Methods that might reduce recidivism
In England and Wales, the proven reoffending rate for all adult offenders is 28.8%, but the rate jumps to 37.1% for adults released from custody. In Scotland, the equivalent reconviction rate is 27.1% for all offenders, and 42.8% for those specifically released from custody.
These are alarmingly high figures if one of the main purposes of imprisonment is to cut crime. We can see real successes in reducing recidivism in other countries, but there are important caveats about what the evidence actually shows.
The most-cited international example is Norway. Before Norway's prison reforms in the 1990s, the country had a recidivism rate in the range of 60-70%. Today its reconviction rate within two years is around 20%, one of the lowest recorded anywhere. The system is built on a ‘normalization principle,’ the idea that prison should replicate ordinary life as far as security allows, with inmates working, studying, and taking responsibility for their environment. Bastøy, an island prison, has prisoners maintain the island themselves through farming, woodcutting, and fishing, with the explicit goal of teaching responsibility and creating people who can contribute to society on release.
That said, the causal story is contested. In Denmark, Finland, and Sweden, roughly one in three prisoners is reconvicted within two years, not dramatically different from comparable countries. Norway's advantage over its Nordic neighbours could be partly attributable to non-rehabilitation factors, and much of the apparent gap with the United States may reflect differences in underlying crime rates rather than the rehabilitation model itself. So Norway's figures are real but shouldn't be read uncritically as proof that the specific programme design transfers elsewhere.
Cognitive Behavioural Therapy (CBT) is the most extensively tested intervention worldwide. Some analysis has suggested that CBT programmes might reduce recidivism by 20-30%, and the evidence for effectiveness is strongest for high-risk offenders, and extends to those with substance-related offences and sexual offences.
However, when only the strictest randomized controlled trials are considered, prison-based CBT appears to have little or no effect on recidivism, with researchers speculating that the programmes are not linked with psychosocial support upon release, and do not address the housing, employment, and financial difficulties that contribute to reoffending risk. The lesson seems to be that CBT works only when it is part of a broader reintegration structure, not as a standalone in-custody programme.
Mentoring and peer support is used in a European programme called Revion, designed with criminological expertise and evaluated through a randomised controlled trial. It demonstrated remarkable effectiveness, halving recidivism rates for high-risk offenders. Trained and supervised mentors, operating as non-authoritative figures rather than authority figures, provided psychosocial and practical support while challenging criminal thought patterns. Volunteerism was a cornerstone of its success.
‘Restorative justice conferencing’ sounds bizarre, but a recent finding from a US programme has obviously transferable lessons. Youths assigned to the ‘Make-it-Right’ restorative justice programme were 19 percentage points less likely to be rearrested within six months, a 44% reduction. The effects persisted, and participants were 15 percentage points less likely to be rearrested within three years, and 27 percentage points less likely within four years. The mechanism involves the offender taking responsibility and making amends rather than simply receiving a sanction.
Education and vocational training play a useful part. Studies suggest that basic education programmes reduce recidivism by around 5%, vocational educational programmes by around 12%, and post-secondary education by around 25%. These are relatively modest but consistent effects, and they compound with employment outcomes.
The clearest finding across the international evidence is that what happens at the point of release matters at least as much as what happens inside prison. European reintegration programmes have achieved around a 12% reduction in recidivism rates, with significant positive effects on housing and social support, though effects on employment and substance use were less consistent.
Programmes that treat release as a cliff-edge rather than a managed transition tend to underperform. The countries that do best, Norway prominently, treat custody and reintegration as a single continuum rather than two separate systems.
For the UK, the tools exist, including support, mentoring, restorative justice, and education. The evidence for each is positive but not overwhelming in isolation, and the structural question of whether release is properly resourced and supported is probably more important than the content of any specific in-prison programme.
If someone leaves prison with no job, no income, and perhaps even nowhere stable to live, and if they have acquired no marketable skills and have not been prepared and advised on a life back in the community, the chances of re-offending will be higher than they need otherwise have been.
Madsen Pirie