Inside England and Wales's prisons crisis: How has performance changed over time?

Performance declines are widespread if not universal – though the gap between best and worst has grown substantially 

Declining performance is not universal across prisons. But it is widespread, with rates of violence, protesting behaviours and self-harm all rising in a clear majority of prisons. These measures all began rising sharply from around 2014, fell to varying degrees in response to the pandemic, then began climbing again and are now nearing or surpassing their pre-Covid peaks. 

This pattern is reflected across the system, with prisons across the whole range of performance now having higher rates of violence than they did 10 years ago. In 2013, nearly two thirds of prisons had fewer than 100 prisoner-on-prisoner assaults per 1,000 prisoners. By 2023, less than a quarter could say the same. ³¹ Ministry of Justice and HM Prison and Probation Service, ‘Safety in custody: quarterly update to March 2024’, 2024, www.gov.uk/government/statistics/safety-in-custody-quarterly-update-to-march-2024  

Self-harm rates show a similar picture, though with a much smaller decline during the pandemic. The median has increased fivefold, from 123 incidents per 1,000 prisoners in 2004 to 642 in 2023. ³² Ministry of Justice and HM Prison and Probation Service, ‘Safety in custody: quarterly update to March 2024’, 2024, www.gov.uk/government/statistics/safety-in-custody-quarterly-update-to-march-2024  Some of this may be down to improved recording practices, but interviewees suggested it reflects a real and sustained increase.

The trend is even more pronounced for ‘incidents at height’, the most common form of protest in prisons. This includes things like climbing on netting or roofs and is sometimes used by prisoners to get an opportunity to speak directly with prison officers, as it typically means they will be taken to segregation afterwards. Incidents at height have dramatically increased since 2012/13 in almost all prisons. Then, 13 incidents per 1,000 prisoners would have put a prison in the worst 20% of prisons. In 2023/24, the same rate would put it in the best performing 20%. ³⁴ Ministry of Justice and HM Prison and Probation Service, HMPPS Annual Digest, April 2023 to March 2024, 2024, ‘Chapter 7 Tables – Protesting behaviour’, www.gov.uk/government/statistics/hmpps-annual-digest-april- 2023-to-march-2024

“[As a prisoner] there are so few staff on the wing, so [incidents at height are] the lever you have to pull to get their attention” – voluntary sector organisation

The gap between the better and worse performers has grown in absolute terms. For example, in 2003, the worst 25% of prisons had more than 138 prisoner-on-prisoner assaults per 1,000 prisoners. This was 61 higher than the median prison. In 2023, the rate for the worst 25% of prisons had risen to 324, some 115 above the median. In relative terms the increase has been bigger at the middle: the median assault rate has risen 171% in that time, while the third quartile has risen 134%. Similar patterns apply to incidents at height and self-harm. 

Prisoner participation in ‘purposeful activity’ – including education, employment, offending behaviour programmes and vocational qualifications – has also dropped sharply over the last 14 years. There is limited available data on this at prison level, particularly before 2023/24,* though the prisons inspectorate and Independent Monitoring Boards have repeatedly raised concerns about the lack of activity in prisons. ⁴¹ HM Inspectorate of Prisons, Purposeful prisons: time out of cell, September 2024, https://cloud-platform-e218f50a4812967ba1215eaecede923f. s3.amazonaws.com/uploads/sites/19/2024/09/Purposeful-prisons-time-out-of-cell-web-2024.pdf ⁴² International Monitoring Boards, National Annual Report 2023, May 2024, retrieved 23 December 2024, https://cloud-platform-e218f50a4812967ba1215eaecede923f.s3.amazonaws.com/uploads/sites/13/2024/05/IMB-2023-National-Annual-Report-.pdf  Interviewees also highlighted that this is a widespread problem, with many prisons offering at best a ‘part-time’ regime where prisoners may have just two half-day sessions of activity a week. ⁴³ Institute for Government interviews, 2024.  

Some outcomes following release have improved substantially, bucking the general trend 

Not all performance measures are moving in the wrong direction. There have been notable improvements in some outcomes after release, specifically the proportion of prisoners housed on their first night in the community and the proportion in employment six weeks later. 

In the median prison, 92% of prisoners were housed on their first night after release in 2023/24. This compares to just 76% in 2019/20. The worst performing prisons have improved the most, with the minimum rate rising from just 42% in 2021/22 to 64% in 2023/24. ⁴⁴ Ministry of Justice, ‘Annual Prison Performance Ratings 2023-24 Supplementary Tables’, 2024, www.gov.uk/government/statistics/prison-performance-ratings-2023-to-2024  Perhaps surprisingly, this increase comes despite resettlement support within prisons having declined in recent years and the probation service being extremely overstretched and rationing support and supervision. ⁴⁵ Comptroller and Auditor General, Improving Resettlement Support For Prison Leavers To Reduce Reoffending, Session 2022-23, HC 1282, National Audit Office, 2023.  

The improvement is likely to have been driven by the expansion of the Community Accommodation Service (CAS-3), piloted in five areas from 2021 and rolled out more widely in July 2023. Whereas traditional accommodation support within prisons involves referring prison leavers to local council homelessness teams and commissioned support services, CAS-3 directly provides up to 12 weeks of emergency housing for those at risk of being released homeless. This has changed the types of accommodation people have on first leaving prison. 

Nationally, there has been a large rise in prison leavers spending their first night in short-term probation-provided accommodation, up from 11% in 2019/20 to 25% in 2023/24; at the same time, the proportion in settled accommodation has fallen from 49% to 45%.** ⁴⁶ Ministry of Justice, ‘Offender Accommodation Outcomes, update to March 2024’, www.gov.uk/government/statistics/offender-accommodation-outcomes-update-to-march-2024

_{* We look at variation in the most recent year below.}

_{** Settled accommodation is long-term, stable accommodation.}

The improvement in first night housing is unfortunately not sustained longer term: data for individual prisons is not available, but nationally the proportion of prisoners in settled accommodation three months on from release has fallen slightly, from 75.3% in 2021/22 to 73.5% in 2023/24. ⁵⁰ Ministry of Justice, ‘Offender Accommodation Outcomes, update to March 2024’, www.gov.uk/government/statistics/offender-accommodation-outcomes-update-to-march-2024  This suggests that the short-term housing support via CAS-3 has had a positive impact, but problems remain when it comes to longer term outcomes.

Employment outcomes after release give a more consistently positive picture, however, with substantial and sustained improvements for most prisons since 2019/20. The best performers are invariably, and unsurprisingly, open prisons – given 25–30% of prisoners tend to be employed in the community during their sentence through release on temporary licence. ⁵¹ Ministry of Justice, ‘Annual Prison Performance Ratings 2023-24 Supplementary Tables’, 2024, www.gov.uk/government/statistics/prison-performance-ratings-2023-to-2024  The average employment rate six weeks after release in open prisons is more than twice the national average at 43%. Improvements over the past four years have been widespread, though, with almost all prisons roughly doubling their six-week employment rates since 2019/20. Employment gains are also spread across all sentence lengths and offence types: while those serving longer sentences typically have higher employment rates after release, prisoners serving all sentence lengths have seen substantial improvements. ⁵² Ministry of Justice, ‘Offender Accommodation Outcomes, update to March 2024’, www.gov.uk/government/statistics/offender-accommodation-outcomes-update-to-march-2024  

Only among the worst performers, which includes all category B and high-security prisons, has there been no improvement. But the employment rates for these prisons may be distorted by the relatively small number of prisoners who are released directly into the community from these prisons. Specialist sex offender prisons, such as Stafford, also typically have very low rates of employment after release.

Given the severe limitations on purposeful activity in prisons and resettlement support, as well as the rationing of probation supervision, this is a surprising finding – especially when conditions within prisons themselves have declined so substantially. The improvements that have been seen in prisoners’ employment outcomes after release may be the result of the government focusing much more on this since 2018. ⁵⁵ Ministry of Justice, Education and Employment Strategy 2018, 2018, retrieved 23 December 2024, www.gov.uk/government/publications/education-and-employment-strategy-2018  

The New Futures Network was set up within HM Prison and Probation Service (HMPPS) to engage directly with employers to find job opportunities for prison leavers. This team is likely to have been less affected by capacity pressures and staff shortages than the resettlement offer provided by front-line prison and probation officers, which may have made it more effective. Other initiatives launched as part of the programme include dedicated ‘employment leads’ to provide the equivalent of a Jobcentre in prison. These were set up in 92 prisons by July 2023, and early evidence suggests that those prisons with employment leads have improved employment rates relative to those without them. ⁵⁶ Comptroller and Auditor General, Improving Resettlement Support For Prison Leavers To Reduce Reoffending, Session 2022-23, HC 1282, National Audit Office, 2023.

Despite these improvements, the employment rate of prison leavers remains low overall: 19% at six weeks after release and 31% at six months. The overall rate is raised by a small number of high-performing prisons. Only 10% of prisons had a six-week employment rate above 36% in 2023/24, and these are all open prisons (including both women’s open prisons). Not all prisons are likely to be able to achieve such high post-release employment rates, as those serving short sentences or convicted of sexual offences are likely to struggle more to secure and hold down employment. Nonetheless, widespread improvements in recent years suggest even the prisons with the most challenging cohorts can improve employment outcomes, so the government should maintain and build on recent successful programmes.