Over the past year, pretty much everyone seems to have come round to the idea that the way we finance higher education needs change.
The Chancellor called student loans “broken”. The Education Secretary wants it to be “fairer”. The Labour deputy leader called repayments “endless and unfair”.
We’re reminded of all this and more in the Treasury Committee’s Student loans: Broken and unfair?, its first report of the 2026–27 session.
The Committee seems to agree with all of them, but its core recommendation turns out to be tinkering rather than transformation – reversing Reeves’ three-year threshold freeze worth £355 million a year by 2029–30 which kicked all of this off.
That gap – between the size of the verdict and the size of the remedy – is the whole report in miniature.
Broken and unfair, you say
The Committee’s method here is mostly to hold ministers to their own words. Rachel Reeves called the system broken at her Mais lecture in March. Baroness Smith called it unfair in an April press release announcing the 6 per cent interest cap. The Committee’s conclusion, in as many words, is that it’s untenable for government to say these things publicly and then not act on them in successive Budgets.
Which is true, and also not much of a finding. The government told the Committee the system was broken before the Committee had taken any evidence. What the Committee adds is 52,000 survey responses – “one of the largest ever responses to a Select Committee survey,” as it says of itself – plus 300-odd written submissions, along with two oral evidence sessions that between them produce most of what’s actually useful here.
A helpful and often counter-intuitive finding in the report comes from DfE’s own modelling – for the 2024–25 student intake, only the top three income deciles will repay their loan in full. Everyone else has some or all of their balance written off after thirty or forty years, depending which plan they’re on.
That means most of the popular demands – cut the interest rate, cut fees, shrink the loan – help only the people who were going to pay it all back anyway.
Witnesses walked the Committee through this decile by decile – cut interest rates and the highest earners pay off faster, some people at the margin tip into full repayment, and everyone below that simply never notices, because the number that gets written off shrinks instead of the number they actually pay each month.
Cut fees and the same thing happens. Shorten the write-off term, on the other hand, and it’s the lower earners who benefit, because their repayments simply stop sooner – the top decile, having already cleared the debt years before, feels nothing at all.
The Committee’s own conclusions land on this too, more or less – the government’s current interest rate cap at 6 per cent is “a step in the right direction” but “will have no impact on the majority of students because the majority have their loan written off before they finish paying it off.”
Written off in the stars
On the great hybrid see-saw, the sense here is that a tilt towards characteristics of a tax rather than a loan is the way forward – but that would mean the usual anger on interest and notional headline debt levels. So it’s coupled with some recoms on the burden overall.
The 60:40 graduate:state split that Vince Cable promised the Commons in 2010 has, on the IFS’s current modelling, become something close to the opposite. For the 2022–23 cohort, the IFS’s Kate Ogden told the Committee the long-run cost to the taxpayer is actually negative – minus £0.8 billion – meaning graduates as a group are expected to repay more in real terms than they borrowed.
Of the total cost of financing that cohort’s higher education, including direct grants to universities, only 3 per cent is projected to be met by the taxpayer. The other 97 per cent falls on students.
Philip Augar, who chaired the 2019 Post-18 Review and recommended a 50:50 split, told the Committee that individuals now hold “something like 95 per cent of the responsibility, and at least 75 per cent.” The Committee’s headline recommendation is to get back to 50:50 in the long term.
Fifty shades of graduate
The problem is that even 50:50 for the cohort doesn’t answer how far you push one end of the see-saw to relieve the burden for the other end. Within a 50:50 cohort split, you can have everyone pay back half, or have poorer grads pay back 5 per cent while the richer ones pay back 95 per cent.
The Committee’s recommendation 20 explicitly punts on subject-level cost variation – “we have not scrutinised the relative value of different higher education courses” – on the grounds that it sits outside Treasury’s remit, and hands it instead to DfE and the Education Select Committee.
As a result even a verty-hard-to-get-to 50:50 split on all the existing debt pile is a single number standing in for wildly different underlying realities depending on subject, institution and mode of delivery, and the Committee’s own evidence base – the IFS figures it relies on for the aggregate – already implies that variation exists without the Committee going anywhere near it.
Institution-level and subject-level analysis of access and participation has been the more honest way to read this system for years – this report, however useful elsewhere, reinforces the aggregate-only framing instead.
The same remit problem shows up on the other side of the ledger too. Because Treasury oversees the graduate-repayment end of student finance rather than the maintenance and access end, the report has almost nothing to say about what a loan actually buys a student while they’re studying – whether the maintenance support that comes with it is enough to live on, whether the experience it funds bears any relationship to what full cost-sharing implies students should expect, or how the collapsing state contribution on the repayment side connects to a maintenance system that hasn’t kept pace with either rent or wages.
You can conclude, as the Committee does, that individuals now bear 95 per cent of the cost of their education without ever asking what that 95 per cent is actually purchasing for them day to day. That’s not a criticism of the Committee doing its job. It’s a structural gap in what any single inquiry, boxed into Treasury’s terms of reference, was ever going to be able to see.
Read the small print
The Watchdog-style evidence in the report comes from DfE roadshow slides, obtained after a BBC investigation and reproduced in the report itself, comparing loan repayments to a mobile phone contract – £15 a month on earnings of £27,000, set next to £14 for a mobile contract and £17 for cinema tickets.
None of the promotional material the Committee saw disclosed that government retains the power to change repayment terms retrospectively. Later versions softened to “all policies are kept under review,” and the Committee notes explicitly that “no emphasis was given to this text.”
Asked directly whether this amounted to mis-selling, Philip Augar said he was “thinking immediately of the car loans scheme or the payment protection insurance scandal, which produced exactly the outcome that you have described, yes.” Baroness Smith disagreed, arguing the terms and conditions “existed at the time” even if “difficult… to see.”
Ministers were defending disclosure that was technically present rather than disclosure that was actually noticed – which is exactly the distinction the “red hand doctrine,” invoked by the Committee via a Mishcon de Reya opinion, exists to catch out.
The Committee’s most serious recommendation follows from this – that future student loans should be issued as contractual agreements rather than statutory instruments, specifically so that government can’t alter the terms of an existing loan without paying compensation.
If that ever happened, most of the report’s other consumer-protection recommendations – disclosure wording, FCA Consumer Duty compliance, clearer “speedbumps” in the application process – would become largely unnecessary, because the legal exemption they’re all working around would disappear.
The report doesn’t flag that its own recommendations are, in effect, several patches sitting on top of the one fix that would make the patches redundant. Nor does it note that when repayments don’t come in at the value expected, the Treasury is going to fight very hard to keep its options open on fiddling with the terms to make graduates pay for what amounts to bad forecasting and a faltering economy.
Two ducks, one statement
The thing walks like a loan and quacks like a tax, and government has spent fifteen years insisting it’s a duck.
The Committee’s own preferred fix for the mis-selling and confusion problem is Recommendation 97 – get DfE and SLC to add wording to annual statements giving graduates “an approximate indication of how much of their total loan is likely to be written off.” It’s a sensible ask. It’s also a modest one – a single static probability, rather than anything that tracks a person’s actual circumstances as they change.
There’s an older way of putting this that explains why the confusion is so durable rather than just describing it. The system, structurally, behaves like a defined contribution pension and a defined benefit pension bolted together, and which half you notice determines whether you think you’ve been lied to.
The principal and the interest rate are the defined contribution bit – you borrowed a specific amount, it accrues, you watch the balance. The threshold, the repayment rate and the term are the defined benefit bit, run in reverse – some graduates end up paying back far more than their education cost, some far less, depending on their earnings rather than on what they borrowed.
The write-off term is the one variable that straddles both, because almost nobody hits it exactly – you either clear the balance early, or the remainder simply disappears at the end.
Politically, that hybrid design has always been useful to obscure, because – as has been argued on here before – the language of debt is far more legible, and far more frightening, than the language of an income-contingent threshold. Freeze or lower the threshold, extend the term by a decade, and on average the total contribution barely moves – but almost nobody notices that, because what people track month to month is the principal and the interest, the defined-contribution-shaped half of the duck.
Turning the volume up on that half, while adjusting the defined-benefit-shaped half underneath it, is how you can tighten the system on graduates in their twenties and again on middling earners in their fifties, while allowing the highest earners to pay proportionately less across their lifetimes – and have most of the public blame the interest rate for all of it.
Which is what makes Recommendation 97 feel like the wrong scale of fix. If the actual problem is that borrowers can’t tell which half of the duck they’re looking at, a single probability of write-off doesn’t solve that – it just adds one more static number to a pile of static numbers already misread as debt.
Compare this with what a defined contribution pension provider is required to do – give you an annually updated projection of what your specific pot is likely to be worth at a future date, recalculated as your contributions and the market actually move. Nothing in the report considers whether SLC could do the equivalent – an individually modelled, annually refreshed projection of a graduate’s own likely repayment path and eventual write-off point, rather than a generic proportion handed over once at the point of signing.
DWP already produces individualised state pension forecasts on broadly the same logic, so the precedent for government running person-specific financial projections isn’t hypothetical. It would be a harder technical ask than a pension projection, because it depends on forecasting one person’s future earnings rather than applying a stable actuarial assumption across a pool. That’s a reason to test the idea, not a reason nobody in this 97-paragraph report even raises it.
Cooking the loan books
None of this is new context for regular readers, and the account of the 2012 reforms as a private-money story, rather than a straightforward continuation of public spending routed differently, is the same one we’ve made before.
What this report adds is confirmation, via oral evidence rather than document analysis, that the OBR’s own November 2025 figures show the recent threshold freeze producing a one-off £5.6 billion improvement in the fiscal position – booked the year the legislation passes, not the year graduates actually pay more. Asked about a second, entirely undisclosed freeze of the interest-rate income bands, worth a further £1 billion and mentioned nowhere in Budget documents, the Chief Secretary to the Treasury told the Committee she wasn’t aware of it.
What the report never does is ask why that keeps happening. Every one of the retrospective changes documented here – the 2016 threshold freeze after a 2011 promise to uprate it, the 2020 freeze, the 2022 RPI-linked rise, the 2025 freeze again – moves in the same direction, because tightening terms on existing borrowers makes the loan book look healthier on the government’s own books, while loosening them makes it look worse. The Committee catalogues the instances without identifying the incentive that produces them.
That incentive is simple – worsening loan terms is fiscally attractive to the Treasury, because it inflates the value of an asset that sits on the government’s balance sheet, and improving those terms is correspondingly unattractive, because it triggers what the accounts record as a “destruction” of that asset – even where the underlying cash position is unchanged or better.
Successive governments haven’t stumbled into repeatedly tightening the screw on graduates. They’ve been following the only lever the accounting framework rewards them for pulling, and Recommendation 97’s fix for confusion doesn’t touch that lever at all – you can hand every graduate a perfectly clear, annually updated projection of their own repayment path, and it changes nothing about why the threshold keeps getting frozen in the first place.
Burnham after reading
This report appears in the middle of Andy Burnham’s leadership run, which is either extremely bad timing for him or extremely good timing for everyone else, depending on how he answers it. Because on the face of it, its central recommendation – get back to a 50:50 split between graduate and state, rather than the 95:5 the Committee’s own evidence describes – costs money. Not a trivial amount, and not money that’s obviously sitting around waiting to be spent.
That would be an awkward inheritance for any incoming leader picking through what he can and can’t afford in his first weeks. It’s a specifically difficult one for Burnham, because he’s been here before. In 2015, unveiling his own leadership manifesto, one of his five headline pledges was to scrap tuition fees altogether in favour of a graduate tax – lifting what he called the “millstone of debt” weighing on young people starting their careers.
No rate was specified – no threshold, no costing. Alongside the pledge, his campaign proposed a Beveridge-style commission to review “potential” graduate tax design, which is an odd thing to need if you already know what you’re pledging.
Eleven years on, this report’s evidence is essentially confirming the diagnosis behind that 2015 pledge, even if nobody on the Committee would put it in those terms – an income-contingent system that behaves like a tax has been dressed up as a loan, run in a way that lets government quietly worsen the terms while the public blames the interest rate rather than the accounting.
Burnham doesn’t need reminding that the millstone is still there. What he’s never had to answer is whether fixing it is worth the price tag now that he’s the one who might have to sign the cheque, rather than the one making the pledge from opposition.
More on that, and on what it would actually take to break the pattern, over on the Post-18 project.
