How best to target agricultural subsidies? The case for an indicator-based targeting system in Malawi

Over the past few years Malawi made remarkable progress toward increasing its national maize production and achieving food security owing to its long-running policy of subsidizing fertilizer. The implementation of these subsidy programs is continuously being improved upon as the country learns from past experiences.1 Recent evaluations of the current Farm Input Subsidy Program by Dorward and Chirwa (2011, 2012) suggest that various components of the program have been redesigned over time. These include timeliness of fertilizer delivery, fertilizer coupon receipts, regional distribution, area targeting, allocation and distribution processes, and coupon use and re-demption. However, the issue of beneficiary identification and targeting remains a challenge for the program. Draw-ing on recent research by Houssou and Zeller (2010, 2011, 2012), this note proposes an alternative approach to effec-tively target the poor within the group of potential program beneficiaries. In our analysis we are mindful of the fact that targeting criteria for the current program are vague at best. One interpretation of these criteria is that the pro-gram intends to target the “productive poor”, and since poverty is widespread in Malawi, some would argue that the focus of the targeting mechanism should be on identifying farmers who would make best use of subsidized fertilizer (e.g., as measured by their marginal productivity per unit of fertilizer applied). Many evaluations, however, have criti-cized the program for not being pro-poor enough; hence aNon-PRIFPRI1; MaSSPDSG