News Release

College-in-prison program found to reduce recidivism significantly

Peer-Reviewed Publication

Crime and Justice Research Alliance

A new study sought to determine the effects of a college-in-prison program, the Bard Prison Initiative (BPI). The study found a large and significant reduction in recidivism rates across racial groups among those who participated in the program. It also found that participants with higher levels of participation had even lower rates of recidivism. In light of their findings, the authors offer several policy recommendations in support of college-in-prison programs.

The study, by researchers at Yale University and BPI, appears in Justice Quarterly, a publication of the Academy of Criminal Justice Sciences.

            “Incarceration is bound with systems of poverty and a lack of access to opportunity, especially education and socioeconomic mobility,” notes Matthew G.T. Denney, a PhD student at Yale University, who coauthored the study. “Participation and intensity of engagement in programs like BPI might disrupt these cycles.”

Beginning in 1965, people incarcerated in the United States were eligible to receive Pell Grants to fund college courses in prison, and federal funds supported higher education programs in prison for decades. But in 1994, the Violent Crime and Law Enforcement Act banned incarcerated students from receiving Pell Grants, effectively shutting down many college-in-prison programs. It was not until 2016, when the Obama Administration started the Second Chance Pell Pilot Program, that federal aid was offered to select colleges with prison programs. While studies have examined the effectiveness of such programs, few have addressed selection and self-selection bias.

This study, which the authors call the largest and most rigorous to date on the effects of a college-in-prison program on recidivism, examined BPI. The program has offered college courses to incarcerated students since 1999, operating campuses in six New York correctional facilities. The study leveraged BPI’s multistage admissions process to compare BPI participants to a similar group of people who did not participate. Researchers worked with the administration of BPI to collect data on BPI applicants and with the New York State Department of Corrections and Community Supervision to collect administrative data. The authors then merged the two sets of data, accounting for self-selection bias by limiting the group of people studied to those who applied to participate in BPI.

            The study found that participation in BPI reduced recidivism 38 percent, and greater levels of participation correlated closely with even lower rates of recidivism. Rates fell across racial groups, even as BPI enrolled a student body broadly reflective of the prison population, which is disproportionately Black and Latinx.

            BPI is a rigorous college program, and the authors caution that college-in-prison programs that lack rigor may not have the same effect on rates of recidivism.

“By reducing recidivism across racial groups, BPI enables greater educational opportunities and addresses racial inequality in prison populations at the same time,” explains Robert Tynes, associate director of research at BPI, who coauthored the study. “In light of these findings, policymakers should promote the flourishing of these programs and encourage more opportunities for incarcerated students.”

            Specifically, the authors recommend that:

  • State and local governments fund college-in-prison programs and expand programs within their jurisdictions,
  • Corrections departments increase support for individual students and college-in-prison institutions, and
  • States and colleges promote rigorous, high-quality college-in-prison programs, including colleges maintaining the same self-assessment standards at prisons that keep them accredited at their home campuses.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.