Abstract: This thesis consists of three essays in development economics. The first essay uses the air-bridge denial policy imposed by Peru in 1995, which shifted the production of coca leaves from Peru to Colombia, to analyze how an exogenous decrease in the demand for coca affects children's allocation of time. The Peruvian Living Standard Measurement Survey allows a comparison before and after the counter-narcotics policy for rural children living in coca-growing regions to children in non-growing regions not able to produce coca leaves due to environmental conditions and thus not affected by the change in policy. After different sensitivity checks, the results indicate that child labor increased in coca-growing states. Work hours and hours engaged in household chores increased as well. The second essay tests empirically the latest theories in which child labor does not have a linear relation with household income. Using data from Peru and adopting a multivariate nonparametric approach, with continuous and discrete covariates and cross-validation methods for optimal bandwidths, the results reveal interesting patterns not detected in the child labor literature. The estimated elasticities are larger than those from previous studies and suggest that changes in income have a heterogeneous effect on employment and school participation with higher effects for low-income households. The third essay examines heterogeneous effects of a conditional cash program in Nicaragua, which includes the random assignment of communities to a treatment and control group so that outcome differences represent the true mean impact of the program. Using quantile regression, the results show that the program had heterogeneous effects across the distribution of the untreated outcome distribution. In particular, the program impact in food and total per capita expenditures is lower for households who were at a lower level of expenditures prior to the program. In addition, the program impact is higher for households who had lower levels of food shares prior to the program. Finally, the program has greater impact in increasing schooling for boys and reducing child labor for girls, and children located in a more impoverished comarca received less impact of the program in schooling and greater impact on working hours.