Rural poverty remains a pervasive problem in Malawi, and many poverty reduction interventions focus on transitioning rural households out of subsistence agriculture. In a country like Malawi, where smallholder farmers constitute roughly 80 percent of the population and produce about 80 percent of the country’s food, this approach is unlikely to be effective, particularly in the short term. Susan Godlonton (Williams College), and colleagues Kate Ambler and Alan de Brauw at IFPRI explored the potential of an alternative solution using an intervention inspired by the Brazilian Fomento program, which connected beneficiaries of the Bolsa Familia cash transfer program with extension agents.
The study, “Agriculture support services: Direct effects, complementarities, and time dynamics,” used a randomized control trial to evaluate an intervention combining cash and input transfers and agricultural extension services on the crop production and its total value in Malawi. The intervention was founded on the premise that insufficient farm management and technical skills, as well as limited resources, constrain agricultural productivity. The study’s cash and input transfers sought to address resource constraints, while different models of extension support were used to enhance farmers’ capacity to implement improved agricultural practices and farm management skills.
Working in collaboration with the National Smallholder Farmers Association of Malawi (NASFAM), 1,200 households from 120 recently registered farmers’ clubs received one of three transfer treatments in the first year of the program:
- Cash transfer group: received a series of three cash transfers at planting time, mid-season, and harvest time
- Input transfer group: received a package of in-kind inputs of seeds, hoes, inoculants for nitrogen fixation, storage sacks and string, and cash for ganyu (hired day labor) and transportation
- Control group: did not receive any transfer; received standard NASFAM support
Transfers were only given in the first year of the project to encourage farmers to make similar investments in the second year. The size of the transfers was intentionally large – valued at roughly US $84 – to enable farmers to make substantive investments. Transfers were also intentionally spaced out to address constraints at key points throughout the agricultural season.
With regards to extension support, the study differentiated between types of agricultural extension by randomizing provision of “intensive extension” and standard NASFAM extension services. Standard NASFAM extension is based on the lead farmer model, in which selected community members are placed in leadership roles and serve as liaisons between hired extension agents and other farmers. The “intensive extension” model employed in this study gave farmers additional support via professional extension agents who conduct three, one-on-one visits with each farmer.
First-year intensive visits were used to develop farm management plans for the entire agricultural season, with the two follow-up visits used to reference this plan and provide additional technical support. Farmers were randomly assigned one form of extension support in year 1, with extension re-randomized in year 2 based on farmers’ year 1 assignment. (For more information on the research design and randomization, see the figure below, the draft report, and the PowerPoint presentation).
The research design made it possible to separate the impact of different types of transfers and agricultural training. The authors found that both cash and in-kind transfers increased crop production, as well as the value of that production. Total value of production increased by 25 percent in the first year for those in the cash group. In the second year, overall crop production increased by 21 percent for those receiving in-kind inputs, and 36 percent for those receiving cash (the significance of the difference between cash and input packages was limited).
Production gains were attributed to both multi-seasonal and once-off household investments, in particular increased investment in agriculture-specific assets and use of pesticides in the first year. There was also an increased use of and expenditure on ganyu (hired day labor) in both years. While impacts of the transfers were similar for male and female farmers, there was a general pattern suggesting that the impact was greater for the women involved.
Regarding the provision of agricultural extension, the study found no evidence that intensive extension led to increased agricultural production during the year of implementation, and no direct evidence that intensive extension is more effective than lead farmer extension. However, the results did highlight important time dynamics related to extension service provision, as well as important complementarities between receipt of extension and transfers. Impacts were greater for farmers who received both the transfers and extension service in the first year, with an 82 percent increase in the gross value of crop production among farmers receiving either of the transfers and extension service in relation to the control group.
These results have important implications for the design of future programming to increase agricultural productivity. As transfers are growing in popularity as a means to eradicate rural poverty, this study provides important evidence about the size, modality, and timing of transfers, as well as questions for further research about time dynamics and complementary interventions.
