Gender in higher education

• Autor(a) principal: Borges, Bruna Pugialli da Silva
• Data de Publicação: 2021
• Tipo de documento: Tese
• Idioma: eng
• Título da fonte: Biblioteca Digital de Teses e Dissertações da USP
• Texto Completo: https://www.teses.usp.br/teses/disponiveis/12/12138/tde-27052021-215611/
• topic: Economia; Economics; Efeito de pares; Ensino superior; Escolha de curso de graduação; Exames importantes; Gender; Gênero; Glass-ceiling; High-stakes assessments; Higher education; Labor market; Major choice; Mercado de trabalho; Peer effects; Teto de vidro

This doctoral dissertation comprises three research papers on gender issues at the higher education level. In the first paper, we investigate if women underrepresentation in Economics influences female students\' labor outcomes. We use administrative data from the Department of Economics at the University of São Paulo, linked with formal labor market data. To deal with endogeneity problems, we exploit the random assignment of students to sections in first-semester courses from the undergraduate program. We find that higher shares of female classmates increase female participation in the labor market. Also, higher percentages of female professors increase the likelihood of a female student working as a top manager. In the second paper, we use data from the UNICAMP admission exam to assess gender differences in reaction to an increase in the examination\'s stakes. At the UNICAMP admission exam, all applicants answer the same exam in both stages. Still, in Phase 2, there is a higher weight to priority subjects, defined according to the applicant\'s major choice. We find that women underperform in priority subjects. When we look at heterogeneous effects, we do not verify systematic heterogeneity in the impacts across subjects, but we see more substantial gender gaps for high-achieving students. The evidence suggests that gender differences in perfectionism and confidence could explain our results. Finally, the third paper examines if higher shares of female schoolmates in high school affect the students\' major choices. Using UNICAMP admission data combined with data from the Brazilian School Census, the identification strategy relies on idiosyncratic changes in cohorts\' gender composition within schools. The share of female peers during high school reduces the likelihood of selecting math-intensive majors. Besides, the percentage of own-gender peers increases the probability of choosing male-dominated fields and reduces the likelihood of choosing gender-balanced majors. Lastly, women exposed to higher shares of female schoolmates choose more competitive majors.