How War in Taiwan Ends
If Deterrence Fails, Could America Thwart China?
In recent years, many in Washington have focused on deterring China from invading Taiwan. Before taking office earlier this year, Elbridge Colby, the U.S. undersecretary of defense for policy, asserted that Taiwan should be “laser focusing on implementing a denial defense against invasion.” Indeed, an array of small, inexpensive weapon systems holds great promise for repelling a Chinese amphibious landing. The Trump administration’s new National Defense Strategy is therefore correct to embrace a strategy of denial for stopping an invasion of Taiwan.
But rebuffing an invasion might not end the war. Joel Wuthnow, an expert on the Chinese military, has warned, “There is no scenario in which China, following an unsuccessful invasion, accepts responsibility, acknowledges that military solutions are impractical, or pivots to a fundamentally different set of political objectives toward Taiwan.” In the wake of a failed invasion, Chinese leader Xi Jinping (or his successor) would be unlikely to simply pack up and go home. Instead, Chinese leaders might reason that they have less to lose by continuing the fight.
This is why the political scientist Michael Beckley has argued that “war over Taiwan likely would become protracted, as nearly all great power wars have since the Industrial Revolution.” World War II ended only when Allied forces captured Germany’s capital and the United States dropped nuclear weapons on Japan. Neither option seems advisable in the context of a U.S.-Chinese war; Washington needs to find other ways to end it. And so, in the years to come, the United States must prepare two forces: one to stop a Chinese invasion and another to end the conflict. Preventing a war from starting in the first place will rely to some extent on the innovative forms of deterrence by denial on which the Trump administration and others have focused. But denial capabilities on their own will not be enough. Ending a war that churns on even after a failed invasion will also require old-fashioned power projection.
IN DENIAL
In the twentieth century, the United States perfected the art of projecting power around the globe. A combination of forward bases and aircraft carriers allowed U.S. forces to operate worldwide. With the fall of the Soviet Union in 1991, the U.S. military’s dominance also meant that one set of forces could employ two distinct forms of deterrence simultaneously: denial and punishment.
Consider the role of U.S. aircraft carrier strike groups during the 1996 Taiwan Strait crisis. At the time, China was staging military exercises and testing missiles in the waters around Taiwan. As tensions rose, Washington maneuvered two carriers near the island. Those strike groups practiced deterrence by denial by threatening to physically repel an attack. But they also performed deterrence through punishment by threatening severe consequences if Beijing went through with it, since carrier-based aircraft could strike ships heading toward China and even targets on the Chinese mainland.
In the last few years, however, the United States has begun tailoring its forces—and those of its allies and partners—for more specific missions. Forward bases and aircraft carriers are expensive to build and maintain, yet still vulnerable to ballistic missiles and other asymmetric systems. Pentagon officials are therefore pushing to acquire more “attritable” systems, which are relatively cheap to produce and designed to be expendable, for use by small units operating within the expanding area that China threatens. As David Berger, the former commandant of the U.S. Marine Corps, explained at a defense industry conference in 2021, the United States must “get comfortable with throwaway things.”
But attritable systems are of limited use against China’s day-to-day coercive operations in the air and sea around Taiwan. Last year, Taiwan detected 5,105 Chinese sorties into its airspace. Defending against these aircraft requires expensive jets rather than low-flying drones. In the maritime context, responding to Chinese naval incursions in the waters around Taiwan will require vessels that can monitor those activities and challenge Chinese forces if necessary.
Even after open conflict begins, denial is still only a partial answer. U.S. mines and missiles can sink Chinese vessels, killing thousands of troops in the process, but Chinese leaders might still seek at least a partial victory. The People’s Liberation Army could attempt to seize Taiwan’s outlying islands or conduct a maritime blockade while its military arsenal makes the waters around Taiwan a no man’s land. “There is no path to U.S. victory that does not include the long blockade,” the former intelligence officer Lonnie Henley has argued.
That is why the United States must be able to convince China that it will face unacceptable costs if it continues fighting in the wake of an unsuccessful invasion. A strategy of denial is only step one; the threat of punishment will be the United States’ ultimate trump card.
CAN’T STOP, WON’T STOP
The war in Ukraine illustrates the difficulty of terminating a conflict even after an initial invasion has bogged down. With small and cheap systems such as drones and mines, Ukraine was able to deny Russia a swift victory but has failed to impose costs high enough to convince Russian President Vladimir Putin to stop fighting. Russia has suffered terrible losses in the war, but Putin appears to have calculated that the costs of continuing are lower than the costs of admitting defeat.
Russia’s example serves as a warning about China’s likely behavior. Ideally, the prospect of a failed invasion of Taiwan would deter China, but Chinese leaders might perceive several incentives for protracting a war following an initial loss. First, China’s industrial capacity far outstrips that of the United States, so it could recapitalize its forces more rapidly. Over the last three decades, China has undergone a massive military buildup. The Office of Naval Intelligence has assessed that China has over 230 times the shipbuilding capacity of the United States. Meanwhile, U.S. armed forces face significant munitions shortages, with some experts predicting that American stockpiles would be expended after just weeks, or even days, of a conflict with China.
Beijing might believe it can outlast Washington and Taipei in terms of other supplies, as well. Getting provisions across land into Ukraine has proved challenging; delivering even basic necessities over water to Taiwan amid a conflict with China would be an order of magnitude more difficult. Taiwan is a relatively small island with limited food and energy stockpiles. Conversely, Beijing’s rapid expansion of wind, solar, and nuclear power would help insulate it against a U.S. energy blockade.
A conflict over Taiwan could eventually become a contest of wills—which Beijing believes it could win. Chinese officials have described Taiwan as “the very core of China’s core interests.” U.S. President Donald Trump’s take is decidedly different: “Taiwan is 9,500 miles away,” he said in an interview with Bloomberg Businessweek. “It’s 68 miles away from China. I just think we have to be smart . . . it’s a very, very difficult thing.” The American people support Taiwan, but many do not want a direct conflict with China: when asked in 2024 by the Chicago Council on Global Affairs about their commitment to defending Taiwan from a Chinese invasion, the majority of Americans surveyed either opposed such a policy or were unsure.
A Chinese failure in a conflict over Taiwan could also threaten Xi or his successor’s hold on power and undermine the legitimacy of the Chinese Communist Party. Xi would want to avoid an admission of failure and thus might order the Chinese military to keep waging an unsuccessful war. Without the prospect of punishment, the CCP might decide that failure poses a greater risk than continuing the conflict.
For all these reasons, Chinese leaders might prefer to keep fighting even after an initial defeat. To bring the conflict to a close, the United States would need to credibly threaten punishment. Colby and other Trump administration officials clearly recognize this; he co-authored a 2022 report advocating “selective punishment operations” and “cost-imposition to favorably manage escalation and seek to terminate a war with China.” These operations could include an embargo or the seizure of Chinese assets held abroad. But Beijing has been insulating itself against political and economic pressure, so military escalation might well be required, including strikes on critical infrastructure and parts of China’s defense-industrial base. These moves would raise the costs for China of continuing a conflict, but they also present a strategic dilemma.
THE GOLDILOCKS PARADOX
A number of factors would complicate any threat of punishment. First is what researchers at the RAND Corporation have termed the “Goldilocks challenge”: threats of punishment must be high enough to persuade Beijing to end a conflict in which it is deeply invested but low enough to avoid provoking unacceptable escalation, such as nuclear use. Finding this middle ground would not be an easy task.
It will therefore be important to try to keep an initial fight over Taiwan limited in order to provide Chinese leaders a pathway for deescalation. Chinese leaders might back down after claiming to have taught Taiwan a lesson or taken some contested territory. Yet China’s own messaging before a conflict could set a higher bar: Chinese leaders might demonize Taiwan and the United States to rally public support, while trumpeting the PLA’s military superiority and China’s great rejuvenation. An invasion of Taiwan might start with Beijing metaphorically burning its ships so there would be no turning back.
In the aftermath of a conflict in the Taiwan Strait, politicians in Washington, Taipei, and elsewhere might themselves raise the stakes. They could seek to use Beijing’s moment of weakness to constrain China’s ambitions, formalize Taiwan’s independence, or undermine the CCP’s hold on power. There would be a fine line between “too hot” and “too cold” policies, and the tradeoffs would become more difficult as the war grew longer, bloodier, and more destructive.
A second challenge is that U.S. “horizontal escalation”—widening the scope of the conflict—may not be as effective today as it once might have been. Devoting more funding to denial capabilities risks cannibalizing resources for military platforms more capable of threatening punishment, such as stealthy bombers and submarines armed with cruise missiles. And although American strategists have discussed the possibility of a blockade to prevent China from importing energy supplies, the country’s nuclear power plants and renewable energy sources now account for a third of its energy production, so that Beijing is less vulnerable to an energy blockade. Cutting China’s fossil fuel imports would hurt over time, but Taiwan would be in a far more dire position.
Therefore, the biggest challenge for the Trump administration’s National Defense Strategy is not whether it allocates resources to a strategy of denial but how it integrates denial and punishment into a holistic deterrence framework. Rebuffing an initial attack on Taiwan is necessary but not sufficient. Without a plan for terminating a war, Washington would risk repeating the pattern of U.S. strategy in Iraq and Afghanistan that many Trump officials critiqued: winning the first battle but losing the war. If the United States is to deter China, it will have to persuade Chinese leaders that Washington has a strategy not only for the early stages of a conflict but also for the end stage of a war.