Analysis of Possible Underlying Factors

Although there is not a substantial variation in outcome rates across jurisdictions, NZ and the NT do show slightly higher completion rates for total orders (over a 10 year period for the NT and the six-years that the program has been operating in NZ) as well as highest or equal highest rates for the most current year (2005-05). The focus of the following analysis is on selected features that may, given the findings of the literature review and outcomes of discussions with program managers, underlie performance variation and that are shared by these two jurisdictions but which are unique to the two when compared with other jurisdictions with lower completion rates.

Type of program

The type of HD program may be a significant factor in completion rates. For example, bail supervision in SA has a much lower completion rate than either frontend or back-end detention in that jurisdiction (57%, 100% and 90% respectively in 2004-05 and similar variation in previous years). NZ front-end and NZ back-end HD completion rates differ, but not consistently on a year-by-year basis. Type of program alone does not distinguish between jurisdictions with higher and those with lower completion rates. In 2004-05, NZ’s back-end HD program completion rate was 86%, lower than the equivalent SA rate of 90%.

Program intensity

Differences in the extent of restrictions on behaviour as a condition of the order combined with variation in intensity of supervision may affect completion rates. For example, an analysis of NSW breach rates reported in verbal evidence to the Standing Committee on Law and Justice enquiry shows 46% of detainees over an 18- month period in 2001-02 breached at least one condition, with the most common being urinalysis testing, although a breach did not automatically revoke the order (59% were officially sanctioned by a warning or a change to case management or supervision practice but did not have their order revoked).

In NZ, restrictions on the use of alcohol or drugs and a requirement to submit to testing are not standard conditions of the order, so are not universally applied to all detainees as is the case in other jurisdictions, instead, being applied at the discretion of the decision-making body according to individual detainee circumstances. However, in practice, a substantial proportion of detainees have this as a condition of their order, and, given that the other jurisdiction with the highest completion rates (NT) does apply this condition universally and shares similar program completion rates to NZ, this does not of itself explain variation in program outcome.

Intensity of supervision could be a contributing factor to program completion rates. All jurisdictions other than the two showing the highest rates (NT and NZ) document a minimum contact frequency by caseworkers of 20 or more contacts in the most intensive supervision regime stage. In comparison, NT operates on an initial weekly contact frequency for urban and monthly for rural areas and the highest minimum contact frequency for any phase and management regime for NZ is 12. However, in NZ all detainees are also subject to continuous active electronic monitoring while at home and work and alternative monitoring while on approved absences by the contracted monitoring company.

Given that the other jurisdictions with similarly high minimum contact frequencies during the most intensive stage of the monitoring regime still show substantial variation in completion rates, intensity of supervision does not in and of itself appear to be a predictor of program completion when taking both frequency and nature of monitoring into account.

Offender populations

There is some empirical research evidence in some reports (although the findings are not consistent across studies) for higher program completion and lower recidivism rates to be related to differences in detainee age, ethnic background, prior imprisonment history, employment status, motivation, current drug abuse history (but not alcohol abuse), offence type, and living arrangements (eg, better outcomes if residing with spouses or same-sex roommates and assessed as having a stable home environment). There is also evidence from research on other offender and prisoner populations showing lower recidivism rates among different types of offenders, eg, violent offenders reported as showing lower rates of reoffending than property offenders in various analyses and those with drug offences showing higher rates.

Jurisdictions vary in whether HD is available to individuals with a history of violent offences under the particular jurisdiction’s legislation. In NSW, Victoria, and the ACT most violent offences preclude eligibility, SA excludes only one (homicide), while Queensland, NT and NZ do not have explicit offence exclusions. In NZ, 27% of detainees on orders in November 2003 (the date of the most recent census of prisoners and detainees) had a violent or sexual offence as the major offence. In the NT, 13% of offenders commencing HD orders in 2004-05 had a violent, sexual or robbery conviction as the most serious offence. Information is not available on the proportion of violent offenders in Queensland. NZ and NT, despite sharing similarly high program completion rates, vary markedly in the proportion of detainees with a drug offence as the most serious/major offence (23 and 2% respectively). Comparable information is not available for other jurisdictions. Type of offence per se is unlikely to explain differences in program completion rates.

Females have been reported as showing lower recidivism rates than males in the research and practice literature, which may be a contributory factor to higher program completion rates where revocation is on the grounds of reoffending. NZ shows a much higher proportion of female detainees than all other jurisdictions (with the exception of Victoria and ACT in some years where the proportions will be skewed by small number effects, eg, the 50% rate for ACT in 2002-03 represents only a single female detainee). However, NT’s proportion of female detainees is lower than that of jurisdictions showing lower program completion rates, and the jurisdiction with the second highest proportion of females shows the lowest completion rate.

Indigenous status has also been linked to higher recidivism rates, which may contribute to lower program completion rates where revocation is on the grounds of reoffending. NT and NZ show the highest percentage of Indigenous detainees and the table below shows these two jurisdictions have higher 2004-05 rates of Indigenous home detainees per 100,000 Indigenous people in the general population, both compared to other jurisdictions and relative to rates for community corrections offenders and prisoners (noting that these detainee rates can still be affected by small number effects even among the jurisdictions with larger detainee numbers below). While there are jurisdictional differences in the percentage and rate of Indigenous home detainees between jurisdictions with higher and lower program completion rates, this is not in the expected direction indicated by the research and practice literature.

Age has also been linked to recidivism, with older age groups showing a lower risk of reoffending. The 2003 NZ Prisoner and Detainee Census notes that the home detention program is weighted toward an older age group relative to sentenced prisoners. One-quarter or more detainees in NZ and NT are aged 40 or over. However, no comparable information was available to the study for other jurisdictions.

These analyses of selected detainee population characteristics, although limited in scope, do not provide any strong evidence for differences in population characteristics such as gender, Indigenous status, or most serious offence being predictors of program outcome. However, analysis of data not available to the study, eg, other population characteristics such as prior correctional history, or a combination of factors could underlie variation in program outcome between jurisdictions.

The most appropriate comparison (supported by findings from the research and literature review) would be a comparison on risk assessment scores. However, this is outside the scope of the current study, and is also limited by jurisdictional differences in the type of assessment instrument used. It may be the case that jurisdictions with higher program completion rates also have detainee populations that comprise lower risk offenders. Both NT and NZ have the highest proportions of detainees to total corrective services populations, which could, if assuming that criminal justice system practices are reasonably similar across jurisdictions in producing comparable client populations, be interpreted as reflecting greater application of HD to lower-risk offenders compared to those jurisdictions with a relatively low proportion. However, these differences could equally reflect a greater willingness of the decision-making body to grant orders to a wider range of offenders, including higher-risk groups.

Caseload and resource factors

There is no obvious correlation between either caseload or unit cost and program completion rates based on the information available to the study. The two jurisdictions with the highest completion rates (NT and NZ) have markedly different detainee to operational staff ratios (attributable at least in part to the use of a contracted company to monitor detainees in one jurisdiction and the scope of non-metropolitan geographic coverage required in the other). Unit cost between the two jurisdictions with the highest and the lowest completion rate was almost identical in 2004-05.

Other program features

No single program feature was identified that is shared by NT and NZ but is unique to only those two jurisdictions.

Generic practices

The performance variation in home detention completion rates is not attributable to generic policy or practice features common to the management of other offenders in the community. For example, the completion rate of supervision orders ¹ over the last three years consistently shows NSW with the highest completion rate of those jurisdictions on which the home detention outcome

analysis above is based, while NT and NZ have the lowest and equal second lowest rates, the direct opposite to the pattern for home detention program completion rates. The same ranking is found for completion of parole orders over the past two years, with NT and NZ show lower completion rates and NSW the highest.

Overview

Overall, on the basis of the information available to the study, there are no obvious factors explaining performance variation on the jurisdictionally comparable measure of program completion when considering either individual program features or detainee characteristics. However, it is possible that combinations of program features may contribute to variation in program completion rates, especially if combined with differences in the level of risk across the detainee population.

For example, jurisdictions vary in whether there is a requirement for all detainees not to use alcohol or drugs throughout the order as a legislatively-prescribed standard order condition (as is the case in, eg, NSW and Victoria), in the minimum frequency of case worker contact (see earlier sections), and in testing regimes (eg, Victoria applies random urine/breath testing for all detainees on a weekly basis in the first phase reducing to monthly in the least intensive phase of the order while NZ does not conduct urinalysis). Jurisdictions with higher proportions of detainees with a higher risk of reoffending such as drug users that also apply more intensive monitoring and substance testing regimes may show lower completion rates because breaches of conditions relating to alcohol/drug use are more likely to be detected.

However, information is not available to the study to assess whether such multiple factors of program features combined with detainee characteristics can predict program outcome.

* SA excludes bail supervision detainees – including this group increases the rate to 14,587

1 Supervision orders comprised offenders other than those on restricted movement orders (ie home detention, curfew orders) or reparation orders (ie offenders with a community service bond or order requiring that they undertake paid work and fine options).