Denial Is the Worst Except for All the Others: Getting the U.S. Theory of Victory Right for a War with China


While Washington continues to debate its Ukraine policy, everyone can be relieved that no side has employed any nuclear weapons yet. When the United States and its partners intervened after Russia’s full-scale invasion, there were serious and well-reasoned concerns about the extent to which the conflict could escalate. These worst-case scenarios never unfolded partly because the United States and its partners calibrated their intervention, rejecting proposals like a no-fly zone that could have brought the U.S. and coalition militaries into direct contact with Russian forces. This proxy warfare strategy helped the United States manage escalation in a similar fashion to various proxy conflicts throughout the Cold War.

In contrast to this indirect defense of Ukraine, President Joe Biden has repeatedly threatened to defend Taiwan directly with U.S. forces. A direct conflict with a nuclear-armed great power like China would push the United States into untested waters that it managed to avoid during the Cold War and create escalation risks comparable to worst-case fears about the war in Ukraine. The protracted conflict in Ukraine should serve as a stark reminder that wars are easier to start than to end and that fighting a nuclear-armed great power requires a fundamentally different mindset than what the United States and its allies became accustomed to over the past three decades.

The United States should enter any conflict with a nuclear-armed great power like China with a theory of victory that outlines how the war will end and how it will manage escalation. Theories of victory are causal stories about how to defeat an adversary. They are the principal tenets of a strategy rather than strategies themselves, and U.S. presidents have historically created them with their most senior military advisors. Developing a theory of victory requires identifying the conditions under which an enemy will stop fighting and then outlining how to shape the conflict in a way that creates those conditions. China’s burgeoning nuclear arsenal, long-range conventional strike capabilities, and cyber exploits against U.S. critical infrastructure are strengthening its ability to escalate in diverse ways, including striking the U.S. homeland. To avoid a Pyrrhic victory, theories of victory against nuclear powers must consider how to keep the war limited.

This article outlines several potential theories of victory for a U.S.-Chinese war over Taiwan, focusing on denial and military cost imposition because they are the most viable and influential. We argue that a denial theory of victory is the best way to strike the balance between the desire to maximize the chances of U.S. success and the imperative to manage escalation. The U.S.-led coalition should avoid theories of victory that rely on military cost imposition, especially because of the difficulties of finding a “sweet spot” of targets that are valuable enough to influence Beijing’s decision-making but not so valuable that attacking them causes unacceptable escalation. This is a dilemma we call the “Goldilocks challenge.”



Potential Theories of Victory

Drawing from research on strategy and coercion, we identify five potential military theories of victory that are universal to all countries and all conflicts. The first candidate, dominance, relies on brute force to eliminate the enemy’s physical ability to continue fighting. Like the U.S. defeat of Germany in 1945 and Iraq in 2003, dominance aims to comprehensively defeat the Chinese military and potentially impose far-reaching surrender terms, such as regime change or forcing Beijing to recognize Taiwan’s independence. Despite the emotional and domestic political appeal, dominance is simply not viable against nuclear-armed great powers. A defining feature of the nuclear age is that the nuclear-armed “loser” of a conventional war can still reach out and annihilate the “winner” even after the defeat of its conventional forces. Destroying China’s military and industrial capacity to the point that it could not keep fighting would plausibly cause nuclear escalation by threatening the Chinese Communist Party’s vital interests, if not survival.

The U.S. coalition should therefore rely instead on coercive theories of victory that persuade China to stop fighting even while it retains the ability to continue the conflict. Because the decision to stop fighting is ultimately a cost-benefit calculation, these theories focus on manipulating different aspects of the costs and benefits. Additionally, for coercion to work, the United States and its allies should define their war aims narrowly, such as preserving Taiwan’s de facto independence even without a formal settlement, so that Beijing is less likely to view the costs of accepting defeat as intolerable. Keeping these war aims limited will be a key challenge. There will likely be strong domestic pressure to adopt expansive goals, like “punishing” China for starting the conflict, as seen in the Russian-Ukrainian context, and preventing Beijing from attempting this sort of aggression ever again. The United States and its allies must resist these pressures because they would make war termination more difficult and escalation more likely by putting China in an increasingly desperate position.

The two most viable coercive theories of victory are denial and military cost imposition. Denial focuses on reducing China’s benefits from continuing the war. The storyline for denial is that destroying the power projection capabilities that China is using to seize Taiwan can persuade Beijing that it is unlikely to accomplish its objectives and that it is better off ending the war because further fighting will not change the eventual outcome. This would likely involve interdicting the air- and sealift assets that China needs to transport and sustain forces on Taiwan. While China might pay a high price for the privilege of seizing Taiwan while it seemed feasible, denial aims to provide new information to the Chinese leadership that the benefits needed to justify its war-time losses will not materialize.

Military cost imposition focuses on increasing China’s costs from continuing the war. The storyline is that military measures like a maritime blockade of China’s seaborne trade or strategic air attacks against other pressure points, like war-supporting industry and political leadership, can convince Beijing that the war is too costly to continue. The critical source of leverage is not China’s beliefs about whether it can achieve its war aims but whether the U.S. coalition can make the process so costly and painful that Beijing concludes it is no longer worth trying. Denial and military cost imposition are not mutually exclusive, so the United States can combine them, but doing so creates additional escalatory risks.

The Case Against Military Cost Imposition

Military cost imposition is a viable theory of victory under three conditions. Satisfying all three in a war against China would be very difficult.

The biggest obstacle is navigating the Goldilocks challenge of finding a coercive sweet spot. The central dilemma is finding targets that are sufficiently valuable to persuade Beijing to abandon its military campaign for Taiwan but not so valuable that Beijing risks significant escalation to retaliate and to compel the U.S. coalition to stop attacking those pressure points.  The most influential coercive levers to pull are therefore the ones that generate the highest escalation risks. Different U.S. administrations may have a range of views on what constitutes “unacceptable” escalation, but examples include nuclear use and widespread conventional attacks on the U.S. homeland. The primary risk is not an immediate Chinese nuclear use against the U.S. homeland. Rather, it is that limited retaliatory measures, like Chinese conventional strikes on the U.S. homeland or tactical nuclear use, could set in motion an “escalation spiral” where both sides engage in tit-for-tat retaliation that becomes increasingly severe and difficult to control.

Finding targets that provide sufficient coercive leverage is challenging because China values Taiwan so highly. For example, China is unlikely to abandon Taiwan to save its overseas military bases in Africa. While China might currently prioritize economic growth over controlling Taiwan, a scenario in which Beijing has rolled the iron dice might reflect a change in the regime’s priorities and a greater willingness to run risks. Proponents of a blockade argue that it would provide enough leverage because it targets the “Chinese economy and hence the legitimacy of the Chinese Communist Party,” which could jeopardize the regime’s domestic control. But if strangling China’s economy threatens the regime’s survival, U.S. and coalition leaders cannot be confident Beijing would not resort to significant escalation to compel the United States to stop attacking these targets.

Identifying sensitive targets that would not provoke unacceptable escalation creates its own challenges. Beijing has good reasons to exaggerate its red lines, and escalation is unpredictable. While some targets, such as China’s leadership and nuclear forces, clearly cross the line, others are ambiguous. In fact, Chinese decision-makers themselves may not know how they will react in advance given the emotions and imperfect information of war. Beijing’s red lines might shift unpredictably over time, and escalation spirals could also make it hard to anticipate the ultimate endpoint of retaliation. U.S. and Chinese leaders historically struggled to understand their adversary’s intentions and coercive thresholds during crises such as the 1962 Cuban Missile Crisis and the 1969 Sino-Soviet border conflict.

The United States has faced this Goldilocks challenge before. In 1980, U.S. officials debated how best to prevent the Soviet Union from seizing Iran and its oil reserves. Because the joint staff doubted that it could militarily defeat an invasion, it proposed a strategy to strike Soviet targets unrelated to the invasion in order to impose coercive pain (cost imposition) and to raise the prospect that further fighting would lead to escalation (brinkmanship). After the undersecretary of defense for policy asked the joint staff to specify exactly “what escalation and where” would accomplish U.S. objectives, the joint staff failed to identify a sweet spot that could change Soviet calculations but avoid escalation: “The only category of ripostes which has the possibility of raising Soviet costs to a level commensurate with the gains of occupying Iran involves major escalation of the conflict,” risking “a worldwide NATO-Warsaw Pact war with the attendant risks of nuclear escalation.”

Recent events also illustrate that the Goldilocks challenge is a recurring dilemma. In Ukraine, Russia’s attacks on Ukrainian population centers and energy infrastructure and Ukraine’s attacks on Russian energy infrastructure have both thus far failed to compel an end to the war. In the Middle East, the United States often faces a challenge in imposing enough costs on Iran to compel it to rein in its proxies but not so much that it provokes a wider war. Iran and Israel have faced similar challenges in calibrating military strikes and retaliatory measures, almost losing control during an escalation spiral in April 2024 that narrowly avoided further escalation.

The second major obstacle for military cost imposition is generating the necessary leverage quickly enough. If the coercive sweet spot does exist, the U.S. military needs to attack the targets fast enough and at a large enough scale to build pressure within an operationally relevant timeframe. If policymakers want to use military cost imposition as an independent theory of victory, the relevant timeframe is how long it takes China to seize Taiwan in a fait accompli. This might take only a few months or even weeks without a direct U.S. denial defense against Chinese invasion forces. Once China holds Taiwan, rolling back its control is extremely risky because endowment effects could make Beijing willing to escalate dramatically (including up to nuclear use) before giving up such a hard-earned prize. The problem with military cost imposition is that blockades are slow-moving and graduated measures that need many months or years to reach their greatest coercive impact. On the other hand, measures that can quickly generate large costs, like widespread air attacks on China’s critical energy and transportation infrastructure, carry much larger up-front escalation risks. History suggests that even strategic air attacks build pressure much more slowly than proponents envision. For example, despite initial optimism of rapid success, the U.S. coercive bombing campaign against Kosovo in 1999 took 78 days to work against a vastly weaker regional power, which capitulated only after its own counter-coercion campaign failed.

The final challenge is providing Beijing credible assurances that the pain will stop if it complies with U.S. and coalition coercive demands. If Beijing does not believe U.S. assurances that compliance will bring relief, it has no incentive to stop fighting. It may even fear that concessions will embolden further coercion and invite harsher demands. This requirement is concerning because China is highly suspicious of U.S. motives, and military measures designed to maximize Beijing’s pain could convince its leaders they were fighting a total war for the regime’s survival. Wartime fog and friction often cause leaders to judge their adversary’s intentions through a worst-case lens, such as interpreting attacks on domestic targets like critical infrastructure or domestic security services as attempts to destabilize Chinese society and thereby facilitate regime change. This could make the costs of accepting defeat appear intolerable or even existential.

The Case for Denial

A denial theory of victory’s core requirement is to convince Beijing that it lacks the ability to seize Taiwan militarily at this time and that further fighting will not change the war’s eventual outcome. While great powers like China will always have the ability to protract or escalate a conflict, denial aims to persuade Beijing that these are bad options that will not solve its core problems. By avoiding broader cost imposition at the start of the war, a denial theory of victory gives Beijing space to decide to stop the war after it realizes that its military operation has failed.  A historical example would be Argentina’s decision to end the Falklands War after it realized that it could not hold the islands. Notably, the United Kingdom did not require that Argentina renounce its claims to the islands.

A denial theory of victory requires having a coalition able to defeat a Chinese amphibious invasion of Taiwan, but it does not rest on the assumption that China will immediately stop fighting after the invasion fails. It is certainly possible that China would shift to a blockade or strategic bombing of Taiwan to see if it could still achieve its original political objective of unification. If the United States and its coalition partners succeed in putting China in this position, they have already significantly narrowed China’s plausible pathways to military victory: Taiwan would be the first nation in history to surrender its sovereignty to a blockade or strategic air attack. Faced with mounting losses and without a clear pathway to achieve its original objectives, the ideal scenario is Beijing then claims “victory” by “teaching the enemy a lesson,” as it did with Vietnam in 1979, and stops fighting.

If the war becomes protracted, skeptics may doubt that a denial theory of victory could place enough pressure on Beijing to compel an end to the war, increasing the temptation to turn to military cost imposition. But the United States should at least wait to see if it can first end the war through denial before it deliberately widens the conflict by striking Chinese pressure points unrelated to the defense of Taiwan. Additionally, Beijing would still face strong pressures for war termination even if the United States did not use military strikes to maximize its pain. War is inherently costly, and those costs may look increasingly untenable and risky to Beijing if it lacks a clear pathway to seizing Taiwan. By drawing out the conflict, China could face tens of thousands of casualties and extensive losses of platforms it took decades to build and that it may want to preserve for a future attempt to take Taiwan. The U.S. coalition could also use economic sanctions and diplomatic isolation as non-military sources of pressure. If all these measures still fail to end the war, the United States has the option to use military cost imposition as a last resort, though this would still require navigating the Goldilocks challenge.

There exist scenarios where Beijing’s initial invasion attempt fails and the war turns into a wide and grinding struggle between the two camps. In such scenarios, both sides may prefer to have begun their cost imposition efforts as early as possible to give them maximal time to accumulate. A denial theory of victory accepts the risk of delaying the start of this broader type of conflict (and the escalation danger it creates) in exchange for the opportunity to discover whether Beijing wants to stop fighting before the war turns into a hegemonic cage match. The alternative is to risk turning a limited war into an existential fight from the outset.

Other skeptics may agree that denial would be ideal but argue that China’s growing military power has made it operationally infeasible. The sense that denial is too hard has contributed to growing interest in military cost imposition as an alternative to the hard work of preparing the joint force for denial. This counter-argument is too pessimistic. Denial remains feasible even if it has become more difficult. China is at a structural disadvantage in that amphibious assaults are immensely challenging operations with which China has no real-world experience. While the U.S. military is increasingly at a quantitative disadvantage, it retains a qualitative edge in key areas such as undersea warfare and penetrating long-range strike. The U.S. government assesses that an invasion “would likely strain” the Chinese military and remains “a significant political and military risk.” Unclassified wargames support this assessment.


Any war between the United States and China would involve significant costs and risks, including potential nuclear use. These costs and risks are why U.S. policy aims to deter such a war from happening in the first place. But civilian analysts and military officers still need to prepare for the possibility that deterrence might fail. If the United States chooses to intervene in defense of Taiwan, a denial theory of victory focused on narrowly scoped war aims can help reduce but not eliminate escalation risks. While there will be strong temptations to reach for all available tools, including the kinds of cost-imposing measures the U.S. military has used in past conflicts, fighting a nuclear-armed great power requires a fundamentally different mindset. The United States may also have to persuade its allies to adopt similar restraint in their war aims and operations, especially given Taiwan and Japan’s growing long-range strike capabilities that could enable their own punitive strikes against the Chinese mainland. To channel Winston Churchill’s famous quotation about democracy, denial is the worst theory of victory except for all the others. A denial theory of victory does not guarantee success, but it offers the best chance to strike an effective balance between the desire to win the war and the imperative to manage escalation.



Jacob Heim is a senior researcher at the RAND Corporation. He served in the Office of the Secretary of Defense and holds degrees in international relations and mathematics.

Zachary Burdette is a Ph.D. candidate in political science at the Massachusetts Institute of Technology, a pre-doctoral research fellow at Harvard’s Belfer Center for Science and International Affairs, and an adjunct researcher at RAND.

Nathan Beauchamp-Mustafaga is a policy researcher at the RAND Corporation, where he focuses on Asian security issues. He holds degrees in international affairs and Chinese language.

This article is based on a recent RAND report that assesses alternative theories of victory for a hypothetical war between the United States and China over Taiwan.

Image: IceUnshattered via Wikimedia Commons