There is a growing bipartisan consensus in Washington on a tighter embrace of Taiwan, which may soon become a stronger implied US commitment to go to war in the event of a Chinese invasion.

Taiwan matters to US security and the regional order, and the United States should continue to make clear that aggression is unacceptable. But those advocating a stronger US security commitment exaggerate the strategic consequences of a successful Chinese invasion. The stakes are not so high as to warrant an unqualified US pledge to go to war. American decision-makers, like their forebears confronting the seeming threat of communism in Indochina, may be trapping themselves into an unnecessarily stark conception of the consequences of a successful Chinese invasion of Taiwan.

It would be irresponsible for the United States to leave itself no option in the event of Chinese aggression other than war. But nor should Washington abandon Taiwan. There is a prudent middle way: the United States should act as armourer, but not guarantor. It should help prepare Taiwan to defend itself, to raise costs against aggression, and develop means of punishing China with non-military tools.

Faced with the possibility of another Taiwan Strait crisis, more and more observers in Washington and elsewhere are making the case for an unambiguous US commitment to defend Taiwan in the event of a Chinese attack. This essay contends that the United States has options between total commitment and abandonment. There is a prudent middle way in which the United States, while reserving the right to intervene if it so chooses, focuses on helping Taiwan to defend itself while building a menu of options for deterring and punishing Beijing’s aggression without fighting.This essay first argues that the case for Taiwan’s strategic significance is often overdrawn. Any Chinese attack would be a tragedy and a crime, and the United States should make clear that such a step is unacceptable and would destroy the Chinese Communist Party’s ambitious development plans. But it need not destroy the US position in Asia or produce a wave of successful Chinese adventurism. The essay then points to the intense dangers of a war for the United States, including outright loss, a crippled military, large-scale attacks on the homeland, and nuclear escalation. We make the case for an alternative to the binary choice of all-out war or desertion: Taking some additional risk to help Taiwan prepare for its own defence, combined with the development of multiple options short of outright war for punishing China in the event of an attack.


The intensifying US–China rivalry has brought new danger to a long-simmering flashpoint: Taiwan. China’s rising power and regional influence, ambitions to gain control of the island, and threats to resolve the issue in military terms[1] may lead to a growing risk of conflict.[2] While Beijing appears wary of the costs of a potential invasion and may see little need to act in the near future,[3] the intensity of its intimidation has certainly grown.[4] Meantime demographic and political trends in Taiwan suggest growing support for independence,[5] and policies that edge in this direction — economic diversification, more open alignment with the United States, and an increased defence budget — may exacerbate the sense of urgency felt by Xi and the Chinese Communist Party (CCP).[6]

The risk of war may not be imminent. But in the face of Beijing’s increased verbal threats and intimidatory activities against Taiwan, a growing chorus urges Washington to abandon the ambiguity of its security promise to Taiwan in favour of an unqualified commitment.[7] Short of such a step, a more general trend towards a tighter embrace of Taiwan is manifest in Washington’s 2018 “Indo-Pacific Strategy”,[8] high-level official visits to Taiwan, new arms sales, the lifting of restrictions on official engagement, and promises of bolstered economic and diplomatic ties. The Biden administration has cautiously endorsed stronger ties.[9] The growing bipartisan consensus on a tougher stance extends to legislative proposals as well.[10] All this amounts to what may soon become a stronger implied US commitment to go to war in the event of an invasion — a trajectory that appears to be triggering rising fears in Beijing that the United States is rethinking its One China policy.[11]

Three pillars that have upheld cross-Strait stability — China’s geostrategic patience, Taiwan’s agnosticism on independence, and the ambiguity of the US One China policy — are shaking. This is happening amidst a shifting global balance of power, always perilous episodes that tend to aggravate security competitions between rising revisionist powers and incumbent “system leaders”.[12] These forces produce a danger of major war not present in almost any other international dispute.

Between these two poles, of abandonment and an unqualified

US security pledge, there is a prudent middle way.

We argue that the United States should reject the emerging belief that it confronts a binary choice over Taiwan — go all-in on a defence of Taiwan in the event of attack, or see its position in the Indo-Pacific crumble. This Manichean view represents an incoherent mixture of alarmist pessimism about what is at stake with a striking overconfidence in the US ability to deter or win a conflict that would represent an existential challenge to the CCP.

Between these two poles, of abandonment and an unqualified US security pledge, there is a prudent middle way. The United States should act as armourer, but not guarantor. It should help prepare Taiwan to defend itself, to raise costs against any aggression, and develop means of punishing China with a range of non-military tools, but not commit in advance to wade in militarily — and consciously equip itself with a menu of options for deterring and punishing Beijing’s aggression without fighting. There are intermediate options that would allow a US administration to reserve the right to directly defend Taiwan — but also to choose not to become directly involved in a war and yet still enhance deterrence before conflict, ensure that any attack was a disaster for Beijing, and make it even more difficult for China to dominate the Indo-Pacific afterwards.

We do not advocate ruling out any future policy choice.[13] We cannot forecast what a US president might decide in the event of a crisis, and we do not suggest that the United States should publicly abandon Taiwan. The chance that the United States would join any war over Taiwan will always exist. Increasingly, however, the scope of choice in the event of an attack appears to be narrowing to two extremes — either fighting, or cutting and running.[14] We argue that the United States should widen the scope of options for US presidents over Taiwan to avoid that dilemma.


The argument for a more unconditional security promise assumes that Taiwan’s security is essential to the United States and the Indo-Pacific regional order. As the argument goes, successful Chinese invasion, especially if unresisted by Washington, would wreck the US strategic position in Asia and lead the People’s Republic of China (PRC) to further military adventurism. Taiwan, in this telling, is the dam holding back a flood tide of aggression.

We disagree. The United States should not abandon Taiwan outright, and we support continued US efforts to bolster its defence. But arguments that the United States should consider the security of Taiwan a vital or even existential interest and promise unequivocally to defend Taiwan rely on implausible assertions about what will happen in the event an attack succeeds.[15] This is true about each of three such arguments: the claim that a PRC occupation of Taiwan would destroy US credibility, spark an accelerating drive by Beijing for military hegemony over the region, and provide Beijing with a decisive regional military advantage.[16]

To take these rationales in order, China hawks claim the fall of Taiwan, undefended by the United States, would ruin allies’ and partners’ faith in US guarantees and credibility. It would represent a major step towards US eviction from the region, they suggest, either in the form of countries bandwagoning under China’s dominance or scrambling to secure themselves independently of America. Elbridge Colby and Jim Mitre assert, for example, that “Other states regard it as the canary in the coal mine — a strong indicator of how far the United States would go to defend them against China. If China were able to suborn Taiwan, the US and allied defence position would be substantially compromised, and US credibility seriously diminished.”[17] Other observers agree that the United States has a massive credibility stake in the issue.[18]

China hawks claim the fall of Taiwan, undefended by the United States,

would ruin allies’ and partners’ faith in US guarantees and credibility.

There are reasons to doubt these claims. Taiwan is not a treaty ally of the United States and choosing not to fight carries no automatic lessons for what the United States will do in other contingencies. The best evidence from history is that credibility and reputation tend to be issue-specific, rather than essential qualities of a great power. Other setbacks, abandonments, and stalemates — in Korea, Vietnam, Lebanon, Kuwait in 1990, and Iraq in 2003–2011 — have caused only temporary and partial effects on US credibility. Giving ground on one issue need not spark a comprehensive loss of faith.[19]

The reputation argument also oversimplifies how allies think about crises. History suggests that allies want reliable support, but do not require (and indeed usually do not want) their patron to run serious risks of immolating itself for others in order to prove itself, especially if doing so entraps those allies in a conflict. If the United States were to fight over Taiwan, it would put regional states in the dilemma of choosing between going to war with China or deserting the United States — on which they would now have to rely, more than ever, for security guarantees.

One of the few studies of allied perceptions of historical Taiwan crises comes to precisely these conclusions. During the First Taiwan Strait Crisis of 1954–1955, US allies were less concerned about American credibility than about how a US response would affect their security. The worse the crisis got, the more nervous allies became about entrapment. The case suggests that states are not looking for signs of blanket loyalty to commitments, but, as the historian Iain D Henry has written, for “proof that the ally’s interests align with its own.” Sometimes the state not directly involved in the war roots for its ally to fight — but “in other cases, where loyalty would not be in the observer state’s interest, it will want its ally to be disloyal to the other ally”. In the 1958 crisis over the offshore islands of Quemoy and Matsu, had the US fought for the islands, some allies might have gone along, “but others might have ended their alliance with the United States.”[20]
A US Navy boatswain scans the horizon from the bridge wing aboard Arleigh Burke-class guided-missile destroyer USS McCampbell
while conducting operations in the Taiwan Strait in March 2020. Image: MCS 2nd Class Markus Castaneda/US Indo-Pacific Command/Flickr.

If US allies thought that way when the military balance was lopsided in America’s favour and China lacked nuclear weapons, how much more anxious would they be in a crisis today, when a stronger and nuclear-armed China could threaten greater losses? Most US allies and others in the region may prefer a United States that stood back from a fight over Taiwan to one that tried to drag them into it. Many, lacking direct interest in the Taiwan issue and confronting massive costs from Chinese retaliation, could refuse to participate even if the United States did fight.[21]

This issue points to a larger problem: the tendency to frame security issues in terms of America’s duty to reassure allies of its resolve and credibility. The balance of power in Asia, and the prevention of China’s dominance, is a shared responsibility. The shift to a more discriminating American world role will require recasting US relationships in the region beyond Taiwan, putting the question back on allies. If the United States is willing to help make it hard for any one power to threaten them, as well as scrutinising US credibility, what are they willing to do?[22]

Finally, even if doubts about US credibility did grow — and choosing not to fight over Taiwan would raise doubts — this need not collapse America’s position. Predation by a belligerent power is more likely to generate balancing than bandwagoning in cases where there is enough capability to counterbalance.[23] Over the last year, even in response to less violent insults, China’s regional belligerence has produced an increasing willingness to push back and grow defence spending throughout the region.[24] Asia contains well-armed, wealthy countries capable of forming a balancing coalition. In the case of an assault on Taiwan, security concerns would mushroom in Japan, Australia, India, Indonesia, and other countries. The result of an outright invasion would likely be increased defence spending and efforts to strengthen the already “thickening web” of security relationships throughout the region.[25] An America energetically coordinating and leading such responses would be perceived as acting responsibly, not forfeiting its regional commitments.


Proponents of a full US commitment to Taiwan frequently make a second, related claim: failure to defend Taiwan will turn China loose to ravage the Indo-Pacific. For Colby and Mitre, “If China is able to suborn Taiwan, it will almost certainly lift its gaze farther afield to countries like the Philippines from an even stronger position.”[26] Yet there is no evidence that China seeks to invade and conquer the Philippines — or Indonesia, Japan, Malaysia, or any country in the region. In bidding for hegemony in Asia, China’s likely approach is to overawe its neighbours and create a hierarchical power relationship, not to occupy them.[27]

China’s regional ambitions, moreover, are surely a function of its growing power and self-defined identity as the natural hegemon in Asia. Taiwan is not the key that unlocks these desires. China is already coercing Japan, the Philippines, Indonesia, Vietnam, Australia, and others without the benefit of owning Taiwan. As its military power grows, it will be able to take more direct action against others regardless of what happens with Taiwan.

If anything, an invasion of Taiwan would likely hamper and delay further aggression, not accelerate it. The PRC would likely suffer huge losses in the process and spend years pacifying Taiwan, militarily and politically; it would be depleted, not energised for further conquest. Meanwhile, others are likely to redouble their defensive measures in response to a Taiwan invasion. There is significant untapped, latent balancing power in the region and, with US assistance, an attack on Taiwan will bring more of it to the fore.

If they plan correctly, therefore, the United States and its allies and partners can ensure that the forcible seizure of Taiwan would impede, not advance, China’s goal of regional hegemony. An invasion of Taiwan is hardly a route to speedy expansion — more likely, it represents a crippling dead end for Beijing’s regional ambitions.[28] We do not mean to suggest that a Chinese invasion should be viewed callously as a potential strategic benefit to the United States. A war would be a devastating tragedy for all involved. But the strategic consequences of a war do not demand that the United States place itself at risk to prevent such an outcome.


Those arguing for a more explicit US strategic promise make a third major argument, claiming China would gain decisive military operational advantages through control of the island. “Suddenly, China’s sweeping ‘nine-dash line,’ would become even more real and more easily enforceable by Beijing”, Joseph Bosco argues, and it could relocate ballistic missiles now targeting Taiwan and the US Navy to other ships, territories, and sea-lanes.[29] For John Mearsheimer, Taiwan is “effectively a giant aircraft carrier sitting off China’s coast”, whose control would “enhance China’s ability to project military power into the western Pacific Ocean”.[30]

Possession of Taiwan would be militarily useful to Beijing. It would allow the PLA to base sensors and air defence systems on the eastern side of the island, extending the effective vision of China’s air defence network further into the Pacific. Access to Taiwan’s ports and airfields would improve the PLA’s anti-submarine coverage, and possession of the island would allow more access to the open ocean for the PLA Navy (PLAN)’s own submarines.
Multiple aircraft from Carrier Air Wing 5 fly in formation over the US Navy's forward-deployed aircraft carrier USS Ronald Reagan,
the flagship of a combat-ready force that protects the Indo-Pacific region. Image: MCS 2nd Class Kaila V Peters/US Navy/Flickr.

These gains, while real, would not provide strategically decisive military advantages in any plausible contingency in the region.[31] It is almost 1000 kilometres from Hong Kong to Luzon in the Philippines, and about 2500 kilometres from the mainland to Indonesia. Possession of Taiwan — roughly 160 kilometres from the mainland’s coast — would only slightly close those gaps. In other cases, the PLA already has mainland locations better suited to support military operations: Shandong Province and cities like Qingdao and Yantai sit right across the Yellow Sea from Korea; Hainan Island and a range of mainland military bases are closer to Vietnam and Thailand. In its Eastern Theatre Command, the PLA already has an abundance of facilities and forces able to execute military operations against any portion of Japan.[32]

In fact, the PLA is expanding a roster of power projection capabilities which give it enhanced striking power regardless of Taiwan’s status. This includes an increasingly sophisticated PLAN with over 350 ships, including more than 50 cruise-missile firing submarines; advanced, increasingly long-range anti-ship cruise missiles; and a growing amphibious and expeditionary warfare capability. Beijing is improving its H-6 bomber fleet and its air-launched cruise missile inventory and is developing a new generation of stealth bomber which the US Department of Defense suggests could have a range exceeding 8500 kilometres.[33] The PLA has a growing suite of long-range conventional cruise and ballistic missiles,[34] which provide the ability to strike targets several thousand kilometres from the mainland. In no case does the extra 320 kilometres of range that would be provided by basing on Taiwan make the decisive difference in the ability of a large set of weapons systems to hit any potential adversary.[35]
The Xian H-6K is a licence-built version of the Soviet Tupolev Tu-16 twin-engine jet bomber,
built for China's People's Liberation Army Air Force (PLAAF). Image: Wikimedia Commons.

Finally, assessments of the military outcome of a Chinese invasion of Taiwan typically consider only half the equation — the enhanced posture China would gain. [36] They overlook two factors which will shape the post-war regional military balance: the price China pays to acquire (and keep) Taiwan, and the responses of others to an act of naked aggression.

There is no way to know precisely how costly an invasion of Taiwan would be for the PLA, but a range of historical rules of thumb ought to give Beijing pause. One 2000 analysis conducted by RAND of the air campaign concluded that the invasion would be “a very bloody affair” likely to witness “extraordinarily high attrition rates”. Projected Chinese losses after just four days of combat were as high as 75 per cent of the committed aircraft.[37] In terms of maritime assets, Michael O’Hanlon argued in 2000 that the PLA would lose at least 20 per cent of its forces just to get its invasion fleet ashore. Losses in the airborne component of an invasion could reach 50 per cent.[38] Admittedly, those analyses are somewhat dated; China’s capabilities have advanced since then, but so have Taiwan’s — and they could improve more. As the US Department of Defense forecast in 2020, an attempt to invade Taiwan would “likely strain China’s armed forces” by inflicting significant combat force attrition. Even assuming a successful landing and breakout, moreover, an invasion would likely plunge China into complex urban warfare and counterinsurgency.[39]

There is no way to know precisely how costly an invasion of Taiwan would be for the PLA, but a range of historical rules of thumb ought to give Beijing pause.

A Chinese military hobbled by such losses, as well as the costs of an ongoing pacification campaign, would likely appear less intimidating, rather than more so. The PLA’s power projection capabilities would be depleted, and it would take years for Beijing to produce new systems to replace much of this equipment. The political effect inside China of such casualties is difficult to anticipate, but the physical attrition would hardly leave the CCP spring-loaded for fresh military adventures beyond Taiwan. China has not known war since 1979, and the loss of tens of thousands of young men and women from largely one-child families could create dangerous political dynamics for the CCP.

Then, too, as argued above, an invasion would likely spur many countries in the region (and even beyond) to boost their defence spending and intensify defence collaboration. China’s provocations have already led Australia, Japan, South Korea, and Vietnam to announce defence budget increases of between seven and nine per cent, and new plans to produce advanced missiles, unmanned aerial vehicles (UAVs), and other systems.[40] If an invasion of Taiwan prompted Japan, Australia, New Zealand, Vietnam, and India to increase their defence spending by an additional 20 per cent and the United States to boost its defence budget by half that percentage, the result would be over US$100 billion in additional annual defence investments.

Possession of Taiwan would provide some extended range for weapons systems, and a wider arc for China’s intelligence, surveillance, and air defence networks. But to obtain this, Beijing would trade a significant proportion of its power projection forces and trigger a substantial military build-up in the region. This road leads to setback, not advantage, for China.


The stakes involved in a Chinese invasion of Taiwan, including the likely outcomes of such an action, therefore do not mandate the US making an unambiguous security pledge.[41] Our argument is not that there are no stakes involved in deterring Chinese aggression, only that the stakes for the United States are less than existential, and therefore smaller than China’s stakes. The security of Taiwan is important enough for the United States to take many actions, perhaps even including war depending on the context of the invasion. But it is not so important that the United States should rule out any option other than going to war.

In addition to being unjustified, an unqualified US promise to defend Taiwan could be strategically disastrous. Arguments for going all-in combine a worst-case account of the consequences of not doing so and a best-case account of the chances of success. They frame China as a dangerous, risk-acceptant aggressor, but confidently bet that Beijing is deterrable before a crisis, or will back down on the verge of attacking when confronted by the American superpower.

The first problem with this argument is that China’s interests in Taiwan are intense. For Beijing, Taiwan’s status is of first-order, existential importance. Once committed to an invasion, it is entirely possible that the CCP “could not afford to accept defeat”.[42] Washington will never care as much about the fate of Taiwan as China does.[43] Before and during a war, Chinese leaders will assume they can and must outlast the United States. The result is an unbridgeable imbalance of resolve that will make it harder than in most crises to deter China before or during a conflict, and shape any crisis over Taiwan.

For Beijing, Taiwan’s status is of first-order, existential importance. Once committed to an invasion, it is entirely possible that the CCP “could not afford to accept defeat”.

A US decision to go ‘all in’, moreover, does not by itself solve the credibility problem. The risks attached to fighting and losing, or winning a pyrrhic victory, are greater than not fighting at all. The United States would have tried to stand up to China’s aggression, and either failed or been badly bloodied — and the lesson of failure may be starker than of the US standing aloof. This risk highlights a contradiction in recent US assessments of the issue. Some official and expert statements alike have averred both that China might attack Taiwan within the next few years, and that given the present-day balance of forces and geography, the United States “gets its ass handed to it”[44] in such scenarios. If both claims are true, for the United States to make an unambiguous commitment to fight a war now would be to commit to a war it is likely to lose.

A related issue is the problem of allied responses. If the United States decides to fight, turns to allies for help — and they refuse, deterred by China’s threats of economic and military punishment as well as a lack of public support among their own people — then US forces would be left out on a limb. The American people would resent countries that deserted the United States in this desperate hour, and may well lose their stomach for further sacrifices to protect them. Choosing to fight an unpopular war for Taiwan could become a more direct avenue to the destruction of US relationships in the Indo-Pacific than not doing so.

US domestic political dynamics would pose other dangers. US public support for a war over Taiwan is modest: one recent poll found significant majorities of foreign policy elites supporting US intervention, but only about 40 per cent of the American public backing it.[45] (Even that figure likely reflects soft and reflexive backing; the country has not fought a major war against a peer adversary for two generations, and in the event of catastrophic US losses amidst perceptions that the war was unnecessary, public support could collapse very quickly.) If a president’s appeal to Congress for a war authorisation sparked a contentious debate and a close vote, that would partly undermine any intended message of US reliability. If Congress refused to endorse the war, it would jeopardise US credibility more comprehensively than US restraint. And if the United States were dragged into a lengthy, destructive conflict, the American people might demand an exit before winning, which would ruin the point of fighting in the first place.

Once the United States went ‘all in’ on a conflict, however, its national security decision-makers would surely consider America’s place in the Asian order and its general reputation as a security guarantor to be at stake. They would resist public pressure to back out — and, if it went badly, would confront pressure to escalate. Sooner or later, in the spiral of hostilities, several possible pathways to nuclear use would open.[46] Even short of a decision to employ nuclear weapons as the conflict continued, the United States would face intense operational military pressures to extend its military strikes to China’s mainland, where China’s nuclear and conventional forces are co-mingled.[47] This, in turn, could perilously resemble an attack to degrade or destroy China’s nuclear arsenal, triggering inadvertent nuclear escalation. Such escalation could also be sparked by attempts at nuclear coercion or from counterforce nuclear attacks. Or it could be accidental or unauthorised.[48]
China's arsenal includes the Dongfeng-31 nuclear missile launcher, on display at the Chinese People’s Revolutionary

Military Museum in 2017 to commemorate the 90th anniversary of the founding of the PLA. Image: Tyg728/Wikimedia Commons.

Any conflict over Taiwan would probably draw in the US homeland. It would most probably escalate in cyber terms, with China going after a range of US critical infrastructure, especially anything directly supporting military deployments. Emerging technologies would also allow Beijing to strike kinetically at the United States — with long-range hypersonic weapons, ‘Trojan Horse’ missiles hidden in shipping containers prepositioned in US ports,[49] or very long-range drones. Discussions of a Taiwan scenario sometimes treat it as a sort of twenty-first century Gulf War, with US forces deployed far from home to win a conflict from which the American people are largely insulated. A war with China will be nothing like that. If mutual homeland attacks do begin, the conflict is bound to escalate.

Finally, there is the issue of the military cost of a conflict. No one can be certain how badly the US military would be mauled in a war with China, but the effects would likely be severe: a significant proportion of the committed forces could be destroyed or disabled, losses that could cripple the US military for years. One recent study indicates that, due to constraints on the US defence industrial base, it could take between 5 and 50 years for the United States to replace current inventories of major weapons systems.[50] For years if not decades after a war, then, the US military would be a fraction of its present strength. So would China’s, but this would reduce the credibility of other US pledges around the world.[51]

Some believe that all of these considerations will be moot if the United States is willing to issue a credible commitment to go ‘all in’. Such a step will deter an attack, the argument goes, which is the best way to prevent war. Yet such a policy would commit the United States to a perpetual arms race to shape a military balance more than 10 000 kilometres from home against a peer competitor who has made this contingency the centrepiece of its military planning. Such an argument also oversimplifies the logic of deterrence. Countries choose war not just when they think they can get away with it, but occasionally when the strategic imperative for conflict becomes irresistible. The risks and costs of an attack on Taiwan, whether full-scale invasion, blockade, or selected strikes, are obvious. If Beijing decides to attack anyway, it will have decided that it has no choice — and be, at that moment, essentially undeterrable.[52] An added increment of US military commitment will not make a decisive difference. Taiwan may be a case where, when the deterrent power of countervailing military power is most needed, the desperate urgency of one party to the dispute makes it essentially irrelevant.

The risks and costs of an attack on Taiwan, whether full-scale invasion,
blockade, or selected strikes, are obvious.

Finally, the hefty price tag of an unqualified commitment to protect Taiwan would need to be traded off against other potential investments. The United States must keep its eye firmly on the larger contest: the broad-based competition of systems and societies emerging with the PRC. Historical lessons on the character of such competitions,[53] the geo-economic character of a wider contest defined by the difficulty of military conquest, and China’s emphasis on gradual, non-military tools of influence suggest that global economic engagement, technological rivalries, investments in diplomatic capability, and information security and cyber capabilities will be more important to the outcome of the larger competition than the status of Taiwan. At a time, for example, when the central office charged with overseeing US cybersecurity has been described as “underfunded, outmatched and exhausted”,[54] it does not make sense to pour tens of billions into capabilities oriented for the defence of Taiwan.

No one can be sure whether the United States would fight for Taiwan.[55] The political and diplomatic pressures weighing on a US president — not to mention the specific circumstances of an attack, just how unprovoked and cold-blooded it turns out to be — cannot be predicted. But going all-in on Taiwan stands a decent chance of failing to deter China’s adventurism. If it did fail, the outcome of the resulting conflict may well accomplish what the hawks fear most, eviscerating America’s position in Asia.[56] It would be a strategic error of the first order for the United States to leave itself no options when that moment arrives besides abandonment or all-out war. The United States has other, less absolute choices available.


Taiwan’s security is not so essential to US interests as to demand a US military commitment, and a shift in US policy in the direction of a more formal promise would be dangerous. Yet, we also oppose a simple abandonment of Taiwan, whether on the verge of war or in the form of some geopolitical bargain.[57] While many arguments for Taiwan’s strategic weight overstate the case, it remains true that if China swallowed Taiwan, unmolested by outside powers and watched passively by the United States, the result would raise serious fears about US dependability, attract widespread condemnation, and furnish Beijing with a stronger regional military position acquired at a lower cost. It would be a tragic error for the United States to fight a catastrophic war over Taiwan — but it would also be a mistake to signal to Beijing that it could subjugate Taiwan without paying a high price.

Our main argument, therefore, is not a plea for abandonment. It is about trying to find some way to mitigate the painful dilemma of the US position — to avoid war and yet deter and punish aggression, while making further aggression harder. This requires increasing the options available to a US president in the event of growing risks of Chinese action against Taiwan, based on the possibility that the United States could fulfill its essential interests in such a contingency without going ‘all in’.[58] This approach is designed in part to intensify the stakes of the choice China would confront in attacking Taiwan — to make clear to Beijing that it can either have “national rejuvenation” or take Taiwan by intimidation or force. But it cannot have both, at least on anything like the schedule the CCP has set for itself — a schedule designed to culminate by 2049 with China being firmly established as a fully developed, prosperous, regionally and globally predominant world leader.

It would be a tragic error for the United States to fight a catastrophic war over Taiwan — but it would also be a mistake to signal to Beijing that it could subjugate Taiwan without paying a high price.

Beijing’s aggression against Taiwan can take many forms. We focus on the most dangerous scenario because it is the one most consequential, namely some kind of direct assault on Taiwan. But the most likely route to subjugating Taiwan might be an attempt at coercion or strangulation short of direct conquest, via air and maritime blockade, or an air, missile, and cyberattack to coerce Taiwan and break its will.[59] One value of the options we propose below is that they provide ways of responding to such actions up and down the spectrum of conflict. The availability of grey zone and partial attack options for China demands a more complex menu of US response options.

That menu, and the broader US policy, should have the goal of sustaining the basic US policy in place since at least 1979: to make clear to Beijing that its interests will only be served by accepting a peaceful interpretation of its essential goal — ending what it views as an unresolved civil war and achieving a form of unification. Ruling out military action is likely to require compromises by Beijing in both the timing and substance of its concept of unification. The goal of American policy should be to make those compromises, painful as they may be, more attractive than the alternative, while relying predominantly on responses short of the threat of military intervention.

Some will argue that only such a threat of US involvement in the war could divert the CCP from its demand to absorb Taiwan. Yet the United States and others have significant leverage because China has other goals. It is on a long-term trajectory towards leading-nation status, both objectively and in its own formulation. Even short of going to war, the United States and others could impair that trajectory. Part of the muscle of this approach comes from the message, broadcast to leaders in Beijing from many other capitals, that the CCP will put its long-term development goals at profound risk with any military action against Taiwan.
The Foundation: Strengthening Taiwan

The first and most important component of such a US approach would be for the United States to serve as armourer rather than guarantor, to help Taiwan strengthen its own ability to defeat an invasion. Taiwan is, in fact, an eminently defensible nation. Both geography and weather are unfriendly to aggressors and tend to channel possible invasions into narrow apertures of space and time. Surprise is nearly impossible: China would need to build up massive forces in ways that would provide Taiwan with ample warning time.[60] If Taiwan acquires, over roughly the next five years, large numbers of additional anti-ship missiles, more extensive ground-based air defence capabilities, smart mines, better trained and more effective reserve forces, a significantly bolstered capacity for offensive cyber warfare, a large suite of unmanned intelligence, surveillance, and reconnaissance (ISR) and strike systems, and counterstrike capabilities able to hit coastal targets on the mainland, it will continually increase the price China will have to pay to win a war.[61]

Taiwan is gradually embracing such an approach, but even given recent increases, it does not yet spend enough on defence to give such a strategy teeth.[62] This has been a problem for some time: Taiwan’s defence cuts prior to 2018, one analyst argued then, after earlier cuts between 1993 and 2002, suggested “either a state of denial about the threat, a gridlocked political system, misplaced faith in current systems and geographic advantages, or perhaps most disturbingly, a belief that the United States is certain to provide timely military assistance.”[63] This is now changing, but slowly. Taiwan’s military reserves, which ought to be a leading tool of an island nation’s self-defence strategy, are poorly trained and equipped.[64] Open sources suggested that as of 2018, Taiwan was only able to fill 153 000 of 188 000 active-duty billets; reports suggest that even frontline units are only manned at 60 to 80 per cent strength. Taiwanese military exercises are rigid, scripted, and of low value.[65] Logistics shortfalls have reportedly even led to suicides by despondent Taiwanese soldiers.[66] Taiwan needs a serious domestic debate over the steps required to more fully provide for its own defence.
President Tsai Ing-wen posted this image to her Twitter account on 24 March 2020, accompanied by the text, "As the world grapples

with the severity of the #COVID19 pandemic, China's military maneuvers around #Taiwan have continued unabated. Whether it's national defence or preventing the spread of disease, our armed forces remain as vigilant as ever." Image: Tsai Ing-wen/Twitter.

Just as important as the amount Taiwan spends on defence is how it spends those funds. Military analysts almost universally agree that Taipei has over-invested in large, expensive, exquisite combat systems. Weapons like modern fighter aircraft, large surface combatants, and even large anti-aircraft missile systems like the Patriot will be vulnerable to the barrage of Chinese air and missile attack.[67] Happily, this is now changing: Taiwan’s 2017 Overall Defense Concept (ODC) pointed the country in the direction of what has been termed a “porcupine strategy” with a shift away from legacy systems to smaller, cheaper, more numerous and survivable, and more explicitly defensive capabilities.[68] This new approach has been manifest in some recent arms purchases[69] and will reportedly be confirmed in Taiwan’s 2021 Quadrennial Defense Review.[70]

The United States could encourage Taiwan’s planned evolution to such asymmetric capabilities in more direct ways. It could offer advice and direct aid to help Taiwan with purely defensive steps, such as national infrastructure resilience, stockpiles of key materials to deal with a potential blockade, and cyber resilience. More ambitiously, it could encourage and directly support development of technologies geared less towards power projection and more towards repelling invasion, such as cheap anti-ship missiles and UAVs,[71] the potential for which was again recently demonstrated in the fighting over Nagorno-Karabakh.[72] China would object, but the United States could argue that this cooperation on lower-technology systems is fully in line with its long-term commitments to reduce arms transfers.

US efforts to strengthen Taiwan may not mitigate one source of instability in cross-strait relations — Taiwan’s potential to take a unilateral stance, given its growing sense of distinct identity. Indeed, any effort to strengthen Taiwan carries an unavoidable risk of emboldening its governments and increasing the perceived space for pro-independence behaviour. This is a risk that broader US policy will have to address, through strengthened efforts to make clear that many of the additional steps below hinge on Taiwan’s own restraint.
Options for US Support Short of Warfighting

The foundation for a wider US option set is therefore supporting Taiwan in an enhanced investment in its own defence. Yet, in the event of Beijing’s coercion, blockade, or attack, the United States will require more specific responses.

There are ready-made models for an alternative approach, ones which reflect a well-established US tradition of supporting clients or allies short of going to war. Before entering the Second World War, the United States provided Lend-Lease assistance to support Great Britain’s effort to hold out against Germany. During the early Cold War, it answered Soviet blockades of Berlin with airlifts of supplies rather than military action. It played armourer and supplier to Israel during several Middle East wars. During the 1980s, it provided support to resistance and insurgent groups — most notably in Afghanistan. Since 2014, it enhanced Ukraine’s resistance capacity against Russian aggression. Taiwan would merely be the latest in a string of US partners who benefited from these less-than-absolute types of support.

The elements laid out briefly below certainly require more analysis and elaboration. As it develops such ideas, Washington should seek the greatest multilateral cooperation possible. Japanese, Australian, and European attitudes towards China have been hardening. None of these countries is anxious to join a war for Taiwan, but all might be increasingly willing to push the boundaries of their policy to help deter and, if necessary, punish such aggression in other ways.[73]
Extend Warning Time

While some commentators speak in terms of a no-notice Chinese attack, most recognise that there would be considerable warning time before an invasion.[74] US policy should seek to extend it. Such an effort could include expanded intelligence, surveillance, and reconnaissance (ISR) assets (including persistent drones),[75] proliferated sensors, and space-based surveillance. Washington could join with allies to help fund a Taiwanese equivalent of Bellingcat[76] to watch open-source intelligence for signs of impending invasion or blockade as well as grey zone harassment campaigns, and help rally global opinion if war or blockade loomed. This element of the strategy could be thoroughly multilateral, with countries such as Japan, Australia, France, and Great Britain joining a “strategic stability” endeavour of sensing and warning to make aggression less feasible.
Unmanned aerial vehicles (UAVs) are a key surveillance capability. China's WJ-700 high-altitude, long-endurance and high-speed
armed reconnaissance drone makes its maiden flight on 11 January 2021. Image: China Aerospace Science and Industry Corporation (CASIC).

The purpose of such enhanced warning time would be not only military — giving Taiwan more time to mobilise reserves and prepare for missile and air strikes — but also political. If China confronts the prospect of being bombarded with warnings and condemnations as it plans for war, it may be less likely to start down that road. Conversely, were China to avoid such scrutiny in the interim pre-war crisis by launching a sudden surprise attack, it would have to risk doing so without proper material preparation, from a standing start. Beijing would face a dilemma that is familiar in any Taiwan scenario, whereby “surprise and deception would come at the expense of preparedness and speed of engagement with large-scale forces, and vice versa.”[77]
Prepare Approaches to Deal with a Chinese Blockade of Taiwan

As noted above, CCP decision-makers could turn first to a blockade and associated harassment to compel Taiwan’s surrender. Such a move would transfer the burden of escalatory risk to Washington, which would have to force the blockade if it chose to deliver supplies, thus risking conflict. The United States could help Taiwan prepare for such a contingency by enhancing Taiwanese self-sufficiency in food, energy, and other necessities. It could rally multilateral statements opposing any such blockade and promising to deliver, through UN or other nongovernmental auspices, specific critical supplies such as medicine. It could develop and communicate specific punishments it would levy on China in the event of such a scenario and coordinate such punishments with other countries.
Develop Non-Combat Means of Supporting Taiwan’s Defence

In the event of coercion or invasion, the United States could support Taiwan in ways other than committing the US military. Especially during a campaign of coercion short of war, it could offer non-military supplies and assistance ranging from food and medicine to support in global financial markets. During the warning time before an attack, it could rush hundreds of precision weapons — in Lend-Lease fashion — to Taiwan, bolstering Taipei’s ability to impose costs on China. Washington could share its own intelligence and battlefield awareness during both crisis and wartime.[78]

The US effort to assist Israel during the 1973 war, Operation Nickel Grass, offers an interesting parallel. The Nixon administration organised a massive relief effort of military supplies — over 55 000 tons, delivered by air and sea, including transferring possession of more than 100 US F-4 fighter-bombers, A-4 Skyhawks, and C-130 transports. Conducting such resupply for Taiwan during wartime would probably not be possible without engaging in the war, but the United States (and others) could be poised to conduct such a support operation in the window of strategic warning before an actual attack.[79] Elements of US military posture in Asia could be specifically designed to provide stockpiles of the kinds of arms that would bolster Taiwan’s defence (or, indeed, the defence of any regional partner) most powerfully if airlifted or sealifted in during a warning period. Such a concept could be linked to enhanced roles for Taiwanese reserve forces, some of which could be trained to operate these systems.
Taiwanese President Tsai Ing-wen inspects the Han Kwong 33 Exercise. Image: Office of the President, Republic of China (Taiwan)/Flickr.

An especially useful form of assistance would be for the United States to enhance Taiwan’s cyber capabilities to help cripple a Chinese attack.[80] Such a program could bring Taiwanese cyber experts into international information security networks, fund the training of cyber warriors, and even offer zero-day exploits[81] and other directly useful intelligence and information. Openly or covertly, the United States might help Taiwan to build a network of state-supported hackers abroad, groups that could continue to impose costs on China for months or years after an invasion or occupation has occurred.

Even a United States that remained out of the fight, moreover, would impose operational military costs on the PLA. Regardless of its expectations of US involvement, Beijing would have to account for the possibility that the war would escalate accidentally, or that the United States would decide at some point during the fighting to jump in. To hedge against such risks, even a PLA fully committed to a Taiwan blockade or invasion would have to reserve significant forces to mitigate the danger of a delayed American blow. The United States could develop concepts to posture its forces before and in the early stages of a war in ways that maximised these hedging requirements. This would reduce the scope and quality of capabilities the PLA could devote to the actual attack, and indirectly support Taiwan’s defence.
Conduct Clandestine Attacks on Chinese Assets

A riskier step would be to launch deniable blows of various kinds on Chinese forces, to disrupt and weaken their invasion. These could include cyberattacks launched from global locations, with the US hand in them concealed, made to look like part of Taiwan’s own cyber campaign. They could include covert intelligence programs to sabotage Chinese infrastructure directly supporting the invasion. They could involve kinetic attacks from partly deniable capabilities: the growing potential of unmanned systems, for example undersea drones, could offer new possibilities in this area, as could unmanned systems for deploying mines, either prepositioned or moved into position in a crisis. The United States could employ a small number of torpedo- and missile-firing mini-submarines which Taiwan would claim as its own.
In August this year, the American Institute in Taiwan (AIT) shared a photo on their Facebook page that shows a
US Air Force Boeing KC-135 Stratotanker refuelling a Taiwanese F-16 fighter jet. Image: AIT/Facebook.

Such covert but direct involvement would risk escalation if the US role were uncovered, or suspected. It could be justified on the expectation that, as in many Cold War cases of proxy conflict such as US support for the Afghan mujahideen, China would prefer to join in the pretence — to absorb these blows rather than generate a direct crisis and bring down the effect of the whole US military. Such a dynamic might emerge only if the United States stayed out of the conflict in formal terms. China would have an incentive to accept a certain amount of harassment rather than escalate and risk drawing the United States in. Still, such direct but covert involvement would carry risk.
Prepare Non-Military Retaliation

A central part of any US and international strategy to deter and punish Chinese aggression against Taiwan must be broad-based, non-military punitive measures. To set the stage for these, the United States should continue recent efforts to increase Taiwan’s international engagement.[82] The more Taiwanese students travel abroad (and international students visit Taiwan); the more Taiwan-sponsored cultural exhibitions or sporting events occur; the more international organisations Taiwan can join; the more international firms do business in Taiwan and the more trade other countries have with it, the greater will be the shock of a Chinese attack. Breaking China’s effort to strangle Taiwan’s interaction with the world — without violating the core of the US One China policy — is thus an important foundation for this approach.[83]

In the event of blockade or attack, the United States could lead an international coalition in applying economic, political, and cultural sanctions. This could include a program for gradual economic decoupling — something few countries are willing to consider today, but which could be promoted as a geopolitical necessity in the wake of an attack. It could include a freeze on all US assets in China and an end to dollar-based transactions. It could seek to confiscate selected Chinese-operated and owned investments around the world. It would involve a new effort to acquire iron-clad commitments freezing Chinese information technologies companies out of the digital infrastructures of other countries.

The United States could begin to lay the groundwork for these steps before a crisis or war. US officials could generate draft prohibitions on contact with Chinese industry, scholars or scientists, and a total ban on educational exchanges, which would be put into effect after an attack. It could design and threaten a global public relations campaign to condemn the Chinese leadership and lead an effort to throw China out of a host of international organisations. The United States might even indicate that an attack on Taiwan would be the one event which would cause it to abandon its One China policy, formally recognise Taiwan as an independent state, and support a government in exile.

The prospect of these actions alone may not deter China,[84] but they would influence the balance of power after an attack. Rising powers want and need legitimation narratives that justify their growing influence to the international community.[85] As well as inflicting material attrition, a comprehensive, persistent effort to deny China the recognition it craves would constrain its ability to exercise any power other than coercive influence. Robert Blackwill and Philip Zelikow suggest that the goal of such reputational punishment would be “to develop a picture in Beijing of the world that could follow a local war over Taiwan… China would have to redefine its future where it had provoked a division of the world in which a large part mobilized against China to an extent that had never happened before.”[86]
Help Establish a Swift Multilateral Counterbalancing Presence in the Event of Invasion

The United States could also take steps to ensure that a Chinese attack would usher in a more elaborate and forward-leaning posture by the United States and its partners, which along with enhanced local defence efforts would confront Beijing with a stronger multilateral blocking position in the region. Most countries in the region do not favour more US forces or facilities on their territory, or steps to cooperate too closely with the United States in preparing for war against China. This would presumably change to some degree in the aftermath of an attack. Washington can work now to establish the outlines of what such a bolstered US and multilateral position would look like, along with added mutual support among the countries of the region. It could recruit European help for such a mission, grounded in Europe’s economic interests, and interest in the well-being of hundreds of thousands of EU citizens in the region.[87] The United States and others could also make clear that a Chinese attack on Taiwan would likely lead them to take stronger positions on territorial disputes in the region, specifically the South China Sea.
Bolstered Options Need Not Undercut Deterrence

Some observers fear any agenda which provides the United States with options short of war. Former US National Intelligence Officer Paul Heer has written that, “Chinese leaders — like their Taiwan counterparts — have long presumed and planned that the United States would intervene militarily in response to a Chinese use of force against the island.”[88] Some could argue that a US effort to strengthen its options short of major war would signal that it intends to stand aloof, encouraging China to believe that it could invade Taiwan without US intervention and undercutting deterrence. We disagree. First, options from our list, starting with significant new efforts from Taiwan, would enhance deterrence in concrete ways not likely to be outweighed by any doubts about US intentions. Second, if it chooses, the United States can reaffirm its current policy of ambiguity in many ways, from public statements to forward-postured naval forces and specific warnings in a crisis.

Most fundamentally, it would be irresponsible for the United States to leave itself no option in the event of Chinese aggression other than total war; if the development of a wider set of credible deterrent and punitive options carries risks, those should be accepted and minimised. It would be strategic folly for the United States to implicitly burn its bridges in service of a commitment that is itself qualified and uncertain.


This is not the first time the United States has confronted a threat to a far-off nation and backed itself into a Hobson’s choice[89] between passivity and tragedy. US officials and national security experts once spoke zealously about the importance of Indochina and the need to affirm US credibility by holding the line against communism. And yet, presidents from Dwight Eisenhower through Richard Nixon doubted that even a vigorous US effort would bring success. Eisenhower saw the misery awaiting his country in a grinding counterinsurgency war almost 10 000 kilometres from home: “No military victory is possible in that kind of theater”, he worried. “Even if Indochina is cleared of Communists, right across the border is China with inexhaustible manpower.”[90] Yet a consensus arose within the American national security establishment that the United States must stop communist progress in Southeast Asia. It was either go all-in and fight, or “abandon” South Vietnam — and the region. This mentality led to the agony of Lyndon Johnson, speaking to Senator Mike Mansfield in June 1965. “Mike, you either get out or you get in”, Johnson lamented. “I don’t think there’s much more neutral.”[91]

The parallels between the dilemmas of Vietnam and Taiwan are instructive. A growing certainty in the United States that regional and global security hinges on a single contest. Claims that America’s rival has an unquenchable appetite in a region ripe for dominance. The conviction that there is no middle ground—that to deter and punish aggression, the United States must prepare to fight the war itself. The result in Vietnam was to compel the fighting of an unnecessary war, and the United States may be sliding towards a similar abyss over Taiwan — but this time the stakes are far higher. This time, the dilemma could trap the United States in precisely the outcome it seeks to avert — a major war with a nuclear-armed China. It could lead to the devastation of the US military, a collapse of public support for power projection in Asia, an America fatally weakened for the broader competition with China and in its position elsewhere, and —perhaps— a nuclear exchange. Washington ought to think along indirect lines, rather than planning to directly engage with every aggression it hopes to deter. Such an approach demands fresh thinking, before it is too late.