Does admission type matter? An analysis of the performance of federal high school students in Brazil

Abstract

This study aims to verify whether a difference exists between the forms of entry (lottery draw vs. merit-based) on the performance of students entering the federal education schools. The data used were collected from students in their last year of high school in relation to mathematics and Portuguese. As each federal education unit has administrative independence in terms of the selection method, the identification strategy is based on assigning students to two groups: treatment (lottery draw) and control (merit-based admission process). Three matching methods were employed to estimate these effects. The empirical results indicate that an effect exists between 0.14 and 0.40 standard deviation smaller in mathematics for students who enter by lottery draw, with differences in gender and the distribution of grades. Among the female students, the effect was greater in Portuguese, and among the male students, the effect was greater in mathematics. Students in the lower quantiles of grade distribution showed the greatest difference.