Taiwan inaugurated Lai Ching-te as president on 20 May after his victory in January’s election, extending the Democratic Progressive Party’s control of the presidency into a third consecutive term. Four days later, China concluded two days of major military exercises around Taiwan involving aircraft, naval vessels and simulated strikes, presenting the drills as punishment for what Beijing considers separatist rhetoric. Chinese authorities were especially angered by Lai’s statement that the two sides of the Taiwan Strait are “not subordinate to each other”, while Lai’s government has continued to argue that Taiwan is already a sovereign political entity under the Republic of China and that its future should be determined by the people living there. The immediate crisis looks military, but the deeper contest is political: China is attempting to use coercive pressure to prevent Taiwanese identity from moving further away from eventual unification, while that same pressure may be making such unification less politically attractive inside Taiwan.
That creates one of the clearest examples of a security dilemma becoming an identity dilemma. Beijing sees Taiwan’s political evolution as progressively weakening the possibility of peaceful unification. From its perspective, stronger military signalling is therefore intended to deter formal independence and remind Taiwanese leaders that constitutional or symbolic moves towards permanent separation could carry enormous costs. Taipei sees those same military exercises as evidence that China is willing to threaten a democratic society for making political choices Beijing dislikes. The more coercion increases, the easier it becomes for Taiwanese politicians to argue that closer political integration with China would endanger the autonomy and democratic institutions voters currently possess. A policy intended to preserve the possibility of unification can therefore reduce the willingness of the population whose consent would make peaceful unification possible.
This feedback loop matters because Taiwan’s status cannot be understood only through military capability. China possesses vastly greater armed forces and has never renounced the use of force to bring Taiwan under its control. Taiwan, however, has developed a distinct democratic political identity over several decades. Its citizens elect presidents, legislators and local governments under institutions separate from the People’s Republic of China, while Beijing has never governed Taiwan. The central strategic problem is therefore not simply whether China can impose enough military costs to defeat Taiwan. It is whether coercive power can produce a political outcome that remains governable afterwards.
Military conquest and political integration are different tasks. A state can destroy an opponent’s armed forces more easily than it can make millions of people accept the legitimacy of the new political order. This distinction becomes particularly important when the target population already possesses functioning institutions and a separate political identity. If Beijing eventually used force successfully, it would inherit not an administrative vacuum but a society whose electoral institutions, media environment and civic expectations differ fundamentally from those of the mainland. The cost of occupation or political pacification could therefore become much larger than conventional military calculations alone suggest.
This is why deterrence should be analysed through expected political outcomes rather than battlefield victory alone. China may ask whether it can seize Taiwan. The more important question is what happens after seizure. Would economic infrastructure remain intact? Would highly skilled workers stay? Would international sanctions follow? Would semiconductor production continue? Would civil resistance emerge? Would neighbouring states increase military alignment with the United States? A successful invasion could therefore produce a strategic outcome much less favourable than the territorial victory suggests.
Taiwan’s semiconductor industry magnifies this problem. TSMC is the world’s largest contract chipmaker and sits at the centre of advanced semiconductor supply chains used in artificial intelligence, smartphones, data centres and military technology. Taiwan’s economic importance therefore extends far beyond its population of roughly 23 million. Any major conflict would impose costs on companies and governments across the world, including China itself, whose industries remain deeply connected to Taiwanese and global semiconductor networks. Reuters reported that China remained Taiwan’s largest trading partner, with bilateral trade of about $224 billion in 2023, illustrating how extraordinary the economic contradiction has become: two economies can remain deeply interdependent while their security relationship deteriorates simultaneously.
This is another reminder that interdependence does not eliminate conflict. Liberal international-relations theory often assumes trade raises the opportunity cost of war because governments stand to lose economically. That mechanism is real. It does not make war impossible because states value security, territory and political legitimacy alongside income. Europe traded extensively with Russia before the invasion of Ukraine. China and Taiwan maintain enormous commercial relationships while preparing for scenarios in which that commerce collapses. Economic interdependence raises the cost of conflict; whether it prevents conflict depends on whether leaders believe the political stakes are worth paying that cost.
The stronger version of the economic-peace argument would therefore be incorrect. Trade does not automatically create peace. It creates hostages to peace: assets, supply chains and future income that would be damaged by war. Those hostages strengthen constituencies favouring stability but cannot guarantee they dominate national-security decision-making. If governments begin believing the expected cost of inaction exceeds the expected economic loss from war, interdependence can become insufficient.
China’s military drills are designed partly to alter exactly that calculation on Taiwan. They demonstrate that Beijing can increasingly operate aircraft and ships around the island, pressure Taiwan’s air defences and rehearse missions involving blockade or strike operations. The People’s Liberation Army described this week’s exercises as a warning against separatism and state media presented them as testing capabilities around key areas. The strategic message is that Taiwan cannot assume geographical separation or American support makes independence costless.
Yet coercive signalling contains an information problem. China wants Taiwan to infer that escalation beyond certain political limits would be dangerous. Taiwanese voters may instead infer that China threatens them regardless of moderation. If the perceived probability of coercion remains high under both moderate and independence-leaning governments, the incentive to moderate declines. Why sacrifice preferred political identity for concessions that do not appear to purchase greater security?
This is the same mechanism sanctions sometimes create. If a target believes punishment will continue regardless of behaviour, changing behaviour produces little benefit. Effective coercion therefore requires not only threats but credible reassurance: if the target complies with specified conditions, costs will actually decrease. Without that second half, coercion becomes punishment rather than bargaining.
Beijing’s current Taiwan strategy struggles with this because the minimum political condition China seeks, acceptance that Taiwan belongs within one China, is increasingly difficult for major Taiwanese parties to endorse in the meaning Beijing prefers. Lai has said he intends to maintain the status quo rather than formally declare independence, while Beijing still views him as a dangerous separatist. If even explicit commitments to preserve the status quo fail to reduce Chinese pressure substantially, Taiwanese voters may conclude that moderation changes very little.
This creates what could be called a credibility gap of accommodation. China wants Taiwanese leaders to believe restraint reduces risk. Taiwan needs observable evidence that restraint produces a safer environment. Repeated large-scale drills after elections or political speeches can weaken that belief by making coercion appear structural rather than conditional.
The Kuomintang, or KMT, faces this problem especially sharply. It traditionally advocates closer cross-Strait dialogue and argues that communication reduces the risk of war. Former president Ma Ying-jeou met Xi Jinping in Beijing in April and called for peaceful relations, while Xi reiterated that external interference could not prevent eventual “reunion”. The KMT therefore offers Taiwanese voters a theory of security based partly on engagement. The DPP offers a theory based more heavily on resilience, international support and protecting Taiwan’s existing political autonomy.
Elections become tests between those competing security models. The DPP’s third consecutive presidential victory suggests many voters remain unconvinced that greater political accommodation with Beijing provides enough security benefit to justify the associated political risks. That does not mean Taiwanese society is uniformly pro-independence. Preferences are more complicated, and many citizens favour maintaining the ambiguous status quo precisely because both formal independence and unification appear risky. The important point is that Chinese military pressure does not seem to have produced a straightforward shift towards accommodation.
This is where coercion can become counterproductive through endogenous preferences. Governments often assume the target’s political preferences are fixed and coercion merely changes the cost of acting on them. In reality, coercion can alter the preferences themselves. A voter previously indifferent towards national identity may become more strongly Taiwanese after repeated military threats. Another may become more cautious and favour accommodation. The aggregate outcome depends on which reaction dominates.
Hong Kong has probably affected this calculation too. Beijing once promoted “one country, two systems” as a framework for eventual Taiwan unification. The political changes in Hong Kong following the 2019 protests and the 2020 national-security law weakened the attractiveness of that model among many Taiwanese, because assurances of long-term political autonomy became less credible. This is another example of reputation spillover. A government’s treatment of one territory changes how another population evaluates promises made to it.
Credible commitment problems therefore sit at the centre of cross-Strait relations. China can promise Taiwan a degree of autonomy after unification, but once unification occurred Taiwan would lose much of the bargaining leverage making that promise necessary in the first place. Taiwanese voters therefore need confidence that Beijing would continue honouring autonomy after the balance of power shifted permanently. Hong Kong provides evidence they can use when estimating that probability.
This is a classic time-inconsistency problem. Before integration, the stronger actor promises restraint to make agreement attractive. After integration, incentives can change because the weaker actor loses the ability to exit. Institutions can solve such problems when neutral enforcement exists. No mutually trusted supranational institution exists capable of enforcing a China-Taiwan political settlement against Beijing after unification. That makes credible guarantees extraordinarily difficult.
Military deterrence therefore substitutes partly for missing political trust. Taiwan invests in defence because it cannot rely entirely on Chinese promises. The United States supplies weapons because American strategy seeks to preserve enough uncertainty around the outcome of coercion that Beijing avoids using force. Washington does not maintain formal diplomatic relations with Taiwan, but under the 1979 Taiwan Relations Act it is required to provide Taiwan with the means to defend itself. This arrangement creates deliberate strategic ambiguity: the United States supports Taiwan’s defence without providing an unconditional treaty guarantee that it would intervene militarily in every scenario.
Strategic ambiguity attempts to deter both sides simultaneously. Beijing should fear that attacking Taiwan might bring American intervention. Taipei should not become so certain of American intervention that it takes political risks Washington considers destabilising. Ambiguity therefore creates a middle equilibrium between abandonment and unconditional commitment.
The problem is that ambiguity works only while every actor remains sufficiently uncertain in the intended direction. If Beijing becomes convinced Washington would not intervene, deterrence weakens. If Taipei becomes convinced intervention is guaranteed, incentives for restraint could change. If American commitments become so explicit that Beijing believes peaceful unification is permanently impossible, Chinese incentives to act earlier could increase.
This is why statements about Taiwan from American leaders generate disproportionate attention. One sentence can affect beliefs about where the actual red lines sit. Diplomatic language becomes part of deterrence architecture.
The Taiwan Strait is therefore not simply a military balance. It is a belief system. Beijing estimates American resolve and Taiwanese resistance. Taipei estimates Chinese willingness to use force and American willingness to help. Washington estimates whether Beijing believes coercion can succeed and whether Taiwan’s government will preserve the status quo. Stability depends on these estimates remaining within a range where nobody sees immediate escalation as optimal.
Military exercises change beliefs.
Weapons sales change beliefs.
Election speeches change beliefs.
Diplomatic visits change beliefs.
Every actor therefore adjusts not just capabilities but perceptions of capabilities and intentions.
This is why apparently symbolic language such as Lai saying the two sides are “not subordinate to each other” matters strategically. Beijing interpreted the formulation as suggesting two separate states, while Taiwanese officials regard the underlying idea as consistent with the island’s existing constitutional reality. One sentence acquires military significance because sovereignty itself remains disputed.
The semantic disagreement reflects a deeper structural one. Beijing considers Taiwan part of China even though the People’s Republic has never governed the island. Taiwan’s current government considers the Republic of China already sovereign and therefore argues it does not need a formal declaration of independence to become something it believes already exists. These positions create very little conceptual overlap.
The status quo survives partly because both sides avoid resolving that contradiction definitively.
Ambiguity therefore is not diplomatic failure.
It is the equilibrium.
Taiwan operates as a self-governing democracy without declaring a new Republic of Taiwan. China claims sovereignty without attempting immediate military annexation. Most foreign governments recognise Beijing diplomatically while maintaining substantial unofficial relations with Taipei. The United States supports Taiwan’s defence without recognising it as an independent state.
Every participant accepts an arrangement containing contradictions because resolving them may be more dangerous than maintaining them.
This is a useful reminder that political stability does not always require conceptual clarity. Sometimes ambiguity allows actors with incompatible preferences to coexist because each can maintain part of its preferred interpretation. International lawyers may dislike unresolved status. Strategists sometimes depend on it.
The danger lies in shrinking ambiguity. If Taiwanese politics moves towards explicit permanent separation, Beijing may conclude the status quo is disappearing. If Chinese pressure becomes indistinguishable from preparation for invasion, Taipei may conclude ambiguity no longer provides security. If Washington clarifies its military commitment completely, Beijing’s assumptions change again.
Stability therefore requires several actors to tolerate statements and behaviours they dislike while avoiding actions that make the existing compromise impossible.
That is difficult because domestic politics rewards clarity. Nationalist constituencies want leaders to state exactly who owns the territory. Military planners want clear commitments. Voters dislike feeling threatened indefinitely. Yet strategic ambiguity asks governments to preserve uncertainty intentionally.
The result can feel politically unsatisfying precisely because it may be safer.
Taiwanese public calm during this week’s drills demonstrates how normalised the threat has become. Reuters reported that daily life and financial markets remained relatively steady even while Chinese forces conducted exercises around the island. Taiwan’s benchmark stock index and TSMC shares remained close to record levels, with many residents describing military intimidation as something they had experienced repeatedly over decades. This resilience reduces Beijing’s coercive leverage because threats create less immediate economic panic than they might if the population believed invasion were imminent.
But habituation has its own danger. If businesses, citizens and governments become too accustomed to military exercises, they may underreact when preparations eventually differ meaningfully from previous drills. This is the cry-wolf problem of grey-zone coercion. Repeated threats can normalise military presence and exhaust attention. The aggressor gains familiarity with operating around the target while the target becomes less sensitive to each individual escalation.
Grey-zone pressure is effective partly because it changes the baseline gradually. An aircraft crossing a previously respected line once appears dramatic. Repeated crossings make the activity routine. Naval exercises expand geographically. Coastguard patrols become more frequent. Each individual step remains below the threshold of war, yet the cumulative strategic environment changes.
This is salami slicing. No single action appears serious enough to justify a major response, but accumulated actions alter facts on the ground or sea. China has used versions of this approach in the South China Sea as well as around Taiwan. The target faces a difficult choice: respond strongly to a relatively limited incident and appear escalatory, or tolerate it and allow the new behaviour to become normal.
Taiwan therefore needs resilience not only against invasion but against permanent coercive pressure. That includes military readiness, cyber defence, energy security, communication networks and social trust. China’s objective need not always be conquest. Pressure can aim to exhaust Taiwanese defence resources, undermine confidence and persuade citizens that resisting Beijing creates endless instability.
Resilience changes that calculation by lowering the political return from coercion. If drills do not crash markets, divide society or alter elections in Beijing’s preferred direction, their effectiveness declines even if their military value remains.
This creates another interesting economic problem. Taiwan needs enough defence expenditure to make coercion costly while avoiding militarising society so heavily that fear itself damages economic attractiveness. The island’s strategic advantage comes partly from being a prosperous, technologically advanced democracy. Defence policy needs to preserve the society being defended.
The same applies to semiconductor diversification. Taiwan and TSMC are investing in production abroad, including in the United States and Japan, partly because customers and governments want supply-chain resilience. Diversification reduces the global economic catastrophe associated with a Taiwan conflict. It can also weaken the informal idea of a “silicon shield”, according to which the world’s dependence on Taiwanese chips gives foreign governments unusually strong incentives to protect the island.
Taiwan therefore faces another paradox. Diversifying semiconductor production makes global supply chains safer while potentially reducing one source of strategic leverage derived from concentration.
The silicon-shield concept should not be exaggerated. No government would base war decisions entirely on one company. Yet economic importance affects political attention. Taiwan’s technological centrality gives governments and businesses around the world concrete interests in stability beyond abstract democratic solidarity.
This is economic deterrence through network centrality.
The more indispensable a node becomes, the greater the costs imposed by its disruption.
But excessive dependence creates vulnerability for everyone else, giving them incentives to diversify.
Successful strategic importance therefore contains incentives for others to make you less strategically indispensable.
Again, power changes the behaviour of actors around it.
The semiconductor issue also makes a Chinese attack economically self-harming. China itself depends extensively on advanced chips produced through Taiwanese and international supply chains. War could therefore destroy access to technologies central to Chinese industry. Economic interdependence imposes costs in both directions.
Yet governments sometimes accept self-harm for strategic objectives. Russia sacrificed extensive European economic integration to invade Ukraine. The lesson should not be that economic costs are irrelevant, but that analysts should avoid assuming leaders value GDP above territorial and political goals automatically.
Xi Jinping has repeatedly emphasised that unification remains a historic objective. That means Taiwanese and American strategy must treat the risk seriously even without assuming an invasion is imminent. The objective of deterrence is precisely preventing events whose probability is uncertain but whose cost would be enormous.
This requires avoiding two analytical extremes. Treating every Chinese exercise as evidence that invasion is about to begin creates permanent alarm and eventually reduces credibility. Treating exercises as meaningless theatre because previous ones did not produce war ignores how military preparation and coercive baselines can change gradually.
The correct approach is comparative signalling analysis: what changed from previous exercises? Which units participated? How close did forces operate? Were civilian shipping routes affected? Were mobilisation patterns different? Did exercises rehearse blockade, missile attack or amphibious operations? Intelligence matters more than headlines.
This is another domain where public rhetoric and professional risk assessment diverge. News organisations need a dramatic event. Military planners need deviations from baseline. A drill can be politically dramatic while operationally familiar, or publicly quiet while containing strategically important changes.
Taiwan’s citizens appear to understand part of this distinction intuitively because repeated exercises no longer automatically produce panic. That social resilience is valuable. It should coexist with serious preparation.
The broader international response also needs calibration. Excessive symbolic support for Taiwan can create political satisfaction while increasing pressure without improving Taiwan’s actual defence. Visits by foreign politicians receive headlines. Submarines, anti-ship missiles, hardened infrastructure, energy reserves and trained personnel produce less immediate political visibility but may contribute more to deterrence.
This is the difference between performative commitment and credible capability.
The United States and its allies should therefore judge Taiwan policy according to whether actions make coercion less likely rather than whether they demonstrate rhetorical toughness towards Beijing. China should apply the same logic in reverse: military punishment may demonstrate resolve while making Taiwanese voters less willing to consider political accommodation.
Every side can become trapped by signalling intended primarily for domestic audiences.
Taiwan’s election provides another source of institutional resilience because Lai won the presidency but his DPP lost its parliamentary majority. The opposition KMT became the largest party in the legislature, while the Taiwan People’s Party holds enough seats to influence outcomes. Presidential authority therefore coexists with a divided legislature. This can create domestic gridlock, but it also demonstrates that Taiwanese voters did not hand one political faction complete control of the state.
That matters when Beijing portrays the DPP as synonymous with Taiwan. It is not. Taiwanese society contains substantial disagreement over China policy, economic policy and domestic governance. A democratic political system allows those disagreements to remain visible instead of compressing them into one national position.
This pluralism complicates Chinese strategy because there is no single Taiwanese actor Beijing can coerce into determining the island’s preferences. Even a president willing to make substantial concessions would need political support sufficient to implement them.
Democracy therefore creates commitment complexity. Negotiating with an authoritarian state can appear easier because one leader may possess enough authority to implement an agreement. Democracies contain elections, legislatures, courts and public opinion capable of reversing policy. That can make agreements harder. It can also make coercion less effective because replacing one leader does not guarantee changing the entire political system.
Taiwan’s political identity has become institutional rather than merely partisan.
That is why the inauguration of one president can trigger Chinese military exercises while not determining the underlying strategic trajectory by itself.
Lai matters.
The institutions and voters producing Lai matter more.
This returns to the central dilemma. Beijing wants to prevent permanent separation. Military pressure signals the cost of crossing its red lines. Yet the same pressure can make Taiwanese citizens value their existing autonomy more strongly and distrust promises of peaceful integration. The policy can therefore succeed tactically while failing politically.
Taiwan wants to preserve autonomy without triggering a war it cannot win alone. The United States wants to deter China without encouraging moves towards formal independence. China wants to compel restraint without convincing Taiwanese society that coercion is permanent regardless of restraint.
All three therefore depend on credible limits.
China needs to communicate what behaviour would actually reduce pressure, not merely what increases it.
Taiwan needs to communicate that maintaining its democracy does not require a unilateral declaration designed to humiliate Beijing.
Washington needs to communicate that supporting Taiwan’s defence is not equivalent to supporting every possible change in Taiwan’s constitutional status.
That is an uncomfortable equilibrium because nobody receives everything they believe history or sovereignty entitles them to.
It may be precisely why the equilibrium has survived.
The military exercises this week are a warning that the margin for error is narrowing. Every new exercise makes Chinese forces more experienced around Taiwan. Every Taiwanese election reinforcing a separate political identity makes peaceful unification harder to imagine under Beijing’s preferred terms. Every American security commitment increases Taiwan’s defensive confidence while increasing Chinese anxiety about permanent separation.
The system therefore contains three individually rational strategies that can collectively make confrontation more likely.
Taiwan strengthens defence because China threatens it.
China increases pressure because Taiwan strengthens separate political identity and foreign ties.
The United States deepens support because Chinese pressure increases.
China interprets deeper American support as further evidence that peaceful unification is being prevented.
The loop repeats.
Breaking it does not require pretending the underlying disagreements disappear. It requires enough reassurance that defensive actions do not become indistinguishable from offensive intentions.
That is difficult because the political objectives themselves overlap geographically.
Taiwan wants to preserve the government that exists.
China ultimately wants that territory under another sovereign framework.
There is no technical confidence-building measure capable of eliminating that contradiction.
The achievement of cross-Strait stability has therefore been managing an unresolved conflict rather than solving it.
That may continue to be the least bad option.
Lai Ching-te has now taken office.
China has responded with war games.
Taiwan has continued functioning.
No side has crossed into war.
For the moment, deterrence is still working.
But deterrence should not be measured by how many missiles can be intercepted or how many ships can surround an island.
Its real success is the crisis that stops before anybody needs to discover whether those capabilities work in combat.
The danger is that repeated coercion teaches every side the wrong lesson: that because escalation was controlled this time, it can safely be repeated next time.
It cannot be assumed.
The Taiwan Strait remains stable partly because governments have spent decades refusing to resolve its central contradiction completely.
The most important task is keeping political pressure from turning that ambiguity into a deadline.
