Despite achieving a significant cost reduction over the past two decades, the absolute cost of food subsidies in Egypt is still high relative to the benefits received by the poor. There is scope for better targeting these food subsidies, in particular for targeting cooking oil and sugar ration cards, both because reforms in this area are perceived to be far less politically sensitive than adjusting subsidy policy for bread and wheat flour and because higher income groups presently receive a significant percentage of the benefits. Targeting the high-subsidy green ration cards to the poor and the low-subsidy red ration cards to the nonpoor will require identification of both poor and nonpoor households. An International Food Policy Research Institute research team in Egypt, in collaboration with the Egyptian Ministry of Trade and Supply, developed a proxy means test for targeting ration cards. The paper describes the process of moving from the optimal income-predicting model to the final model that was both administratively and politically feasible. An ex-ante evaluation of the levels of accuracy of the proxy means testing model indicates that the model performs quite well in predicting the needy and nonneedy households. An effective and full implementation of this targeting method would increase the equity in the ration card food subsidy system, and, at the same time, the total budgetary costs of rationed food subsidies would decline. Moreover, the experience gained under this reform would facilitate targeting future social interventions to reduce and prevent poverty in Egypt. (C) 2002 Elsevier Science Ltd. All rights reserved.
