the training. Furthermore, researchers are mapping the social networks within groups in order to understand how relationships affect technology adoption, evaluate any differences in how the program affects men and women, and conduct a cost-benefit analysis.

The evaluation is led by the World Bank’s Development Impact Evaluation Initiative and the South Asia Agricultural Development team, in collaboration with and external research partners, the Yale University School of Management and IPA Bangladesh.

Results and Policy Lessons:

Project ongoing, results forthcoming.


[1]U.N. Food and Agriculture Organization. “Food Security in South Asia.”

[2]U.N. Food and Agriculture Organization. “The State of Food Security in the World.” 2013

[3]U.N. Food and Agricultural Organization and World Food Program (WFP). “FAO/WFP Crop and Food Supply Assessment Mission to Bangladesh.” 2008.

Measuring the Demand for Aflatoxin Tested Maize in Kenya

Aflatoxin is a toxin produced by a fungus that grows on certain crops, such as maize and groundnuts. Consumption of high levels of aflatoxin can be fatal, and chronic exposure has been linked to liver cancer, suppressed immune response, and child stunting. Kenya has among the highest rates of aflatoxin exposure globally.1 This study evaluates whether there is sufficient demand in the Kenyan market for flour that has been tested for aflatoxin and how demand might be stimulated.

Policy Issue:

Aflatoxin is a toxin produced by the Aspergillus species of fungus. Consumption of high levels of aflatoxin can be fatal, and chronic exposure has been linked in numerous studies to liver cancer, suppressed immune response, and child stunting. Aflatoxin can be present in a variety of crops and animal products, but the consumption of maize and groundnuts is the most common source of exposure worldwide. While the developed world has reduced the risk of aflatoxin contamination from the human food supply through the use of preventive controls, including routine testing and channeling highly contaminated grain to less susceptible species, exposure remains a risk in many developing countries where maize is the staple food for humans.  More than 1.2 billion people in Sub-Saharan Africa and Latin America rely on maize as a staple crop.2

This “proof-of-concept” project will assess whether the Kenyan market will reward maize millers who implement a process control system to manage aflatoxin risk. This information will assist millers in deciding whether to invest in these improved practices. Demonstrating that a potential market exists for labeled process-verified flour is expected to catalyze provision of safer maize meal by the private sector.

Context of the Evaluation:

Maize is the staple for 96 percent of Kenya’s 40 million people,3 and is the primary source of aflatoxin exposure in that country. The legally allowable level of aflatoxin contamination in food for human consumption set by the Kenyan regulatory authority is no more than 10 parts per billion. However, one study by Center for Disease Control and Prevention (CDC) scientists showed that 65 percent of maize samples collected from 20 major millers did not meet the national standard.4 The Cereal Millers Association (CMA) of Kenya has expressed strong interest in improving food safety standards in the maize-processing sector and IPA is partnering with one of their member companies for this study.

The study is taking place within the current distribution network of the miller and includes Machakos, Kitui, Nairobi, Meru, Embu, Nyeri and Murang’a counties.  Machakos and Kitui have a particularly high incidence of aflatoxin contamination in maize.

Details of the Intervention:

The two-year study evaluates whether there is sufficient demand in the Kenyan market for aflatoxin-tested flour and how demand for the flour might be stimulated. It involves working closely with a miller and 72 small shops and supermarkets to produce and market a flour that has been tested, labeled, and verifed by a third party as aflatoxin-free.

Working with the Texas A&M Agrilife Research Office of the Texas State Chemist through the Aflatoxin Proficiency Testing for Eastern and Central Africa (APTECA) program, IPA will monitor collaborating millers’ compliance with the APTECA protocol. Texas A&M AgriLife Research will verify the accuracy of the aflatoxin test results at their accredited laboratory in Nairobi.

The research team will begin collecting sales data from 72 shops before the tested maize flour is officially launched and will continue tracking sales for the duration of the study.  One-third of shops in the study will display a poster explaining the aflatoxin-safe label, with no discounting or additional promotion. Consumer response to safety labeling of flour at these shops will inform researchers about the impact of such labeling under typical market conditions.

Two-thirds of shops, 48 in total, will be randomly assigned to receive leaflets promoting the tested flour for a week at the time the tested flour is launched. At half of these shops, leafleting will be repeated at four-week intervals (intensive marketing). Through this marketing intervention, researchers aim to evaluate the impact of information on consumer behavior over three distinct time periods: during, weeks after, and several months after an informational campaign. 

Half of the shops assigned to the marketing intervention (including half of those assigned to receive intensive marketing) will also be assigned to a discount group. These shops will be asked to offer customers a one-month discount for the tested, labeled, flour at the time the product is launched. This will allow researchers to measure the long-term impact of temporary discounts on consumer behavior over the longer-term. 

Customers at each participating shop will be interviewed before the intervention and nearby market where unbranded flour is sold and again at the end of the study period about their purchases. This customer-level data will complement the shop-level data on sales and shed light on the socio-economic profile of customers who purchase the tested flour in each treatment group.

Results and Policy Lessons:

Results forthcoming.

[1] Liu, Y. and F. Wu. (2010). Global Burden of Aflatoxin-Induced Hepatocellular Carcinoma: A Risk Assessment. Environmental Health Perspectives. 118(6): 818–824.

[2] IITA (2009). Accessed August 24, 2013 at: 

[3]Water Efficient Maize for Africa (2010).  Reducing Maize Insecurity in Africa: the WEMA Project.
Policy Brief of the African Agricultural Technology Foundation (AATF) and the Kenya Agricultural Research Institute (KARI). Accessed 17 February 2014 at

[4] Gathura, G. “Study finds 65 p.c. of flour unfit for eating.” Daily Nation. March 16, 2011. Accessed December 15, 2014 at:

Agricultural Microfinance in Mali

Evidence suggests additional investments in agriculture could increase income for subsistence farmers, potentially improving the livelihoods of millions of people. In rural Mali, giving some farmers unrestricted cash grants led to significantly higher productivity and profits, suggesting farmers would invest more in their farms if they had more capital. Providing farmers with an innovative loan product also led to a significant increase in farm investments and expenditures, suggesting agricultural loans tailored to farmers’ seasonal cash flow may be an effective way to increase investments in agriculture. In addition, this research suggests farmers vary in the returns they are able to generate from inputs, and agricultural loans attract clients with a better-than-average ability to grow their farms.

Policy Issue:

Agricultural productivity in Africa is very low despite the existence of modern agricultural inputs such as improved seeds, fertilizers, and pesticides. As a much of the population works in agriculture, encouraging the adoption of these technologies could raise agricultural productivity and in turn reduce poverty and encourage economic growth. But why do farmers fail to invest in these potentially profitable inputs? One reason may be that they do not have enough cash on hand when they need to purchase inputs and they lack access to credit. Microcredit organizations have attempted to address this problem, but the typical microcredit loan contract—where clients must start repayment after a few weeks—is ill-suited for agriculture. Providing farmers with loans at the beginning of the planting season, to be repaid in a lump sum at the time of harvest, could facilitate investment in inputs and increased profitability. However, there is little evidence on this type of microcredit.

Context of the Evaluation:

In Mali, 80 percent of people work in agriculture, most of them in subsistence farming. The sector accounts for about 37 percent of Gross National Product.1 This study takes place in the region of Sikasso, within the areas of Bougouni and Yanfolila, in southern Mali. In these areas, farmers grow cash crops like cotton, maize, sorghum, millet, and groundnuts.

Soro Yiriwaso, is a microfinance institution in Mali whose mission is to increase economic opportunities for poor Malians, especially women, by offering financial services. Soro Yiriwaso offers a loan product called Prêt de Campagne, or “countryside loan,” to women who join local community associations. Unlike most microloan products, it is designed specifically for farmers. The loans are dispersed at the beginning of the agricultural cycle between May and July and clients must repay loans on one lump sum immediately after the harvest. Soro Yiriwaso issues the loan to groups of women organized into village associations. Each individual woman then establishes a contract with the association for her loan.

Details of the Intervention:

To evaluate the impact of the Prêt de Campagne loan product on agricultural productivity and farm profits, researchers partnered with Soro Yiriwaso to conduct a two-stage randomized evaluation in 198 villages in rural Mali.

In the first stage of the study, Soro Yiriwaso offered their standard agricultural loan in 88 randomly selected villages. In these villages, only women who joined a local community association were eligible to receive loans. In the remaining 110 villages, no loans were offered.

In a second stage of the study, researchers offered grants in the 88 loan villages to a random subset of the households who chose not to take out loans.  The grants were worth 40,000 FCFA (US$140)— about the size of the average loan provided by Soro Yiriwaso and equivalent to around 70 percent of what average households spent on agricultural inputs. In the 110 villages where no loans were offered, households were also randomly selected to receive the grants. Similar to the loans, these grants were issued directly to a female household member.

Over a two-year period, researchers measured changes in farmers’ cultivated area, input use, and production output. They also collected data on food and non-food expenses of the household as well as on financial activities (formal and informal loans and savings) and livestock holdings.

Results and Policy Lessons:

Take-up and Use of Loans: About 22 percent of the women chose to accept the loan in treatment villages, which is a take-up rate similar to other microcredit products. Households offered loans increased investments and expenditures on agricultural inputs. Households in villages which were offered loans spent on average US$10.35 more on fertilizer and US$5.08 more insecticides and herbicides than the households in villages that did not get loans. Offering loans led to an increase in the value of agricultural output by US$32, and an increase in the value of livestock by US$168. The loans did not have a significant effect on profits, consumption, whether the household has a small business, nor educational expenses.  The return to agricultural investment varied across farmers, with more productive farmers generating higher returns to agricultural inputs.

Grants: In villages where loans were not offered, providing grants to women in households led to an increase in agricultural investments and, ultimately, profits. Households randomly selected to receive grants cultivated 8 percent more land and invested 14 percent more on inputs than households that did not receive grants. Output and farm profits among women who received grants also increased 13 percent and 12 percent, respectively.  

Self-Selection: Productive farmers—households that invested more in agriculture, had above-average agricultural output and profits, or had more agricultural assets and livestock than average before the program—were more likely to borrow and demonstrated higher returns to investment. Moreover, in the villages where loans were offered, households who opted not to take out a loan and instead received a grant did not generate the same returns as those in no-loan villages who were offered grants. This suggests that households that applied for loans were those with the high returns to capital.

Retention and repayment: The repayment rate among women who elected to take out loans was perfect and 50 percent of clients chose to borrow money again, which is on par with typical client retention rates for sustainable, entrepreneurially focused microcredit operations.

Overall, this research suggests that agricultural lending tailored to the farmers’ seasonal cash flow may be an effective way to increase investments in agriculture and improve yields and profits. Furthermore, the evidence that productive farmers are more likely both to apply for agricultural loans and to generate higher returns on the agricultural investments they make has important implications for credit markets. In short, farmers vary in the returns they are able to generate from inputs, and agricultural loans attract clients with a better-than-average ability to grow their farms. 

Related Paper Citation:

Beaman, Lori, Dean Karlan, Bram Thuysbaert, and Christopher Udry. Self-Selection into Credit Markets: Evidence from Agriculture in Mali. No. w20387. National Bureau of Economic Research, 2014.

[1] International Fund for Agricultural Development (IFAD). “Mali Statistics.” Rural Poverty Portal. Accessed: 4 March 2015.

Improved Sorghum Technology Adoption in Burkina Faso

Despite the availability of new agricultural technologies, which may increase yields and household income, few sorghum farmers in the Sahel region of Sub-Saharan Africa are using improved seeds and fertilizer. This study evaluates if providing seeds, fertilizer and training to sorghum farmers on “microdosing” increases productivity, and aims to identify the factors that prevent farmers from adopting new technology both on the demand side and the supply side.

Policy Issue:

Agricultural output has accelerated in Sub-Saharan Africa the last two decades, but agricultural productivity in the region remains low, especially compared to other regions of the world.1 Greater use of inorganic fertilizer and improved seeds are widely considered necessary to ensure African farmers boost production and farm profitably, but adoption of these technologies has been slow in the region.2  Previous research indicates that farmers invest in fertilizer when provided with small, time-limited discounts just after harvest, when they have income. To contribute evidence on the mechanisms that help people save enough money to buy fertilizer and seeds, this evaluation replicates this influential fertilizer study in the West African context with a fertilizer and seed, as opposed to only fertilizer. This study also investigates if people are more likely to adopt the technology if they receive it free of charge and how knowledge passes from farmer via social networks.

Context of the Evaluation:

Sorghum is the main food staple and most widely cultivated dryland crop among rural people of the West African Sahel. Despite the potential to attain over two tons per hectare with improved varieties, average sorghum yields in Burkina Faso are estimated at less than half that much. In addition, adoption of sorghum agricultural technology is far less than it is for rice, maize or specialty crops. To explore the reasons behind low adoption, and identify ways to promote higher adoption, researchers at Michigan State University partnered with IPA and the following organizations: AGRODIA, an agricultural non-profit based in Burkina Faso, and INERA, the Environmental Institute for Agricultural Research in Burkina Faso.

Details of the Intervention:

Researchers are carrying out a randomized evaluation in 165 villages in northern rural areas of Burkina Faso to evaluate the impact on farmers of providing a “micro-pack” of seeds and fertilizer, along with training on fertilizer “microdosing,” a technique that involves the application of small, affordable quantities of fertilizer using a bottle cap. Researchers are measuring impacts on the sorghum farmers’ productivity, and identifying the factors that prevent farmers from adopting new technology. Villages were randomly selected to one of the following seven groups. Households were randomly-selected in Group 1 only.

1) Free distribution of the micro-pack to randomly-selected households. (15 households in 80 villages)

2) Free distribution of the micro-pack, with households selected based on the degree to which a farmer is connected to other farmers. (15 households in 16 villages)

3) Free distribution of the micro-pack, with households selected based on the degree to which a farmer influences other farmers. (15 households in 16 villages)

4) Early commitment input fairs and training: Farmers were offered fertilizer at the market price and free seeds right after the harvest. (all households, 12 villages)

5) Late commitment market price input fairs and training: Farmers were offered the micro-pack at the market price of fertilizer, with free seeds, just before the planting season. (all households, 12 villages)

6) Late commitment subsidy input fairs: Farmers were offered the micro-pack with a 20 percent subsidy on the market price of fertilizer. (all households,12 villages)

7) Comparison group (20 villages)

Groups 1-6 are offered training in addition to the micro-pack.

The study design enables researchers to determine if and to what degree the micro-pack improves farmers’ productivity; estimate potential gains from adoption of new sorghum technology; and to estimate if targeting based on social network characteristics, in this case more connected or influential farmers, increases diffusion of fertilizer micro-dosing. In addition, the study design will allow researchers to determine whether the price of products or simply lacking cash (before the planting season) plays more of a role in preventing farmers from buying improved seeds and fertilizer. Furthermore, this study will measure different effects on male and female farmers and measure the presence of secondary effects of technology adoption, namely changes in household composition and adult and child labor supply.

Results and Policy Lessons:

Results forthcoming.

[1]Fuglie, Keith, and Nicholas Rada. "Resources, Policies, and Agricultural Productivity in Sub-Saharan Africa." USDA-ERS Economic Research Report 145 (2013).

[2]Druilhe, Zoé, and Jesús Barreiro-Hurlé. "Fertilizer subsidies in sub-Saharan Africa." Food and Agriculture Organization of the United Nations (2012): 2. 

Andrew Dillon

Environmental Investments on Private Land: Planting Trees in Chipata

Many climate change programs that target small-scale farmers seek to change farmers’ agricultural practices, whether to sequester additional carbon or to improve climate resiliency. Farmers are often hesitant to adopt new practices as they often entail high upfront costs. Climate change programs therefore generally provide inputs or incentives for adopting and complying with the program’s objectives. Yet it is unclear how much farmers respond to inputs and incentives, and to which they respond more favorably. This study in Zambia assesses the impact of two programs, one that provides incentives to farmers to grow a nitrogen-fixing tree, and another that provides inputs.

Policy Issue:

Financing for carbon offset investments is growing quickly. The voluntary market for carbon offsets traded over 700 million dollars worth of emissions reductions in 2008, a third of which came from land use projects.[1]  These investments have the potential to benefit smallholder farmers, not only by creating revenue from selling carbon credits, but also by incentivizing more climate-resilient agricultural practices and technologies to increase production. Many climate change programs that target smallholder farmers seek to modify current agricultural practices, whether to sequester additional carbon or to improve climate resiliency. Because these changes often impose costs on the farmer, many programs provide upfront inputs or incentives for adopting and complying with the program’s objectives.

However, in spite of a growing number of NGO- and government-led adaptation and climate resilience projects, farmer adoption remains a challenge and concerns persist due to the high cost of inputs, training and monitoring in comparison to the value of the credits earned from sequestered carbon. A more rigorous understanding of the relationship between input costs, compliance incentives and program outcomes may help improve the success and cost effectiveness of both carbon offsets and climate resiliency programs. To date, none of the numerous programs that offer landholders inputs or performance payments have systematically varied contract design to generate causal evidence on the determinants of program success.   This study proposes address this knowledge gap in the context of a program promoting fertilizer trees in Eastern Zambia.

The project implementation is designed to allow the researchers to investigate (a) the role of option value in shaping farmer decisions, and (b) the effect of cost sharing and performance incentives on selection into the project and on long-run performance under the contract.

Details of the Intervention:

The partner organization, Mitengo Zambia promotes a fertilizer tree (Faidherbia albida), locally known as the masangu tree, both for its carbon sequestration potential and its ability to help farmers adapt to a changing climate.  Faidherbia fixes nitrogen in its leaves, providing benefits to farmers, including better soil fertility, maize yields and resilience to climate change. To grow the tree, seedlings must be purchased and raised, and then planted among low growing crops, weeded and watered. These adoption costs are highest in the first year and tree survival is low.

Mitengo Zambia  has partnered with Dunavant Cotton to investigate the carbon sequestration and soil fertility potential of encouraging agroforestry adoption among Dunavant farmers. Findings from the research phase will be incorporated in program scale-up with Dunavant Zambia Limited, a leading cotton ginning company in Zambia.

Description of the Intervention:

Around 2,000 outgrower farmers associated with Dunavant cotton in Chipata, Eastern Province, Zambia receive training and subsidized inputs (seedlings) for growing Faidherbia on their land. Most Dunavant farmers produce on a small-scale, with a mean landholding size in the study sample of around one hectare, and have access to loans for cotton inputs from Dunavant.  The company organizes the farmers into groups of approximately 15 geographically clustered farmers. Each group has one lead farmer who, under the Dunavant system, is responsible for training his farmer group on cotton production and, under the Mitengo Zambia program, on Faidherbia planting and management.  

Lead farmers organize trainings on Faidherbia for their groups of farmers, which are attended by Mitengo Zambia and IPA staff who assist with administration of the treatments and the baseline survey.

After their training, farmers decide whether to join the program based on two factors:

(1) Variation in input prices – Farmer groups will be randomly assigned to receive one of four input prices that range from fully subsidized (free) to the cost-recovery price for the implementing organization (approximately $2.50 US). A transport allowance (of $2.50), provided to the farmers to remunerate any transportation costs of attending the lead farmer's training, ensures that farmers have enough cash to make a participation decision based on willingness to pay, not on liquidity constraints. Variation in input prices allows researchers to test hypotheses on risk and on cost-sharing. Specifically, how the probability of take-up changes as the input prices increase, controlling for individual characteristics and incentives, will be assessed.

(2) Variation in incentives – Individuals will be randomly assigned to receive different levels of incentive pay, which farmers are informed of either before or after making their take-up decision. The range of incentives is based on project pilots from the previous year, which ranged from $0 - $30 (0 - 150,000 ZMK). The use of scratch-off cards to reveal the incentives ensures that incentives cannot be manipulated and that the variation is perceived as fair by the participants.  Incentives will be paid after one year, conditional on 70% tree survival.  All farmers received 50 seedlings. The variation in incentives will allow researchers to test the causal effect of incentives, by comparing the probability of take-up and the rate of tree survival for farmers at different incentive levels, controlling for individual characteristics.

At the time of training, farmers receive a detailed baseline household questionnaire that includes modules on demographics, socioeconomic status, agriculture and environmental knowledge. The survey is administered to all farmers who attend the training, regardless of their decision to participate. One year after contracts are initiated, all participating farmers will be visited and the number of surviving trees recorded, an incentive payments delivered on the basis of tree survival.

Results and Policy Lessons:

Results forthcoming.

Read detailed findings from an in-country event with cross-sector stakeholders from the Zambian government, private sector, international donor and research community, and leading non-governmental organizations here.

[1]Conte, M. and M. Kotchen. Explaining the price of voluntary carbon offsets. Working Paper. (2009)


Examining Underinvestment in Agriculture: Returns to Capital and Insurance Among Farmers in Ghana

Risk, rather than a lack of capital, appears to drive underinvestment in agriculture in Northern Ghana - when farmers were provided with weather insurance they spent more on inputs such as chemicals, land preparation, and labor.

Policy Issues:

Underinvestment in agricultural inputs such as fertilizer, hybrid seeds, or labor is thought to drive low crop yields in Africa and other parts of the developing world. Several factors may help explain why farmers fail to invest in such potentially profitable inputs. It is possible that they are wary of the riskiness of adopting new agricultural methods or tools—if they invest and their crops still fail, they will have even less money than if they had not invested at all. Farmers may also lack the capital necessary to purchase these inputs, and be unable to obtain credit to finance investment in their farms. Because the returns to using new technologies can be so high, encouraging use among farmers has the potential to greatly improve their welfare, but financial institutions and policymakers need to first understand what factors are truly driving underinvestment in agriculture.


The climate of northern Ghana’s savannah region has a single short wet season, with high annual variation in rainfall. This kind of weather pattern creates great risk for farmers who depend on the weather for their livelihood, particularly when agriculture is primarily rain-fed, as it is in this area. There is strong evidence that shocks in the amount of rainfall translate directly into consumption fluctuations for farmers, and so investment in new agricultural technologies or methods has the potential to significantly affect welfare. Throughout Ghana, the average farmer uses only 7.4 kg of fertilizer per hectare, while in South Asia fertilizer use averages more than 100 kg per hectare. Initial surveys in northern Ghana revealed that the median farmer participating in this study did not use any chemical inputs on their crops, often citing lack of money or concerns regarding weather risk as key obstacles preventing investment.

Description of the Intervention:

In the first year of the study, researchers tested the relative importance of capital and risk in driving farmers’ investment behavior. From a total of 502 households, 117 were randomly selected to receive a cash grant to fund agricultural inputs; these farmers received GHC 60 (approximately US$45) per acre for up to 15 acres, delivered at a time of their choosing. Another 135 randomly selected households received a grant for an insurance scheme that paid roughly GHC 100 (US$75) per acre of maize if rainfall at a local weather station went above or below specified thresholds. Ninety-five households received both the cash grant and the insurance grant, while 155 households received no additional services and formed the comparison group.

In the second year, researchers tested different prices for rainfall insurance among the original sample households, plus households in an additional twelve communities. Households were visited up to four times by marketers: during the first visit they were informed about the product, during the second visit they were asked to sign the contract and pay premiums, during the third visit the marketer issued a policyholder certificate, and during a fourth visit an auditor verified their understanding of the product. The price that people were offered for insurance was randomly assigned at the community level: households in the original sample would be offered rainfall insurance at a cost of either 1 GHC or 4 GHC (approximately US$0.75 or US$3), while in the newly added communities, households would be offered insurance at either the market price of GHC 12-14, or the actuarially fair price of GHC 8-9.5.

In year three, the pricing experiment continued in collaboration with the Ghana Agricultural Insurance Programme (GAIP), to market their drought-indexed insurance. Because this product was more complex, scripts used at the four marketing visits were updated to make it more understandable. Pricing of the insurance was again randomized at the community level, with 23 communities receiving the market price, 23 communities receiving the actuarially fair price, and 26 communities receiving a subsidized price.


Importance of Capital vs. Risk: Results from the first year suggest that risk, rather than capital, was the major constraint on investment among farmers in this sample. Farmers who received the insurance grant increased their expenditure on farm chemicals, and also brought more acres of land under cultivation. If the primary constraint on investment was a lack of capital, then the insurance product, which offered no up-front payouts, would not have affected their ability to purchase materials. Many farmers appeared to recognize the value of the insurance product, with a significant proportion choosing to purchase insurance in years two and three. Take-up of the insurance product did not change when a token price of GHC 1 per acre was charged, and even at the actuarially fair price of almost GHC 10 per acre, 40-50 percent of farmers purchased insurance.

Impacts of Weather Insurance: Farmers with weather insurance invested more in agricultural inputs, particularly in chemicals, land preparation, and hired labor. Total cultivation expenditures were more than GHC 250 (US$188) higher for farmers with insurance, representing a 33 percent increase relative to the comparison group. These impacts were even larger among farmers who received both insurance and a capital grant. Despite the increases in production, it is not clear that investments were actually profitable for farmers: the additional expenditures may have increased by more than the value of the additional output, depending on how household labor is valued.

Trustworthiness of Insurance: Results suggest that how much farmers trust the insurance scheme has a large impact on their take-up and response to rainfall insurance. Take-up of insurance was considerably higher among farmers who also received a capital grant, but it was not higher among households who were wealthier. This suggests that farmers might not have been entirely confident that the promised insurance payouts would be made when trigger events occurred, and so they were more willing to take the risk of purchasing when they had been given extra cash. Similarly, individuals who were familiar with others who had received insurance payouts in previous years were significantly more likely to take-up insurance themselves.


Read about the new agricultural insurance product launch as a result of this study here, and about its' 2012 payout to farmers here.

Click here for a Q&A with researcher Chris Udry

The Impact of Food and Cash Loans on Smallholder Farmers in Zambia

Many farming households turn to off-farm work to make ends meet between harvests, reducing the amount of time they can invest in increasing their farms’ productivity. In ongoing work in Zambia, researchers are testing the relationship between scarcity, labor supply and agricultural productivity. Selected farmers receive access to either cash or food loans during they lean season that they are responsible for repaying at harvest. Researchers will track a variety of outcomes including yields, labor supply, food consumption and how impacts vary within the household and by loan type. In a pilot, already completed, food loans were shown to increase food consumption during the lean season, reduce the portion of households engaging in off-farm work, and increase wages. In the pilot, both take up and repayment rates were over 90 percent and results clearly established the relevance of hunger as a driver of labor supply decisions. 

Policy Issue:
Many farming households turn to casual day labor, typically on other better-off farms, to make ends meet between harvests, which often means that farmers do not have time to invest in  their own land. If farmers neglect on-farm activities that are important for production, such as weeding or application of fertilizer, then off-farm work can result in lower yields on farmers’ own fields. With a poor harvest, families are less able to set aside resources to last through the next season, increasing their reliance on off-farm work. Providing credit, either in the form of food or in cash, could allow farmers to spend more time on their farms. Farmers would not be forced to find off-farm income to feed their families between harvests, and would therefore be able to spend additional time applying fertilizer, weeding, or harvesting the crop, which would increase yields. In the long run, this gain in productivity might increase incomes by more than farmers could earn through casual labor. Although existing research looks at the impact of agricultural loans on crop productivity, this is one of the first studies to look at the impact of credit on how farmers allocate labor.
Context of the Evaluation:
Small-scale farming is the primary source of income in rural Zambia, and 72 percent of the work force is employed in agriculture. Most farmers are poor, and in Chipata District, where this evaluation takes place, the average income is less than US$500 per year for a household of six people. Sixty-three percent of households in rural Chipata are classified as “very poor” and almost all households lack electricity and piped water.
Zambia’s long dry season allows for only one harvest per year, which means that the harvest must generate income to last the entire year. Payments for input loans and other debts are often due at the time of the harvest, exacerbating the difficulty of setting aside resources for the next year. As a result, many households turn to off-farm, casual labor during the hungry season (January to March) to cope with short-term financial needs. In the study sample, 60 percent of young men reported engaging in casual labor during the previous season.
Description of Intervention:
Researchers are testing the effects of food and cash loans on labor supply and agricultural productivity. They have completed a preliminary, pilot study to establish the relationship between food shortages and labor supply. A larger study is ongoing.
Pilot study: In the pilot study, researchers tested the effect of  food loans on farming households’ food consumption, casual labor supply, and wages. The loan program offered households one 25-kilogram of ground maize flour per month during the lean season (January to March). Farmers were expected to repay the loan at harvest (in June) with three 50-kilogram bags of unground maize. Each bag of ground maize was worth about 45 Zambian kwacha (US$9) and each bag of unground maize was worth 50-65 kwacha (US$10-13), resulting in an interest rate of 11-44 percent over the loan period.
Researchers evaluated the loan program in 40 rural villages, each within a day’s drive of Chipata town and with 15-25 small-scale farmers. In 10 villages, all eligible households (those with 1-5 hectares of land) were offered the loan (“full treatment villages”). In 20 villages, households that expressed interest in the loan entered a lottery, through which half were selected for the loan program (“partial treatment villages”). Ten additional villages served as the comparison group and did not receive loans.
Households that took up the loan could pick up the maize flour at one of ten distribution centers in January, February, and March. To repay the loan, households brought maize for repayment to a central point in each village in June.
Full study: Over the next two planting seasons, researchers will compare the in-kind maize loans with cash loans and test the impact on farmers’ labor allocation and crop yields. As in the pilot study, loans will be disbursed in January and repayment will be due in June. Regardless of loan type, borrowers will be able to repay with either maize or cash. In order to measure how the effect of receiving loans persists over time, some villages will not receive loans during the second year of the full study.
Results and Policy Lessons:              
Pilot study: There was high take-up of the loans, and villages where loans were offered saw an increase in food consumption, a decline in the number of households engaging in casual day labor, and an increase in wages for those who did engage in casual labor. The impacts were larger in villages where all eligible households could take out a loan.
Take-up and repayment: In full treatment villages, 90 out of 104 eligible households (87 percent) took out a loan. In partial treatment villages, all 104 households that were offered the loan program took out a loan. Ninety-eight percent of borrowers in full treatment villages and 95 percent in partial treatment villages repaid their loans in full.
Food consumption:In full treatment villages, the probability of missing a meal in the previous week was 16.5 percentage points lower than in comparison villages, where 36 percent reported that a family member had to miss a meal. In addition, the number of meals eaten in the previous 24 hours increased by0.2, from a base of 1.7 meals in comparison villages. The loans had no significant effect on the probability of missing a meal or the number of meals eaten in partial treatment villages.
Labor supply and wages: The share of households working in casual labor declined in both full and partial treatment villages. In comparison villages, 60 percent of households engaged in casual labor during the agricultural season, but this figure was 9 and 8 percentage points lower in full and partial treatment villages, respectively. Wages also increased in villages where lean season casual labor fell.
Full study: The study is ongoing. Results are forthcoming.

Improving Measurement of Farmers’ Skills in Western Kenya

Farmers in western Kenya have limited knowledge of farm inputs suitable for the agro-ecological conditions they face, and use of agricultural inputs in the region is low. This study in western Kenya evaluates how farmer experimentation and learning about different inputs impacts farmers’ subsequent use of high-quality inputs and the productivity of their farms.

Policy Issue:

Rising world food prices and fears of food shortages have refocused attention of many policymakers on the challenge of increasing agricultural productivity in developing countries. While many parts of the developing world have experienced boosts in food production in recent decades, agricultural productivity has remained relatively low in Sub-Saharan Africa, which may be attributable to low adoption of new technologies. Understanding the reasons underlying this lack of adoption is key for the design of interventions or policies to increase agricultural productivity. This research focuses on information constraints as a possible barrier to technological adoption. Specifically, this research examines how giving farmers the opportunity to experiment and learn about new technologies affects future adoption and farm productivity.

Context of the Evaluation:

Farmers in western Kenya have limited knowledge of farm inputs suitable for the agro-ecological conditions they face, and use of agricultural inputs is low. Yet the opportunity to experiment with different products, such as parasite-resistant maize seeds, and to learn about integrated soil fertility management, may improve farming practices and increase agricultural productivity.

Integrated soil fertility management (ISFM) is a set of agricultural practices adapted to local conditions to maximize the efficiency of nutrient and water use and improve agricultural productivity. ISFM strategies center on the combined use of mineral fertilizers and locally available soil amendments (such as lime and phosphate rock) and organic matter (crop residues, compost and green manure) to replenish lost soil nutrients.

This study evaluates how farmer experimentation and learning about different inputs impacts subsequent use of high-quality inputs and agricultural productivity. The evaluation is being carried out in partnership with the International Institute of Tropical Agriculture (IITA) among 960 smallholder farmers in 96 villages in Siaya County, in rural western Kenya.

Through a public lottery, half of the villages were randomly assigned to serve as a comparison group and the other half were assigned to the treatment group. Ten households in each treatment village (five randomly selected and five selected by village members) were invited to participate in agronomical trials for three seasons with either maize, soy, or a soy-maize intercrop. The trials began in two waves, with half of the treatment villages starting during the long rains in spring 2014, and the other half starting during the short rains in fall 2014.

Farmers in the treatment group received free seeds, fertilizer and assistance with the preparation of demonstration plots of about 12 meters by 17 meters. These plots were subdivided into six subplots, of which one was a comparison plot.  Each of the treatment plots have a different set of fertilizers and seed varieties, chosen in such a way that it should be possible to discern the effect of each product separately and in combination with others.

Results forthcoming.


Karen Macours

Aflatoxin Exposure and Child Stunting in Kenya

Child stunting has been associated with exposure to aflatoxin, a toxin produced by a fungus that affects crops such as maize, groundnuts, and sorghum. However, the causal relationship between aflatoxin exposure and height-for-age child growth has not been demonstrated.  This project seeks to reduce exposure of aflatoxin through a maize-testing and swapping program, while assessing the effects of post-harvest and storage technologies, which may reduce aflatoxin contamination in home-produced maize.

Policy Issue:

Aflatoxin is a toxin produced by the Aspergillus species of fungus. It is present in a variety of crops and animal products, though the consumption of maize and groundnuts is the most common source of exposure worldwide. Chronic consumption can cause cancer of the liver, and consumption in large quantities may cause aflatoxicosis, which is fatal. What is not known is the impacts of chronic exposure on child growth.  While a few studies have found some association between aflatoxin exposure and child stunting, these studies were not randomized evaluations and results remain inconclusive. This research aims, therefore, to provide rigorous evidence on the causal relationship between aflatoxin exposure and child growth.

Context of the Evaluation:

While developed countries have largely removed the risk of aflatoxin contamination from the human food supply through the use of modern drying and storage systems and routine testing, exposure remains a risk in many developing countries where maize is the staple food.  More than 1.2 billion people in Sub-Saharan Africa and Latin America rely on maize as a staple crop. In Kenya, maize is the staple for 96 percent of the country’s 40 million people, 1 and is the primary source of aflatoxin exposure.

This project is taking place in Meru and Tharaka-Nithi counties in eastern Kenya. High levels of both child stunting and aflatoxin exposure have been found in this region making it an appropriate site to study whether reducing aflatoxin exposure will improve child growth.

Details of the Intervention:

This project attempts to removed aflatoxin-contaminated maize from the diet of young children in eastern Kenya and evaluates the impact that reduced exposure has on child growth. This project also identifies and promotes improved post-harvest and storage technologies and evaluates the impact on aflatoxin contamination of smallholder farmers’ maize. The study is taking place across 71 villages in Meru County and Tharaka-Nithi county in eastern Kenya.

The villages are randomly assigned to one of two treatment groups or a comparison group. The first treatment group is designed to examine the effects of improved post-harvest and storage practices on aflatoxin levels in maize stores. Households in this group will receive intensive training regarding the hazards of aflatoxin consumption, how maize and other crops become infected with aflatoxin, and which storage and post-harvest techniques can reduce aflatoxin levels in their food stores.2 Farmers in these households will also be given access to significantly discounted improved maize drying and storage equipment throughout the harvest season.

The second treatment group is designed to examine whether aflatoxin exposure impacts height-for-age child growth. Households in this group will be able to buy certified aflatoxin-safe maize provided by the study via a local stockist.3 They will also be offered testing of their household maize stores at least every two months over a period of two years, and will be given the opportunity to switch out any aflatoxin-contaminated maize for aflatoxin-safe maize. The linear growth of children born during the study period in these households will be compared to growth of those in the comparison group. Blood samples and measurements are collected from pregnant mothers at baseline, while child blood sample and measurements will be collected during endline. Blood sera will be analyzed for a biomarker indicating recent exposure to aflatoxin.

At the conclusion of the intervention, study results will be reported back to all villages; the research team will provide information regarding the most successful post-harvest and storage techniques, as well as the cost of those techniques.

Results and Policy Lessons:

Results forthcoming.


[1]Liu, Y. and F. Wu. (2010). Global Burden of Aflatoxin-Induced Hepatocellular Carcinoma: A Risk Assessment. Environmental Health Perspectives. 118(6): 818–824.

[2]Given the known health effects of aflatoxin exposure, an intervention benefitting only a subset of study households would raise ethical issues. The research team is therefore providing information on how to reduce aflatoxin exposure to all study households, both those in the intervention group and those in the comparison group.

[3]Aflatoxin-safe maize is contaminated at a rate below 10ppb, the government-designated threshold safe for human consumption.

Vivian Hoffmann

The Impact of Inventory Credit on Food Security and Rural Livelihoods in Burkina Faso

Rural finance is considered a key tool to reduce poverty, yet more evidence is needed on how to make financing work best for poor farmers. This study with farming communities in Burkina Faso evaluates the impact of inventory credit on agricultural production, food security and resilience, as well as on informal systems for lending, borrowing and saving.

Policy Issue:

Small-scale farmers in Africa often have little cash at hand and have difficulty accessing credit to invest in their farms. They typically sell their crops en masse right after harvest to meet their immediate cash needs – either to pay for school fees, medical expenses, or to reimburse informal high-interest rate loans taken from traders to invest in inputs in the early agricultural season. A few months later, supply shrinks and prices rise, but farmers typically lack enough grain to sell at that point and have little income to invest in agricultural inputs for the following season. An inventory credit system, or warrantage, which allows farmers to stock part of their harvest in a warehouse for several months, and use the bags as collateral for a loan if they choose, may give the smallholders the means to buy essential inputs for the next planting season and to hold on to their food until the lean season, when food stocks start to run low and prices climb. Storing grain for several months also has the potential to smooth yearly food consumption and improve food security. This research will complement existing work in the small but growing evidence base on inventory credit.

Context of the Evaluation:

Southwestern Burkina Faso is a region with high agricultural potential, where grain production is high, but where acute malnutrition and households debt is common. To improve livelihoods, three large farming cooperative unions and their partners organized an inventory credit system in 2007 as part of a European Union “Food Facility” project, which has since grown into a large agricultural cooperative, COPSA-C, that specializes in training producers and providing access to inventory credit from sensitization campaigns on villages to follow-up of stocks and warehouses management. COPSA-C is the main implementing partner in this study.

The rural credit system, or warrantage, guarantees farming cooperatives and/or their members credit when they store part of their harvest in a warehouse as collateral. The value of the agricultural products is expected to increase between the deposit and the withdrawal of the products during the lean season, enabling farmers to earn a higher value for their crops in addition to being able to access a loan if they choose.

Details of the Intervention:

Researchers are evaluating the impact of inventory credit on small-scale farmers’ income and crop decisions in two provinces of Burkina Faso, as well as impacts on food security and risk-sharing mechanisms.

After an informational campaign was performed by COPSA-C in 39 villages in the provinces of Tuy and Ioba, IPA collected information on the households’ composition, demographics, and interest in the program. A random sample of about 40 households from each village was drawn from producers who were interested in participating in warrantage, and a random sample of about 10 households from each village was drawn from producers who were not interested in participating. Interested households were randomly assigned either to the warrantage program or to a comparison group through a village-level public lottery.

Treatment intensity was also randomized across villages with one-third of interested households getting access to warrantage in 20 villages and two-thirds getting access in the remaining 19 villages.

Households participating in warrantage then stored their bags for a 6 to 8 month period. They could then take up a group loan from the commercial bank partner with the bags serving as collateral. The producers were given the opportunity to borrow up to 80 percent of the value of stored bags (a safety net in case of large depreciation of crops prices). COPSA-C followed up this phase by fixing the value of crops bags. To prevent default and promote investment in income-generating activities, COPSA-C advised producers during the storage period. In case of credit default, bags can be sold right after the storage period ends.

Results and Policy Lessons:

Project ongoing; results forthcoming.

The Impact of Contractual Partnerships on Small-Scale Rice Growers in Ghana

In Ghana, and in much of Sub-Saharan Africa, a large portion of the extreme poor live in rural areas and work in agriculture. Higher productivity in the agricultural sector could reduce extreme poverty, while improving food security and economic growth. This study measures the impact of an “out-grower scheme,” a contractual partnership between growers and agribusiness companies, on small-scale rice growers’ productivity, income, and quality of life, and intra-household dynamics. This study aims to contribute to agricultural development policy in Ghana and beyond, as well as to the body of evidence on how to close the gender gap in productivity.

Policy Issue:

Low agricultural productivity in Sub-Saharan Africa has led to increasing food imports and the loss of competitiveness for domestic producers. Spending on agricultural research and extension in Africa, especially regional research, remains low, and links between farmers and agribusiness are weak.[i] Agriculture and rural development have been identified as priorities for reaching the Millennium Development Goals,and regional bodies have formulated agriculture development plans that prioritize sustainable land and water management, access to markets, and the reduction of hunger. Individual countries, such as Ghana, have also implemented programs to address these issues. Additionally, while almost half of the agricultural workers in Africa are women, productivity on their farms is significantly lower per hectare compared to men. [ii] Addressing the agriculture gender gap will help drive economic growth across Africa, but more evidence is needed on how best to do this. This study evaluates a farm “out-grower scheme,” a contract between farmers and agribusiness companies with potential benefits to both parties. Evidence from the study aims to contribute to Ghana’s agricultural development program, and to similar programs in other parts of Africa and beyond.

Context of the Evaluation:

As part of Ghana’s strategy to increase food security and agricultural development, it has launched the Ghana Commercial Agriculture Project (GCAP). The program, which is funded by the World Bank and USAID, aims to promote agricultural productivity and livelihoods, specifically by linking small farmers into the agricultural value chain through encouraging large-scale commercial agricultural firms to establish outgrower schemes.[iii]  The study partner, a commercial rice producer, GADCO, is offering contracts to small-scale rice producers in the Accra Plains area. IPA and the World Bank are collaborating to evaluate the impact of the contracts on productivity, income, and general quality of life of the rice growers.

Thirty percent of the plot cultivators in this study are female. Of these, half are in male-headed households, and half are in female-headed households. And many of the rice plots being farmed in this study were distributed by the Government of Ghana to individuals who signed up for them, as part of the development of the Kpong and Weta Irrigation Schemes.

Details of the Intervention:

The study looks at the impact of the contracts, or out-grower schemes, on plot cultivators. Researchers will compare plot-cultivators who participate in the out-grower scheme to a comparison group of growers who do not participate.

Following completion of the baseline survey, 50 percent of the study sample farmers will be assigned to participate in the scheme, and 50 percent will be assigned to a comparison group that will not participate. The commercial rice producer will then offer contracts to farmers in the treatment group. A farmer who signs the contract receives inputs on credit for planting, fertilizer application, and crop protection as well as extension advisory services and mechanized harvesting. On harvesting the rice, GADCO buys the output, accounts for input costs, and then pays the farmer the difference between the value of output and input cost.

This evaluation will look at the impact of the contracts on farmers’ productivity, quality of life and intra-household gender dynamics. Researchers will measure, for example, crop production, yield, profits, sale price, extension, technology adoption, hired and family labor use, and market access. With regard to impact on the household, researchers will look at indicators such as labor and time use, employment, business activities, assets and household investments in education and health. In addition, the study will assess the impacts on male versus female farmers.

Results and Policy Lessons:

Results forthcoming.


[i] "Agricultural Development in West Africa: Improving Productivity through Research and Extension."World Bank, 2013. Available at:

[ii] “Levelling the Field: Improving Opportunities for Women Farmers in Africa.” World Bank and One Campaign, 2014. Available at:

[iii] "World Bank Approves $100 Million for Scaling Up Commercial Agriculture."  World Bank, 2012. Available at:

Selective Trials for Agricultural Technology Adoption and Experimentation

Farmers may vary in their suitability as experimenters with new agricultural technologies. For example, farmers likely differ in their beliefs regarding the possible returns of the technology, in their willingness to take risks with an unknown product or process, and in their social skills and willingness to share information with others. Researchers are investigating how important these differences are when it comes to increasing the use irrigation pumps in western Kenya, and whether subsidies for experimentation can be targeted based on certain skills or traits to more effectively increase technology adoption.

Policy Issue:

Local experimentation is believed to be a necessary condition for smallholder farmers to learn about the potential benefits and costs of a new agricultural technology. Governments and development organizations often indirectly subsidize experimentation through extension workers, whose demonstrations can at times substitute for local experimentation. Another option is to directly subsidize experimentation either by providing discounts on technologies to a few farmers per village. Little evidence exists, however, as to whether farmers vary in their suitability as experimenters. That is, farmers likely differ in their beliefs regarding the possible returns of the technology, in their willingness to take risks with an unknown product or process, and in their social skills and willingness to share information with others. Researchers are investigating how important these differences are when it comes to spreading the use of new technologies, and whether subsidies for experimentation can be targeted based on certain skills or traits to more effectively increase technology adoption.

Context of the Evaluation:

This study takes place in Busia and Bungoma counties in rural western Kenya, among small-scale farmers who farm on plots 20 meters or less from a water source. Even during long dry spells, there is low adoption of efficient irrigation technologies among these farmers. Many farmers use “bucket irrigation”—carrying buckets of water from a well, river, or other water source to their fields—which is extremely time-consuming and limits the area that can be irrigated. In Busia county, only 3 percent of farmers in the study sample were using a manual pump when the study began, and none had a motorized pump.

Details of the Intervention:

Researchers are conducting a multi-arm, multi-stage, randomized evaluation to study experimentation with, and adoption of, small-scale irrigation pumps among smallholder farmers in western Kenya. This study aims to determine if significant differences exist across farmers in experimental skills; if these skills are known to farmers, and/or to their neighbors; if information about these skills can be elicited; if targeting based on experimental skills promotes information creation and diffusion; and whether better information diffusion leads to faster spread of technology. IPA field officers are collecting all the data for this study.

A total of 191 villages, each with an average of 23 participants, were randomly assigned to either a pure comparison group or one of four experimental arms. In all arms (including the comparison), farmers were told that a study to facilitate experimentation with a new technology in the neighborhood was occurring, and each farmer received a voucher for a slightly discounted pump, redeemable locally within 6-9 months.  

Each of the four experimental arms is a randomized evaluation itself, with one or two individual farmers per cluster assigned to treatment—access to a free pump—while the rest of farmers have access to pumps for cash at a small discount. Thus, within each group, researchers can estimate the magnitude of the “first stage” effect—the difference in short-run adoption and experimentation rates between free pump recipients and non-recipients.  The differences in the magnitude of the first stage across these groups will be one of the outcomes of interest – it will tell researchers how the mechanism used to select free trial recipients affects take-up of the free trial itself.

The experimental groups vary in how the experimentation promotion is introduced into clusters. Following an initial survey, on average 1.4 farmers per cluster are chosen as “experimenters” to receive a free pump. These experimenters are either selected completely randomly, as in a standard lottery with equal odds, or through what we call a “selective trial,” where the odds of being selected increase with a participant's likely experimental skills. Three different selective trials are being tested, each corresponding to a different way of eliciting participants' proclivity for experimentation with the technology: one is based on whether farmers are willing to pay for a higher chance of winning the free pump (in cash or grain), one based on whether farmers are willing to work for a higher chance of winning the free pump, and one based on how many votes farmers receive from their neighbors when their neighbors are asked to vote for the best experimenter.

There are two main sets of outcomes of interest. The first is the rate of pump usage among those farmers sampled to receive the free technology sample, as well as changes in yields and income. The second and main outcome of interest is the village-level rate of pump usage among other farmers (those not sampled for a free pump). Knowledge of how to use the pump and beliefs about the returns to using a pump will also be measured for all farmers.  These outcomes will be measured through follow-up surveys to be started in February 2015.

The project is being implemented in two waves. Wave 1 began in November 2013 with approximately 1,700 participants in 85 villages in Busia county. Wave 2 began nine months later among 2,400 participants in 106 villages in Bungoma counties.

Results and Policy Lessons:

Results forthcoming.



Land Leases to Semi-Nomadic Herders in Peri-Urban Areas of Mongolia

In the last two decades, large numbers of Mongolian herders have migrated to the outskirts of the country’s major cities, which has led to conflicts over land and  overgrazing of common pastureland. Herders may change their herding practices to better sustain the land if they own rights to it, which could also translate into bigger and healthier animals, and more income for the herders. In this study, carried out near two cities in Mongolia, researchers evaluated the impact of private property rights on land use and herder income.

Policy Issue:

Animal herding has been the major source of income and a way of life for rural people in Mongolia for thousands of years, and nomadic and semi-nomadic herders still make up a third of the population.1 Yet since the transition to a market economy in the 1990s the number of livestock in Mongolia has more than doubled, putting a strain on the common use grasslands in peri-urban areas. Overgrazing has led to severe degradation of the rangeland on which these herders depend. Property rights may help address the problem of land degradation as people may be more motivated to conserve the renewable resources on land they have exclusive use. If herder are granted long-term rights to the land, including the ability to exclude use by other herder groups, they may have greater incentives to reduce over-grazing and make long-term investments in the land and their herds. More productive pastureland could also translate into bigger and healthier animals, and increased income for the herders. This evaluation in Mongolia will contribute evidence on this topic.

Context of the Evaluation:

The Millennium Challenge Account-Mongolia, the partner in this study, launched the Peri-Urban Rangeland Project (PURP) to conserve pastureland in peri-urban areas by granting exclusive-use leases to herding groups. Since private ownership of land is new in Mongolia, part of the goal of the five-year project was to serve as a proof of concept that private pastureland rights can lead to sustainable land management and improved outcomes for the herders. The project, which consists of leasing rights, material support, and training,targets semi-nomadic herders groups who lived in areas surrounding two Mongolian cities, Choibalsan and Kharkhorin.

Details of the Intervention:

Researchers measured the impact of the Peri-Urban Rangeland Project (PURP) on herder income and land quality (as measured by pastureland productivity) over a two-and-a-half-year period.

Through a lottery, 165 herder groups—with two to seven households in each group—were randomly assigned to receive exclusive grazing rights to a section of pastureland for fifteen years. The same number of herder groups were also randomly assigned to serve as a comparison group. Herders selected to receive exclusive grazing rights were also offered material support in the form of well construction, seeds for growing fodder plants, and materials for building fencing or animal shelters. The material assistance was donated up front, but approximately 50 percent of the cost was required to be repaid by the group over a period of 15 years. The groups also received free training on veterinary care, productive breeds of cows, sustainable pastureland management, and proper preparation for harsh winters.

Researchers measured the impact of the PURP program on herder income and land productivity. Productivity was measured using grass clippings at different times of the year.

Results and Policy Lessons:

Results forthcoming.                

Building Market Linkages for Smallholder Farmers in Uganda

Prices of staple foods like maize, beans, and rice vary substantially in Sub-Saharan Africa, depending on the season, country, and region. Addressing the imbalance in food supply and increasing farmer income may require a multi-pronged approach that tackles multiple barriers at once. Researchers will evaluate the impact of contract farming services and a mobile technology-enhanced trader alerts system on food markets across Uganda.

Policy Issue:

Prices of staple foods like maize, beans, and rice vary substantially in Sub-Saharan Africa, depending on the season, country, and region. Some countries face food shortages when there is ample supply in the country as a whole, and the same is often true on a regional level.1 The inability of markets to efficiently move food from surplus to deficit regions has a major effect on both farmer incomes and food security. Seasonally, farming families tend to sell when prices are low in the months following the harvest, and buy when prices are high in the months preceding the harvest. Within countries, some fertile and rainy regions experience a food surplus while other regions are chronically in crisis. This imbalance in food supply is often attributed to poor roads and infrastructure, as well as high transportation costs, but other factors likely hinder efficient market integration. Small-scale farmers often lack accurate or sufficient information about prices in other geographical areas, or any guarantee that they will be able to sell their goods elsewhere. Yet, research suggests that better information alone is not enough to persuade farmers to move their food surpluses to areas where there are deficits.2 Addressing the imbalance in food supply, and simultaneously increasing farmer income, may require a multi-pronged approach that tackles intermediation, information, and contracting barriers at once.

Context of the Evaluation:

A large portion of people in Uganda face food insecurity, though rates vary regionally. According to a 2013 report, almost half of Ugandans are food energy deficient.3 Over a third have low dietary diversity, but that number reaches 55 percent in the western region—a region also burdened with the highest rates of childhood stunting.     

AgriNet, Ltd. the main project implementer, is the largest private-sector market brokerage firm in Uganda. AgriNet signs contracts with major buyers and then organizes supply from farmers. The company also has a “trader alerts” system, which feeds basic market information to farmers to help them bulk and sell at higher prices than are available on spot markets.

Details of the Intervention:

Researchers will carry out a randomized evaluation in 230 markets in 110 sub-counties in northeast, western, and central Uganda. The aim is to evaluate the impact of AgriNet’s mobile technology-enhanced trader alerts system on farmers and intermediaries’ profits, intermediaries’ method of connecting with farmers, additional market linkages, and price dispersion. Three institutions—AgriNet, Kudu, and IPA—will collaborate to implement and evaluate the program. 

The study team will randomly assign half of the sub-counties to the treatment group and half to the comparison group. In each of the 130 markets in the treatment group, AgriNet will recruit, train, and certify four commission agents.These agents, recruited from a pool of local traders, will actively link farmers to large buyers and operate the enhanced trader alerts system. The amount of grain farmers have for sale will be uploaded onto the trader alert platform and broadcasted to potential buyers and vice versa. Kudu, the mobile application integrated in the trader alerts, will serve to electronically bulk the grain, allowing buyers to view what is available for purchase at the village-level. In addition, traders will send information on local market prices to Kudu via SMS. Kudu will post this information daily, enabling farmers to sell when prices are most favorable and to negotiate for the prevailing market price. 

In the 55 sub-counties that serve as a comparison group, AgriNet will not execute any activities for the duration of the study.

To evaluate the impact of the project, IPA will conduct three types of surveys—an initial survey of farmer households, a survey of traders, and a market price survey—in both treatment and comparison markets. The trader survey will track the revenue, income, volumes, crops, and locations of trading for five buyers in each studied market as well as for commission agents in the study.  The agricultural household survey will be administered to 11 households in each catchment area that are active in agricultural markets (i.e. not subsistence farmers) around study markets, for a total of 3,000 households.

In addition, the market price survey will measure the impact of the project on price dispersion. For each market, IPA will recruit two traders, who are aware of market prices, train them to use the SMS survey system, and pay them (with mobile phone credit) to answer a detailed market survey every two weeks.

Results and Policy Lessons:

Results forthcoming.


[1] Feed the Future, East Africa Regional Profile.

[2]Fafchamps, Marcel, and Bart Minten. "Impact of SMS-based agricultural information on Indian farmers." The World Bank Economic Review 26, no. 3 (2012): 383-414.

[3] World Food Program; Uganda Bureau of Statistics (UBOS). “Comprehensive Food Insecurity and Vulnerability Analysis, Uganda,” April 2013.

Tracking Agricultural Households in Sierra Leone

Sierra Leone’s government made agriculture its top priority beginning in 2008, but policymakers lacked information about the status of the sector. In August 2009, the Government of Sierra Leone commissioned a large-scale survey to obtain accurate and credible agricultural data that could serve as a baseline for years to come.

Policy Issue:

Agriculture is the backbone of Sierra Leone’s economy. Agricultural production amounted to about 57 percent of Sierra Leone's GDP in 2011.1 In 2008, the Government of Sierra Leone made agriculture its top priority. Yet, policymakers lack information about the status of the sector, potentially inhibiting effective policy formulation, planning, implementation and performance evaluation. Therefore, in August 2009, the Government of Sierra Leone requested and supported an independent exercise, the Agricultural Household Tracking Survey (AHTS), to obtain accurate and credible agricultural data that could serve as a baseline for longitudinal analyses of Sierra Leone’s progress in agricultural development over the next several years. AHTS was commissioned and overseen by the Office of the President, and implemented collaboratively by the Ministry of Agriculture, Forestry, and Food Security, IPA/J-PAL.

[This project was not a randomized controlled trial.]

Details of the Intervention:

The main goal of the AHTS was to provide accurate and credible information about the agricultural sector in Sierra Leone from the household level. The AHTS was designed to capture decision-making by farmers, data on all aspects of cultivation from planting to harvest, and their access to services, technology, markets and infrastructure.

First, 920 areas were randomly selected out of 9,671 areas used in the 2004 census. Within each area, 10 agricultural households were selected to be interviewed. The total target sample size for the survey was 9,030 agricultural households. Twenty surveyor teams of five people collected data between March and May 2010. Field monitors, district coordinators, and ultimately, the AHTS technical team supervised the data collection.

Results and Policy Lessons:

The survey results confirm much existing knowledge about agriculture in Sierra Leone, for example, the national importance attached to rice and cassava, the concentration of cacao and coffee in the East of the country, and the existence of a seasonal hungry season. The results also point to various challenges faced by farming households. If these can be addressed, there is potential for large-scale improvements across the country. The results suggest that policy interventions in the agricultural sector should be intensified and expanded, particularly in the following areas:

  • Technology adoption:Levels of fertilizer use adoption of improved seed varieties were low, particularly for the main staples, rice and cassava;
  • Access to markets:Households' access to and interaction with markets remains low. For example, 92 percent of sampled households reported that their main point of sale for threshed rice was at farm gate (64 percent for clean rice). AHTS communities reported an average distance of 6.6 miles to the 
nearest market and 8.8 miles to the nearest permanent market.
  • Rural infrastructure:49 percent of farmers harvesting cereals stored their cereals in a room inhabited by the household; only 13 percent of households used a cement floor for drying. Out of 880 communities surveyed as part the AHTS community module, 25 percent reported a walking distance to the nearest road of more than 30 minutes during the dry season.
  • Financial access: The majority of farmers with existing loans borrowed on the informal market. Sixty-eight percent of households who did not borrow money said they had nobody to apply to for a loan.

[1] World Bank. “Sierra Leone at a Glance.”


Contractual Partnerships for Bean Growers in Central Uganda

Small-scale farming accounts for over 90 percent of agricultural output in Sub-Saharan Africa, and agricultural productivity on these farms is low, on average.  Contractually linking farmers to buyers may improve farmer profits and stimulate economic growth, but more evidence is needed on how these agreements impact farmers’ livelihoods and the crops they grow. This pilot evaluation with bean farmers in Uganda measures the impact of contracts between farmers and buyers on income, beans sold, inputs used, and a variety of other outcomes.

Note: This is a pilot study.

Policy Issue:

Small-scale farming characterizes agriculture in much of Sub-Saharan Africa. More than two-thirds of holdings are less than one hectare on average and these small plots produce over 90 percent of agricultural output.1 Agricultural productivity in Africa is, on average, low.2 Transitioning from low productivity, semi-subsistence agriculture to high productivity, commercialized agriculture has been a core theme of development. Since poverty is concentrated in rural areas among small-scale farmers, reducing overall poverty is thought to start with these farmers. Beyond increasing farm incomes, growth in agricultural productivity may also stimulate linkages to the rural economy outside the agricultural sector, causing further economic growth.3

Programs that contractually link farmers to buyers have become popular in the development sector, as such programs may reduce intermediaries and increase farmers’ profits. However, more evidence is needed regarding the impact of these contracts on farming practices and farmers’ livelihoods.

Context of the Evaluation:

In Uganda, agriculture employs 70 percent of the labor force and accounts for 20 percent of GDP, making it the backbone of the economy.4 Ninety two percent of the poor live in the countryside, and 89 percent of the population is classified as rural.5  In 2010, Uganda was ranked second in production of beans after Tanzania in the East Africa.

Myanzi Area Cooperative Enterprises, a partner in this evaluation, is a Uganda-based agricultural cooperative that works to create market linkages for rural producers, gather and disseminate market information, bulk agricultural products, and promote technology use to increase production, among other objectives.

Details of the Intervention:

This pilot study aims to determine how access to improved markets, by contractually linking bean farmers to bulk buyers, impacts farming practices. Data collection for the study will be accomplished through household surveys carried out before contracting and after the harvest.

The study includes 500 bean farmers in the Central Region of Uganda randomly assigned to a treatment group or control group. The treatment households will receive a visit from project staff encouraging them to sign up for the contract that links farmers to the bulk buyer.

If a large enough proportion of the households in the encouraged or treated group sign up for beans contract, the analysis will proceed by comparing income, beans sold, inputs used, and a variety of other outcomes, on the group that signs up for the contracts relative to the one that is not offered the contracts. Differences in outcomes across groups constitute unbiased estimates of the effect of the intervention on outcomes at “complying” households, i.e., households that were motivated to apply for the program by the encouragement.

Results and Policy Lessons:

Results forthcoming.

[1] International Food Policy Research Institute. “Ending Hunger in Africa: Prospects for theSmall Farmer.” p. 2

[2] International Center for Trade and Sustainable Development. “Raising agricultural productivity in Africa: The energy challenge.” May 1, 2008.

[3] Haggblade, Steven, Peter Hazell, and James Brown. "Farm-nonfarm linkages in rural Sub-Saharan Africa." World Development 17, no. 8 (1989): 1173-1201.

[4]Feed the Future. Uganda, country profile.

Muthoni Ngatia

Hidden Information and Market Failures for Crop Insurance in the Philippines

Farmers across the developing world face risk from hazards such as weather, pests, and crop disease, but largely lack insurance to manage these risks. One reason for this lack of viable insurance products may be that farmers know their plots and risks better than insurers, and react accordingly. In the Philippines, researchers offered insurance on randomly assigned plots to farmers, and found that farmers preferred to insure the plots that faced more risk. Farmers also invested less in fertilizer for insured plots, and those plots suffered from more preventable (pests and crop disease) than natural (flood and typhoon) damage. The findings suggest that information that is unavailable (at least in part) to insurance providers – the susceptibility of agricultural plots to damage and farmers’ effort to prevent damages – may be a substantial barrier to the functioning of crop insurance markets. 
Policy Issue
Farming households in developing countries face enormous risks from natural hazards such as pests, droughts, and floods. Not only are these constant threats to farmer livelihoods, but they can limit farmers’ access to credit, and lower their investment into planting. Despite the prevalence of these risks, insurance markets have largely failed to emerge, potentially because of information problems in the markets. 
This study examined two possible causes for this failure, associated with information asymmetries, or hidden information, when farmers know more about their plots and investments than insurers. The first, adverse selection, can occur when farmers choose to buy insurance on plots that are most likely to suffer losses (e.g., low-lying, flood-prone plots), but the associated risk characteristics of these plots may be difficult for the insurance company to identify. The second, moral hazard, is a result of insured farmers applying less effort to taking care of their farm when aware that it is insured against certain risks. 
This study was designed to test for these possibilities by using randomization to offer partial insurance to groups of farmers, first eliciting their preferences for which plots to insure, then measuring if having insurance changed the level of investment into different plots. Understanding how these asymmetries affect farmer behavior may be a crucial part of developing functioning insurance markets for farmers in the developing world.
Context of the Evaluation
Rural poverty accounts for 80 percent of the Philippines’ overall poverty rate, and agricultural production remains a major source of livelihood for the rural poor. Recognizing vulnerability to risk as a major constraint to agricultural productivity and improved welfare, the Philippine government created the Philippine Crop Insurance Corporation (PCIC) in 1989 to provide a “multi-peril” crop insurance product for rice and corn farmers, designed to help ameliorate the consequences of the many agricultural risks posed by typhoons, floods, droughts, and various pests and crop diseases. In a country that is affected by about 19 typhoons in a typical year, the need for such insurance is especially pertinent. However, take-up remains low, and no private market for crop insurance has developed. 
The study took place in the Bicol region of the Philippines, situated among a “typhoon belt” where risks to agricultural production are particularly high. The research focused exclusively on rice cultivation, which is the major agricultural activity of the region. 
Details of the Intervention
Over three consecutive farming seasons (two dry seasons and one wet season), the research team invited farmers tilling at least two irrigated rice plots to participate in the study. Participants could then enter a lottery where there was about a 70 percent chance that at least one of their plots would receive free crop insurance for that season. To see if farmers preferred to have certain plots insured over others, before the randomization, the farmers were asked to rank their plots in order of priority to receive insurance. Each farmer’s first-choice plot was allowed a slightly higher chance of receiving free insurance to motivate them to rank plots in accordance with their actual preferences. Farmers were then entered into a lottery and randomly allocated to one of three groups:
Group A (66.5%; Full Randomization): Received insurance on a random half of their plots.
Group B (3.5%; Choice): Received insurance on a first-choice plot and a random half of remaining plots.
Group C (30%; Comparison group): Received no insurance. 
The product paid out per hectare of insured land in proportion to the share of harvest lost to specific causes. As most farmers had limited or no experience with crop insurance, a member of the research team explained the product and claims process in person while providing informational materials, including an explanatory comic strip. The research team also followed up with insured participants through in-house visits and text message reminders throughout each season. 
Independent surveyors conducted three surveys throughout each farming season: a baseline before the randomization, a follow-up survey after planting, and an endline survey after harvest. 
Damages from preventable causes (i.e. pests and crop diseases) were higher on randomly insured plots by about 25% when compared to non-insured plots. In contrast, there was no difference in typhoon and flood damage across insured and non-insured plots of the same farmer. This difference suggests that farmers may respond to risk-mitigating insurance cover by reducing effort to prevent damages (i.e. moral hazard). 
The findings also suggest that farmers used less fertilizer on insured plots. Taken together, these findings suggest innovations in monitoring that limit moral hazard while keeping transaction costs low may be necessary for insuring against preventable kinds of losses.
Moreover, farmers preferred insurance on plots that were low-lying and prone to floods, characteristics that are mostly unobserved by the insurance company. The plots selected by the farmers to be prioritized for insurance coverage had 20% higher damages from flooding than other plots in their portfolios, suggesting the presence of adverse selection as farmers preferred to insure their more vulnerable plots. This finding suggests that insurance companies may need better assessment tools to help them identify risk characteristics (e.g. more precise altitudinal data to assess flood risk) and vary premiums based on this information. 
The identification of both moral hazard and adverse selection at work help explain why insurance markets have failed to develop and point the way to creating more viable products for farmers in the developing world.

Understanding Technology Adoption: Fertilizer in Kenya

Why do so few people use fertilizer even though it can considerably improve yields? This project measures the increase in yield due to fertilizer and hybrid seed use in Western Kenya. It found that fertilizer is profitable, and providing information goes part of the way towards increasing fertilizer adoption. Programs that help farmers commit when they have money to use fertilizer in the future have a very large impact on fertilizer adoption. 

Policy Issue:

By some estimates, approximately 1.4 billion people lived on less than $1.25 a day in 2005, and many of the poor are farmers. Identifying ways to increase agricultural incomes is crucial to alleviating poverty. Such strategies are especially important in sub-Saharan Africa, a region that has suffered decades of decline in per capita food production.

Context of the Evaluation: 

An estimated 50 percent of the population of Kenya’s Western Province lived below the poverty line around the time of this study, which often means they are unable to afford enough food to meet their basic calorie requirements as well as their non-food needs.The majority of Kenyan subsistence farmers grow maize as their staple crop, but many have only small amounts of land and are actually net buyers of maize, purchasing it when their own supply runs out immediately before a harvest. In such an environment, improving agricultural productivity could substantially benefit the farmers’ livelihoods. A potentially important input into increasing productivity is chemical fertilizer. Numerous agricultural trials on experimental farms suggest substantial returns to fertilizer, and improved fertilizer use has been associated with the increase in agricultural incomes during the Green Revolution in South Asia. However, only 37 percent of sampled farmers in the Busia district of Western Kenya report ever having used fertilizer. 

The overall goal of this research program is to understand why farmers do not invest in fertilizer. This part of the project first investigates whether the returns to fertilizer are actually substantial on real-world farms in real conditions.

Details of the Intervention:

In collaboration with International Child Support (ICS), an NGO, researchers set out to experimentally measure the returns to fertilizer amont area farmers. Farmers were selected from lists of parents at local schools, and ICS paid for fertilizer and hybrid seeds, delivered materials, helped these farmers apply fertilizer and seeds, and assisted them with the harvest. On each farm, a comparison plot was kept directly next to treatment plots, which was farmed using traditional methods. The type of seed and amount of fertilizer applied to each plant was varied by plot (see below), but farmers were instructed to tend all plots exactly the same.



Time of Application

# of Plots


¼ tsp Calcium Ammonium Nitrate

2 months after planting



½ tsp Calcium Ammonium Nitrate

2 months after planting



1 tsp Calcium Ammonium Nitrate

2 months after planting



Hybrid Seeds

1 tsp Di-Ammonium Phosphate

1 tsp Calcium Ammonium Nitrate

(the “full package” recommended by the Kenyan Ministry of

At planting

At planting

2 months after planting



Results and Policy Lessons:

Impact on Crop Yield: All fertilizer treatments led to increases in yield, though in different amounts. Interventions A, B, and C led to yield increases of 28 percent, 48 percent and 63 percent respectively, relative to comparison plots. Intervention D, the Ministry of Agriculture recommended package, led to an average 91 percent increase in yield relative to comparison plots. These increases in yield are generally consistent with the results obtained in experimental farm trials.

Rates of Return:  Rate-of-return calculations suggest that Intervention B is highly profitable, with mean returns of 36 percent over a season and 69.5 percent annualized. Implementing Intervention B on a 0.93 acre area of maize cultivation (the average acreage under maize cultivation in this sample) would increase agricultural income net of fertilizer cost by about 1,100 Kenyan shillings (US$33 PPP) compared to traditional methods—this represents a 15 percent increase in net income and more than a month’s agricultural wages. This evidence demonstrates that fertilizer use can have substantial returns, even in the absence of any changes in other farming practices, on real-world farms. However, other levels of fertilizer use, including the official recommendations of the Ministry of Agriculture, are unprofitable for the average farmer in this sample. Thus, while fertilizer can be very profitable when used correctly, one reason why farmers may not use fertilizer is that the official recommendations are not adapted to their specific context. This also suggests that fertilizer is not necessarily easy to use correctly, and may not be profitable for many farmers who do not use the right quantity.


1 Shahua Chen and Martin Ravallion (2008). “The Developing World Is Poorer Than We Thought, But No Less Successful in the Fight against Poverty,” World Bank Policy Research Working Paper #4703.
2 National Coordinating Agency for Population and Development (NCAPD) [Kenya], Ministry of Health (MOH), Central Bureau of Statistics (CBS), ORC Macro. 2005. “Kenya Service Provision Assessment Survey 2004”. Nairobi, Kenya: National Coordinating Agency for Population and Development.

Selected Media Coverage:

Savings, Subsidies and Sustainable Food Security in Mozambique

When smallholder farmers see how fertilizer increases their yields, they may continue using it. In this study in Mozambique, where very few farmers use agricultural inputs, researchers evaluate if giving farmers fertilizer subsidies encourages them to continue using fertilizer when subsidies run out. This study also measures the impact of coupling the subsidies with different types of savings accounts. Do subsidies, savings accounts, a combination of both, or none of the above, lead farmers to invest in their farms, grow more food, and earn more income?

Policy Issue:

Motivated by the recent escalation in food prices around the world, several countries, including Kenya, Malawi, Rwanda, and Zambia, have implemented large-scale fertilizer subsidy programs to boost food security and small farm productivity. If people are unaware of the benefits of using fertilizer, or do not know how to use it, then subsidies may be a useful tool to give people experience with using fertilizer, and promote adoption. However, a long-standing question is whether one-time or temporary provision of subsidized fertilizer can get households to adopt it long-term, or whether input use and farm production eventually return to previous levels after subsidies are phased out. The key to determining whether provision of subsidies can lead to long-term growth, even after the subsidies are no longer in effect, is to discover if farmer practices change fundamentally, or whether these practices change only (if at all) when subsidies are being offered.


Large-scale emigration, economic dependence on South Africa, and a prolonged civil war hindered Mozambique’s development until the mid 1990s. Agriculture accounts for almost 29 percent of the country’s GDP, however agricultural technology adoption has been slow in Mozambique compared to other counties in the region. Most of the farmers interviewed for this study had little or no experience with application of chemical fertilizers and other agro-chemical inputs.

Description of Intervention:

Researchers are investigating the impacts of fertilizer subsidies on smallholder farmers in rural Mozambique, and in particular, whether providing farmers opportunities for savings accounts can help subsidies achieve a greater sustainable impact. Vouchers for fertilizer were distributed randomly to a sample of farmers. In partnership with Banco Oportunidade de Moçambique (BOM), researchers also randomized offers of one of several different savings accounts interventions, to see how the subsidies and savings accounts complemented one another.

The sample comprises farmers with access to some type of agricultural extension service, either through an NGO or government entity, so that they have access to information on how to use fertilizer if they choose to use it. Researchers worked with two sub-groups of farmers. The voucher randomization (VR) sample is comprised of farmers randomly distributed (or not distributed) vouchers for fertilizer. The VR sample enabled researchers to examine the interaction between voucher receipt and savings incentives.

Treatment Groups:


No savings offered

Offered regular interest rates

Offered individual savings with 50% match

Offered group savings with match

Received voucher for fertilizer

Treatment Y-0

Treatment Y-1

Treatment Y-2

Treatment Y-3

Did not receive voucher for fertilizer

Treatment N-0

Treatment N-1

Treatment N-2

Treatment N-3

As shown in the table, the VR sample consists of three treatment groups which received different combinations of interventions, and a comparison group which did not receive an intervention. In treatment group one, farmers were offered a savings account with standard BOM interest rate. Treatment group two was offered “matched savings” accounts, where farmer received matched funds equal to 50 percent of his or her average savings balance (up to 3,000 MZN, or US$112) during a defined match period. (The match rate is the percentage of the average balance in the account that is contributed by the project at the end of the match period, not an annual percentage rate.) In treatment group three, farmers were offered a savings match with a group incentive, where the match rate rises or falls in accordance to the average account balance of the entire group. Farmers are not required to use the match for fertilizer, yet the match amount does allow each farmer to afford the inputs provided in the fertilizer package, which many farmers could not afford otherwise.

During meetings with farmer groups, project staff discussed the importance of savings and keeping part of one’s harvest proceeds for fertilizer and other agricultural inputs for the next season. Farmers were also given specific instructions about using the fertilizer package for maize, and information on BOM savings services and locations. After farmers completed the baseline survey, savings accounts were offered, and project staff assisted interested farmers in filling out the forms to open an account. Farmers then could make their initial deposit at a BOM branch or a Bancomovil, a mobile bank that services many of the sites.

During follow-up surveys planned for 2012 and 2013, researchers will collect data on per-capita income and expenditures, maize yields and use of seed varieties and fertilizers, and the creation and use of savings accounts.


Results forthcoming.

For more on Michael Carter's research, click here.

Insurance, Credit, and Technology Adoption in Malawi

Poor subsistence farmers often see adoption of new technologies, such as hybrid seeds, as risky because they fear the up-front investment will not pay off and they could be worse off, particularly in the case of drought. Yet new technologies can help farmers produce more food. So what happens if the risk of trying a new technology is removed? Few studies have evaluated whether providing insurance can increase the adoption of profitable agricultural technologies. In this study, researchers examined whether bundling rainfall insurance with a credit program–intended to finance adoption of a new crop technology—increase demand for credit.

Policy Issue:

The classic economic view of poor farmers is that their lack of a savings cushion causes them to prefer agricultural approaches with more reliable, but lower average returns. Farmers may see adoption of new technologies as risky, especially early in the adoption process when proper use and average yields are not well understood. Weather and environmental factors can also pose significant risks. Risk and uncertainty can lead to low levels of technology adoption, particularly where resources to help farmers deal with risk, such as insurance, are not available. However, few studies have evaluated whether providing insurance can increase adoption of profitable agricultural technologies.

Context of the Evaluation:

Nearly all Malawian households (97 percent in 2004-2005) are engaged in maize production, but only 58 percent use hybrid maize varieties. Hybrid maize adoption in Malawi has lagged behind adoption in Kenya, Zambia, and Zimbabwe. Aversion to risk, credit constraints and limited access to information are among the most cited reasons why hybrid seeds and other technologies have failed to take hold in Malawi. Within Malawi, varying rainfall risk is by far the dominant source of production risk, followed by pests.

Description of Intervention:

Researchers conducted a randomized evaluation in the field to determine whether bundling rainfall insurance with a credit program (intended to finance adoption of a new crop technology) increased demand for credit. Researchers randomly selected 16 localities in central Malawi where farmers were offered credit to purchase high-yielding hybrid maize and groundnut seeds for planting in the November 2006 crop season. In another 16 localities, farmers were offered a similar credit package, but if taking the loan were also required to purchase (at actuarially fair rates) a weather insurance policy that partially or fully forgave the loan in the event of poor rainfall.

The microfinance institutions Opportunity International Bank of Malawi (OIBM) and Malawi Rural Finance Corporation (MRFC) offered loans for the hybrid seeds based on group liability contracts for clubs of 10-20 farmers. Take-up of the loan was an individual decision, but the subset of farmers who took up the loan were told that they were jointly liable for each others’ loans.

The weather insurance policy was customized to each of the four project regions (Lilongwe North, Kasungu, Nkhotakota, and Mchinji). Payouts were based on the rainfall readings at the closest weather station to the individual in question, and premiums were lower in places where the likelihood of a bad rainfall shock was lower. Compared with the annual interest for the uninsured loan (27.5 percent), a farmer taking out an insured groundnut loan faced an effective interest rate ranging from 37.8 percent to 44.4 percent, depending on the area.

All farmers in the study were administered a household survey that covered income, education, assets, income-generating activities (including detailed information on crop production and crop choice), measures of risk aversion, and knowledge about financial products such as credit and insurance.

Results and Policy Lessons:

Loan Take-Up:Take-up was 33 percent among farmers who were offered the basic loan without insurance. Take up was lower, at only 17.6 percent, among farmers whose loans were insured against poor rainfall. A potential explanation is that farmers already were implicitly insured by the limited liability inherent in the loan contract, so that bundling a loan with formal insurance (for which an insurance premium is charged) is effectively an increase in the interest rate on the loan.

It is also possible that farmers may have been uncertain about the risks associated with the hybrid seeds. For those in the treatment group, the fact that they were offered insurance may have served as a signal that the seeds were a risky investment. Lower take-up of the credit plus insurance product would then be a rational response.

Analysis indicates that farmers who are wealthier and more educated were more likely to take up the insured loan. By contrast, there is no indication that farmer education, income, or wealth is related with loan take-up in the uninsured loan group.

Xavier Giné, Dean Yang

Promoting Sustainable Farming Practices in Malawi

Policy Issue:

Low productivity in agriculture is a pressing challenge in the developing world.  The compound effects of farming with mechanized soil tillage, climate change, and increasing urbanization are adversely affecting the long-term productivity of soil worldwide.  As a result, crop yields in developing countries are often many times lower than those that could be achieved using readily available technologies and farming techniques. Improving food security and agricultural incomes therefore depends on farmer adoption of these tools and techniques. A critical determinant of new technology adoption is the learning process through which information on these techniques is disseminated, understood, and applied. Although the importance of information flow through existing village and social networks in developing countries is well-documented, to what extent these networks can be used to disseminate new information from public sources—such as agricultural extension officers—remains unclear.


A lush climate and rich soil make Malawi well suited for agriculture, which is central to the country's economy and national life, making up 36 percent of its GDP[1] and occupying more than 80 percent of its workforce.[2] Tobacco is the leading export crop, followed by tea, sugar and cotton. The staple food crops are maize, cassava and millet, grown by smallholder farmers mostly at the subsistence level. Most rural families have too little land to produce sufficient food and too little income to buy extra. According to some reports, a quarter of the population runs out of food only five months after the harvest.[3]

Description of the Intervention:

Researchers investigate how new information on agricultural technology from outside sources, such as government-employed agricultural extension agents, is transferred through existing village and social networks. Understanding how gender and relationships affect communication between extension officers and farmers, as well as between farmers is crucial to designing effective information-based interventions to promote technology adoption. Both the technologies and methods for disseminating information are evaluated in the context of the Malawi Agricultural Development Programme Support Project (ADP-SP). This project is intended to support the efforts of the Malawi Ministry of Agriculture to achieve sustainable productivity growth in smallholder maize production systems.

Each village assigned to receive the intervention will be randomly assigned a dissemination method as well as a farming technique that is to be promoted. The two techniques for dissemination are:

(1)   Fertilizer Nutrient Management (FNM), which includes the use of efficient combinations, timing, and spatial concentration of fertilizer application and results in short-run increases in land productivity.

(2)   Conservation Agriculture (CA), which includes pit planting, minimal tillage, and mulching. CA is associated with long-run returns, as CA practices increase the biodiversity of farm ecosystems, allowing non-chemical organisms to take a role in soil maintenance.

The randomly varied dissemination methods are:

(1)   Extension officers, working through their existing channels and provided with incentives based on the adoption of the techniques in their areas.

(2)   Extension officers collaborating with lead farmers in each village, whom they select in consultation with the community. Lead farmers will be provided with incentives based on the adoption of the techniques in their villages.

(3)   Extension officers collaborating with peer farmers, selected through focus groups in each village and representing average farmers spatially dispersed throughout the village. Peer farmers will receive incentives based on adoption in their neighborhoods.

Across all three types of dissemination methods, incentives were sometimes randomly assigned to the communicators.  In the lead and peer farmer villages, the gender composition of the message sender will be varied. The random variation in the technology promotion agent and in the technique being promoted will allow researchers to determine the potentially differing outcomes of various components of information dissemination programs.


Results forthcoming.

[1] US Department of State, Bureau of African Affairs. “Background Note: Malawi.” January 11, 2011

[2] World Vision. “Country Profile: Malawi.”

[3] ibid

Technology Adoption and Diffusion through Social Networks – Evidence from a Coffee Training Program in Rwanda

Policy Issue:

In sub-Saharan Africa, a significant proportion of the poor are small-scale farmers with productivity levels among the lowest in the world. Agronomic practices, such as the correct combination of organic and inorganic fertilizer or pruning of tree crops, have the potential to substantially increase yields. However, application of such practices, and adoption of productivity-enhancing agricultural technologies in general, remain infrequent in most of sub-Saharan Africa. Assuming that knowledge deficits are the main hindrance to the adoption of best agronomic practices, training on agricultural technologies could promote the take-up of these practices, increase productivity, and improve economic outcomes for these poor farmers. This project aims to evaluate the impact of agronomy trainings on small-holder productivity and adoption of best agronomic practices. The study will also measure the diffusion of the program’s effects, if any, within farmers' social networks, by linking each grower to the people with whom they might share the information provided in the trainings.  

Evaluation Context:

According to Rwanda's Ministry of Agriculture, the main reason why the country is currently falling short of the coffee sector targets for 2010 is an insufficient level of cherry production[1]. In Rwanda, one of the most densely populated countries in sub-Saharan Africa, land parcels are extremely small – averaging only a few acres. Most farmers live off subsistence agriculture, and therefore usually give priority to food crops, allocating only a small fraction of their land to coffee. Since their revenues from coffee sales are limited by production levels, finding ways to improve their yields could make coffee a more attractive crop choice and help small-scale farmers shift out of subsistence farming into more profitable activities.

TechnoServe, an agri-business NGO aiming to provide business solutions to those in rural areas, is conducting agronomy trainings in several coffee growing regions in Rwanda and other East African countries. IPA is conducting a randomized evaluation of one of TechnoServe’s agronomy training programs in Nyarubaka, a rural sector (a geographical unit covering thirty villages) in the Southern Province of Rwanda. Similar to other coffee farming regions in Rwanda, the majority of farmers in Nyarubaka live off the food crops they grow and few households have electricity.

Description of the Intervention:

In Nyarubaka, the agronomy training evaluation sample is composed of the 1600 farmers (from 27 villages) who signed up for Technoserve’s agronomy trainings. As Technoserve only had resources to train a limited number of farmers, households in the sector were randomly assigned to a treatment (50%) or comparison (50%) group after receiving a baseline survey. The baseline survey collected crop production and profit information (across all crops), information on consumption, household demographics and assets, and a detailed social network map (including information on who farmers talk to about coffee).

Villages in the treatment group were then randomly assigned to different intensity levels: in some villages ¼ of registered households received treatment (i.e. ¾ comparison), others ½, and others ¾.

Training sessions took place once a month for 11 months in the first year of the program and an additional 5 review training sessions were done over the course of the second year. These trainings covered a number of coffee best practices: rejuvenation and pruning; nutrition (organic and inorganic); pest, disease and weed management; mulching; soil and water conservation (water traps, terracing); shading; and record keeping. Trainings were attended by groups of around 30 farmers and took place on the plot of a designated "focal farmer" who was generally chosen because of the accessibility of his/her plot.

Farmers in the 1600 farmer sample were surveyed every few months to gather data on their coffee yields, sales and practices, as well as their harvests of other crops during the two rainy seasons. In addition, there is a consumption survey administered once a year. Take-up of the agronomy practices taught in the trainings is monitored closely through bi-yearly plot inspections. Farmers in both treatment and comparison groups were given scales for weighing their coffee harvests and calendars to record their harvests and sales.

All the coffee farmers in the sector who are not part of the direct agronomy training evaluation are also surveyed at least once a year to collect data on their own coffee practices. Annual follow-up of these additional 1400 coffee farmers in the region allow us to identify their “coffee friends” (individuals who they talk to about coffee) to assess the diffusion of these practices through the social networks of trained farmers. This context is a particularly interesting one to learn about diffusion, as there is variation in the type of technologies included in farmer training. Some of the best practices result in outcomes that are visibly evident (i.e. coffee trees just look much better/healthier) whereas others are less visible. So, we may see differences in the relative speeds of diffusion of these various practices (with some diffusion even outside the social networks, but within geographical space).

Both geographic and social network mapping of all coffee farmers in Nyarubaka is carried out regularly to understand how geographic and social infrastructures affect adoption and diffusion of the agronomic practices.

Results and Policy Lessons:

Results forthcoming.


[1]Ministry of Agriculture and Animal Husbandry and Ministry of Trade and Industry, “Rwanda National Coffee Strategy 2009-2012.” December 2008.

Esther Duflo, Tavneet Suri

Contract Farming, Technology Adoption and Agricultural Productivity: Evidence from Small Scale Farmers in Western Kenya

Researchers study how cash advances and information delivered via text messages can encourage farmers to adopt efficient agricultural practices and new crop varieties.

Policy Issue:

Many farmers in the developing world practice subsistence farming, a mode of agriculture in which a plot of land produces only enough food for personal consumption. Moving from pure subsistence crops to high-value cash crops can help poor farmers gain income, by enabling them to sell their produce in markets. However, a number of obstacles may prevent subsistence farmers from adopting high-value crops. First, long lag times between effort and rewards may reduce farmers’ willingness and ability to grow certain crops (sugarcane, for example, has an 18 month average harvest cycle in Kenya, the country targeted by the study). Second, it can be costly to diffuse information about high-value crops.


The Mumias Sugar Company (MSC), a leading sugar producer in Kenya, is located in Mumias District, in Kenya’s Western Province. MSC has the largest sugarcane factory in Kenya, and works with approximately 70,000 out-grower farmers, whose plots encompass an area of 400 square kilometers, in several districts in the province. The company is strongly dedicated to innovation and experimentation. Over the past decade, the MSC agronomy department has been active in testing new cane varieties, fertilizers, and herbicides.

Description of Intervention:

Researchers partnered with the Mumias Sugar Company (MSC) to work with smallholder farmers who, as out-growers, sell sugarcane to the firm. Working with the company management, researchers will explore two interventions: (i) cash advances conditional on farmers' performance at intermediate stages of the harvest cycle, (ii) cell-phone text-message reminders.

The first intervention will introduce rewards for intermediate outcomes for a subsample of farmers. MSC field assistants will assess fertilizer application and weeding accuracy. If credit constraints play a large role, anticipated payments tied to intermediate outcomes should increase farmers’ effort and plot management quality.

The second intervention will aim to reduce costs and information problems related to extension services. Researchers will exploit the large number of cell phones in the area of study (MSC estimates that 70 percent of farmers have access to a cell phone) and use them to deliver information about agricultural practices. Together with MSC agronomists, researchers will develop an SMS reminder system. Reminders will concern planting, weeding, fertilizer and herbicide application, cane fire prevention, and harvesting.

The evaluation will assess the impact of the treatments on plot management quality,  good agricultural practices knowledge, yields and crop variety adoption.


Results forthcoming.


Providing Collateral and Improving Product Market Access for Smallholder Farmers: A Randomized Evaluation of Inventory Credit in Sierra Leone

Policy Issue:

Inter-seasonal fluctuation of agricultural prices is widespread throughout the developing world. For many crops, prices decrease at harvest season, owing to the availability of large quantities of crop, while prices increase in the lean season.  However, small farmers are often unable to benefit from this price increase due to a lack of proper storage facilities and credit constraints. Inventory credit products address both storage and credit constraints by allowing small farmers to store their harvest in a secure warehouse as collateral for a loan. Such products have had successful small-scale test cases in West Africa, including Ghana (Technoserve), Niger (Food and Agriculture Organization) and Mali (World Bank). Yet, to date there have been no rigorous evaluations to assess inventory credit’s cost-effectiveness and sustainability.

Context of the Evaluation:

In Sierra Leone, palm oil is an essential component of rice consumption and exhibits large and predictable seasonal price changes, creating inter-temporal arbitrage opportunities, which remain largely unexploited by small farmers. Pilot data shows 72% of farmers sell a majority of their output within two months of harvest, despite an expected price increase of over 70% within six months.

The Sierra Leone National Program Coordinating Unit (NPCU) at the Ministry of Agriculture plans to implement a palm oil inventory credit scheme in collaboration with three Rural and Agricultural Banks (RABs). IPA will use the rollout of this program to conduct a randomized evaluation of the intervention.

Details of the Intervention:

The three participating RABs identified 120 communities that would be eligible to receive the inventory credit product. These communities will be randomly assigned to three groups with 40 communities each: The first will receive the inventory credit product, the second will receive assistance with management of a community storage space, but no access to inventory credit, and the third will serve as a comparison group.

In inventory credit communities, farmers will receive harvest time loans in exchange for storing their palm oil as collateral. The loan amount will be 70% of the palm oil’s harvest-time value, or around $5.50 per 5 gallon container of palm oil. The bank will store the collateral in a secure room provided by and located within the community, which will have two locks: the key for one will be controlled by the community; the key for the other will be controlled by bank staff. The banks will provide containers for the storage space, in which the palm oil will be stored. In the lean season (nine months later), when prices are typically more than 50% higher, the banks will assist the farmers with selling the collateral. The bank will recoup its loan and interest. The farmer will keep all additional revenue.

In storage communities, farmers will receive storage containers for the community store space and assistance with management of the space: NPCU staff will control the key to one lock, while the community will control the key to a second one. However, no inventory credit will be offered.

The study seeks to rigorously evaluate this intervention to examine the relation between storage, credit and access to markets for small farmers. The questions it seeks to answer are: a) Do farmers’ take-up the credit product? b) To what degree do farmers modify their sales patterns when using the credit product? c) Does inventory credit affect prices received by farmers in different seasons of the year?


Results forthcoming. 

Lorenzo Casaburi

Encouraging Adoption of Rainwater Harvesting Tanks Through Collateralized Loans in Kenya

Policy Issue:
Over the last decade, billions of dollars have been invested in the development of new agricultural technologies, which have the potential to contribute to economic growth and poverty alleviation amongst the poor. However, in Sub-Saharan Africa, the adoption of these promising agriculture technologies has been limited. Low adoption rates may be explained in part by credit constraints, which prevent many smallholder farmers from making the often large, initial investment in a new technology. A sizeable initial deposit or a guarantor is often required to receive a loan, restricting access for most. In developed countries, guarantor requirements are uncommon; instead, loans for purchase of houses, vehicles, and small business equipment often use the assets themselves as loan collateral. Using assets as collateral is rare in developing countries, where credible and efficient institutions and processes to seize collateral may be lacking. But in certain environments or for certain products—like water tanks, which are large and difficult to hide or relocate, making repossession feasible—can asset collateralization expand access to credit without affecting repayment rates?
Context of the Evaluation: 
Almost 80 percent of the Kenyan population lives in rural areas, where many suffer from significant water constraints. About 96 percent of cropland in Kenya is rain-fed and the rains are both seasonal and unpredictable. In recent years, much of rural Kenya has been hit by extreme droughts, which can be particularly problematic for dairy cattle, which require a consistent and regular water supply. Dairy farming is common in Kenya’s Central and Rift Valley provinces, the sample area for this study. The smallholder dairy farmers in the study sample had an average herd size of two cows, but 32 percent of households reported loosing at least one cow to drought in the year prior to the study, and 52 percent of households had a cow get sick. Without easy access to water, sample farmers reported spending a considerable amount of time (ten hours per week) taking cows to a water source. Technologies, such as rainwater harvesting tanks, that help to give dairy farmers reliable and convenient access to water, could improve their productivity and other economic outcomes.
Details of the Intervention: 
Researchers used a randomized evaluation to examine the effect of asset collateralized loans on the take-up of rainwater harvesting tanks and any subsequent impacts on dairy production, time use, or girls’ enrollment in school.
In partnership with a local dairy cooperative, researchers offered smallholder dairy farmers in central Kenya a loan to purchase a durable, 5,000-liter rainwater harvesting tank. The market price of the tanks at the time of the study was Ksh 24,000 (US$320), or roughly 13 percent of annual total household consumption in the sample. A full tank would last between 17 days and 6 weeks depending on whether the water was used for livestock, domestic use, or both.
From a sample of 3,000 farmers who sold milk to a local dairy cooperative, researchers randomly assigned 1,804 to receive one of four initial loan offers. The value and terms of the loan were the same for all four offers: the loan could only be used to purchase a water tank, the value of the loan was set equal to the market price of the tank (US$320), and all borrowers had to make a minimum deposit of Ksh 1,000 (US$15).
The four different loan offers varied in the deposit and guarantor requirements, and whether the asset was pledged as collateral. 
(i) 100 percent secured joint-liability loan - required borrowers to make a deposit equivalent to one-third of the value of the loan in their savings account. Three guarantors were also required to insure two-thirds of the loan amount through savings or shares in the dairy cooperative. This offer was essentially the same as the standard contract used by the cooperative when it gave out loans.
(ii) 4 percent deposit loan - used the water tank asset as collateral but carried no other up-front requirement apart from the small, Ksh 1000 (US$15) deposit, equivalent to 4 percent of the loan amount, required of all borrowers.
(iii) 25 percent deposit loan – used the water tank asset as collateral and required a larger deposit of Ksh 6,000, equivalent to 25 percent of the loan amount.
(iv) 25 percent guarantor loan – used the water tank asset as collateral and also required a single guarantor to pledge 21 percent of the loan amount (Ksh 5,000), in addition to the Ksh 1,000 deposit from the borrowing farmer.
Researchers anticipated that take-up rates would vary across these groups due to differences in the perceived profitability of borrowing under the given terms, differences in access to guarantors, and differences in farmer productivity and characteristics. In order to distinguish between these factors, farmers in groups iii and iv were randomly divided into two subgroups. Half of the farmers who received offer iii or iv had the requirements waived after the loan contract was signed; in group iii, the deposit amount in excess of the minimal 4 percent deposit was returned, and in group iv, the guarantor was absolved of any responsibility for the loan and his deposit was returned.
For all six resulting treatment groups, the timeline for the loan was 24 months and monthly repayments consisted of a fixed amount (Ksh 1,000) plus interest (1 percent per month) on the declining principal balance.  
The default procedures were the same across all treatment groups. When a farmer fell two months behind, he received a letter warning him that he was pending default and had two months to pay off the late amount. If the payment was still outstanding after 60 days, the cooperative applied any deposits held to the balance. If a balance still remained, the farmer was given an additional 15 days to clear it, after which the cooperative would repossess the tank and sell it to cover the farmer’s loan obligations. 
Results and Policy Lessons: 
Take-up rates: Take up of the 100 percent secured joint-liability loan was low, with only 2.4 percent accepting the offer. Asset collateralization significantly increased take up—nearly 28 percent of those who received the 25 percent deposit loan offer accepted, 23.5 percent of those who received the 25 percent guarantor offer accepted, and 44 percent of those who received the 4 percent deposit offer accepted.
Loan repayment rates and timing: About 63 percent of the farmers who took out a loan were late on their payments at least once during the course of the loan, but repayment performance did not vary significantly between treatment groups. The only major difference was the time taken to repay the loan: the average time taken to repay the 100 percent joint-liability loan was only nine months, while the other loans took between 17 and 22 months to pay off. No farmers defaulted on their loans and no tanks were repossessed. These preliminary results suggest that access to credit can be improved at little cost to lenders.
Impact on economic and welfare outcomes: Preliminary results show that access to asset-collateralized loans (through the 4 percent deposit offer) improved milk production specifically for those without access to piped water before the start of the study. In addition, it reduced the time girls in treatment households spent fetching water by 35 percent, and increased the probability that girls were enrolled in school by 4 percentage points. This latter effect is striking because enrollment rates in the sample area were already high—girls’ enrollment at baseline was 94.6 percent. This was not the result of households being able to borrow to pay school fees (a common use of microfinance loans), but rather the ability to invest in an asset that reduced the demands on girls’ time.

Promoting Adoption of New Rice Varieties: Addressing the Costs of Early Adoption in Sierra Leone

Can improved seed varieties benefit poor farmers in Sierra Leone? Can price subsidies and agricultural extension training lessen the costs of early adoption?

Policy Issue:

Agricultural productivity has stagnated in much of sub-Saharan Africa, while many other regions of the world have seen dramatic productivity improvements in recent decades. New agricultural technologies, such as high-yielding crop varieties, offer the promise of improving productivity and hence the welfare of farmers. But adoption of these technologies has often been low in countries where dissemination programs have been conducted. First adopters of new technologies play an important role in the spread of technology as they take on the burden of experimentation—testing whether and how a new variety works in local conditions. This is particularly important in much of sub-Saharan Africa where a multiplicity of micro climates within a small area means that experimentation is essential for farmers to learn which crop varieties are best for their particular land. There is also concern that early subsidization to increase adoption of new technologies will lead to expectation of continued subsidies, depressing demand at market prices.

Details of the Intervention:

Researchers sought to test whether improved seeds are beneficial for the poor in Sierra Leone and how best to promote uptake given the high costs of early adoption. Early adopters generate a positive externality to surrounding farmers and communities by delivering information on the effectiveness of new varieties and how to make the most of them in local conditions.

Two types of incentives will be tested: (1) a price subsidy scheme allowing farmers to purchase new seeds at a price lower than the market price; and (2) targeted agricultural extension work involving community demonstration plots and practical advice on how to use these seeds. Two types of new seed varieties will be used in the intervention. The first is one of the NERICA varieties and the other is a local variety developed by the Rokupr station, ROK-16, which has proven popular in early participatory variety selection tests.

The first stage of the intervention will pilot the agricultural extension training and subsidy incentives. One treatment arm will receive free seeds and training, and take up and yields will be compared to a comparison group. In the second year, researchers plan to test a more complex set of alternatives involving six treatment arms, each with approximately 35 communities, under three different schemes:

Pricing Scheme:

  • T1: Farming households offered ROK-16 variety at 0 percent subsidy (market price).
  • T2: Farming households offered ROK-16 variety at 50 percent subsidy.
  • T3: Farming households offered ROK-16 variety at 100 percent subsidy (free).

Training Scheme:

  • T4: Farmers receive a targeted training program without a formal opportunity to purchase the new seeds.
  • T5: Farmers receive a targeted training program and are offered the chance to purchase ROK-16 seeds at one of the three subsidized prices.

NERICA Scheme:

  • T6: Farmers are offered one upland variety of NERICA rice (NERICA-6) for free (as is currently the practice of the government’s NERICA project) as well as the targeted training program.

The pricing scheme aims to test the hypothesis that a one-time subsidy can reduce the adoption cost for early-adopters and have a long-lasting effect both on the beneficiary and their neighbors. The training scheme aims to reduce the cost of learning by providing information on how well the seed works in the community (through a demonstration plot) and on how to cope with some of new features of the rice. Both ROK-16 and NERICA are included because there is little information on the relative productivity of each variety of seed in upland conditions.

Key outcome variables measured at endline include: (1) the amount of improved rice variety seeds (NERICA and ROK-16) purchased and planted; (2) planting of other rice varieties and other crops; (3) amount of family and hired labor used on the farm; (4) consumption and food security.

Results and Policy Lessons:

Study in implementation, results forthcoming.

Community Based Rangeland Management in Namibia

We conduct an impact assessment of the Community Based Rangeland and Livestock Management (CBRLM) program in Namibia.  This program is part of a larger set of interventions in the agricultural sector designed to reduce poverty among the population of the northern regions of the country.  Many people in the area rely on cattle production for their economic livelihoods, however overuse of the communal grazing areas and suboptimal grazing practices threaten the long-term viability of the land and contribute to persistent poverty.  

To increase the productivity of livestock and other animals using the land, the Namibia Millenium Challenge Compact funds a pilot program designed to help communities improve their livestock practices, address rangeland degradation, and improve market access.  

The evaluation is designed to test the impact of the various activities within the CBRLM intervention on household income, cattle productivity, and the condition of the rangeland. The intervention targets both inadequate information about appropriate cattle production practices and the social or other behavioral preferences of farmers.  At the moment, there is a collective action or “tragedy of the commons” problem – individual farmers are hesitant to reduce their herd’s impact on the rangeland because they are fearful that others will not follow suit which often results in overuse and degradation of the land.

Barriers to Fertilizer Use: Evidence from a Field Experiment in Kenya

Policy Issue:

Agricultural outputs in Africa have stagnated over the past decades: although total output has risen, food production has not kept up with the increase in Africa’s population. The number of chronically undernourished people in Africa has increased to 200 million in 1997-99.1 When used correctly, chemical fertilizer can substantially raise agricultural yields, yet usage of fertilizer remains low in Sub-Saharan Africa. Past studies suggest that usage is low because farmers have difficulty saving harvest income to purchase fertilizer for the next growing season, have limited information on the benefits of using fertilizer properly, and the fact that knowledge about fertilizer is not passed from one farmer to another. This project attempts to address all three issues.

Context of the Evaluation:

This project focuses on small-scale subsistence farmers in rural Western Kenya, many of whom grow maize as their staple crop. All farmers in this population are extremely poor subsistence farmers, earning on the order of $1 per day. Previous research in this area has shown that when used correctly, top dressing fertilizer can increase yields by about 48%, amounting to a 36% rate of return over just a few months. However, only 40% of sampled farmers in the Busia district of Western Kenya report ever having used fertilizer.

Details of the Intervention:

This experiment looks at a complex intervention with several components meant to increase fertilizer use and dissemination of knowledge. Farmers were recruited to the study through meetings at primary schools and randomly divided into four groups.

The first group received small, time-limited discounts which were valid within a 3 week window right after harvest, redeemable at a local shop. Farmers received coupons for a discount of about 15% of the price of fertilizer, for up to 25 kilograms.

The second group was encouraged to form farmers’ cooperative with their friends and neighbors to talk about fertilizer and agricultural practices. The researchers organized the groups and coordinated the first few meetings, but did not provide any direct information to the groups.

The third group participated in both the coupon scheme and the cooperatives.

A fourth group received none of these services, and served as a comparison.

Researchers will examine the changes in fertilizer usage between the different groups and whether farmers in the treatment groups talk to each other about agriculture more than others.  

A separate intervention was designed to investigate the spread of information and technology when provided only to a subset of farmers in the treatment and comparison groups. The research team visited the randomly selected farmers and provided them with ½ teaspoon measuring spoons, as well as information about the returns to using ½ teaspoon of fertilizer per plant. To enable diffusion of this technology to others in the community, the spoons were made available in nearby fertilizer shops to other farmers for a nominal fee. In addition, when distributing the measuring spoons, the farmers were given vouchers for spoons which they could give to their friends. This intervention will test the hypothesis that the fertilizer discount intervention and the cooperative intervention could lead to greater diffusion of information about fertilizer.

Results and Policy Lessons:

Results forthcoming.


1 Ernest Harsch, “Agriculture: Africa’s Engine for Growth”, available at

2 Duflo, et. al., “Nudging Farmers to Use Fertilizer: Experimental Evidence from Kenya”, available at

Paying for Environmental Services: An Experimental Study in Bolivia

Can financial incentives and information influence farmers to account for spillover effects of their cattle management practices?

Policy Issue

Some agricultural and farming practices create spillovers that affect others or the environment. These spillover effects, known as externalities, can create a wedge between the benefits a farming practice has to individuals and the effects it has on society as a whole. Although adoption of agricultural technologies that reduce the production of negative externalities, such as pollution or deforestation, is beneficial to society, such technologies will not be adopted if they don’t bring benefits to individual users. The standard policy solution in the face of such externalities is to change incentives so that private individuals benefit from use of socially responsible practices. In recent years, policymakers have advocated payments for environmental services as an incentive-based approach to internalizing the externalities of land use decisions, but there is little empirical evidence on the impacts of such programs.

Context of the Evaluation

This study takes place in Bolivia’s Rio Grande Protected Area, where cattle are both a  source of private income, and of negative agricultural productivity, health and environmental externalities. Most participants in the study area own both cattle and forest land. Through forest clearing for pasture and free range grazing in the forest, local cattle management practices generate significant negative externalities at local (watershed) and global (biodiversity, carbon) scales. Free range grazing cattle spend a significant amount of time standing in streams, where much of the best vegetation grows. This causes erosion of the stream banks, and contributes to landslides that close roads and block access to markets.

Details of the Intervention

Researchers partnered with Fundación Natura Bolivia, an organization that has worked in the study area for six years implementing payment for environmental services projects. This study will evaluate interventions designed to increase the adoption of more environmentally friendly cattle management practices, including the establishment of pastures with trees and water troughs, which reduce negative externalities related to health, productivity and environmental processes by reducing the time cattle spend grazing near streams.

In half of the eligible villages, financial incentives will be introduced in the form of conditional contracts for cattle management. The other half of the villages will be provided only with information about the harms associated with poor cattle management. In addition to evaluating effectiveness of incentives for conservation outcomes, data collection will focus on ancillary measures that help shed light on the process of collective action within the community.

In the short run, adoption of these management technologies will be costly to individual users, but has the potential to create benefits for everyone in the village by reducing erosion. In the long run (>5 years), the technology is expected to be both privately and socially beneficial, as the reductions in land degradation bring benefits to individual users. By varying the price incentives for adoption and the information about the externalities, the intervention will offer insights into whether incentives and information can help farmers adopt technologies that prevent negative externalities.

Results and Policy Lessons

Results forthcoming.

Farmer Decision-Making and Technology Experimentation in Indonesia

Policy Issue: 

Despite regularly making decisions that affect their crop yields, farmers in the developing world may lack information about how to appropriately use farming inputs or techniques. For instance, a farmer must decide on the optimal timing of planting, the depth and spacing of seeds, the amount of fertilizer to use, and the timing of fertilizer use, among other things.  Farmers may be reluctant to engage in potentially risky experimentation on their own plots, which prevents them from gathering new information about efficient agricultural practices. And even when farmers have information available to them about ways to increase their yields, they may not adopt optimal agricultural practices. One explanation for this is that farmers fail to notice certain details about the cultivation process simply because they may not believe that information to be useful. This suggests that it could be possible for farmers to learn about optimal farming techniques if they are presented with information that helps them notice new or previously neglected dimensions of the production process. 

Context of the Evaluation: 

Seaweed farming has been prominent in Nusa Penida district in Bali, Indonesia since it was introduced during the 1980s. Most seaweed is cultivated by taking raw seaweed and cutting it into pods, which are then planted at intervals along the ocean floor. The size of the pods and distance between them is determined by the farmer. Due to its relatively short crop cycle of 35 to 40 days, seaweed cultivation provides farmers with ample opportunity to learn through experimentation and implement new techniques. However, 86 percent of seaweed farmers sampled were unable to provide information about the pod size they used on their plots or what they believed to be the optimal pod size for seaweed cultivation.

Details of the Intervention: 

From a group of 232 seaweed farmers, researchers randomly selected 117 to participate in an experimental trial to determine the optimal pod size for seaweed cultivation on their plots. Farmers in the treatment group assisted an agricultural extension worker to vary the seaweed production methods on one of their plots and received the trial results afterwards. The remaining 115 farmers served as the comparison group and received no new services. 

The farmers in the treatment group were randomly assigned to one of two sub-groups that varied the size and weight of seaweed pods.

  • Pod Size Treatment: Researchers collected data on the variation in pod size within each farmer’s plot to understand whether farmers could achieve a higher yield by systematically planting pods of a specific size. 
  • Pod Weight Treatment: Researchers collected data on how variation in the weight and distance between pods affected the yield. In both cases, farmers observed as enumerators planted the seaweed pods. 

All farmers in the treatment group received compensation for their participation in the form of farming inputs, a guaranteed income from the plot they cultivated as part of the trial, and a small gift worth US $1.  

Researchers conducted a follow-up survey from April to May 2008 to test whether farmers changed any of their methods after participating in the trial. After the first follow-up survey, farmers were given summarized trial results, which included information on the returns from different farming methods and highlighted which pod size and pod weight produced the highest yields, respectively. Enumerators talked through the results with the farmers, in addition to providing them with a written summary. Two months after the results were distributed, researchers conducted a second follow-up survey to determine whether farmers had changed their methods as a result of having received the trial results. 

Results and Policy Lessons: 

Researchers found that farmers neglect certain dimensions of seaweed cultivation and fail to use the optimal level of inputs along those neglected dimensions. For instance, while most farmers are attentive to the optimal distance between seaweed pods, very few farmers had consciously experimented with pod size prior to the trial. 

Participation in the trial alone did not induce a significant change in farming techniques, which suggests that learning through observation and experience does not always guarantee effective use of technology. However, farmers in the treatment group report making large and significant changes in their production techniques after receiving summarized trial results and specific recommendations about which pod size or weight improved yields. The number of farmers who reported changing their techniques increased by 16 percentage points between the first and the second follow-up surveys.  Similarly, pod size increased by approximately 7 grams in the pod-size treatment sub-group relative to the comparison group. This was consistent with the average change in pod size recommended by enumerators. 

Thus, while farmers did not appear to consider pod size to be an important part of the production process prior to the trials, providing summary information on the optimal pod size appeared to change their use of this production input. This suggests that farmers who fail to notice may not learn even when they are actively experimenting and, as a result, may not notice the very features of a technology that make it profitable. Training programs for farmers may be useful, not only for new technologies, but also for existing technologies that individuals may have had prior experience with.  


Irrigation and Property Rights for Farmers in Mali

Farmers face many challenges as they try to grow and sell enough crops to support their families. Uncertain rainfall, potential crop failure due to natural distasters or disease, unpredictable crop prices, and shaky land tenure all contribute to the difficulties and risks inherent in farming. Improvements in the production processes and productivity of farmland could help many poor families achieve a better life. The Alatona zone is one of the most disadvantaged zones in Mali. The main livelihood sources of the population in the zone are rain-fed agriculture, pastoralism, and wage labor in the Office du Niger.

The Alatona Irrigation Project is a component of the MCC Compact in Mali, and like the larger mission of MCC, the AIP seeks to reduce poverty through economic growth. The AIP is focused on increasing production and productivity, increasing farmer's income, improving land tenure security, modernizing irrigated production systems and mitigating the uncertainty from subsistence rain-fed agriculture. AIP seeks to develop roughly 14,000 hectares of net irrigated land in the Alatona perimeter, representing an almost 20% increase of "drought-proof" cropland and a 7% increase of the country's total stock of fully or partially irrigated land. The AIP will introduce innovative agricultural, land tenure, and water management practices, as well as policy and organizational reforms aimed at realizing the Office du Niger's potential to serve as an engine of rural growth for Mali.

The AIP project will create an additional 14,000 hectares of irrigated land in the Office du Niger (ON) zone. Project affected people currently residing or tending land in the project zone will be given 5 hectare plots and new settlers wishing to move to the area will be required to purchase plots of 5 hectare or greater. Women will have access to 500 m² market gardens. Institutions and management processes will be improved relative to comparable institutions in existing ON areas. One such example is land titles: all farmers in the Alatona will be given full land titles for their land, including rights to sell the land. This feature of the intervention will increase property rights and security for AIP farmers in contrast to the short or longer-run lease system currently used by the ON. Farmer organizations will also be encouraged and supported to maximize local capacities.

Additionally, water management and the collection of water fees will be administered through local farmer associations; whereas in existing ON areas, the ON is directly responsible for the collection of water fees. Financial services including starter kits (fertilizer, seeds) for the PAP concessions and agricultural credit will also be developed in the area. Finally, since AIP land is sold to beneficiaries, the land revenues may provide opportunities during and most likely after the Compact period for subsequent development of the zone.

Lori Beaman, Andrew Dillon

Nudging Farmers to Use Fertilizer: Experimental Evidence from Kenya

Use of inorganic fertilizer has the potential to dramatically increase yields and, if used correctly, is a highly profitable investment. So why do so few farmers in sub-Saharan Africa use it? Is it lack of information about profitability, lack of money to purchase the product, or an inability to save for the purchase? Researchers designed an intervention to offer farmers the ability to save harvest income for future fertilizer purchase. An ICS officer visited farmers immediately after the harvest, and offered to sell them a voucher for fertilizer, at the regular price, with free delivery later in the season. The program was popular, and in the first season increased usage by 14 percentage points. In the second season, the increase was even bigger, increasing usage by 18 percentage points. These effects are comparable to those obtained from a 50% price subsidy.

Policy Issue:

By some estimates, there are approximately 1.4 billion people living on less than $1.25 a day,1many of whom are farmers. As such, identifying ways to increase agricultural incomes is crucial in meaningfully alleviating poverty. Such strategies are especially important in sub-Saharan Africa, a region in which agricultural yields have been low and remained stagnant for many years. Use of inorganic fertilizer has the potential to dramatically increase yields and, if used correctly, is a highly profitable investment. Why do so few farmers in sub-Saharan Africa use fertilizer- is it lack of information about its profitability, lack of money to purchase the product, or an inability to save for the purchase?

Context of the Evaluation: 

An estimated 66% of the population of Kenya’s Western Province lives below the poverty line2where the majority of small farmers grow maize as their staple crop. Improving agricultural productivity through increased use of inorganic fertilizers could have substantial benefits for the livelihoods of these subsistence farmers. Numerous agricultural trials on experimental farms, as well as experiments where farmers were given full fertilizer, showed that adding fertilizer once the plants had sprouted (as top dressing) generated a 70% annualized return in Western Province. However, only 40% of sampled farmers in the Busia district report ever having used fertilizer. The overall goal of this research program is to understand why farmers do not invest in fertilizer. This part of the project investigates whether difficulty in saving harvest income until the time that inputs are needed is a significant barrier to adoption.

Details of the Intervention:

In collaboration with the NGO International Child Support (ICS), researchers designed an intervention to test if providing mechanisms to save harvest income for future fertilizer purchase could be effective in increasing usage. The intervention was called the Savings and Fertilizer Initiative (SAFI). The design of the experiment allowed researchers to test the impact of the SAFI program against various other strategies to improve usage, in particular fertilizer subsidies.

The following interventions were tested over two seasons among a sample of farmers:

  1. Basic SAFI: An ICS officer visited farmers immediately after the harvest, and offered to sell them a voucher for fertilizer, at the regular price, with free delivery later in the season. The farmer had to decide during the visit whether or not to participate in the program, and could buy any amount of fertilizer.
  2. SAFI with ex ante Choice of Timing: An ICS officer visited the farmers before the harvest and offered them the opportunity to decide when, during the next growing season, they wanted the officer to return to offer them the SAFI program. They were then visited at the specified time, and offered a chance to buy a voucher for future fertilizer use (as in the Basic SAFI program, as described above).
  3. Free Delivery Visit Later in the Season: Same as SAFI program, but farmers were visited later in the season. An ICS officer visited farmers 2-4 months after the harvest (when it is time to apply fertilizer as a top-dressing to the next crop), and offered them the opportunity to buy fertilizer, at the regular price, with free delivery. This program was identical to SAFI, except that it was offered later.
  4. Subsidy Later in Season: An ICS officer visited the farmers 2-4 months after the harvest (when it is time to apply fertilizer to the next crop) and offered to sell them fertilizer, at a 50% subsidy, with free delivery.


Results and Policy Lessons:

The SAFI program was very popular. The basic SAFI was offered in two seasons. In the first season, the program increased usage by 14 percentage points, on a base of 23 percentage points. In the second season, the increase was even bigger, increasing usage by 18 percentage points. SAFI with ex ante timing choice was also successful, increasing usage by 22 points.

These effects are comparable to those obtained from a 50% subsidy offered later in the season: the subsidy increased usage by 14 points. They are also larger than an undiscounted offer later in the season: the free delivery visit later in the season had no significant effect on usage.

Consistent with a savings problem, enrollment in the SAFI program did not cause farmers to use fertilizer in subsequent seasons (as would be predicted if farmers were learning about fertilizer through the program). This suggests that it was the lack of commitment mechanism that was preventing farmers from purchasing and using fertilizer.

Overall, the results suggest that offering farmers small, time-limited discounts on fertilizer may substantially increase usage without inducing overuse among farmers who are already using fertilizer, at relatively low cost.

1 Shahua Chen and Martin Ravallion (2008). “The Developing World Is Poorer Than We Thought, But No Less Successful in the Fight against Poverty,” World Bank Policy Research Working Paper #4703.
2 National Coordinating Agency for Population and Development (NCAPD) [Kenya], Ministry of Health (MOH), Central Bureau of Statistics (CBS), ORC Macro. 2005. “Kenya Service Provision Assessment Survey 2004”. Nairobi, Kenya: National Coordinating Agency for Population and Development.

Syndicate content
Copyright 2014 Innovations for Poverty Action. All rights reserved.