Peer effects in early schooling: Evidence from Brazilian primary schools

Empirical investigations of peer effects on student achievement are often challenged by methodological problems related to the simultaneous influence of peer interactions and selection bias. To minimize these problems, I estimate peer effects controlling for school fixed effects and using educational data at school entry, when students' baseline performance has not been influenced by their peers and ability grouping is less common because schools have unclear information about the ability level of incoming students. Using longitudinal data from Brazil, I compare students at the beginning and at the end of first grade to see whether classroom composition with respect to peers' academic performance impacts students' achievement gains in their first year in school. Consistently with previous literature, results suggest that students learn more when they attend classrooms with higher achieving peers. Whereas classroom heterogeneity in terms of peers' ability tend to have a negative but small effect on achievement gains in first grade. Empirical investigations of peer effects on student achievement are often challenged by methodological problems related to the simultaneous influence of peer interactions and selection bias. To minimize these problems, I estimate peer effects controlling for school fixed effects and using educational data at school entry, when students' baseline performance has not been influenced by their peers and ability grouping is less common because schools have unclear information about the ability level of incoming students. Using longitudinal data from Brazil, I compare students at the beginning and at the end of first grade to see whether classroom composition with respect to peers' academic performance impacts students' achievement gains in their first year in school. Consistently with previous literature, results suggest that students learn more when they attend classrooms with higher achieving peers. Whereas classroom heterogeneity in terms of peers' ability tend to have a negative but small effect on achievement gains in first grade. (C) 2017 Elsevier Ltd. All rights reserved.