Research

Articles backed by data. Findings verified against source. Scrutinise the methodology, check the sources, draw your own conclusions.

Latest

Questions answered

All questions →
FindingSpending3 September 2026The Home Office publishes about £11 in every £100 it spends on asylum11%Asylum spending itemised in transparency data, 2024/25The Home Office spent about £4.9 billion on asylum in 2024/25. Its own published record of payments over £25,000 accounts for £540 million of that. The rest is reported in totals, not payment by payment.Read more →AnalysisAccountability3 September 2026The asylum accommodation contracts have been changed 257 times257Modifications to the seven asylum accommodation contracts, 2019 to January 2025Seven regional contracts, modified 257 times in under six years. The ten-year cost estimate has risen from £4.5 billion at award to £15.3 billion.Read more →AnalysisAccountability3 September 2026Clearsprings Ready Homes is absent from two years of published Home Office spending2 yearsPeak hotel years with no published Home Office payment to ClearspringsClearsprings Ready Homes holds two of the seven asylum accommodation contracts, covering southern England and Wales. It does not appear in the Home Office's published spending data for 2022/23 or 2023/24, the two years when hotel use peaked.Read more →FindingAccountability3 September 2026£883 million went to councils for asylum. £1.6 million of it is findable£1.6mOf £883m in council asylum grants, the amount traceable by supplier nameThe Home Office paid councils about £883 million in 2024/25 for dispersal accommodation and for children who arrive alone. Searching every supplier name in its published spending file finds £1.6 million of it.Read more →

Money and contracts

LatestSpending27 September 2026Asylum seekers could be charged about £10,000 for their support. The Home Office has not said what it would raise£10,000Expected flat-rate charge per adult, Immigration and Asylum Bill 2026The Immigration and Asylum Bill would let the Home Office charge adults who received asylum support a flat rate expected to be around £10,000, if they can afford it, before they can settle. Its impact assessment puts no figure on what the charge would recover, and it applies only to claims made after it comes into force.Read more →LatestAccountability27 September 2026Military sites house 1,162 asylum seekers, about 1% of those on support, against 16,021 in hotels1,162People at Wethersfield and Crowborough, 30 June 2026At 30 June 2026 the two large sites in use held 1,162 people: 744 at Wethersfield in Essex and 418 at Crowborough Training Camp in East Sussex. That is about 1% of the 93,293 on asylum support, against 16,021 in hotels. Cameron Barracks was dropped without being used, and three more military sites are proposed.Read more →FindingAccountability3 September 2026MHCLG labels 63% of everything it publishes 'Grant payable by the Exchequer'63%Share of published MHCLG spending carrying one uninformative labelHomes for Ukraine and the Afghan resettlement schemes are paid by MHCLG, not the Home Office. Across 22 months of its published spending, only £22.8 million mentions refugees at all, while £51.9 billion carries a single catch-all narrative.Read more →AnalysisAccountability3 September 2026Five of the seven asylum accommodation contracts have exceeded their profit-share thresholds5 of 7Asylum accommodation contracts above their profit-share thresholdSupplier profit on the asylum accommodation contracts has averaged 7%. Five of the seven contracts have passed the point at which profit must be shared back with the Home Office.Read more →AnalysisSpendingUpdated 31 August 2026Asylum hotels cost £5.77 million a day in 2024/25. The hotel population has since halved.£5.77MDaily asylum hotel cost, 2024/25The House of Commons Library puts Home Office spending on asylum hotel accommodation at £2.1 billion in 2024/25, which is £5.77 million a day, inside £4 billion on asylum support altogether. That is a 2024/25 figure and it is the most recent hotel-specific one published. The population it paid for has since fallen from 30,657 people in hotels to 16,021.Read more →InvestigationSpending14 April 2026Hotels house 29% of asylum seekers and take 76% of the contract cost76%Hotel share of accommodation contract cost, first seven months of 2024-25197 asylum hotels house approximately 31,000 people, 29% of the 107,003 on the supported asylum population at 31 December 2025. Hotels took around 76% of the cost of the accommodation contracts in the first seven months of 2024-25, £1.3 billion of an estimated £1.7 billion. Dispersal housing is roughly 10x cheaper per person per night. The hotel-heavy model is a policy choice, not a necessity.Read more →InvestigationAccountability14 April 2026Three companies control every asylum hotel in Britain3Prime providersSerco, Mears Group, and Clearsprings Ready Homes operate regional monopolies over asylum accommodation under the AASC. Each controls a geographic region. All hotel procurement, dispersal housing, and support services flow through three companies. The 10-year contract: £15.3 billion.Read more →

Arrivals, decisions and returns

LatestRoutes27 September 2026One in, one out has returned 1,087 people to France, against 27,920 small boat arrivals3.9%Small boat arrivals matched by a return to France, 6 August 2025 to 30 June 2026 (our calculation)The UK returned 1,087 people to France under the one in, one out agreement between 6 August 2025 and 30 June 2026, and took 1,117 in return. Over the same dates 27,920 people crossed the Channel in small boats. Returns are rising, and the pilot's current extension ends on 1 October 2026.Read more →FindingRoutesUpdated 27 September 2026Small boat arrivals are down 42% on last year, and still 88% of illegal entry18,609Small boat arrivals, 1 January to 24 September 202618,609 people were detected arriving by small boat between 1 January and 24 September 2026, against 32,188 over the same period of 2025, a fall of 42.2%. The complete 2025 year was 41,472. Small boats still account for 87.6% of detected illegal entry in the year to June 2026, so the route has shrunk without being displaced.Read more →FindingRoutes31 August 2026Four Lancashire authorities are in the national top nine for asylum support per head4 of 9Lancashire authorities in the national top nine by rate, 30 June 2026Counted per head of population rather than in absolute numbers, Hyndburn is second of 361 UK local authorities on asylum support, Burnley fourth, Blackburn with Darwen sixth and Pendle ninth. Only Glasgow City is above them. None of the ten largest caseloads in absolute terms appears in the top nine by rate.Read more →FindingRoutesUpdated 31 August 202693,293 on asylum support. One in five is in ten councils.19.6%Share of the supported population in 10 councils, 30 June 202618,301 of the 93,293 people on asylum support at 30 June 2026 were in ten local authorities, 19.6% of the national total. Glasgow City carries 3,938, more than twice the next area. Dozens of councils carry nobody. The supported population has fallen from 107,003 at the end of 2025, but the concentration has barely moved.Read more →AnalysisRoutesUpdated 31 August 2026Refugee family reunion peaked in 2024 and is now falling12,888Refugee family reunion grants, year ending June 202612,888 refugee family reunion grants in the year to June 2026, down from 20,643 in the year to June 2025. The route peaked at 19,258 in calendar 2024 after running between 3,700 and 7,500 a year for the previous decade. It is still 73% above the 2019 level, and the direction has reversed.Read more →FindingRoutesUpdated 31 August 2026The North West carries 1 in 5 people on asylum support. More than London.20.3%North West share of the supported population, 30 June 202618,962 people on asylum support in the North West at 30 June 2026, 20.3% of the 93,293 national total. London carries 12,140, or 13.0%. The North West has held the largest regional share through every quarter of the current decline, and its share has risen slightly as the national total has fallen.Read more →FindingBacklogUpdated 31 August 202640,168 asylum seekers waiting for an initial decision40,168Awaiting an initial decision, 30 June 202640,168 people were waiting for an initial decision on an asylum claim at 30 June 2026. The queue has fallen in six consecutive quarters, from a peak of 175,457 in June 2023 and 64,426 at the end of 2025. It is falling because decisions are outrunning claims: 22,176 initial decisions against 17,233 new claims in the quarter.Read more →ArticleRoutes27 May 2026Asylum grant rates rise sharply on appeal, across 101 nationalities+14.6ppAverage appeal uplift across 101 nationalitiesOf 105 nationalities tracked in the Asy_D04 outcome analysis, 101 see asylum grant rates rise between initial decision and final outcome after appeal, by an average of 14.6 percentage points. Sri Lanka 26.9 percent initial to 60.1 percent after appeals. Iraq 30.9 percent to 57.0 percent. Russia 57.8 percent to 79.4 percent. The published initial grant rate of 39 percent for YE March 2026 understates the true protection rate by roughly 14 percentage points.Read more →ArticleRoutes27 May 2026Pakistani UK asylum decisions: 989 in 2022, 17,592 in 202517.8xIncrease in Pakistani UK asylum decisions 2022 to 2025Pakistani UK asylum decisions rose from 989 in 2022 to 17,592 in 2025, a 17.8 times increase in three years. Initial grant rate fell from 60 percent (2022) to 32 percent (2025) as throughput rose. The true grant rate after appeals stabilises at 49.8 percent. Pakistan is now the largest single source of UK asylum claims. The Pakistani political situation since 2023 (Imran Khan jailed, PTI persecuted, blasphemy laws, religious minority attacks) is the demand-side driver. The Home Office's own Country Policy and Information Note February 2026 documents the case-handling framework.Read more →ArticleRoutes27 May 2026Britain returns 39,000 people a year. Every political manifesto promises more.39,000UK total returns year ending March 2026The Home Office's headline return figure for year ending March 2026 is 39,000 people removed from the UK. The Reform UK 2024 manifesto wanted approximately 250,000 returns a year. Labour's Border Security Command produced a 13 percent rise in enforced removals and a 5 percent rise in voluntary returns, but no order-of-magnitude shift. The 39,000 ceiling is structural, produced by limited return-agreement coverage, tribunal pipeline length, and identity verification constraints.Read more →FindingRoutes14 April 2026Yemen 95%, India 3%. the grant rate gap the Home Office won't publish as a league table107Nationalities assessedFirst full grant rate league table from Home Office decision data. 107 nationalities. 11 have 75%+ initial grant rates (genuine refugees). 41 have under 25% (overwhelmingly refused). The gap between the two groups is the clearest evidence that the system processes genuine refugees and economic migrants alike, then treats them identically until decision.Read more →

Local pressure

FindingPressureUpdated 12 October 2026Blackpool's pressure is adult social care, not asylum or SEND growth4thBlackpool's rank for adult social care spend per adult, of 153 councilsCorrected 12 October 2026. On published figures Blackpool has the 4th highest adult social care spend per adult of 153 councils in England. Its growth in pupils with an EHC plan ranks 56th of 144, and its asylum support rate 94th of 361.Read more →AnalysisSocial careUpdated 12 October 2026Adult social care spend does not track demographic change: the councils compared are mostly one county£687Lancashire, spend per adult 2024-25, both comparison groupsCorrected 12 October 2026. This piece compared the five councils with the fastest White British decline (£612 per head) with the five with the slowest (£472), on figures that could not be traced to any published table. Eight of those ten are Lancashire districts, which run no adult social care. On DHSC's 2024-25 figures their care is all Lancashire County Council's, at £687 per adult, so the comparison does not hold.Read more →AnalysisCrimeUpdated 12 October 2026The 0.68 correlation between asylum dispersal and recorded crime published here is withdrawnWithdrawnCorrelation of 0.68 published April 2026, withdrawn 12 October 2026Corrected 12 October 2026. This piece reported a correlation of 0.68 between asylum dispersal and recorded crime across 25 areas. The crime figures it used could not be traced to ONS tables, so the correlation is not supported by published data and should not be cited.Read more →FindingPressureUpdated 12 October 2026Recomputed on published data: Manchester, Liverpool and Middlesbrough carry the highest combined pressure87.3Manchester, top composite pressure score of 285 English authorities, recomputed 12 October 2026Corrected 12 October 2026. The index is recomputed on four published domains, asylum support, demographic change, SEND growth and adult social care spend, across the 285 English authorities that hold all four. Manchester is first on 87.3, Liverpool second on 85.9 and Middlesbrough third on 83.6. Pendle is the highest in Lancashire, fourth.Read more →FindingSchoolsUpdated 12 October 2026Pupils with an EHC plan rose 65% in five years in England, with no link to demographic change+65.4%Pupils with an EHC plan, England, 2020/21 to 2025/26Corrected 12 October 2026. Pupils with an EHC plan in England rose by 65.4% in five years, from 325,618 to 538,547. This piece originally said 42% across 25 areas, on figures that could not be traced to DfE's tables. On the published figures, growth does not follow the pace of demographic change.Read more →AnalysisPressureUpdated 31 August 2026Ribble Valley and Hyndburn: nine miles apart, thirteen times the load13.7xHyndburn's asylum support rate against Ribble Valley's, 30 June 2026Ribble Valley has 24 people on asylum support, 3.6 per 10,000, ranking 230th of 361. Hyndburn, which it borders, has 426 at 49.0 per 10,000 and ranks 2nd in the UK. Same county, same region, same prime provider, thirteen times the rate across a boundary you can walk over.Read more →

Demographic change

FindingDemographicsUpdated 12 October 2026Blackburn projected minority White British by 2027. Three more Lancashire towns follow by 2050.2026-2028Blackburn WBI <50% yearBlackburn with Darwen: 56.9% White British in Census 2021. The projection crosses 50% in 2027, and on the model's own error the crossing sits between 2026 and 2028. Pendle follows around 2033-2035, Preston 2035-2039, Burnley 2048-2053. Four Lancashire towns crossing the same threshold within a generation.Read more →AnalysisDemographicsUpdated 12 October 2026What is really driving ethnic change: national trend, not local migration81%Change from national trendShift-share decomposition of 307 local authorities shows that 81% of Burnley's White British decline is explained by the national trend, not local immigration. Most areas are changing because the whole country is changing, not because of exceptional local migration.Read more →AnalysisDemographicsUpdated 7 September 2026The most deprived fifth of England records 25.5% of births. The least deprived records 15.4%.25.5%birth share, most deprived fifthONS 2024 births by IMD deprivation decile: the most deprived quintile records 25.5% of births, the least deprived 15.4%. That is a ratio of birth counts, not a fertility rate, because the published table carries no count of women aged 15-44 per decile. Deprivation and ethnicity are heavily confounded, so neither gradient can be read as the independent effect of the other.Read more →AnalysisDemographics14 April 2026Albanian asylum seekers don't change ethnic minority statistics. 90% identify as White89.7%Albanian → White OtherCensus 2021 data shows 89.7% of people identifying as Albanian classify themselves as White Other. Albanian asylum seekers (the second-largest claiming nationality with 57,360 decisions) reduce White British percentage but do NOT increase ethnic minority statistics. This distinction matters for accurate demographic analysis.Read more →AnalysisDemographics14 April 2026Arab population projected to triple in 20 councils by 2051. a group no model has tracked before319KArab population (Census 2021)Our model is the first to project the Arab ethnic group separately. Census 2021 counts 319,452 people identifying as Arab. previously hidden inside 'Other ethnic group'. In areas like Brent (5.3% to 12.9%) and Westminster (7.6% to 9.2%), the Arab population is projected to grow faster than most other groups.Read more →FindingDemographics14 April 2026White British fertility has fallen to 1.31. lower than any demographic model assumed1.31WBI TFR (ONS 2024)ONS 2024 births data reveals White British fertility has fallen to an empirical TFR of approximately 1.31. well below the 1.55 assumed by academic models and the 1.71 implied by Census 2021 child-woman ratios. Pakistani fertility remains at 2.52. The fertility gap between ethnic groups is wider than any published model uses.Read more →AnalysisDemographics14 April 2026Roma and Gypsy/Traveller populations now separately tracked. concentrated in specific boroughs162KRoma + Gypsy/Traveller combinedOur model separates Roma (98,785) and Gypsy/Irish Traveller (63,348) from the White Other category for the first time. Roma are concentrated in London boroughs (Hammersmith 0.81%, Haringey 0.76%). Gypsy/Travellers are concentrated in rural LAs (Maidstone 0.58%, Fenland 0.57%). These populations were previously invisible in demographic projections.Read more →FindingDemographics11 April 2026Academic ethnic projections over-predicted White British share in 95% of areas3.95ppNEWETHPOP MAE vs Census 2021 actuals (296 areas)The NEWETHPOP cohort-component model, the UK's most cited academic ethnic projection, over-predicted White British population share in 282 out of 296 local authorities. NEWETHPOP MAE against the Census 2021 actuals: 3.95pp across 296 areas. A previous version of this piece set that against a 1.71pp score for our own model and claimed we were 33 per cent more accurate. That comparison has been withdrawn: our 1.71pp came from a backcast fitted on the same two Censuses it was tested against, so it measured our own guardrails rather than forecast accuracy. On a genuine out-of-sample test, fitting 2001 to 2011 and forecasting 2021, our model scores 1.56pp.Read more →