Each year, more than 600,000 people are released from prison. The obstacles to successful reentry are daunting, starting with the challenge of finding stable work. Indeed, a large proportion of released inmates return to prison within a relatively short time. In recognition of the enormous human and financial toll of recidivism, there is new interest among researchers, community advocates, and public officials in prisoner reentry initiatives, particularly those focused on employment.

In May 2006, the National Poverty Center at the University of Michigan hosted a meeting — “Research on Prisoner Reentry: What Do We Know and What Do We Want to Know?” — to discuss the state of research on employment-focused prisoner reentry programs. This paper, written as background for the meeting, reviews previous research, describes some planned or ongoing evaluations, and proposes some ideas for future research.

There have been few rigorous studies of employment-focused reentry models, and there is a pressing need for more definitive evidence of what works. A meta-analysis of eight random assignment design studies of postrelease community-based programs found that they did not reduce recidivism. There are only a few experimental studies on the effects of in-prison services, and it is hard to draw lessons from the nonexperimental research. Many experts believe that the most promising reentry models provide coordinated services both before and after inmates are released. There have been a few studies of such models to date, including two that used random assignment; the results were only somewhat positive.

It is clearly difficult to increase employment and earnings for disadvantaged men. Yet the results described above do not support the view that “nothing works.” Some programs seem to be modestly successful: those for older ex-prisoners, integrated services both before and after release, and perhaps models using financial incentives. Further attention to design and evaluation of prisoner reentry programs may produce useful results, as most of the studies are quite old, and both the economic and criminal justice contexts have changed dramatically in recent years.

Some large-scale studies now under way will dramatically expand the knowledge base, but some important gaps will remain. Approaches that might be tested in the future include: earnings supplements and work incentives; employer-focused strategies, such as the federal bonding program and supported placements to give employers incentives to hire ex-prisoners; in-prison vocational training; performance goals for parole officers that emphasize parolees’ employment; and programs that address motivational issues, such as faith-based initiatives, therapeutic models, and those that engage ex-prisoners’ families.