Universities UK’s latest financial pressures survey is dominated by dramatic institutional issues – talk of mergers, shared procurement and digital transformation.
But buried within is a far more telling set of numbers.
The share of universities that have cut student bursaries and scholarships has jumped to 27 per cent, up from 15 per cent a year ago. And the share cutting hardship funding has risen from 9 per cent to 13 per cent.
In a survey ostensibly about how the sector keeps the lights on, support for students is now one of the things being switched off.
It’s notable how easily it can be done. In some cases, cutting a bursary that formed part of a provider’s Access and Participation commitments may amount to a stealth revision of its Access and Participation Plan without ever troubling OfS for approval.
Notwithstanding OfS’ bizarre lack of interest in inflation over time when approving APPs, that this can happen at all tells us most of what we need to know about how this sort of spending is regarded.
It is, in the end, treated as the university’s own money – there to offer when times are good, and to withdraw when they are not.
I’ve only just been paid
To understand why a university can cut financial support to students, we have to go back to the bargain struck when tuition fees were introduced.
Ever since variable “top-up” fees arrived, there has been an assumption built into the system that at least some of that money should be spent on student financial support, over and above the national statutory schemes.
In 2004, partly to get top-up fees through Parliament at all, the then education secretary Charles Clarke created a new Office for Fair Access, and made charging the higher fee conditional on an access agreement that committed a portion of the extra income to bursaries and outreach. A proportion of fees, in other words, has been earmarked for disadvantage from the very beginning.
But the design carried a tension that has never gone away. Some argue that local decision-making – institutional autonomy – is better at getting the money to where it is really needed. Others point out that redistributing fee income within a provider, rather than across the country, means support ends up being based not on need, but on how many other students at your own university happen to need it too.
The old Office for Fair Access could at least show us the shape of the problem – it tracked how many “OFFA countable” students each provider had and what it was spending, and it tended to confirm the uncomfortable truth that the institutions doing the most on access were the ones with the least to spend per head.
Over time, though, direct financial support fell out of fashion – nudged along by research that struggled to find much effect on applications or retention, and by a regulatory turn that prized continuation and completion metrics over anything as old-fashioned as what students actually needed to do anything other than survive.
When OFFA was folded into the Office for Students, provider-level reporting on financial support pretty much disappeared with it. The expectation that a slice of fees goes to disadvantage survived. The means to see whether it actually does did not.
Which is why, for the fifth year running, I’ve prised the underlying numbers out of the Office for Students via a Freedom of Information request. The result is not pretty – and what’s most alarming is where the falls are sharpest.
You’re broke
If we just look at cash help, in 2024/25 just under £476m went to just under 295,000 students across England – a spend per head of around £1,614 in cash terms.
On the face of it that looks stable. The year before it was a touch over £496m to around 311,000 students, or £1,598 a head, so the nominal per-head figure has even nudged up a little.
But two things moved the wrong way at once. Strip out Sheffield – a large Russell Group university that appears not to have returned cash data this year, and so drops out of the figures – and the pot still shrank by around £12m, with roughly 8,000 fewer students getting any cash help. (Leave Sheffield in and the headline falls look bigger, about £20m and 16,000, but around half of that is the missing return rather than a real cut.) For several years now, rising student numbers had quietly disguised the squeeze – more and more people sharing a pot that was flat in cash and shrinking in real terms. This is the year the headcount fell too.
Does any of that make sense given what we actually know about how students are doing? Nope. Maintenance support has spent years drifting below inflation, the household income thresholds that govern who gets what have been frozen since 2008, rents have run away from everything else, and campus food banks have gone from novelty to fixture.
If institutional help were responding to need, it would be going up. It is going down.
I’m working for the cash machine
In previous years I’ve compared the cash figures year to year without adjusting for inflation. This time I’ve deflated all ten years to roughly today’s prices using ONS CPI, so the bars in the chart are all in 2026 money.
On that basis, cash help per student helped has fallen from about £2,273 in 2015/16 to £1,672 in 2024/25 – a drop of around a quarter. The real-terms pot is down about 17 per cent across the decade, even as the number of students helped has risen 13 per cent.
So this isn’t just a story about money being spread a little more thinly. It’s a story about there being meaningfully less of it.
One health warning that cuts in only one direction. CPI almost certainly flatters the numbers. The things that dominate a student’s outgoings – rent, energy and food – have risen a good deal faster than the headline index. A student-specific deflator would make these falls look worse, not better. So read the real-terms numbers here as the optimistic version.
I scratch a living, it ain’t easy
Ever since the days of the Office for Fair Access, HESA has collected, for each provider, the amount spent on student financial support and the number of students it helped. It splits into four types of spend.
- Cash is any bursary, scholarship or award paid to students with no restriction on how it’s used.
- Near cash is voucher schemes or prepaid cards that can only be spent in defined places.
- Accommodation discounts are reductions on university halls and residences.
- Other is the catch-all – travel, equipment, printer credits, subsidised field trips and meals, and so on.
The collection covers undergraduates and postgraduates, and UK-domiciled and international students alike. Where a student receives help from more than one provider they appear more than once, awards to fewer than ten students are suppressed, and Disabled Students’ Allowance has been stripped out. It’s England only.
That postgraduate and international presence matters for how you read the per-head numbers, because the denominator includes heads who sit on very different support footings. “Per student helped” is a rough divisor, then, not a clean measure of what a needy home undergraduate receives. Bear that in mind throughout – it’s one of several things this data can’t tell us, of which more below.
The interactive lets you pick the national picture, a provider group or any individual institution, and toggle between the four types of support. The bars show real-terms spend per student helped – the line shows how many students received it.
I’m always paying, never make it
Spend per head is really just a pie – the total amount spent – cut into slices, one per student helped. A falling per-head figure can mean a smaller pie, more slices, or both. And it’s the difference between groups of universities that tells the story.
Nationally, the pie shrank about 17 per cent in real terms while the number of slices grew 13 per cent, which is how you get to that 26 per cent fall per head.
But split it by mission group and it all gets uncomfortable. In the Russell Group the real pie barely moved – down only about 4 per cent – even as it took on around 8 per cent more students, so per-head help slipped about 11 per cent, to roughly £2,675. The richest institutions, in other words, protected the pot. (Sheffield, which appears not to have returned cash data this year, is left out of the Russell Group figures so the comparison is like-for-like.)
In the post-92 universities, the opposite. There the pie was cut by about a quarter and they helped about a quarter more students at the same time – the double squeeze – so per-head cash help fell around 40 per cent, to roughly £1,122.
The pre-92s saw the pie slashed by around 41 per cent and per head down 38 per cent, to about £1,315 – the newer universities were down about 39 per cent, to around £1,186. The only group moving the other way is small and specialist institutions, where per head is actually up about 12 per cent – but off a low base and a small population.
The net effect looks like this – the gap between the Russell Group and the post-92s, measured per student helped, has widened from about 1.6 times to about 2.4 times in a decade, and the squeeze is broad – of the roughly 110 providers we can track all the way back to 2015/16, real cash help per head has fallen at around seven in ten.
None of which means the sector divides tidily into rich and poor, or that any one university is being careless. Institutions run very different blends of cash and non-cash support, serve very different students and face their own financial constraints, and the group averages iron all of that out. But the direction is hard to miss.
It pays to handle the individual numbers with care because per head can mislead in both directions. Last year I pointed to students at Salford getting £358 each in cash while their neighbours at the University of Manchester got far more. This year Salford’s figure has “risen” to around £551 – but only because it helped a third fewer students than the year before, even as its own pot shrank. The per-head number went up precisely as the support came down. Handled carefully, though, the contrast still hurts – a supported student at Manchester now gets around £2,368, and one at Birmingham close to £4,891, while a couple of miles from Manchester their counterpart at Salford gets a small fraction of that.
It pretty much tracks not need, but how many other students at your university also need help – which was always the objection to moving fee income around a provider rather than around a country.
My credit’s in the red
Cash is the bulk of it, but the other categories are revealing in their own right.
Near cash has all but disappeared. Real per-head spend is down about 54 per cent, and the national pot has fallen from around £31m to under £9m. It was overwhelmingly a post-92 instrument – Manchester Met, Bournemouth, Anglia Ruskin and the like – and that has been dismantled almost entirely, with post-92 near cash collapsing from roughly £25m to barely £2m.
Accommodation discounts are shrinking overall – real per head down about a third – and it’s becoming an almost exclusively Russell Group habit. The Russell Group now accounts for roughly two-thirds of the category, with York, King’s, Oxford and Liverpool to the fore, and is about the only group where it has grown. Ignore the 2020/21 bulge in the chart – that’s pandemic rent rebates, not a trend.
Other is lumpy and heavily concentrated, driven by one or two big institutions in any given year – Durham in 2022/23, the Open University in 2023/24 – so read its line with care rather than as a clean trend. The one durable signal underneath the noise is post-92 “other” support, down around 40 per cent.
You can’t look back
There’s a lot this doesn’t tell us. We don’t know whether falling per-head cash is retrenchment or better targeting – fewer, larger, sharper awards aimed at the students who most need them would look identical in this data to a straightforward cut.
It can’t tell us whether any of it is working, and we’d first have to agree what “working” means – continuation and completion rates, or whether students are eating properly and learning anything (or, God forbid, making the most of the wider experience.)
Most important of all, it can’t always tell us who the “students helped” really are. The figures count undergraduates and postgraduates, and home and international students, all together – and the number being helped at the post-92 and newer universities has jumped by a quarter or more, almost all of it since 2022/23, exactly when taught-postgraduate and international recruitment surged at those same institutions.
Some of the per-head fall, then, may be more heads of a different kind – overseas or postgraduate students on modest or fee-discount awards – diluting the average, rather than home students in hardship visibly getting less. A change to how the data was collected from 2022/23 nudges the recorded numbers around too. We can’t cleanly separate any of that here.
Nor does it tell us the home/international split or whose need is actually being met – or how much of any provider’s spend is need-based in the first place. A large cash-per-head figure, of the kind the most selective universities post, may reflect generous merit and international scholarships every bit as much as hardship support for the poorest.
It can’t show us these sums as a share of each provider’s total or fee income, the only way to tell a deliberate choice from a casualty of wider pressure. And it can’t be squared against Access and Participation Plan commitments or student premium funding, because the published finance data records expenditure without the student numbers, the pre-pandemic baselines or the premium allocations that would let us compare like with like.
Where I want to be
I regard the previous section as a set of fair questions, and all represent reasons for OfS to publish this material properly rather than leave us to ask for it every spring.
But strip the caveats away and one thing is not in doubt, because it doesn’t depend on the denominator at all – the real-terms pot is shrinking. Nationally, cash support is down about 17 per cent in real terms over the decade. It is broadly flat in the Russell Group, down a fifth to a quarter across the post-92 and newer universities, and down more than 40 per cent among the pre-92s. The money is being withdrawn fastest from the kinds of universities that enrol the most students who need it – and however you read the per-head figures, that much holds.
It was never easy to justify asking a provider to redistribute its fee income among its own students. On these numbers it is harder than ever. Redistribution doesn’t achieve much when the richest institutions have the fewest poor students to spend it on, and the ones doing the most for access have least left to give.
Don’t make plans
None of the pressures pushing the other way are letting up. Maintenance remains adrift of inflation, the household thresholds remain frozen, and rents, energy and food costs keep climbing – with headline food inflation alone set to hit 10 per cent by December thanks to Trump’s Iran folly.
And the UUK survey we began with shows the cutting has not only started but is accelerating – the share of universities cutting bursaries has nearly doubled in a year, the share cutting hardship funding is up by almost half, and more than nine in ten institutions say the recent fee uplift will not fully offset the policy changes bearing down on them.
Set against rising employer national insurance, a weaker international recruitment market and stubborn cost inflation, student support is an easy line to trim. So the pot is far more likely to shrink than to grow over the next few years – precisely as student need climbs. Whatever this data says about the last decade, the next one starts from a worse place.
Ticket inspector’s there
If we genuinely wanted the most money to reach the students who need it most, we wouldn’t start here. The way to do it would be to act at least regionally – pooling bursary and hardship spend across the providers in an area, rather than leaving each to recycle its own fee income among its own students.
It remains vanishingly unlikely that OfS’ regional partnership structures will go anywhere near that far, but the logic is sound – it is the only way to stop support being rationed by the accident of which university a student happens to attend.
Even then, beware the agglomeration trap. The English region with the lowest higher education participation rate is the North East, at around a third – London’s is closer to two-thirds. Pool only within regions and you simply move the unfairness up a level – the areas with the most students who need help would still have the least to share around. The money ought to flow from the well-resourced, high-participation corners of the country towards the ones doing the hardest work, which means acting nationally, not locally.
And there is a fairness problem the within-provider model never escapes – one owed to fee-paying students themselves.
When fees were a modest part of the funding mix, it was reasonable to tell students that the state would concentrate its help on those in need. But as the public subsidy has thinned, and graduates are now expected to repay almost everything they borrow, we are increasingly asking students’ own debt to do the job the state used to do.
It is one thing for your fees to help another student at your own university. It is quite another for them to prop up a national safety net that general taxation has opted out of funding.
You don’t get by
Strip the politics away and there are really only two designs on offer. Top-slice fee income and let each provider spend it on its own students – the “we know best” model, defended on grounds of institutional autonomy and local knowledge. Or top-slice it and spread it across England on the basis of need – blunt, but fairer.
Reasonable people can make the case for the first. But it becomes indefensible the moment you get the differentials seen here, and when it turns into cover for everyone quietly cutting both the pot and the per-head amount while the regulator notices nothing at all. That, on five years of this data, is roughly where we have ended up.
The least OfS could do is make providers show their working. The most it could do is accept what plenty of people warned about when the original settlement was struck – that progressive funding of student support, much like progressive taxation, works better done across a whole country than on one campus at a time.
