England’s student loan system died gradually, then suddenly

Two ways. Gradually, then suddenly.

Of course, Hemingway wasn’t really writing about bankruptcy. Campbell’s financial ruin was a stand-in for the broader post-war disintegration of his class, his relationships, and his sense of self.

It captures something about how we experience decline – we normalise each small deterioration as it happens, adjusting our baseline incrementally, until the cumulative weight of all those small shifts produces a catastrophic and apparently sudden collapse.

Was the moment that, at PMQs, Keir Starmer said that the vetting of Peter Mandelson had included an “ongoing relationship” with Jeffrey Epstein the fatal one?

People knew that beforehand, and the jettisoning of close aide Morgan McSweeney may delay the inevitable. But most now think that the Starmer premiership is over.

By common consent, Keir Starmer is a decent man but one who arrived in office with no real politics of his own, and an assumption that competent management would be enough. He outsourced strategy, values and judgment to others – and when those borrowed convictions collapsed, he had nothing to fall back on.

But he didn’t just inherit other people’s baggage. He actively chose to embrace it – appointing Mandelson to Washington when he could have kept his distance. Every decision point was a chance to build something of his own – but he never took it.

Now everyone’s unhappy and nobody’s loyal. The Blairites think he’s a cold fish who doesn’t understand politics. Labour’s left never trusted him. The media have locked onto the contradictions and each week brings a new story. His achievements, such as they were, are swallowed whole by the Epstein taint.

Starmer is neither one thing nor the other, and when you occupy the centre on credibility alone, the moment the credibility goes, there is nothing left.

Keep on tweaking

I think similar things about England’s (and, by Treasury proxy, devolved nations’) student loan system.

Starmer’s government arrived with no real conviction behind the existing system, and an arrogant assumption that tweaking the inherited model would be enough.

Successive governments outsourced responsibility for making its madness make sense – to universities, to regulators, and to students themselves – and when the system’s contradictions became impossible to ignore, there was no underlying vision to fall back on.

Then Labour didn’t just passively inherit a broken model, it actively made it worse – squeezing maintenance in real terms, ignoring the concentration of numbers into high tariff institutions, and now freezing thresholds to make a spreadsheet add up. Every decision point was a chance to reform. They never took it.

Now everyone’s unhappy and nobody’s defending it. The right hates it because it’s not a real market. The left hates it because it loads debt onto graduates for a public good. Those who like loans complain that interest “punishes success”. Those who would prefer a progressive tax bemoan that the principal never seems to reduce.

Meanwhile, universities can’t deliver what they promised on a frozen fee. Graduates aren’t getting what they were promised either – the premium is contested, the debt feels real, and the maintenance never covered the rent.

HE expansion was an achievement – but the funding model has tainted everything it enabled. The loan system is neither loan nor tax, neither market nor public service, and when you’re in the middle with nothing but credibility and delivery, the moment both fail, there is nothing left.

It’s over

For me, the moment it was all over was IFS’ publication of this graph. I’d had a run at modelling something similar a while ago, but seeing it interactively laid out like this suddenly made all of the tweets, the LBC callers and the newspaper profiles seem much more real:

Some are suggesting, as some Labour backbechers are reported to be pushing for, that reversing the buried “Plan 2” freeze will be enough to stave off crisis. Others, like the campaign in the Sunday Times, posit that cutting interest rates or introducing a lifetime cap on total interest paid will arrest the system’s decline.

But as John Blake, the new Director of the Post-18 project pointed out this week, it matters not that a spreadsheet will show that reducing interest only benefits the rich, and only then in the long-term. It feels, to almost six million people, unfair.

Some suggest that time will heal, that Plan 2 grads will get better jobs, and that the Rishi Sunak/Michelle Donelan changes in Plan 5, with lower interest traded off for a 40 year term, will feel much better.

But even then, the contradictions will continue. The graduate premium, buffeted by a fast-rising minimum wage and turbulence in white collar recruitment, will look increasingly shaky.

And the numbers are still a problem. On a debt of £60k with a threshold of £25k, even RPI only interest means I’d need to be earning £52,333 to make a dent. And then there’s the postgrads.

Plan 3 graduates that take up a PGT will also need to earn north of £50k to make a dent – but their repayments pile up. A graduate on £40k a year will pay £2,490 in today’s money – more than their total NI contributions this year. And in that scenario, their statement will show their combined principal growing by about £900 a year. It’s over.

The wider problems associated with a voucher-based higher education funding system – the “market”, the lack of planning, the geographical impacts, the psychology of all it – they’ll all be debated and theorised over in any ensuing review.

It’s certainly the case that other countries seem to have managed to run an income-contingent student loan scheme without the worst of the press coverage this week – perhaps because few ever went as far as putting so much of the “public goods” funding into the hand of the “private” voucher holder.

It’s also the case that to the extent to which money can be found down the back of a sofa to apply fixes to Plan, that won’t be money flowing in the coffers of embattled and squeezed students or their universities.

But understanding how we got here will matter if we’re about to embark on the slow process of identifying an alternative.

Fag packets out

Imagine that you, tomorrow, were asked to design an income contingent student loan scheme.

First, you’d decide how much to lend to students to pay for their education. You have calls to make about the extent to which that amount is universal, or differentiated on the basis of how long it takes a student to complete, or which subject they study and where. They all have implications.

Then, you decide how much to lend to students for their living costs. Again, you have calls to make – should some of it be a grant, how much should the maximum be, whether and how to means test, and so on. They also all have implications too.

Next, you decide how much of that borrowing you want back – the rest is your subsidy. All governments subsidise HE – because there are public goods and benefits. But the cost-share has multiple and compounding implications.

Then, you decide when you want the repayments in by – 10 years, 20 years, 30 years? Lifelong? Debts chaseable after death? And you’ll need to decide whether you want more back when graduates are older, or younger. All of those calls have implications as well.

You’ll need to decide whether you want the same percentage back from everyone, or whether you want more from rich graduates to help pay for less from poorer graduates. Implications abound – progressive on paper might feel like “punishment for success” in the Times, for example.

You may not, perhaps, issue an unlimited number of these loans – doing so has implications – but if not, whether and how you control that supply has all sorts of implications to consider.

And then once you’ve done that, you might make a series of promises about index linking the loans you issue out, and doing something similar for wage growth when indexing the repayment thresholds. Each of those promises, whether in law or in speeches, have implications and no end of potential payloads.

Then, when your demand modelling goes wrong, you might be tempted to fiddle about with any or all of the above to keep the system “sustainable”. And when the wider economy goes south, you might fiddle about again with promises, again to keep the system “sustainable”. Doing so, even if done stealthily, has implications.

You may then need to act surprised and/or baffled when a student nurse who has four years of debt because they helped out during Covid is upset that they have to be earning about 80k to stop their loan balance getting bigger. But it won’t solve your problem.

You only get what you give

As such, as the debate kicks off again about the size and shape of England’s (and by HMT proxy, the rest of the UK’s) graduate contribution scheme, you would be wise to think through far more of the implications than appears to have ever been attempted previously.

But you might also remember something important – stage 3 of that multi-step process I laid out. The truth is that each time the sector and its students have found themselves to be in a crisis, the solution has been to transfer some of the costs away from the state and onto graduates.

That is partly why reform of England’s iteration of an income-contingent loan scheme seems so intractable. The state may be running a better loan operation than you’d find on the high street – but once graduates have paid back what they will pay back, the government’s getting a bargain.

The cost-share matters whether you design a graduate tax, a loan scheme or any other wheeze. Ideologically, politically, behaviourally and morally, the less you put in, the more chaos you generate.

And while that split is so opaque and never quite at the surface of the debate, remember this.

If we use RPI-X to calculate in today’s money, in 1998 the HEFCE teaching recurrent grant was around £6.6bn, teaching capital and special funding was around £370m, publicly-paid (ex-LEA) tuition fees were around £2.1bn, legacy maintenance grants were around £1.5bn and intended maintenance loan subsidy was around £370m – totalling approximately £10.8bn.

Today, the OfS recurrent grant is around £1.3bn, and the estimated transfer element of student loan outlay (ie the bit that won’t be paid back) is around £7.7bn – a total of approx £9bn. Reducing your public expenditure in real terms while almost doubling student numbers has been quite the achievement – but it can’t go on.

Cost-sharing is always the buried political choice, and systems designed to obscure it eventually collapse when people do the maths themselves. Labour had 14 years in opposition to develop a higher education funding policy, and chose not to. Whatever else is coming next, it now pretty much has to.