Program evaluation of agricultural input subsidies in Malawi using treatment effects: methods and practicability based on propensity scores

The study evaluates the impact of two agricultural input subsidies in Malawi during the 2003/04 and 2006/07 production periods on household income. The study employs quasi-experimental econometric techniques that use propensity score matching to control for selection bias on beneficiaries. A household model for each dataset is estimated together with Average Treatment Effects on the Treated. The evidence suggest that the matching mechanism performs well in evaluating the impact of the starter pack program which had a significant negative impact on household income compared to the refined agricultural input subsidy program which showed significant positive impacts on household income.