An Analysis of Prisoner Reentry and Parole Risk Using COMPAS and Traditional Criminal History Measures

The California Department of Corrections and Rehabilitation has adopted Correctional Offender Management and Profiling Alternative Sanctions (COMPAS), an actuarial risk- and needs-assessment instrument, as part of its reentry supervision and parole planning procedure. A large-scale 3-year prospective study was conducted to assess the instrument with regard to how well it predicted whether a parolee would be rearrested for (a) any crime and (b) a violent offense. This study followed, for up to 2 years, a total of 91,334 parolees who had been assessed with COMPAS prior to reentry into the community. The instrument achieved an acceptable level of predictive validity in general rearrests with an area under the curve value of 0.70, but its predictive power for subsequent violent offenses fell short of this conventional threshold. Moreover, a parsimonious model using four known risk factors from existing official records (i.e., gender, age, age of first arrest, and the number of prior arrests) performed just as well in predicting subsequent arrests. Findings from this study illustrate the challenges in applying group-based attributes to predict individual criminal behavior and suggest that, although COMPAS has other attractive features such as case management capability, existing official records may offer a lower cost alternative for assessing the risk of reoffending for community reentry purposes.