NCJ Number
224645
Date Published
July 2008
Length
11 pages
Annotation
The purpose of this study was to apply the Group Risk Assessment Model (GRAM) to measure reoffending among offenders released from prison in New South Wales (NSW).
Abstract
The strongest independent predictor found of reconviction was the number of times the offender had been convicted in the past. The odds of reconviction for an adult released prisoner who had 11 or more prior convictions were more than 6 times greater than a released prisoner with only 1 prior conviction. The analyses also revealed that several other characteristics made significant independent contributions to the predicted likelihood of reconviction. These were being younger at release, being indigenous, having been released to parole supervision, having served between 2 and 12 months in prison, and having had a prior conviction for breaching a justice order, a nonaggravated violent offense, or a theft offense. Offenders who had all of these risk factors were estimated to have a 94 percent chance of reconviction within 2 years of release. Validation procedures demonstrated that the model was able to adequately discriminate recidivist from nonrecidivist released prisoners. Internationally, governments are making renewed efforts to reduce rates of reoffending. The GRAM, developed by the United Kingdom, is a statistical technique designed to obtain more accurate estimates of trends in reoffending by adjusting for the characteristics of offenders coming through the justice system. This study describes a model built to adjust for the characteristics of prisoners released from the NSW adult prison system to enable accurate estimates of trends in reconviction over time for prisoners released. Tables, figure, references, notes and appended tables