Summary statistics.

This study tests the role of violence and policing in predicting student absences at the school level in New York City. It uses a large dataset on daily attendance over six school years (2013–2014–2018–2019) across all New York City public schools, and operationalizes policing by arrests, and violence by reported violent crime. While much literature focuses on the impact of violence on student outcomes, this study finds that arrests are, in fact, a relatively strong predictor of school absences. Perhaps more importantly, nearby arrests have a uniquely strong association in schools with greater numbers of Black students. In high schools, nearby arrests negatively affect school absences, but only in schools with a high proportion of Black or low-income students. Nearby violence has a small association with absences in K-8 schools, but no significant association with absences in high schools. These findings underscore the need to understand policing’s disparate impact on educational outcomes.