Skip to main content

An experimental evaluation of a comprehensive employment-oriented prisoner re-entry program (Cook et al. 2015)

Absence of conflict of interest.

Citation

Cook, P. J., Kang, S., Braga, A. A., Ludwig, J., & O’Brien, M. E. (2015). An experimental evaluation of a comprehensive employment-oriented prisoner re-entry program. Journal of Quantitative Criminology, 31, 355-382.

Highlights

  • The study’s objective was to examine the impact of the Milwaukee Safe Street Prisoner Release Initiative on employment, earnings, public benefit receipt, and recidivism.
  • Drawing on administrative data from the Wisconsin Department of Corrections, the study used a randomized controlled trial to estimate program impacts. The authors used statistical models to compare the outcomes of treatment and comparison group members.
  • The study found that those in the treatment group had significantly higher earnings and rates of employment and significantly lower likelihood of new arrests than those in the control group. The study found no significant differences between the groups in the likelihood of reimprisonment in the one year of follow-up.
  • The quality of causal evidence presented in this report for the recidivism outcomes is moderate because it was based on a randomized controlled trial in which the authors did not account for the different probabilities of assignment to the treatment group; nevertheless, this study can be considered to have employed a well-implemented nonexperimental design. This means we are somewhat confident that the estimated effects are attributable to Milwaukee Safe Street Prisoner Release Initiative and not to other factors. The quality of causal evidence presented in this report for the employment, earnings, and public benefit receipt outcomes is low because the authors did not ensure that the groups being compared were similar before the intervention. This means we are not confident that the estimated effects are attributable to the Milwaukee Safe Street Prisoner Release Initiative; other factors are likely to have contributed.

Intervention Examined

The Milwaukee Safe Street Prisoner Release Initiative

Features of the Intervention

The Milwaukee Safe Street Prisoner Release Initiative was a comprehensive prisoner re-entry program focusing on employment. Services began while participants were incarcerated and continued after they had been released. Six months before their scheduled release, participants were moved to facilities in Racine, Wisconsin, and completed an assessment with a social worker to identify their needs, skills, and risks. They were offered opportunities to complete soft skills and vocational training as well as restorative justice circles. The program included the 12- to 16-week Breaking Barriers curriculum, which intends to help people identify and change behavior, thinking, and attitudes related to criminality. Finally, participants were provided opportunities to use reach-in services from the Community Corrections Employment Program, alcohol and drug treatment, remedial education, and work release (if they were minimum-security offenders).

The program staff team met with participants monthly after their release, and participants could access the Community Corrections Employment Program, which provided job search assistance and gave job creation subsidies to employers. In addition, participants could receive substance-abuse treatment and other wraparound services.

Features of the Study

The study used a randomized controlled trial to estimate program impacts. Over a seven-month period, eligible people were randomly assigned to either the treatment group, which was offered the opportunity to participate in the Milwaukee Safe Street Prisoner Release Initiative with guaranteed access to all services, or the control group, which could access the same services as the treatment group but had to join a waitlist. There were 106 men randomly assigned to the treatment group and 130 to the control group. The authors do not report the sample size by condition for the analysis.

Both groups were predominantly African American (84 to 86 percent), with 8 to 10 percent of the sample identifying as Hispanic. Sample members were 28 years old on average. Over one-third of the sample had less than a high school degree and on average read at a 9th-grade level. The sample had between 2.2 and 2.7 prior felony convictions on average and 1.1 to 1.2 prior misdemeanor convictions on average.

The study used administrative data from the Wisconsin Department of Corrections from January 2009 to December 2011. People were followed for one year after their release for the employment, earnings, and recidivism data. To measure earnings and employment, the study used parole officers’ recorded updates, collected weekly for treatment group members and every one to two weeks for control group members, for one year following release. The authors used statistical models to compare the outcomes of treatment and comparison group members. The analysis accounted for race, age, education, reading level, release location, juvenile record, risk level, and length of incarceration.

Findings

Employment

  • The study found that treatment group members were 20 percentage points more likely to be employed during the four follow-up quarters compared with the control group. This is a significant difference.

Earnings

  • The study found that the treatment group earned $408 more across the four follow-up quarters than the control group. This is a significant difference.

Public benefit receipt

  • The study found that more treatment group members received Supplemental Security Income benefits than the control group (7 percent compared with 2 percent). The authors did not assess the significance of this finding.

Recidivism

  • The study found that treatment group members were 39 percent less likely to be arrested during the four follow-up quarters compared with the control group. This is a significant difference. The study found no significant differences between the groups in the likelihood of reimprisonment during the follow-up period.

Considerations for Interpreting the Findings

The study was a randomized controlled trial, but the authors did not clarify the number of people in each group for whom there was outcome data. In addition, initially eligible participants had a 33 percent probability of being assigned to the treatment group. The authors realized, however, that they would not have as many participants as planned, so they decided to randomly select 16 people previously assigned to the control group and move them to the treatment group. Although this change does not compromise the validity of the random assignment process, it does mean that different members of the sample had different probabilities of assignment to the treatment group, and the authors did not account for this in their analysis.

For the public benefit outcome, the authors did not account for other factors that could have affected the difference between the treatment and comparison groups, so this outcome receives a low rating. For the employment and earnings outcomes, the authors did not account for existing differences between the groups before program participation. These existing differences between the groups—and not the program—could explain the observed differences in outcomes, so these outcomes also receive a low rating.

In total, 9 percent of the sample did not receive any of the services, as they were not transferred to the Racine facilities before their release; these people are still considered part of the treatment group in the analysis. The control group could access the services offered to the treatment group, but they had to join a wait list (the treatment group was allowed guaranteed access to the services). The control group accessed some services before their release (3 percent completed their high school education, 25 percent received counseling, and 13 percent completed an alcohol and other drug abuse program). Six percent participated in a work release program.

Causal Evidence Rating

The quality of causal evidence presented in this report for the recidivism outcomes is moderate because it was based on a randomized controlled trial in which the authors did not account for the different probabilities of assignment to the treatment group; nevertheless, this study can be considered to have employed a well-implemented nonexperimental design. This means we are somewhat confident that the estimated effects are attributable to Milwaukee Safe Street Prisoner Release Initiative and not to other factors. The quality of causal evidence presented in this report for the employment, earnings, and public benefit receipt outcomes is low because the authors did not ensure that the groups being compared were similar before the intervention. This means we are not confident that the estimated effects are attributable to the Milwaukee Safe Street Prisoner Release Initiative; other factors are likely to have contributed.

Reviewed by CLEAR

January 2020

Topic Area