Abstract:

This dissertation explores three issues in high-school education in Colombia. Chapter one examines the gender gap in test scores in the Colombian standardized test used to enter post-secondary institutions. I find that males outperform females even after individual, household and school characteristics are taken into account. I then go on to examine the influence of alternative activities to education on the self-selection of males. Using mortality indicators as a proxy, I find that 28% percent of the gender gap in test scores is explained by the large degree of illegal activities in Colombia that disproportionately affects the selection process of men.

In chapter two, I examine the advantage private-school students have on public-school students when it comes to two educational outcomes in Bogot?: post-secondary test scores and college entrance rates. I find that after controlling for socioeconomic status, the advantage of private-school students halves for college entrance and becomes statistically insignificant on tests scores. Using distance to schools as instrumental variable, I also find that there is no evidence that the estimators are biased. Instead, the advantage of private schools is the result of more resources available to students. Specifically, private schools have a higher computer to student ratio and more language laboratories than do public schools. In short, the advantage of private schools over public schools may be due to in part to self-selection, but it is also due to better resources.

In chapter three, I examine whether competition among schools improves school quality in Bogot?. Evidence of competition is first examined. While theoretical models suggest that competition results in segregation, a descriptive spatial analysis of school quality in Bogot? shows a polarization of schools where high quality schools cluster in the north and low quality schools cluster in the south. Further analysis shows that private and public schools are segregated by income but not by ability. The spatial models confirm that spatial order is explained by the socioeconomic spatial segregation in Bogot?. Furthermore, the close proximity of private schools increases the quality of a public school even after controlling for household characteristics.