06 May 2025
We have plenty of ideas about how to reduce recidivism. What we need now is evidence
Michael Sanders and Julia Ellingwood
At a time of fiscal restraint, we need to be sure that money is spent on interventions that are actually going to work

On an episode of The Rest is Politics podcast last week, Rory Stewart and Alastair Campbell talked about the Twinning Project, a charity that supports people in prison by teaching them to become football coaches when they leave. Having been to an event for the charity, the hosts were full of praise for the work and its impact, and it is clear that there are a number of high-profile supporters of the work.
At the same time, they were critical of the lack of money received either from the Premier League or from government. Given that what both of us know about football could be written on the back of a postage stamp in wax crayon, it is the latter that we shall focus on here.
There are a number of similar interventions to this one – whereby people are taught a trade during a custodial sentence, which hopefully helps them find and retain employment after they leave prison.
This would be good for everyone – the probationers themselves, who would be diverted from crime and avoid subsequent sentences; their would-be victims, who avoid the material and psychological consequences of crime; and society at large, which would both be safer and avoid the incredibly steep cost of keeping someone in prison.
Other schemes include Oxfordshire’s Tap Social, which teaches brewing skills; Redemption Roasters, which trains people to roast and make coffee; and The Clink, which runs restaurants inside prisons that train people to be chefs.
The logic behind these approaches is clear: giving people a sense of dignity, a skillset, and gainful employment is likely to divert at least some of them from committing crimes, particularly those born of economic desperation.
Nonetheless, the landscape of social policy is littered with ideas that should have worked but ultimately were found not to be (cost-) effective. Given the incredibly high levels of recidivism and re-incarceration that we see in Britain, we can’t say that any individual intervention doesn’t work, but we can certainly conclude that our collective efforts at the moment are not working as well as they should.
While Stewart is dismissive of HM Treasury’s demand that interventions to be funded should have an evidence base (specifically a randomised controlled trial), the alternative is either eminence-based policymaking (Rory and Alastair are super smart, so we should fund what they think works), or vibes-based policymaking (this seems nice, so we should fund it). Neither is a good idea under any circumstances, and both are untenable in an environment of fiscal constraint.
When every pound must be fought for, the Treasury is right to demand evidence. Where we will take Stewart’s argument is that it is not enough for RCTs simply to be demanded – someone has to actually run them, and government is best placed to make this happen. As well as randomised trials, quasi-experimental approaches have already been used to understand interventions to help prison-leavers, including the Peterborough Social Impact Bond and the Ministry of Housing, Communities and Local Government’s Accommodation for Ex-Offenders programme. The problem isn’t that evidence in probation is especially hard to produce; it’s that too little energy and treasure has been expended by government on producing it.
The What Works Network has, over the last decade and a half, made substantial inroads into building an evidence base in a number of policy areas, from crime and policing to education. As we have written previously, given the amount of money and human misery at stake, probation is a glaring omission from the current list of What Works Centres.
The government wants to reduce the number of people in prison and the rate of recidivism that leads the majority of people released from prison to be back inside within 12 months. A what works centre that combines scientific rigour and passionate impatience could help the Treasury and the Ministry of Justice to achieve a step change, as well as creating a case for funding projects like the Twinning Project – if they work.