The EU settlement scheme student loan scam
How an absurd immigration policy has opened up massive abuse of taxpayers
Five years after the “deadline” to apply, new applications for “EU Settled Status” are going up, not down - and large numbers of new grants are being made. We are awarding tens of thousands of grants of EU Settled Status to nationals not of the EU, but of India. And billions of your taxes are being “loaned” to people in other countries - and will not be repaid.
Why?
This article is the result of five separate rounds of Freedom of Information Requests and Parliamentary Questions, so I apologise that it is quite data heavy.
Executive summary
In 2019 the Home Office estimated that between 3.5 million and 4.1 million people would be eligible for EU settlement. But as of March 2026 4.4 million people had been granted settled status and a further 1.5 million held pre-settled status - so 5.9 million in total.
Over 160,000 “late applicants” have been granted settled status since the deadline of the 30th June 2021, along with a further 54,000 family members joining them - so 220,000 people. 124,000 people applying after the deadline got pre-settled status, with 202,000 family members joining them1.
In the past 12 months alone, around 66,000 completely new applicants gained residency via this scheme, five years after the deadline.
Some nationalities are much more likely to apply. At least one in every 20 Romanian citizens has settled or pre-settled status in the UK.
Rather than tailing off as you’d expect, both the number of late applicants and joining family members actually rose sharply in the last quarter.
EU citizens granted settled or pre-settled status via the EUSS are able to bring family members under conditions much more favourable than those available to British citizens bringing their spouses or children. They face no salary test or language test or fee, and extended family members can be brought over, not just direct family.
In the past 12 months, there were more EUSS applications from Indian (15,164) and Pakistani (9,318) nationals, than from French (9,178), German (6,009), Belgian (2,467) or Dutch (5,003). In total there have been nearly 50,000 grants of settled status to Indian nationals via the EUSS. Indian nationals are now the second-largest nationality among new applicants.
People on EUSS are eligible for student loans.
In 2024/25 foreign nationals received around £4bn in student loans. One in five people getting student loans were non-UK nationals. 5% of all first time student loans in the UK went to Romanians alone.
The government has no grip on the losses on these loans.
Even if non-UK nationals were as likely as UK nationals to repay, we would expect to lose 30% of the £4 billion a year.
But in practice, almost all nationalities have a repayment rate significantly lower than UK nationals. We can see from new data I obtained that the average repayment rate for non UK nationals is just 51% - much lower than the UK repayment rate (68%).
This means that UK taxpayers are losing between £1.2 and £2.2 billion a year on loans to non-UK nationals - probably more like the latter.
Losses on loans to Romanian nationals alone probably cost around £650m a year.
As well as any eventual losses from people not earning enough, the British taxpayer is already owed £1.5bn by a total of around 50,000 overseas “non-compliant” borrowers. “Non-compliant” nationals from tiny Cyprus alone (with a population similar to Leeds) owe over £300m.
How on earth did we get here?
--------
In the beginning
As part of the Brexit withdrawal agreement signed in January 2020, the EU Settlement Scheme (EUSS) was set up to allow European nationals residing in Britain to apply to stay here permanently. The deal was: we don’t disrupt the lives of people here - and they don’t disrupt the lives of Brits abroad. Fine.
EU nationals were granted either pre-settled status or settled status, depending on how long they had been resident in the UK before the end of the transition period on the 31st December 2020. People with pre-settled can later move to settled.
In practice, settled status functions as a more generous form of indefinite leave to remain, in that it granted immediate access to the NHS and welfare, while also granting foreign nationals family reunion rights which are more generous than for British citizens.
This is a crucial point in the story.
To bring a family member to the UK, a British citizen must meet a £29,000 minimum income threshold. The family member must pass a Secure English Language Test and pay nearly £2,000 in application fees, with the visa initially granted for under three years.
EUSS joining family applications are free, permanent, and open to extended family members. British citizens can only sponsor partners, children, and adult dependents.
So someone with EUSS has great ability to bring more people to the UK.
This is pretty remarkable in and of itself, but arguably, these concessions were a necessary price to leave the EU. However, we are now a decade on from the Brexit vote, and five years on from the deadline to apply to the scheme, and people are still getting new grants in their tens of thousands.
So what on earth is going on?
EUSS: the big picture.
First, it’s worth looking at the scale of this scheme.
The Home Office Impact Assessment for the EU settlement scheme published in March 2019 stated that: “the total number of EEA citizens and their family members eligible to apply for the EU Settlement Scheme by the end of the planned implementation period on 31 December 2020 is likely to be between 3.5 million and 4.1 million.”
But as of March 2026 4.4 million people had been granted settled status and a further 1.5 million held pre-settled status - so 5.9 million in total2.
This means that, in part because of the unlimited number of family members those who received grants were able to bring, each of the top six nationalities receiving settled status has received more grants than the number of passport holders from that country recorded as resident in Britain in 2021.
For Polish nationals alone, grants of settled status have already exceeded the 2021 passport holder count by nearly 200,000.
The door is still open
Five years on from the application deadline, you’d expect most of these grants to have been awarded long ago, and for applications to have ended. Remarkably, this isn’t the case.
In the past 12 months alone, 440,000 applications to the EUSS were processed.
Now, most of these applications are what’s known as “repeat applications”, where people are moved from pre-settled status (temporary) to settled status (permanent). This represented around 275,000 applications, with the majority being granted. But that means there were still 165,000 other applications. How come?
First, there are those who are applying late. The Home Office say that they assess these applications on a case-by-case basis, considering “reasonable grounds” for missing the deadline. The numbers suggest that this interpretation is a very generous one - in the last 12 months over 36,000 late applications were successful, with only around a half of all these applications refused.
I have asked the Home Office for a breakdown of why so many late applications are still being allowed - but they refuse to say. Surely the window for reasonable exceptions has long passed.
Second, there is also a route open for family members. The only requirement for EU citizens bringing their extended family is having settled or pre-settled status, and saying that the relationship with the family member predates 31 December 2020. There is no cap, and no fee, on how many family members can be brought over. There is also no deadline for joining family members to apply via the EUSS, and in the last 12 months, around 30,000 family members were granted settled or pre-settled status via this route.
Between these two routes, around 66,000 completely new applicants gained residency via this scheme in the past 12 months. And this number is now beginning to rise. The chart below strips out people moving from pre-settled to settled status.
Alarm bells
The longer it is allowed to continue, the further the scheme drifts from its original purpose. And several things in the data ring alarm bells, and suggest it is being abused.
If you missed the deadline by a week that would be one thing. But missing by five years and still being waved through is another.
There’s no reason some nationals should be worse at remembering to post the letter. And yet we see that while for countries like Germany or France, roughly a quarter of grants of settled status have come since the deadline, for Romania it is nearly half. For non-EU citizens (mainly family members joining) it is higher still, at 54%.
We can also see that later cohorts are more likely to be dominated by family members.
Over 160,000 “late applicants” have been granted settled status since the deadline of the 30th June 2021, along with a further 54,000 family members joining them - so 220,000 people, and a quarter are family members.
124,000 people applying after the deadline got pre-settled status. They are people who are less far down the track, likely later cohorts. They have 202,000 family members joining them. So 62% are family members for this mainly-later group.
Then there is the growth in the number of applications from countries outside the EU. In the past 12 months, there were more EUSS applications from Indian (15,164) and Pakistani (9,318) nationals, than from French (9,178), German (6,009), Belgian (2,467) or Dutch (5,003). And in total, non-EEA nationals now account for more than 20% of all new applications.
It’s not just applications but grants. Nearly 50,000 Indian nationals in total have gained settled status since the scheme opened. For context, this is more than double the number of grants given to Danish citizens (23,793) and comfortably more than the number received by Belgian citizens (30,769).
While the majority of these grants will be to those joining family members, government figures show that there have also been more than 2,000 Indian nationals who applied successfully in their own right as late applicants. How on earth is this being allowed to happen3?
Another clear outlier in the data is the volume of applications coming from Romania. In the last 12 months alone, there were nearly 120,000 applications concluded from Romanian nationals, more than triple the next highest, from Italy, at 37,000. As a share of the population, applications from Romania are about five times the average rate for the EU.
One possible explanation is that Romania grants citizenship and a Romanian (and EU) passport to neighbouring Moldovans if they have ancestors in Greater Romania, and this has been the basis of years of scams linked to this route. But there may be other reasons.
Although around half of applications from Romania are repeat applicants (i.e. those moving from pre-settled to settled), nearly 56,000 are new (i.e. those applying as late applicants or joining family members). And of these new applicants, around 16,000 were successful.
So what share of Romanians have settled status in the UK now?
Romania's resident population is about 19.0 million as at December 2025 on the Eurostat measure. But Romania's own statistics office publishes a second, larger figure: a legally resident population of 21.6 million as at January 2026. The gap between the two is the emigrant population: people still registered in Romania but living elsewhere. As of March 2026 over a million Romanians had Settled or Pre-Settled status in the UK, so nearly one in twenty (4.7%) of all Romanians have settled or pre-settled status in the UK - though not all are necessarily resident here.
Why does this matter? The link to student loans
It seems that people are increasingly gaming the system and it is drifting away from its original intent. Why does this matter, beyond increasing immigration?
As we have already seen, a grant of EU settled status unlocks a whole host of privileges including access to benefits, the ability to bring over family members free of charge and the immediate use of the NHS.
Perhaps less well known is the access granted to the British student loan system.
From academic year 2021/22, new EU students in England (without a ‘settled’ or ‘pre-settled status’) were no longer eligible for tuition fee funding.
But those with EU settled status become eligible for tuition fee and maintenance support on the same terms as a British national. Those with pre-settled status can access tuition fee support after three years, and the full maintenance loan as well, if they or a family member is in employment..
After much back and forth with the Student Loans Company (SLC) I have begun to uncover some of the scale and costs involved.
In total, if we look at unique individuals who received their first ever loan from the Student Loans Company, there are just over a million non-UK nationals who have received UK student loans, 423,000 of which have been in the last five years. Given that EU students without settled status are no longer eligible for loans since 21/22, you might have expected the share of loans going to non-UK nationals to fall dramatically. But it’s actually gone up a bit:
In the 2024/25 academic year alone, there were 160,941 foreign national borrowers with EU Settled Status and 34,879 who were classified as EUSS Migrant Workers. A further 24,344 were eligible via standard Settled Status or Indefinite Leave to Remain. (There are also a whole host of smaller residency categories which qualify foreign nationals for loans).
A separate dataset breaks down the number of students and the total amount loaned by nationality rather than immigration status.
In the same academic year, some 280,000 foreign nationals received loans[5] worth over £4bn in total.
The breakdown by nationality also shows some trends that suggests people have learned to play the system. For example, Romanian nationals are receiving loans at a completely unprecedented rate compared to their EU counterparts.
In less than a decade between 2016 and 2025, the number of Romanians receiving loans each year has grown nearly eightfold. In 2024/25 alone, 86,000 Romanian nationals were loaned a total of £1.3bn. While grants to Polish people have barely moved in the last five years, grants to people from Romania have doubled. The number of borrowers from Romania has gone up each and every year:
It is the same trend if we look at the amount borrowed each year:
If we look at first-time borrowers last year, non-UK nationals now account for a fifth of all loans handed out and a similar share of the value. Romanians alone account for 5% of all UK student loans:
What’s the problem?
The whole purpose of the student loan system is to invest in people who will build careers here and repay through higher earnings over time.
People who are not UK nationals are more likely to leave, so British taxpayers’ investment will get up and walk out the door.
This would be an issue even if we were sure borrowers from overseas were as likely to repay from abroad - we are no longer getting the social benefit.
But there is a practical problem. Loan repayments are deducted automatically from UK payslips. There is no equivalent mechanism to recover money owed from those who move abroad, and repayment rates are far lower.
How much will not be repaid?
The government has no grip on this. For student loans as a whole the government publishes measures of how much of the money it expects to get back. For example, around 30% of all the money lent to full time undergraduates is scored as public spending, because it won’t be paid back.
So even if loans to non-UK nationals were paid back at the same rate - 30% of £4 billion a year would mean a £1.2 billion subsidy being paid by British taxpayers to people who are not British.
But this 30% figure isn’t broken down by nationality. In response to my Parliamentary Question, the government published what it considers to be the “repayment rates” for student loans, broken down by nationality.
In simple terms, this shows what percentage of borrowers who are past their Statutory Repayment Due Date made a repayment in the financial year, broken down by nationality.4
The chart below sets out this repayment rate against the amounts loaned last year. This shows that the most significant borrower nationalities have repayment rates far lower than British nationals. Overall, the weighted average repayment rate for non UK nationals is just 51%, substantially lower than the UK repayment rate (68%).
The official student loan forecasts are that around 30% of all new (Plan 5) student loans will not be repaid (that’s including UK and non-UK borrowers).
The government publishes two similar measures called the “RAB charge” and “transfer proportion” which describe the cost to taxpayers of this expected non-repayment - the difference between the two is about how you measure interest costs.
As a result of this 30% expected loss, the government scores £8-9 billion a year of student loans as spending rather than simply lending.
Given their lower repayment rate, the share of losses to the taxpayer accounted for by non-UK borrowers is likely going to be bigger than their share of borrowing. On the basis of the weighted average above non-UK students are 19.3% of new borrowing - but more like 27% of non-repayment.
If we apply that 27% to the expected losses on student loans (£8.3 billion overall in 2030/315 ) then on the face of it, the cost to taxpayers of losses on lending to non-UK nationals could easily be £2.2 billion a year.6
More detail on variations by nationality
Of the seven nationalities who received more than £100m in student loans, four had a repayment rate of less than 50%, including Romania, which received £1.25bn despite a meagre repayment rate of 45%. This implies that losses on loans to Romanian nationals alone probably cost around £650m a year.
The bubble chart above only includes the larger nationality groups. If we include some of the smaller borrower groups we see even lower repayment rates. The repayment rate for Kuwait, for instance, is at 17%, while only a quarter of liable borrowers from Cyprus repaid anything at all:
Why aren’t non-UK nationals repaying? Are they not earning enough? Or have they vanished?
It looks like it’s a mix. We have also been able to extract data from the SLC on the number of, and amount owed by, overseas “non-compliant borrowers”.
The Student Loans Company say of them that:
“Non-compliant customers are those that are in repayment status, have a balance outstanding, and have not made a repayment in the last 12 months, excluding those with an agreed repayment arrangement in place”
On that definition, there are currently 54,537 non-compliant overseas borrowers from EU countries, and they owe a combined £1.5bn. This is money we may be unlikely to see again.
This figure is itself likely to be an underestimate, for two reasons:
First, it excludes “unverified customers,” which refers to overseas borrowers whose current financial status SLC cannot establish at all[2]. A recent FOI by journalist Matt Davis put this figure for EU borrowers at 42,000, owing nearly £893m. SLC caution that not all of these are necessarily liable to repay, some may be unemployed, on benefits, or between jobs, but the category is growing and the outstanding balance has more than doubled since 2020/21.
Second, the data we have on “non-compliant” borrowers only capture those who have moved abroad; UK-settled foreign nationals who fail to repay are not reflected here at all.
Even with these caveats, the scale of non-payment for those who we know are not repaying is shockingly high for some nationalities.
Cypriot nationals, for instance, are recorded as having over 10,000 overseas non-compliant borrowers, against a total of only around 27,000 Cypriot borrowers since 2005 (so 37%). This is so high that I double checked it with the SLC - but it is correct. In total, the bill from non-compliant Cypriots who are now overseas is more than £300m.
Conclusions
The EU Settlement Scheme was designed to give certainty to people who had made their lives here before Brexit, and enable the same for Brits in Europe. It quickly achieved that aim, and as a result the lives of millions of EU and British citizens were not disrupted by Brexit in the way some had suggested they would be.
However, we are now five years on from the deadline, and the system is riddled with signs of abuse. I am always the first to say that the last Conservative government should have been tougher on all aspects of migration, and this is no exception. But at least in 2022 or 2023 I can see how ministers might have looked at the data and thought this scheme was going to tail away. But when applications are rising in 2026 you can’t possibly defend it.
Shabana Mahmood has always been good at managing the media, and I guess she benefits from a contrast with the most extreme open-borders people in Labour. But really.
It’s not just that that she is preparing to import tens of thousands of people from Sudan and Eritrea to try and fiddle the figures on the small boats. She is also presiding over a system in which new EUSS grants are beginning to rise, fully six years after Brexit. You don’t have to be an immigration sceptic to think that’s nuts.
Non-EU nationals are taking advantage, and countries with less rigorous citizenship processes are now dominating the applications.
It is time to draw a line under this. The government wouldn’t even need new legislation to end new EUSS claims tomorrow. They just need to stop having an absurdly generous over-interpretation of rules. The fact that the Home Office won’t release any information on the reasons why these new claims are being accepted is very telling.
The same logic should apply to family members. If you want to bring over a partner, you need to prove you were in an established relationship at the end of the transition period, on 31 December 2020. So if you are showing up suddenly asking to bring over your girlfriend six years later, Home Office officials need to start being a bit more sceptical.
Likewise on student loans. If European citizens long-established in Britain, who were going to stay here wanted to do a serious degree and repay, that would be one thing.
But my worry is that too many of these people will leave and not repay, or are getting the loan to top up their income while mainly working and doing a bit of studying. My worries are compounded by the facts that (a) I can see the DFE are not even measuring the problem, and (b) the data I can get from FOI and PQs looks really bad.
If we stop the flow of new EUSS claims, the flow of EUSS student loans is likely to dry up too. And we should stop giving student loans to non-EU people with ILR too.
We can then end any remaining abuses with a wider crackdown on low value courses. Even for UK students there’s a long tail of students doing low quality courses who won’t repay. The Conservatives have said we will reintroduce number controls and shut these courses down.
To conclude: we simply cannot afford for Home Office officials to be this “generous” in handing out tickets for a scheme carrying so many entitlements to benefits, student loans and and so on. It is insane that new claims are going up six years after Brexit.
The cost of just one aspect of this, the student loan system, puts the cost into perspective. Foreign nationals received £4bn in student loans in the last academic year. Yet DFE has no idea how much of this we will get back or not - but we could easily be losing over £2 billion a year on loans which we should not be making.
This is not right, not fair, and not sustainable.
It’s time for this loophole to be closed, and this scam to be ended.
Particular thanks to James for his dogged pursuit of the FOIs and WPQs in this article.
Source: EUSS_D03
Source: EUSS_03 and EUSS_RA_04
India does not permit dual citizenship, so it can’t be the case that these people are dual nationals.
Ministers say: “The repayment rate was calculated as the percentage of borrowers whose loan has passed its Statutory Repayment Due Date (SRDD, normally the April after the end of the course) and have made a payment in the financial year out of all borrowers whose loan has passed its SRDD. Although borrowers become liable for repayment from the SRDD, repayments are only due where earnings exceed the relevant threshold. Borrowers who remain below this threshold, or experience fluctuating incomes, are not required to make repayments on a continuous basis. Additionally, periods of further study, unemployment, or other life events may reduce earnings below the threshold, resulting in no repayments being collected despite being past the SRDD.”
OBR: EFO March 2026, page 114).
The measure of non-repayment here is similar to but not the same as the RAB charge or transfer proportion. So without some deeper analysis by HMT and DFE we can’t know exactly what share of that £8.3 is accounted for by overseas borrowers. But the difference in repayment rates suggests something like 2.2bn.
An extraordinary piece of research – thank you!
Fabulous piece of evidence based research, which reinforces that our nation is truly run by idiots. Trump would ban Romanians from applying for student loans and ban flights to and from Romania if they complained. Job done