A Policy Evaluation of Home Detention Sentencing: Evidence from New Zealand
On 1 October 2007, New Zealand authorised home detention as a stand-alone sentence, permitting a judge to directly sentence offenders to home detention without requiring an initial imprisonment sentence. In this thesis, I estimate the causal effect of New Zealand’s home detention sentence on first-time offenders’ recidivism rates. Using population-wide linked administrative data from Statistics New Zealand’s Integrated Data Infrastructure, I identify the local average treatment effect for first-time offenders sentenced around the home detention policy reform of 1 October 2007. By focusing on offenders sentenced within a narrow time window of the policy implementation, I construct a treatment and comparison group such that differences in recidivism rates can be used to recover the causal effect of home detention. In contrast to a policy evaluation of home detention carried out by the Ministry of Justice, I find no evidence that home detention affects the recidivism rate of first-time offenders, relative to short-term imprisonment, community detention or intensive supervision sentences. This conclusion is robust across a range of fuzzy regression discontinuity design and instrumental variables specifications. Additionally, I analyse the effect of a home detention sentence on offenders’ attachment to the labour market. I run a difference-in-differences regression analysis of average monthly employment rates, earnings and benefit receipt, controlling non-parametrically for calendar time trends. The regression results provide no evidence that a home detention sentence increases offenders’ labour market attachment relative to offenders at the margin of home detention eligibility. Overall, results in this thesis provide little justification for promoting home detention as a means for reducing crime or improving offenders’ short-term or long-term labour market positions.