Planning to Evaluate a Reentry Program? What Are Some Challenges?

Experimental designs are quickly becoming the standard for evaluations of reentry initiatives.

The most effective outcome evaluations of reentry initiatives employ experimental designs that randomly assign prison releasees to receive reentry services. Control groups consist of prison releasees who receive regular parole supervision, or no supervision at all in the case of direct release from prison. While prison and parole administrators and judges may prefer to select the "neediest" offenders for reentry services, there have not yet been enough evaluation studies to determine how effective these services are and which offenders can benefit most from them. When random assignment is not possible, evaluators have tracked recidivism and other outcome measures for individuals who received reentry services and those who did not, and attempted to control for differences between the two groups using statistical techniques.

The voluntary nature of many reentry initiatives poses particular challenges for evaluators.

Implementing an experimental or quasi-experimental design can be difficult when the reentry initiative being examined is voluntary for participants. If a program is initiated but few inmates opt in, then the power of the evaluation design to discover statistically significant group differences when they are present is compromised. In addition, the initial equivalency of groups achieved by the random assignment may be compromised by differential dropout rates if reentry offenders find the conditions of the program too challenging. The voluntary nature of these initiatives may also limit the external validity of the evaluation design in that volunteers tend to have different characteristics than nonvolunteers, and these, rather than participation in the reentry initiative, may account for observed differences between them and offenders in the control group.

The difficulty programs have in successfully providing services requires evaluators to focus on implementation issues in process evaluations of reentry programs.

Reentry programs face a formidable array of obstacles to successful implementation, including lack of community resources and services, inadequate interagency coordination, and difficulties in keeping families involved in the reentry process. For this reason, it is critical that evaluators carefully assess the degree to which the program is implemented as originally planned; that is, how well actual activities match proposed activities. Reentry program outcomes should not be assessed until the evaluator is confident that services are being delivered as intended.

| | |
| | |