Are recent immigrants a "ticking time bomb" for British public finances?

At a few recent policy events, I keep hearing that dependents on visas are a Problem. Indeed, I keep hearing specific complaints that this was particularly bad during the tail end of the Conservative government. During this time, there were - as I keep being told - a huge number of low wage migrants brought dependents with them who do not work.

If these new immigrants are a significant fiscal drain, the UK probably does not want them to stay long-term. Thus, there have been moves to reform the indefinite leave to remain system, so that the UK is not stuck with a large non-working population.

I decided to check the data on how worried the UK should be about this. Is the UK actually the world’s welfare state, having admitted vast numbers of people who will take out more in benefits than they pay in taxes? Will the British-born be on the hook for the astronomical cost of mass migration?

I decided to look at the data. Handily for me, the Home Office has collected data about this specific wave of immigrants, including information on their wages and labor force participation.

tldr: I think these concerns are overblown. Even considering only new Boriswave immigrants, I calculate that the average new immigrant salary was still higher than the average British national’s. This is true even when one considers dependents and corrects for their likelihood to take paid work.

Considering that immigrants are also cheaper to the state than the native-born, and they outearn the native-born, migrants who arrived 2019-2023 will make a net contribution to British finances.

I do not think that the UK should restrict immigration for fiscal reasons. If the UK does choose to restrict immigration - as it seems inclined to do - it should be clear about why. It is not because immigrants have placed “public services and housing access… under too much pressure” and immigration threatens the welfare state. Rather, it is simply because voters don’t like it very much.

First, let’s start with basic statistics.

The UK Labor Force

Before discussing migrants, it is useful to calibrate expectations.¹ If one is going to talk about the wages and labor force participation of a new group, one must also know what the labor market looked like before.

The median employed worker made about £613 a week, across full- and part-time workers.² This is equivalent to an annual salary of £31,891.³

This data excludes the self-employed. I don’t have great data about what self-employed residents make, so I've assumed that they make similar amounts to those who are not-self-employed. This is almost certainly too generous, because 2023 data suggests that the self-employed generally make less than the non-self-employed, but it’s what I have, so I’m going with it.

Of course, not everyone is currently in work. According to ONS, about 75% of UK citizens aged 16-64 were employed in 2024. The remaining 25% might be unemployed, in school, providing care, or indeed anything else that takes you out of the labor force.

Considering this, I estimate the median working-age UK citizen made about £23,918 in 2024. This is probably somewhat of an overestimate, though.

What happened with migration from 2019 to 2024?

During this period net migration to the UK was very high by historical standards. In the mid 2000s, net migration hovered around 250,000; in 2023, it hit a peak of 906,000. That is a significant change; perhaps it is not surprising that there has been equally-significant backlash.

Between 2019 and 2023, new entrants to the UK⁴ were:

• ~11% people on skilled worker visas

• ~8.5% dependents of people on skilled worker visas

• ~15% people on health and care worker visas

• ~18% dependents of people on health and care worker visas

• ~12% family visas (e.g. the spouse of a UK citizen)

• ~7% other visas (business mobility, investor, global talent, high potential individual, etc)

• ~4% dependents on other visa types

• ~24% people on humanitarian visas (Ukraine, Hong Kong)

Below, I split this into three categories: main applicants, dependents, and people on humanitarian visas.

Main Applicants

I’ve included four types of visa in this category: skilled worker, health and care worker, family (usually spouses of British citizens), and other main applicants.

There is considerable variation in salary by visa type. The highest median incomes are for skilled workers and business mobility visas, at £56,600 and £47,900 respectively.⁵ The lowest is for those who are joining family, at £20,200.⁶

Some of this variation is structural; a skilled worker or health worker must have a job offer in hand in order to acquire a visa, but someone on a family visa does not.

Overall Wages

I calculate the average employed main applicant’s annualized salary at £37,168.⁷ Since not all main applicants are required to be in work, the average income is slightly lower, at £34,376.⁸

Still, this is higher than the median UK worker (£31,891), let alone the median person in the UK (when one corrects for labor force non-participation). Given that the British state doesn’t have to pay to educate migrants, this is clearly an excellent fiscal deal.

My calculation isn’t perfect; the Home Office provides data on only one kind of “other main applicant”, so I’ve applied that income mean to the whole category.

Still, even if we set the median wage for “other main applicants” to £0, the overall main applicant average would be above the average wage for the British-born, solely on the strength of skilled worker incomes. The median skilled worker makes more than twice what the median British citizen does.

Wait a second - a ton of these people don’t have any income.

According to the Home Office, ~20-30% of main applicants have no PAYE earnings.⁹ Are there actually vast numbers of people moving to the UK on skilled worker visas that then proceed not to work?

Probably not, no.

This is for two reasons:

First, and most simply, the Home Office’s matching process with PAYE data isn’t perfect. They even say that themselves; there are definitely people working (legally!) in the UK that the data merge just didn’t catch.¹⁰

However, I think the more significant reason is that many of the migrants who don’t have UK earnings have already left the UK.

Most types of main applicant require a job offer in order to move, so one cannot enter without a job. Quitting your job and living on welfare isn’t much of an option either, because 1) once you quit your job, you have 60 days to leave the country, 2) you and your family aren’t permitted to access most types of assistance.¹¹

As anyone who lives in the UK will tell you, it is an expensive place to live. It is unlikely that many people have moved to the UK, with a job in hand, then quit that job, ignored the requirement to leave and decided to become an illegal immigrant instead.

Rather, many of the main applicants - on routes that require jobs - do not have PAYE earnings because they either have not yet entered the UK or have already left.

The latter is particularly likely because return migration - that is, leaving the country one immigrated to and returning to one’s native country - is quite common. Based on older UK data, 10% of immigrants leave within a year; 40% within five years.

Given this, it seems very likely that many of ~20-30% of main applicants that are required to have a PAYE job while in the UK that do not appear to have a PAYE job are simply… not in the UK anymore. Thus, our conclusion that migrants are very good for public finances is unchanged.

Let’s Talk About Dependents

OK, so main applicants earn reasonable wages. However, much of the public concern about immigration has not been about the main applicants, but their dependents.

Who are dependents?

To be a dependent, one must be the spouse, unmarried partner,¹² or a minor child of the main applicant. Note that parents, cousins, siblings, adult children¹³ or other relatives cannot qualify for a dependent visa.

Over 2019-2023, there were 537,900 dependent visas issued to 776,900 main applicants. That is, the average migrant to the UK entering in 2023-2024 had 0.69 dependents, though, in some years, there were more dependents than main applicants.

Are dependents dependent?

Does this mean that - as Katie Lam MP would have - that millions came here without intent to work? Not really. The term dependent is a bit of a misnomer here. They are a dependent in terms of immigration status only.

In some countries (cough cough, the US), dependents are barred from work. This is not true in the UK. Indeed, in a lot of ways, your options for employment are better as a dependent than as the main applicant.

For instance, I am on a skilled worker visa. Any job I take must:

• Be for a UK employer that’s been approved by the Home Office

• Be for an employer that has a ‘certificate of sponsorship’ (CoS)¹⁴

• Be on the list of eligible occupations

• Be paid at least a minimum salary at the start of the job (in 2023, £26,200; in 2024, £38,700; in 2025, £41,700, or the “prevailing wage” in your field, whichever is higher)

My employer must also pay the certificate of sponsorship fee (£525) and the Immigration Skills Charge (£364 - £1,000 for each year they intend to employ me).¹⁵

By contrast, the requirements for my (nonexistent) spouse's job would be:

• It is a job in the UK.

Their employer also does not have to pay any fees to employ them. This leads to an interesting conclusion: if you are a couple where both people could plausibly get UK work visas, it is probably a better idea for one person to get a work visa and one to enter as a dependent. Being a dependent for the purposes of immigration does not mean that you are economically dependent.

Labor Force Participation Rate

The fact that dependents can work does not necessarily mean that they do, though. Every UK citizen is eligible to work; this does not mean that every UK citizen does. Indeed, a significant fraction of people are not in work at any given time.

To figure this out, I again use Home Office data. I calculate the labor force participation rate for each category from:

1. The percentage of dependents with PAYE earnings in that tax year (by visa category).

2. The percentage of main applicants with PAYE earnings in that tax year (in a visa category)

The latter adjustment is because it seems unlikely to me that the dependents of a worker will receive PAYE earnings if the main applicant (required to have a PAYE job) is not receiving PAYE earnings.¹⁶

As above, the majority of those on visas with no PAYE earnings have probably left the country. Their spouse and children are quite unlikely to have stayed in the UK if they have departed. Therefore, I adjust the labor force participation rate of dependents by the percentage of main applicants with PAYE earnings.

I calculate the labor force participation rate for dependents in this sheet. All of these are likely to be underestimates, as the Home Office data excludes self-employment,¹⁷ but it does appear to vary substantially by visa type. Only 37% of dependents of those on business mobility visas work, while 83% of dependents on health and care worker visas have some earnings.

That last point bears some emphasis, as there has been considerable discourse about the dependents of low-wage care workers being a drain on the British economy. And it is true that there were many such dependents; a full third of non-student visas issued in 2023-2024 went to dependents of someone on a health and care worker visa. But very few of these migrants seem to sit at home - indeed, the dependents of those on health and care worker visas are considerably more likely to be in work than British citizens.

Labor force participation seems to be quite sensitive to the earnings of the main applicant. Adult dependents are very likely to work in migrant households where the main applicant is not a particularly high earner (e.g. health and care worker visa holders), but are less likely to work the more money the main applicant makes. 83% of health and care worker dependents work, while only 60% of skilled worker dependents do. This seems reasonable; a health worker on £30,000 can’t afford a stay-at-home spouse, but a skilled worker on £56,600 might be able to.

Weighting by the number of dependents of each type, 70% of dependents that are likely to be in the country do have PAYE earnings. This is similar to the percentage of the overall population with PAYE earnings.¹⁸

Individual Wages

There is less variation in wages by type of visa among dependents than main applicants. This makes sense - as noted above, main applicants can have significant restrictions on where they can work and what types of jobs they can take, while dependents experience the general UK labor market.

Dependents of those on skilled worker visas are the highest earners, taking jobs paying £30,200 on average; dependents of those on health and care worker visas make the least, taking jobs that pay £22,100. As above, these are underestimates; this excludes any self-employment incomes, such as Deliveroo earnings.¹⁹

The average employed dependent makes £25,534; the average dependent (accounting for labor force non-participation) makes £17,898. This is lower than the average UK wages, by £6,020.

However, that’s not as concerning for public finances as it sounds for two reasons.

1. Our data excludes self-employment earnings. I’ve included an imputed self-employment earning for UK residents (indeed, I’ve probably overestimated self-employment earnings for UK residents), but not for migrants. Dependents are free to start businesses, and this could make up some of the gap.

2. Dependents - and other migrants - are quite a bit younger than the British-born. I couldn’t find data for 2023-2024, but in 2020, the median new migrant was some 14 years younger than the median British person. One would expect a younger group to have lower current earnings even if they have similar lifetime earnings.

Considering the average wage for a person in their 30s is some £7,600 more than someone in their 20s, recent immigrants actually seem like they’re on a perfectly reasonable earnings trajectory. In 14 years, it seems reasonable that they will close the £6,020 gap with the native-born.²⁰

Household Wages

Of course, one does not have just main applicants or just dependents. People are part of families; you cannot simply say “I would like your high-earning spouse but the low earning one can stay in Pakistan”. (Or you can, but then you take the risk neither spouse will move to the United Kingdom.)

From our wage calculations, we can estimate the average household wages for skilled workers, health and care workers, and those on other visas.

Skilled worker households are relatively small, at 1.8 individuals. For every four households with a skilled worker main applicant, three will contain a dependent; one will not. The average household income is about £70,203. This puts the average skilled worker household at about the 70th percentile in the UK.

Despite the stereotype, the average health and care worker household isn’t that much larger, at 2.24 individuals. Assuming that households have two adults and 0.24 children, they have an average household income of around £49,180. This is very similar to the UK average - Britain-wide, the average household contains 2.36 individuals and has an income of £53,140.

Other work visa holder households are smallest, at 1.6 individuals. As noted above, I’m making a bit of a heroic assumption, applying business mobility visa wages to all other work visas - but if this holds, these households have an income of £54,213. This is nearly identical to the average British household.

Humanitarian Visas

All of the above ignores the largest single visa category for 2019-2023: humanitarian visas. I always feel a bit weird writing about the fiscal dynamics for humanitarian visas, because that is not why countries take humanitarian migrants.

It is true that the median Ukrainian migrant doesn’t make very much money. The government itself estimates the median Ukrainian only makes £16,000 per year, less than even the much-complained-about dependents of care workers.

This is one of the lowest wage categories; only the dependents on other visas have lower expected wages, and that number is a little bit fictional. (I only had data for the dependents of those on business mobility visas; I’ve assumed the dependents on other types of more obscure work visas had similar labor force participation, but that’s… not terribly well justified.)

The good news for British public finances is: if one wants to limit the number of new refugees entering the country, that problem has already been solved. Refugee numbers spiked in 2022 due to the war in Ukraine, but have declined quickly. The number of humanitarian visas issued in 2023 was just a quarter of the number issued in 2022. It continued to decline in 2024.

As long as Vladimir Putin manages to avoid invading any other countries, the 2022 spike in humanitarian visas is unlikely to be repeated.

Summary

• The average main applicant who received a visa 2019-2023 made about 45% more than the average person in the UK.

• The average dependent who received a visa 2019-2023 made about 25% less than the average person in the UK.

• The average person who received a humanitarian visa made about 33% less than the average person in the UK.

When we take a weighted average of all categories of 2019-2023, the average new migrant who arrived 2019-2023 made £24,881 in FY 2023-2024. This is about 4% more than average earnings - even when I have made optimistic assumptions about native-born earnings and excluded any self-employment income for migrants.

It is difficult to see how admitting people who outearn the native-born would be catastrophic for public finances, particularly given that the British state need not pay to educate them. The data certainly does not seem to suggest this cohort is a “ticking time bomb” for British public finances.

1. Particularly for the Americans, who are going to read this post and go “??? what are these salaries???”

2. This is based on ASHE data, which may be upwardly biased. Therefore, I may overestimate the median native-born salary.

3. ONS has a slightly lower number, at around £2300 pcm in 2023. I’ve chosen the more optimistic estimate.

4. Note: I exclude students from this entire post, because students are not expected to contribute to public finances via income taxes. Furthermore, to obtain ILR, they must spend at least some time on other visa types; the students that stay long-term will thus be counted in those visa types.

5. These numbers are all annualized. Almost all people actually make less than this their first year, because they arrive partway through the year.

6. One might reasonably categorize these people as dependents of the British citizen, but since they aren’t dependents of someone on a visa, I’ve called them main applicants.

7. This is technically a mean of medians, which isn’t ideal, statistically. Unfortunately, this is the data I have. Given that median incomes tend to be lower than mean incomes, I expect this is an underestimate of the mean of means.

8. I assume that all main applicants are working where they are required to have a job (HCW, SW, other work), but adjust the family visa income by observed labor force participation.

9. For the Americans: Pay As You Earn; think W-2 wages.

10. “Of the approximately one million Sponsored Work and Family visa records granted between April 2019 and March 2023, extracted for this report from Home Office case working systems, 93% were successfully matched to HMRC’s Migrant Worker Scan (MWS) database, enabling their records to be linked to HMRC PAYE RTI data. Further development work is ongoing to improve the quality of matches across additional visa routes.”

11. This data is only for people whose visas were granted after 2019, and uses their 2023-2024 earnings; they will not have been eligible for ILR/able to access to public funds.

12. With a relationship duration of >2 years.

13. Adult children can continue to have a dependent visa if they have turned 18 during their parents’ visa duration. This does technically mean you can have an adult child on a dependent visa, but this is likely to be a relatively small percentage of children on dependent visas. Since visas are for a maximum of five years, a 17-year-old would have a maximum of four years on a dependent visa; a 16-year-old a maximum of three, a fifteen-year-old a maximum of two years, etc.

If we assume children are equally distributed in age (that is, 6% of children who enter the UK are at each age), and all visas are five years, approx. ~10% of person-years of dependents would be as adults (>= 18). Since not all visas are five years long (and a shorter visa length will also shorten the amount of maximum time a dependent can be in the country after turning 18), I feel comfortable calling this basically a rounding error.

14. Many UK employers simply do not sponsor visas because it’s a hassle.

15. This serves to disincentivize hiring immigrants, because they’re more expensive to employ than the native-born at the same pay scale.

16. This isn’t perfect; as the Home Office themselves note, it is possible to be on a visa but be long-term sick and thus not have PAYE earnings. I think this is likely to be a relatively small group, though.

17. Since about 14% of UK workers are self-employed, this underestimation could be considerable - but it’s what I have, so it’s what I’m going with.

18. Note that this is lower than overall labor force participation because it excludes self-employed people. In August 2023, there were 30.2 million people in the UK with PAYE earnings. There were 42,861,264 people aged 16-64 in the UK, for a PAYE participation rate of 71%.

19. There seems to be particular concern about migrants working for delivery companies; see Neil O’Brien’s post about “the Deliveroo visa”.

20. Indeed, we know that the wages of migrants increase quite sharply the first few years after arrival.