THREE EMPIRICAL ESSAYS IN DEVELOPMENT ECONOMICS

This dissertation studies three empirical questions on development economics. The first two chapters investigate the impact of the Mexican health program Seguro Popular on vulnerability to poverty and educational outcomes. Seguro Popular is a health program that provides healthcare services and medicines to households not covered via salaried jobs and does not provide cash transfers or income to disabled individuals. Enrollment is free for households in the poorest three income deciles. Through November 2018, it covered about 50% of the population. The last chapter investigates the extent to which the Peru’s 2010 Progress out of Poverty index (PPI) can be used to predict poverty transitions of households. The PPI has become a preferred proxy means test used by government agencies and non-governmental agencies for its accuracy in measuring poverty and as it is quick, inexpensive, and easy to implement.Currently, the Mexican government evaluates poverty using a national income poverty line and monitors poverty using a multi-dimensional poverty index that tries to account for other aspects of wellbeing beyond a monetary metric. However, there is no estimation of the level of exposure to future poverty, which is an element that should be relevant for anti-poverty and human development programs, especially because households around the poverty line are likely to become poor after they experience a negative shock. Chapter 1 investigates the extent to which Seguro Popular can be used to reduce vulnerability to poverty in Mexico. The empirical analysis provides evidence that Seguro Popular helps reduce vulnerability to poverty. A more granular analysis also shows that households are exposed to different factors, such as macro-economic conditions, labor income changes, and weather events, and that Seguro Popular only helps households cope with negative shocks related to health-related expenses. Thus, when analyzing the program’s impact on vulnerable households, Seguro Popular does not have an impact on the most vulnerable households, households with the fewest assets, individuals with the least education, or individuals working mostly in non-salaried jobs. Chapter 2 evaluates the impact of Seguro Popular on education outcomes. Previous studies have helped elucidate the unintended effects of Seguro Popular on labor outcomes decisions, fertility, and healthcare services use, among others. Overall, this study validates the hypothesis that Seguro Popular has a positive impact on educational outcomes for children aged six to 11. On average, participation increases enrollment in elementary school and increases the time spent on schoolwork. Additionally, among the unintended effects, the labor time of children decreases for households enrolled in Seguro Popular. Finally, Chapter 3 investigates the extent to which the PPI can be used to predict poverty transition. Chapter 3 analyzes the PPI indicators that are used to estimate the likelihood that households can be identified as living in poverty. In doing so, it quantifies both the relevance of each variable and the predictive power of the PPI score to determine poverty status. For policymaking, a metric that can help predict future status and measure graduation out of anti-poverty programs, while remaining accurate and limiting measurement errors (which are a waste of the financial budget), is necessary. The results show the effectiveness of using PPI to predict poverty transition in terms of targeting accuracy and measurement errors. However, as with previous results, this study shows that such an approach should be complementary to additional information.