U.S. President Joe Biden speaking virtually with Chinese President Xi Jinping, Washington, D.C., November 2021
U.S. President Joe Biden speaking virtually with Chinese President Xi Jinping, Washington, D.C., November 2021
Jonathan Ernst / Reuters

The United States is getting serious about the threat of war with China. The U.S. Department of Defense has labeled China its primary adversary, civilian leaders have directed the military to develop credible plans to defend Taiwan, and President Joe Biden has strongly implied that the United States would not allow that island democracy to be conquered.

Yet Washington may be preparing for the wrong kind of war. Defense planners appear to believe that they can win a short conflict in the Taiwan Strait merely by blunting a Chinese invasion. Chinese leaders, for their part, seem to envision rapid, paralyzing strikes that break Taiwanese resistance and present the United States with a fait accompli. Both sides would prefer a splendid little war in the western Pacific, but that is not the sort of war they would get.

A war over Taiwan is likely to be long rather than short, regional rather than local, and much easier to start than to end. It would expand and escalate, as both countries look for paths to victory in a conflict neither side can afford to lose. It would also present severe dilemmas for peacemaking and high risks of going nuclear. If Washington doesn’t start preparing to wage, and then end, a protracted conflict now, it could face catastrophe once the shooting starts.


A U.S.-Chinese war over Taiwan would begin with a bang. China’s military doctrine emphasizes coordinated operations to “paralyze the enemy in one stroke.” In the most worrying scenario, Beijing would launch a surprise missile attack, hammering not only Taiwan’s defenses but also the naval and air forces that the United States has concentrated at a few large bases in the western Pacific. Simultaneous Chinese cyberattacks and antisatellite operations would sow chaos and hinder any effective U.S. or Taiwanese response. And the People’s Liberation Army (PLA) would race through the window of opportunity, staging amphibious and airborne assaults that would overwhelm Taiwanese resistance. By the time the United States was ready to fight, the war would effectively be over.

The Pentagon’s planning increasingly revolves around preventing this scenario, by hardening and dispersing the U.S. military presence in Asia, encouraging Taiwan to field asymmetric capabilities that can inflict a severe toll on Chinese attackers, and developing the ability to blunt the PLA’s offensive capabilities and sink an invasion fleet. This planning is predicated on the critical assumption that the early weeks, if not days, of fighting would determine whether a free Taiwan survives.

Yet whatever happens at the outset, a conflict almost certainly wouldn’t end quickly. Most great-power wars since the Industrial Revolution have lasted longer than expected, because modern states have the resources to fight on even when they suffer heavy losses. Moreover, in hegemonic wars—clashes for dominance between the world’s strongest states—the stakes are high, and the price of defeat may seem prohibitive. During the nineteenth and twentieth centuries, wars between leading powers—the Napoleonic Wars, the Crimean War, the world wars—were protracted slugfests. A U.S.-Chinese war would likely follow this pattern.

If Washington doesn’t prepare for conflict now, it could face catastrophe once the shooting starts.

If the United States managed to beat back a Chinese assault against Taiwan, Beijing wouldn’t simply give up. Starting a war over Taiwan would be an existential gamble: admitting defeat would jeopardize the regime’s legitimacy and President Xi Jinping’s hold on power. It would also leave China more vulnerable to its enemies and destroy its dreams of regional primacy. Continuing a hard fight against the United States would be a nasty prospect, but quitting while China was behind would seem even worse.

Washington would also be inclined to fight on if the war were not going well. Like Beijing, it would view a war over Taiwan as a fight for regional dominance. The fact that such a war would probably begin with a Pearl Harbor–style missile attack on U.S. bases would make it even harder for an outraged American populace and its leaders to accept defeat. Even if the United States failed to prevent Chinese forces from seizing Taiwan, it couldn’t easily bow out of the war. Quitting without first severely damaging Chinese air and naval power in Asia would badly weaken Washington’s reputation, as well as its ability to defend remaining allies in the region.

Both sides would have the capacity to keep fighting, moreover. The United States could summon ships, planes, and submarines from other theaters and use its command of the Pacific beyond the first island chain—which runs from Japan in the north through Taiwan and the Philippines to the south—to conduct sustained attacks on Chinese forces. For its part, China could dispatch its surviving air, naval, and missile forces for a second and third assault on Taiwan and press its maritime militia of coast guard and fishing vessels into service. Both the United States and China would emerge from these initial clashes bloodied but not exhausted, increasing the likelihood of a long, ugly war.


When great-power wars drag on, they get bigger, messier, and more intractable. Any conflict between the United States and China is likely to force both countries to mobilize their economies for war. After the initial salvos, both sides would hurry to replace munitions, ships, submarines, and aircraft lost in the early days of fighting. This race would strain both countries’ industrial bases, require the reorientation of their economies, and invite nationalist appeals—or government compulsion—to mobilize the populace to support a long fight.

Long wars also escalate as the combatants look for new sources of leverage. Belligerents open new fronts and rope additional allies into the fight. They expand their range of targets and worry less about civilian casualties. Sometimes they explicitly target civilians, whether by bombing cities or torpedoing civilian ships. And they use naval blockades, sanctions, and embargoes to starve the enemy into submission. As China and the United States unloaded on each other with nearly every tool at their disposal, a local war could turn into a whole-of-society brawl that spans multiple regions.

Bigger wars demand more grandiose aims. The greater the sacrifices required to win, the better the ultimate peace deal must be to justify those sacrifices. What began as a U.S. campaign to defend Taiwan could easily turn into an effort to render China incapable of new aggression by completely destroying its offensive military power. Conversely, as the United States inflicted more damage on China, Beijing’s war aims could grow from conquering Taiwan to pushing Washington out of the western Pacific altogether.

All of this would make forging peace more difficult. The expansion of war aims narrows the diplomatic space for a settlement and produces severe bloodshed that fuels intense hatred and mistrust. Even if U.S. and Chinese leaders grew weary of fighting, they might still struggle to find a mutually acceptable peace.


A war between China and the United States would differ from previous hegemonic wars in one fundamental respect: both sides have nuclear weapons. This would create disincentives to all-out escalation, but it could also, paradoxically, compound the dangers inherent in a long war.

For starters, both sides might feel free to shoot off their conventional arsenals under the assumption that their nuclear arsenals would shield them from crippling retaliation. Scholars call this the “stability-instability paradox,” whereby blind faith in nuclear deterrence risks unleashing a massive conventional war. Chinese military writings often suggest that the PLA could wipe out U.S. bases and aircraft carriers in East Asia while China’s nuclear arsenal deterred U.S. attacks on the Chinese mainland. On the flip side, some American strategists have called for pounding Chinese mainland bases at the outset of a conflict in the belief that U.S. nuclear superiority would deter China from responding in kind. Far from preventing a major war, nuclear weapons could catalyze one.  

Once that war is underway, it could plausibly go nuclear in three distinct ways. Whichever side is losing might use tactical nuclear weapons—low-yield warheads that could destroy specific military targets without obliterating the other side’s homeland—to turn the tide. That was how the Pentagon planned to halt a Soviet invasion of central Europe during the Cold War, and it is what North Korea, Pakistan, and Russia have suggested they would do if they were losing a war today. If China crippled U.S. conventional forces in East Asia, the United States would have to decide whether to save Taiwan by using tactical nuclear weapons against Chinese ports, airfields, or invasion fleets. This is no fantasy: the U.S. military is already developing nuclear-tipped, submarine-launched cruise missiles that could be used for such purposes.

A local war could turn into a whole-of-society brawl that spans multiple regions.

China might also use nuclear weapons to snatch victory from the jaws of defeat. The PLA has embarked on an unprecedented expansion of its nuclear arsenal, and PLA officers have written that China could use nuclear weapons if a conventional war threatened the survival of its government or nuclear arsenal—which would almost surely be the case if Beijing was losing a war over Taiwan. Perhaps these unofficial claims are bluffs. Yet it is not difficult to imagine that if China faced the prospect of humiliating defeat, it might fire off a nuclear weapon (perhaps at or near the huge U.S. military base on Guam) to regain a tactical advantage or shock Washington into a cease-fire.

As the conflict drags on, either side could also use the ultimate weapon to end a grinding war of attrition. During the Korean War, American leaders repeatedly contemplated dropping nuclear bombs on China to force it to accept a cease-fire. Today, both countries would have the option of using limited nuclear strikes to compel a stubborn opponent to concede. The incentives to do so could be strong, given that whichever side pulls the nuclear trigger first might gain a major advantage.

A final route to nuclear war is inadvertent escalation. Each side, knowing that escalation is a risk, may try to limit the other’s nuclear options. The United States could, for instance, try to sink China’s ballistic missile submarines before they hide in the deep waters beyond the first island chain. Yet such an attack could put China in a “use it or lose it” situation with regard to its nuclear forces, especially if the United States also struck China’s land-based missiles and communication systems, which intermingle conventional and nuclear forces. In this scenario, China’s leaders might use their nuclear weapons rather than risk losing that option altogether.


There is no easy way to prepare for a long war whose course and dynamics are inherently unpredictable. Yet the United States and its allies can do four things to get ready for whatever comes—and, hopefully, prevent the worst from happening. First, Washington can win the race to reload. China will be much less likely to go to war if it knows it will be outgunned as the conflict drags on. Washington and Taipei should therefore aggressively stockpile ammunition and supplies. For the United States, the critical assets are missiles capable of sinking China’s most valuable ships and aircraft from afar. For Taiwan, the key weapons are short-range missiles, mortars, mines, and rocket launchers that can decimate invasion fleets. Both nations also need to be ready to churn out new weapons in wartime. Taiwanese factories will be obvious targets for Chinese missiles, so the United States should enlist the industrial might of other allies. Japan’s shipbuilding capacity, for example, could be retooled to produce simple missile barges rapidly and on a massive scale.

Second, the United States and Taiwan can demonstrate their ability to hang tough. In a long war, China could try to strangle Taiwan with a blockade, bombard it into submission, or take down U.S. and Taiwanese electrical grids and telecommunications networks with cyberattacks. It could use conventionally armed, hypersonic missiles to attack targets in the U.S. homeland and flood the United States with disinformation. Countering such measures will require defensive preparations, such as securing critical networks; expanding Taiwan’s system of civilian shelters; and enlarging the island’s stockpiles of fuel, food, and medical supplies.

China might use nuclear weapons to snatch victory from the jaws of defeat.

Breaking a Chinese campaign of coercion also requires threatening Beijing with painful retaliation. A third objective, therefore, is to own the escalation ladder. By preparing to blockade Chinese commerce and cut Beijing off from markets and technology in wartime, the United States and its allies can threaten to turn an extended conflict into an economic catastrophe for China. By preparing to sink Chinese naval vessels anywhere in the western Pacific and destroy Chinese military infrastructure in other regions, Washington can threaten a generation’s worth of Chinese military modernization. And by developing the means to hit Chinese ports, airfields, and armadas with tactical nuclear weapons, the United States can deter China from initiating limited nuclear attacks. Washington should confront Beijing with a basic proposition: the longer a war lasts, the more devastation China will suffer.

Because controlling escalation will be essential, the United States also needs options that allow it to dial up the punishment without necessarily dialing up the violence. By subtly demonstrating that it has the cyber-capabilities to cripple China’s critical infrastructure and domestic security system, for example, the United States can threaten to bring the war home to Beijing. Similarly, by improving its ability to suppress Chinese air defenses near Taiwan with cyberattacks, electronic warfare, and directed-energy weapons, the United States can increase its freedom of action while limiting the amount of physical destruction it wreaks on the mainland.

Any escalatory moves risk ratcheting up the intensity of a conflict. So the final preparation Washington must make is to define victory down. A war between nuclear-armed great powers would not end with regime change or one side occupying the other’s capital. It would end with a negotiated compromise. The simplest settlement would be a return to the status quo: China stops attacking Taiwan in exchange for a pledge that the island will not seek formal independence and that the United States will not endorse it. To sweeten the deal, Washington could offer to keep its forces off Taiwan and out of the Taiwan Strait. Xi would be able to tell the Chinese people that he taught his enemies a lesson. The United States would have saved a strategically positioned democracy. That may not be a satisfying end to a hard-fought conflict. But in a long war between great powers, protecting vital U.S. interests while avoiding Armageddon is good enough.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now