When allocating contracts, governments decide between exercising hiring discretion or allowing higher competition without firm selection. Ex-ante, it is not clear which allocation format leads to better outcomes. The trade-off depends in part on the government’s ability to select the best firms when restricting competition and on the probability that this practice leads to corruption. In the first chapter, I study the allocation of street pavement contracts in Mexico. I combine auction methods with a productivity analysis to indirectly test whether local governments select firms with low excess costs when restricting competition. This indirect approach allows the monitoring of contract allocation in settings with little information on firm reputation. I find that firms selected to settings with less competition have lower costs in complex pavement contracts, but higher costs in simple ones. Contrary to the current practice, the results suggest that the government would benefit from using public auctions for simple pavement contracts.

The next two chapters study investment in human capital, where I study the impact of a child sponsorship program on psychological constructs and educational aspirations. Previous work finds positive effects from international child sponsorship programs on educational attainment, employment, and adult income. I study whether elevating aspirations and socio-emotional skills among sponsored children work as mechanisms through which sponsorship programs impact children. For a sample of 2,022 children in Kenya, Indonesia, and Mexico, I use an age-eligibility rule applied during program rollout to identify the sponsorship effect. I document that sponsorship increased indices of self-esteem (0.25 std. dev.), optimism (0.26 std. dev.), aspirations (0.29 std. dev.), expected grade of completed education (0.43 years), and actual grade completion by 0.56 years.

For the Mexican subsample, I further study both the impact of subjective expected educational returns on aspirations and the correlation between selection to sponsorship and the sponsorship effect. I find that the average treatment effect on the treated is positive and consistent with previous studies, although not statistically significant. The sponsorship effect is higher for children most likely to be selected to the program. Finally, I document that children in rural settings have realistic, although heterogeneous income expectations, and present clear gender gaps.