Skepticism often means asking whether something really works. Whether it’s a trendy herbal remedy or the criminal justice system, the key question is: does it do what we want it to do? Unfortunately, while government officials often talk about getting tough on crime, there is little concern to back up current approaches with evidence of effectiveness. That may be because the evidence suggests that current approaches have the opposite of their desired effect, and that there are better alternatives that face resistance from entrenched interests.
Before we assess whether our criminal justice systems work, we have to understand what we want them to be doing. There are at least five distinct goals one could have for a criminal justice system, though most systems theoretically serve multiple goals: retribution, deterrence, rehabilitation, restoration, quarantine.
Of these, deterrence, rehabilitation, and quarantine are the most self-explanatory, as they are all just approaches to harm prevention. Conversely, retributive justice is entirely about figuring out who deserves to be punished and harming them proportionally to the harm they caused, even if doing so serves no other function. Finally, restorative justice focuses on the needs of all involved with the goal of healing past harm and preventing future harm.
While retribution remains a strong desire for many people, defenders of modern criminal justice systems typically argue that these systems are functioning properly by claiming they deter crime, rehabilitate criminals, or keep them quarantined if the first two options fail. For example, the UK’s Justice Secretary Alex Chalk stated in 2023: “We must ensure dangerous criminals face the full force of the law by serving their time in prison. But we must also be honest: if we don’t use prison in a targeted way, we will run out of it.”
Advocates often speak confidently that these systems are working as intended.
For those fortunate enough to be unfamiliar with the detail of risk reduction interventions within UK prisons, ‘Building Choices’ is a Risk-Needs-Responsivity course based in cognitive behavioural techniques that has been rolled out in the last year aimed at people who have been convicted of violent and/or sexual offences. The Ministry of Justice (MoJ) describes the course as follows:
The Building Choices programmes has been designed with a person-first approach at its core, recognising and addressing the individual needs of participants in order to address both offence-specific and wider offending behaviour risks and needs to target the versatility of offending behaviour.
This sounds very much like it will fit the bill if the aim is rehabilitation and offence reduction. The course is also ‘accredited’ with the Ministry of Justice, which they state signifies confidence that a programme is designed based on the best available evidence, monitored to make sure it is delivered as intended, and evaluated to show the outcomes.
However, thanks to a FOIA request, we can see that, far from having strong evidence to support their claims, in many cases the UK government failed to even study the issues. When asked to provide the evaluation justifying the rollout of ‘Building Choices’, the MoJ responded in December 2024 as follows:
No such evaluations exist. To be fair, it will be difficult to prove if a course reduces reoffending until people have gone through it and time has passed to see if they go back to their crime committing ways. Building Choices has only been rolled out estate-wide in the last couple of months.
This invites the question: what is the evidence for other, older courses? Building Choices is being rolled out as a replacement to other courses which have been in place since around 2017. In fact Building Choices is set to replace almost all accredited programmes, with the exception of those designed to address specific sexual paraphilias and extremist ideology.
The main predecessor course for Building Choices is a course called Kaizen. So, what is the evidence for Kaizen working as a rehabilitative tool? Again it appears that no such evaluations exist.
Going back further – the reason Kaizen was rolled out in 2017 was in part because one of its predecessors, the less euphemistically named Sex Offender Treatment Programme (SOTP), had been shown to at best make no difference and at worst increase reoffending rates. SOTP had been running in UK prisons from the early 1990s. An evaluation of SOTP was undertaken in 2012 which indicated that it was counter-productive. However this information was not published and the course was not pulled until 2017, with the MoJ insisting that it was effective until very shortly before it was pulled.
Things are not looking good for accredited programmes aimed at offending behaviour.
There is also a similar story in relation to other interventions designed to address offending in those diagnosed or believed to have personality disorders. The Offender Personality Disorder (OPD) Pathway is a collection of interventions both in custody and the community. An evaluation of the effectiveness of the OPD Pathway between 2012 and 2017 was published in late 2022. The outcome of this evaluation was that the interventions were found to have no quantitative impact on risk reduction. There was no published reason given as to why the publication of this evaluation had taken five years – a similar length of time to the SOTP delay.
But it’s not just about the programmes themselves. We also need to consider the broader environment in which these interventions are delivered.
The UK prison population has been rising steadily and is forecast to exceed 100,000 by 2026. Overcrowding is now endemic: as of 2023, around two-thirds of adult male prisons were officially overcrowded. This directly impacts the ability to deliver interventions in any meaningful way. Her Majesty’s Chief Inspector of Prisons (HMCIP) noted in the 2022–23 annual report that “crowded conditions continued to undermine decent living conditions and access to purposeful activity and rehabilitation services”.
Staffing shortfalls further exacerbate these problems. HMCIP reported that acute staff shortages were “widespread” and that many prisons were “unable to deliver even basic regimes,” with some residents locked in their cells for over 22 hours a day.
Housing is another key issue. Stable accommodation on release is strongly linked to lower reoffending, but many people leave prison homeless or in precarious temporary arrangements. In 2022–23, only around half of people released from prison had confirmed accommodation upon release. The situation is especially dire for people on probation, many of whom face weeks in hostels or nights spent rough sleeping.
Mental health provision is similarly inadequate. The prison population has a high prevalence of mental health needs – estimated by the Ministry of Justice to be as high as 70% – but access to treatment is often limited or delayed. HMCIP found in multiple inspections that services were “under pressure, with long waiting lists and insufficient support for those with complex needs”.
Recall rates are also climbing. In the year ending September 2023, over 11,000 people were recalled to custody, often for non-criminal breaches, such as missing appointments. This revolving door undermines continuity of care and disrupts engagement with any rehabilitative services that do exist.
When these structural issues are taken together, it becomes difficult to see how any rehabilitative intervention – no matter how well-designed – could be expected to succeed. If the conditions are so poor that programmes can’t be delivered properly, people can’t access housing or treatment on release, and many are recalled for administrative reasons, then the entire rehabilitation project begins to look more like rhetoric than reality.
If rehabilitation is unlikely to succeed in the absence of social reforms, are the prospects for restorative justice approaches any better? Restorative justice programs, which arose as a response to the adversarial nature of modern justice systems, aim to provide alternative methods of accountability and harm prevention.
Restorative justice relies on mediated discussions with offenders and sometimes between offenders and victims where participants share their understanding of the harm caused by the actions and the personal and community needs that have to be met to satisfy the need for genuine accountability and to reduce the risk of future harms. Advocates of restorative justice claim it is both a more humane and more effective method of dealing with harmful behavior than the current systems, while critics object that it is too offender-centric and is only functional for addressing minor harms.
As with any form of justice, the effectiveness of restorative programs depends on the specific methods they use, but a recent meta-analysis of restorative justice research found that “restorative justice programs resulted in greater victim and client satisfaction, victims’ views of procedural justice, and client accountability compared to traditional legal system approaches”. Increased emphasis on victim needs over punishment is the most likely explanation for the greater victim and client satisfaction, as well as improved views on procedural justice as a concept.
Restorative justice also does better on more subjective assessments of accountability, such as “self-report assessments of accountability for the criminal act, victim perceptions of client accountability, and whether or not the client expressed feelings of accountability, such as through apologizing”. So, there do appear to be some legitimate benefits of restorative justice as an alternative to traditional approaches.
It is likely that research will continue to emphasize the broader truth that criminal justice systems are not a solution to social injustice
Unfortunately, when we take into consideration the quality of restorative justice studies, the current quantitative data suggests that restorative justice practices do not have a significant impact on recidivism. The authors of the meta-analysis caution though that we should be careful not to read too much into these results, as the primary aim of restorative approaches is not to reduce recidivism, and those approaches could be functionally combined with more empirically supported rehabilitative models like Risk-Need-Responsivity to more reliably reduce recidivism.
Risk-Need-Responsivity models also focus on addressing the needs that lead to harmful behaviors, while assessing the risk posed by the individual when determining how significant the intervention needs to be in terms of intensity and duration. However, as we see from the FOIA results above, the effectiveness of Risk-Need-Responsivity approaches and empirical support for them may also be overstated. Progress towards more humane criminal justice is likely to involve synthesising various needs-based approaches, while continuing to weed out the retributive urges that create a persistent demand to focus on punishment.
More research is clearly needed to determine what balance of criminal justice principles and practices works best, which in turn depends on what we want that system to accomplish. However, it is likely that any such research will continue to emphasise the broader truth that criminal justice systems are not a solution to social injustice; at best they can ameliorate the symptoms long enough for us to correct the systemic causes of harmful behavior.
