The Treasury exerts control where it can, while HE gets trashed in the process

On the OBR’s spring forecast figures, the student loan system costs the taxpayer about £7.6 billion a year.

When the government lends money to a student, it knows some of it will never come back. The student might not earn enough to repay, or they might repay for a while and then hit the write-off point with a balance still outstanding.

The government adds up all those expected losses across every loan issued that year and books the total as a single charge.

That is not to say that £7.6 billion is nothing. But notwithstanding that the system involves administering a loan system over what is now 40 years, the idea that student loans are some kind of runaway fiscal disaster is wrong.

£6.3 billion

For the past few weeks, a number has been doing the rounds in Parliament. Whenever an MP raises the idea of unfreezing the student loan repayment threshold and reversing Rachel Reeves’ wheeze at the budget, ministers have been pushing back with “be careful, that’s worth £6.3 billion.”

Sometimes, for reasons that are not clear, the figure is a little different. DfE minister Josh MacAlister said this week:

The threshold freeze raises £5.9 billion next year, and it is incumbent on any party that is serious about fiscal prudence to set out how it would pay for changes.

Whether it’s £5.9bn or £6.3bn, the thing is real – it comes directly from the OBR’s own forecasts. But what is it?

It’s a calculation of the total extra repayments that the threshold freeze will generate over the entire 30-year life of the affected loans. To make that comparison fair, future repayments are discounted – a pound received in 2055 is worth less than a pound today, so it counts for less in the total. Add it all up, apply the discount, and you get £6.3 billion.

What you do not get is £6.3 billion arriving at the Treasury next year, or the year after, or during this Parliament. The extra cash repayments generated by the freeze in any single year are a few hundred million pounds at most, and the bulk of the £6.3 billion sits in the 2030s, 2040s and 2050s – dependent on what graduates happen to earn across four decades that nobody can predict with much confidence.

So why does it matter so much in 2025/26? Well, the change means that the “partition” has to be adjusted. When a loan is made to a student, one part sits on the balance sheet as an asset – the bit the government expects to get back. Another bit sits on the in-year expenditure line. That’s the bit the government doesn’t expect to get back.

So if you pull a stunt like future-freezing the threshold for a few years, suddenly your asset is bigger and your in-year spend is lower. And because you can’t turn back time, the full saving shows up in the year that you push the legislation through.

In other words, the £6.3bn saving on write-offs shows up in the budget this year, which helps explain why ministers are keen to point out that if it wasn’t there, other things the government has spent money on this year would have been unaffordable.

Whether you think asking private loan holders to pay for public goods through worse terms is moral or not is one question. The other one is whether accounting treatment should be driving decision making.

Rear projections

Interest is interesting – as is inflation. For a long time now, both on maintenance and fees, “index linking” has meant using the projected RPI-X inflation rate for Q3 in that academic year. The OBR undershooting that projection has led to compounding problems in the value of the maintenance loan.

This year both max fees and maintenance (and stuff like PG loans) have increased by 3.1 per cent, which is what the OBR thought RPI-X would be in Q1 of this when that decision was made.

In the latest forecast, the OBR expects RPI-X this current quarter to be 4.09 per cent. So even if you thought RPI-X was a good measure of inflation, the increase being felt by students and universities is not matching it.

But outlay isn’t the only thing the Treasury is worrying about. Student loan receipts are £1.4 billion lower in 2025-26 than in the November forecast, “reflecting the impact of lower-than-expected RPI on student loans.”

If the interest income expected over the life of the loans is lower, the loans are worth less as an asset, and the RAB charge – which measures that expected loss – rises accordingly.

Normally, lower inflation is good news for the public finances. But because the government charges students RPI or RPI+3 on student loan balances, lower RPI means lower receipts, which means a higher write-off, which means a higher cost.

So decisions start being made to manage the number rather than to achieve anything useful.

As such, student loan policy in England is largely a history of parameter adjustments made to keep the taxpayer cost at a level that will not cause a scene in the Treasury. The switch from Plan 2 to Plan 5 – lower interest, lower threshold, longer repayment window – was partly about reducing the optics of graduate debt, but the detailed settings were clearly chosen to reduce the RAB rate, not to reflect a settled view about what graduates should contribute, what they should get out of HE or how many should go.

The result is that nobody is making a coherent decision about what universities are for. Is higher education a public good that the state should fund generously, because a more educated population benefits everyone? Is it a private investment, where the person who benefits should pay? Something in between, varying by subject or institution? Those are the real questions. Instead, the question being asked is – what settings keep the RAB charge from causing a problem?

Trading places

I raise this because lots of folk in HE like uncapped recruitment and a complete absence of student number planning and controls. The trouble is that inside the Treasury, if the political position is “no meaningful controls on who goes, where they go or what they study”, all you have left when advising the Chancellor to exert control is the terms of the loans.

Whether you’re a fan of uncapped recruitment or more sceptical, there’s a trade-off. And the downsides are starting to play out now in the press in a month when the value of HE couldn’t look worse.

At the British Academy’s Shape conference the other day, Adam Tickell, vice chancellor of the University of Birmingham, argued that the system is “just not working” for students, universities or taxpayers.

Calling for a proper review, he said that students without a single A level or equivalent are accessing the student loan book for courses where, he argued, the investment is unlikely to pay off:

We are investing so much money in people who we are not really capable of graduating.

Vivienne Stern of Universities UK disagreed with his call for a full review. The post-16 education White Paper published in October already contains, in her words, “about 35,000 policy recommendations” – opening another review risks more chaos on top of existing chaos. If a review does happen, she said, its scope should be tightly controlled.

Stern is wrong because tweaks won’t fix a system which is now clearly being driven by arcane accounting rules and stealth attempts by the Treasury to reduce the costs of loans already made when some of them cost much more than the old projections.

Tickell is wrong because he seeks to control the cost of the system via the “quality” of applicants instead of the “quality” of admissions decisions and programmes – a slippery slope if ever I’ve heard one.

I wouldn‘t start with arts

Shadow ministers repeatedly raise the question of whether students on creative arts courses – drama, music, fine art – are generating taxpayer costs that can’t be justified. The argument goes that graduate earnings from those subjects are lower on average, which means a higher proportion of loans are written off. The proposed solution – cap or restrict loan access for students choosing those subjects.

The problem is that restricting loans based on subject-level RAB rates is not policy analysis. It is reading a spreadsheet and calling it a conclusion. It reproduces the class structure of the existing labour market – the jobs that show up well in PAYE earnings data versus the jobs that don’t – and uses it to decide who deserves an education.

It’s a policy idea that invites the question – which students are worth the money? We’d be better to ask – which provision is worth the money? And when Tickell says “we are investing so much money in people who we are not really capable of graduating”, one of the questions is “who is the we”.

As we’ve noted before, over the past decade, a group of for-profit private colleges in large urban areas has expanded rapidly by doing the following – partnering with registered universities, who lend them their OfS registration, and enrolling large numbers of students onto Business and Management courses. The student loan money flows to the university, which takes a franchising fee and passes the rest to the college. The college provides the teaching.

OfS data shows around 102,000 students enrolled on these subcontracted Business and Management courses in 2023-24 – up from fewer than 6,000 ten years ago. Last year we identified a core group of 16 such providers that enrolled more than 40,000 full-time first-degree students in 2022 alone.

The outcomes are poor. Only 70 per cent of students continue past the first year, against a sector average of 88 per cent and a minimum regulatory standard of 80 per cent. Only 53 per cent progress to a graduate job or further study, against a sector average of 71 per cent and a minimum of 60 per cent.

Of every 40,000 students who start at these providers each year, roughly 16,000 drop out – mostly in year one – with loans they are almost certain never to repay. Of those who complete, only about half get graduate jobs. Many of the rest end up in work that pays below or just above the £25,000 repayment threshold, meaning their debt will sit largely unrepaid for 40 years before being written off.

Taking all of that together, about 70-80p in every pound lent to this sub-sector will never be repaid. The system average much much lower. The government lends these circa 100k students about £2bn billion a year. The annual taxpayer cost is around £1.75 billion a year.

The same group of providers made £504 million in gross profit in the most recent year for which Companies House figures are available, on total income of £815 million – a 62 per cent gross margin.

Exerting control

OfS has regulatory thresholds for continuation, completion and graduate employment. Providers that fall below those thresholds are supposed to face consequences – improvement conditions, sanctions, and ultimately removal from the register of approved providers.

In practice, providers dispute the findings. Course portfolios change between when the data is collected and when the case is heard, making comparisons harder. By the time any enforcement action is resolved, the students whose poor outcomes triggered it have long since left – one way or another. The system is too laggy to exert control.

One alternative approach would be a tax rather than a regulatory process. The government already collects detailed data on what every graduate earns, year by year, through HMRC and the Student Loans Company. It knows, for each provider, how much of the money it lent out is actually being repaid.

That data could be used to calculate a tax on private provision – if your graduates are repaying at a 71 per cent write-off rate against a Business and Management benchmark of 50 per cent, you pay a levy of 21 percentage points on every pound lent to your students.

But the earnings data is five years behind at the point it becomes usable. A provider can register, enrol 40,000 students, generate £500 million in gross profit over three or four years, and close – directors resigned, company dissolved – before the first tax assessment arrives. The possible business model of the worst actors in this group is built around that timeline – grow fast, extract profit, exit before accountability catches up.

You could try to fix this with an upfront bond. When a provider registers, it deposits money with DfE or the Student Loans Company, calculated based on predicted write-off rates for its subject mix and student profile. The deposit sits in escrow and is drawn down against actual outcomes as the data comes in. The Student Loans Company already sits between the loan flow and the provider – holding back a portion of fee payments into a provider-specific reserve is operationally straightforward.

But correctly calibrated, the bond required to cover the expected excess losses from a single year’s cohort at a 75 per cent write-off rate is larger than the entire annual profit. Which means the bond would not regulate the sub-sector. It would close it down. These private providers would not post the bond. They would not operate. At which point we have to ask the obvious question.

Who’s asking?

When Tickell asks whether students without A levels should get loans, he is asking a question about students.

When the Conservatives ask whether creative arts degrees are worth subsidising, they are asking a question about subjects.

When ministers invoke the £6.3 billion figure to justify threshold reform, they are asking a question about spreadsheets.

Nobody is asking why that private franchised business provision exists, who it serves, and whether anyone other than its owners benefits from it continuing.

The students at these London Business and Management colleges are not, for the most part, making well-informed choices in a functioning market. They are responding to marketing – sometimes from agents working on commission, advertising maintenance grants of up to £15,000 on pull-up banners in shopping centres.

They are told about the partnering university’s outcomes, not the college’s. The college’s actual continuation, completion, and employment data is either not published separately or not findable in any useful form. They take on £60,000 of debt without being able to find out that only 53 per cent of students at those providers end up in graduate jobs.

That is not a story about the wrong students making bad choices. It is a story about a system that has allowed private profit to operate on some of the most financially vulnerable students in higher education, with almost no accountability, while the costs pile up in the student loan book only for the Treasury to wreak revenge in a way that a) makes the asset look bigger b) delivers a false in-year budget bonanza and c) wrecks the reputation of HE in the press in the process.

I say again – why does DfE think it’s worth the risk to keep open the student loan book to private providers through franchise agreements for non-specialist subject higher education?

The faster the government changes course, the faster all of us can turn our attention to improving higher education’s contribution to society and economic growth – rather than chasing around owners of colleges who, collectively, are getting rich off outcomes which OfS says are unacceptably poor while the rest of the sector’s reputation and funding gets trashed in the process.