Peter Turchin’s empty prophecies

The social scientist has made a career from predicting global instability. But in the new book End Times, his analysis produces nothing but banalities.

By William Davies

In January 2010 the scientific journal Nature did what publications so often do at such moments in the calendar, and invited a range of scientific experts and policymakers to speculate on “where their fields will be ten years from now”. The various responses, which were collected under the heading “2020 Visions”, included a brief contribution from the biologist and complexity theorist Peter Turchin.

Turchin, who is based at the Complexity Science Hub in Vienna, is a pioneer of the field of cliodynamics, in which scientists seek to establish long-running historical trends by studying large statistical datasets. He had spent the previous few years trying to spot patterns of political instability across societies. His 2010 response to Nature included what, from a historian’s or sociologist’s perspective, would have appeared a simplistic observation: “In the United States, 50-year instability spikes occurred around 1870, 1920 and 1970, so another could be due around 2020.”

Turchin is certainly not the first social scientist to identify long-term historical patterns. The Soviet economist Nikolai Kondratiev (who Turchin cites approvingly) had noted how new technologies unleash long-term waves of productivity growth, which then dissipate until a new one comes along, at regular intervals. The economic anthropologist and historian Karl Polanyi introduced the concept of the “double movement”, whereby long phases of market expansion eventually hit their limit, and are then rebuffed by equally long phases of social protection.

More ambitiously, the Marxist historian Giovanni Arrighi observed how, ever since the Renaissance, the global economy has witnessed the rise and fall of successive state hegemonies, with each phase shorter than its predecessor, and each terminating in a crisis of “financialisation”. Arrighi’s The Long Twentieth Century (1994) made no concrete prediction, but nevertheless interpreted the rising power of Wall Street since the 1970s as a symptom of imminent US decline, followed perhaps by an era of Chinese global leadership.

[See also: Is neoliberalism really dead?]

When 2020 finally arrived, Turchin lucked out. The combination of Covid-19 with the protests that followed the murder of George Floyd by Minneapolis police officers – which escalated to become one of the largest social movements in US history – added to the already unhinged presidency of Donald Trump. In witnessing these events unfold, few would disagree with a diagnosis of American “instability”. Yet what followed is a morality tale about the impoverishment of public intellectual discourse, and the mystical cult of “Science” that continues to exert its hold over the media.

“This researcher predicted 2020 would be mayhem,” announced Time magazine excitedly. “Ten years ago Peter Turchin, a scientist at the University of Connecticut, made a startling prediction in Nature,” declared the Economist. The year 2020 was turning out “roughly as the Seer of Storrs foretold ten years ago”, gushed the Atlantic. “The scientist who predicted 2020’s political unrest reveals what comes next”, ran a Vice headline. Turchin leaned into his new celebrity: “The science behind my forecast for 2020”, he headlined one of his blogposts.

Few remembered (or cared) that Turchin had only been responding to Nature’s invitation to speculate about 2020, at a time when the US was already pulling apart under the strains of the 2008 financial crisis and deepening political polarisation. Turchin had attained the status of a prophet. In 2023 we are presented with his new book, the none-too-subtly-titled End Times, which promises to reveal how he foresaw the future, and what else we (or the US) might now be in for.

Predictions are fun. When sports pundits and pollsters stick their necks out to predict the outcomes of contests, it’s all a pretty good laugh, especially when someone ends up with egg on their face. Who can forget the sight of the academic and commentator Matthew Goodwin eating his own book live on Sky television in 2017, as he’d promised to do if Labour polled more than 38 per cent in that year’s general election. Twitter has added a playfulness to all of this, where people’s spookily good, and comically bad, calls can be dredged up from the archive.

Yet is it the job of social science or history to try to predict the future? Some would say that it is. In his 1953 essay “The Methodology of Positive Economics”, Milton Friedman argued that the value of any science, social or otherwise, lay entirely in its capacity to make accurate predictions, and owed nothing to its descriptive plausibility. Thus, regardless of whether human beings are as economists portray them, it only matters if things turn out as economists predict. A theoretical model is not a representation of the world, but a tool for controlling it – an engine not a camera.

Friedman’s behaviourist manifesto aimed to position economics as closer in spirit to the natural sciences than to its “social” cousins, anthropology and sociology, and to have nothing to do with such disciplines as history or philosophy. The urgency generated at the time by the Cold War saw the Pentagon funnel billions of dollars into fields of psychology, mathematics and computer science that promised to render the social world predictable. It has since become a commonplace for statisticians to offer forecasts for phenomena such as inflation, GDP growth or flu outbreaks.

[See also: The triumph of cash]

This faith in the predictive power of quantitative methods has experienced various surges of optimism (and investment) over the decades, the most recent of which is the one surrounding algorithmic data science. It was around 2008 (the same time, not coincidentally, that Google and other tech companies were collating the world’s data) that breathless anecdotes began circulating in Wired magazine and management journals, of the algorithm that knew a woman was pregnant before she did, or the marketer who could predict a consumer’s shopping preferences based on their music downloads. The journal Cliodynamics (which Turchin edits) was founded in 2010, during the same hype cycle.

Elevating prediction over description comes at some cost, however. Mathematical models are esoteric and opaque to the public, especially when (as in the case of surveillance capitalism) those models are proprietary secrets. Description, on the other hand, produces shared narratives and explanations. There is a reason why works of history command prominent positions in the windows of Foyles bookshop, while econometric models circulate in specialist journals and expert networks. When experts do seek to make the transition from specialised intricacy to a general readership, they often end up resorting to a mixture of mysticism (“This is all based on Science!”) and crude credentialism (adding “PhD” in their byline) to get the layperson to trust what they can’t really understand.

This is the dilemma that runs through End Times, a book that alternates between superficial bird’s-eye views of historical epochs, derived from prominent historians and social scientists, and bold declarations of what the author’s all-powerful (but largely unexplained) model and database can yield. As a descriptive account of what’s wrong with the US, Turchin’s story often repeats the economist Thomas Piketty’s account of inequality trends, and the economists Angus Deaton and Anne Case’s findings on life expectancy and morbidity, and the recent revival of the sociology of elites.

Turchin does cite lavishly, quoting long passages of contemporary historians, but the implication is that conventional history is not what he’s here to do. There is little to learn from such observations as “Race has been one of the most important issues in American politics, all the way from its beginning to this very day.” Meanwhile, the claim that “Privatisation [in the former Soviet Union] was an incredibly corrupt and violent process, with the winners literally walking on the corpses of their less lucky competitors” belongs in Private Eye’s Colemanballs.

So what does Turchin have to offer? The central theory is a relatively simple one. There are four drivers of political instability, gleaned from cliodynamics: “popular immiseration”, “elite over-production”, “failing fiscal health and weakened legitimacy of the state”, and “geopolitical factors”.

The first two (and especially “elite over-production”) receive the most attention, but as Turchin proceeds in his discussion of the present day, it becomes less and less clear what he is adding to Piketty’s finding, that wages have stagnated while the number of billionaires has grown. The political analysis then comes down to a crude economic determinism, that at some point people will not stand for this any longer, though this is hidden behind jargon regarding a mysterious “wealth pump” that transfers money from the poor to the rich.

Should anyone doubt this analysis, it comes verified by “CrisisDB”, the statistical database containing everything that Turchin’s team has collected from the study of societies throughout history. On the issue of “elite over-production”, Turchin fears that the US is now churning out too many graduates for what its labour market can support. “History (and CrisisDB) tells us that the credentialed precariat (or, in the jargon of cliodynamics, the frustrated aspirant class) is the most dangerous class for social stability.” Flatlining wages (which he attributes to high levels of immigration, citing Angela Nagle) makes the US ripe for a revolution.

In case any of this sounds too apocalyptic (or empirically falsifiable?), Turchin throws in some caveats for good measure. It’s not enough for unwieldy economic forces to be present: there also needs to be a large proportion of “radicals” in the population, and the masses need to be organised, further variables that cliodynamicists need to feed into their model. Fortunately, for the US, neither of these things has happened. Crisis averted, or at least delayed. One wonders if the real insight of CrisisDB is that political instability is most likely when politics becomes unstable.

The climax of the book, the money shot of this whole exercise, occurs on page 200: “Now that we have the MPF engine [Turchin’s model, still not really explained by this point], let’s use it to investigate the possible trajectories that the American social system could take beyond the 2020s.” Finally, after considerable limbering up, the main event arrives.

Spoiler alert: MPF’s big reveal is disappointing. “Elite over-production” and “popular immiseration” will continue through the “turbulent Twenties”, producing greater and greater violence. Yet the effect of this violence is, counter-intuitively, to turn “radicals” into “moderates”, until the country calms down for 40 years before erupting again in 2070. The problem, Turchin explains, and the reason why society will become unstable every 50 years is the “wealth pump”, which continues to divert money from the poor to the rich. Only with the kinds of liberal reforms witnessed in Britain in the late 19th century, or the US after 1929, will the “wealth pump” be switched off and political instability put to rest.

Rather like Deep Thought, the super-computer in The Hitchhiker’s Guide to the Galaxy, CrisisDB and MPF appear to have been fed the most fundamental, existential questions of politics, and answered with a banality: moderate redistribution or barbarism. More than once, I wondered if this whole project was a hoax. (If rationalists can pull pranks on critical theory journals by sending in fake papers to see if they get published, why shouldn’t critical theorists do the same to Nature?)

Who really needs End Times, when one can read the sweeping historical work of Piketty, Arrighi or Eric Hobsbawm instead? This is TED-Talk-lit that so vastly over-promises that the under-delivery is baked in from the start. The insistence that nothing can escape statistical model-building is ultimately a power grab, a determination to own and to control, not just the future but also adjacent scholarly disciplines. As Theodor W Adorno wrote of philosophical system-builders, this is a symptom of the “belly turned mind”.

End Times: Elites, Counter-Elites and the Path of Political Disintegration
Peter Turchin
Allen Lane, 368pp, £25

Purchasing a book may earn the NS a commission from Bookshop.org, who support independent bookshops