When Transform Justice was researching courses for those who committed domestic abuse, we asked what proof there was that they changed behaviour. In the case of the main statutory course – Building Better Relationships (BBR) – there was no evaluation which measured its impact on offending. 9 years since the course started, and with 15,180 people having completed it, there still isn’t. The course is accredited, but the accreditation process is totally opaque, with no public information on the panel or the process.  Nearly all offender behaviour courses run by prison and probation are accredited, but this gives little reassurance that they work. Accreditation involves approving the theory behind the course, and the manual.

People became concerned when they heard that the sex offender treatment programme was counter-productive – that the reoffending of those who completed it was higher than those who didn’t. Subsequently a researcher who tried to whistleblow about this sued the MoJ for discrimination. Kathryn Hopkins revealed that evidence showing the programme was ineffective was available years before it was halted.

You would have thought this experience would prompt the Ministry of Justice (or rather HMPPS, which runs prisons and probation) to evaluate all their programmes asap and to publish the results. But they seem to have become paralysed – most prison and probation accredited programmes still don’t have impact evaluations, and HMPPS has suppressed the evaluation of the OPD pathway. This is an approach to those with personality disorder which involves prison ‘therapeutic communities’, and classroom-based cognitive-behavioural psychological courses.  Professor Paul Moran was commissioned to assess whether those assigned to this pathway were more or less likely to reoffend. Unfortunately he found those who follow the OPD pathway are more likely to offend. That research was finished in 2019 and presented at a conference a year ago but attendees were not allowed copies of the slides. HMPPS still hasn’t published the research.

In the case of another programme – Resolve – they have published research which indicates that it doesn’t seem to work. The course was designed for medium and high-risk prisoners convicted of violent crimes. It reduced general offending but made no impact on whether people committed further violent crimes after release. It wasn’t counter-productive but was ineffective. Its not clear whether HMPPS is going to stop it.

I’ve been concerned about the lack of evaluation of the programmes for a while and have regularly asked HMPPS for an update; most recently in January this year. Amy Rees, Director General for Probation pointed out that “a substantial amount of [evaluation] work was initially paused or delayed to support the Department’s response to COVID-19 during the first lockdown and subsequent Tiers”.  Fair enough, but incredibly little progress seemed to be made since her previous letter in July 2019 – bar the publication of two studies including the unsuccessful Resolve. Against a number of studies the January letter says “Scoping has been delayed due to Covid-19 issues and restrictions and is planned to restart during 2021/22” (this document details the evaluation situation in July 19 and Jan 21). Call me churlish but I can’t quite see how scoping could not be done by researchers working from home? Nor why evaluations which have been done, could not be published? In the case of BBR, HMPPS wrote in early 2018 that “Timings for evaluations of BBR and Kaizen have not yet been finalised; plans will be developed following scoping work to establish the most appropriate methods and timing for evaluation”. We are still waiting for that scoping document.

I’m sure few prisoners have done courses during the pandemic, but they will soon restart and we’ll again be making prisoners and those on probation do courses which may not work and may be counter-productive. And some of these courses have been running for years. Justin Russell, Chief Inspector of Probation, has said he is pleased that BBR has continued running during the pandemic but, without an impact evaluation, we have no idea whether it is doing any good.

For fear of getting technical, many of these programmes do have process evaluations, but these are about implementation, not about whether a programme has an impact on reoffending. I also want to emphasise that many, many programmes and interventions do have a positive impact on reoffending. Many of the most successful are not offender behaviour programmes – for instance the Justice Data Lab found that grants from PET (the Prisoners’ Education Trust) had a positive impact on reoffending. So some things do work.