Absence of conflict of interest.
Citation
Wiegand, A., & Sussell, J. (2016). Evaluation of the Re-Integration of Ex-Offenders (RExO) program: Final impact reports. Retrieved from the Department of Labor website: https://wdr.doleta.gov/research/FullText_Documents/ETAOP-2015-10_The-Evaluation-of-the-Re-Integration-of-Ex-Offenders-%28RExO%29-Program-Final-Impact-Report_Acc.pdf
Highlights
- The study’s objective was to examine the impact of Reintegration of Ex-Offenders (RExO) programs on employment, earnings, public benefits receipt, and recidivism outcomes.
- The study was a randomized controlled trial whose authors employed a statistical method to estimate the impact of RExO using data from four sources, including qualitative data generated from an implementation study, administrative data on criminal justice outcomes, administrative data on employment and earnings from the National Directory of New Hires, and a follow-up survey of participants.
- The study found that the RExO program did not increase the long-term employment, earnings, public benefit receipt, or recidivism.
- The quality of causal evidence on employment, earnings, public benefit receipt, and recidivism outcomes presented in this report is high because the evidence is based on a well-implemented randomized controlled trial. This means we are confident that any estimated effects would have been attributable to the RExO program and not to other factors had the study found statistically significant effects.
Intervention Examined
Reintegration of Ex-Offenders (RExO) programs
Features of the Intervention
The RExO program started in 2005 through a collaborative effort from the U.S. Department of Labor, Employment and Training Administration and the U.S. Department of Justice to assist urban communities with a large number of prisoners reentering the community. The RExO programs aim to assist such communities by providing ex-offenders three types of services typically for 12 weeks: (1) mentoring, (2) employment services, and (3) case management and other supportive services. A few programs also offered general educational services. Mentoring included one-on-one or group mentoring. Employment services included work readiness services, job training or placement, transitional employment, and follow-up after people were placed in jobs. Case management services included referrals for substance use disorder treatment, health, or mental health services, transitional housing, court assistance (including financial support for fees), and financial assistance for transportation or to support obtaining key milestones.
People in this study were placed in work readiness training at the start of the study. They were then matched to either an individual mentor or to group mentoring, meeting with a case manager biweekly or weekly to obtain employment services. People were eligible to participate in RExO programs if they are age 18 or older, were convicted as adults and incarcerated for at least 120 days, and enrolled in a RExO program within 180 days upon release from incarceration. People who were convicted of a sex-related offense, or whose most recent offense was a violent crime, were not eligible to participate in RExO programs. Programs were located at 24 organizations in urban U.S. settings.
Features of the Study
The study was a randomized controlled trial that took place in 24 sites across 18 states: Arizona, California, Colorado, Connecticut, Florida, Illinois, Iowa, Louisiana, Maryland, Massachusetts, Michigan, Missouri, New Jersey, Ohio, Oregon, Pennsylvania, Texas, Washington.
From a total of 4,655 people, 2,804 were assigned to the treatment group to use the RExO programs and 1,851 were assigned to be part of the control group. The majority of study members were ages 24 to 54: about a third were ages 24 to 34 or 35 to 44 years old, and about 20 percent were ages 45 to 54. Most study members were male (80 percent); 51 percent of study members were African American, and 33 percent were White. Most of the study members had some high school education (43 percent) or a high school diploma or GED (42 percent). After release, study members were mostly on probation (28 percent) or parole (50 percent).
The authors randomly assigned 60 percent of all people to the treatment group and 40 percent to the control group. There were three different processes to randomly assign people. The first random assignment process, which was used at one site, assigned people while they were still incarcerated. The second random assignment process, used at 15 sites, was conducted when people signed on to the study. The third random assignment process, used at 8 sites, took place after screening assessments to determine people’s eligibility.
The authors used a statistical model to compare differences between the treatment and control groups, accounting for race and ethnicity, gender, age, prior education, unemployment rate at the time of enrollment, and measures of prior criminal history. The authors also examined differences between subgroups of people considered by the authors to be at higher and lower risk of recidivism based on their number of past convictions. The study included data from four sources, including qualitative data generated from an implementation study, administrative data on criminal justice outcomes, administrative data on employment and earnings from the National Directory of New Hires, and a follow-up survey of participants. Administrative data on criminal justice outcomes included arrest and conviction data from criminal history repositories in 16 of 18 states where RExO sites were located, and state prison incarceration data from the department of corrections in 16 of 18 states. Administrative data on employment and earnings from the National Directory of New Hires included new hire data, quarterly wage data, and unemployment insurance data. Survey data included information on arrests and prison admissions.
Findings
Employment
- People who participated in the RExO program were no more likely to have been employed in the years following the program (through the fourth year after program enrollment) than the control group.
Earnings
- People who participated in the RExO program were no more likely to have higher earnings in the years following the program (through the fourth year after program enrollment) than the control group.
Public benefit receipt
- People who participated in the RExO program were no more likely to receive unemployment insurance benefits in the years following the program (through the fourth year after program enrollment) than the control group.
Recidivism
- Overall, based on measures from both the administrative and survey data, RExO program participants did not differ significantly from the control group in their likelihood of arrest, admission to prison, or the number of incarcerated days for the three years following program enrollment. In the third year after enrollment, administrative data showed that RExO program participants were more likely to be convicted of a crime than those who did not participate in the program. The authors suggest that this finding is spurious, however, because it was not supported by analyses of similar outcomes. Lastly, self-report survey data showed RExO program participants were less likely to be arrested in the third year after enrollment, but the authors note that analyses using administrative records data show that is likely because of recall bias or otherwise inaccurate reporting. Findings for subgroups at higher and lower risk of recidivism were mixed and did not show consistent patterns of impacts.
Considerations for Interpreting the Findings
The authors noted that, although the study was evaluating impacts of the RExO program, the services offered to people varied substantially across the grantees, including initial requirements for participation (such as meeting basic eligibility and simply expressing interest as opposed to needing to complete assessments and workshops before enrollment). In addition, program models delivered by grantees varied. Some grantees (about two-thirds) focused initially on stable employment, which involved delivery of work readiness training and job leads shortly after enrollment. Other grantees (about one-third) focused first on supportive services (including ensuring stable housing and ability to pass drug tests) before referring people to jobs. Some of these variations are discussed in the subgroup analysis, which is not covered extensively in this review.
Causal Evidence Rating
The quality of causal evidence on employment, earnings, public benefit receipt, and recidivism outcomes presented in this report is high because the evidence is based on a well-implemented randomized controlled trial. This means we are confident that any estimated effects would have been attributable to the RExO program and not to other factors had the study found statistically significant effects.
Additional Sources
Wiegand, A., Sussell, J., Valentine, E., & Henderson, B. (2015). Evaluation of the Re-Integration of Ex-Offenders (RExO) program: Two-year impact report. Oakland, CA: Social Policy Research Associates.