Appendix Two: Recidivism Index and Rehabilitation Quotient
The Recidivism Index (RI) is the measure Corrections uses to represent progress in reducing in re-offending.
At one level the RI is fairly straightforward: it is the percentage of offenders in any given cohort that is reconvicted within a given period of time (the follow-up period), and who receive either a prison sentence (RI reimprisonment) or any Corrections-administered sentence (RI reconviction). Corrections’ RI figures are produced on a monthly basis, using a 12 month follow up period and a rolling cohort of all offenders.
When RI rates are reported, they relate to offenders who were released from prison or started a community sentence over a 12 month period, ending 15 months ago. The difference in time between the offender release or sentence start period and the reporting of the re-offending, is to allow for the follow-up period to run its course, and to ensure that convictions for new offences have been recorded by the courts.
The Rehabilitation Quotient (RQ) measures the impact of the department’s rehabilitative programmes, through comparing the rates of reconviction and reimprisonment among ‘treated’ offenders (who completed a rehabilitative intervention) with the rates observed among ‘untreated’ offenders (offenders who are matched based on a range of risk-related factors, but had no involvement in that specific programme).
RQ scores are calculated separately for programmes delivered in prison and in the community. The cohort of prisoners are those who completed programmes in prison and were released in the 12 months ending 31 March 2014. We analyse their re-offending over the 12 months following their individual release dates.
The cohort of community offenders are those who completed a programme on a community sentence, where the programme end date occurred within the 12 months ending 31 March 2014. Corrections then measure their re-offending over the 12 months following completion of the programme.
The rates are represented by percentage-point changes in the rates of reimprisonment or reconviction of ‘treated’ offenders compared to the equivalent ‘untreated’ offenders. A reimprisonment score of -10.0 indicates that the rate of reimprisonment for ‘treated’ offenders was 10% lower than for the comparable ‘untreated’ offenders (for example, 12% compared to 22%). ‘Untreated’ includes those who did not receive any form of treatment, and those who received other forms of intervention but not the specific one being tested. The statistical method used in the analysis controls for the influence of these factors.
Many prisoners and community-based offenders participate in more than one programme. Where this occurs, the effects of participation in multiple programmes are not double-counted in each of the different programme RQs. The rates of some programmes reported are small and below the level of statistical significance; however, this does not necessarily mean that the particular programme has no impact on re-offending.