China's Favorite American Idea

the "Thucydides Trap" shows how bad history can succeed if it serves a purpose

A few days ago Xi posed a question to Trump inside the Great Hall of the People: Can China and the US overcome the Thucydides Trap?

Not since George W. invoked democratic peace theory to justify invading Iraq has an obscure social science concept been put to such productive misuse. Xi deployed a piece of academic jargon to frame the entire bilateral relationship, and the US president walked right into it.

I say “walked right into it” because besides Graham Allison himself, the biggest beneficiary of the Trap has been...Beijing. The framing does three things that reinforce Xi’s view of global politics. It flatters China’s rise, places the burden of accommodation on the US and, best of all, accomplishes all this using the comfortable non-threatening language of American social science.

The infamous Trap refers to the observation that “the growth of the power of Athens, and the alarm which this inspired in Sparta, made war inevitable.” Allison, a Harvard political scientist, used this line in an Atlantic essay later expanded into a 2017 book, Destined for War. When a rising power threatens to displace an established one, he argues, war results about 75 percent of the time. Allison identified sixteen such cases over five hundred years, twelve ending in war.

The BBC once noted that Allison’s book “has become required reading for many policymakers, academics and journalists.” And as others have complained, this is a huge problem because the concept is a mess. Lawrence Freedman called it “an unhelpful construct” whose case studies tell us “very little of value.”

Criticism of the Thucydides Trap has been so vicious and thorough that I can organize it into five different categories:

The intellectual theft. The idea that power transitions produce war began with A.F.K. Organski, who laid out power transition theory in 1958. His students Jacek Kugler and Douglass Lemke refined the theory into a research program. Then there was Charles Doran with power cycle theory; George Modelski with long cycle theory; Robert Gilpin with hegemonic war theory. In graduate school I had a passing interest in these cyclical theories of war, so I can attest the literature is well developed. All of them preceded Allison’s contribution and engaged with arguments that Allison’s work does not.

In an H-Diplo roundtable on Steve Chan’s book-length critique of the Trap, Kugler was barely able to contain his frustration at Allison’s “rebranding of near seventy years of research.” The moderator, Jonathan DiCicco, put it diplomatically: there was “surprisingly little acknowledgement” of the power transition research program in Allison’s work. Erik Gartzke put it less diplomatically last week, posing a hypothetical: imagine if a student submitted a paper that reproduced an existing literature’s core logic with novel labels but without citing it.

Allison, Gartzke noted, is not an obscure figure and could have afforded the intellectual generosity.

The methodological problems. Steve Chan’s Thucydides’s Trap? (2020) spent a lot of time on the book’s methodological issues. Jack Levy, one of the discipline’s eminence grises, called it “the most thorough critique to date of Allison’s argument and the reasoning and evidence underlying it.” Chan showed that Allison’s interpretation of his own cases was often wrong, his reading of Thucydides was superficial, and the causal logic was underspecified to the point of unfalsifiability.

Richard Hanania (yes, that one) published a takedown in Strategic Studies Quarterly likewise identifying three big problems: unclear definitions (what counts as a rising power or a ruling power?), omitted variable bias (the cases span centuries of different international contexts collapsed into a single data point), and selection bias (Allison chose cases that fit the thesis).

Thucydides didn’t say what Allison thinks he said. Thucydides makes at least two contradictory claims about power and war. The famous line about Athens and Sparta suggests that war erupts when a rising power threatens an established one. But the Melian Dialogue implies the opposite: “the strong do what they can and the weak suffer what they must.” That’s a prediction about powerful states being tempted into war through hubris brought on by strength. During the Cold War, IR theory followed the Melian Dialogue’s logic: parity meant stability and unipolarity would mean over-reach. That was the Waltzian world of realism. Allison simply picked the Thucydides quote that suited his thesis, goes the critique, and ignored the one that contradicted it.

The Eurocentrism. David Kang and Xinru Ma titled their critique in the Washington Quarterly “Power Transitions: Thucydides Didn’t Live in East Asia.” Allison’s sixteen cases are drawn almost entirely from European great power politics. But East Asian power transitions like the Imjin War or the Ming-Qing transition operated by different logics entirely, Kang and Ma argue. They suggest that power vacuums can be as dangerous as power transitions, that internal decline is more threatening to hegemons than external challengers, and that the challenge can come from smaller powers rather than the obvious rising rival.

It may be backwards. This is the Brands and Beckley critique, who in their essay “China Is a Declining Power — and That’s the Problem” and their book Danger Zone argue the most dangerous states are not the ones still rising. If you’re rising, you can afford to wait. The most dangerous states are ones that feel they’ve peaked, and fear the window is closing. Germany in 1914 didn’t think it was going to be endlessly ascendant; it thought the moment of relative advantage was slipping away. Japan in 1941 faced the same problem. If China’s growth is slowing, its demographics collapsing, and its geopolitical environment hardening, then the Thucydides Trap misjudges the question. The danger is not a confident rising power but the frightened peaking power lunging in panic.

So. The concept is stolen, methodologically broken, historically selective, Eurocentric, possibly backwards, and probably useless. Why does it persist?

Partly because it’s a great brand. “Thucydides Trap” is a five-syllable shortcut that flatters the audience and compresses decades of scholarship into a talking point. From a marketing perspective it’s up there with “Clash of Civilizations.”

I don’t mean to sound dismissive. That’s not a trivial achievement. Allison understood what Organski and Kugler and Lemke did not, which is that good packaging will win you more fans than good content.

But another reason for the concept’s persistence is that it does work. It performs diplomatic work and rhetorical work. And the diplomacy and rhetoric it serves is Chinese.

Did you know that Xi Jinping has been invoking the Thucydides Trap since at least 2013? That was only a year after Allison first came up with the concept, and well before it entered common discourse. He used it with Obama. He used it with Biden in Lima in 2024. He used it with Trump last week.

Each time, the framing accomplishes three things:

First, it positions China as the rising power, the Athens to America’s Sparta. For domestic audiences, this is flattering and affirming. China is ascending and the old hegemon must reckon with its arrival. Never mind the debates around China’s slowing GDP growth, shrinking population, or troubled real estate sector. The Thucydides Trap presupposes the rise.

Second, it frames the US as the status quo power whose job is to accommodate the challenger. There is a policy demand hiding inside the historical analogy; once you accept the framing, the burden of adjustment implicitly falls on Washington. The established power has to adapt or risk catastrophe.

Third, it wraps a Chinese narrative in American academic language. Xi is not citing Mao or Sun Tzu but a Harvard man. The concept arrives pre-legitimized with the institutional imprimatur of the American establishment: birthed in the Belfer Center, endorsed by Henry Kissinger, reviewed in the New York Times. It sounds like neutral social science instead of an instrument of Chinese framing.

Allison seems unbothered by all this. He’s described himself as pleased that Chinese leaders engage with his work, and has met with Xi to discuss it. There’s nothing sinister about this of course. Allison wants to prevent war, and academic engagement with policymakers is a good thing. But it seems a little naive. When a concept you created is deployed by one side to frame the terms of competition in their favor, and you respond by feeling flattered, you’ve become part of their game.

There should be a German term for this. Something like Begriffspolitik. The politics of concepts. When Xi invokes the “Thucydides Trap” he is not making a neutral observation about IR theory but placing a bet that American elites will accept an unfavorable framing partly because it comes wrapped in their own prestige.

And you know what — it works! Every cable news segment that explains the Trap implicitly reproduces the Chinese framing. Every article that treats the concept as settled analytical wisdom reinforces the premise that China is rising and America has to adjust.

After Xi’s question in the Great Hall, Trump responded on Truth Social that Xi had called America a “declining nation” — but only under Biden. “Two years ago, we were, in fact, a Nation in decline,” Trump conceded. “On that, I fully agree with President Xi.”

The response shows how completely the frame mogged him. Trump couldn’t reject the premise, only haggle pathetically over the timeline.

I don’t think Allison is wrong that power transitions are dangerous, or that US-China competition carries risks of escalation. But the Trap as a concept has become something distinct from the scholarship it once strip-mined. It’s now a diplomatic weapon that an American academic built and the Chinese leadership has deployed. At this point the real trap might be the Trap itself.