Deprecated: Hook custom_css_loaded is deprecated since version jetpack-13.5! Use WordPress Custom CSS instead. Jetpack no longer supports Custom CSS. Read the documentation to learn how to apply custom styles to your site: in /var/www/wp-includes/functions.php on line 6078
The prospects for war with China: Why I see a serious chance of World War III in the next decade - Chris Blattman

Chris Blattman

Close this search box.

The prospects for war with China: Why I see a serious chance of World War III in the next decade

When I wrote my book on war, I avoided ongoing conflicts because I didn’t want the book to be dated the moment it came out. The roots of war and the paths to peace are timeless, and I wanted examples that made this permanence clear. Still, it was hard not to read obsessively about the conflicts facing today’s world and develop opinions. Over the next few months I plan to write a series on some of the most troublesome conflicts, and what I think social science has to say about them.

I’ll begin with the one that worries me most, by far: a Chinese invasion of Taiwan, one that draws the United States and its Pacific allies into prolonged fighting. In other words: World War III.

As many of you know, I am no China expert. Instead, this is what I gleaned from trying read as much as possible. My hope is that some of these ideas are as surprising to you as they were to me, including:

  • Why America’s policy of strategic ambiguity is targeted at the Taiwanese, not China
  • Why most of the things we read about in the news—from Pelosi’s visit, to Biden’s gaffes—are distractions from the important policy choices
  • Why I think the two most significant events of the 21st century so far are China’s Hong Kong security law and Xi Jinping’s centralization of power (on par with, or even more important than, 9/11 or Russia’s invasion of Ukraine)
  • Why the United States has been remarkably restrained and stable (even naive) in its China policy—maybe to its detriment—and much larger shifts seem inevitable to me.

I am keen for feedback. Comments are open.

The dispute

The core issue is how autonomous Taiwan gets to be. In principle, there are lots of options. At one end of the spectrum is an independent and sovereign state of Taiwan, at the other is complete unification into mainland China under the current regime. The status quo is, of course, in between—Taiwan is de facto politically independent, while most countries and the UN acknowledge that Taiwan is part of China (some more reluctantly than others).

This status quo worked for a long time for a few reasons:

  • China never had the military capability to invade, not least because amphibious landings are really difficult
  • Occupying a huge and possibly hostile island is also hard, especially because the style of Chinese rule is to try to control society through Communist Party cells
  • Besides the usual costs of war, invasion could unravel China’s economic entanglements with the rest of the world, and that could endanger growth
  • Deterrence: Taiwan’s allies supplied it with weapons and soft assurances that, if invaded unprovoked, Taiwan could probably expect military support

Another way to put this: everyone knew that invasion was the worst possible way to settle this dispute. War would be ruinous for both sides. Even if China’s leaders hated the status quo, it made sense for them to grudgingly accept it. Same for Taiwan. Political leader seldom get what they want.

Mapping it out

It’s useful to use a diagram here, because it will clarify why the status quo may be unsustainable.

Let’s take the spectrum from total unification to complete independence and draw it as a line. For argument’s sake, let’s also suppose China and Taiwan + allies have had even odds of winning a war (hence the little black marker at the middle of the spectrum). I’ve put the status quo is just to the right of that for illustrative purposes, but the exact position doesn’t matter. The key intuition: rivals usually find a compromise that’s roughly proportional to their ability to burn the house down. That’s bargaining power.

Arguably, 50 years ago, the balance of military power and status quo was further to the right (more autonomous). This was before China’s rise—back when many countries recognized the government in Taiwan as the sovereign of the island, if not all of China. But as mainland China’s economic and military might grew under the Communist Party, the balance of power shifted leftwards. China pressured the rest of the world to recognize its rights over the island, and most nations complied. China hauled the status quo in their direction.

Now, if that was so successful, why didn’t China just invade, and pull the status quo all the way to the left? Or why hasn’t a fervent Taiwanese president come to power and declared independence, pulling it all the way to the right? The answer is simple: war would be costly, not just in terms of lost life and armaments, but destroying a lot of what makes Taiwan valuable (like microchip factories, and highly skilled workers) and possibly crushing Chinese exports and growth. Costs like these are why diplomacy, cajoling, and dirty tricks usually win out over war.

The cost of the war creates a whole range of compromises both sides prefer to fighting. Suppose the cost is really high, represented by the bracket centered around the 50-50 split in military might. Those costs mean that China would fight rather than accept any Taiwanese autonomy to the for right of the bracket. Taiwan would fight rather than accept anything to the far left of the bracket. In between, there’s a whole range of deals. If one side were to find a clever way to change the status quo, the other side would grumble (or conduct military exercises, or seek sanctions, or something) but they wouldn’t invade.

(By the way, I’m pulling this example from a terrific 2016 article on China-Taiwan relations by Scott Kastner. Some of you will recognize the bargaining approach to war, which comes from Jim Fearon. But the idea has been around much longer than that, and goes back to a literature on strikes, litigation, and business, including the famous Coase Theorum. It is a very deep and important idea in political economy: striking and battling are inefficient and so, most of the time, they’re avoided.)

So what happens when one side—China—experiences the greatest and most sustained economic miracle in human history, spending some of that wealth on weapons, military research, troops, and information control? Maybe China advances to a 70% chance of victory. If that’s true, the balance of power shifts to the left, taking the range of compromises with it.

The status quo is still preferable to fighting, but (let’s be honest) it looks more fragile.

Why war wasn’t an issue up to now

This helps explain Beijing’s patience up to now. For years, China experts have been telling us that Beijing was happy to tolerate the current status quo but leave the future open. They were betting on the continued rise of China, and the decline of the West. The bargaining range would keep shifting leftwards, and eventually the Taiwanese would have to concede some autonomy. Because the cost of not doing so was too dear.

This also helps explains Taiwanese and Western patience. For four decades, the liberal-capitalist world made a bit bet that massive development, economic integration, and information technology would eventually make China a more open society. If that happened, mainland Chinese wouldn’t care so much about unification, and Taiwanese wouldn’t find somewhat closer relations so objectionable. I think that was the right bet to make at the time, given what we knew. But it is a bet the West lost.

Both sides have now begun to lose their patience. The West lost their bet. And, on the Chinese side, growth has slowed under Xi Jinping. A lot. (This slowdown was happening before the tech company crackdown and before zero covid policies. Those policies have probably made growth even worse. Lately the government has even resorted to delaying the release of economic statistics. That seems desperate.) If China enters a low-growth state, then the leftward shift of the bargaining range no longer looks so inevitable.

Still, even if it’s goodbye to patience, that doesn’t mean war. These are painful realities for all sides to accept. But that’s politics. Like I said, no government gets what it wants.

How each side is responding strategically

Instead, both sides are doing their best to improve their bargaining position—to move that balance of power in the direction of their preferred pole.

Take Taiwan’s military strategy. It involves buying conventional planes, tanks, and missile defense systems, plus training a reserve army, and (increasingly, it seems) trying to make a landing and occupation very painful. These are attempts to keep the status quo in scope. The same goes for Taiwan’s allies. When the United States puts an aircraft carrier closer to the South China Sea, bans the sale of microchips to mainland China, urges Japan to re-arm, or sells nuclear submarines to Australia, it’s trying to nudge the balance of power rightwards.

The other thing all these military investment do is increase the costs of war. If the expected destruction from war rises faster than the value of controlling Taiwan, that widens the bracket, creating more and more buffer room for the status quo. Some of you will recognize this as the logic of deterrence. It feels paradoxical: arm in order to avoid fighting. The logic works, often. But often is not always. This is terrifying, because if deterrence ever failed, the resulting war is going to be a doozy.

Fortunately there are other ways to make war costly other than arming. These include sanctions or the promise to decouple China from the world economy. Those policies are designed to make war more costly to one side, and to keep the status quo feasible, even as China grows in power.

So far, this tale of politics should make us breathe a little easier. It implies we should see lots of provocative military exercises, bombast, armaments, tense negotiations, and hostile rhetoric on both sides, but no war. That’s true most of the time. Peace is almost always your safest prediction.
But there are a few reasons I am worried. Normally an optimist, I see a real chance of a Chinese invasion in the next 10-20 years. And I see a significant chance that the United States, Japan, Korea, and Australia get pulled in.

The path to war

How does that war happen? I see six dynamics. Three involve a shrinking in the relevant costs of war—the costs that one side or the other actually takes into consideration, weighed against the benefits of fighting. Going back to our diagram, I worry about a situation that looks like this:

The costs of war have shrunk, and so the status quo is outside the new range of mutually acceptable bargains. Here is how that could happen:

  1. Increasingly unchecked Xi. Xi is centralizing and personalizing power. Personalized autocrats do not need to consider all the costs of war. The death of soldiers or the burdens on ordinary people—he is increasingly insulated from all that. So the costs he weighs are less than the full costs of war.
  2. Ideological Chinese. Xi, many members of the ruling party, and even many Chinese seem to really, really want Taiwan in their fold. I am not surprised. If, at the end of the U.S. civil war, the confederates had retreated to the Florida keys and set up a rump state, 150 years later I am confident most Americans would refuse their right to exist–especially if that rump state represented a model of society most Americans found threatening or repugnant. They’d be willing to pay some costs to win on sheer principle. Those nationalistic, ideological benefits would shrink the net cost of war, making Americans more willing to wage it. Obviously this analogy is a stretch. I make it only to communicate to Westerners in some small way why so many Chinese might be willing to bear some costs of warfare.
  3. Decoupling. The Chinese economy is already getting decoupled (a little) from the world economy. America is trying to reshore industry. China is rushing to become self-sufficient in microelectronics. China is still the world’s workshop, Taiwan and China are still major trading partners, and social and familial ties between the two polities are in some ways stronger than ever. But thing are moving in a direction that makes war less disastrous.

Still, better to deal than fight

Now here’s something important: Normally, none of these three would be a problem for peace. Power balances change all the time, and opponents compromise and relinquish power because fighting is ruinous. In the scenario I just mapped out, we should expect Taiwan to grudgingly accept a loss of autonomy, like this:

This is what most rivals do most of the time. If Russia grows more powerful vis-a-vis Belarus, it expects (and has gained) a greater say in local politics. If World War II transforms the United States into the world’s superpower, it asks (or pays) the Egyptians and the Colombians and the Kenyans to adopt their policies, and (to some degree) these countries comply. We don’t have to like it, but that is the realpolitik world we live in.

In the case of China, there’s precedent for the new more unified status quo: “One China, Two Systems”. Hong Kong occupied that position.

The obvious problem with this example is that One China, Two Systems didn’t last. For years, protesters in Hong Kong tried to move their autonomy to the right—not past the red line, but very close to it, trying to get the most freedoms possible under official Chinese sovereignty. (This is what adversaries do—they connive politically to improve their position without courting war.) But Beijing outmaneuvered them. It imposed a new set of security laws and a crackdown that severely pulled Hong Kong towards the unification pole. Again, Beijing didn’t do this so severely that the new status passed Hong Kong’s red line. But pull it towards unification they did.

In principle, the same could happen to Taiwan. With sufficient Chinese economic and military growth, we could all wake up one morning to a blockade and an ultimatum, a tense standoff that, after some months, leads to Taiwan acceding to a different arrangement. There are dozens of ways this could unfold. In many of them, the West would grow irate, and sanction the bejeezus out of China. But so long as a degree of freedom and autonomy was preserved, the Taiwanese might not revolt, and the world might not aide them—so long as China’s actions didn’t cross Taiwan’s red line. Thomas Schelling, the famed game theorist, called such a maneuver a “strategic move”. Every rival would love to pull one off.

The question, then, is why could this strategy fail?

The further path to war

I see three main ways these events lead to war.

  1. Ideological Taiwanese and Americans. To some people, liberty is not divisible. They will reject realpolitik and fight on principle. Damn the costs of war; some things are worth fighting for. Now, it must be said: most oppressed peoples don’t behave like this. The North Koreans, Russians, and Belarusians have not risen up against their overlords. Throughout history, most of the repressed do not revolt. Only some do. In the 18th century the American revolutionaries refused semi-sovereignty (while my Canadian ancestors accepted it). In the 21st century, Ukrainians rejected an offer of quasi-independence from Russia. Perhaps the Taiwanese would do the same. If they did fight, I expect a great many Americans would support them (including me)—not only for strategic reasons, but because of a belief that some things are worth that price. Biden and others increasingly frame this rivalry in ideological terms—as a fight between autocracy and democracy. This is both admirable and risky. We don’t talk about these ideological sources of bargaining failure much, but they’re powerful.
  2. A Chinese commitment problem. The crackdown on Hong Kong strikes me as one of the most significant and saddening foreign policy events of the past decade. Why? Because in a matter of weeks, China lost all credibility that it could offer Taiwan a One China Two Systems deal like the new status quo I illustrated above. As a result, the Taiwanese can reasonably worry that any concession to China would lead to another demand, then another, until freedoms are sliced off slice by slice, like a salami slowly being carved. If true (and that’s a big, unknown “if”) then it’s better not to concede that first slice.
  3. A deluded Xi. Finally, there’s a chance that an increasingly autocratic Xi grows more and more isolated from the truth—that Chinese generals who naysay get early retirement. This happens in many countries, including democracies, but personalized autocrats often end up the most insulated from accurate information. That’s why, to me, the other incredibly significant policy event this century is Xi’s personalization of power. Especially because (like many personalized rulers) he seems to rate his own judgment highly and like taking risky bets—like zero covid, or cracking down on the Uyghurs or tech barons. The chances that Xi underestimates the costs of war, or overestimates his chances of victory, should worry all of us (especially his fellow Chinese).

(One cautionary note: As far as I can tell, foreign governments and intelligence have almost no idea what the Chinese leadership really thinks or plans. The circle of insiders around Xi seems to be unusually small and tight. This means that outsiders have to rely on political rhetoric and propaganda, and to try to discern policy from it. According to this terrific interview with Chris Turner: all the talk you hear about China’s timeline for reunification with Taiwan—that must be coming from Americans reading tea leaves, because he doesn’t see that conversation or information coming from within China.)

Altogether, I worry that some combination of these six factors could bring about a war in the next decade. The clearest case I’ve seen comes from Oriana Skylar Mastro. Even if war doesn’t result, she lays out the military options available to China short of full invasion (such as a naval blockade). These might extract concessions from Taiwan—forcing it to shift to the left on the unification-independence spectrum.

(By the way, for the political science nerds in the audience, all I’ve done here is take the arguments from China experts and done two things: framed them in the light of “rationalist warfare”, while also drawing attention to a couple of under-appreciated failures outside that paradigm—ideologies and misperceptions. Merging these different social science traditions was really the whole point of my book.)

Judging the US response

Let’s evaluate some US policies in this light:

  • Increasing US (or other) military presence in the South China Sea: This includes Australia, Korea, or Japan arming. But it also includes people arguing that the United States should make serious defensive investments in the Pacific. Arming the region increases Taiwan’s bargaining power versus China, and increases the cost of any war, and thus probably lowers the likelihood of an invasion. Also, it likely accelerates the arms race with China (which will be expensive for China, especially if the days of high growth are gone). The main down side: if an invasion does happen, there’s a greater chance America/Japan/Korea/Australia get drawn in, and the war would likely be longer and bloodier.
  • Western sanctions on Russia, and continued support to Ukraine. If this were a one-off war, it would make sense for the United States and others to try to pressure Ukraine into a deal, or at least not subsidize their fight. But Ukrainian resistance is offering the West a sinister gift: a chance to signal to future aggressors liberal-democratic unity and the cost of illegal invasions. This is useful because Western unity and cost was in doubt. Putin crossed red lines in Syria and Crimea with no consequences. That impunity was arguably dangerous and ill-considered by the West. Whether you like it or not, America and Europe have a strategic interest in fueling the Ukraine conflict for reputation-building purposes (in addition to any ideological motives). Frankly, this is why I think the US ought to escalate support for Ukraine.
  • Strategic ambiguity. If Biden wants deterrence to work, it makes little sense to be equivocal about American support for Taiwan. So why stay ambiguous? One possibility is that China was never the target of strategic ambiguity. It was the Taiwanese. America does not want a fervent nationalist President to declare independence from China (crossing the red line), expecting American military support. Ambiguity meant “we will help you if you are invaded, but not if you provoke invasion by crossing Chinese red lines.” I’m not certain this is true, but it is the most sensible interpretation of the policy I have heard.
  • Senior Republicans and Democrats visit Taiwan. I don’t think this does much, other than signal America’s bipartisan resolve, reminding China of the credible threat of deterrence. Arguably it compels China to respond symbolically. Saber-rattling all around. Dangerous but maybe necessary to prevent Xi from thinking invasion would be easy.
  • Re-shoring manufacturing and strategic trade/tech bans. In the short run, American economic nationalism lowers China’s perceived cost of war, because it removes one of the incentives they had for peace and integration. In the longer run, however, China would probably develop a lot of these technological capabilities anyways. This new chips policy slows them down, possibly delaying the risk of an invasion. Maybe even delaying it to the point where there really is a regime change in China. Not a great bet, but probably better than the miserable alternatives.
  • Biden’s contradictory statements. I don’t think these supposed gaffes do much. That said, “increasing uncertainty” is seldom a good thing. I think the most charitable interpretation is that the United States is inching towards ending the policy of strategic ambiguity and these are advance warnings and test balloons and slips.
  • Ending strategic ambiguity. Some argue that the shift in Chinese power and aggression is so great that it’s time to abandon ambiguity and make American commitments clear. Reduce uncertainty and pull the status quo back in the independence direction. Presumably, they’re not worried about Taiwanese idealists declaring independence and forcing America into a war. I don’t know if that is a risk or not. If not, then I agree less ambiguity is usually good.

All of these US policy changes, however, strike me as surprisingly small in comparison to the MASSIVE changes in Chinese strength and policy. In the past two decades we have seen incredible military build-up, an effective end to the two systems policy in Hong Kong, increasingly hostile rhetoric on Taiwan, and Xi cementing personal control of the Chinese Communist Party. American policy seems amazingly (maybe even naively) slow and restrained in comparison.

Why We Fight - Book Cover

Subscribe to the Blog