It’s September and the students are going back to Uni. Half a million new graduates will be getting their student loan documents.
I have been writing about student loans for a while. Back in 2018 I wrote The Great Student Loans Swindle, followed a month later by Student Loans Swindle 2: Cash from Chaos. Earlier this year I returned to the subject in Student Loans: The Accounting Trick That Got Out of Hand. I had thought I might be running out of things to say about them, but apparently not.
In July the House of Commons Treasury Committee published the results of its inquiry into student loans. More than 52,000 people responded, making it one of the largest responses ever received by a Select Committee inquiry. Its conclusion was that the student loan system was “unfair and broken”. More remarkably, it concluded that some of the ways in which student loans had been promoted to young people amounted to “mis-selling”. Mis-selling is quite a word to use about a financial product promoted by the Government, particularly when most of the people being sold it were 18 years old.
The student loan mis-selling scandal
Plan 2 student loans were introduced alongside the increase in maximum tuition fees to £9,000 in 2012. The proposition was relatively straightforward. Graduates would repay 9% of their earnings above £21,000, any balance remaining after 30 years would be written off, and interest would vary according to income, reaching RPI plus 3% for higher earners.
There was another important part of the deal. When the new system was announced, the Government said that the £21,000 repayment threshold would be “uprated annually in line with earnings from April 2016”. This wasn’t some incidental detail. If you are going to repay 9% of everything you earn above a threshold for potentially 30 years, what happens to that threshold over those 30 years makes an enormous difference to how much you eventually pay.
It didn’t happen. The £21,000 threshold remained frozen in 2016/17 and 2017/18. Theresa May’s Government subsequently reversed the freeze, increasing the threshold to £25,000 in 2018 and restoring annual earnings indexation. It then started happening again. The threshold reached £27,295 in 2021/22 and was frozen there for three years. At the same time the Government changed the default basis for future increases from average earnings to RPI inflation.
The threshold eventually increased to £28,470 in April 2025 and £29,385 in April this year. It is now going to be frozen again. Under current policy, Plan 2 borrowers will continue to repay 9% of everything they earn above £29,385 for three years from April 2027. Without the freeze, the Institute for Fiscal Studies estimates that the threshold would have reached about £32,265 by 2029/30.
This isn’t an obscure administrative adjustment to the student finance system. It is a decision to make graduates pay more. An affected borrower earning above the threshold will pay an additional £93 in 2027/28, rising to £259 a year by 2029/30. More importantly, the effect doesn’t necessarily disappear when the freeze ends. Borrowers can continue making higher repayments until either their loan is cleared or it is eventually written off.
The IFS estimates that the combined effect of the latest threshold changes will increase average lifetime repayments for the 2022 university-entry cohort by around £3,000. The effect isn’t evenly distributed. Some lower-middle lifetime earners lose considerably more, with borrowers in the third decile of lifetime earnings estimated to repay around £5,000 extra.
There is also another freeze which attracted rather less attention. Plan 2 interest varies according to income. Borrowers currently earning £29,385 or less are charged 4.1%, with the rate increasing until it reaches 6% for people earning £52,885 or more. The Government is freezing those interest thresholds as well. Oddly, this wasn’t explicitly announced in the Treasury’s Budget documents, although it did appear in the OBR costings. Kate Ogden of the IFS told the Treasury Committee that the two freezes together generated around £1 billion of savings. When the Committee asked the Chief Secretary to the Treasury about the second freeze, she said that she wasn’t aware of it, which is reassuring.
All of this matters because of what students were originally told when they took out the loans. The Treasury Committee identified three particular problems. Department for Education videos and presentations promoting student loans didn’t make sufficiently clear that future governments could retrospectively alter their terms. Some promotional material compared student loan repayments with relatively mundane monthly expenditure such as mobile phone contracts or trips to the cinema, which the Committee concluded could substantially understate the burden for higher earners. Even during the Student Loans Company application process, borrowers weren’t given sufficiently prominent warning that governments could subsequently change the terms. The Committee concluded that these practices amounted to mis-selling.
Imagine Barclays doing this. It lends an 18-year-old tens of thousands of pounds, explains how repayments will be calculated and says that the repayment threshold will increase with earnings. Several years later it decides not to increase the threshold, meaning that its customer has to repay more. I suspect that the Financial Conduct Authority might take an interest.
Student loans are different, because they aren’t regulated in quite the same way. Indeed, during the inquiry the Government and the Student Loans Company managed to disagree about something fairly fundamental. The Government referred to borrowers signing a loan contract. The Student Loans Company pointed out that student loans aren’t actually contractual loans in the conventional sense. Their terms are determined by legislation, which means that Parliament can change them. The Treasury Committee rather reasonably suggested that the Government and its own Student Loans Company might like to agree on whether or not their nearly £300 billion loan book consists of actual loan contracts.
The Government published its response this week and has agreed to change the information provided to future students. Prospective borrowers will be told more explicitly that the terms of their loans are governed by legislation and can therefore be changed by future governments. Parliament has concluded that student loans were mis-sold partly because students weren’t adequately warned that governments could change the deal later. The Government’s solution is to make sure that future students are warned that governments can change the deal later.
The Government rejected the Committee’s recommendation that student-loan promotions should be subject to the same FCA Consumer Duty standards expected of private lenders. It also rejected the recommendation to stop using RPI to calculate student-loan interest. It hasn’t quite said no, however, to reversing the latest freeze in the £29,385 repayment threshold. The Treasury Committee says that there is a “moral obligation” to reverse it, while the Government says that the student-finance system is under review. I will be watching the budget to see if we get an announcement at the next fiscal event.
Whether the freeze survives or not, the argument about it exposes something much more fundamental about the student-loan system. We keep talking about people repaying loans, but that isn’t really how the repayment system works.
Student loan repayments aren’t really loan repayments
Suppose I owe my bank £5,000 and you owe the same bank £150,000. We both earn £40,000. It would be rather strange if the bank required both of us to make exactly the same monthly repayment. How much you borrowed, how much you still owe and the interest being charged are normally fairly important factors in determining how much you repay on a loan.
Student loans don’t work like that. The Government’s own guidance says that the amount you owe has no impact on how much you repay each year. Instead, repayments are determined by your income. Plan 2 borrowers currently pay 9% of everything they earn above £29,385. Plan 5 borrowers pay 9% above £25,000. Someone with a postgraduate loan pays another 6% above £21,000. The money is normally deducted from salary through payroll alongside income tax and National Insurance. If your income falls, your repayments fall; if it falls below the threshold, you pay nothing; if your income increases, your repayments increase. Until you reach the point where you actually clear the balance, how much you owe makes almost no difference to what comes out of your salary.
There is a name for a compulsory payment to the Government which increases according to how much you earn.
It’s a tax.
The Treasury Committee has effectively reached the same conclusion, describing student loans for many borrowers as a 30- or 40-year supplementary income tax. Once you start looking at student loans in those terms, some of our supposedly familiar income-tax rates begin to look rather different.
Take someone in England with a Plan 5 student loan. In 2026/27 they start repaying the loan once they earn £25,000. Between £25,000 and £50,270, each additional pound they earn is subject to 20% income tax, 8% employee National Insurance and a 9% student-loan repayment. They therefore lose 37p of the next pound they earn and keep 63p.
Once they earn more than £50,270, income tax increases to 40%, National Insurance falls to 2%, and the student-loan deduction remains at 9%. For every additional pound they earn, 51p disappears and they keep 49p. We describe this person as paying 40% income tax, which is legally correct, but it doesn’t provide a particularly accurate description of what happens to their additional earnings.
It gets worse if somebody has both an undergraduate and a postgraduate loan. The postgraduate loan takes another 6% of earnings above its threshold. Between £25,000 and £50,270 the marginal deduction can therefore reach 43%; above £50,270 it reaches 57%.
Then we arrive at one of the sillier parts of the British tax system. Once somebody earns more than £100,000 their personal allowance is withdrawn at a rate of £1 for every additional £2 of income, producing an effective 60% income-tax rate between £100,000 and £125,140. Add 2% National Insurance and a 9% Plan 5 student-loan repayment and the marginal deduction becomes 71%. Add a postgraduate loan and it becomes 77%. For every additional £100 earned, that graduate can be left with £23.
There are qualifications to this. Pension contributions and individual circumstances can change the calculation, and student-loan repayments aren’t legally income tax. But if we’re interested in the economic effect on someone’s payslip rather than the terminology attached to each deduction, those are the numbers.
Looking at student loans as a supplementary income tax also explains why changes to the repayment thresholds matter so much. Freezing the Plan 2 threshold is economically the same as freezing an income-tax allowance. As wages increase, more income is dragged above the threshold and the Government collects 9% of it. It is fiscal drag for graduates. The useful bit, from the Treasury’s point of view, is that because the additional deduction is described as a loan repayment rather than a tax, a Chancellor can increase the amount collected from millions of graduates without announcing an increase in the basic or higher rates of income tax.
It also explains one of the stranger arguments about student loans: the obsession with the interest rate. For some borrowers the interest rate barely matters. If somebody is ultimately going to reach the end of the repayment period with a substantial balance outstanding, adding another £10,000 of interest to that balance doesn’t necessarily cost them another £10,000. They were never going to repay it anyway. Conversely, reducing the interest rate can disproportionately benefit higher earners, precisely because they are the borrowers who are most likely to repay their loans in full.
Plan 5 partly recognises this. Its interest rate is normally RPI rather than the potential RPI plus 3% charged under Plan 2. The current Plan 5 rate is 4.1%. That sounds considerably more generous, but the other side of the bargain is that Plan 5 graduates start repaying at £25,000 rather than £29,385 and can continue paying for 40 years rather than 30. The Government has reduced the importance of the enormous nominal balance accumulating interest while extending and broadening the thing which actually matters to most graduates: the 9% deduction from earnings.
Plan 5 therefore looks even more like a graduate tax. The Government currently expects 55% of full-time undergraduates beginning courses in 2025/26 eventually to repay their loans in full, considerably more than under Plan 2. Even so, 45% still aren’t expected to repay everything, despite potentially making repayments for 40 years.
This brings us back to the other reason why student loans aren’t really loans. They don’t behave much like conventional loans for the Government either.
Student loans aren’t really loans for the Government either
At the end of the 2025/26 financial year, outstanding English higher-education student-loan balances had reached £294.6 billion. A year earlier they stood at £266.6 billion, an increase of £28 billion in twelve months. Back in 2013/14 the entire balance was £54.4 billion. Undergraduate lending during 2025/26 alone was £20.5 billion. Including postgraduate lending, total higher-education lending was £21.3 billion, while total student-loan outlay was £21.4 billion. The Department for Education expects annual student-loan outlay to reach £25.2 billion by 2030/31.
These are enormous numbers, and they help explain why governments became so enthusiastic about student loans in the first place. Suppose the Government gives a university £10,000 to educate somebody. That looks like £10,000 of government spending. Suppose instead that the Government “lends” a student £10,000, which is immediately transferred to the university. The Government has still handed over the same £10,000, but it has also created a £10,000 financial asset representing the student’s promise to repay the money at some point in the future.
For years, the accounting treatment concentrated heavily on that second part of the transaction. Government lending was treated as the acquisition of a financial asset rather than ordinary expenditure. There was an obvious problem: the Government knew perfectly well that a substantial proportion of the loans would never be repaid. Nevertheless, the accounts treated the loans as assets and even accrued interest as government income where there was little realistic prospect that some of that interest would ever actually be received.
This was the subject of my original Great Student Loans Swindle article in November 2018. I argued that billions of pounds of what was economically government expenditure had effectively disappeared from the deficit by being transformed into student loans. A month later something slightly inconvenient happened for the Government. The Office for National Statistics decided that the existing accounting treatment didn’t adequately reflect economic reality.
The ONS adopted what it calls the “partitioned loan-transfer approach”. Behind the slightly impenetrable name is a remarkably straightforward idea. When the Government lends £1 to a student, estimate how much of that pound will ultimately be repaid. The part expected to come back is genuinely a financial asset and can reasonably be treated as a loan. The part which isn’t expected to come back is government expenditure, recorded as a capital transfer. The ONS describes this as conceptually equivalent to cancelling that portion of the loan at the moment it is issued.
The change also stopped the rather wonderful practice of the Government recognising as income interest which it didn’t realistically expect ever to receive. When the new accounting treatment was incorporated into the public finances in 2019, it increased recorded public-sector net borrowing for 2018/19 by £12.4 billion. That isn’t exactly a rounding error. When I described the original system as an accounting swindle, I may actually have been being a little restrained.
There is an important distinction to make here. It would be wrong to claim that the Government is still simply hiding all student-loan expenditure off the books. The ONS fixed the most egregious part of that problem. Student loans are now split in the national accounts between genuine lending and government expenditure according to the proportion expected to be repaid.
That makes the history of the policy more interesting rather than less. The student-loan system was dramatically expanded under an accounting treatment which allowed huge amounts of government support for universities to appear as the creation of financial assets rather than spending. Eventually the statistical authorities called time on the fiction. The policy survived, and governments instead repeatedly changed its terms so that graduates would repay more of the money.
We can see the consequences in the Government’s own forecasts. For every £1 of new full-time Plan 2 lending in 2025/26, the Department for Education estimates that the Government will ultimately subsidise 39p. For Plan 5 it expects to subsidise 33p. Even after reducing the repayment threshold, extending the repayment period to 40 years and substantially increasing the proportion of graduates expected to clear their balances, the Government doesn’t pretend that all of these “loans” will come back.
The nearly £295 billion headline figure for outstanding English higher-education student debt is therefore simultaneously enormously important and rather misleading. Some of it is a genuine financial asset which graduates will eventually repay. Some of it is expenditure which the ONS recognises from the outset will never come back. For many individual graduates, meanwhile, the precise size of their nominal debt makes little difference to what comes out of their salary each month. Their income determines that.
So what exactly is a student loan?
This leaves us with something rather peculiar. The Government calls student finance a loan, but Parliament has just concluded that some of those loans were mis-sold because borrowers weren’t properly warned that governments could retrospectively change their terms. Normal commercial consumer protections don’t apply in quite the same way because student loans aren’t conventional contractual loans; their terms are determined by legislation and can be changed by Parliament.
For many graduates, meanwhile, repayments behave much more like an additional income tax. They pay 9% above an earnings threshold, usually through payroll, regardless of whether their outstanding balance is £20,000 or £100,000. From the Government’s perspective, the ONS says that the proportion which isn’t expected to be repaid isn’t really lending at all. It is government expenditure.
We have therefore achieved something quite remarkable. To the graduate, much of a student loan behaves like a tax. To the national accounts, a substantial part of it is government spending. Its terms can be changed by Parliament because it isn’t a conventional commercial contract. Yet throughout the entire process we continue to insist on calling it a loan.
Perhaps that is the real student loans swindle. It isn’t simply that the original accounting was manipulated; the ONS eventually fixed that. It isn’t simply that graduates pay a lot of money, because somebody has to pay for universities. It isn’t even simply that governments have retrospectively changed repayment terms, although Parliament now regards the way some of those loans were originally promoted as mis-selling.
The more fundamental problem is that we have spent more than a decade pretending that student finance is something it isn’t. If we want graduates to pay an additional 9% of their earnings for 30 or 40 years to fund higher education, we can make that argument and call it a graduate tax. If we think taxpayers should meet a third of the cost of higher education because society as a whole benefits from having doctors, engineers, teachers, scientists and, heaven forbid, economists, we can make that argument too and call it government spending.
What we shouldn’t do is construct a system which behaves like a tax when graduates repay it, behaves partly like government expenditure when the national statisticians account for it, can have its terms retrospectively changed in ways that wouldn’t be acceptable for an ordinary commercial loan, and then insist that an 18-year-old has simply borrowed some money.
Because it isn’t really a loan. It never was.