Identifying targeting with nonparametric methods: An application to an Indian microfinance program

We discuss nonparametric methods and statistical tests that are appropriate to assess poverty targeting in public programs. These methods explicitly account for the possibility that the population distributions of participants and non-participants cross. Crossing points provide us with upper bounds on the income of those who have been excluded from the program. Applying these methods to data from a microfinance program in the state of Jharkhand in India, we find evidence that very poorest households are largely excluded from the program.