Revisiting the Thucydides Trap

Are we better off if we hang together? Or might we avoid hanging altogether by separating?

Thucydides Mosaic from Jerash, Jordan, Roman, 3rd century AD, at the Pergamon Museum in Berlin

---

A decade ago, I was very worried about the possibility of America falling into what Graham Allison termed “the Thucydides Trap,” a situation where a rising power and an established power facing a possible power transition each make moves that, while individually rational, ultimately lead to a catastrophic war that proves ruinous to both powers’ fortunes. Avoiding that outcome seemed to me the most important goal of American statesmanship, and I was actively seeking innovative ideas on how to do so. I started, logically enough, with Allison’s own book about the applicability of the Thucydides Trap framework to the Sino-American rivalry, which I reviewed for The American Conservative.

At the time, America was transitioning from engagement and integration to something more wary and complex. In the first years of Xi Jinping’s premiership, the Obama administration began their “pivot to Asia,” which involved reducing commitments to other parts of the world (the Iran deal and the muted response to Russia’s seizure of Crimea were parts of this) along with greater investments in the Indo-Pacific region. A key element was the Trans-Pacific Partnership, which would have deepened America’s economic ties to a variety of countries on China’s periphery and created an alternative architecture to compete with China’s Belt and Road Initiative. The objective was to create a situation where China saw more benefit to working within an American-led order, and more risk to challenging it, but the ultimate end-game was a cooperative relationship that would not prevent China from rising to the limits of its potential provided that rise was peaceful. As such, it was plausibly aimed at avoiding the Thucydides Trap, though of course it might have failed to achieve that aim.

We never got to find out if it would have worked, because in the Trump years America changed strategy. We abandoned the TPP, tore up the Iran deal, and launched a trade war with China. Trump’s policy was not as simple as overt confrontation; the trade war ended with only limited concessions, and Trump’s threats were often paired with conciliation and vice versa. We never got to see where precisely Trump was headed (assuming he even knew) for a variety of reasons: because he was never fully in charge of his own administration, because COVID upended things so dramatically (with China in particular self-isolating to an extraordinary degree), and because Trump left office after one term. But the Biden administration came in with a decidedly more confrontational posture vis-a-vis China: deepening our security commitment to Taiwan, establishing the “quartet” of India, Japan, Australia and the U.S., imposing serious export controls on key technologies, investing in rebuilding American industry in strategic sectors, and so forth. The new policy—a response both to an understanding of the nature of Xi Jinping’s regime and to the scale of the China challenge—was overtly aimed at limiting China’s rise as well as containing it.

There was a clear logic to this strategy, which was a response to events (changes in the nature of the Chinese regime under Xi and a recognition of the scale of China’s economic development and military modernization), and I recognized that logic. But the thing about the Thucydides Trap is that it is composed of entirely logical actions—yet each one prompts an escalating reaction. Export controls, for example, are ways of slowing a rival’s technological development, yes. But they also let the rival know that a peaceful rise is now off the table—they are a threat, and a direct one. So they will prompt the rival to redouble their efforts to prevent you from stifling them, and if those efforts succeed then you’ll have even less leverage than you had before, and a rival more convinced than ever that they will have to be confrontational to achieve their aims. New military alliances, for another example, arguably strengthen your position by presenting a united front against a rival that the rival would have a hard time matching in the event of a conflict. But they also prompt your rival to assemble offsetting alliances of their own among disgruntled powers. The result is more opportunities for conflict and limited diplomatic flexibility (because your credibility as an ally is now on the line when those conflicts present themselves). The Biden administration strategy—for all its logic—amounted to steering the ship of state right into the Thucydides Trap, and hoping to emerge victorious.

It’s too soon to know whether that strategy would have worked—and we’ll never know because Trump is back in charge, and already brusquely revising international relationships just as he is domestic ones. But the early returns from the Biden strategy should be sobering to anyone paying attention. By the end of Biden’s single term, China and Russia were more closely allied than ever, with both Iran and North Korea increasingly integrated into a common security and economic bloc. China’s military capabilities have grown significantly, to the point where the chorus warning that we could lose a shooting war with them over Taiwan has become deafening. They have parlayed their civilian manufacturing prowess into the leading global position in crucial sectors with military applications like electric vehicles and drones, and export controls have not stopped them from making huge advances in fields like artificial intelligence. America retains a significant lead in a variety of areas, including aerospace, and of course our military remains vastly more battle-tested. But the last decade’s developments should leave us little confidence in our ability to sustain our lead over the long term.

Matt Yglesias, in an excellent post from last week, points out how much of a surprise it is, or should be, that we’ve come to this pass. It was an article of faith for a long time that Chinese economic development and integration into the world economy would lead both to liberalization and to an increasing stake in a peaceful international order. When both of those proved untrue, the conviction remained that an illiberal China would inevitably hamstring its own rise—because an authoritarian political system could not foster sufficient technological and economic innovation, and because China’s arrogant and self-interested behavior internationally would push other countries increasingly into America’s arms. At this point, we can say that both of those convictions are also false. China’s political system has not stopped its economic rise (and if it slowed it, that just means that China would have risen even farther and faster with a more open political system). And America’s severe financial sanctions on Russia and support for Israel’s war in Gaza have led much of the world to prefer neutrality to shelter under the American aegis.

Where does that leave my quest for a way out of the trap? Yglesias ends his, as I say, excellent post by saying how vital it is now for America to cooperate with our allies to confront the China challenge, since if we don’t hang together we will surely hang separately, and bemoaning that Trump is doing the opposite right out of the gate. I’m not going to defend Trump’s behavior, much of which I find baffling. But I do have to question the “now more than ever” assumption that underlies Yglesias’s conclusion. The Thucydides Trap is still out there. If steering into it hasn’t achieved the desired results, it’s not clear to me why gunning the engines on the same course would achieve better ones.

The key problem is that America’s primacy may no longer be sustainable, and our hegemony almost certainly is not. We may not be able to remain “top nation” in all respects—indeed, we already aren’t, as China is already by far the dominant manufacturing power on the planet, and in purchasing power parity terms has the largest economy. (Side note: PPP isn’t a perfect way to compare economy sizes, but neither is nominal GDP based on current exchange rates; they each have some validity depending on what you’re focusing on.) And while we still have the world’s most capable military, we quite plainly no longer are the sole world power shaping or responding to events around the globe. The Biden strategy was to use alliances as a force-multiplier, which is how America approached (and won) the Cold War. But the more dynamic and economically important China becomes, the more expensive it is for other powers to choose to affiliate exclusively with an American-led bloc rather than side with China or—best of all—work with both sides. To hold its alliances together, then, America has to offset their costs—and that cost to us we have no way of offsetting.

This is the central dynamic in Europe today. Germany, Europe’s manufacturing heartland, is at least as threatened by China’s emerging dominance in a sector like electric vehicles as America is. But despite the size of its economy, Germany is a much smaller economic power than either America or China. Does it make rational sense for Germany to ally with America to aggressively restrict Chinese electric vehicle exports? Or does it make more sense for Germany to bandwagon with China and look to become its European subsidiary? Germany’s own behavior over the past year—it was the key vote against tariffs on Chinese EVs within the EU—suggests the latter. Meanwhile, America’s main leverage to move Germany toward the American position is the promise to continue bearing the largest burden of supporting Ukraine in its war with Russia. In other words, the way our alliance with Germany works is that we buy cooperation by extending our security commitments. Is that a net positive trade for America in its contest with China? I don’t think so.

America’s Asian allies are more directly threatened than Europe is by China’s rise from a security perspective, which is why America has deepened its relationship with Vietnam, supported Japan in its efforts to rearm, etc. But some of the same conflicts apply even more dramatically in the economic sphere. It’s notable that Chinese investment in Vietnam has ballooned massively since the United States began its decoupling from China proper. If America tries to block Vietnamese exports because they are effectively an economic arm of China, won’t that undermine our security relationship? And if we don’t, won’t our economic “containment” of China leak badly? Even Japan cannot afford to wall off China.

The problem across the board is that most distinctive asset America brings to the table in these negotiations is a willingness to extend ourselves militarily, while our economic interests are not obviously aligned with our own allies. Maintaining a broad, united front against China therefore puts America in the position of bribing other countries to maintain that front. This costs America more than it does them, and thereby strengthens China’s relative position. China can bide its time, invest in its own growth and development, and benefit from cheating by our allies.

America was victorious in the Cold War in large part because we were the dominant center of by far the most dynamic economic bloc in the world. If we re-run that same play against the most dynamic economic power in the world, should we expect the same results? I don’t see why we should.

Now, one answer is to focus on building up American power directly: making our economy more efficient and resilient, making our military more effective, investing to maintain or restore our technological edge and fostering demographic expansion. All of these are valid and important goals regardless of our alliance strategy, and if they work then America may wind up retaining or regaining a large measure of dominance. But are we confident enough that we will succeed that we don’t want to rethink the current trajectory toward Cold War II?

Consider our relationship with India, because doing so highlights an alternative possibility for the world’s future than bipolar competition between America and China. India is the only major power that could become a true rival to China in some number of decades. It is considerably behind China in industrialization, but it is moving rapidly up the curve, and China’s own rise proves that even very large and poor countries can make extraordinary strides in a shockingly short time. Not that they always do—but we have proof of concept. And India has a notable demographic advantage over China, with a much younger overall population and a fertility rate that has not yet collapsed to dangerously low levels. For all these reasons, India figures heavily in armchair strategists’ dreams of a “free world” bloc truly capable of rivaling China on every dimension.

Except . . . why would India want to be part of such a bloc? India, unlike Japan and South Korea, pointedly refused to line up behind America’s severe financial sanctions against Russia, preferring instead to maintain its distance from the conflict and maintain a cordial relationship with Russia. (It’s worth noting that India has similarly maintained friendly relations with both Iran and Israel, and that Israel tried for as long as possible to maintain friendly relations with both Russia and Ukraine.) Quite sensibly, India pursued its own national interest over America’s conception of the collective interest. If India grows more powerful economically and militarily over time, we should expect it to become even less willing to line up consistently with any other power’s interests, or with their notion of what that collective interest might be. India will happily benefit from our interest in building up rivals to China—but it isn’t going to be an ally of anything but convenience.

Does that mean the United States should be wary of building India up, or that we should try to tether our support to commitments to increasingly integrate into an American-led bloc? Perhaps—but perhaps not. Maybe we should welcome and even assist India’s rise precisely because its stubborn independence complicates the international picture and makes it less likely that China will ever come to dominate the future—even though it also means that America will be less likely to dominate it. In a sense, that was the logic behind our original opening to China in the Nixon years. We had no reason to think China would thereby become a reliable ally to the United States. But we knew that our opening would seriously complicate the Soviet Union’s own strategy, and that in itself would open up more room for our own maneuver.

And perhaps we should bring that same perspective to bear on our existing alliances. I’ve been beating this drum for a long while, but its urgency has only grown with time. Our alliance system is structured to turn our allies into dependencies. America protected Europe, but also prevented Europe from becoming a serious rival to the United States, and over time, Europe atrophied to the point that it was less and less useful even as an ally. So long as America was the overwhelmingly dominant world power, that wasn’t necessarily a bad trade, but I do think it’s a bad trade for a world where we face a rival that is potentially stronger than we are. A stronger Europe would inevitably be less pliant, less easily bribed. But it would complicate our rivals’ calculations more than a pliant but weak Europe does—and that matters more, doesn’t it?

We do need good relations with other powers, big and small. But maybe our alliances should be alliances of convenience in the sense that they extend as far as mutual interest dictates, as our relationship with India does. Could we build more constructive relationships with other powers, including ones that are currently hostile, on that basis? I don’t know. Nor do I deny that a world in which America, China, India, Europe, Russia, Japan, Brazil, Turkey, Iran and who knows what other mid-sized powers jockey for position could be very messy and unpredictable. But if we aren’t sure we are going to win a game of bipolarity—and I don’t think we should feel sure—maybe we should be trying to figure out how to foster a true multipolarity instead.

On some level, that means avoiding the Thucydides Trap by giving up; the reason it’s a trap is that the dominant power wants to remain dominant and therefore looks to weaken its rising rival and augment its own power with alliances so as to remain so. But there is a difference, I think, between abandoning the quest to maintain dominance ourselves and letting the rival become the new dominant power.

At least I want to believe there is. I don’t know precisely how that difference plays into specific policies. I certainly don’t see how it implies that we should just pick fights with our allies just to show them who’s boss; trading a fading world hegemony for a more obnoxious dominance of our own hemisphere doesn’t feel like a very good trade. But it has to imply some kind of change, in the terms of our relations with our long-time allies and in terms of our approach to other rivals. Doesn’t it?