摘要
This study determines the factors underlying the allocations of food aid in Ethiopia. We focus on regional differences in targeting criteria, and targeting accuracy according to per capita income. Data are drawn from two linked rural household surveys in 1995–96. We find large differences in food aid allocations across regions that cannot be explained by observable regional characteristics such as per capita income and rainfall. These differences are consistent with speculation that food aid is being used by the Ethiopian government to transfer resources to favored regions. We also find wide variations in the criteria used to identify recipient households across regions. We identify measurable indicators that could be used by food aid authorities to improve targeting effectiveness in the future, both across and within regions. Finally, we present simulation results on targeting accuracy under various targeting strategies and discuss potential benefits and shortcomings of those strategies.