UK Immigration and Public Services
The UK Home Office has released a white paper on changes to the UK immigration system. It opens with a rather provocative set of claims:
“In 2023, under the previous government, inward migration exploded to over a million people a year – four times the level compared with 2019. This was a political choice that was never put before the British people. In fact, quite the opposite – the previous government repeatedly promised inward migration would be brought under control. Instead, Britain became a one-nation experiment in open borders.
The damage this has done to our country is incalculable. Public services and housing access have been placed under too much pressure. Our economy has been distorted by perverse incentives to import workers rather than invest in our own skills.”
It seems worth evaluating some of these claims. Despite the white paper’s claim that such damages are incalculable, these are empirical questions.1 So: how have changes in immigration in the UK affected public services in the past, and how are they likely to do so in the future?
First, some context:
Britain tracks net migration.
This spiked in the year ending 2023, though it has since declined a bit.2 Who are these migrants? And has it caused damage to Britain?
What do we know about UK migration?
How many migrants are there (and is that a lot)?
The white paper says that “Britain became a one-nation experiment in open borders”. This seems a bit of an exaggeration. At peak, the net population increase in the UK (in the year ending 2023) was 906,000. Since the UK had a population of 68.3 million, immigration alone caused a population increase of 1.3%.
This is high by comparison to recent UK migration numbers; in general, the UK has seen fewer than 500,000 new residents per year. It is also relatively high in historical terms; it is on a similar level to that seen in the US from 1880 to 1914.
However, it is not the highest in the rich world. Canada, Australia and Germany had higher migration per capita in 2022; both Canada and Australia also did in 2023. Indeed, Canada has expanded immigration much more dramatically than the UK. In 2023, Canada’s population grew by 3.2%, 98% of which was from immigration; this is >2x the UK rate.
Thus: yes, the UK has had high immigration in recent years. It is, however, not accurate to say that the UK has had open borders during this period.
What are migrants doing in the UK?
In general, the largest single category of visa are study visas, at around 410,000 per year. Most of the study visas are for masters level courses, rather than undergraduate courses. This is followed by work visas, though that number has varied substantially per year as immigration restrictions have changed.3
Despite the near endless press coverage on asylum seekers, the number of refugee/humanitarian visas granted is small by comparison.4 Most people who move to the UK come to work or to study.5
Where are they from?
Most new migrants are not from EU countries. Indeed, for the last several years, more EU migrants resident in the UK have moved back to the EU than new migrants have entered.
The most common source countries are not overly surprising - large countries with a British colonial history are well-represented. In the year ending 2024, the largest contributors are:
India (240,000 people)
Nigeria (120,000 people)
Pakistan (101,000 people)
China (78,000 people)
Zimbabwe (36,000 people)
No rich country makes this list, but that is not unusual; in general, rich countries tend to have relatively small diasporas relative to country size.
How much money are migrants earning in the UK?
In general, the average migrant earns a similar amount to the average Brit.
Upon first arrival, a new migrant often makes less than a Brit. This is likely because they’re in a new labor market, may not speak the language perfectly, and their experience is likely to be less easy to understand if you are a recruiter. Wages for migrants increase pretty sharply after arrival, as they gain UK-relevant experience.
By year 2-4, immigrant wages look relatively similar to native-born wages. Beyond this point, immigrant wages generally exceed native-born wages. Immigrant wages seem likely to continue to be higher than the wages of the native-born, as recent immigration changes have substantially increased the minimum salary needed for a skilled worker visa.6
The difference between immigrant and native-born wages isn’t huge, though. In 2022, the median foreign-born person made £33,000, while the median British-born person made £32,000. A 3% difference in average wages is not large.7
Immigrants are also over-represented at the very top of the income distribution. This is particularly important to consider when one thinks about the UK, as the UK has a progressive taxation system. A person who earns 4x the average wage pays far more than 4x the taxes.8
Given a slightly higher median wage, and a significantly higher likelihood of being in the top 1% of earners, migrants pay (on average) more tax than the native-born.
What does this mean for the state?
What are the fiscal impacts of migration to the UK?
Generally, a state is stuck with its citizens for their entire life cycle. Immigrants, however, may or may not be resident in a state during particular periods. A solely fiscally oriented state would like immigrants to pay as many taxes as possible and consume as few state services as possible, so that they contribute to the state without taking much out.
Ideally, this would mean immigrants move in after university, make a lot of money, pay a lot of taxes, do not have children who would cost money to educate, and at retirement, promptly return to their own country (not drawing a pension or incurring healthcare costs in the country where they immigrated). States generally don’t get that, even for migrants, but migrants are generally a better fiscal bet than citizens.
This is for three reasons:
As noted above, migrants make (marginally) more than the native-born and thus pay more taxes.
Migrants pay visa fees. Generally, each visa carries a fee (varies based on visa from £200 to £2000) and then one also has to pay the Immigration Health Surcharge (IHS). The IHS is £776 - £1035 per person per year of visa.9
Generally, after five years, one can apply for indefinite leave to remain.10 At this point, a migrant is no longer liable for additional fees. I’d estimate that each migrant ends up paying some ~£6000 or so.
Migrants are in country for more profitable portions of one’s lifespan.
People are highly (fiscally) net negative at both the beginning and end of their lives. At the beginning of one’s life, the state pays for education; at the end, the state pays for a pension and (generally the bulk of) healthcare. It is in the middle part of life - generally ages 25-65, or so - where one is net positive.
UK immigrants, on average, arrive at age 25. This means the state does not have to pay for educating migrants, which makes them generally cheaper than the native-born.
All three of these mean that, on average, migrants are net contributors to Britain’s treasury. Indeed, the Office for Budget Responsibility estimates that when there are more migrants, the state has to borrow less and Britain is better off.
That is not to say that every migrant is a net contributor to the economy; somewhat inevitably in a welfare state, some people will end up as net recipients. However, most migrants are contributors, and some are significant contributors.
OBR’s work is illustrative on this point. They split migrants into three categories: low-wage (25th percentile of first year wages), median wage (average UK wages) and high-wage (75th percentile of first year wages).11
(OBR)
In this model, average-wage migrants and high-wage migrants are net contributors. Low-wage migrants are not, but there might be other reasons one still thinks low-wage migrants are valuable to admit into a country.
For instance, low-wage migrants are particularly likely to work in the care sector. One might think that hiring relatively cheap labor for care work frees up higher productivity workers to work elsewhere. This appears to be what happens in the US, where undocumented migrants often work in care. When there are fewer undocumented migrants, more highly educated mothers end up staying at home to care for children, and thus overall wages decline.
Is this the case in the UK? I don’t know, but it seems possible. This example gives some sense of why modeling the exact effects immigration will have on the economy is difficult. Immigrants do not participate in a static economy; their participation in the labor force changes it.
Furthermore, there are other assumptions one must make when modeling the fiscal effects of immigration. We know that most immigrants will probably move back to their home country at some point,12 and that many will not retire in the UK, but parameterizing the exact timing involved can be difficult. Overall fiscal impact is also quite sensitive to how long people live - and frankly, we don’t yet know how long today’s 25 year olds will live.13
Still, we can draw some broad conclusions. Given that migrants make a similar amount to (or more than) the British-born, and generally cost less, migration is net positive for the British economy.
What about migrants’ children?
When considering the fiscal impact of migration, some commenters are worried about the children of current migrants also being net negative. I think this is not too much of a concern for two reasons:
Migrants, on average, make about the same as the native-born. For their children to be net negative, one would need to believe that the children of migrants would perform worse (on average) in the labor market than their parents. This seems unlikely to be true.
What data we have - which is admittedly limited - suggests that second-generation immigrants have similar-to-slightly-better wages. Dustmann, Frattini, and Theodoropoulos 2010 finds a small wage advantage for second-generation migrants relative to the native-born.
Interestingly, this is not because there is no discrimination against ethnic minorities. In the same job and location, ethnic minorities are likely to make slightly less. But second-generation migrants are more educated than the average non-minority Brit, and second-generation migrants are more likely to live in places where wages are high.14 This more than compensates for any disadvantages in the job market, and second-generation migrants still do well in the labor market.
There simply isn’t evidence for a cycle of dependency among British migrants.
How do migrants affect public service provision, sector-by-sector?
Since migrants are, on average, net tax contributors, they are, on average, also net contributors to public services. This is relatively simple arithmetic; the more money the government has, the more it can spend on the NHS, defense, and/or a public service of choice.
However, this could be negated if migrants are (for some reason) much more costly than average. OBR doesn’t think that they are - they model positive net fiscal contribution for most immigrants - but we can also consider the impact sector-by-sector.
How do migrants affect the justice system?
I think it is very likely that immigrants to the UK commit less crime than the native-born. I’ve previously written several thousand words on why I think this, but in short:
Migrants are substantially underrepresented in prisons and jails.
What academic work there is suggests that immigration has little to no impact on crime rates in the UK.
Therefore, migrants do not appear to be incurring disproportionately high costs in the justice system.
How do migrants affect the NHS?
Staffing
There is a stereotype that the NHS runs on foreign labor. While this isn’t entirely true, the NHS does employ more foreign workers than average. About 16% of the UK population is foreign-born, while about 19% of NHS staff are foreign-born.
Furthermore, this is not driven by low-wage workers, as the percentage of foreign-born workers rises with the level of education required. 30% of nurses in the NHS are foreign-born; 36% of doctors in the NHS are foreign-born.
Given the NHS has a perpetual shortage15 of workers, and that health workers likely produce social surplus beyond their tax receipts, the UK benefits from the large number of foreign-born health workers currently practicing in the country.
Costs
Furthermore, immigrants are not disproportionate users of the NHS.
NHS use is highly correlated with age.
(UK HSA)
Immigrants are, on average, younger than the native-born and therefore consume fewer healthcare resources. The NHS thus benefits from a young(er) immigrant population that also provides a considerable fraction of its staff.
How does migration affect social housing provision?
The white paper specifically notes that immigration has strained housing access.
Social - or subsidized - housing is relatively common in the UK, with about 17% of households living in this type of housing.16 Immigration status is taken into account when housing allocations are made,17 but there are some circumstances in which an immigrant can occupy social housing.18
Overall, those born outside the UK are slightly less likely to live in social housing than the native-born. though the difference is small (16% vs 17%).
This might be changing over time, though; immigrants are somewhat overrepresented in new social housing tenancies. I’d say we need more data to determine if this pattern holds, but it is possible that in future, immigrants will incur more social housing cost than the native born.
What about London?
In the UK property market, there are essentially two regions: London and not-London. The average home in London costs £545,000; the average home outside London costs £186,000.19
London also has vastly more immigrants than the rest of the UK - it is 40% foreign-born. One could imagine that if immigrants were disproportionately likely to use social housing in London, that could make social housing for immigrants disproportionately expensive even if immigrants are not overall more likely to use social housing.
I don’t think this is true, though, at least if one considers recent immigrants.
There are three categories of resident in social housing:
UK-born UK nationals (about 84% of the UK population, about 60% of the London population)
Non-UK-born UK nationals (about 7% of the UK population, about 17% of the London population20)
Non-UK born non-UK nationals (about 9% of the UK population, about 23% of the London population)
In London, there are about 780,000 households living in social housing. British-born British nationals are underrepresented in social housing (405,000 households, 52%). So are foreign-led households (115,000, or 14%).21 Households headed by a British national who was born outside the UK are substantially overrepresented (about 260,000 households, or 33%).
This immigration white paper is largely concerned with the explosion of immigration from 2020 onwards. These are not the people in London social housing, because it generally takes at least five years to get British citizenship. London social housing is under pressure - waitlists keep growing - but it does not seem to be driven by post-2020 migration.
One could argue that past immigration puts London social housing under strain, because non-British-born UK citizens are overrepresented, but this feels like an odd argument. Once someone is a UK citizen, they are entitled to all of the same things as any other British citizen; unless one is planning to strip people of their UK citizenship, any changes to immigration law are completely irrelevant to this group.
Still, it seems that social housing probably is put under some strain by non-natives who have become citizens. I think it’s plausible that immigrants, on average, do consume more social housing than the native-born.
Overall Impact
Still, social housing is the only area where I think is plausible that immigrants cost more than the native-born - and social housing is quite a bit cheaper than healthcare.22
There is a reason that OBR models the average migrant as a net contributor: they pay similar amounts in taxes (due to similar earnings levels) and consume fewer services. Public services are not put under more strain from high migration; rather, public services would be in better shape if there were even more migrants.23
The author’s firm belief is that nothing is incalculable.
There is a significant caveat in these numbers - all net migration numbers after 2020 were calculated with a different methodology than before 2020. (Hence the dotted line at 2020.)
The pre-2020 methodology was known to underestimate migration, particularly from eastern European countries. It remains to be seen how accurate the current methodology is, but early indications are that it, too, may underestimate migration. Both the 2022 and 2023 estimates have had significant revisions (by >15%); I would not be surprised to see the same for 2024 and 2025.
This is a correction; I originally stated work visas were the largest category, as I looked at 2023 numbers instead of 2024 numbers. I have offered Rasool from the EA forum a $10 bug bounty for catching this error.
This peaked in 2022, at 263,000 visas granted due to the start of the war in Ukraine; it has since declined to just 40,000 visas per year.
While many visa types do allow a migrant to bring their spouse or children, and those people are not required to be working or studying, there are many more main applicants (479,000 in the most recent year of data) than there are dependents (163,600).
Specifically, in April 2024, the median full-time worker made £728 a week, or £37,856 a year. This would result in a tax bill of £7,078. A person who makes 4x the median would make £151,424 and pay £59,383 in taxes. This is 8x the tax bill of the median worker, even though the salary multiplier is only 4x.
No, paying this does not exempt migrants from paying for the NHS via National Insurance.
Soon to be 10 years, if the immigration white paper suggestions go through.
I think their modeling is a needlessly conservative estimate for the fiscal contributions of all types of migrants, because - as discussed above - first year wages are generally quite a bit lower than wages that a migrant once a migrant has been in the country for a bit, but OBR assumes that wages for migrants stay at the same percentile as year 1.
Many people in rich countries assume that if a migrant moves to a rich country, they will definitely stay their entire life. This is not true. Based on prior trends, less than one-third will end up staying in the country long-term.
At current tax rates / state pension age, the UK would be in serious trouble if everyone lived to be 100. I think it is likely that tax rates and pension age will change as medical science advances.
London
OK, so it’s a little more complicated than that; there is some concern that the UK would be better off expanding internal medical training places rather than importing workers from abroad. However, given the NHS’ difficulty retaining staff, the size of the shortages, and the UK’s aging population, I think it is likely that the NHS will continue to need to rely on immigrant labor.
By contrast, only 2% of US households receive housing assistance, and this is mostly in the form of vouchers. It is very rare for a local authority in the US to own and operate subsidized housing themselves.
This is a correction from the original version of the post, which incorrectly said that immigration status was irrelevant. I have offered Lauren Thomas of Priced Out UK a $30 bug bounty for this correction.
Generally if they have lived in the UK >5 years or if they have refugee status.
My calculation, given that London is home to 3.5 million of the 28.3 million households in the UK (12.6%) and that the average London home price is £545,000 compared to average home price in the UK of £282,000.
My calculation, given that London is 23% foreign passport holders and 40% foreign born.
This seems logical. Many local authorities - including those in London - require you to have lived in an area for five years before you qualify to even get on the waiting list for social housing. UK nationals are more likely to have lived in the UK for five years than non-UK nationals.
Assuming the marginal migrants had similar observable characteristics to the current immigrant population.
Great analysis work. I wanted to ask how much faith you put in the data though. I recall one dimension of the "UK grooming gangs" stories is that UK police were treating immigrant crimes different than domestic out of fear of political backlash, stories which seemed plausible if not definitive from across the pond.
I also wonder if aggregate data is obscuring local issues. For example, I found no evidence that the wave of US immigration since COVID changed anything about the town I live in. But when I visited my brother in Denver for Christmas, the local cathedral and grocery stores had armed guards, apparently after a number of violent encounters. I've never seen anything like it, and it has made me more sensitive to community-level concerns that might be more leading indicators of problems.
The biggest difference is that the US has no safety net. We must pay for and compete with immigrants for every service whether we’re citizens or not. This creates a huge resentment problem especially when you have poor immigrants moving here and having the state pay for them to have babies when the average American pays $20000 out of pocket just to have a baby now. Our public housing system has mostly been eliminated due to the demand. Landlords can rent anything so there is no reason for them to partake in the section 8 program (a rent subsidized program where the gov reimburses the land lord the difference between fair market value and the persons ability to pay). I live close to the northern border so I don’t see very many, but I can tell you there is a huge resentment towards immigrants in the US now. We never should have allowed this to happen as it will be very painful to reverse. Without a safety net and adequate wages for the local population it’s a recipe for disaster.