Can Targeted Food Programs Improve Nutrition? An Empirical Analysis of India's Public Distribution System

Despite the widespread belief that a substantial assault on poverty requires targeting scarce resources toward the poor, practitioners frequently claim that targeted programs deliver fewer benefits to poor households than do universal programs. This article evaluates this concern through an analysis of India's targeted food distribution program, the Public Distribution System. It first examines the case for targeting by assessing the responsiveness of caloric intake to the amount of the food grain subsidy; a low elasticity suggests that improvements in nutrition may require large subsidies and, hence, a targeted program. It then examines whether targeting adversely affects the probability of poor households participating in the program. The empirical analysis identifies the effect of the program by combining time-varying policy changes in the value of the program with cross-sectional variation in program benefits generated by variation in market prices. The results confirm the low responsiveness of caloric intake to food grain subsidies and hence suggest the need for a relatively generous program. However, I also find that, under the current design of the program, targeting reduces the probability of participation by poor households, relative to a universal program.