Three State Recidivism Study

September 2001:

Submitted to the Office of Correctional Education United State Department of Education
EXECUTIVE SUMMARY
The Correctional Education Association conducted the Three State Recidivism Study for the United States Department of Education Office of Correctional Education. The study was designed to see if education, independent of other programs, could have significant impact on the behavior of inmates after release. Data on about 3,200 inmates, who were released from Maryland, Minnesota and Ohio prisons in late 1997 and early 1998, are reported in this longitudinal study. The research design, which uses educational participation while incarcerated as the major variable, measures the impact of education while incarcerated on post release behavior, primarily recidivism and employment. The states pooled their data in a format that allows for individual state as well as aggregate reports. Within each state the correctional, parole and probation, education and work force agencies cooperated in the data collection.
When the study began each state determined 1,000 or more people about to be released from their state prisons. This type of research design, called a release cohort, provides for a framework with internal control groups.

The research tools were designed by a number of correctional researchers. The Pre-Release Survey was developed to gather information from the inmates themselves on factors closely correlated with criminal behavior. These included information on socio-economic factors, criminal behavior, family life, educational experiences and work history. The Educational/Institutional Data Collection Form contained questions that included criminal behavior, demographic information, institutional behavior and education history during incarceration. The Parole/Release Officer Survey contained questions designed to collect data on subsequent criminal behavior, employment and educational experiences. Post Release Criminal History Data was collected to measure recidivism. Employment data was collected to find out about the kinds of jobs and the amount of wages earned after release.

Inmates, about to be released, were assembled in the various institutions where the Pre-Release Survey was administered by trained staff. Those who did not want to take the survey were excused. The refusal rate was very low. Inmates put their answers on scan sheets. Criminal history and educational data from the education and correction records were collected from the various agencies using the Educational/Institutional Data Collection Form in either an electronic or a scan sheet format. After release the parole officers were sent the Parole/Release Officer Survey for behavioral, educational and employment information for the ex-offenders under their supervision. In two states, the state departments of labor were able to contribute Employment Data for post release employment and wage information. Finally, re-arrest, re-conviction and re-incarceration data were collected from state Criminal History Data files. The federal Department of Justice Bureau of Justice Statistics commonly uses these three areas to describe recidivism.

All the information obtained from the surveys and state databases were entered into one large database to allow for individual state and aggregate reports. Over 500 variables were collected on each of the study participants, resulting in a very large quantity of information. There are demographic data on family and community background, economic status and employment, educational experience, offender perspectives on education and motivational factors that correlate highly with criminality as well as educational, correctional and criminal history. This is the first study to collect extensive information from the inmates themselves.

Another unique aspect of the study is the collection of individual wage data from two of the states labor databases. Data was collected from each state, but only Maryland and Minnesota was useable. The Ohio data was in a format that could not be read by the researchers.

The analysis of the data indicates that inmates who participated in education programs while incarcerated showed lower rates of recidivism after three years. For each state the three measures of recidivism, re-arrest, re-conviction and re-incarceration were significantly lower. The employment data shows that in every year, for the three years that the study participants were followed, the wages reported to the state labor departments were higher for the education participants compared to the non-participants.

The following report narrates the overall design, execution of the study, data analysis, discussion of the results and conclusions. ..more.. by Dr. Stephen J. Steurer -and- Dr. Linda Smith -and- Dr. Alice Tracy, Correctional Education Association