The Surplus Elite: Why 2026's Record Graduate Glut Is a Warning the Ancients Already Wrote Down

There is a particular kind of disappointment that never shows up in a headline unemployment rate: the disappointment of the over-qualified. In the summer of 2026 it is arriving at scale. In China, a record 12.7 million new university graduates are walking into a labour market where the urban youth jobless rate — for 16-to-24-year-olds, excluding students — climbed to 16.9 per cent in March, up from 16.1 per cent the month before. In India, the State of Working India 2026 report finds that close to 40 per cent of graduates under 25 are unemployed, and that fewer than 7 per cent of young male graduates land a permanent salaried job within a year of finishing their degree. A diploma was meant to be a ticket. For millions, it has become a queue.

It feels like a uniquely modern failure — of economies that learned to educate faster than they could employ. It is not. The gap between the number of people trained to rule, manage and prosper and the number of seats actually available for them is one of the oldest destabilising forces in political history. The ancients watched it topple republics and gave it names. A modern data-scientist has given it a mathematics. And on the cyclical view of history — the view this column defends — the graduate glut of 2026 is not a glitch in the machine. It is the machine doing what it has always done.

What does a surplus of graduates have to do with the fall of Rome?

The bridge between a frustrated law graduate in Chengdu and the death of a classical republic is a mechanism the historian Peter Turchin calls elite overproduction. His definition is blunt: "the presence of more elites and elite aspirants than the society can provide positions for" is, he argues, "inherently destabilizing." Societies can absorb a few disappointed strivers. They struggle to absorb a generation of them.

Turchin's data traces how the surplus builds. The engine, on his account, is what he calls the "wealth pump." "The productivity of American workers has been increasing," he told the Niskanen Center, "and for a while until 1970s, the wages were increasing completely in parallel, and then something, boom, happened in the late 1970s" — productivity kept climbing, wages stalled, and the difference was pumped upward. The result is a swelling class of the very rich and the credentialed who expect to join them: between 1980 and the 2010s, Turchin notes, "the numbers of decamillionaires, people who owned a wealth of 10 million or more in real terms… it increased tenfold." The aspirant class grows faster than the positions it is trained to fill, and the competition for those positions turns vicious.

This is Turchin's Structural-Demographic Theory in miniature: two forces — elite overproduction at the top and the popular immiseration the same wealth pump produces below — push a society from an integrative phase, when it is broadly cohesive, toward a disintegrative one, when it is not. The forces are slow, structural and, crucially, measurable. They build for decades before anything visibly breaks, which is exactly why each generation experiences the crisis as unprecedented. It never is.

Why the frustrated graduate is the dangerous part

The crucial move in the theory is what happens to the aspirants who do not get the seat. They do not vanish. Some become what Turchin calls counter-elites — talented, credentialed, organised, and now carrying a grievance against the order that educated and then rejected them. "We are overproducing lawyers by a factor of three to one," he argues, and with AI able to do a large share of legal work, "we would be overproducing lawyers by a factor of six to one." His chosen emblem of the surplus is pointed: "probably the institution that produces the most elite aspirants is Yale Law School."

History, he notes, is written disproportionately by disappointed lawyers and intellectuals: "Lenin was a lawyer, Mao, Castro, Abraham Lincoln… was also a lawyer." This is the hinge of the whole theory. A queue of 12.7 million graduates is an economic statistic. A cohort of educated young people who have concluded that the social contract was broken with them specifically — that the credential they were promised would open doors instead opens onto a wall — is a political one. The poor, in Turchin's grim arithmetic, supply the numbers; it is the surplus elite who supply the leadership. Aristotle saw the same fuel two and a half thousand years ago, and called it by a different name.

Rome ran this experiment first

The clearest worked example is the one Turchin himself returns to: the late Roman Republic. In Secular Cycles, the book he co-authored to test the theory against the historical record, the long crisis that destroyed the Republic is read as elite overproduction made flesh. By the second century BCE, Rome's conquests had enriched a senatorial class faster than the Republic could create the magistracies — the consulships, praetorships and governorships — that its sons were raised to expect. Too many ambitious aristocrats chased too few positions of honour.

The pressure-valve blew first in 133 BCE, when the tribune Tiberius Gracchus — impeccably elite, grandson of Scipio Africanus — broke the rules of senatorial consensus to push land reform and was clubbed to death by a mob of senators. His brother Gaius followed a decade later, and met the same end. The Gracchi are Turchin's archetypal counter-elites: insiders who turned on the order that bred them. What followed was a century of escalating political violence — the wars of Marius and Sulla, Sulla's proscriptions, the rise of Caesar, the civil wars of the Triumvirate — until the Republic itself was quietly retired and a single ruler, Augustus, restored stability on autocratic terms. Polybius's wheel, turning in real time: a republic of the few, decaying into faction, ending in a restored monarchy. It took roughly a hundred years. The graduate in the queue and the senator's frustrated son are, structurally, the same person.

The ancients gave the pattern a name

The cyclical theorists of antiquity were not mystics; they were pattern-spotters working from the cases in front of them. Plato, in Republic Book VIII, sketched a downward sequence of regimes — aristocracy decaying into timocracy, then oligarchy, then democracy, then tyranny — each transition driven by who is shut out of power and wealth. Aristotle, in the Politics, classified six constitutions (three sound forms and their corrupt shadows) and devoted Book V to stasis: the factional strife that erupts when a class feels unjustly excluded from honour or property. His remedy was a large middle class to dilute the resentment at both extremes — advice that reads uncomfortably well in 2026.

The most systematic was Polybius. In Book VI of his Histories he laid out anacyclosis, an explicit cycle of governments: kingship degenerates into tyranny, is overthrown into aristocracy, which rots into oligarchy, which gives way to democracy, which decays into ochlocracy — mob rule — before a strongman restores monarchy and the wheel turns again. The engine in every case is the same one Turchin models: each ruling group eventually serves itself, a rival group with a grievance organises, and the form of government flips. Polybius thought he had watched the early stages of it in Rome — and, as it turned out, he had.

The lawyers' revolutions

If the surplus-elite mechanism is real, you would expect to find disappointed, over-credentialed professionals at the head of the great upheavals — and you do. Scholars who study elite overproduction routinely count the English Civil War of the 1640s and the French Revolution of 1789 among its clearest episodes. Both were led not by the starving but by the educated and ambitious, and in the French case the leadership was so heavily legal it reads like a bar association.

Maximilien Robespierre was a provincial barrister from Arras. Georges Danton was a struggling Paris advocate. Camille Desmoulins was a briefless lawyer turned pamphleteer. Jacques-Pierre Brissot was a writer scraping a living at the edges of respectability. The Old Regime had produced a glut of educated, talented young men — lawyers, journalists, second sons — with elite expectations and no elite positions to absorb them, and when the monarchy's finances finally cracked, that frustrated stratum supplied the Revolution its orators, its journalists and, eventually, its Committee of Public Safety. It is no accident, on Turchin's account, that revolutions are so often a settling of scores by the over-educated and under-employed. The credential is a loaded instrument: it raises expectations precisely as fast as it multiplies the people holding it.

The seventeenth century broke everywhere at once

The theory that underwrites this essay did not begin with Turchin; it began with the sociologist Jack Goldstone, whose 1991 Revolution and Rebellion in the Early Modern World is its founding text. Studying the wave of upheavals that struck Eurasia in the mid-1600s — the English Civil War, the French Fronde, revolts that shook the Ottoman empire, the collapse of Ming China — Goldstone noticed they shared a structure. In the decades before each, populations had grown faster than economies and offices could absorb: a youth bulge, rapid urbanisation, price inflation, and — the familiar refrain — elite overproduction, as prosperity bred more educated aspirants than there were high-status positions. The state, squeezed between rising costs and flat revenues, reached for unpopular expedients such as debasing the coinage or raising taxes, lost legitimacy, and broke.

What makes the seventeenth-century case so telling is its synchrony. Societies with little contact — Stuart England, Bourbon France, the Ottomans, Ming China — convulsed within the same few decades, because the same demographic and fiscal pressures were maturing in each. That is the deep claim of the cyclical view, and the reason it is more than antiquarianism: not that one civilisation copies another, but that similar structures, left to run, tend to produce similar crises. It is precisely the kind of pattern that a person watching only their own country, in their own decade, will always experience as singular — and never is.

America's first age of discord

The United States has been here before, and Turchin's reconstruction of it — in the book he aptly titled End Times — is the most data-driven case of all. In the decades before the Civil War, American higher education and commercial wealth both boomed, throwing up a swelling class of ambitious, credentialed men. "Sons of merchant families chose to go into the law profession," as one summary of his argument puts it, and "the growing number of lawyers included Abraham Lincoln and his peers, who ran for political office." There are only so many offices to go around.

Through the 1850s, intra-elite cooperation came apart. The competition for power fused with the slavery question, and the norms that had held the political class together dissolved: in 1856 a congressman, Preston Brooks, beat the senator Charles Sumner nearly to death with a cane on the floor of the Senate itself. By 1860 the established parties had splintered so badly that four major candidates contested the presidency, and within months the Republic was at war with itself. The bloodiest conflict in American history was not, at root, a poor man's revolt; it was a crisis of a divided, over-supplied elite that could no longer share the spoils. Turchin's data placed the United States on the cusp of another such turbulent peak around the 2020s — a forecast he published, strikingly, in 2010.

China has run it before, too

If the Roman and American cases feel remote, China's own history offers a parallel so exact it is almost uncanny — and it bears directly on the 2026 anchor. For thirteen centuries, imperial China selected its elite through the keju, the civil-service examination. It was a magnificent engine of meritocracy and, eventually, a machine for manufacturing frustrated aspirants: as the population grew and the number of official posts did not, the pass rate at the higher levels collapsed to a sliver. The empire was producing degree-holders it could not employ — a credential glut with a Confucian face.

One of the disappointed was a village schoolteacher named Hong Xiuquan, who failed the provincial examination four times, the last in 1843. After his final failure he suffered a visionary breakdown, emerged convinced he was the younger brother of Jesus Christ, and built that conviction into the Taiping Heavenly Kingdom. The rebellion he launched against the Qing dynasty ran from 1850 to 1864 and killed an estimated 20 to 30 million people — by some counts far more — making it among the deadliest conflicts in human history. The most destructive war of the nineteenth century was set in motion by a man who could not pass an exam. That is the dark logic of elite overproduction at full extension; and it is worth holding in mind when contemplating a single graduating cohort of 12.7 million entering a market with one job for every several applicants.

The modern symptom already has a name: 'lying flat'

The Chinese state, of all governments, knows this history; the keju is its own. And the contemporary response to the credential crush has produced two words that would have made perfect sense to Ibn Khaldun. The first is neijuan, 'involution' — a vicious cycle of escalating competition with diminishing returns, in which students and workers feel compelled to run ever harder for rewards that keep shrinking. The second is tangping, 'lying flat': a withdrawal, a refusal, a decision by a visible cohort of young urban Chinese to opt out of the grind for what they call a 'low-desire life,' reducing their work, their spending, and their willingness to marry, have children or buy property.

Lying flat is what elite overproduction can look like before it turns political — frustration metabolised, for now, as apathy rather than revolt. It is a reminder that the cycle has more than one possible expression. The surplus graduate may become a counter-elite, as Hong Xiuquan did; or may simply disengage, quietly withdrawing the energy and consent on which a fast-growing state depends. Neither is what a government wants from the most educated generation in its history. Both are what the histories predict when the number of aspirants outruns the number of rungs — and a demographic withdrawal, compounding over a decade, can hollow a state as surely as a rebellion, only more slowly.

The Arab Spring was a graduate revolt

The most recent rehearsal is within living memory. The uprisings that swept the Arab world in 2011 are often remembered as bread riots, but the demographic that lit and led them was the educated young. As the UNU-WIDER research programme summarised it, "many observers see youth unemployment as the major reason behind the recent popular uprisings," with frustration running highest "especially among university graduates." Across the region, upwards of 40 per cent of young people were neither in employment, education nor training — and it was, in the words of one analysis, "particularly better-educated youth with high expectations" who took to the streets.

The pattern fits the theory with unsettling precision. Governments in Tunisia, Egypt and beyond had expanded universities and implicitly promised graduates the salaried, white-collar futures their parents had been guaranteed — and then could not deliver them. A generation invested years and savings in credentials that bought nothing. The gap between expectation and reality, not absolute poverty, is the combustible material; a society that educates a cohort for a future it cannot provide is manufacturing counter-elites by the hundred thousand. Aristotle would have recognised the grievance instantly, and Ibn Khaldun the brittleness of the regimes it toppled.

Ibn Khaldun and the solidarity that dies in comfort

Fourteenth-century North Africa supplied the cycle's subtlest theorist. In the Muqaddimah, Ibn Khaldun built his history around asabiyyah — group feeling, the social cohesion that lets a community act as one. His insight was that asabiyyah is self-consuming. A hungry, cohesive group rises from the periphery, seizes the centre, and then — comfortable, fragmented, "increasingly lax and more concerned with maintaining their lifestyles" — loses the very solidarity that won it power. As one dynasty's asabiyyah decays, he wrote, "another more compelling asabiyyah may take its place," and the cycle restarts, typically over the span of a few generations.

The modern rhyme is not a desert dynasty but an incumbent elite that pulls the ladder up behind it — capturing institutions, hoarding credentials and assets, and losing cohesion and legitimacy even as it clings on. The wealth pump is asabiyyah running in reverse: a ruling stratum optimising for its own comfort while the bonds that justified its rule quietly corrode. Ibn Khaldun would not have needed the V-Dem dataset to tell him what a self-dealing elite does to a state. He would have asked only how many generations it had been since the people at the top last had to fight for anything.

The rhyme is not only in the job market — it is in the bond market

Elite overproduction is one face of the cycle; fiscal strain is another. When a political centre can no longer make hard distributive choices, it does what late republics and empires did — it borrows, and eventually it debases. Rome progressively stripped the silver from the denarius until the coin was a token; the modern equivalent is a sovereign-debt market that loses patience with a state living beyond its political means.

In 2026 the warning lights are flashing well outside the United States. France ended the first quarter of 2025 with government debt at 114.1 per cent of GDP — the third-highest in the eurozone behind Greece and Italy — and the political turmoil since its June 2024 snap election is estimated by OMFIF to have added 6 to 7.5 billion euros in extra interest costs over the life of the debt issued. That is the bond market pricing political dysfunction directly, in basis points. In Japan, the 40-year government bond yield broke above 4 per cent on 20 January 2026, for the first time since the maturity was launched in 2007, as investors recoiled from a 21.3-trillion-yen stimulus package. The bond market is simply the modern augur, reading the entrails of a state that cannot reconcile what it has promised with what it can pay.

And it is in the health of the constitution itself

The most direct echo of Aristotle is institutional. V-Dem's Democracy Report 2026, published on 17 March 2026, found that nearly a quarter of the world's nations were autocratizing in 2025, and — strikingly — that "six out of the ten new autocratizing countries… are in Europe and North America," naming Italy, the United Kingdom and the United States among them. "Freedom of Expression… shows the most drastic global decline," the institute reported, and the rule of law was deteriorating in 22 countries. Within a single year, the United States' score on V-Dem's Liberal Democracy index fell 24 per cent and its world rank dropped from 20th to 51st.

Honesty requires a caveat here, and the cyclical case is stronger for making it: Aristotle's stasis meant open factional strife, not slow erosion by censorship and norm-bending. The analogy is real but loose. What the ancients would recognise instantly is not the precise mechanism of decline but its location — the decay now runs through the mature democracies that had assumed themselves immune, the ones that believed they had stepped outside the cycle altogether.

Machiavelli's escape hatch: a pattern is not a sentence

If the essay stopped here it would be fatalism, and fatalism is both bad history and bad politics. The cyclical tradition itself contains the antidote. Machiavelli, reading Polybius in the Discourses on Livy, accepted anacyclosis as the natural tendency of simple constitutions — and then argued it could be broken. Rome's genius, he held, was the mixed constitution: by balancing the one, the few and the many — consuls, senate and tribunes — against one another, Rome slowed the wheel for centuries. Polybius had said much the same: that a mixed government "might help to stabilize republics and prevent permanent anacyclosis."

That is the practical payload of the cyclical view, and the reason it is not a counsel of despair. The pattern is a tendency, not a fate. The forces that drive it — a surplus of frustrated elites, a wealth pump redirecting gains upward, a fiscal centre that will not choose — are visible and, in principle, governable. Throttle the wealth pump so wages track productivity again; widen the number of genuine elite positions, or cool the credential arms race that manufactures aspirants; restore the fiscal compact — and you do to the cycle what Rome's tribunes did for a few hundred years: you buy time. The ancients did not believe in inevitability. They believed in good institutions, badly maintained.

Where the analogy honestly breaks

The strongest objection to a 'history rhymes' essay comes, fittingly, from the man whose data this one leans on. Turchin is emphatic that he is not doing cyclical history in the old sense. "These oscillations are not strictly periodic cycles," he writes, "because there are always exogenous influences that continuously perturb the trajectories," and because societies evolve, "many oscillations we see in history occur around a moving target." Worse, he says, traditional cyclical history fails on its own terms: its "mechanisms producing cycles are either entirely missing, or inadequately specified," and its theories are "not subjected to empirical tests with independently gathered data." The honest version of the thesis claims structural rhyme, not the closed loop of a clock.

There are concrete cracks, too, and intellectual honesty demands naming them. India's graduate-unemployment rate has hovered near 35–40 per cent since the 1980s — a chronic structural feature, not a fresh cyclical inflection, which genuinely complicates any tidy 'turn.' Critics such as Yascha Mounk argue there is no real surplus elite in wealthy democracies at all, only a familiar churn of ambition. And the modern world has brakes the ancients simply lacked: sustained industrial and technological growth that keeps lifting the ceiling Malthus assumed was fixed; nuclear deterrence that makes great-power war ruinously expensive; mass social-mobility channels and welfare states that drain off some of the grievance before it organises. The Arab Spring's graduates, notably, mostly did not seize lasting power; the wave broke, and several states restabilised or relapsed into autocracy rather than completing any tidy 'cycle.' "This time is different" is the most expensive sentence in history — but it is not always wrong, and a column that argues the cycle is real owes its readers the cases where the cycle might, this once, be cheated.

The accelerant the ancients lacked

There is one feature of the 2026 data with no clean classical precedent, and it cuts against the optimists. In every previous cycle, the surplus of aspirants was eventually absorbed — by war, by emigration, by new industries, or by the slow attrition of disappointed people settling for less. Artificial intelligence threatens to run the mechanism the other way: to enlarge the surplus by automating precisely the credentialed, white-collar work that graduates spend years training for. This is the sharp edge of Turchin's lawyer example. "ChatGPT-4 already at this level of technology can take over 45% of what lawyers do," he argues, so that an economy already "overproducing lawyers by a factor of three to one" could soon be overproducing them "by a factor of six to one."

Generalise that across the knowledge economy and the implication is stark. The very technology hailed as the next great engine of growth may, in the medium term, widen the gap between the number of people educated for elite work and the number of elite positions that still require a human to fill them — pouring accelerant on the exact mechanism the histories flag as combustible. The ancients watched arable land run short relative to people; the distinctively modern worry is that meaningful positions run short relative to graduates, and that our cleverest machines are speeding the shortfall. It is the oldest problem in political history, arriving in a genuinely new key — which is also why the comforting thought that technology has freed us from the old cycles deserves to be held up to the light, rather than assumed.

So where does this trajectory point?

Put the brakes and the engine together and the cyclical lens predicts something specific and, mercifully, unspectacular: not collapse, but a sustained season of turbulence. Expect more counter-elite politics as surplus graduates seek a return on their credentials through disruption rather than careers — increasingly online, where a frustrated aspirant can find an audience faster than a job. Expect more volatile bond markets as fiscal centres dither between promises they cannot keep and cuts they cannot pass. Expect more institutional stress in precisely the mature democracies that assumed they had exited history. The cycle, on this reading, resolves not in apocalypse but in a reset — either an orderly one, in which elites accept a smaller share and the ladder is deliberately reopened, or a disorderly one, imposed by the turbulence when reform comes too late.

The histories even tell you what to watch. The leading indicator is not GDP but the ratio between what elites expect and what society can supply: graduate-employment rates, the gap between top incomes and the median, the share of young people who have concluded that effort no longer pays. The lagging indicators are the dramatic ones — polarisation hardening into faction, norms of restraint breaking down, bond markets demanding a premium for political risk. By the time those arrive, as Rome discovered after the Gracchi, the cheap reforms have usually been refused and only the expensive resolutions remain.

The encouraging half of the story is that the expensive resolution is not the only one on offer. Societies have repeatedly throttled the wealth pump and widened the elite before the wheel completed its turn — broadening the franchise, taxing concentrated fortunes, rebuilding the institutions that re-open the ladder. That is the whole difference between a republic that reforms and one that ruptures, and it is a choice, not a verdict. The cyclical view, properly understood, is the opposite of fatalism: it holds that the danger is real, legible and recurrent, which is precisely why it can be met.

For 2026 specifically, the instrument reads early-to-middle, not terminal. The graduate glut is building the pressure; the bond markets are pricing the strain; the democracy indices are registering the institutional wear — but the surplus elite is still, for the most part, lying flat rather than taking up arms, and the rich democracies still hold the fiscal and institutional capacity to act. That is the window the ancients would urge their readers not to waste. Polybius set down his cycle not to induce despair in his Roman patrons but to warn them; Machiavelli catalogued the wheel in order to jam it. This column stands with them: the rhyme is real, the mechanism is old, and the right response to recognising a pattern is not to surrender to it but to break it while breaking it is still cheap.

The value of reading Polybius in 2026 is not a prophecy of doom; it is an early-warning system. A society that recognises elite overproduction for what it is can act on it — as Rome did, for a while, and as imperial China, fatally, did not. The graduate in the queue is not merely a casualty of a slow economy. She is a reading on an instrument the ancients built and a modern scientist recalibrated, and the needle is moving in a direction the histories teach us to respect — not to fear blindly, but to take seriously enough to act before the wheel turns on its own.