Build 10,000 new detention places and expand prison capacity
Reform UK · what the evidence says
An independent, source-checked look at Reform UK’s policy “Build 10,000 new detention places and expand prison capacity” — what it would actually do across the things that affect your life. Every claim below quotes the source behind it. How this works.
Public finances & the next generation — Hurts
major · moderate confidence
Building 10,000 new prison places would cost roughly £5 billion in capital alone, plus hundreds of millions a year to run, with no funding source stated in the policy. This passes a large, ongoing bill to future taxpayers without any offsetting investment return.
The evidence
- The policy commits to building 10,000 new state-built and managed detention places, including commissioning disused military bases. — reformparty.uk (manifesto) — “Reform UK will start building 10,000 new state-built and managed detention places and expand prison capacity, commissioning disused military bases if needed”
- The upfront capital cost per new closed prison place is estimated at £500,000 in 2024-25 prices, implying ~£5 billion for 10,000 places. — assets.publishing.service.gov.uk (government) — “The upfront capital cost per new closed prison place is estimated at £500,000 (2024-25 prices)”
- The government's existing plan for 20,000 new prison places has escalated from £5.2 billion to between £9.4 billion and £10.1 billion — an 80% cost overrun. — theguardian.com (media) — “initially estimated at £5.2 billion in 2021, is now projected to cost between £9.4 billion and £10.1 billion, representing an increase of at least £4.2 billion (80%) over original estimates”
- Construction inflation (up 40%) and planning delays are key drivers of cost escalation on existing prison expansion. — nao.org.uk (institutional) — “construction inflation (prices up by 40%) and delays in securing planning permission as key contributors to these escalating costs”
- The expanded existing estate is estimated to cost approximately £670 million per year more to run. — nao.org.uk (institutional) — “the current government's expanded estate will be approximately £670 million per year more expensive to run (2024-25 prices)”
- The average annual cost to keep a prisoner in England and Wales was close to £57,000 in 2023/24. — en.wikipedia.org (media) — “The average cost to keep a prisoner for a year in England and Wales was close to £57,000 in 2023/24”
- The NAO found previous government prison expansion plans were unrealistic, ran years late, and were billions over budget. — theguardian.com (media) — “previous government prison expansion plans were "unrealistic," "not prioritised," ran years late, and were billions over budget”
- Additional detention places would require commensurate increases in healthcare and staffing, historically a challenge. — gov.uk (media) — “More detention places would require a commensurate increase in healthcare provision and staff, which has historically been a challenge”
- The policy's 'life means life' emphasis contributes directly to rising prison population projections, as offenders serve longer in custody. — commonslibrary.parliament.uk (government) — “This approach contributes directly to the rising prison population projections, as offenders serve longer periods in custody”
Biggest unknown: Whether any fiscal offset (reduced reoffending, redirected budgets, or explicit tax/borrowing plan) would be attached to the policy at implementation — none is stated here.
Our reading: The policy commits to building 10,000 new state-built prison places with no stated funding mechanism — no tax, no borrowing rule, no offset. The evidence grounds a substantial fiscal impact: at the government's own estimate of £500,000 per place (E8), the capital bill alone is approximately £5 billion, consistent with the trajectory of the existing 20,000-place programme which has already blown past its original £5.2bn estimate by at least £4.2bn (E6). Construction inflation of 40% and planning delays (E7) mean real costs are likely to exceed even the headline figure. On the revenue side, each additional prisoner costs roughly £57,000 per year (E10); 10,000 new places filled to capacity add roughly £570 million annually in running costs, in line with the NAO's estimate that the current expansion already costs £670m/year more to operate (E9). The 'life means life' commitment (M) directly drives up the prison population by extending sentences (E21), compounding both capital and recurrent costs over time. None of this spending finances productive investment — it finances incarceration, which is consumption expenditure. There is no cited evidence of an offsetting mechanism such as reduced reoffending savings or a dedicated funding source. The NAO's finding that past prison expansion was 'unrealistic' and 'billions over budget' (E28) is a strong implementation-risk signal that actual costs could be higher still. The verdict is 'worsens/major': a large, unfunded, recurrent spending commitment that expands consumption borrowing and passes a multi-billion-pound long-term liability to future taxpayers, with no evidenced fiscal offset.
Crime, justice & national security — Helps
moderate · low confidence
Building more prison places addresses a real capacity crisis and keeps dangerous offenders off the streets for longer, which directly improves public safety through incapacitation. However, severe delivery risks — massive cost overruns on similar programmes and staffing shortfalls — mean the safety benefit may be much smaller or much slower than stated.
The evidence
- The policy commits to building 10,000 new state-managed detention places, including on disused military bases, and to 'life means life' sentences for those who endanger the public. — reformparty.uk (manifesto) — “Reform UK will start building 10,000 new state-built and managed detention places and expand prison capacity, commissioning disused military bases if needed, to ensure 'life means life' for those who endanger the public.”
- England and Wales prisons are almost at breaking point: as of June 2026, the prison population stood at 87,441 against a usable capacity of 89,694, leaving only 2,253 spare places. — insideoutjustice.com (media) — “the total prison population in England and Wales stood at 87,441, with a usable operational capacity of 89,694, indicating an overcrowding rate of 97.5% and limited available spaces (2,253).”
- Nearly three-quarters of prisons were overcrowded in 2024–25, affecting over 21,600 people — a quarter of the prison population. — prisonreformtrust.org.uk (media) — “almost three-quarters (72%) of prisons in England and Wales were overcrowded, affecting over 21,600 people—a quarter of the prison population.”
- The prison population is projected to rise further to between 98,000 and 103,600 by March 2030, deepening the capacity shortfall. — commonslibrary.parliament.uk (government) — “more recent projections anticipate a rise to between 98,000 and 103,600 by March 2030, with a central estimate of 100,600.”
- While incarceration prevents crime during the sentence, approximately half of adult prisoners reoffend within one year of release, limiting longer-term crime-reduction effects. — parliament.uk (government) — “Approximately half of adult prisoners reoffend within one year of release.”
- The existing government's plan for 20,000 new prison places has already ballooned from £5.2 billion to between £9.4 billion and £10.1 billion — an increase of at least 80%. — theguardian.com (media) — “The current government's existing plan to deliver 20,000 new prison places, initially estimated at £5.2 billion in 2021, is now projected to cost between £9.4 billion and £10.1 billion, representing an increase of at lea…”
- The NAO found the previous prison expansion plans were 'unrealistic,' ran years late, and were billions over budget, attributing capacity problems to governments' failure to align places with criminal justice policy. — theguardian.com (media) — “The National Audit Office has been particularly scathing, concluding that previous government prison expansion plans were "unrealistic," "not prioritised," ran years late, and were billions over budget.”
- Longer sentences contribute directly to rising prison population projections, as offenders serve more time in custody. — commonslibrary.parliament.uk (government) — “This approach contributes directly to the rising prison population projections, as offenders serve longer periods in custody.”
- Overcrowding is linked to deteriorating safety, increased use of force, and reduced opportunities for rehabilitation. — prisonreformtrust.org.uk (media) — “Overcrowding is linked to deteriorating safety, increased use of force, and reduced opportunities for purposeful activity and rehabilitation.”
Biggest unknown: Whether the programme can actually be delivered at scale and on time, given that the existing government's comparable 20,000-place plan has already run 80% over budget and years late.
Our reading: The core safety mechanism here is incapacitation: keeping dangerous offenders off the streets for longer directly reduces the crimes they would otherwise commit while free. This is the clearest O5 gain. The policy also responds to a genuine and worsening capacity crisis — at 97.5% utilisation, the system is already forcing compromises (including early releases) that undermine public safety, and population projections point to a further rise to over 100,000 by 2030. Adding 10,000 places would provide meaningful headroom against that trajectory. However, several factors limit confidence in the magnitude and timing. First, delivery risk is extreme. The comparable existing government programme for 20,000 places has run 80% over budget, and the NAO explicitly found such plans 'unrealistic' and years late. There is no credible reason to expect a further 10,000-place programme via disused military bases to perform better. The safety benefit only materialises when places are actually built and staffed. Second, 'life means life' sentencing — the policy's demand-side driver — pushes population projections up further, meaning the 10,000 places may be consumed by sentencing policy rather than relieving the crisis. Third, the ~50% one-year reoffending rate indicates that incapacitation without rehabilitation does not translate into durable long-term crime reduction; the overcrowding that persists in this scenario actively degrades rehabilitation programmes. The net O5 verdict is 'improves' in direction — incapacitation is real and the capacity shortfall is real — but low confidence given delivery history, and long-term because any places commissioned now are years from completion.