The California Department of Corrections and Rehabilitation (CDCR) today released the 2012 Outcome Evaluation Report, the third in a series of annual reports tracking the recidivism rates of adult offenders released from state prison. The report shows that recidivism rates have declined for the second straight year.
The focus of the 2012 report are the inmates released from CDCR during fiscal year 2007-08, a pivotal year when CDCR began using risk and needs assessments to better rehabilitate and supervise its offender population. These inmates had a 63.7 percent three-year recidivism rate, down from 65.1 percent the year before. They also committed 1,450 fewer crimes than those released the year before, despite being a larger group of inmates.
“We’re pleased to see that recidivism rates are improving and that the reforms we undertook in 2007 and 2008 are working,” CDCR Secretary Matthew Cate said.
In 2007 CDCR began using the Correctional Offender Management Profiling for Alternative Sanctions (COMPAS) and the next year began using the California Static Risk Assessment (CSRA) and the Parole Violation Decision Making Instrument (PVDMI).
COMPAS is a research-based risk and needs assessment tool used by CDCR in the placement, supervision and case management of offenders. It helps CDCR staff assign the right inmates to the right programs at the right time based on individual risk and needs assessments, reducing the likelihood of reoffending. All inmates are assessed using COMPAS before their release from prison. The report finds that, of those inmates with a substance abuse need as identified by COMPAS, those who receive in-prison substance abuse treatment and aftercare recidivate at less than half the rate of those who receive neither (30.7 percent compared to 62.7 percent, respectively).
CDCR partnered with the University of California, Irvine, to create a validated risk assessment tool to inform decision-making for parolees. The CSRA is a 22-item actuarial risk prediction instrument that predicts the likelihood to recidivate and moved CDCR from the use of an offense-based system. The report finds that the CSRA performs well at predicting the risk for recidivism.
The PVDMI, launched statewide in November 2008, assesses a parolee’s risk for recidivism as calculated by the CSRA and the severity of the parole violation, based on a severity index, to determine a consistent and appropriate response to the violation. Since the implementation of the PVDMI, fewer parolees have been returned to prison.
The report also contains a new section on the Prison University Project at San Quentin State Prison, a college program that began in 1996. Inmates who graduated from the program had a very low rate of recidivism after one year out of prison, as compared to a matched comparison group of similar inmates (5.4 percent compared to 21.2 percent, respectively).
In fiscal year 2007-08, 116,015 people were released or re-released from state prison. The in-depth 2012 report focuses on the 73,885 inmates who returned to prison within three years of release. It also looks at demographics, including gender, age, ethnicity, sentence, length of stay, mental health status, and other factors. CDCR measures recidivism by arrests, convictions and returns to prison and uses the latter measure – returns to prison – as its primary measure of recidivism. CDCR’s return-to-prison measure includes offenders released from prison after having served their sentence for a crime as well as offenders released from prison after having served their term for a parole violation.
Future reports will monitor how the implementation of realignment legislation impacts recidivism.
The 2012 Outcome Evaluation Report is published by CDCR’s Office of Research, which provides research data analysis and evaluation to implement evidence-based programs and practices, strengthen policy, inform management decisions and ensure accountability.
To view the entire report, please visit http://www.cdcr.ca.gov/Adult_Research_Branch/Research_Documents/ARB_FY_0708_Recidivism_Report_10.23.12.pdf