Absence of conflict of interest.
Citation
Tyler, J. H., & Kling, J. R. (2007). Prison-based education and reentry into the mainstream labor market. In D. Weiman & S. D. Bushway (Eds.), Barriers to reentry?: The labor market for released prisoners in post-industrial America (pp. 227-256). New York: Russell Sage Foundation.
Highlights
- The study’s objective was to examine the impact of obtaining a general education diploma (GED) while in prison from 1994 to 2000 in Florida on male prisoners’ post-incarceration employment, earnings, and recidivism.
- The authors used a nonexperimental design to compare outcomes for those who obtained a prison-based GED (the treatment group) with those that did not (the comparison group). Drawing on administrative data from the state of Florida and Unemployment Insurance (UI) records, the authors estimated impacts on earnings and compared the means of the two groups for the employment and recidivism outcomes.
- The study reported no statistically significant findings on earnings in the first and third years after release or on recidivism or employment outcomes. The study found that, in the second year after release, people who obtained a GED in prison earned $114 more per quarter than those who did not; this was a significant difference.
- The quality of causal evidence presented in this report is low because the authors did not ensure that the groups being compared were similar before the program and did not eliminate concerns about self-selection into the prison-based GED program. This means we are not confident that the estimated effects are attributable to the prison-based GED program; other factors are likely to have contributed.
Intervention Examined
Prison-based program to obtain general education diploma (GED)
Features of the Intervention
The intervention was the receipt of a GED through a Florida prison-based GED program for prisoners who entered without a high school diploma. The sample included 12,956 men who were newly incarcerated in a Florida state prison between 1994 and 2000 and who lacked a high school diploma when entering prison. During this period, 1,967 people obtained a GED in prison (the treatment group) and the remaining 10,989 people did not (the comparison group). Overall, the sample was 47 percent White, 44 percent African American, and 8 percent Hispanic. The average age was 28, and the average length of education was 9.5 years. In all, 31 percent of the sample were employed before their incarceration.
The authors examined the impact of receiving a prison-based GED on post-release earnings, employment, and recidivism outcomes using administrative data provided by the state of Florida and UI data. To estimate impacts on earnings, the authors used a statistical model to control for individual factors and background characteristics. For the employment and recidivism outcomes, the authors presented the mean outcomes for both groups without any statistical controls. The authors assessed the recidivism and earnings outcomes annually for three years after release, and the employment outcome for one year post-release.
Findings
Employment
- The study did not find a statistically significant relationship between obtaining a GED in prison and employment one year after release.
Earnings
- The study found that, in the second year after release, those who had obtained a GED in prison earned $114 more per quarter than those who had not obtained a GED. This was a statistically significant difference. The study did not find any statistically significant differences in earnings between the two groups in the first or third year after release.
Recidivism
- The study found no statistically significant relationships between obtaining a GED in prison and recidivism, defined as a conviction resulting in a return to prison or probation, up to three years after release.
Considerations for Interpreting the Findings
In this study, participants elected to receive the intervention (obtaining a GED) rather than being assigned to it, for example, by random assignment. In studies in which participants elect to receive an intervention, there is always a risk that those who do differ from those who don’t in important ways that influence study outcomes. For example, a prisoner that opts to obtain a GED while in prison could be more motivated than other prisoners, which might affect their later employment outcomes. In this study, the authors control for these possible mechanisms to a large extent in the earnings analysis using a statistical model, but there might still be bias from members electing into the intervention group. Moreover, the authors do not control for such mechanisms in the analyses of employment and recidivism outcomes.
In addition, the study noted large dissimilarities between the two groups on race and age. The treatment group had a significantly higher proportion of White people than the comparison group (60 percent and 45 percent, respectively) and a lower proportion of African Americans (32 percent and 46 percent, respectively). The treatment group was also significantly younger, with an average age of 25.6 years old compared with 28.8 years old for the comparison group. Although the authors accounted for race and age in their statistical models for the earnings outcomes, it is uncertain whether the analysis could fully account for existing differences this large between the groups. These large existing differences between the groups—and not the prison-based GED program—could explain the observed differences in outcomes.
Causal Evidence Rating
The quality of causal evidence presented in this report is low because the authors did not ensure that the groups being compared were similar before the program and also did not eliminate concerns about self-selection into the prison-based GED program. This means we are not confident that the estimated effects are attributable to the prison-based GED program; other factors are likely to have contributed.