Does U.S. food aid have disincentive effects on local food production?

Food aid is an important tool to fight poverty and avert humanitarian crises in developing countries. Although multiple studies on U.S. food aid have been carried out in the past few decades, the answer to the question— what is the impact of food aid on agricultural production— remains unclear. Many empirical studies focus on the household level, and their results vary widely depending on the characteristics of the country as well as the time period of interest. Our study, instead, uses country-level data in 118 recipient countries from 1961–2006. This paper examines the patterns of food aid allocation using a random effects tobit model, and provides policy-makers causal evidence of the average effects of food aid on food production. To address reverse causality between food aid and local production, we adopt the instrumental variable method as the identification strategy. Specifically, we use a three-way interaction term of U.S. wheat stocks, the dollar amount of U.S. military assistance measured in logarithm, and a measurement of a country's alliance with the U.S. as the instrument for the endogenous variable food aid. In addition, we reconcile commodities in production and food aid data to make sure that exactly same commodities are included in the aggregate group. We find that U.S. food aid shipments are responsive to donor's political interests as well as recipient countries' needs. U.S. cereal aid may negatively affect cereal production in recipient countries. When the U.S. doubled food aid donations, production would decrease by 1.5% on average, which means that an increase in the amount of U.S. food aid by a mean value of 70,832 metric tons would result in an average reduction in production of 173,952 metric tons. This disincentive effect of food aid on production is particularly significant for Sub-Saharan African countries, low-income countries, and regular recipients of U.S. food aid. However, food aid generates insignificant adverse impact on production when food aid is used for humanitarian needs. The heterogeneous effects of food aid on production are important for policy makers in order to evaluate and redesign the food aid programs