Student loans are still broken and unfair

The government has “replied” to the Treasury Committee’s Student loans: Broken and unfair?

The report concluded – with the help of 52,000 survey responses, two ministers and the Chancellor’s own vocabulary – that the system is both.

HM Treasury (HMT) and the Department for Education (DfE) have written back jointly (Whitehall’s version of divorced parents) and the result runs to nine pages.

Eleven recommendations get an answer, sixteen numbered conclusions get nothing at all, and the upshot of what has been accepted is that the Student Loans Company (SLC) will change a bit of wording on its application journey in the spring of 2027.

It opens as these things do, by welcoming the report and “the detailed consideration it has given to the operation of the student finance system”, before explaining that the costs and benefits of that system “must be considered in the round, taking account of the interests of students, graduates, higher education providers and taxpayers alike”.

The evidence for the “round” turns out to be Universities UK’s own £265 billion economic impact figure, which is a nice touch – a Treasury document defending the graduate share of costs by citing the sector lobby’s estimate of how much the sector is worth.

The Committee had asked whether the deal was fair to the people repaying. The government answers a different question – which is whether the deal is good for everybody else.

Ninety five, thirty five, whatever

The headline recommendation was that the split between graduate and state should go back to 50:50 in the long term, on the basis of evidence from the Institute for Fiscal Studies (IFS) that for recent cohorts the taxpayer’s share of financing higher education, grants included, has fallen to around 3 per cent, and of Philip Augar telling the Committee that individuals now hold “something like 95 per cent of the responsibility, and at least 75 per cent”.

Unsurprisingly, the government “does not accept this recommendation”. Its counter is that:

…as Minister Smith explained to the Committee, we estimate the overall government contribution to be in the region of 35–40%… depending on year of issue.”

That’s a gap of more than thirty percentage points between two figures, and at no point does the response acknowledge that the Committee’s number exists, let alone try to reconcile it – which is more than a smidge ironic.

The government’s number is the RAB charge, the accounting estimate of the share of a single financial year’s lending that won’t come back, discounted at the Treasury’s chosen rate and resting on the Treasury’s chosen earnings assumptions, with teaching grants and the strategic priorities grant bolted on to get from thirty-odd to the advertised 35–40 per cent. DfE’s own statistics are explicit that this is a financial year measure which includes “student loan borrowers across multiple cohorts who receive outlay in the same year”, and in 2025-26 roughly three quarters of that £21.4 billion of outlay is Plan 5 lending to people who started in 2023 or later.

The Committee’s number is a cohort measure. The IFS took everybody who started in 2022-23 – the last Plan 2 intake – and worked out what the state will have contributed to their higher education across their working lives once their repayments are counted in real terms, and the answer was £0.7 billion out of £23.8 billion, with the loans themselves running at a long-run surplus of £0.8 billion to the Exchequer.

Both things can be true at once, because they are not measures of the same thing. One is a forecast of what this year’s lending will cost under today’s terms. The other is what a particular group of borrowers will actually hand over under the terms as they now stand – which are not the terms they signed.

That last clause is the whole bloody point. The 2022-23 cohort’s number is close to nothing because of what has been done to Plan 2 since those loans were issued – the four year freeze of the repayment threshold, the February 2022 decision to switch its uprating from earnings growth to inflation for good, the fresh three year freeze from April 2027, and the undisclosed freeze of the interest rate bands that went with it. IFS and London Economics both put the effect at something like £13,400 more over a lifetime for men in that cohort and £16,900 for women, set against the terms they were sold. DfE puts it more delicately in the small print of its forecasts, observing that previous cohorts on Plan 2 “were charged interest above RPI so could repay more than 100% of their loan in real terms”.

When ministers quote 35 to 40 per cent, in other words, they are quoting the subsidy on the loans the state has not yet got round to fiddling with the terms of.

It would be useful at this point to check the ones it has, and you can’t. DfE’s student loan forecasts, republished on 9 July, two days after the Committee reported, do carry a cohort table – average balances, average lifetime repayments and the proportion of outlay repaid in real terms, broken down by lifetime earnings decile – but only for students starting in 2025-26, every one of them on Plan 5, of whom the lowest earning tenth will repay 8 per cent of what they borrowed and the cohort as a whole 69 per cent. The equivalent table for the final Plan 2 intake appeared in the 2023 edition and has not been published since. For the eleven entry cohorts between 2012 and 2022, five and a half million borrowers, there is nowt. I wonder why.

What there is for Plan 2 is a RAB charge of 39 per cent for 2025-26, forecast to reach 50 per cent by 2028-29 – but that is the charge on new Plan 2 outlay, the £4.6 billion still going out to students who started before August 2023 and haven’t finished yet, and it says nothing at all about the loans already made. Beyond that there is a single stock charge of 41 per cent covering the entire outstanding Plan 2 book on the same discounting that produces the 30 per cent, blended across every cohort at once. Nothing published shows what any Plan 2 year group was scored at when it borrowed set against what it is now expected to pay. The one statistic that would reveal the effect of a retrospective change on the people it was done to – surely the stat the inquiry ought to have demanded – is the one the department does not produce.

Oh, the irony. A subsidy quoted at the point of lending is a forecast made under terms that the government spends three pages of the same document defending its right to change, on the grounds that it “needs to be able to adapt the system to changing economic circumstances”. Adapting the system is precisely how a third at issuance becomes 3 per cent by the end – it’s not an unfortunate side effect of the design, it’s the mechanism. The Committee asked for contracts so that the number quoted to somebody when they sign is the number that survives, and the government refused, in the same nine pages in which it offers that number as proof of its own generosity.

Half and half

Moving to 50:50, meanwhile, “would carry a substantial fiscal cost”, we are told, though not how substantial.

The 50:50 rejection also leans on the IFS finding from June that the average graduate ends up around £100,000 better off over a lifetime after tax and loan repayments. The same IFS study found a quarter of graduates financially worse off for having gone, one in ten men more than £90,000 worse off, and four in ten degrees not paying for themselves from the Treasury’s point of view.

The average does a lot of work in that paragraph, and the response does at least concede that “outcomes and returns vary across individuals and subjects of study”, before promising to keep the evidence “under review”.

Then there’s the freeze. The Committee said the government “must reverse the repayment threshold freeze in the Autumn Budget”, costed the reversal at £355 million a year by 2029–30, and said ministers had “a moral obligation” to do it in order “to honour the terms and conditions under which those loans were sold to students”.

The response “recognises the cost-of-living challenges faced by many graduates, including those with Plan 2 student loans”, and then explains that repayment decisions “must be considered alongside wider fiscal priorities, the long-term sustainability of the higher education funding system, and the need to ensure value for money for taxpayers”. It’s off the table.

On interest rates the Committee “reiterated” the recommendation its predecessor made in 2018 to drop RPI for CPI and said it was “disappointed that eight years later this recommendation has not been implemented”. The government “notes” it, points out that the ONS will fold CPIH methods into RPI from February 2030 anyway, and will “continue to consider the appropriate treatment of inflation measures”.

Thus the answer to eight years of not doing it is four more years of not doing it, after which the problem solves itself. Glacial.

Quack quack oops

The Committee’s most serious recommendation was that future loans should be issued as contracts rather than under statute, so that governments could no longer rewrite the terms of an existing loan without paying compensation.

The response rejects this at length – and in doing so gives part of the game away. Student loans are “unlike any other form of conventional loan”, the taxpayer subsidy is “a conscious and important investment in the long-term skills capacity, people and economy of the country”, and:

To ensure that the taxpayer contribution to student loans remain financially stable in the long term and available to all as intended by Parliament, the Government needs to be able to adapt the system to changing economic circumstances.”

The examples offered of this adaptability are the Plan 2 threshold rising to £28,470 in April 2025 (“its first increase since 2021”) and to £29,385 this year. Those are the two occasions in a decade on which the power has been used in graduates’ favour, both of them restoring indexation after a four year freeze, cited in a document that declines to reverse the next freeze of the very same threshold.

They’re also utterly irrelevant, because the Committee’s recommendation was for contracts that would stop governments changing terms “retrospectively to the detriment of the borrower without paying compensation”.

Nobody has ever objected to a lender writing to say your repayments are going down. The only flexibility a contract would take away is the flexibility to make things worse, so the government has defended its power to make things worse by pointing to the two times it didn’t. It is also the power that turns 35 to 40 per cent into 3 – so defending it and citing the 35 to 40 per cent in the same document is quite the achievement.

The contractual route is also dismissed as “administratively prohibitive” because it “would require the Secretary of State to enter and manage individual agreements with large numbers of borrowers”.

Two recommendations later, resolving what the Committee thought was a contradiction between DfE calling the loan a contract and the SLC saying it wasn’t, the response explains that the 2011 Regulations “require eligible students to enter into a loan agreement with the Secretary of State”. In other words, the Secretary of State already manages individual agreements with millions of borrowers – it’s just that the government wants agreements it can vary.

Best of all is the reason given for keeping the Financial Conduct Authority’s Consumer Duty away from loan marketing. That regime was “developed for commercial financial products offered within competitive markets”, whereas:

The income-contingent repayment system means that monthly loan costs are independent of both the amount borrowed and the interest rate, which makes them very different to commercial loans.”

Well, yeah. It walks like a tax and quacks like a tax, and here the government is explaining that it quacks, in order to argue that the people who were sold it as a loan don’t need the protections you’d get if you’d bought a loan. The Committee, lest we forget, found that DfE’s own roadshow slides compared repayments to a mobile phone contract and cinema tickets and said in terms that this “amounted to mis-selling“, three times over. The response never uses the word.

What is accepted is that the SLC will make it “more prominent that student finance is governed by legislation and that regulations may be amended by Government and Parliament”, that its “speedbumps” will say so from spring 2027, and that DfE and the SLC will sort out which of them is legally responsible for the application form in time for the 2027–28 cycle.

In other words the one part of the report the government likes is the part that tells future borrowers, clearly and up front, that the terms can be changed on them afterwards. The Committee offered disclosure as the fallback if contractual protection was refused. The government has taken the fallback and dropped the protection, and from 2027 every applicant will have been warned.

Potentially misleading

Recommendation 27 asked for something small – wording on the annual statement giving borrowers “an approximate indication of how much of their total loan is likely to be written off”. The Committee had anticipated the objection, saying in conclusion 26 that it disagreed “with the government that additional information would confuse loan holders”, and had identified the harm – people making voluntary repayments on balances that would have been written off anyway.

The response agrees “more can be done” and then says no, because any such forecast “would require too many assumptions about an individual’s future circumstances, including career progression, earnings growth, periods of childcare or caring responsibilities, changes in working patterns etc” and so “would be, potentially, misleading”. Instead there’ll be some illustrative material for prospective borrowers showing how a career break affects a repayment trajectory. The graduates who are confused now get nothing, the applicants who haven’t signed yet get a PDF.

Three things make that harder to take seriously than it already is. The first is that on 27 August, eleven days before the response was signed, DfE published an ad hoc statistical release delivering half of what Jacqui Smith promised the Committee in June – a table of the proportion of borrowers in each earnings band forecast to repay in full at three, five and ten years after becoming liable.

The other half, how much people at different earnings levels actually end up paying, and the seven deciles below the three John Glen was told repay in full, didn’t make it. Five years out, 50 per cent of Plan 2 borrowers earning £45,000 to £50,000 are forecast to clear the loan, rising to 89 per cent above £80,000, while for Plan 5 it’s 56 per cent at £25,000 to £30,000 and 95 per cent above £80,000.

That is an approximate, band-level indication of exactly the kind the Committee asked for, built, assured and published as official statistics by the same department that told the Committee it would mislead. The response doesn’t mention it.

The second is that the government itself relies on the same forecasts, made for everyone at once, to book billions against the fiscal rules. The modelling is reliable enough to score in a Budget and too unreliable to show a graduate a range.

The third is that this has all been done before. In 2019 MoneySavingExpert and the Russell Group piloted a redesigned “graduate contributions statement” that led with what you’d earned and paid, demoted the balance to the foot of the page, projected your likely future contributions from your own HMRC data and ONS earnings trajectories, and then listed its assumptions – including that career breaks would mean paying less than shown – on a page of their own with a warning attached.

It was designed around the government’s objection seven years before the government made it. Martin Lewis said at the time that the existing statement was “misleading and dangerous” and had “panicked many into poor decisions that cost some £1,000s”. The government’s position in 2026 is that the cure would be misleading, while the disease is fine.

Crap in

None of this is a surprise, and it isn’t entirely the government’s fault that the response is this thin. The Committee’s report catalogued every retrospective change since 2016 and noted they all went the same way. It got onto the record that the OBR scores the threshold freeze as a one-off £5.6 billion improvement in 2026–27, booked in the year of enactment on the strength of repayments graduates won’t make for decades. It surfaced a second, undisclosed freeze of the interest rate income bands worth around £1 billion that the Chief Secretary to the Treasury said she wasn’t aware of. And then failed to ask why any of that keeps happening.

The answer, of course, is that making the terms worse frees up money in the year you do it. Because the freeze is scored as a capital transfer at the point of enactment, the OBR records it as a £5.6 billion reduction in borrowing in 2026–27 – headroom against the fiscal rules that the Chancellor can spend on something else this year.

Making the terms better does the reverse, writing down the value of the loan book in year and eating headroom before a single graduate has paid a penny less.

So successive governments have pulled the one lever the accounting pays them to pull, and the graduates picking up the tab are the only people in the transaction who don’t get to see the ledger.

That is a fiscal problem, and the one body in Parliament whose remit is fiscal accounting – the discount rate, the RAB charge, the ONS partition, the way a terms change scores against the fiscal rules – is the Treasury Committee.

It has investigated what has been done, but not why HMT has been doing it. Having diagnosed a communications problem, it recommended communications remedies, and communications remedies are the ones a government with the incentive intact can accept for free, which is exactly what it has done.

The Committee had all of this in front of it but wrote a report about form wording. The government has accepted some of the form wording.

Who pays for the headroom

Which brings it back to the question the Committee’s started with. The response’s whole answer on fairness is that the system “is designed such that lower earners receive a larger subsidy, and higher earners make a larger contribution”, which is true in pounds and beside the point.

The Committee’s evidence was that only the top three income deciles of the 2024–25 intake will ever repay in full, and that the people at the top, clearing early, pay less interest and can end up handing over a smaller share of what they borrowed than middling earners who pay 9 per cent for the full term. Larger in cash and larger as a proportion are different claims, and the response answers a challenge framed in the second with the first.

Look at who the year’s two adaptations actually land on. Freeze the threshold and everyone above it pays more each month – but for a high earner who was going to clear the loan anyway, paying more per month just means clearing it sooner, and the lifetime cost barely moves. For the teacher or the nurse on £35,000 who was never going to clear it, every extra pound taken before the write-off is a pound of lifetime cost they didn’t have last year.

Under the much-trumpeted interest rate cap, the reverse happens – the Committee’s own conclusion 10 says the 6 per cent cap “will benefit only students who will pay back their loan in full”. So the tightening that produced £5.6 billion of headroom is paid for by the middle, and the loosening that costs next to nothing is enjoyed by the top.

It’s why “cancelled at the end of the loan term with no detriment to the borrower”, which the response reaches for twice, grates. For someone who has paid 9 per cent above the threshold for 30 or 40 years the write-off isn’t a gift, it’s the point at which the levy stops, and calling it a subsidy at “no detriment” to a person who may have repaid more than they borrowed only makes sense if the balance is a real debt – which the same document says doesn’t determine what anyone pays.

The Committee asked whether the system was broken and unfair. It still is.