11 out of 12 rehabilitation programmes in prison not working – at a cost of $159 million a year

Eleven of the twelve prison rehabilitation programmes currently offered by Corrections have almost no impact on reoffending rates according to the Department’s own figures.  As a result, the Department has little hope of achieving its stated goal to reduce reoffending by 25% by the year 2017.

Eighty per cent of prisoners in New Zealand commit their offences under the influence of alcohol and drugs. In response, successive governments have slowly increased the availability of addiction treatment in prison by setting up drug treatment units (DTUs). In 2009, there were six  such units.

Evaluation process

That year, the Department evaluated the effectiveness of the DTU programme and claimed they reduced an inmate’s risk of reoffending by about 13%. At the time, that made drug treatment the most effective rehabilitation programme in the entire prison system. There are now nine DTUs but since then, their effectiveness has declined dramatically. The Department’s Annual Report for 2015, shows they now reduce reoffending by only 5%.

Even that figure is optimistic – for a number of reasons. First the 5% reduction only applies to the first 12 months after the prisoner is released.  The Department is too embarrassed to publish figures which would no doubt show that two years after release, there is no reduction at all, at least not one that is statistically significant.

Statistical significance means the outcome is caused by attendance at the rehabilitation programme rather than being the result of random variation. However, reductions of 5% or less are unlikely to be significant – and hidden in the Appendix at the end of the Annual report, the Department acknowledges:

“The rates of some rehabilitation programmes reported are small and below the level of statistical significance.”

In fact the rates for 11 of the Department’s 12 prison programmes are small – averaging between 4 and 6%. In 2015, only one programme (the Special Treatment Unit Rehabilitation programme for violent offenders) made any significant difference – reducing reoffending by 17% (see chart below). That appears to be a statistically significant result.

Improper matching

These poor results are partly the result of dodgy methodology used by Corrections. In order to measure whether a particular programme made any difference, prisoners who attended that programme need to be compared with a control group that didn’t. As the Department notes:

“Outcomes are measured by… comparing the rates of reconviction and reimprisonment for offenders who completed a rehabilitative intervention with the rates of a matched group who did not complete that intervention.”

‘Matching’ means the group of offenders who completed a particular intervention had the same, or very similar, characteristics to a group of inmates who did not.  People in the two groups need to be roughly the same age, the same sex, have the same length of sentence, the same risk of reoffending, the same support on release and so on. Otherwise they are not ‘matched’ and any apparent improvement by prisoners in the group undergoing rehabilitation is random rather than statistically significant.

When it comes to matching prisoners who receive drug treatment, prisoners in the group that don’t receive treatment (the control group) need to have equally serious addictions as those who do receive treatment (the intervention group).

Of course it would be pretty stupid, not to mention unethical, for Corrections to refuse treatment to a group of offenders with severe addictions just so they could use them as a control group. So what the Department does is compare a group of prisoners with addictions with a group of offenders that don’t have addictions, but have a similar risk of reoffending (according to a mathematical formula called the RoC*RoI). But if addicts are ‘matched’ with offenders who are not addicts, the two groups are not matched at all. In fact they could hardly be more different.

This means the Department’s claim that drug treatment in prison reduces reoffending by 5% in comparison with a group of offenders who did not attend treatment is fabricated nonsense .

Reintegration strategy missing

But that’s still only half the story.  In 2011, in Flying Blind, I made the case that more rehabilitation programmes were needed in prison. But I also said that Corrections does not have a reintegration strategy; and that providing rehabilitation in prison is a waste of taxpayer’s money if inmates have no accommodation or are not supported when they get out.

This is especially true of offenders with addictions. If they attend treatment in prison but are then released into the same impoverished, binge drinking, drug taking environment that helped create or cause the problem in the first place, guess what happens. They relapse – and then they reoffend.

In 2015, Corrections spent $169 million on rehabilitation and reintegration programmes, but only about $10 million of this went into reintegration.  That means $159 million was spent on rehabilitation. This is almost a total waste of taxpayer’s money if the Department does not have a realistic, financially supported reintegration strategy. It didn’t have one in 2011. Five years later, it still doesn’t have one.

It was also 2011 when the Government introduced the goal of reducing reoffending by 25% – which was supposed to reduce the total number of people in prison by about 600. But in June 2016, Police Minister Judith Collins was forced to admit that, after five years, it had been reduced by only 6.7%, and the prison population was at an all-time high of 9,400 inmates. Clearly, the Department’s rehabilitation programmes in prison are not working.