This paper presents early results from an evaluation of the Center for Employment Opportunities (CEO) in New York City, a highly-regarded employment program for former prisoners. The evaluation is part of the Enhanced Services for the Hard-to-Employ Demonstration and Evaluation project, sponsored by the U.S. Department of Health and Human Services, with funding from the U.S. Department of Labor. The project is led by MDRC, a nonprofit, nonpartisan education and social policy research firm, along with the Urban Institute and other partners.
More than 650,000 people are released from prison each year. These ex-prisoners, many of them parents of children receiving welfare, face serious obstacles to successful reentry, and rates of recidivism are high. Most experts agree that finding steady work is one of the central challenges they face. CEO uses a distinctive transitional employment model. After a four-day job readiness class, participants are placed in temporary, minimum-wage jobs with crews that work under contract to city and state agencies. Within weeks, they receive help finding permanent jobs and, later, services to promote employment retention.
The evaluation targets a key subset of CEO’s population — ex-prisoners who showed up at the program after being referred by a parole officer. It uses a random assignment design: in 2004 and 2005, nearly 1,000 people were assigned, at random, to the regular CEO program or to receive basic job search assistance (this is called the control group). The research team is following both groups for several years, using surveys and administrative data to measure the program’s impact on employment, recidivism, and other outcomes. At this point, data on employment covered by unemployment insurance (UI) and several measures of recidivism are available for one year.
For the full research sample, CEO generated a large, but short-lived increase in UI-covered employment. By the end of the first year of follow-up, the program and control groups were equally likely to be employed. There were also small but statistically significant decreases in two key measures of recidivism — felony convictions and incarceration for new crimes — during Year 1, but no effects on other measures.
Most CEO participants enroll just after release from incarceration. However, a large fraction of the study sample came to CEO many months after release. Because the program was designed to serve recently released ex-prisoners, results were examined separately for those who came to CEO within three months after release (about 40 percent of the sample) and those who came later. These results are less certain because the sample sizes are relatively small, but they show a potentially important pattern: Among those who came to CEO within three months after release, program group members were significantly less likely to have their parole revoked, to be convicted of a felony, and to be reincarcerated. Effects on these measures are rarely seen in rigorous evaluations. In contrast, there were no significant impacts on any recidivism measures for those who came to CEO more than three months after release.
The CEO evaluation is still at an early stage. The study will eventually include three years of follow-up and will incorporate several additional data sources, including records of New York City jail stays and a survey of more than 500 sample members. A full report is planned for the first half of 2009.