Which Asylum Cost Figure Is Real?

£4bn, £4.7bn, £5.4bn, £15.3bn, £700m — nearly every figure quoted about asylum and Rwanda scheme costs is real, sourced from an official document. They're just measuring different things: one year vs ten, hotels alone vs all accommodation, a scrapped scheme vs the ongoing system. Here's what each one means.

At the end of 2018, the Home Office was accommodating around 47,500 asylum seekers; by March 2023 that had risen to around 112,000. Annual spending told a similarly steep story: asylum support costs were around £3.6bn in 2022/23 (nearly double the year before), rising further under the previous government, with hotels alone costing over £2bn a year at the peak — hotels were only ever meant to be "contingency accommodation" for when other housing couldn't be sourced, but were used at scale and for far longer than originally intended from 2020 onward.

The most current picture, as of March 2026, is genuinely improved on the peak. The total number of people requiring accommodation support has fallen 13% since March 2023, to around 98,000. The number in hotels specifically has fallen much further: from 29,561 (29% of supported asylum seekers) in June 2024, when the current government took office, to 20,885 (21%) by March 2026 — and the number of hotels in use fell from 213 to 170 over roughly the same period. The Home Office's own accounts show hotel spending fell by £0.9bn in 2024/25 compared to the year before.

That said, the improvement is very unevenly spread. As of March 2026, London still had 58% of its accommodated asylum seekers in hotels — nearly three times the national rate — followed by the South East (39%) and South West (33%). The national figure falling to 21% partly reflects a genuine shift toward "dispersal" housing elsewhere in the country, not an even reduction everywhere.

The current government's Spending Review 2025 set a target to end the use of hotels entirely by 2029, funded partly through a £3.25bn Transformation Fund, with ministers estimating this would save £1bn a year once achieved, plus a further £1bn a year in savings from wider asylum system reforms by 2028/29 compared to 2024/25 prices.

The government's claim: that it has "restored grip on the system" — citing a 52% increase in asylum decision-making and 24,000 removals of people with no right to remain, with hotel numbers at their lowest since the last election. This claim is now well-supported by the current data: hotel use genuinely has fallen substantially (29,561 → 20,885 people, 213 → 170 hotels) since the government took office in mid-2024.

The critical framing: that regardless of which government is in office, the underlying £15.3bn ten-year contract cost trajectory reflects years of accumulated system strain, and that London specifically has seen almost no improvement (58% of its asylum seekers remain in hotels, the highest rate anywhere in the country) — meaning the national improvement is real but geographically lopsided, and "forecast savings" for the remainder of the decade are still projections, not money already banked.

Separate from ongoing accommodation costs, the previous government's Rwanda deportation scheme — intended to deter Channel crossings by sending asylum seekers to Rwanda for processing instead of the UK — is its own significant, and highly contested, cost story. The scheme was signed in April 2022. A 2024 government watchdog report estimated it could cost around £630m in total. In practice, by the time the newly-elected Labour government scrapped it in July 2024, the Home Office confirmed £700m had already been spent to relocate just four people, who went voluntarily.

Home Secretary Yvette Cooper called it "the most shocking waste of taxpayer money I have ever seen," and revealed the previous government had planned to spend a further £10bn on the scheme had it continued. Cooper said scrapping it would save £220m in further direct payments to Rwanda, plus up to £750m already set aside to cover the scheme for that year alone. The then-shadow home secretary, James Cleverly, who had overseen the scheme, accused Cooper of using "hyperbole and made-up numbers" in response.

One official calculation, tied to a threshold written into the UK-Rwanda deal, illustrates just how expensive the scheme's structure was: if exactly 300 people had been sent (the point at which the UK owed Rwanda an extra one-off £120m payment), the cost would have worked out to roughly £2 million per person once earlier lump-sum payments were included — falling to around £450,000 per person only if numbers reached as high as 2,000. The government's own justification at the time was that "illegal migration" could otherwise cost the taxpayer up to £11bn a year by 2026, a projected future cost used to argue the scheme's expense could be worthwhile if it deterred crossings — though the previous government's own impact assessment separately admitted there was little to no evidence policy changes in a destination country actually deter people from leaving their country of origin.

The story isn't fully closed even now. Despite Starmer declaring the scheme "dead and buried" on his first day in office, the government didn't formally exit the deal until December 2025 — and Rwanda is currently pursuing the UK through international arbitration at the Permanent Court of Arbitration in The Hague, claiming it is still owed roughly £100m in promised payments plus a further £6m in compensation for the UK's failure to accept an agreed number of refugees under the deal's terms. The UK government disputes this, arguing the two countries agreed in November 2024 that Rwanda would forgo the payments — a dispute Rwanda denies, meaning the final total cost of the scheme is not yet fully settled.