Overview of: “For whom does a transitional jobs program work?

Examining the recidivism effects of the Center for Employment Opportunities program on former prisoners at high, medium, and low risk of reoffending”

Research Summary

This study documents that a transitional jobs program for former prisoners had its strongest reductions in recidivism among those in the program with the highest risk of reoffending. The New York City-based Center for Employment Opportunities (CEO) is a transitional jobs program designed to help former prisoners increase longer term employment and, consequently, reduce recidivism. Interim results from MDRC's rigorous impact evaluation of CEO showed reduced recidivism in both the first and the second year of follow-up. The current study

• • 
  expanded on the interim results by using regression-based analysis to identify whether CEO had its greatest impact among low-, medium-, or high-risk offenders—with risk levels defined by participants’ characteristics before random assignment that are associated with recidivism after random assignment.
• • 
  found that CEO had its strongest reductions in recidivism for former prisoners who were at the highest risk of recidivism. For high-risk former prisoners, participation in CEO reduced significantly the probability of rearrest, the number of rearrests, and the probability of reconviction 2 years after random assignment to the program.

Policy Implications

The findings suggest important implications for policy, practice, and future evaluation research.

• 1
  The limited resources available to transitional jobs programs for former prisoners should be targeted toward people at the highest risk of recidivating because they are helped most by this intervention.
• 2
  Age and criminal history are critical determinants of recidivism risk, and programs for former prisoners should consider assessing the likelihood of reoffending using tools that measure both characteristics. Although specific to our sample, the average-aged offender (33 years old) was considered at a high risk of recidivism if he had nine or more prior arrests; similarly, those with seven prior arrests (the sample average), were at high risk of recidivism if 28 years old or younger.
• 3
  Because CEO interim evaluation results did not show an effect on increases in unsubsidized employment, it is not clear what is causing the recidivism effects. Thus, future evaluation research should examine the mechanisms by which this transitional jobs program reduces recidivism among its clients, and particularly, the highest risk clients.