Aim: To compare estimates of re-offending obtained from two sources, police data on persons of interest proceeded against and court finalisation data.
Method: Offenders who were convicted in a NSW Local or Higher Court in 2009 and received a non-custodial penalty were identified. For these offenders, rates of re-offending at 12 and 24 months were estimated using police and court data. Differences between the estimates obtained from these data sources, and the effect of offender and offence characteristics on these estimates, were investigated. In addition, the time taken for a re-offence to be proceeded against by police or finalised in court was examined and the impact of allowing shorter versus longer periods of time for re-offences to be captured in the data sources was explored.
Results: As at 30 June 2011, the proportion of offenders estimated to have re-offended within 12 months was 20.1 per cent using proven offences in court, 21.7 per cent using court finalisations (regardless of outcome), 21.9 per cent using police data where persons of interest were proceeded against to court, and 23.1 per cent where persons of interest were proceeded against more generally (i.e., to court, criminal infringement notice, cannabis caution). Most, but not all, offenders (94%) identified as having re-offended using court data were identified as having re-offended using police data. Estimates of re-offending by offender and offence characteristics were similar across data sources. The median time between a re-offence occurring and police commencing court proceedings against the person of interest was 1 day, while the median time between a re-offence occurring and being finalised in court as a proven re-offence was 86 days.
Conclusion: Data on persons of interest proceeded against by police may provide a more timely measure of re-offending than court finalisations data, potentially enabling program evaluations to be conducted 6 months earlier than current practice.