Students across the United States experience high levels of contact with the police. To clarify the causal relationships of this contact with educational outcomes and the mechanisms by which such relationships arise, we estimate the effects of arrest on student engagement with school using daily attendance data. Recently arrested students missed significantly more school than did students who would be arrested later in the school year. The effects of arrest on attendance can be attributed to suspensions and court appearances; we found little evidence of changes in absences due to health or skipping school. These results suggest that institutional, not student centric, mechanisms drive the relationship between arrest and educational outcomes. Were it not for institutional channels, particularly exclusionary discipline, arrested students would likely remain more engaged in school. Estimates are similar for white and black students, but black students are differentially affected because they are arrested at higher rates.