We study the impact of a child sponsorship program on the intended choice of schooling of children in the states of Oaxaca and Chiapas in Mexico. In order to account for the program's procedure to select the children, we estimate a binary Roy type model, with unobservables generated by a one-factor structure. We use both carefully scripted survey questions to elicit the subjective-expected returns to schooling, and the results in Aakvik et al. (2005) to estimate the sponsorship effect on the revealed intended choice of acquiring a technical degree or university studies. The sponsorship effect is positively correlated with selection to the program, with an average sponsorship effect on the sponsored of 20 percentage points and a null effect on the average sponsorship effect. Nevertheless, the standard deviations are large, and both the average sponsorship effect on the sponsored and on the population are not statistically signicant.
Presented in Session 8. Economy, Labor Force, Education, & Inequality