UK immigration

813,000 people moved to the UK for a year or more in 2025, down from the peak of 1,469,000 in the year to March 2023. It is still higher than every year in the series before 2018. Averaged over the chart's own data, the 1990s ran at about 325,000 a year.

Net migration fell from its peak of 944,000 in the year to March 2023 to 171,000 in 2025: 171,000 is 18.1% of 944,000. The Migration Observatory puts the 2025 level below the levels of the 2010s, while gross arrivals remained high. The peak is a rolling year ending in March; the annual chart shows calendar years, so its highest point is lower.

Adding the chart's calendar-year values for 2021 to 2024 gives about 4.8 million long-term arrivals and about 2.5 million net migration; for 2022 to 2025 the totals are about 4.7 million and 2.2 million. The two measures answer different questions. Gross arrivals count every person who came, including British nationals returning and students who later left. Net migration counts what was added to the population once departures are subtracted.

Net migration was negative in every year from 1964 to 1978, in 1980 to 1982, and in 1988, 1992 and 1993. In each of those years immigration was below a third of its 2025 level, and more people left than arrived. British nationals have had negative net migration in every year since 1964 except 1985, so those years mainly reflect British emigration outweighing a smaller inflow.

The sustained rise is clearer after 1997 than at 1990. ONS identifies an increase in non-EU net migration after 1997. The Migration Observatory links the wider rise to a growing economy and low unemployment from the late 1990s, followed by a more open non-EU work-permit system and efforts to attract international students in the early 2000s. These changes developed over several years; the graph cannot attribute the turn to a single election or policy. The dashed line near 1991 instead marks a measurement change.

The earlier rise began before eastern European countries joined the EU. In 2004 the UK, unlike most existing members, opened its labour market immediately to workers from the new accession countries. This helped produce the subsequent increase in migration from eastern Europe. It explains a later step, not the beginning of the late-1990s rise.

The dip in net migration in 2020 reflects the disruption caused by COVID-19. It is an exceptional baseline for judging the subsequent rise.

The surge was driven by non-EU arrivals. The government restored a post-study work route, opened routes for people from Ukraine and Hong Kong, and made care workers eligible for work visas. Universities recruited overseas as their finances deteriorated, while care employers recruited to fill vacancies. The route chart below shows how much each contributed.

Work visas began falling in late 2023, especially in care. The Home Office says closer scrutiny of sponsoring employers and enforcement against non-compliance likely contributed to the initial decline. The decline began before the 2024 rule changes, so those changes cannot explain its start.

Restrictions on bringing family members for care workers and most international students then helped reduce arrivals. Further restrictions introduced in the second half of 2025 are not yet fully reflected in this chart.

Non-EU+ work-related arrivals fell by 47% in 2025, the main reason net migration continued falling. Overall emigration fell slightly that year, so the latest fall came from fewer arrivals rather than more departures. Departures had risen between 2022 and 2024, so the explanation differs across the decline.

The immigration line remains well above its earlier levels after dividing by the UK population. A larger country would receive more people at an unchanged arrival rate; this reading shows that arrivals also grew relative to the population.

At the peak, work, study, the Ukraine scheme and the BN(O) route for Hong Kongers all added large numbers at once, while EU+ net migration had already turned negative. The Ukraine scheme line jumps in 2022 and has fallen back to around zero since; the BN(O) line is smaller and steadier. A policy aimed only at work or study would not have touched the two humanitarian routes.

The Ukraine scheme's contribution peaked in 2022, after the humanitarian routes opened following Russia's full-scale invasion. Work and study kept rising into 2023 while the Ukraine line fell, so Ukraine explains part of the spike, not the wider surge.

Study was the largest reason for non-EU+ arrivals in 2025, at 47%, but its net contribution is much smaller than its arrivals because many students leave: 159,000 non-EU+ nationals who came to study emigrated in 2025, 57% of all non-EU+ emigration.

Asylum-related immigration was broadly stable at 88,000 in 2025, and few people who came for asylum leave, so its net contribution has stayed high while work and study fell. These estimates include people who claimed asylum after arriving on a visa as well as those arriving by small boat.

In 2025 net migration was 350,000 for non-EU+ nationals, negative 136,000 for British nationals and negative 42,000 for EU+ nationals. The British figure grew after 2022 mainly because fewer Britons returned; ONS says British emigration itself has been stable, and the British estimates are particularly uncertain.

Several changes increased non-EU arrivals together: the government restored a post-study work route, care workers became eligible for work visas in early 2022, and new routes admitted people from Hong Kong and Ukraine. Demand reinforced those policy changes: universities recruited more overseas students as their finances deteriorated, while social care employers recruited abroad to fill vacancies linked to low pay and poor working conditions.

627,000 of the 813,000 people who moved to the UK long-term in 2025 were non-EU+ nationals, down from 780,000 the year before but still the great majority.

EU+ arrivals remain positive, but more EU+ nationals are leaving than arriving: net migration was -42,000 in 2025 and has been negative in every rolling-year estimate since June 2022. Arrivals and net migration measure different things.

EU net migration began falling after the 2016 referendum, before free movement ended in January 2021. Recent departures include people who arrived during the earlier period of higher EU immigration.

In the year to mid-2025 the UK population grew by 227,700. Net international migration was 222,900 and natural change only 4,900, so migration supplied 98% of the growth (222,900 divided by 227,700). The Migration Observatory puts migration's share at 68% of growth from 2005 to 2024 and 99% since 2020.

Natural change has declined since 2011 and was negative in the year to mid-2023. Net migration has been the larger component in every year since mid-1999 except mid-2012.

ONS now assumes long-term net migration of 230,000 a year, down from 340,000 in its previous projections. On that basis deaths exceed births from mid-2026, and net migration is the only source of population growth over the next 25 years. The population peaks in the mid-2050s and then declines.

By mid-2049, the population would be about 72.4 million on the principal assumption, 78.9 million with high migration (455,000 a year), 68.8 million with low migration (105,000 a year) and 65.5 million with zero net migration. With zero net migration the population falls from mid-2025 onward.

On this chart's fixed-age measure there were about 302 people aged 65 and over per 1,000 aged 16 to 64 in 2024. By 2074 the ratio reaches about 457 with high migration, 525 on the principal assumption, 580 with low migration and 641 with zero net migration. Higher migration slows the rise but does not stop it, and it does so by making the population larger. On ONS's own pension-age measure, the principal projection goes from 280 in mid-2024 to 329 by mid-2049.

ONS says its long-term migration assumption "should not be viewed as a forecast". Each line shows what a migration level implies if it is sustained, not how likely it is.

In 2024 the UK received the fourth-largest number of new permanent-type migrants in the OECD, but per head of population it ranked 21st of the 37 countries shown, next to Norway, Greece and Finland. Luxembourg, Iceland, Switzerland and Ireland took the most relative to their size; Japan and Mexico the least. The OECD notes that UK inflows per head "remain modest despite the recent increase in inflows in absolute terms".

UK permanent-type migration fell by 41% between 2023 and 2024, the second-sharpest fall after New Zealand; the OECD links both to unusually high permanent labour migration in 2022–2023.

The OECD counts foreign nationals granted permanent-type status and does not subtract anyone leaving, so a country with heavy emigration can rank high here and low on net migration. No comparable cross-country net migration series was found: the UN and World Bank figures disagree with ONS by about half for the UK. Temporary migrants such as students are excluded.

The comparison starts with non-EU+ foreign nationals. Switching to EU+ or all arrivals changes the ranking: these are different questions, so a high total does not by itself establish a high non-EU+ inflow. This is gross arrivalsGross ArrivalsThe number of people moving into a country, before subtracting anyone leaving. In these UK charts, this means people moving to the UK for at least 12 months, including returning British citizens. It is the immigration line, not net migration.×: countries with substantial outward migration can have high arrivals without equally high net migrationNet MigrationImmigration minus emigration: the number moving in, less the number moving out. Net migration can fall because fewer people arrive, more people leave, or both. It does not count births or deaths.×.

A modest number of arrivals is a larger share of a small country's population. The nationality breakdown also differs sharply: EU+ nationals form a much larger share of arrivals in Luxembourg than in Malta. Nationality and previous residence are different measures, so this chart uses citizenship consistently.

Poland and Slovakia exclude Ukrainians under temporary protection from these migration statistics. Their low bars therefore cannot be read as evidence that they received very few people fleeing Ukraine. Differing coverage is one reason the chart is an approximate comparison rather than a precise league table.

For the eleven years to mid-2021 there were fewer than 200,000 work visas a year, partly because EU nationals needed no visa while free movement lasted. Health and Care Worker and Skilled Worker visas drove the rise to 613,627 in 2023, and together made up 76% of work visas that year. The total was 234,841 in the year to June 2026: 158,722 main applicants and 76,119 dependants.

The post-Brexit system cut the skill threshold for a work visa from degree level (RQF 6) to A-level equivalent (RQF 3), opening sponsored visas to mid-skilled occupations; care workers were added to the Health and Care Worker route separately, in February 2022. On 22 July 2025 the threshold went back to RQF 6, which the Migration Advisory Committee describes as ending the 2020 relaxation.

These routes (Global Talent, High Potential Individual, Innovator Founder, Scale-up and their predecessors) were under one in twenty-five work visas in every year shown, and their number in 2025 was below 2023's. The chart counts visas applied for abroad only, so it misses people who switch onto these routes from inside the UK.

The Migration Advisory Committee found that Health and Care Worker visa holders brought 1.21 dependants each between October 2022 and September 2023, against 0.77 for other Skilled Worker occupations: 173,900 dependants on that route in a year. The dependant line fell after care workers lost the right to bring family in spring 2024, but more slowly than main applicants, because people recruited before the change could still bring family from abroad. In the year to June 2026, 79% of Health and Care Worker visas went to dependants.

Main applicants on the route peaked in the third quarter of 2023 and dropped by four-fifths within two quarters. Dependants stayed high into early 2024, then fell by about two-thirds in the quarter after care workers lost the right to bring them. The government's technical annex puts the 2023-to-2024 fall in the whole route at 68%, to 110,800, and attributes it to Home Office scrutiny of employers, compliance action and the spring 2024 dependant rule.

Grants to sponsored care workers fell by about nine-tenths between the third quarter of 2023 and the first quarter of 2024. The Home Office attributes the start of the fall to closer scrutiny of health and social care sponsors and compliance action against employers, before the 2024 rule changes took effect. By the time the route closed to new overseas care recruits in July 2025, grants were already about a thousand a quarter; the government's own impact estimate for the closure was about 7,000 main applicants a year. The Home Office summary nonetheless describes the fall from 107,772 to 344 as following "the end to overseas recruitment of care workers in July 2025".

Health and Care Worker visas for nursing professionals fell 95%, from 22,858 to 1,125, between 2023 and the year to June 2026; the Department of Health and Social Care links this to the end of its centrally supported international recruitment programme and lower demand. Doctors fell less, and were 59% of the route's visas in the latest year.

Skills for Care puts the 2021/22 vacancy rate at 10.4%, the highest since its records began in 2012/13, and the 2025/26 rate at 6.1%, one point below 2018/19. It credits international recruitment with part of the fall, and changes in economic inactivity and fewer vacancies elsewhere in the economy with another part; it does not say how much each contributed.

About 105,000 people arrived from abroad and started direct care jobs in 2023/24, 50,000 in 2024/25 and 30,000 in 2025/26. Around 2,000 were recruited between April and June 2026, and the monthly vacancy rate stayed between 6.4% and 6.1% from March to August 2026. That is under a year of data after the route closed, so the effect of the closure on vacancies is not yet known.

The vacancy rate among employers with internationally recruited care staff fell from 9.9% in March 2022 to 5.6% in March 2026.

Posts filled by British nationals fell from 85% of the workforce in 2020/21 to 67% in 2025/26, about 130,000 fewer posts, 40,000 of them in the last year. Elsewhere in the same report, measured from 2021/22 across local authority and independent employers only, the fall is 83% to 65%, or about 107,000 posts. Skills for Care attributes the decline to fewer British people starting care jobs, not to more of them leaving. About 150,000 posts were held by people on Health and Care Worker visas in March 2026.

By December 2025 a quarter of jobs held by employees who registered for National Insurance with a non-EU nationality were in health and care, up from 22% in February 2022; 161,000 Skilled Worker visas had gone to non-EU care workers by the end of 2025.

The median care worker earned £12.60 an hour in March 2026, 39p above the National Living Wage. Workers with five or more years' experience earned about 12p an hour more than new starters, down from 33p in 2016. The chart shows the average premium over the minimum wage narrowing since 2022/23, as the minimum wage rose faster than care pay.

The committee that recommended adding care workers to the route said "the underlying cause of these workforce difficulties is the underfunding of the social care sector" and "immigration could not solve these workforce issues alone". Its main recommendation, a minimum of £1 an hour above the National Living Wage for publicly funded care, was not implemented. It also said low pay and visas tied to one employer left migrant care workers open to exploitation.

The Home Office revoked 1,948 worker sponsor licences between July 2024 and June 2025, against 937 the year before and 261 and 247 in the two years before that. It names adult social care among the sectors with the most abuse, and gives underpayment, helping people get round the rules and failing to provide promised work as reasons. The series records licences, not the workers employed, and has no sector or reason breakdown.

Registered Skilled Worker sponsors roughly quadrupled between 2019 and 2025. Measured against that larger base, revocations in 2025 were about the same share of sponsors as in 2014 to 2017, after very low rates during 2021 to 2023 while sponsor numbers were climbing fastest.

UK-educated joiners to the nursing register rose 11.9% in 2023-24 and 5.9% in 2024-25, while internationally educated joiners reached their 2023-24 peak and stayed high; internationally educated professionals were 49.4% of new joiners in 2023-24 and 26.4% in 2025-26. The two counts moved together, which does not by itself show whether one affects the other.

In 2025-26, 10,961 internationally educated professionals joined, against 20,669 the year before, and UK-educated joiners fell 4.9% to 30,581. Acceptances onto nursing degrees at English universities in 2025 were 18% below their 2021 peak. Two of the three largest sources of international joiners, Nigeria and Ghana, are on the list of countries where active recruitment is barred under ethical codes.

The Migration Advisory Committee's 2018 review found "no evidence that migration has had a negative impact on the training of the UK-born workforce" under EU free movement. The 2025 white paper reads the recent period the other way, from the two trends moving in opposite directions rather than from a study. The 2018 research covered EU free movement, before sponsorship and salary thresholds applied; see Immigration and Domestic Workforce Training.

In June 2025, 21.3% of hospital and community health staff with a recorded nationality were not British: 36.3% of doctors and 30.0% of nurses. The EU share rose in the early 2010s and stopped rising after the 2016 referendum; the recent rise is in staff with Asian and African nationalities.

Of 85,891 people who claimed asylum in the year to June 2026, 41% (34,999) had held a visa or other leave, slightly more than the 40% (33,949) who arrived by small boat. The Home Office notes that the rise coincided with the rise in visa grants up to 2023, and that a claim can come long after the visa was granted, during or after the period of leave.

In the year to June 2026, claims from people who last held a study visa fell 35% and claims from work-visa holders rose 16%; Graduate visa holders' claims more than doubled to 5,130 while claims from other work visas fell 9% to 9,021. Each route's claims are a small fraction of the visas it issued; rough ratios are in About this data. The dataset records the last leave held, not how the person first entered.

The UK (England and Wales) has the sixth-highest share of foreign-trained doctors among the 28 countries the OECD reports, about twice the OECD average of 20%. It admitted more foreign-trained doctors in 2023 than in any earlier year on record. Across the OECD, domestically trained doctors still made up most of the growth in physician numbers between 2010 and 2023.

The OECD average share of foreign-trained nurses was 8.8% in 2023; the UK ranks fourth. Foreign-trained nurses accounted for 83% of the increase in UK nurse numbers from 2010 to 2023. Registries differ in whether they count only practising nurses, and Germany counts nationality instead of place of training.

The median Graduate visa holder in work earned £1,750 a month in the first year (£21,000 a year), rising to £2,020 a month (£24,240) in the last three months; UK graduates earned a median £26,000 about 15 months after graduating. The MAC expected some gap because international graduates have less experience of the UK labour market. 19% earned at least £2,500 in their first full month, and 38% by the twelfth month. These are first-year figures only; nothing published follows the same people further.

69% of people who switched from the Graduate route to a sponsored work visa went into managerial, professional or associate professional jobs, against 75% of UK graduates and 42% of people who switched straight from a student visa. 49% of direct student switchers became care workers or senior care workers, against 20% of Graduate route switchers and 6% of UK graduates in the whole caring, leisure and other service group. Students are also staying longer than they used to: 41% of the 2020 study cohort still held leave five years later, against 17% to 18% for the 2012 to 2016 cohorts; see Student Migration and Subsequent Residence.

In pounds per person at 2023 prices, UK GDP per head was 40,125 in 2019 and 40,546 in 2025. It fell to 36,033 in 2020, recovered to 40,554 in 2022, then slipped to 40,126 in 2023 and 40,153 in 2024. Between 2007 and 2019 it had risen from 37,625 to 40,125, with the 2009 recession in between.

The 2019–2025 window contains the 2020 pandemic fall in output as well as other shocks. A series of GDP per head cannot by itself say how much of the stall, if any, is due to migration; that needs a model, and the OBR's scenarios are further down. No study attributing the stall among its causes was found.

Total GDP rose from 2,673,465 to 2,817,414 (£ million, 2023 prices) between 2019 and 2025, while output per person barely moved. A larger population raises total output; whether it raises output per person is the separate question the OBR scenarios address.

The World Bank series uses market exchange rates, so the gaps between countries' levels move with currencies. The ranking of growth since 2019 is in the bar chart below.

On the World Bank series, GDP per head rose more between 2019 and 2025 in the United States, Italy, Japan and France than in the UK; Canada grew less and Germany shrank slightly. The UK's stall was shared by other large economies, so on its own it is not evidence about migration. Net migration per head for these countries over the same years is not shown here (see Data still needed).

In the OBR's March 2024 scenarios, 200,000 more net migrants a year leave GDP per person "largely unchanged" if they are as productive as residents, 0.4 per cent lower if the capital stock does not adjust, and 0.8 per cent higher if nearly all adults arrive on work visas. The lower-migration scenario is the mirror image.

In March 2026 the OBR summarised the literature: the effects on real GDP per person "are relatively small in either a positive or negative direction". It adds that migration is likely to raise participation, because migrants are younger, but that the effect on productivity "is much more uncertain".

After the ONS revised up emigration by British nationals, the OBR cut its net migration forecast by around 60,000 people a year, lowering the adult population in 2030 by around 200,000. Because the extra emigrants were assumed to be younger and lower-paid, the change left potential output per person "broadly unchanged".

In the same scenarios total GDP in 2028-29 is 1.5 per cent higher with migrants as productive as residents, 1 per cent higher if capital does not adjust, and 2.2 per cent higher with very high participation.

With departmental spending left as planned, 200,000 more net migrants a year cut borrowing in 2028-29 by £19.9 billion, and 200,000 fewer raise it by the same amount. If spending is raised to keep services per person unchanged — about £6.1 billion extra in the higher scenario — the effect is £13.1 billion either way.

The Migration Observatory notes that short-term forecasts "are particularly likely to find a positive impact from migration, because they do not account for the costs that come later as migrants age". Over fifty years the OBR's 2023 projection had higher net migration lowering the primary deficit by 1.1% of GDP by 2072/73, without stopping debt rising towards 300 percent of GDP. In 2013 the OBR described higher migration as "delaying some of the fiscal challenges of an ageing population rather than a way of resolving them permanently".

In its 2024 long-run scenarios, a shorter average stay lowered debt by about 23 per cent of GDP by 2073-74 and higher migrant earnings by about 40 per cent of GDP; in neither case did debt stop rising. Migrants earning half the average instead raised debt to 351 per cent of GDP, 76 per cent of GDP above the baseline.

Oxford Economics, expecting net migration to stay low, estimated in February 2026 that if the OBR followed its forecast, borrowing would be £19bn (0.5% of GDP) higher. That is a private forecast, not an OBR figure.

The Migration Advisory Committee models the 329,200 people who arrived on the Skilled Worker route in 2022/23, including Health and Care visas and dependants, at a net £47.1bn over their lifetimes, or +£143,000 each. Main applicants outside health and care average +£689,000; Health and Care main applicants +£54,000, split between care workers at −£36,000 and other health and care occupations at +£166,000. Adult dependants average +£3,000 on the Skilled Worker route and −£67,000 on the Health and Care route.

The MAC calls the route "clearly fiscally positive", which it says is "almost inevitable" given the salary thresholds.

On the same method the whole UK resident population has a mean lifetime contribution of −£39,000 and a median of −£145,000. Residents with the same starting ages as Skilled Worker main applicants average +£117,000 (median −£47,000). Care workers are "broadly similar to the UK median for the age group". Health and Care dependants fall short of their own age-matched comparator of +£84,000.

Partners of British citizens and settled residents arriving on the Partner route in 2022/23 are modelled at −£109,000 each.

The top 10% of earners outside health and care (salary at least £131,000) account for 39% of that group's contribution. Main applicants who stay are modelled at +£931,000 and those who leave at +£174,000. The MAC concludes that for lower earners temporary migration is fiscally better than permanent, and for higher earners the reverse.

The MAC uses a 3.0% real discount rate, "broadly in line with HMT Green Book". Without discounting, the future contribution of a 25-year-old UK resident would be −£186,000 rather than +£184,000, because old-age costs fall late in life. Any lifetime figure has to state its discount rate to be compared with another.

The MAC has not estimated lifetime effects for refugees, students or other routes; the Migration Observatory expects refugees to be more negative because of lower employment and greater need for support.

In the OBR's stylised profiles, a migrant arriving at 25 on average pay is a net contributor "by their second year", while a resident of the same age becomes one only in their 40s, because the migrant's childhood education and health were paid for elsewhere. A migrant earning 30 per cent above average stays a net benefit even living to 100; one on half the average falls behind the average resident in their early 40s.

The Centre for Policy Studies read these OBR profiles to age 82 for its 2025 estimate — for example, about £405,000 for an average-wage migrant arriving at 25 — and applied them to projected settlement numbers.

Every line falls after state pension age as health, care and pension costs mount. The correction of November 2024 made the low-wage migrant "significantly less fiscally beneficial, and substantially so at older ages". The OBR's representative migrant has no dependants; the OBR says it chose "a more typical migrant on a work visa and no dependants".

In 2023/24 median annual earnings were £56,600 for Skilled Worker main applicants, £47,900 for Senior or Specialist Workers and £30,900 for Health and Care main applicants; care workers and home carers earned £23,100. Dependants earned £30,200 (Skilled Worker) and £22,100 (Health and Care); family-route visa holders £20,200. The ONS median for full-time employees was £37,430 in April 2024, a different measure shown only for reference.

At least 75% of Skilled Worker main applicants and at least 45% of their dependants had PAYE earnings; the Home Office counts these as minimums because self-employment is not included. It also advises caution in comparing routes, because work eligibility differs.

HMRC payroll data show each non-EU cohort from 2021 to 2023 earning around 95% of the UK median after one full year, falling to 90% for the 2024 cohort. The 2021 and 2023 entrants passed the overall median within two years; 2015 entrants took about six. Entry-year pay fell sharply in 2025. The Migration Observatory says faster catch-up may reflect lower earners leaving the payroll, and the data mix full-time and part-time work.

Of non-EU migrants first on payroll in 2015, 60% were still there in 2025; of EU migrants, 45%. These data are by nationality, not by visa route.

The MAC found migration "is not a major determinate of the wages of UK-born workers", with some evidence of a negative effect on lower-skilled and a positive one on higher-skilled workers. Its update implies EU immigration over 1993–2017 lowered UK-born wages by 5.2 per cent at the 5th percentile and raised them by 1.6 per cent at the median and by 4.4 and 3.1 per cent at the 90th and 95th percentiles, as short-run estimates.

Expressed per year, a 2022 study found immigration over 1994–2016 cut the hourly wage at the 5th percentile by about half a penny a year and raised it by 1.7p a year at the 90th. In low-wage service jobs, a one-point rise in the migrant share lowered average occupational wages by about 0.2%. An earlier estimate for semi- and unskilled services, a 1.9 per cent fall per 10-point rise, revised down an even earlier 5.2 per cent.

One study found the main impact of immigration from 1975 to 2005 fell on migrants already in the UK. The Migration Observatory knows of no study of wage effects under the post-Brexit system.

The Office for Students found 35.8 per cent of institutions in deficit in 2024-25, forecast to rise to 42.7 per cent in 2025-26. It lists "an overreliance on fee income from international students" among the risks, alongside over-optimistic recruitment forecasts.

In 2024/25 UK-domiciled students paid 53.7% of higher-education course fees at UK providers; the rest came from students domiciled elsewhere. Fees are only part of university income, so this is not a share of income.

England added 208,600 net dwellings in 2024-25. The ONS projects an average of 242,000 additional households a year over 2022–2032, up from 158,000 a year in its 2018-based projections for the same period. A household is not a dwelling — some homes stand empty or are second homes — so the two numbers are not an exact measure of shortfall, and the ONS says its projections are not a forecast of how many homes should be built.

Net additions fell to 130,610 in 2012-13 after the financial crisis, rose to 248,590 in 2019-20, and have fallen in each year since 2021-22. Net migration peaked at 944,000 in the year to March 2023, so the latest fall in building came while net migration was at or near its peak.

Annual private rent growth peaked in March 2024 and stayed near that level through 2024, then fell through 2025 to 3.8% in the year to September 2026 — less than half its peak. Over roughly the same period net migration fell from 944,000 in the year to March 2023 to 171,000 in 2025. Rent growth turned about a year after net migration did.

Rent levels have not fallen. The average UK private rent was £1,406 in September 2026, 3.8% higher than a year earlier. Growth was fastest in the North West (5.7%) and slowest in London and the South East (3.5%).

There were 112,200 fewer pupils in England's schools in January 2026 than a year earlier. DfE attributes the fall to fewer births since a 2012 peak and projects primary and secondary numbers to keep falling to 2030. In the state-funded chart, primary numbers peaked in 2019 and the total in 2024.

England had 30,183 fully qualified full-time-equivalent GPs in June 2026, 1,381 more than a year earlier and 2,382 more than in December 2019. There is no GP chart here: the official monthly series, which starts in September 2015, could not be downloaded.

The number of GPs rising does not show that GPs kept pace with patients. Patients per GP varies by area, from 2,566 in North East London to 1,798 in Bristol, North Somerset and South Gloucestershire after adjusting for need. The national trend in patients per GP is listed under data still needed.

Social rent completions fell from 39,562 in 2010-11 to 5,827 in 2016-17, and were 12,267 in 2024-25 — less than a third of the 2010-11 level. Total affordable completions held up because the new affordable rent tenure, at higher rents, replaced much of social rent from 2011-12.

Local authorities completed 10,480 affordable homes in 2024-25, the most since the series began in 1991-92. The release described 12,198 social rent homes as the most since 2012-13; the June 2026 revision puts the figure at 12,267.

Real school funding per pupil was flat at around £7,700 from 2010-11 to 2015-16, fell 3.6% to 2018-19, then rose, to £8,390 in 2026-27 in 2025-26 prices. Real health spending per person was almost £3,500 in 2022, almost double its level at the turn of the century, though still below the average of 19 advanced economies.

The OBR projects real day-to-day departmental spending per person to reach its highest level in two decades outside the pandemic, around £7,800 in 2028-29. It attributes the rise from about £7,700 in the June 2025 Spending Review plans to a lower population forecast, "largely due to lower projected net migration".

UK government investment as a share of GDP was below the OECD median in every year from 1995 to 2024. In the decade to 2017, the OBR ranked the UK between 23rd and 27th of 30 OECD countries. UK investment has been higher as a share of GDP since 2020 than in most of the 2010s, but the OECD median rose too.

In 2021–2024 the UK invested 0.4–0.5% of GDP a year in health capital, below the OECD average, after a clear decline through the mid-2010s and a rise since.

The OBR notes that much public investment is fixed in cash terms for each Spending Review period, so it "does not respond immediately to changes in the size of the population"; whether it catches up later is uncertain.

The ONS main projection adds about 2.4 million households in England between 2022 and 2032. Its zero net migration variant has 1,287,752 fewer by 2032, so in the projection migration accounts for just over half the growth; the rest comes from the existing population ageing and forming smaller households.

These are projections of past demographic trends, not forecasts of need, and they leave out local economic and housing policy. Net migration has since fallen well below the levels of 2022–2024, so the migration share will depend on which variant turns out closest.

In housing markets most like the UK's, the median study finds a one-point rise in the immigrant share raises house prices by 0.65%. Applied to the UK's foreign-born share, the review says this "can explain approximately 4-6%" of the three-decade rise in real house prices. It calls immigration "unlikely to be a dominant driver" of UK housing costs, while finding the effect positive.

Across all studies the median rent effect is 0.64%, against 0.36% for house prices; in UK-like countries it is 0.81% for rents and 0.65% for prices. The review produces no rent equivalent of the 4–6% figure.

Prices respond more where planning limits construction. Studies of small areas find weaker or negative effects because existing residents move away; in the UK, Sá (2015) finds a 1.7% fall in local house prices per point of immigrant share. Migration also supplies the building workforce in London, where migrants are around 46% of construction employment.

Around a third of people in privately rented homes in England and Wales were born abroad in 2021, against around 10% of owner-occupiers. Home ownership was 17% for migrants who had arrived in the past five years and 68% for those here more than 20 years, against 67% for the UK-born.

Lead tenants with a non-UK nationality took 6% of new social lettings in England in 2007/08 and 10% in 2022/23. People on work, study and family visas are not eligible, because of the no recourse to public funds condition, and at the 2021 Census non-UK passport holders were slightly under-represented in social housing. The MAC's 2018 review judged that, with little social housing being built, new lettings to migrants are "inevitably at the expense of other potential tenants".

On the OECD's own national accounts, the UK ranked 27th of the 36 members with 2024 data, close to Germany and Denmark and above Spain, Portugal and Ireland.

Grants peaked in 2010 at almost 250,000 and then averaged 111,000 a year from 2011 to 2024. The Home Office attributes the 2010 spike to 82,295 discretionary grants made mainly to clear a backlog of unresolved cases.

The fall from 2010 to 2016 was partly due to the July 2012 change that raised the family route's qualifying period from two to five years, delaying grants that would otherwise have come earlier. The Home Office's cohort data show people arriving on family routes after the change typically took longer to settle.

Grants have risen since 2017, mostly to people whose last leave was on a work route: 80,409 in the year to June 2026, up 36%. Grants to Hong Kong BN(O) visa holders, whose route opened in January 2021, rose to 38,373 as the first cohorts completed five years. Grants to refugees fell 40% to 22,799, reflecting fewer protection decisions during the pandemic in 2020.

The Home Office forecasts that between 1.3 million and 2.2 million people will settle in 2026–2030 (central estimate 1.6m), peaking at 359,000 to 620,000 in 2028 (central estimate 450,000). It assumes people settle at historical rates; the range reflects uncertainty about that. It names three drivers: 154,000 visas granted to Hong Kong BN(O) holders in 2021–22 (an extra 92,000 applications in 2026 and 54,000 in 2027); 616,000 Health and Care entry visas in 2022–24 (384,000 extra applications in 2027–29, peaking at 210,000 in 2028); and 344,000 Skilled Worker entry visas in 2022–24 (around 222,000 applications). The forecast is under current rules, before any of the proposed changes, and excludes the EU Settlement Scheme.

A Centre for Policy Studies projection covering visas issued from January 2021 to June 2024 found more than 801,000 recent migrants would receive indefinite leave in its base case, and as many as 1.24 million in an alternative. It put the lifetime net fiscal cost at £234 billion. After the Office for Budget Responsibility revised the data it used, the CPS said the cost estimates "should no longer be used"; it stands by the visa analysis and the numbers projected to settle. The £234 billion was a lifetime net cost of people settling, not a benefits bill or a saving from removing existing settlement.

The Windrush Lessons Learned Review found the scandal "foreseeable and avoidable": the government removed and detained people, and denied others work, housing and services, even though they were lawfully entitled to them. The Home Office identified 164 people in the UK before 1973 who had been detained or removed since 2002. By August 2026 the compensation scheme had paid £133,475,539.07 on 4,221 claims; of 11,690 final decisions, 4,009 were value offers and 6,741 found zero entitlement. The Windrush victims were entitled to stay under the law as it stood; the current proposal leaves Windrush grants out of scope.

Just under 4.6 million settled-status grants have been made under the EU Settlement Scheme since 2018. The Migration Observatory estimates that 621,000 to 820,000 non-EU citizens held indefinite leave at the end of 2024 (the summary of the same piece gives the lower figure as 622,000). EU scheme status is outside the government's proposed reforms, under the Withdrawal Agreement.

The non-EU estimate combines the 2021 Census for England and Wales with visa data. It leaves out Scotland and Northern Ireland, and some people counted will have left the UK or become British citizens since.

Around 2.2 million people — 53% of everyone on a temporary visa at the end of 2024 — held visas with a five-year path to settlement. About 4.2 million held a temporary visa in all, around 3.6 million of whom had arrived since 2021. Not all will apply: some will leave first. The Migration Observatory puts the number of children on a path at 430,000 to 520,000.

HMRC data for 2023/24 show around half of adult Skilled Worker dependants were not working, and a third of Health and Care dependants. About a third of adults on family visas were not working by their fifth year, around 45% of refugees, and 30% of adult BN(O) visa holders. On current circumstances, hundreds of thousands would fail the proposed employment requirement.

Of the 344,058 people who began a migrant journey in 2020, 55% still held leave five years later. Of those who came on a work route, 30% had indefinite leave and 28% were still on temporary leave. On family routes, 86% still held leave and 46% had indefinite leave. Less than 1% of those who came to study had settled within five years. Most people granted protection go on to settle: 89% of those who claimed asylum in 2015 had indefinite leave within ten years.

Among people who started on a work route in 2015, 20% had settled after ten years — fewer than the 30% of the 2020 work cohort who had settled after five. Settlement grants recorded in the migrant-journey data fell 9% to 153,204 in 2025, which the Home Office attributes to fewer arrivals in 2020.

No published series reports the employment, earnings or benefit receipt of people with indefinite leave by route. The closest measure is Universal Credit by recorded status: 2.7% of people on Universal Credit in July 2025 were in the indefinite-leave group (not the EU scheme), against 9.6% in the EU Settlement Scheme group. The command paper cites the Department for Work and Pensions as reporting 213,000 migrants with settlement claiming Universal Credit in June 2025. The department's employment rate for the indefinite-leave group on Universal Credit was 33%, the same as for British claimants.

Status is recorded at the claim and the department is not usually told when someone becomes a citizen, so the indefinite-leave group includes some British citizens; the Migration Observatory says its estimates of holders cannot be compared with it. These are "statistics in development". Dividing the 213,000 claimants by the 621,000 to 820,000 estimated holders gives no reliable claim rate, for the same reason.

In England in 2024/25, 89% of lead tenants in new social lettings were UK nationals, 4% were from the European Economic Area and 8% from outside it. The non-EEA share was around 3–4% from 2008/09 to 2020/21, rose to 9% in 2023/24, then fell to 8%. Whether a tenant holds indefinite leave is not collected for new lets.

The fee for indefinite leave is £3,226 against a Home Office estimated unit cost of £310. A Skilled Worker visa of over three years from overseas is £1,618 against £153. Across UK Visas and Immigration, fees recovered 256% of cost in 2024/25. Not every fee is above cost: a short visit visa is £135 against £159, and Global Talent with an approval letter £205 against £488.

Most adults pay a surcharge of £1,035 per year of their visa. The Home Office sets it on "how much healthcare an 'average' Health Charge payer is expected to use", not on each person's use. A Lords committee found the calculation "unconvincing" and slightly high: £1,036 per person where it should have been £1,013. The surcharge is paid on top of tax and national insurance. For a single Skilled Worker, fees before settlement come to around £9,900 over five years and would be £16,900 over ten, mostly because the surcharge is paid every year.

People on temporary visas are usually barred from most benefits until they settle. The 91 councils that reported to the NRPF Network spent £94 million supporting households with no recourse in 2024/25. No official count exists of how many people, or how many children, are subject to the condition; the Home Affairs Committee expects longer routes to increase child poverty.

The Royal Society's 2025 comparison puts the UK second to Switzerland's £3,871 at nominal exchange rates, and first after adjusting for purchasing power. UK settlement costs were 336% higher than the average of the other European countries.

In September 2025, 58% opposed removing indefinite leave from people who already hold it and 29% supported it. Even among those who supported ending new grants, only 41% backed removing it from current holders. Reform UK voters supported removal by 54%; Conservative voters split 44% to 42%; Labour, Liberal Democrat and Green voters opposed it by 71-80%.

Ending new grants of indefinite leave was supported by 44% and opposed by 43%. Majorities wanted to keep permanent residence for people in the UK over ten years, those in full-time work, NHS workers, spouses of British citizens and business founders. This is one poll from one pollster.

In England and Wales, 4.3% of residents were born abroad in 1951, 8.9% in 2001, 13.4% in 2011 and 16.8% in 2021, about one person in six. The largest single increase between the last two censuses was in people born in Romania, after work restrictions on Romanian citizens were lifted in 2014.

The share includes people who arrived decades ago, many now British citizens, and it falls when people born abroad leave or die. It says nothing about how long people have been here, which is what the language and identity figures below are about.

From 2004 to 2022 the employment rate of the combined Pakistani and Bangladeshi group rose from 44% to 61%, the largest increase of any group, but it remained the lowest. The Indian rate overtook the White British rate over the same period.

The overall rate also rose over the period, so part of each group's rise is the general recovery after 2010 rather than anything specific to that group.

In 2019, immigrants from the Middle East, North Africa, Pakistan, Afghanistan and Turkey (the ministry's "MENAPT" group) cost Danish public finances DKK 79 thousand each on average, while immigrants from Western countries contributed DKK 52 thousand each, and people of Danish origin DKK 22 thousand. Immigrants and descendants together cost DKK 16 billion, about DKK 21 thousand per person.

The average net cost of MENAPT immigrants fell by about DKK 37 thousand a person between 2014 and 2019. The ministry attributes about half the fall for non-Western immigrants to stronger attachment to the labour market, during years when employment was rising for everyone. Had MENAPT immigrants aged 25 to 64 worked at the Danish-origin rate, their 2019 contribution would have been about DKK 12 billion higher.

Immigrants who came as refugees cost about DKK 95 thousand each in 2019, while those who came to work contributed about DKK 120 thousand. MENAPT immigrants came mainly as refugees or family members, Western immigrants mainly to work or study. With the Western mix of reasons for migrating, MENAPT immigrants' average would have been about DKK 55 thousand higher, still negative.

Descendants of every origin, including Western, show net costs because most are children in school. At the Danish-origin age profile, descendants' total contribution would be about DKK 13 billion higher and immigrants' about DKK 28 billion lower.

In 2024, 66.1% of Chinese state-school pupils went to university at 18, and 29.8% of White pupils, the lowest rate. Every group's entry rate rose between 2006 and 2024. Research cited by the publisher finds almost all minority groups more likely to go to university by 19 than White British pupils, even after allowing for prior attainment and family background.

81% of adults in England agree their local area is a place where people from different backgrounds get on well together, in a narrow range since 2018/19; the 2024/25 figure is a small but statistically significant fall from 81.4% to 80.6%.

40% say many people in their neighbourhood can be trusted, within a 40% to 42% band since 2016/17 but down from 48% in 2013/14, the first year of the current survey method.

63% of adults in the least deprived tenth of areas say many neighbours can be trusted, against 18% in the most deprived tenth. Trust is also lower among most minority religious groups, including 25% of Muslims.

Most studies find more diverse communities report less trust, and a 2005 British survey linked rising diversity to lower neighbourhood trust. Other British studies find no strong erosion once deprivation is taken into account, and a London study found diversity linked to higher perceived cohesion, with segregation within neighbourhoods linked to lower cohesion.

Between 2011 and 2021 the foreign-born share rose by about 10 percentage points in Barking and Dagenham and about 9 in Watford, Havering and Thurrock, against 3.4 points for England and Wales as a whole. Inner London boroughs that were already the most foreign-born changed little.

EU-born residents accounted for more than half of the rise in 6 of the 20 places: Boston, where they were almost all of it, Peterborough, Thurrock, Luton, Barking and Dagenham, and Watford. In the other 14, including Hillingdon, Dartford, Swindon, Sandwell and Sutton, people born outside the EU accounted for half or more. Across England and Wales the two contributed about equally.

The same Census tables split the non-EU-born by continent. In Hillingdon, Swindon, Sandwell and Leicester most of the non-EU rise was in people born in the Middle East and Asia; in Dartford and Salford a large part was in people born in Africa. The Census groups do not separate Muslim-majority countries, so these figures cannot say how much of the change was from any particular culture or religion.

11% of people who arrived between 2011 and 2021 said they could not speak English well, against 4% of those who arrived before 1971. Across the whole population, 1.5% (880,000) could not speak English well and 0.3% (161,000) could not speak it at all.

84% of residents who arrived before 1971 had English as their main language, against 37% of 2011–21 arrivals; most recent arrivals who do not have English as a main language say they speak it well or very well. The share who cannot speak English at all is around 1 to 2% in every arrival period.

94% of non-EU-born residents who arrived before 1951 described their national identity as UK-only, against 81% of 1981–90 arrivals and 20% of 2011–21 arrivals.

In 2022, 75% of Pakistani and Bangladeshi men were in work and 46% of women. The publisher notes that the Commission on Race and Ethnic Disparities cited cultural and religious reasons for this difference.

At the 2021 Census, 51.4% of Muslims aged 16 to 64 were in work against 70.9% of the population. The ONS attributes the gap to high shares of students and of people looking after home or family. 28.7% of Muslim women were looking after home or family, against 3.9% of Muslim men.

25.3% of Muslims had no qualifications, against 18.2% of the population, but 32.3% had a degree-level qualification, close to the 33.8% overall.

An independent study estimates that, over a lifetime and including the second generation, labour migrants contribute €125,000 each on average, while study migrants cost €75,000, family migrants €275,000 and asylum migrants €475,000.

Western immigrants contribute €25,000 on average and non-Western immigrants cost nearly €275,000; Morocco, the Horn of Africa and Sudan sit at a net cost of "€550,000 to €600.000" per immigrant, in the study's own words. The authors find that where the first generation is a large net cost, the second generation is usually also negative or at best budget-neutral.

Statistics Norway modelled a one-off increase in immigration in 2015. Over 2015–2100 each extra immigrant from the rest of the world (R3) adds NOK 4.1 million of net public spending, while Western immigrants (R1) reduce it by NOK 0.8 million and Eastern EU immigrants (R2) add 0.8 million.

In 2022/23 Chinese (65.5) and Indian (59.4) pupils had the highest average Attainment 8 scores, and Bangladeshi, Black African and Pakistani pupils also scored above White British pupils (44.9). Gypsy and Roma pupils (20.3), Irish traveller pupils (26.8) and Mixed White and Black Caribbean pupils (39.1) had the lowest, with Black Caribbean pupils also below White British. The Education Policy Institute found most ethnic groups achieved higher GCSE grades than White British pupils in 2023.

Among pupils eligible for free school meals, White British pupils scored lower than every group except Gypsy, Roma and Irish traveller pupils. Every group scores lower when eligible for free school meals.

In a 2024 poll, 28% of British Muslims said it would be undesirable to outlaw homosexuality, against 62% of the public. In a 2015 poll in areas where at least a fifth of residents were Muslim, 52% of Muslims disagreed that homosexuality should be legal and 18% agreed; 39% agreed that wives should always obey their husbands.

In the 2015 poll, 86% of Muslims felt a strong sense of belonging to Britain, against 83% of the national control sample.

The one academic study comparing generations, from 2009, found young Muslims more liberal than their elders on gay marriage, with 74 percent supporting same-sex marriage or civil unions, but the most conservative of any group on the morality of homosexuality. No repeated survey tracks these attitudes across generations.

On 30 June 2026, 57% of the 275 people in custody for terrorism offences in Great Britain were categorised as Islamist extremists, 29% as extreme right-wing and 14% as other ideologies. 82% gave their nationality as British, and 60% gave their religion as Muslim, against 6.5% of the population of England and Wales in the 2021 Census. Nationality is self-declared; the statistics do not record country of birth.

Since comparable records began in 2021 the Islamist share has fallen from 70% to 57% and the extreme right-wing share has risen from 22% to 29%; the number of Islamist prisoners has stayed at about the same level.

MI5 has said its counter-terrorism work splits roughly 75% Islamist extremist and 25% extreme right-wing. This comes from a secondary summary of the Director General's October 2024 threat update; MI5's own site refused automated access, so the figure has not been checked against MI5's text.

There were 237 terrorism-related arrests in the year to June 2026 not linked to Palestine Action, against a 10-year median of 238. Of these, where ethnic appearance was recorded, 56% were White and 27% Asian.

Since September 2001, 73% of people arrested for terrorism-related activity have described themselves as British or British dual nationals.

3,630 (94%) of the year's arrests were for supporting Palestine Action after it was proscribed on 5 July 2025. A total that includes them is not comparable with earlier years, which is why the first chart leaves them out.

The Home Office says spikes in racially or religiously aggravated offences "coincided with" events: the EU referendum, the 2017 terrorist attacks, the start of the Israel–Hamas conflict, the disorder after the Southport murders, protests from May to October 2025, and the start of the US–Iran conflict. August 2024 had 10,158 offences, the highest month on record. The Home Office does not test what caused the spikes, and the published statistics do not separate the effect of the events themselves from that of the commentary or protests around them.

The Home Office says police data are heavily affected by improvements in crime recording, so they are limited for judging longer-term trends.

The National Audit found "disproportionate numbers of men from Asian ethnic backgrounds" among suspects in Greater Manchester, West Yorkshire and the National Crime Agency's Rotherham investigation. In Greater Manchester, 52% to 54% of suspects in exploitation investigations were Asian, against 20.9% of residents, with ethnicity recorded for every suspect. In West Yorkshire, 35% of suspects were Asian and 34% White, against 16% and 77% of residents, with 24% of suspects' ethnicity not stated. In Operation Stovewood, nearly two-thirds of 323 suspects and of 42 people convicted were Pakistani, against 4% of Rotherham's population.

Among 908 suspects in all open child sexual abuse cases in Greater Manchester, 44% were White, 16% Asian and 32% unknown; the audit says this is "much closer to that of the local population".

The audit says the national data cannot support "any accurate assessment". Re-based to include suspects whose ethnicity was not declared, the national group-based dataset is 28% White, 2% Asian and 66% not declared; the audit calls presenting it without the unknowns "misleading". So neither "most offenders are Asian" nor "most offenders are White" can be stated for England and Wales from these data. Asian people were 9.3% and Pakistani people 2.7% of the population of England and Wales in 2021.

The national police datasets recorded ethnicity for 31% of perpetrators of all child sexual abuse and 34% of group-based suspects in 2023, so 69% and 66% were unrecorded. The Independent Inquiry into Child Sexual Abuse found "widespread failures" to record ethnicity in all six areas it studied — 86 percent of offenders unrecorded in Tower Hamlets' 2019 profile — and concluded it was "impossible to know" whether any ethnic group is over-represented. Telford still had "a high proportion of missing data" two years after its inquiry.

The inquiry into child sexual abuse said it was "unclear whether a misplaced sense of political correctness or the sheer complexity of the problem" had inhibited data collection. The National Audit found organisations "avoiding the topic altogether for fear of appearing racist, raising community tensions or causing community cohesion problems".

The police recorded 102,878 child sexual abuse offences in 2024, of which around 60% were contact offences; about 17,100 contact offences were flagged as exploitation, and the national group-based dataset found around 700 group-based offences in 2023. The audit estimates that around 500,000 children a year experience child sexual abuse of any kind, most of it never reported. In the 2019 crime survey, 7.5% of adults aged 18 to 74 had been sexually abused before 16; the abuser was most often a friend or acquaintance (37%) and in 30% of cases a stranger, and 7% of victims of rape or assault by penetration told the police. The survey does not record abusers' ethnicity.

The audit says the group-based figure is "highly unlikely" to reflect the true scale, because the abuse is under-reported and definitions are applied inconsistently.

Non-UK citizens were 13.1% of people cautioned or convicted in 2024 and 12.1% of prisoners in June 2025, against an estimated 13.5% of adults; the share would be lower after adjusting for age. They were 18.7% of entries into prison in 2025, because they are less often granted bail.

Among the 50 largest nationalities, Albanians, Iraqis and Afghans were over-represented in convictions and Indians, Pakistanis, French and Chinese under-represented. These are nationality figures: naturalised citizens count as British and country of birth is not recorded, so they say nothing about British-born people of any ethnicity.

After correcting for age, men born in Denmark to non-Western immigrant parents had a Criminal Code conviction index of 262 in 2023 (100 is the average), about 2.6 times what their numbers and ages would predict, against 158 for non-Western immigrants and 88 for men of Danish origin. The research unit also says this group's conviction rate has fallen over time, and that most men convicted are of Danish origin. The index does not correct for income or education, and no comparable British figures exist.

In 2024/25, 21% of referrals (1,798) were for extreme right-wing concerns and 10% (870) for Islamist extremism; of cases adopted for Channel support, 42% were extreme right-wing and 15% Islamist. Most referrals (56%) recorded no ideology. Referrals record what others report, which is a different measure from the prisoner and MI5 figures above.