This report describes changes in the reoffending outcomes observed for 1,272 cases (involving 973 distinct young people) that resulted in one of three types of supervision orders in the Youth Court: ‘stand-alone’ Supervision (SUP), Supervision with Activity (SwA) or Supervision with Residence (SwR). The supervision-type orders examined were imposed between 1 October 2010 and 31 March 2013 – this being the first 30 months under the Fresh Start reforms.  Reoffending was examined for the 12 month period after the orders were served.

Caution must be taken in interpreting reoffending outcomes as they are not always a measure of the effectiveness of an intervention alone. For example, some people may reoffend less often simply due to the fact that they were caught by the Police and made to account for their actions, regardless of the particular intervention applied. Measuring the impact on reoffending of an intervention would require a robust statistical approach such as a matched comparison analysis. This was not in scope for this analysis.

Key findings

Offending patterns across all three orders

• There was an overall reduction in both the frequency and seriousness of offending in the 12 months following all three types of supervision orders compared to the 12 months prior to the orders. 
• These reductions were proportionally larger for SUP and SwA orders than for SwR orders. However, the young people receiving SwR orders had more extensive offending histories than was the case for the other two types of orders.
• Theft-related and burglary offences were the most common offences committed by young people in the 12 months before and after all three orders. These offence types also showed the greatest numerical decreases in the 12 months after the orders.
• Violent offending halved in the follow-up period for all three types of supervision orders.
• Of the young people with SUP orders, 22% did not reoffend within 12 months. The figures for SwA and SwR were 21% and 13% respectively. Seven to eight out of 10 of the young people reduced the frequency or seriousness of their offending in the 12 months after all three types of orders. The figures were lowest for SwR orders.
• European young people had slightly better offending-related outcomes in the 12 months after SUP orders than Māori young people. However, the reverse was the case after SwR orders, with Māori young people generally having slightly more positive outcomes than European young people. The relationship between ethnicity and outcomes was less clear after SwA orders.
• SwA orders had a higher rate of post-order custodial sentences  (35%) than the two other types of orders (25% for SUP and 30% for SwR). The higher figure for SwA appears to be due to a greater likelihood of having a subsequent SwR order imposed – often as a result of breaching the SwA order.
• SUP orders had the highest proportion (62%) of people who did not receive any further supervision-type orders or prison sentences in the 12 months after the order. This proportion was higher than SwA orders (56%) and SwR orders (59%).

Shorter versus longer orders

• The Fresh Start reforms doubled the maximum length of SwA and SwR orders from three to six months. In the first 30 months under Fresh Start:
  • 75% of SwA orders were imposed for longer than three months
  • 67% of SwR orders were imposed for longer than three months.
• Young people who received shorter SwA orders (of three months or less) were more likely to reoffend in the 12 months after the order compared with recipients of longer orders in excess of three months (90% and 76% respectively). Despite this difference, the proportions of young people who reduced the frequency or seriousness of offending did not differ much with the length of the SwA order.
• The findings in the point above suggest that young people with longer SwA orders were more likely to stop offending over the next 12 months than was the case for those with shorter SwA orders – who more often reduced their level of offending, but did not stop completely.
• Forty percent of the young people with shorter SwA orders received a custodial sentence within 12 months, with the figure being lower at 34% for longer orders.
• There was no difference between shorter (exactly three months) and longer SwR orders in the likelihood of not reoffending or reducing the frequency of offending. Those with longer SwR orders (80%) were more likely to reduce the seriousness of their offending than those with shorter orders (72%).
• There was a small difference in the proportion of young people who received a custodial sentence within 12 months of release for shorter and longer SwR orders (29% and 31% respectively).

Further research
There is interest in the extent to which doubling the maximum length of SwA and SwR orders under Fresh Start has led to improved outcomes for recipients of these orders. We observed, for example, that a much lower proportion of those with longer SwA orders reoffended compared to those with shorter SwA orders. However, such a comparison needs to account for any differences in the types of young people getting different length orders. This analysis could usefully be undertaken in the future.

There is also interest in the relative effectiveness of the three types of supervision orders, and in particular whether there are any subgroups of young people for who particular interventions appear more effective than others.

Some of the supervision-type order recipients spent time out of the community on custodial remand or subject to custodial orders which may have had some effect on their rates of offending. The feasibility of calculating an adjusted reoffending rate that reflects actual time spent in the community could usefully be investigated in the future.