NCJ Number
249844
Date Published
December 2015
Length
72 pages
Annotation
Using data from the Florida Department of Corrections, this study examined the effects on recidivism and employment for released offenders under "split supervision" (community supervision post-incarceration included in the sentence) and, "conditional release" (parole) supervision compared to release from prison with no form of community-based supervision.
Abstract
Overall, the findings indicate that post-prison supervision is a significant predictor of reduced recidivism outcomes (rearrest and conviction) among a sample of Florida inmates (n=201,447) released from prison between January 2004 and December 2011. Post-prison supervision also increased the odds of obtaining employment compared to those released with no community-based supervision, but only for those on split supervision. Data were collected up to 3 years after release. These findings are mitigated, however, by the significant increase in the likelihood of supervised offenders returning to prison for a felony conviction compared to unsupervised inmates. Taken as a whole, effects are generally positive for post-release supervision in reducing recidivism and increasing employment, particularly for offenders under split probation or split community-control supervision. By identifying and monitoring offenders at greater risk for reoffending and subjecting them to a swift and certain punishment (return to prison) for technical violations, ex-inmates are less likely to return to serious criminal behaviors. Future research should focus on identifying the differential effects of post-prison supervision on arrest for any crime compared to arrest for a felony offense. Policymakers are cautioned against expanding the use of split supervision until further research has examined the full scope of positive and negative effects of split supervision. Survival analysis, logistic regression, precision matching, and propensity score matching methods were used to examine multiple recidivism outcomes at one, two, and three years after release and to predict the likelihood of obtaining employment after release. 17 tables and 65 references