Essays on University Admission Systems
Sisu; Enem; selection systems; student mobility; student diversity; academic trajectory.
Selection systems organize competition for higher education places and can alter how educational opportunities are distributed across programs and institutions. In Brazil, with the consolidation of Sisu and the growing use of Enem scores as an admission criterion, access has become more nationally integrated, although through distinct logics in the public and private sectors. In this context, this thesis examines how this coordination of admission is related to the profile of incoming students, student mobility, and trajectories of dropout and completion. The first essay analyzes the extent to which the adoption of Sisu is associated with changes in the profile of incoming students and in trajectories of mobility, dropout, and completion in in-person public higher education programs, using data from the Higher Education Census and Sisu (2010–2023) and difference-in-differences estimates. The results indicate a reduction in female participation by 2.2 p.p. and an increase in the share of students admitted through reserved places by 5.1 p.p., with no consistent changes in mobility. Higher dropout is also observed among students admitted through Sisu, by 2.6 p.p. in the first year and 8.1 p.p. in the last observed year, as well as lower completion by 5.7 p.p., compared with students admitted through the vestibular. Placebo tests do not contradict the main interpretation of the findings. For female participation, heterogeneity analyses show a sharper reduction in Education programs and in institutions with higher IGC scores, while the intersectional analysis indicates a reduction in the participation of white female students and an increase in that of black female students from public schools. The second essay examines, in in-person private higher education, how the use of Enem scores in admissions is related to student profile, mobility, and academic trajectories. Using Higher Education Census microdata (2010–2023) and difference-in-differences estimates, the results suggest greater female participation (3.9 p.p.) and younger students (9.7 p.p.), as well as a lower proportion of white students (-8.5 p.p.), public high school graduates (-15.8 p.p.), and students with financial aid (-7.6 p.p.), along with an increase in the presence of students receiving social support (15.5 p.p.). In this case as well, no consistent changes in mobility are observed. In the private sector, however, admission based on Enem scores is associated with lower dropout (-15.2 p.p. in the first year and -9.9 p.p. in the last observed year) and higher completion (11.6 p.p.), compared with admission through the vestibular. The third essay incorporates the time dimension until dropout or completion. Using Higher Education Census microdata (2012–2017), the strategy combines propensity score methods and survival analysis to examine federal universities and private institutions. The findings indicate that, in the federal sector, students admitted through Sisu and Enem tend to drop out earlier and complete later than those admitted through the vestibular. In the private sector, greater persistence is observed among students admitted through Enem, associated with a shorter time to completion. Overall, the thesis contributes to the debate on higher education by highlighting the relevance of selection systems and the importance of articulating these instruments with other educational policies.