Build Four New Prisons and Expand Removals

Public finances & the next generation — Hurts

moderate · moderate confidence

Building four new prisons adds billions in capital spending and hundreds of millions in annual running costs, with no stated funding source and a track record of severe cost overruns. Removing more foreign national offenders could partially offset costs, but the savings are smaller than the new liabilities.

The evidence

• The policy commits to building four new prisons completing 20,000 new prison places by 2030. — conservatives.com (manifesto) — “build four new prisons, completing the programme of 20,000 new prison places by 2030”
• The estimated cost of the 20,000-place expansion has escalated from £5.2bn to between £9.4bn and £10.1bn — an increase of at least £4.2bn (80%). — prisonreformtrust.org.uk (media) — “The estimated cost for the expansion has escalated from £5.2 billion in 2021 to between £9.4 billion and £10.1 billion, representing an increase of at least £4.2 billion (80%).”
• The expansion is expected to increase annual running costs of the prison estate by approximately £670m — an 18% increase on 2023-24. — resolutionfoundation.org (institutional) — “The expansion is expected to increase the annual running cost of the prison estate by approximately £670 million, an 18% increase compared to 2023-24.”
• As of September 2024, only 6,518 of the 20,000 planned places had been delivered, meaning the four new prisons represent a large share of remaining unbuilt capacity. — prisonreformtrust.org.uk (media) — “As of September 2024, only 6,518 of these places had been delivered.”
• The NAO found the 20,000-place commitment was five years late and £4.2bn over budget, describing it as 'unrealistic and not prioritised'. — prisonreformtrust.org.uk (media) — “the government's prison expansion plan was "unrealistic and not prioritised." They found that the commitment to 20,000 new places would be five years late and £4.2 billion over budget.”
• Cost overruns were driven by planning difficulties, unrealistic timelines, and a 40% rise in construction sector inflation. — prisonreformtrust.org.uk (media) — “overestimating the ability to gain planning permission for new prisons, unrealistic timelines, insufficient understanding of programme requirements, and a 40% rise in construction sector inflation.”
• FNO presence in UK prisons costs approximately £537m per year on average since 2019. — ukonward.com (media) — “analyses indicating an average annual cost of approximately £537 million since 2019.”
• Increased FNO removals could reduce ongoing custody costs, but removal numbers remain well below the full FNO population of around 10,321–10,800. — ukonward.com (media) — “foreign nationals constituted approximately 12% of the prison population in England and Wales, numbering around 10,321 to 10,800 individuals.”

Biggest unknown: Whether the prison-building programme would be funded by new borrowing, cuts elsewhere, or tax rises — and whether FNO removal savings materialise at sufficient scale to dent the fiscal impact.

Our reading: The prison-building component of this policy carries a substantial and growing fiscal liability. The capital cost for the broader 20,000-place programme has already ballooned by over 80% to up to £10.1bn, driven by planning failures, inflation, and poor programme management — and the NAO judged it unrealistic. Adding four more prisons extends this pattern. Beyond capital costs, annual running costs rise by around £670m per year indefinitely — a recurrent pressure with no stated funding mechanism in the policy text. This is consumption-adjacent spending (custody is not a growth-enhancing investment in the way infrastructure is), so it worsens the long-run debt path unless offset. The FNO removal strand offers a partial fiscal counterweight: if the roughly 10,300–10,800 foreign nationals in custody were progressively removed, savings against the ~£537m annual baseline could be meaningful. However, removal volumes — 5,179 in the year to July 2025 — are well below the stock, and administrative and legal barriers have historically limited throughput. The partial savings do not plausibly offset the capital and running-cost burden of four new prisons. No funding source is stated for either the capital or running-cost commitments, which is the core O12 concern. The time horizon is long-term because the capital costs compound over the build period and the running-cost increase is permanent. Confidence is moderate rather than high because the final fiscal envelope depends on whether the programme avoids further overruns and on the pace of FNO removals.

Crime, justice & national security — Helps

moderate · low confidence

Building more prisons and removing foreign national offenders addresses a genuine, severe prison capacity crisis that is already undermining the justice system — but the existing programme is already five years late and massively over budget, so whether this policy actually delivers by 2030 is highly uncertain.

The evidence

• The policy commits to building four new prisons, completing 20,000 new prison places by 2030, using planning streamlining and scrapping legacy EU rules. — conservatives.com (manifesto) — “build four new prisons, completing the programme of 20,000 new prison places by 2030, by scrapping legacy EU rules and streamlining the planning system”
• The policy also commits to removing more Foreign National Offenders via the Early Removal Scheme and new Prisoner Transfer Agreements. — conservatives.com (manifesto) — “remove more Foreign National Offenders by increasing removals under the Early Removal Scheme and negotiating more Prisoner Transfer Agreements”
• The adult male prison estate has been critically overcrowded, running at over 99% capacity for much of the period since early 2023. — prisonreformtrust.org.uk (media) — “The adult male estate has run at over 99% capacity for much of the 18 months since February 2023.”
• The prison population reached 85,900 by October 2024. — prisonreformtrust.org.uk (media) — “By October 2024, there were 85,900 people in prison in England and Wales.”
• The prison population is projected to rise to around 100,000 by 2030, outpacing current build plans. — commonslibrary.parliament.uk (government) — “The prison population is projected to increase to around 100,000 by 2030.”
• The MoJ projects a shortage of 12,400 prison places by end of 2027 even accounting for the building programme. — prisonreformtrust.org.uk (media) — “the Ministry of Justice (MoJ) projects a shortage of 12,400 prison places by the end of 2027 if demand for places continues to increase as forecast”
• As of September 2024, only 6,518 of the promised 20,000 new places had been delivered. — prisonreformtrust.org.uk (media) — “As of September 2024, only 6,518 of these places had been delivered.”
• The NAO found the 20,000-place commitment is now not expected until 2031 — five years late — and described the plan as 'unrealistic and not prioritised'. — prisonreformtrust.org.uk (media) — “the government's prison expansion plan was "unrealistic and not prioritised." They found that the commitment to 20,000 new places would be five years late and £4.2 billion over budget”
• Delays in the existing programme stem from overestimating ability to gain planning permission, unrealistic timelines, and a 40% rise in construction sector inflation. — prisonreformtrust.org.uk (media) — “overestimating the ability to gain planning permission for new prisons, unrealistic timelines, insufficient understanding of programme requirements, and a 40% rise in construction sector inflation”
• Foreign nationals constituted approximately 12% of the prison population as of June 2023, numbering around 10,321 to 10,800 individuals. — ukonward.com (media) — “foreign nationals constituted approximately 12% of the prison population in England and Wales, numbering around 10,321 to 10,800 individuals”
• FNO removals have been rising: in the year to July 2025, 5,179 FNOs were returned, a 14% increase on the year before. — prisonreformtrust.org.uk (media) — “In the year to July 2025, 5,179 FNOs were returned, a 14% increase on the year before.”
• The Chief Inspector of Borders and Immigration found the ERS and Facilitated Return Scheme were not administered effectively, with casework delays adding significant costs. — vertexaisearch.cloud.google.com (media) — “the ERS and Facilitated Return Scheme (FRS) were not administered effectively, and casework delays added significant costs”
• Nearly 40% of ECHR-based removal appeals were allowed between 2008 and 2021, representing a structural barrier to increasing FNO removals. — ukonward.com (media) — “nearly 40% of appeals allowed on these grounds between April 2008 and June 2021”

Biggest unknown: Whether planning, cost, and construction barriers can be overcome to deliver on the 2030 target, given the NAO found the existing programme 'unrealistic' and five years behind schedule.

Our reading: The policy addresses a genuine and severe O5 problem: the justice system's ability to sentence, detain, and incapacitate offenders is materially compromised by a prison estate running above 99% capacity, with a projected shortfall of 12,400 places by 2027 even accounting for current building. Critically overcrowded prisons impair safe custody, rehabilitation, and effective sentencing — all direct O5 harms. More prison places, if delivered, would improve the system's protective capacity. The FNO removal strand likewise has a direct O5 rationale: removing convicted foreign nationals frees capacity and removes individuals already shown to have offended. Removals are already rising (5,179 in the year to July 2025), and the Early Removal Scheme is an established mechanism. However, the delivery case is deeply problematic. The NAO found the existing 20,000-place programme 'unrealistic' and five years late; as of September 2024 only 6,518 places had been delivered. The same planning and construction barriers this policy claims to fix (via planning streamlining and scrapping EU rules) are precisely what caused the delays in the first place, and experts note domestic criminal law is largely unaffected by Brexit, casting doubt on the EU-rules framing. The policy therefore offers the right instrument for a real problem, but the evidence strongly suggests it cannot deliver on its stated 2030 timeline — making the O5 improvement real in theory but highly uncertain in practice. The FNO removal strand is more credible as an incremental improvement (trend already positive) but faces structural ECHR and administrative barriers that limit scale. Verdict: improves in direction if delivered, but low confidence given documented delivery failure on the identical programme already underway.

Immigration & border control — Moves toward more control

We don’t call this better or worse — that’s your call; we only show which way the policy moves it.

moderate · moderate confidence

This policy moves toward more controlled immigration by increasing removals of foreign national offenders from UK prisons. The actual scale of reduction depends on how many additional removals are achieved beyond current trends.

The evidence

• The policy commits to removing more Foreign National Offenders via the Early Removal Scheme and new Prisoner Transfer Agreements. — conservatives.com (manifesto) — “remove more Foreign National Offenders by increasing removals under the Early Removal Scheme and negotiating more Prisoner Transfer Agreements”
• Foreign nationals make up around 12% of the prison population in England and Wales. — ukonward.com (media) — “foreign nationals constituted approximately 12% of the prison population in England and Wales, numbering around 10,321 to 10,800 individuals”
• FNO removals have been increasing, with 5,179 returned in the year to July 2025. — prisonreformtrust.org.uk (media) — “In the year to July 2025, 5,179 FNOs were returned, a 14% increase on the year before”
• The Early Removal Scheme already accounts for over half of deportations between July 2024 and January 2025. — ukonward.com (media) — “The Early Removal Scheme (ERS) was responsible for over half of deportations between July 2024 and January 2025”
• Administrative inefficiencies in the ERS and related schemes may limit the additional removals achieved. — vertexaisearch.cloud.google.com (media) — “criticisms from the Chief Inspector of Borders and Immigration suggest that the schemes for FNO removal have historically faced administrative challenges and inefficiencies, potentially limiting their full impact”

Biggest unknown: Whether negotiating more Prisoner Transfer Agreements and expanding the Early Removal Scheme will significantly increase removals above the already-rising baseline, given past administrative inefficiencies flagged by the Chief Inspector of Borders and Immigration.

Our reading: The policy explicitly targets increased removals of foreign national offenders, which directly reduces the number of non-British nationals residing in the UK (in custody). With FNOs at around 12% of the prison population and removals already rising, expanding the Early Removal Scheme and adding Prisoner Transfer Agreements represents a continued push toward more controlled and lower net migration. However, the ERS already handles over half of deportations and has faced administrative criticism, so the marginal increase from further expansion is uncertain.