Students
Who borrows, who repays, and the one thing nobody records.
Britain lends almost £20bn a year to students from England. We know who borrows it, nationality by nationality, only because MPs kept asking. We know who pays it back. What we don’t know, because nobody records it, is whether the students actually turn up.
This page puts the official figures in one place: four parliamentary answers, the Student Loans Company’s data on the weekend courses, and the regulator’s results for every franchised course (one taught by a private college on a university’s behalf). Every figure has its source and every table can be downloaded. There’s a short glossary of the jargon at the bottom.
Nobody records whether students turn up
Student loans are paid on the assumption that students attend. Before the money for a maintenance loan (for living costs) is released, the university or college tells the Student Loans Company that the student is attending. That is essentially the whole check. The lender doesn’t routinely verify it, and I can find no published attendance figure for any course, college or university in England, not from the Student Loans Company, the Department for Education or the Office for Students.
Colleges keep their own registers. What they don’t do is report them anywhere the public, or Parliament, can see. Officials have known this for over a decade:
- 2014. The National Audit Office finds that the Student Loans Company relies on colleges declaring attendance themselves, and can only inspect their records if it already has grounds for suspicion.
- 2017. The same auditors find the company’s attendance data isn’t reliable enough to act on.
- 2024. They find “insufficient evidence that students are attending and engaging with their courses”. The Public Accounts Committee adds that there is “no agreed definition of what constitutes attendance or engagement”.
- 2026. About £190m in student finance is found to have gone to some 22,000 students on courses taught only at weekends, which the rules have excluded since 2011.
Sources: NAO, 2014 (para 5.3), 2017, 2024; Public Accounts Committee, 2024; House of Commons Library.
Who borrows
Pick a nationality. Anyone with a British passport counts as British here, even if they hold another.
Borrowers: PQ 10703 (answered 22 July 2026; first given in the answer to PQ 123974). Amount lent: PQ 296. Repayment: PQ 297. New borrowers by type of loan: PQ 298. All cover loans from Student Finance England and come from the Student Loans Company’s working records, so they can change. “Per 100 passport holders” compares borrowers with the 2021 Census for England and Wales. Treat it as a rough guide: the Census counts all ages, and anyone who also holds a British passport is counted as British.
How it changed
Borrowers from the largest EU nationalities, 2020/21 to 2024/25.
- Romanian borrowers doubled, from 40,949 to 83,789. But the growth was already slowing, year after year: 34%, 25%, 15%, then 7%. That began well before the newspapers got hold of the story in March 2025. In 2024/25 the number of Romanians taking out their first full-time undergraduate loan fell by 11%, while first-time postgraduate loans more than doubled, from 722 to 1,527.
- Slovak borrowers rose almost sixfold and Czech borrowers fourfold. New Slovak maintenance loans jumped from 466 to 2,090 in 2021/22 alone, the first year of the post-Brexit rules. I haven’t seen anyone explain why.
- Ukrainian borrowers rose from 158 to 4,853, as people who came on the refugee schemes became eligible. Not every rise means abuse.
Who repays
The share of borrowers past their repayment date (usually the April after leaving) who paid anything in 2024-25. Only nationalities with at least 1,000 borrowers are shown, because small groups swing about.
It’s tempting to read Romania’s 44.9% as proof of something. But France’s repayment rate is identical and Bulgaria’s is lower. The gap with British borrowers (68.1%) runs across most of the EU. One obvious explanation is that many graduates go home, where collecting from them is harder. A borrower earning below the threshold isn’t expected to repay at all, so a low rate isn’t the same as default. Source: PQ 297.
The weekend courses
In March 2026 the government stopped living-costs loans to about 22,000 students on courses taught only at weekends. It shouldn’t have been paying them at all: since 2011 the regulations have said that attending “on a weekend or during any vacation” doesn’t count as attending. The rule was printed in Student Finance England’s own guidance in 2023. It was only spelt out to universities when the Education Secretary wrote to them in December 2025. These are the 21,380 students the Student Loans Company flagged.
Source: Student Loans Company freedom-of-information response, analysed by Wonkhe (17 August 2026). Figures for colleges are approximate. The data doesn’t by itself show fraud or wrongdoing by any student, university or college, and nine universities have begun legal action against the government’s decision. The figures don’t say how these students qualified for loans, for instance whether through the EU worker route, which requires EU citizens to keep working while they study.
Franchised courses
Much of this happens on franchised courses: a university registers the student and takes the loan money, then pays a private college to do the teaching. In October 2025 the regulator published results for each of these partnerships for the first time. They measure whether students carry on (continuation), finish (completion) and get a graduate job (progression), against a minimum and a benchmark (the result expected for similar students). They don’t measure attendance.
Full-time first-degree students only. Continuation covers students who started between 2019-20 and 2022-23; completion, those who started between 2016-17 and 2019-20. The regulator’s minimums are 80% for continuation, 75% for completion and 60% for progression. Source: Office for Students. Some universities now teach most of their students this way. In 2024-25 the share was 77% at Suffolk, 74% at Canterbury Christ Church, 73% at Bath Spa and 69% at Leeds Trinity (HEPI); Bath Spa’s own accounts give 23,645 of its 32,142 students.
Three times in 25 years
None of this is new. Britain has paid for people who weren’t studying at least twice before, and the pattern repeats: the money follows the student, the middlemen are paid per head, and attendance is vouched for by whoever is being paid.
The 2013 colleges, then and now
In November 2013 ministers suspended student funding at 23 fast-growing private colleges. Several have since taught students on loans as partners of universities, according to the regulator’s figures for 2019 to 2023. Some of their results are good.
Suspension list: Times Higher Education, 29 November 2013; the suspension followed a rapid rise in student numbers. Results: Office for Students partnership data (continuation, full-time; first degree unless marked). This shows the same colleges carrying on, not wrongdoing. Where the company name differs from 2013, I have marked the link as unconfirmed.
What nobody measures
The questions the official figures can’t answer.
The jargon, explained
- Benchmark
- The result the regulator would expect from a course, given the backgrounds of the students it recruits. A course can miss the minimum yet beat its benchmark.
- Completion
- Whether students finish their course. The regulator’s minimum for full-time degrees is 75%.
- Continuation
- Whether a student is still studying a year after starting. The minimum is 80%.
- EU worker route
- EU citizens with pre-settled status who have lived in the UK or EEA for three years can get full student finance, including the living-costs loan, if they keep working throughout their course. The government’s rule of thumb is ten paid hours a week.
- Franchised course
- A university registers the student and awards the degree, but pays a private college to do the teaching, and keeps a share of the fee (12.5% to 30% in the cases the National Audit Office examined). Officially, subcontracted or partnership provision.
- Maintenance loan
- Money for living costs, paid to the student in instalments through the year. The lower the household income, the more you can borrow.
- Progression
- Whether graduates are in a professional job or further study 15 months after finishing. The minimum is 60%.
- Repayment rate
- The share of borrowers past their repayment date who paid anything that year. On the current plan, graduates repay 9% of earnings above £25,000 and anything left after 40 years is written off, so a low rate isn’t the same as default.
- Tuition fee loan
- Pays the course fees, up to £9,535 a year in 2025-26. It goes straight to the university, never to the student.
- Weekend-only course
- Under the 2011 regulations, attending only at weekends or in the holidays counts as distance learning, which doesn’t qualify for a maintenance loan.
Data and method
- The nationality figures are the Student Loans Company’s live management information, as given in parliamentary answers. They can be revised, so other sources may give slightly different totals for the same year.
- The parliamentary spreadsheet for amounts lent is headed “£000s”, but its figures are in pounds, rounded to the nearest £1,000. I show them in £m.
- Average loan is the amount lent divided by the number of borrowers in the same year, taken from two separate answers.
- The department defines the repayment rate as “the number of borrowers who have a loan which has reached its statutory repayment date who have made a payment in the financial year as a percentage of the total number of borrowers who have a loan which has reached its statutory repayment date”.
- Where I compare the weekend cohort (2025-26) with borrower totals (2024/25), the years differ, so the comparison is indicative.
Part of Outgoings, a deKilpeck project by Harry Pye. Corrections and new data are welcome: harry.pye@dekilpeck.co.uk.