Aim: To compare reoffending rates between those who received an intensive correction order (ICO) and those who received short prison sentence (less than two years).
Method: Offenders’ demographic characteristics, index offence characteristics, prior convictions and penalties, LSI-R score and re-offences were extracted from the Re-offending Database maintained by the NSW Bureau of Crime Statistics and Research. Using logistic regression model with inverse probability of treatment weighting, the effect of penalty choice on re-offending was estimated. Doubly robust estimation and bivariate probit model with an instrumental variable were also used to address potential model misspecification and endogeneity of penalty assignment. As a sensitivity analysis, separate modelling was performed for offenders who were in medium to high risk categories and the prison group was restricted to those with 6 months or less fixed term.
Results: There was a 11%-31% reduction in the odds of re-offending for an offender who received an ICO compared with an offender who received a prison sentence of up to 24 months. The bivariate probit model with an instrumental variable did not reveal a significant effect or evidence of endogeneity.
Conclusion: These results further strengthen the evidence base suggesting that supervision combined with rehabilitation programs can have a significant impact on reoffending rates.