The educational consequences of receiving peers from violent areas: Evidence from a Latin American context of escalating violence

This paper estimates the association between receiving new peers from violent municipalities and the academic performance of the classmates who welcome them. Using individual-level administrative data from the Chilean Ministry of Education matched with municipal homicide records, we focus on two post-pandemic cohorts of fourth-grade students who did not change schools between third and fourth grade. Our empirical approach exploits within-municipality variation in the arrival of new peers, distinguishing whether incoming students are highperforming, come from high-violence municipalities, or both, while controlling for student, school, and municipality characteristics and including municipality and year fixed effects. We find that receiving at least one new peer from a high-violence municipality is associated with a 0.03 standard deviation decline in stable students' test scores. This negative association is driven by lower-performing peers from violent areas. We find no evidence that high-achieving new peers from these municipalities influence classmates' performance, suggesting that community violence dampens the academic benefits typically associated with high-achieving peers. These findings extend the literature on peer effects and community violence, largely developed in contexts of historically high violence, to a setting of escalating but comparatively low violence, and provide policy-relevant evidence on an underexamined channel through which Chile's rising insecurity affects children's educational trajectories.