Preventing school-based arrest and recidivism through prearrest diversion: Outcomes of the Philadelphia police school diversion program.

Abstract

Objectives: Created to combat the school-to-prison pipeline, the Philadelphia Police School Diversion Program offers voluntary community-based services to eligible youth accused of minor school-based offeses in lieu of arrest. This study evaluated program effectiveness in accomplishing goals related to reductions in school-based arrests, serious behavioral incidents, and recidivism. Hypotheses: We expected the annual number of school-based arrests in Philadelphia schools to decrease over the program's first 5 years and predicted that the annual number of serious behavioral incidents would not increase. Further, we expected that diverted youth-compared to youth arrested in schools the year before Diversion Program implementation-would have significantly lower rates of recidivism arrests in the 2 years following their school-based incidents. Method: Using a quasi-experimental design, we examined data from 2,302 public school students (67.0% male; 76.1% Black; age range: 10-22 years) who were either diverted from arrest through the Diversion Program or arrested in Philadelphia schools in the year prior to Diversion Program implementation. We compared rate of recidivism arrest, number of arrests, and time to arrest between diverted and arrested youth. We also used district-wide descriptive statistics to examine 5-year trends in school-based arrests and serious behavioral incidents. Results: Since program implementation, the annual number of school-based arrests in Philadelphia has declined by 84% and the number of serious behavioral incidents has declined by 34%. Diverted youth demonstrated less recidivism than arrested youth in the 2 years following their initial incident; however, after propensity score matching, we no longer observed significant differences. Conclusions: Findings indicate that a prearrest diversion program can safely reduce school-based arrests and suggest a need for future research regarding the role of demographic and incident-related characteristics in recidivism outcomes.