It’s not hard to find arguments that suggest that the UK has too many graduates.
And underneath articles of that ilk, we’re usually treated to laments about the decline of polytechnics, the idea that students should learn a trade, and (at best) the lack of investment in skills.
The Sunday Times spent the weekend declaring the graduate salary premium “a myth.” The Conservatives responded with a plan to cut 100,000 university places.
Conservative leader Kemi Badenoch says that New Labour’s 50 per cent target for school leavers was achieved by “creating loads of useless courses”, and shadow education secretary Laura Trott refuses “to keep funding degree courses which are not delivering for young people.”
Badenoch blames Britain’s “low-growth” economy on “the Blairite settlement” while proposing to fix it entirely by tinkering with the supply of graduates.
She treats the student loans crisis as a supply-side problem. Too many graduates, too many courses, not enough quality control. The prescription is familiar – restrict entry, close programmes, steer people elsewhere.
But cross-national evidence tells a very different story. The decline of the UK graduate premium isn’t an inevitable consequence of mass higher education. It’s a peculiarly British problem – and it’s rooted in what happens after graduation, not before it.
Occasionally, pieces like this from the FT’s John Burn-Murdoch offer up a counterpoint.
Just us, then
In 1999, the average UK graduate salary was 80 per cent more than their non-graduate counterparts. In the latest data that’s down to 45 per cent – and that’s before factoring in student loans. The obvious response is to blame oversupply – expand participation, dilute the pool, and watch the premium fall.
Even the Sunday Times’ more conservative measure – a premium of 30 per cent in 2007, down to 22 per cent by 2024, shrinking to 16 per cent after tax and loan repayments – points in the same direction.
But the problem with that argument is it hasn’t happened elsewhere. In the US, the graduate premium has risen from 80 per cent to 92 per cent even as degree-holding has climbed from 27 to 40 per cent. Premiums have held up or increased in Canada, France, Spain and the Netherlands too.
And in the US, a performing arts graduate earns around a third more than their British counterpart after adjusting for purchasing power – rising to 60 per cent more across all subjects.
The travails of Italy’s graduates in the 1990s and 2000s were more a comment on the country’s stagnant economy than its higher education system – the same is true of the UK today.
So the decline in the UK graduate premium isn’t an inevitable consequence of mass higher education – it’s a peculiarly British problem. The question is whether the answer lies in a weak labour market, weak graduates and courses (and their location), or something else entirely.
Use it or lose it
Much of the literature frames the UK issue not as “too many graduates” but as weak skills utilisation – the economy doesn’t generate enough roles, workplaces and progression structures that actually deploy higher-level skills.
A decent chunk of cross-national comparative work comes from Francis Green and Golo Henseke at UCL’s Institute of Education. Their 2016 study in the Oxford Review of Economic Policy used a validated index of graduate jobs across 21 OECD countries and found that graduate underemployment correlates with the aggregate supply-demand imbalance – but not with indicators of labour market flexibility.
You can’t deregulate your way out of this. You need to actually create the jobs.
Their cross-national OECD working paper found that in most countries between 70 and 80 per cent of graduates worked in graduate jobs – but the UK sat at the bottom of the table alongside Japan and Spain, with among the highest proportion of graduates outside graduate-level roles.
Their most recent task-based analysis in Oxford Economic Papers (2024/25) found that the growth of graduate skills requirements in the British workforce slowed markedly after 2006 – the demand side stalling while supply kept expanding. And they make the crucial point that for higher education to deliver “value for money”, employers need to demand distinctly graduate skills. For graduate skills to be rewarded, jobs need to make productive use of them.
There’s the “low-skills equilibrium” tradition, developed through Ewart Keep’s work at SKOPE in Oxford and going back to the Snower/Wilson analysis. That treats weak skills demand as self-reinforcing – large segments of the UK economy compete on cost and standardisation, leading to limited training, limited job redesign, limited internal progression, and a tendency to use credentials for screening rather than redesigning roles to exploit skills.
It produces a stable but low-productivity equilibrium that sticks around even when the education system expands.
The OECD’s own recent work on adult skills reinforces that. Its 2026 report on how workers use (or don’t use) their skills in the workplace separates skills proficiency from skills use at work – and explicitly highlights that high proficiency doesn’t automatically translate into high workplace use. The UK pattern is one where people have the skills but jobs don’t let them use them.
Geography doesn’t help. IFS research found that employment in high-paying occupations grew by 240 per cent in Inner London since 1993 but just 41 per cent in Cheshire. The new high-skilled jobs in IT, business and finance clustered overwhelmingly in London while low-paid service work grew everywhere else.
And the CIPD’s comparative work found that the UK has 58.8 per cent of graduates in non-graduate jobs – exceeded only by Greece and Estonia – while countries like Germany, the Netherlands and Slovenia with strong vocational training traditions have 10 per cent or less.
The usual suspects
None of this means we should ignore what universities do. But the evidence suggests the supply-side story is weaker than the policy debate assumes.
This “wage premium puzzle” paper for the IFS asked how the UK maintained a large education wage premium despite massive graduate expansion – and found that for a long time, endogenous organisational change (firms restructuring because graduates were available) kept pace with supply. The premium held up.
It’s only more recently, as the post-crisis productivity stagnation has bitten and the minimum wage has risen, that the squeeze has become visible.
The Productivity Institute’s work on graduate mismatch adds another idea – it’s not only the number of skilled jobs that matters but the sectoral composition of employment and the pathways into better matches. Graduates can be in the “wrong job” or the “wrong industry”, and the wage penalties for mismatch point to structural issues in how the UK allocates graduate labour, particularly after shocks.
This matters because much of the coverage frames the problem as one of institutional quality – Russell Group graduates earn well, post-92 graduates don’t. But the IFS data on geography tells a different story. If high-paying graduate jobs grew by 240 per cent in Inner London but just 41 per cent in Cheshire, the earnings gap between a London-based Russell Group graduate and a post-92 graduate in the Midlands is substantially a story about where the jobs are, not where the degree came from.
And management matters as well. Research on management practices shows that management capability varies hugely across UK firms and sectors, and that it directly affects task allocation, training, autonomy and whether skilled workers are set up to use their skills. The UK has a particularly long tail of poorly managed firms compared with the US and Germany.
Free riders
Hat tip to the drum that Johnny Rich regularly beats here – employers are probably underinvesting. UK employer investment in training per employee has fallen by 36 per cent in real terms since 2005, according to the Learning and Work Institute – a decline that has not been observed to the same extent in most other European countries.
The IFS confirmed in its 2023 study that participation in adult education and training has declined across every measure – the fraction of workers participating, the average number of hours, and the intensity of training.
If we compare that with Germany, companies invest around €18,000 per apprentice per year and the system generates a positive return on investment when longer-term benefits like lower recruitment costs and reduced turnover are factored in. The CIPD’s evidence to the Industrial Strategy Committee made the connection directly – a “significant part” of the UK’s low productivity compared with Germany can be attributed to poor workforce skills development, linked to business strategies that rely on cheap labour rather than investment.
So when the CIPD says that “the assumption that we will transition to a more productive, higher value, higher skilled economy just by increasing the conveyor belt of graduates is proven to be flawed”, it’s not making a case against higher education.
It’s making a case against the idea that supply alone creates demand. Simply increasing the qualification level of individuals going into a job does not typically result in the skill required to do the job being enhanced. That requires job redesign, management capability and employer investment. And the countries that maintain strong graduate premia during expansion are the ones that pair education growth with skilled job creation and occupational upgrading.
Johnny Rich’s long-held proposal for a graduate employer contribution gets the diagnosis right – the current system doesn’t incentivise the outcomes it needs, and employers are free-riding on public investment in human capital.
But his proposed solution – a 3 per cent levy on employers hiring graduates, paid back to universities – never seems right for the political moment, especially now.
After the employer NICs increase, the Employment Rights Bill costs and rising minimum wage obligations, another percentage-based charge on graduate employment would almost certainly reduce graduate hiring at the margin.
Worse, it gives employers a direct financial incentive to recruit non-graduates or international graduates to avoid the charge. The stronger version of the employer contribution argument would look more like the French model of a legal obligation to invest in workforce training – something that changes what employers do with their workforce rather than what universities charge them for producing it.
Not off the hook
So if the core problem is demand-side, there are limits to what universities can fix. If nothing else, tinkering with course closures based on earnings data that is really measuring the weakness of the economy rather than the quality of the course is addressing the wrong end of the problem entirely.
The Conservative proposal to cut 100,000 places and redirect savings into apprenticeships is exactly this error with a price tag attached – and as I’ve explored elsewhere, the maths doesn’t survive contact with reality either.
But that doesn’t mean universities have nothing to do. And it’s not OK to simply rely on “graduateness” or a resistance to instrumentalism in a mass system where the economy is on its knees. An accommodation is required – and it isn’t always “skills in the module.”
The comparative evidence on work experience is remarkable. Cross-national data from the REFLEX survey of European graduates found that British graduates gathered much less work experience during higher education than their counterparts in the Netherlands or Germany. Credit-bearing, structured work placements and co-op models – not just “employability modules” bolted onto existing courses – seem to matter for how quickly graduates transition into skilled work.
There’s a structural problem with how we do this in England. A sandwich year typically means an extra year of debt. At Manchester Met, the RISE programme enables students to “over-credit” their degrees with additional units based on engagement with employment-enhancing extra-curricular experience, building skills recognition on top of rather than instead of academic credit. It won a Guardian Award for course design and now involves over 3,000 students a year.
But even RISE sits on top of the degree rather than within it – which creates a question about whether “skills” in our system will always mean either extra time (and debt) or extra workload on top of what’s already a compressed experience.
It all connects directly to the evidence on student regret. The Bristol/HEPI/Advance HE “Benefits of Hindsight” research found that while only 2-3 per cent of respondents thought higher education was the wrong path altogether, 52 per cent of graduates said with hindsight they would have done things differently.
Among graduates who wanted to change, the most common reason was that different choices would have led to more career options. Four in ten expressed the need for different or better careers advice. For those employed in highly skilled occupations, satisfaction with their choices was markedly higher than for those in less skilled positions.
The regret isn’t about having gone to university. It’s about what happened – or didn’t happen – within the experience and afterwards. And that’s a gap universities can address, not by turning every course into vocational training, but by ensuring that structured engagement with skills and work is embedded in the credit system (not necessarily the curriculum) rather than offered as an expensive optional extra or a bolt-on employability week.
Credit where it’s due
One of the UK’s open secrets is that across our universities, there is an enormous volume of higher technical provision that the public and policymakers struggle to see. It’s there, but it’s invisible behind university branding and a policy debate that treats all degrees as if they’re the same thing.
On Kuenssberg, even Laura Trott had to concede that:
…universities are often involved in apprenticeships, which is something that we are going to expand quite dramatically for 18 to 21.
When we visited Técnico in Lisbon, taxi drivers knew exactly what the university’s faculty of engineering and technology does. In England, the equivalent provision is buried inside institutions whose names no longer signal what they offer.
But Europe has a complementary secret that is arguably more important. In systems across the continent, enthusiastic Bologna compliance and the credit structure itself does the work that UK degrees try to bolt on.
Technical students take humanities and social science credits. Purely academic students pick up formal credit for internships, work experience, volunteering and skills development. Students gain recognition for learning acquired through employment, service or positions of trust. And all of it sits within the standard 180 ECTS rather than on top of them – no extra year of fees, no optional bolt-on that time-poor students skip.
In other words, countries can maintain stronger graduate premia via systems where technical students do humanities, purely academic students pick up credit for skills and work experience, both have graduateness experiences we call “extra” curricular rewarded, and none of it requires an extra year of fees. A Chemistry student does less Chemistry – but they’re a more rounded graduate, and in a mass system, that trade-off looks increasingly sensible.
Some tell me that the UK’s compressed three-year, 360-credit model makes this harder but not impossible. But the real obstacle isn’t credit volume – it’s programme design philosophy. Our system tends to treat the subject as sacrosanct and everything else as an add-on, whereas the systems sustaining better graduate outcomes build breadth, skills and work into the core credit structure from the start.
Green and Henseke’s cross-national work suggests the mechanism matters. It’s not about ticking an “employability” box – it’s about work-based learning with clear outcomes and assessment that captures problem-solving and responsibility, making skills legible to employers and reducing the risk of graduates starting in low-skill trajectories from which they may never recover. Early mismatch can persist – which means what happens in the transition matters as much as what happens in the classroom.
And even when skills are developed – whether through the curriculum, work experience, or the kinds of extracurricular activity that European systems formally record on the Diploma Supplement – there remains a surfacing problem. As I’ve argued on the site elsewhere, students develop skills but struggle to name and evidence them, employers report attributes as missing that may actually be present but invisible, and the UK’s patchwork of documentation infrastructure fails to record what students have gained.
See also the desperate need for PSRBs to engage with all of this, and the policy disaster that is the Lifelong Learning Entitlement, which ought to make earning while learning a doddle, but in fact is almost as inflexible for full-time students on mixing and matching as the system it’s designed to replace.
Wrong end of the telescope
As well as all of that, across employers, government and universities, there is clearly a coordination failure. If employers don’t redesign work and invest in training, universities can’t “teach” their way out of underutilisation. If government focuses mainly on supply-side reforms, it can expand qualification rates while leaving the job structure unchanged.
And if universities treat skills as someone else’s problem – or as something that can only be addressed by bolting optional extras onto already overloaded students – then the accommodation the economy needs won’t happen either.
Right now, almost everything government is causing in HE policy and regulation – closing courses based on earnings data, restricting international recruitment, tightening loan terms – addresses the supply side of a demand-side problem. And now opposition parties are competing to do the same thing faster.
The countries that don’t have this problem are the ones that paired graduate expansion with skilled job creation, employer investment, and regional economic development.
The government’s own Skills and Productivity Board material is explicit that skills and productivity are connected but mediated by employer strategy, investment, management and local economic structure.
So place-based policy can’t just fund provision – it has to target adoption and diffusion of technology and management practices in lagging sectors and regions, use procurement and regulation to raise job quality in the large low-wage service markets where the state is a major purchaser, and align industrial strategy with occupational upgrading so that growth sectors translate into more skilled roles rather than just more headcount.
If Britain is “short of graduate jobs”, the policy lever isn’t restricting graduate supply. It’s increasing the number and quality of roles that genuinely require and reward higher skills, and accepting that skills do have a place in undergraduate credit.
Universities have a part to play – particularly in how they credit experiential learning, surface what students have gained, and in how they design degrees. Letting students design their own would be a great start.
But they also have a distinctive contribution to make on the demand side itself – surfacing and generating the policy research that government needs to understand why the UK is an outlier, what levers would shift employer behaviour, and what regional economic conditions would turn graduate capability into productive work.
Until that lands with ministers (or maybe, in the future, Mayors), the debate will remain stuck on supply-side tinkering while the real problem goes unaddressed – a surefire way to become the next Italy.











