Targeting the poor in Mexico: An evaluation of the selection of households into PROGRESA

In this paper, we conduct an evaluation of the targeting method used by Health Education and Nutrition Program (PROGRESA) of Mexico to identify beneficiary households. We address two key questions: (a) How well does PROGRESA's targeting perform; and (b) How does the program perform in terms of its impact on poverty alleviation relative to other feasible methods and transfer schemes. The first question is accomplished by comparing PROGRESA's method to an alternative selection method based on household consumption, which is our preferred measure of welfare. We employ the concepts of undercoverage and leakage and find that PROGRESA selection method is more effective in identifying the extremely poor localities or households but less so when it comes to distinguishing among localities or households in the middle of the scale. To address the second question, we compare the potential impact of PROGRESA on poverty alleviation against uniform transfers that involve no targeting at all, targeting based on consumption, and geographic targeting (i.e., targeting at the locality level rather than at the household level). We find that PROGRESA's method of targeting households outperforms uniform coverage and targeting at the locality level in terms of reducing the poverty gap and severity of poverty indices, even after taking into account the economic costs of targeting. But, the closeness of PROGRESA's performance to what could be achieved by geographic targeting alone raises some serious questions about the costs and benefits associated with the practice of household targeting within poor localities. (C) 2001 Elsevier Science Ltd. All rights reserved.