Direct and Indirect Effects of Malawi's Public Works Program on Food Security
Public works programs are common and costly, but there is little direct evidence about their benefits. In Malawi, the cash-for-work program is the largest component of the social protection agenda. The Malawi Social Action Fund Public Works Programme (MASAF-PWP) has been operational since the mid-1990s and aims to provide short-term income activities to poor, able-bodied households for the purpose of enhancing their food security, through increased access to farm inputs at the time of the planting period. The program has undergone several re-designs over the years. In the 2012/2013 season, it expanded both in terms of the duration of exposure and net income earned by participants (from 12 days to 48 days per beneficiary) and the inclusion of work in the second half of the agricultural season (extending the exposure from planting to planting and into the harvest period).
Professor Jessica Goldberg and two colleagues at the World Bank partnered with the Local Development Fund Technical Support Team, the Government of Malawi agency responsible for the PWP, to evaluate the effects of the program and the potential benefits of budget-neutral improvements to its design. The study used two types of random assignment in order to be able to estimate the causal effect of PWP. At the village level, communities that had requested PWP activities were either assigned to the “treatment” group or to the “control group.” Villages in the treatment group all had the same number of days of PWP employment at the same wage, but the research team varied the timing of employment. In the control group, villages did not have any PWP activities. The second level of randomization was at the household level. The program is not able to accommodate all poor households, and the research team randomly assigned some households to be included in the program. The design made it possible to estimate the direct effects of the program for households that were offered the opportunity to participate (“treated” households), and the indirect effects of the program on households in villages that had PWP but who were not invited to participate themselves (“untreated” housholds). The program did not achieve its stated goals: it does not improve the food security or increase the fertilizer use of households that were included in the program. Puzzlingly, there were negative spillover effects on the food security of untreated households in villages that had PWP but were not included in the program in some parts of the country.
Much of the research about PWPs assumes that beneficiaries’ welfare improves, and focuses on improving the targeting of the programs or understanding their effects on labor markets. This study challenges that assumption. Currently, Professor Goldberg and her colleagues are working with the World Bank team and Malawian officials on revisions to the MASAF-PWP. While the study does not provide definitive answers about the optimal design, it suggests that changes such as aligning work opportunities with the lean season and substantially increasing either the potential income or the insurance value of the program may improve its impacts.