Essays on education economics

Abstract

This dissertation is intrinsic related to the literature on Education Economics, where we investigate some inputs of the education production function. We possess data sources comprising characteristics of applicants and students to one of the major flagship universities in Brazil and very restrict data on tax-registered firms, that enable us exploit the influence of two of these inputs on a series of academic and labor outcomes. In the first chapter, we examine the labor market returns to attending free elite higher education. Using restrict-access data from a flagship university in Brazil and from tax-registered firms, we explore an entrance rule that generates exogenous variation close to admission cutoffs, allowing us to compare marginal applicants and to estimate the causal effect of enrollments on salaries and occupations. Our findings indicate that enrolling in the elite free university raises wage premiums on 8% and the likelihood of reach reputable occupations in the future. The benefits are more expressive among female applicants and those with poorer backgrounds. We also provide evidence that, much more than just having the advantage of have a higher degree diploma, the elite wage premiums are guided by better matches on jobs demanding more specialized tasks. Our results are valuable for policy debates related to interventions aiming on promoting access to selective higher education for disadvantaged social groups. In the second chapter, we address the impact of an almost unexplored side of peer effect on students achievement and incentives to graduation regarding social comparative advantages. We estimate the effect of perceived rank in college and show that being last among the best increases the willingness to switch careers and reduces the likelihood of having a more prestigious occupation. To do so, we exploit a discontinuity in the class assignment in a flagship university in Brazil that sends the median student to either a better or a worse class in the same major program. Since the skill difference between classes varies within and between programs, we find that the ranking effect can be cancelled out by a high increase in peer quality. Our findings imply that the perceived rank sends a misleading signal, making similar students in the same program take distinct decisions and have different long-term outcomes. Higher parental education and stronger convictions about future earnings reduce the influence of this signal.