Can Social Programs be Reliably Evaluated with NonExperimental Methods? Evidence on the Performance of Regression Discontinuity Design using PROGRESA data

In 1997 a social program called PROGRESA was introduced in Mexico using a design for a randomized experiment. We exploit a build in, but neglected, discontinuity in the eligibility rule and use the quasi-experimental Regression-Discontinuity design in order to estimate marginal average treatment effects. Our findings show substantial regional variation. Moreover, given that the RDD approach allows us to use only data from the treated sample, we are able to investigate the extend to which the introduction of the program had an effect on ineligible children in the localities it was introduced and compare its performance to the experimental outcomes.