Introduction

146. The majority of crimes are committed by people who have offended before, usually many times. Serious criminal offending is not something that many people in society engage in very often, it is something that a small group of people engage in sporadically, and some repeatedly. Rehabilitation of offenders currently in the justice system has the potential to bring about reductions in offending after sentence completion, and thereby to achieve reductions in victimisation, and, if it is well targeted, reductions in imprisonment.

147. Figure 11 below illustrates 'survival without re-conviction' for offenders released from prison in 1993. As the curves show, the rate of re-offending is strongly influenced by risk score, as measured by the Department of Corrections' risk assessment procedure. Offenders with the lowest 20 percent of risk scores re-offend more slowly, reaching 40 percent re-offending by four years after release. High-risk former inmates re-offend rapidly and reach 25 percent survival, or 75 percent re-conviction after one year, and 10 percent survival, or 90 percent re-conviction, after four years. Overall, all released inmates had a re-conviction rate of around 73 percent after four years.

Figure 11: Survival without re-conviction after release from prison

Source: Wanganui Computer records for 1993

What works to prevent chronic repeat offending?

148. 'What works?' as a way of preventing established offenders from continuing to offend has been one ofthe central questions in offender management for several decades. Until recently, the commonly held view was that 'nothing works' - established adult offenders continue offending until they 'burn out" sometime after the age of 40 years, and attempts to rehabilitate before then are pointless. More recently, however, reliable experimental evidence has accumulated suggesting that under certain circumstances rehabilitation can reduce the frequency and/or the seriousness of future offending, and occasionally shut it down entirely - 'some rehabilitation programmes are successful with some offenders in some settings when applied by some staff.⁴⁵ The 'certain circumstances' that must be provided to reduce the incidence of future imprisonable offences are summed up in the 'risk-needs-responsivity model'.

149. Rehabilitation rates for established adult offenders remain low, however - typically 10 to a maximum of 30 percent. For inmates as a whole (Figure 11 above) rehabilitation has the potential to reduce the expected 73 percent reconviction at the end of four years, to 50 to 60 percent reconviction - perhaps a one-third reduction in offending and also in victimisation.

The risk-needs-responsivity model

150. The mainstream approach to the rehabilitation of es'rablished adult offenders was set out in three key research reports published in 1990 and 1994.⁴⁶ Together these studies established the 'risk-needs-responsivity' model, which suggests that effective rehabilitative services must be matched to each individual offender's risk level, needs profile, and responsivity profile. Rehabilitation must be individualised and carefully targeted.

151. 'Risk' is a measure of an offender's expected frequency and severity of re-offending, as measured by a research-proven actuarial procedure. Under the risk-needs-responsivity model, resources for rehabilitation are allocated in proportion to risk - high-risk cases receive more intensive treatment, and resources may be withheld from low-risk offenders, so that high-risk offenders can receive the necessary intensity of services. Risk assessment is now being introduced nationally as an aid to the management of offenders referred to the Department of Corrections.

Figure 12: Relationship between offender's risk score and the average cost of future offending in the five years following conviction

Source: Wanganui Computer records for 1993