NCJ Number
147058
Date Published
1992
Length
997 pages
Annotation
This study assessed how sentences affected the subsequent recidivism of a New Jersey sample of convicted offenders.
Abstract
The assessment at sentencing of an offender's recidivism risk is a premise of the research. A chapter discusses how this relates to the various sentencing goals. Another chapter addresses several substantive and methodological issues pertinent to the use of actuarial or statistical models to measure and predict recidivism. A review of choices among predictor variables focuses on the logical and temporal nature of such variables. Two chapters detail the sample and data-processing procedures leading to the analyses and enumerate the recidivism indicators selected. A general model of recidivism based on offender characteristics and prior record is developed and tested using a single recidivism indicator (rearrest). Other recidivism measures are similarly modeled. Other chapters assess the impacts of various sentencing intervention, address substantive and methodological issues in the prediction of the probability of recidivism, and draw policy implications. Subject to several important caveats, the findings show effects for the sentences studied. Sentences to State prison are associated with decreased levels of recidivism, although the mechanisms that produce this decrease are unclear. Across most definitions of recidivism, the study found that those sentenced to a State prison were significantly less likely to recidivate. This finding is tempered, however, by the finding that over 62 percent of offenders receiving this sentence were eventually rearrests. The authors conclude that recidivism levels should have been higher given the characteristics of those sentenced to State prison. Given some recidivism definitions, jail sentences were also related to lower recidivism levels. Recidivism levels were still quite high, however, suggesting that a jail sentence only "slows down" offending. There was little evidence that the sentencing of young adult offenders to the Youth Complex at Yardville was effective. Throughout the study, the researchers found that offender and offense characteristics were more predictive of subsequent recidivism than sentence characteristics. Extensive tables and 213 references