Three essays on the indirect impacts of conditional cash transfer programs in Mexico and Nicaragua

This dissertation contributes to the evaluation of Conditional Cash Transfer (CCT) programs by estimating indirect impacts of the Oportunidades and RPS programs in rural Mexico and Nicaragua, respectively, using experimentally designed data. The first essay estimates the impact of early nutrition intervention in the Oportunidades program on the probability of enrolling in primary school on time. Identification is achieved through a cross-sectional double difference estimator on observations from two cohorts of children. Although the results are not strongly significant, they do indicate that early nutritional interventions can have a positive impact on school outcomes and that care giver characteristics affect the magnitude of the impact. The results also highlight that a first difference estimate of the impact would be downwardly biased by unobserved differences in characteristics. The second essay examines the impact of the RPS program on an intermediate fertility outcome, birth spacing, using a stratified Cox proportional hazard model. Birth histories of female household heads and spouses are constructed from household roster information, and to deal with autocorrelation, the data are stratified by parity (the number of live births). The results indicate that the RPS program decreased the hazard of a birth, which resulted in an increase in the space between births of between 6 to 8 months. Further research is required to determine whether the impacts will persist in the long-run. The final essay explores whether the Oportunidades program relaxes liquidity constraints and can lead to increased agricultural production; this is explored by estimating the impact of the program on consumption of food from own production as well as on land use, livestock ownership and spending on agricultural crop production. Impacts estimated using first difference estimators are compared to those from weighted estimators in which the weights are constructed from propensity scores. The program is found to increase the value and variety of food consumed from own production; no differences in the two estimators are observed. In addition, small positive impacts on land use, livestock ownership and crop spending are observed. The implication of these findings is that part of the cash received is invested in agricultural production.