Zero tolerance discipline policies have lost favor in recent years due to concerns that they reduce offending students’ classroom time, beget further misconduct, and decrease student engagement. In the same vein, school police programs often associated with zero tolerance policies are frequently charged with increasing student involvement in the criminal justice system. Despite much discourse on the topic of school discipline, few studies have rigorously examined causal links between zero tolerance, school-based law enforcement, and student outcomes. This dissertation examines the effects of dismantling zero tolerance and reducing police officer presence in Philadelphia schools on school-level rates of student misconduct, administrative responses, and academic achievement. Quasi-experimental methods applied to the data include propensity score matching, generalized difference-in-differences analysis, comparative interrupted time series analysis, and fuzzy regression discontinuity analysis. Results suggest that dismantling zero tolerance did not affect school arrests rates or the rates of incidents involving law enforcement; and while transfers and expulsions decreased two years after the policy change, truancy increased. Limiting school police officer staff positions may have led to declines in the rates of incidents involving law enforcement and arrests, but the evidence is weak due to low statistical modeling power. Areas for future work are discussed.