Publication: Essays in educational and intergenerational inequality
Loading...
Identifiers
Publication date
2019-07
Defense date
2019-09-23
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
In this dissertation, I study different dimensions of socioeconomic inequality. My objective
is to produce empirical evidence to help governments design better policies in the fields of
education and the labor market. The two first chapters are focused on educational inequality and the effect of policies that aim to tacke this issue, while the third chapter aims to present useful solutions to an important conceptual issue for measuring intergenerational mobility.
In the first chapter, ‘Affirmative Action, Centralized Admissions and Inequality in Access
to Higher Education: Evidence from Brazil’, I analyze how two major reforms, introduced
to democratize access to public higher education in Brazil, impacted enrollments of students from a low-socioeconomic status. The first policy centralized applications in a nationwide platform and the second expanded affirmative action quotas to a uniform share of fifty percent of all vacancies offered by each major and institution. Their progressive adoption generates cross-sectional and time variation, allowing the separate identification of their causal effects. Results show that the affirmative action reform increases enrollments of public school, black and low-income students, while the centralized admission system acts in the opposite direction, decreasing their participation. Moreover, the interaction between both policies has a positive and significant effect on enrollments of the vulnerable groups. I then shed light on some mechanisms behind these results. I find that centralization disproportionately increases enrollments of high-SES out-of-state students in the least prestigious degrees, crowding-out low-SES students with mobility constraints. On the other hand, the expansion of affirmative action does not only mechanically improve equity, but also changes application behavior.
In the second chapter, ‘Does Affirmative Action in Undergraduate Education Impact High
Schools? ’, I delve into the analysis of unintended consequences of affirmative action initiatives in higher education, which have been implemented in different countries to improve access of vulnerable groups and to reduce inequality in educational attainment. A growing empirical literature has investigated how such policies impact college students’ outcomes and pre-college human capital accumulation. Yet, little is known about how they affect students’ choice of high school and, consequently, school quality and peer interaction. I study this question in the context of Brazil, one of the most unequal countries in the world, and where the government approved, in 2012, the "Quota Law (QL)”. It established that fifty percent of all vacancies in each major and federal higher education institution, including some of the best universities in the country, has to be reserved to students that attended secondary education integrally in a public school. I show that the adoption of QL increases strategic mobility from private to public schools by 29 percent and that the movers come disproportionately from low-SES and low-quality private schools. Nevertheless, this exogenous influx of private school students increases public school quality, while it also raises inequality within the public school system. Finally, in the third chapter, ‘Correction Methods for Intergenerational Mobility Estimates’,
co-authored with Martin Nybom and Jan Stuhler, we study another dimension of
socioeconomic inequality: the transmission of economic status between generations. The
estimation of standard measures of intergenerational mobility ideally requires the complete income history for two generations to determine their lifetime incomes. However, empirical applications are typically based on snapshots of income over a limited number of observations in the life cycle. If those snapshots do not mimic lifetime outcomes, the estimates are subject to attenuation and lifecycle bias. The literature has followed two different strategies to address this problem. The first models the income processes itself, the second the relation between annual and lifetime incomes over the life cycle. In this paper, we use uniquely long income series from Sweden to study how well these methods approximate the intergenerational elasticity of income. All methods are biased to some degree, because neither accounts for three key components of the income process: (i) income growth explained by observable characteristics, (ii) transitory noise, and (iii) unexplained income growth that correlates within families. We propose a lifecycle estimator that addresses all three components, and which can improve estimates of the intergenerational elasticity in a wide range of settings.