Evaluating the direct and indirect effects of a conditional income support program: the case of Progresa -PhD thesis-
Poverty alleviation programs that target education, health and nutrition are increasingly popular amongst policymakers. Countries implementing these types of programs need to be aware that such policies could have adverse effects on non-targeted outcomes such as migration. Cash transfers may relax the budget constraint faced by poor households to migrate. This paper analyzes spillover effects, anticipation effects and direct effects of a poverty program on domestic and international migration in rural communities in Mexico. Particularly, we study the regional impact of Progresa (renamed Oportunidades) on migration. We apply difference-in-difference propensity score matching techniques to estimate the existence of spillover effects using as a control group a non-experimental sample of non-eligible households. Additionally, we use simple difference-in-difference techniques to evaluate the experimental impact of the program. We compare these results with those obtained using matching methods to learn about possible anticipation effects. The effects of Progresa on migration are ambiguous depending on the region, the type of migration and the amount of the transfer received. We find a few positive and significant spillover effects on eligible and non-eligible households. Positive anticipation effects on eligible households are also present. Overall, the program has a higher impact on domestic migration than on international migration.