Why do we keep treating education like a bond that can only go up? A university degree once looked like a sturdy bridge across the river of working life, but bridges are not immortal. They rust, traffic changes, and sometimes the road on the far bank disappears. The latest numbers suggest the graduate premium is not vanishing everywhere, but it is behaving less like a guarantee and more like a fragile social contract. In Britain, that contract has thinned sharply. In the United States, it is more complicated, which is another way of saying the same comfort story no longer fits the facts.
The useful mistake investors and families make is to confuse a familiar pattern with a permanent one. When a premium persists for decades, people start to assume it is as natural as gravity. Yet economics is full of examples where scarcity, not virtue, creates the reward. Once scarcity fades, so does the rent. A degree can still be valuable, but value is not the same thing as certainty. The bitter lesson is that a credential is only as powerful as the labor market that stands behind it.
The clearest warning comes from the UK. Wonkhe, summarizing the FT’s John Burn-Murdoch, reported that in 1999 the average UK graduate salary was 80 per cent more than that of non-graduates. In the latest data, that gap had narrowed to 45 per cent, and that is before student loans are taken into account. That is not a minor wobble. It is the kind of erosion that changes how households think, how schools sell aspiration, and how policy makers justify the entire financing model.
Other data points point in the same direction. Times Higher Education, citing UK Labour Force Survey data, said the graduate premium fell from 50 per cent to 36 per cent for ages 16-64 between 2007 and 2023, and from 35 per cent to 21 per cent for ages 21-30 over the same period. The difference between older and younger workers matters. It suggests the strain is not only historical memory fading; it is arriving where the promise was supposed to be freshest. That should worry anyone who thinks of education as a reliable hedge against weaker wages.
There is also the uncomfortable fact that nearly 5 million UK graduates work in non-graduate roles, according to The Guardian. That figure does not mean the degree is worthless. It means the labor market is not absorbing the credential into the jobs it was supposed to unlock. In engineering terms, the intake valve is clogging. You can still pour in more graduates, but if the system cannot route them into higher-skill work, the pressure leaks elsewhere. Under those conditions, the premium is squeezed not because education became easier, but because the destination became less scarce.
The obvious explanation is oversupply: too many graduates chasing too few good jobs. But the story is more severe than that neat diagnosis allows. Wonkhe framed the decline as a demand-side problem, not simply an issue of supply, and said the UK sits near the bottom of OECD countries for graduates in graduate-level roles. That matters because it shifts the blame from classrooms to companies and from student enthusiasm to national economic structure. If the issue were just too many people going to university, the fix would be simple. A demand problem means the economy itself is failing to generate enough work that uses graduate skills.
That is why the debate should not be reduced to a moral lecture about “low-value degrees” or an equally lazy defense of higher education as an untouchable good. Markets punish both forms of simplification. A degree is not a sacred relic, and it is not a scam by default. It is an investment whose return depends on the breadth of the economy’s productive base. When that base narrows, even a well-earned qualification can become an underused asset. History has seen similar patterns before: guilds, clerical offices, and state examinations all once conferred outsized status until the social machine around them changed.
Investor psychology offers a useful analogy. People tend to anchor on old averages, even when the underlying regime has changed. They assume the next twenty years will look like the last twenty because that is what their memory supplies. But memory is a terrible forecaster. It sees the premium as a line, when in reality it is the product of shifting bargaining power, technology, and the number of people competing for the same rung on the ladder. Once the ladder gets crowded, the premium can shrink without any scandal, fraud, or headline crash.
This is also where the family story becomes deceptive. A degree is often sold as a defensive move: pay now, protect future income later. That logic is not wrong, but it is incomplete. A hedge only works when the risk it is meant to offset is stable. If the labor market itself is changing the shape of risk, then the hedge may still have value while losing its old symmetry. In plain terms, the degree may still improve odds, but the odds are no longer what parents remember from their own youth.
The United States is less straightforward, and that should make us more suspicious of blanket claims. Wonkhe, again summarizing the FT, reported that the US graduate premium has risen from 80 per cent to 92 per cent even as degree-holding has climbed from 27 per cent to 40 per cent of the population. That looks, at first glance, like proof that more graduates can coexist with a higher premium. But The Economist, citing New York Fed data, reported that the US college wage premium for young graduates fell from 69 per cent in 2015 to 50 per cent in 2024.
These figures are not identical, and they should not be forced into harmony just to preserve a neat thesis. They likely reflect different age groups and time windows. That is precisely the point. The premium is not one fixed thing. It changes with cohort, age, and the stage of the economic cycle. A mid-career degree holder and a recent graduate do not inhabit the same labor market, and the same country can produce opposite-seeming truths depending on who is being measured. The wise response is not to pick the figure that flatters one’s preconception, but to admit the measure itself is conditional.
Still, even the divergent US evidence points to caution. If the premium for young graduates is falling while degree attainment keeps rising, then the market is not delivering a simple story of ever-growing reward. It is sorting winners and losers more aggressively. That is a classic feature of systems under strain. When a credential becomes more common, the premium can survive only if the economy creates enough high-productivity roles to absorb it. If not, the degree becomes a filter, then a sorting device, then a soft credential with hard costs attached.
This is why the policy debate matters even in a story that is really about labor demand. Wonkhe noted two live pressure points: the UK government’s three-year freeze on the student-loan repayment threshold and Conservative proposals to cut about 100,000 university places. Both are responses to the same underlying anxiety: if the payoff is weaker, who should bear the bill? That is the political version of the same market question. When the asset underperforms, does the owner eat the loss, or does the system quietly transfer the cost to the next cohort?
Resolution Foundation research, as reported by The Guardian, found that new UK graduate salaries fell sharply in real terms over two decades while the minimum wage rose. That is an awkward combination for any society that still uses university as the main ladder of upward mobility. If the floor is rising and the graduate starting point is not keeping pace, then the premium compresses from both sides. The result is not collapse but disappointment, and disappointment is dangerous because it accumulates slowly. Families can tolerate a one-off shock more easily than a long erosion of trust.
The deeper fragility is not that degrees have no value. It is that too many people have been taught to think of education as an antifragile asset, one that benefits automatically from time and effort. In reality, the degree premium resembles a moat that must be maintained. If the surrounding economy is not producing enough graduate-level activity, the moat fills with sediment. Then the credential still signals discipline, intelligence, and persistence, but those qualities no longer command the same wage difference.
That should change how we think about human capital. Not every investment in knowledge is meant to pay off in salary, and not every salary premium is a virtue. Some are simply by-products of scarcity. When scarcity eases, the premium should fall. That is not a moral tragedy. It is a re-pricing. The tragedy would be if institutions continued to sell old expectations after the market had already moved on.
The proper lesson is not anti-education. It is anti-complacency. A degree may still be worth having, but not because the world has promised to reward it forever. The real premium belongs to adaptability, and that premium has always been harder to package, harder to market, and harder to borrow money against.
