Introduction
PRC pressure on Taiwan has intensified since 2016. Actions range from suspending PRC tourist visits to Taiwan and imports of Taiwanese goods to increasing military activity around Taiwan, disinformation efforts, and even undue influence in domestic politics (The Economist, 2021). Beijing's intent behind such coercive behavior is to weaken Taiwan's opposition to demands such as accepting its ‘92 Consensus’ and ‘One Country, Two Systems’ formula (Maizland, Reference Maizland2021). The Chinese Communist Party (CCP) leadership appears to believe that these steps are vital to preventing what they see as de jure Taiwan's declaration of independence, although such steps may complicate efforts to compel the island and its population to accept unification. Given the stark asymmetry in economic size, military capability, and official diplomatic isolation, Taiwan provides a good case for studying coercive diplomacy that takes the form of threats to punish. Material differences suggest that Taiwan should capitulate, and ‘cheap talk’ thesis expects threats to have no discernable effect. Instead, Taiwan demonstrates a remarkably persistent and rising will to resist, driven by growing distrust and non-identification with the PRC, perhaps consistent with balance of threat theories.
Behind Taiwan's resolve lies growing public defiance driven by the increasing frequency and expanding scope of direct and indirect PRC coercion designed to spur submission. Even as such sentiments of resistance rise, they are not evenly distributed across society. Taiwanese differ in perceptions of threat from the PRC, which in turn affects the policy responses they prefer. People who perceive high levels of threat from Beijing prefer resistance through balancing with other major powers against Beijing, while those with low levels of threat prefer to cooperate and bandwagon with Beijing. Respondents holding moderate threat perceptions do not see the PRC as an imminent danger but foresee potential problems lean toward hedging, where simultaneous cooperation with Beijing and others seek to keep options open. Others seeing serious but temporary threats from Beijing opt for isolation. Such variegated responses to threats of punishment are consistent with past research reporting differentiated public responses to sanctions.
As coercion from Beijing persists and grows over time, perceptions of threat from the PRC increased among the Taiwanese population and support for resistance through balancing rises accordingly. Concurrently, trust in Beijing's ability and willingness to commit to promises of restraint fell. More Taiwanese became more skeptical of the PRC's ‘One Country, Two Systems’ formula which ostensibly promises the maintenance of Taiwan's ‘capitalist system’ so long as it is part of the PRC. Taiwanese responses run counter to expectations about capitulation to coercion given stark capability asymmetries as well as assertions that coercion triggers a straightforward ‘rally-around-the-flag’ effect.
Despite the extensive literature on coercion, coercive diplomacy, and statecraft, the effects of coercive diplomacy on public opinion remain under-studied. Coercive diplomacy is not just the compellent use of force and violence, where existing studies suggest a strong public ability to withstand physical punishment such as heavy aerial bombardment and even firebombing (Pape, Reference Pape1996; Biddle, Reference Biddle2019). Much of the literature on coercive diplomacy focuses on signaling and pressure among elites or the state in the aggregate rather than public responses to varying levels of coercive threats. Where publics feature, they tend to be part of audience cost models that try to examine public restraints on states and elites that initiate coercion (Slantchev, Reference Slantchev2005; Thomson, Reference Thomson2016). Newer studies that look at public opinion in the face of coercive threats tend to use survey experiments rather than the effects of actual coercion on the public (e.g., Wang and Eldemerdash, Reference Wang and Eldemerdashforthcoming). We consider public responses to threats of force and the non-violent imposition of costs, such as economic punishment and diplomatic isolation.
This piece tests several common claims about coercion and public opinion, noting that public opinion on any issue – including responses to coercion – often varies across a population (Frye, Reference Frye2019; Gueorguiev et al., Reference Gueorguiev, McDowell and Steinberg2020). We evaluate the independent effects of threat perceptions toward the coercer, controlling for the influence of identity and party affiliation, which often account for rallying-around-the-flag effects and public pressure for accommodating coercer demands. We also control for public confidence in self-defense, which may affect preferences, separate from perceived threat. If identity, party affiliation, and public confidence are the main drivers of public opinion toward coercion, perceptions of threat should have little independent effect. In Taiwan's case, identity is especially tied to coercive threats since Beijing claims the island and its people in their entirety.
Taiwan, of course, presents a particularly useful case for studies of coercive diplomacy and its consequences. The cross-Strait relationship is a key potential flashpoint for major power conflict that threatens to embroil at least the USA, PRC, and Japan (Kastner, Reference Kastner2018). Conceptually, the focus existing literature on coercive diplomacy tends to focus on elite or aggregated state responses (George, Reference George1991; Markwica, Reference Markwica2018). Data from the Taiwan case enable an extension of this research to the individual-level, which is particularly helpful for understanding the drivers of public tenacity and debility in the face of threat. Popular political will may be critical for democracies like Taiwan, where leaders rely on voter support, but can also have non-trivial effects for authoritarian regimes given the general importance of public support for all leaders. Since Taiwan appears to be an open-and-shut case for coercive diplomacy given apparently overwhelming odds, explaining its resolve and defiance may shed light on how targets of coercion decide between fight and flight.
Public Taiwanese responses to growing PRC threats further suggest that there should be little concern over Taiwan being a ‘troublemaker’ in cross-Strait relations and that Beijing deploys coercion at its own peril. That PRC threats resulted in rising public support for external balancing rather than retaliation or immediate independence indicates a strong status quo preference among the Taiwanese public. Separate experimental survey research about growing support for defending Taiwan in the event of a PRC attack reinforces this impression (Wang and Eldemerdash, Reference Wang and Eldemerdashforthcoming). These observations imply that political parties have incentives to demonstrate resolve in maintaining Taiwan's de facto independent status and developing defensive capabilities rather than to accommodate Beijing or seek formal independence immediately. They further suggest that PRC threats may be more effective in promoting Beijing's short-term interests over its longer-term aims by deterring de jure independence in Taiwan rather than advancing unification.
Of course, Taiwan's strengthening pro status quo preferences in the face of Beijing's coercive diplomacy may change if PRC threats become more severe. However, the direction of such change is indeterminate from our data and can only be speculative at present. Further, our analysis indicates that even if Chinese coercive diplomacy currently deters or discourages formal Taiwan independence, it is unlikely to resolve differences in Beijing's favor unless the PRC is ready to bear the costs and risks associated with extensive escalation. This disjuncture between the compellent and deterrent effects of coercive diplomacy may represent a more general pattern in international politics, but requires further exploration incorporating cases beyond Taiwan.
Using individual-level data from the multi-year Taiwan National Security Survey (TNSS), this piece measures and analyzes the effects of coercive diplomacy through threat perceptions toward Beijing in Taiwan as well as policy preferences among Taiwan's public (Niou, Reference Niou2020). We use the years 2016, 2019, and 2020 as they provide comparability on the questions asked. They also represent a period where Beijing's coercive diplomacy toward Taiwan both increases in intensity and expands in scope across different issue areas. Based on respondents' answers to the TNSS survey, we formulate four strategic policy options that vary in resistance. The strongest resistance strategy is to align with the USA and Japan, the other major powers active in the region and Beijing's putative rivals, against the PRC, which we classify as ‘balancing.’ A more moderate response from the TNSS is to seek alignment with neither the PRC nor the other major powers but to maintain a distance from all sides, a move we term ‘isolation.’ ‘Hedging’ is when there is a preference to cooperate with all major power actors to open options, preserving possibilities for isolation, bandwagoning, and balancing. The most conciliatory option is to ‘bandwagon’ with the largest threat through cooperation and compliance in the hope of not triggering the latter's ire and any resulting punishment.
After proposing the array of strategies, we provide three indicators to explain respondents' strategic choices. The first indicator is the degree to which respondents see enmity as existing between two sides of the Taiwan StraitQ. We expect that higher perceived threat should make respondents more likely to balance or hedge such a threat. The second indicator is the degree to which respondents believe that economic cooperation with the PRC will result in Beijing using such exchanges to pressure Taiwan politically. We anticipate that respondents are less likely to bandwagon with China if they believe that China would exploit Taiwan's economic dependence on China to coercive Taiwan into political concession. Relatedly, our final indicator is the degree to which respondents believe that economic ties with the PRC should be strengthened or not. Respondents may be more likely to choose the bandwagon strategy if they believe that Taiwan should strengthen its economic ties with China.
The rest of this paper is organized as follows. In the next section, we discuss elaborately on a theory that aims to explain the relationships among threat, coercion, and resistance. In the third section, we briefly describe the evaluation of cross-Strait relations in the first term of President Tsai In-wen in Taiwan. The fourth section includes our research design and empirical findings. Specifically, we show that while there is some ambivalence among the public, increasing PRC pressure over time creates more suspicion toward cross-Strait cooperation and more desire for resistance. Such patterns indicate that Beijing's threats and overwhelming material advantages entrench Taiwanese desire for both separateness and an avoidance of escalation. The final section concludes by highlighting the broader implications of our study for understanding cross-Strait relations, Taiwan politics, and coercive diplomacy in international relations.
A theory on strategic responses to coercive diplomacy
Threat and statecraft
Purposeful deployment of threats, punishment, and reactions to these moves are unsurprisingly well-studied topics in the international security, strategic studies, and foreign policy analysis literature, given their prevalence in the practice of international politics. Work on coercive diplomacy and economic statecraft examines how actors calibrate threat, assurance, credibility, commitment, and perceptions to shape the behavior of others (George, Reference George1991; Christensen, Reference Christensen2011; Markwica, Reference Markwica2018). Deterrence, for instance, seeks to find an appropriate mix of credible assurances and threats to incentivize adherence to the status quo (Schelling, Reference Schelling1966; Lantis, Reference Lantis2016). Compellence, such as sanctions and other punishment, imposes costs on a target that an actor commits to lifting should the former alter behavior (Schelling, Reference Schelling1966). Other forms of coercive diplomacy aim to change a target's behavior by threatening the imposition of cost and promising restraint should such adjustments occur. Thus, attempts at compellence may not involve military forces. For instance, PRC efforts to discourage formal Taiwanese independence through the threat and imposition of costs are an example of deterrence, while using the same to promote unification exemplifies compellence. The two strategies are not necessarily mutually exclusive, although PRC attempts to make Taipei accept its more restrictive version of the ‘92 Consensus’ from 2016 suggest a shift to compelling unification rather than just deterring independence. That Beijing employs coercive diplomacy toward Taiwan indicates some confidence in its efficacy.
Whether the topic at hand is economic statecraft, sanctions, threats of force, deterrence, or crisis management, there tends to be an overwhelming focus on elites and their reactions. This is often the case with the many detailed studies on the use and mechanics of coercion in various contexts, ranging from the Korean War through threats of nuclear weapon use and the use of sanctions as tools for coercive diplomacy (George and Simons, Reference George and Simons1994; Art and Cronin, Reference Art and Cronin2003). Work in political psychology examining decision-makers facing, issuing, and responding to threats likewise concentrates on elite decision-makers (Levy, Reference Levy2008; Markwica, Reference Markwica2018). The roles of the public in many of these accounts tend to be in terms of how elite decision-makers perceive, interpret, or expect public reactions to play out in response to some receipt or issuing of threat (Byman and Waxman, Reference Byman and Waxman1999; Schultz, Reference Schultz2001). Another common approach is to examine how elites mobilize public opinion in the face of threat or to extend some threat to another actor (Onderco, Reference Onderco2017; MacDonald, Reference MacDonald2020). How publics actually understand and act around such threats receive less attention.
More conceptual work examining the dynamics of coercive bargaining, often employing game theoretic and statistical methods, similarly focus their attention on elites or else states (and other entities) as aggregated, corporate actors (Zeng, Reference Zeng2010; Sechser and Fuhrmann, Reference Sechser and Fuhrmann2016). Much as this scholarship draws needed attention to the more fundamental logics associated with threats and responses to them, they tend to pay less attention to the specificities of public opinion in these processes except in abstraction. Such work draws on foundations laid by pioneering scholars like Thomas Schelling and David Baldwin who seek to explain the use and response to threats and coercion to shape the actions of others (Schelling, Reference Schelling1966; Baldwin, Reference Baldwin1985). Public opinion tends to play the role of positive or negative incentives and sometimes signals in efforts to model and approximate state or decision-maker behavior (Gates and Humes, Reference Gates and Humes1997; Schultz, Reference Schultz2001). These perspectives are useful but provide less direct observation on what drives public opinion and how it evolves when facing or wielding threats.
The debate around ‘audience costs’ tries to address the roles of publics play in a more sustained and systematic manner in relation to the issuance of threats, but generally examines how public perceptions and preferences react to elites' actions. Much of the work in this vein surrounds whether, when, and how much audiences – including the public but also others – will punish or reward an actor for delivering or reneging on a promise, as well as if such incentives matter for decision-makers (Krustev and Morgan, Reference Krustev and Morgan2011; Zhang, Reference Zhang2019; Walentek et al., Reference Walentek, Broere, Cinelli, Dekker and Haslbeck2021). Public preferences in these studies tend to be taken as exogenous effects, with less attention paid to how public reactions develop and change. More dynamic approaches to understanding public preferences and what shapes them can augment existing studies by shedding light on specific conditions under which public opinion responds to threats. Such efforts aid in understanding how the public constrains and enables decision-makers in coercive environments.
Newer scholarship uses survey experiments to understand the effects of coercion on a population and broaden the discussion of coercion and threats beyond elites. Such research tends to be historically less common given the challenges of collecting public opinion during crises or episodes where a target faces coercive threats. Nonetheless, extant work suggests that the purposeful deployment of threats by another actor does not necessarily stiffen public resistance. Frye (Reference Frye2019) argues, based on a survey experimental approach, that as publics learn about sanctions, they tend to react strongly to the reasons why the sanction were put in place rather than the existence of sanctions themselves. Gueorguiev et al. (Reference Gueorguiev, McDowell and Steinberg2020) contend that coercion provides information about an issue to publics, which mobilizes public opposition to the sanctions imposed because of a desire to prevent the coercer from gaining an advantage against their state.
Coercion, resolve, and resistance
Our paper adds to existing research by looking at the case of Taiwan, which is the target of expanding as well as intensifying direct and indirect coercion from the PRC since the election of President Tsai Ing-Wen in 2016. Beijing's threats from 2016 to 2020 involved the exacting and promise of increasing costs in the same issue domain (intensification) and across new domains (expansion). These moves seek to force Taiwan to accept a version of the so-called ‘92 Consensus’ that sees Taiwan as part of China represented by the PRC (Chen and Cohen, Reference Chen and Cohen2019). Starting with the suspension of group tours from the Mainland, Beijing later included a halt on the import of Taiwanese pineapples as part of its economic coercion efforts (Cole, Reference Cole2020). Diplomatic isolation attempts grew with Taiwan's exclusion from the World Health Assembly and International Civil Aviation Organization as well as efforts to get states to end official ties with Taipei and have corporations list Taiwan as part of China on their websites (Cole, Reference Cole2020).
PRC actions extended into other domains, such as the rendition of Taiwan citizens from third countries under fraud allegations, cyberattacks, disinformation campaigns, election interference, as well as increasing military aviation and naval activity near Taiwan (Cole, Reference Cole2020). Beijing's crackdown on pro-democracy protests in Hong Kong and its substantive revisions to interpretations of Hong Kong's Basic Law further underlined a willingness to renege on promises made under the ‘One Country, Two Systems’ formula. ‘One Country, Two Systems’ as originally conceived by Deng Xiaoping ostensibly commits Beijing to preserving Hong Kong's capitalist system by granting significant levels of autonomy in exchange for accepting PRC rule. Beijing extends an offer of the ‘One Country, Two Systems’ to Taiwan by supposedly guaranteeing the latter's way of life under CCP rule (Xi, Reference Xi2019). Beijing's actions and pronouncements are observable to everyone on Taiwan via the media and official statements, enabling TNSS to collect public responses to coercion.
Taiwan is particularly apt for understanding the effects of coercion on public opinion and their policy implications. Taiwan maintains an open society and political system which is familiar with robust public opinion research on political matters, including cross-Strait relations. Consequently, there is a wide range of public opinion research – including the TNSS – against which to check general patterns in public preferences. Our research on Taiwan can highlight key patterns in public opinion responses to coercion in ways that lend itself to the making of causal inferences. Past efforts tend to highlight the hypotheticals of PRC coercion or use of force rather than the actual effects of changing coercion levels considered in this study (Wang, Reference Wang2013). How the Taiwanese public responds to threats from the PRC is helpful in addressing the perennial debate over whether the population has the political will to resist pressure and even attacks (Hickey, Reference Hickey2020; The Guardian, 2021). Understanding this debate can help inform USA and Japanese policies toward Taiwan as well as its own efforts at defense.
Research on airpower suggests that the use of compellent force can encourage resolve and resistance from a targeted population (Byman and Waxman, Reference Byman and Waxman2000). The Blitz during World War II reinforced the British public's determination, just as the firebombing of cities did not break the German and Japanese will to fight (Pape, Reference Pape1993). Less clear is whether coercive diplomacy through the deployment of threats has similar effects since such promises of punishment play more on uncertainty and risk aversion rather than the imposition of actual costs that can trigger public anger and desire for revenge. More conventional approaches to coercive diplomacy tend to emphasize the role of elite decision-makers operating in almost clinical environments, at most taking in and responding to public opinion that is exogenously given. By trying to further account for how publics empirically respond to coercion, we hope to move a step toward a more endogenous explanation of coercion in international politics and its effects.
Existing literature seems ambivalent over the role that coercive diplomacy and even coercion play in shaping public responses, however. Classic works on systemic balancing by scholars such as Waltz, Powell, and Schweller argue that actors tend to bandwagon against a state under pressure to enjoy the spoils of victory but are silent about options for the target state (Waltz, Reference Waltz1979; Schweller, Reference Schweller1994; Powell, Reference Powell1999). For our purposes, this means Taiwan has few options for balancing against PRC pressure since other actors would be eager to divvy up the benefits from a Chinese takeover with Beijing. Walt, in comparison, argues that states balance against threat, implying that Taiwan would seek out partners facing similar pressure from Beijing to resist changes to the status quo (Walt, Reference Walt1987). None of these approaches speak to reactions outside of the balancing–bandwagoning dichotomy, although complications and cost may cause under-balancing, where balancing actions are anemic (Schweller, Reference Schweller2004).
Based on these discussions, we identify four strategies public may prefer when responding to coercion. These strategies include descending strength of resolve and resistance, balancing against the coercer, isolation, hedging, and bandwagoning with the coercer. Balancing involves the target working actively with other actors, notably major rivals of the coercing state, to counter the threat from the coercer, including using force for defense if necessary (He, Reference He2009). Isolation is a choice to disengage with all actors, including the coercing state, with the aim of averting friction and confrontation, a policy that Paul Schroeder also describes as ‘hiding’ (Schroeder, Reference Schroeder1994). Hedging is the effort to work with all actors simultaneously, including the coercer, with the expectation that such engagement can dilute animosity while preserving options for resistance in conjunction with some set of actors should relations with the coercing state deteriorate (Medeiros, Reference Medeiros2005). Bandwagoning entails accommodating and cooperating with the threat sending state in the expectation that such action will prompt restraint (Schweller, Reference Schweller1994).
To be clear, we do not intend to invent new theories of balancing, bandwagoning, isolation, hedging, or other responses to coercive threats. We seek to examine public preferences and responses to coercive diplomacy with the goal of explaining these reactions and discussing their implications. Ordinary citizens may not have a developed sense of the precise strategy to take in response to threats but do express general preferences on how to react based on perceptions of the situation. Such perspectives create incentives and constraints for policymakers who must win public support and avoid public ire. In this vein, Table 1 displays a two-dimensional conceptualization for the four strategies that respondents in Taiwan commonly prefer and for which we try to discern patterns in public support.
Table 1. Conceptualization of Taiwanese citizens' strategic choices
From the typology in Table 1, we advance two hypotheses to investigate the determinants of citizens' strategic choices in the target country. First, we hypothesize that shaping strategic choices of citizens in the target country are perceptions about the coercer's intention as well as an evaluation on their country's relation with the coercer. Citizens in a target country who think that the coercer will increase pressure on their country in the future will be less likely to bandwagon with coercer and more likely balance against it by allying with other countries. Note that balancing aims to maintain the status quo rather than escalate. Conversely, citizens would not ally with third countries if they perceive a low level of threat from a coercer and would instead bandwagon. We further argue that the public is most likely to choose the hedging strategy when they perceive that the level of hostility between their country and the coercer is moderate. Citizens who believe threats to be transient, even if serious, seek isolation to avert pressure and avoid escalation by remaining low-key. Therefore, regardless of other traits and markers of identity, we propose:
H1a: Citizens in the target country are less likely to bandwagon with the coercing country if they perceive more threat from the coercer.
H1b: Citizens in the target country are more likely to balance with the coercing country by allying with other countries if they perceive more threat from the coercer.
H2: Citizens are most likely to choose to hedge when they perceive that the hostility between their country and the coercing country is moderate.
If citizens in a target state prefer bandwagoning given greater expectations of threat or balancing if they have lower threat perceptions, it becomes possible to challenge our claims. Should target state citizens prefer strategies other than hedging when perceiving moderate levels of threat, our claims again face dispute. We discuss different levels of perceived threat and their measurement in the section on independent variables.
We use Taiwan to examine our hypotheses because Taiwan experienced increasing political and economic coercion from China, especially after Tsai In-wen won the presidential election in 2016. PRC coercion makes Taiwan an ideal case to test our claims.
Threat and tension across the Taiwan Strait, 2016–2020
Taiwan has a tumultuous relationship with the PRC. Taiwan's current system of government evolved from one imposed by the Kuomintang (KMT, or Chinese Nationalist Party) after they relocated to the Taiwan following defeat to the CCP in the Chinese Civil War (Rigger, Reference Rigger1999, pp. 1–102). The end of KMT-enforced martial law and democratization in the late 1980s witnessed Beijing trying to create conditions supportive of the electoral chances of China-friendly politicians and undermining their opponents, including through large-scale missile tests in 1995–1996 (Rigger, Reference Rigger1999, pp. 103–193). Coming to office after 8 years of fraught ties with a PRC-cautious Democratic Progressive Party (DPP) administration, President Ma Ying-Jeou's KMT administration (2008–2016) sought to strengthen ties with Beijing by enhancing economic exchange and integration via a series of agreements (Cabestan and deLisle, Reference Cabestan and deLisle2014). However, a Taiwanese public wary of Chinese domination staged a series of mass protests culminating in 2014 Sunflower Movement, delivering the KMT defeats in both the 2014 local elections as well as the 2016 presidential and legislative elections (Cole, Reference Cole2017).
Since the election of President Tsai Ing-Wen in 2016, Beijing's pressure on Taiwan steadily mounted. Beijing demanded public acceptance of the ‘92 Consensus’ and, by extension, the PRC's interpretation of ‘one China,’ soon after Tsai's election victory. Seeking to address domestic concerns while mollifying Beijing, Tsai avoided the terms ‘92 Consensus’ and ‘One China’ in her inaugural speech. However, she and her administration acknowledged the ‘historical fact’ of a cross-Strait meeting in 1992, indicating that they would honor the 23 agreements inherited from the preceding Ma administration on that basis. Beijing found this position unacceptable and a series of broadening and intensifying coercive measures from Beijing soon followed, all with the aim of forcing Taipei to accept its preferences. Together, such threats targeted everything from commercial activity, to diplomacy, news and social media, and even Taiwan's domestic politics in line with the media, psychological, and legal components of a supposed ‘three warfares’ approach Beijing adopts toward Taiwan (Kania, Reference Kania2016).
Beijing's threats grew more tangible once the Tsai administration took office. On 24 May 2016, 4 days after Tsai's inauguration, the China's National Bureau for Quality Control and Inspection issued an alert requiring enhanced inspections and testing of all oranges imported from Taiwan. Beijing suspended communication between Taiwan's Strait Exchange Foundation (SEF) and China's Association for Relations Across the Taiwan Strait (ARATS), the two semi-official bodies responsible for cross-Strait ties, permitting only emergency contacts at the lowest level. From 20 May to 23 August 2016, the number of Chinese tourists visiting Taiwan fell by 22.3% relative to the same period in 2015 (Taiwan Tourism Bureau, 2021). Moreover, Beijing sought to bypass the Tsai government through direct exchanges with businesses, students, local politicians, technology, fishermen and farmers' associations, and retired military personnel, averaging about 30 meetings a month from 2016 to 2020 (Mainland Affairs Council, 2020). Additionally, Beijing sought to entice younger Taiwanese to live and work in China and assimilate them as a means to counteract their ‘natural inclination toward independence.’ The TAO established multiple platforms to encourage Taiwanese entrepreneurship, job seeking, and university applications across municipalities and provinces in China (Smith, Reference Smith2017). In return for providing their personal data, Taiwanese would also be granted a digital residence card that functioned similarly to a Chinese identity card and allowed them to enjoy more of the benefits usually extended to Chinese nationals (Spencer, Reference Spencer2018).
Alongside economic measures, China sought to increase diplomatic pressure on Taiwan. The Chinese government pressured governments with diplomatic ties with Taiwan to switch recognition to Beijing. Since 2016, Kiribati, the Solomon Islands, the Dominican Republic, Burkina Faso, El Salvador, Panama, and São Tomé and Príncipe dropped official ties with Taipei in favor of formal relations with Beijing. There are currently 14 countries and the Vatican with formal diplomatic relations with Taiwan, following Taipei's unseating from the United Nations in 1971. Shattuck noted that this round of de-recognition aimed to push the Taipei ‘into accepting the “92 Consensus” or adopting a more pro-China policy’ (Shattuck, Reference Shattuck2020). In response to PRC compellence, Tsai called for the ‘redoubling’ of efforts to develop substantive international economic and security partnerships with like-minded countries to win acknowledgement and support (Office of the President, 2018).
When Tsai's DPP suffered a major setback in the November 2018 local elections, Beijing believed this vindicated its strategy combining diplomatic pressure on the Tsai administration while enticing Taiwanese to vote against the DPP. With the pro-China KMT and an independent now controlling 15 out of 22 metropolitan areas and counties, including the DPP stronghold of Kaohsiung, Beijing stepped up attempts to bypass the Taiwanese government by expanding direct ties with non-DPP districts. Seeking to further rein in Taiwan, Chinese leader Xi Jinping issued a statement in January 2019, insisting that China should ‘explore the “two system” option for Taiwan, and enrich the implementation of peaceful unification (Xi, Reference Xi2019). Xi's statement resulted in alarm on Taiwan since he seemed to define the ‘92 Consensus’ as ‘One Country, Two Systems,’ which he was eager to impose.
Many Taiwanese took Xi's statement to be a warning about China's unification agenda. They were particularly attentive to the linking of ‘One Country, Two Systems’ to Hong Kong and watched developments there closely for implications on Taiwan. The Hong Kong government's introduction of an extradition law in April 2019, which led to massive demonstrations, became a critical test for whether Beijing honored its promises under ‘One Country, Two Systems.’ Protests soon included demands for the universal suffrage Beijing promised under the 1984 Sino-British Joint Declaration and Basic Law. Widespread police brutality against protesters signified an erosion of the freedoms guaranteed under ‘One Country, Two Systems,’ culminating in a harsh, expansive, and vaguely worded National Security Law (NSL) in June 2020 (Young, Reference Young2021). The NSL brought a crackdown on activists, journalists, elected politicians, and protesters, casting serious doubt over Chinese commitments for many Taiwanese.
Sabre-rattling formed another component of PRC threats toward Taiwan between 2016 and 2020. Since 2013, Chinese military aircraft conducted more than 4400 patrols into Air Defense Identification Zones (ADIZ) maintained by Japan, South Korea, and Taiwan, with circumnavigational flights around Taiwan demonstrating growing sophistication. In addition to training pilots for war and probing defenses, China's regular flights into ADIZs remind target countries of its ability to exact costs militarily. While most of PLA flights occurred at edges of Taiwan's ADIZ, China termed these missions toward Taiwan ‘island encirclement patrols,’ designed to deter what they saw as Taiwan independence forces despite the fact that such actions increased after Taipei refused Beijing's exact formulation for cross-Strait relations (Trent, Reference Trent2020).
Finally, Beijing concurrently sought to discredit and stoke public distrust in the Tsai administration and Taiwan's state institutions. Between 2016 and 2020, Taiwanese were deported to China from locations as varied as Armenia, Kenya, Malaysia, Spain, Thailand, and Vietnam for alleged telephone scamming, possibly to demonstrate the Taiwan's inability to protect citizens overseas (Matthew, Reference Matthew2017). Following a 2018 typhoon in Osaka, PRC-related sources and proxies spread a rumor that Taiwanese stranded at Kansai airport had to be rescued by buses chartered by the Chinese consulate, even though local authorities shut down transport to and from the airport. This disinformation unleashed public anger toward the Tsai administration on mainstream and social media for alleged ‘incompetence,’ resulting in the director-general of Taiwan's Osaka Office, Su Chii-Cherng, taking his own life to ‘bear responsibility’ (Wang, Reference Wang2020). Such Chinese-backed disinformation persisted in the lead-up to Taiwan's local elections in 2018 and national elections in 2020, seeking to cast doubt on the electoral process while smearing DPP candidates (Monaco et al., Reference Monaco, Smith and Studdart2020). For instance, Liqiang Wang, an alleged Chinese spy, disclosed details about China's efforts to influence Taiwan's presidential election in 2020 when he fled to Australia (Aspinwall, Reference Aspinwall2019).
Mounting Chinese coercion toward Taiwan occurred against a backdrop of worsening Beijing–Washington ties. The Obama administration's final year saw sharpening superpower differences over expansive PRC claims over the South China Sea, especially as an Arbitral Tribunal ruled that most features in those waters could not generate the territorial waters China asserted (Congressional Research Service, 2016). Donald Trump came to office calling for the punishment of the PRC for unfair trade practices and having taken a congratulatory call from Tsai Ing-Wen following his election, despite the USA having no official diplomatic relations with Taiwan. Beijing's efforts to seek compromise met with limited success. The Trump administration imposed tariffs on PRC goods worth over US$200 billion, while deepening and expanding contact as well as support for Taiwan (Lawder, Reference Lawder2019).
The above developments resulted in further rounds of reactions and counter-reactions among various powers. PRC retaliation came as trade penalties, continuing pressure on Taiwan, persistent contestation of waters disputed with Japan and Korea, as well as the reclaiming and militarization of features in the South China Sea (Chhabra and Hass, Reference Chhabra and Hass2019). The Trump administration sold yet more advanced weapons systems to Taiwan even as it challenged China's South China Sea claims through more frequent, publicly announced Freedom of Navigation Operations in those waters amid Chinese protests (Blanchard, Reference Blanchard2020; Lubold and Youssef, Reference Lubold and Youssef2020). Under Abe Shinzo's premiership, Japan as well became more wary of China and less willing to back down over disputes even as Japanese public opinion toward Beijing hardened (Silver et al., Reference Silver, Devlin and Huang2020). Public views of China grew more negative in South Korea with Beijing's punishment of Korean companies and performers following the Park Geun-Hye administration's decision to deploy an advanced missile defense system to protect against North Korea.
Public opinion in Taiwan likewise became more skeptical toward China between 2016 and 2020. According to the surveys released by Taiwan's Mainland Affairs Council (MAC), respondents who disapproved of ‘China PLA's intrusion and encirclement of Taiwan's nearby water and space’ increased from 82% in May 2019 to 91.1% in June 2020. Respondents who found ‘Xi's rejection of renouncing to use forces against Taiwan and his repeated military threats’ unacceptable rose from 77.2% in January 2018 to more than 89% in October 2019. Surveys between November 2018 and November 2020 showed more than 70% of respondents supporting the view that the Taiwanese government should ‘cooperate with like-minded countries to preserve regional peace’ in the face of growing PRC pressure. As Beijing's crackdown in Hong Kong intensified, respondents opposing China's ‘One Country, Two Systems’ formulation grew from 75.4% in January 2019 to 83.6% in May, 89.3% in October, and 90% in March 2020 (Mainland Affairs Council, 2021).
Research design
Data
To investigate Taiwanese public responses to the PRC's coercion and test our hypotheses, we analyze the data collected by the TNSS in 2016, 2019, and 2020. All waves of surveys were conducted by the Election Study Center at National Chengchi University via telephone interviews. In this article, we only use the data collected in 2016, 2019, and 2020 because the question on allying with the USA is only available in those three waves (and not asked in the TNSS 2017), rather than look at all 13 waves of TNSS data. Analyzing these three waves of data can help us understand how Taiwanese citizens respond to the increasing China's threat to Taiwan during the first term of Tsai In-wen's presidency. In particular, the most recent 2020 wave of TNSS was conducted in late October 2020, 4 months after Hong Kong's adoption of the NSL in response to the territory's Anti-Extradition Law Amendment Bill (Anti-ELAB) Movement. To the extent that the NSL undermines or suppresses Hong Kong's autonomy under the framework of ‘One Country, Two Systems,’ it affects Taiwanese views of Chinese credibility over any cross-Strait arrangements and any future with the PRC. Examining these three waves of data enables a more comprehensive understanding of Taiwanese citizens' perceptions of strategic choices given increasing PRC threats.
Dependent variable
The dependent variable in our empirical analysis is respondents' strategy preference in cross-Strait relations. Based on our conceptualization discussed in Table 1, we use two questions in the TNSS to construct an array of four strategies to respond to China's threat. Hsueh (Reference Hsueh2022) also uses this operationalization when analyzing the TNSS data. First, the TNSS asks respondents' support for Taiwan's economic engagements with China via the following question: ‘In Taiwan, some believe we should strengthen economic and trade relations with Mainland China, and some believe we should weaken these relations. What do you think is the right opinion?’ Second, the TNSS asks respondents' support for Taiwan's alliance with the USA and Japan to counter China: ‘Some people argue, “Taiwan should strengthen ties with the United States and Japan to counter mainland China”. Do you support this statement?’
Taking respondents' answer to both questions into consideration, Table 2 represents the two-dimensional conceptualization of public opinion regarding Taiwan's relations with China. In the upper left cell, respondents neither support Taiwan strengthening economic relations with China nor support allying more closely with the USA and Japan to counter China. We consider this combination of preferences to be ‘isolation’ as it indicates a desire to not engage both China and the USA. The second strategy is to ‘bandwagon’ with China by strengthening cross-Strait economic ties with China but not allying with the USA and Japan to counter China. This approach follows the call by some political observers and scholars for Taiwan to deepen its relationship with China to take advantage of China's domestic market and economic power. A third strategy is to ‘balance’ against China by reducing economic exposure to China while allying with the USA and Japan against Beijing. The last strategy is for Taiwan to ‘hedge’ by deepening economic engagement with China even as it seeks to develop ties with the USA and Japan in preparation to counter Beijing. Some observers claim that ‘hedging’ maximizes Taiwan's interests.
Table 2. Conceptualization of Taiwanese citizens' strategic choices
Figure 1 shows the distribution of support for the four strategies across three waves of the TNSS. There are significant changes in support for these four strategic choices across survey waves. Specifically, supporters of bandwagoning decreased from 48.51 to 31.72% between 2016 and 2020, while supporters of balancing grew from 18.89 to 39.75% in the same period. The shift may be a result of the increasing threat from China to Taiwan, especially after the 2019–20 Hong Kong protests. Additionally, more than 20% of respondents supported hedging between 2016 and 2020. In the following analysis, we further investigate correlates with those four choices.
Figure 1. Strategic choices among respondents.
Independent variables
China's level of threat to Taiwan induces different strategic choices among Taiwanese citizens. When perceptions of China's threat are low, Taiwanese may choose to bandwagon with China for economic benefit. In contrast, Taiwanese may prefer to ally with the USA and Japan to balance if they think China presents a high degree of threat. We use two questions of the TNSS to measure respondents' evaluation of threats from China. The first question asks respondents to evaluate the propensity of China to use economic coercion toward Taiwan: ‘Some say that if Taiwan's economy relies too heavily on Mainland China, then the Mainland will use the economy to coerce Taiwan into making political concessions in the future. Do you agree with this?’ Respondents were asked to indicate whether they strongly disagree, disagree, agree, or strongly agree on this statement.
We use a four-point Likert scale to indicate a higher level of respondents' agreement with this statement to operationalize their perceptions of China's potential use of economic coercion toward Taiwan. Figure 2 displays the distribution of respondents' perceptions across three waves of the TNSS. More than 60% of respondents in each survey believe that China would use economic coercion against Taiwan if the latter is too economically dependent on China.
Figure 2. Perceived propensity of China to use economic coercion toward Taiwan.
The second variable on China's threat is respondents' perception of hostility in cross-Strait relations. The TNSS asks respondents to rate the Mainland–Taiwan relationship on a 0–10 scale, with 0 being very antagonistic and 10 being very peaceful. We reverse this scale to make a higher number indicative of more hostility between Taiwan and China. Figure 3 illustrates that Taiwanese perceive rising hostility in cross-Strait relations from 2016 to 2020. We regard perceived hostility as a proxy of China's threat toward Taiwan given Beijing's growing political and economic coercion since the start of the Tsai administration in 2016.
Figure 3. Perceived hostility in cross-Strait relations.
Results
To test our hypotheses, we employ multinomial logit models to estimate covariates of Taiwanese citizens' preferences on strategy. Multinomial logit models enable us to estimate the probability of strategic choices in various categories compared to the probability of strategic choice in the reference category (i.e., ‘hedge’), conditional on covariates. We include a set of variables on respondents' demographic traits and political attitudes to control for confounding effects on the relationship between China's threat and preferred strategies. We include respondents' age, gender, education, and partisanship. We also control for their identity, position on whether Taiwan should unite with China or declare independence, and whether the USA would help Taiwan if China attacks Taiwan's independence. Table A1 describes the operationalization of these variables.
Table 3 reports the estimation results of multinomial logit models for the TNSS surveys in 2016, 2019, and 2020. To ease interpretation, we report the coefficients of our multinomial logit models as relative risk ratios (RRRs), a concept similar to odds ratios. A coefficient large than 1 indicates a positive relationship, whereas a coefficient smaller than 1 a negative relationship.
Table 3. Multinomial logit estimates for correlates of strategic choices
Note: The baseline category is Hedge. Standard errors in brackets. Coefficients presented as relative risk ratios (RRRs). *P < 0.1, **P < 0.05, ***P < 0.01.
Results are all in reference to the baseline category of hedging. Model 1 suggests that respondents who are more concerned about China's propensity to use economic coercion toward Taiwan are more likely to prefer isolation or balancing. Those less concerned about China's economic coercion are less likely to choose bandwagoning, relative to hedging. The pattern remains in models 2 and 3, especially for those choosing bandwagon or balance rather than hedge. In other words, respondents who perceive a higher propensity of Chinese economical coercion toward Taiwan choose isolation or balancing as strategies over bandwagoning or hedging.
To further clarify the relationship between the perceived propensity of Chinese economic coercion and strategic choices, we calculate the marginal effects and visualize those effects. As shown in Figure 4, the probability of choosing bandwagoning decreases with the perceived China's propensity to use economic coercion, while the probability of choosing balance increases with it. Figure 4 further suggests that the perceived propensity of China's economic coercion induces more support for the balance strategy in 2020. This result supports our second hypothesis, that ‘threat perceptions have a positive relationship with preferences for resistance to coercive diplomacy despite differences in material capacity’.
Figure 4. Marginal effects of perceived China's propensity of economic coercion.
The results of models 1–3 also indicate that respondents' perceptions of cross-Strait hostility are related to their strategic choices, but the relationship may be non-linear because most of the RRRs of hostility are smaller than 1 while the RRRs of hostility squared are always larger than 1. To further illustrate the effects of perceived hostility on strategic choices among Taiwanese citizens, we draw marginal effect plots for hostility (and its squared terms) based on the results of models 1–3. Figure 5 confirms that the relationship between perceived hostility and strategic choices are non-linear, especially for those who prefer balance and hedge.
Figure 5. Marginal effects of perceived cross-Strait hostility.
The inverted U-shaped line representing ‘hedging’ suggests that the support for this strategy is highest when Taiwanese perceive mid-ranging levels of cross-Strait hostility. They would be less likely to choose hedging as a strategy when perceived hostility is too low or too high. This result highlights the rationale behind hedging: there is no need to hedge when the cross-Strait hostility is low, and it is unrealistic to hedge when the cross-Strait hostility is high.
The non-linear relationship between perceived hostility and cross-Strait hostility also exists for ‘balancing’ in 2016 and 2020. Respondents' support for balancing is low when the perceived cross-Strait hostility is at middling levels in 2016 and 2019. However, this non-linear relationship between perceived hostility and the probability of choosing ‘balancing’ becomes more linear in 2020. This change may be a result of increased tension across the Taiwan Strait, which made some citizens who previously perceived moderate levels of cross-Strait hostility to see such animosity as more acute and thus become more supportive of ‘balancing.’
Figure 5 also suggests that support for ‘bandwagoning’ decreases as the perceived cross-Strait hostility increases. Nonetheless, those who perceived relatively high levels of hostility in 2020 are slightly more supportive of ‘bandwagoning’ than those who perceive moderate hostility. Even though a detailed analysis of this result is beyond the scope of this study, we suspect that behind such preferences may be some who are intimated by China's military capacity and become more dovish toward China. If true, this suggests that the intended effects of coercion to cow a target only operate on a segment of the target population.
Influences of the anti-ELAB movement on Taiwanese public opinion
Our claims focus on how coercion from China can affect Taiwanese citizens' strategic preferences toward the PRC, particularly how rising threat perceptions encourage more, not less, resistance. One factor seems to be developments in Hong Kong surrounding the 2019 Anti-ELAB Movement. The 2019 TNSS wave was conducted in January, before the start of the movement and could not capture Taiwanese responses to events in Hong Kong. The 2020 wave for TNSS was conducted in October and enabled investigation respondents' reactions to the Anti-ELAB Movement and its aftermath. Taiwanese citizens grew more suspicious of China as Beijing became increasingly implicated in heavy-handed intervention into Hong Kong's autonomy during the movement, undermining, curtailing, and eroding the territory's autonomy.
Included in TNSS 2020, is the following question: ‘Recently, the Chinese government passed a new Hong Kong National Security Law. With the ongoing development in Hong Kong, does this development make you more inclined to support Taiwanese Independence?’ Respondents were asked to choose from strongly disagree to strongly agree. Based on our argument, we expected respondents to become more resistant to PRC compellent threats. In other words, they would be less likely to bandwagon with China and more likely to choose balancing with USA and Japan against China. Results from our multinomial logit model, as shown in Table 4, confirm this conjecture and offer further supportive evidence to our hypotheses.
Table 4. Multinomial logit estimates for correlates of strategic choices
Note: The baseline category is Hedge. Standard errors in brackets. Coefficients presented as relative risk ratios (RRRs). *P < 0.1, **P < 0.05, ***P < 0.01.
Discussions and conclusions
Our study raises a number of implications for efforts to employ coercive diplomacy, especially when public opinion, preferences, and pressure come into the picture. Should publics play a non-trivial role in constraining policymakers, prompting certain courses of action, or even act as veto players, then how they read and respond to coercion and threats can matter greatly. Even if decisionmakers read the signals sent by a coercer perfectly and have a complete grasp of relative material capabilities, public perceptions and opinion can still shape their reactions (Kirshner, Reference Kirshner2000). This remains the case even if public opinion is split, which may spell restraint and caution in some instances and more opportunism, even assertiveness in others. So long as there is sensitivity to public opinion, such dynamics may compound situations where key decisionmakers vary in their preferences and even operate in non-democracies.
Variation in the effects of coercion across a population can blunt efforts at coercive diplomacy. Coercive signals may elicit resistance as much as they prompt compliance or capitulation, even as costs and risks associated with actual and threatened punishment rise. In political systems where decisionmakers are subject to public opinion, coercion may not result in the sort of clean, unambiguous accommodation a coercer prefers. Such differentiated effects of coercion persist as efforts to coerce expand in scope and intensity, possibly making coercion more costly – even leaving aside any resulting growth in the popularity of defiance. Competing public preferences may replicate in domestic politics the complications associated with trying to coerce alliance systems with diverse members, possibly making autocracies easier to coerce since efforts can be concentrated on fewer actors and public opinion is less salient (Christensen, Reference Christensen2011).
Another takeaway from our findings is that coercive diplomacy may be more effective when the coercer and target share relatively positive relations, with the effects of coercion diminishing as tensions rise and frictions grow over time. As PRC coercive measures continue and expand, there is a noticeable rise in public preferences for balancing along with expectations of economic coercion from Beijing from 2016 to 2019. Beijing had relatively warm relations with the preceding pro-China Ma administration despite its declining popularity during its second term, only implementing clear coercive measures once the Tsai administration took office. Our findings suggest that as coercive attempts persist and intensify, the public in the target may become disabused of goodwill from the coercer and lower expectations for reconciliation, learning to ‘price in’ any imposition of cost as well as expectations of uncertainty. Coercion by perceived ‘friends’ and ‘partners’ may ironically be more effective in getting a target to comply than perceived ‘rivals’ and ‘enemies,’ given expectations about the possibility for repairing ties and the likelihood for punishment.
Availability of partners that can support attempts at resistance may as well inform the willingness and ability of a target to withstand coercion. There was limited appetite for the go-it-alone ‘isolation’ strategy across the three TNSS waves we consider, given that it is consistently the least popular of the four strategies. A partial exception may be in 2016 where slightly more respondents preferred ‘isolation’ when expecting greater economic coercion and perceptions of hostility, possibly because of a belief that tensions would soon subside. This pattern disappears subsequently, likely as more respondents realize that PRC assertiveness is here to stay. ‘Balancing’ in the TNSS was notably in terms of cooperating with the USA and Japan to defy China at a time when both Washington and Tokyo were becoming increasingly bold in confronting Beijing across a range of issues, including support for Taiwan. Support for ‘balancing’ in the abstract or in the absence of options for partnerships is unknowable from the TNSS, although other surveys indicate some willingness by the Taiwanese public to resist Chinese aggression alone (Lin, Reference Lin2018; Wang and Eldemerdash, Reference Wang and Eldemerdashforthcoming).
In other words, the efficacy of coercive diplomacy and threats may decay over time as targets, or at least their populace, may become more accustomed to such situations. Lower expectations for cooperation encourage the targetted state to seek substitutes for what the coercer can offer. With every round of coercion or threat, publics in a target learn and ‘price in’ a higher discount rate on the returns from cooperation with the coercer even if power differentials are substantial. For a given coercer and target, the same form of coercive threat is therefore likely to elicit a lower level of compliance when used again. To achieve the same level of compliance, the coercer may have to engage in more heavy-handed action that comes with increasing cost and risk. Such considerations can complicate and even temper a coercer's willingness to be increasingly provocative and risk escalation should the target does not respond satisfactorily.
Beijing's coercive actions toward Taiwan appear to be making the latter's population warier of China, counteracting simultaneous attempts to entice Taiwanese to accept its rule or even just stop challenging PRC claims publicly. Such trends are persistent despite the PRC's longstanding claims over Taiwan, unwillingness to renounce force to bring the island and its population under Beijing's fold, and the increasingly stark cross-Strait power asymmetry. Beijing's bullying and hectoring of Taiwan seems to be hardening positions on Taiwan and pushing its citizens ever further from accepting Chinese domination, material differences, economic linkages, and cultural ties notwithstanding. Such trends may be especially pronounced given the mere presence of the USA and Japan as security partners for Taiwan. That coercion does not pay as well when facing pluralistic publics and audiences in the Taiwan case seems to anticipate broader public responses to Beijing's ‘wolf warrior diplomacy’-style coercion from Europe and North America to Oceania (Martin, Reference Martin2021).
Why Beijing persists in such behavior despite their apparent counter-productivity remains unanswerable at present given the PRC's opacity, but it suggests a trend toward greater friction in China's external relations. That Beijing today insists on using coercion over a range of issues and interlocutors indicates a conviction in the efficacy of such an approach at the highest levels, which may reflect thinking among other major powers as well. Even as public responses around the world come to demonstrate more resistance toward such actions, the PRC seems unlike to back down on such use of pressure tactics. This likely spells an increased possibility of heightened tensions, even diplomatic confrontation.
Worth noting too is that growing preference for balancing does not translate into support for de jure independence. Balancing aims to preserve a status quo where Taiwan has substantive practical independence and behaves like a sovereign state for all intents and purposes despite limited formal diplomatic recognition. There is little evidence to suggest that any Taiwan administration or the public is making a serious push for a formal, legal declaration of independence. Different opinion polls consistently indicate that an overwhelming majority of Taiwanese (~85–87%) disfavors unification, but that only a large minority favors independence (~29–31%) and a small minority (~6–7%) wishes to do so immediately (Mainland Affairs Council, 2021). These results suggest that while PRC coercion has limited effectiveness in compelling unification, it may have some deterrent effect against independence for now.
Implications that follow from our findings suggest several additional avenues for research. For instance, whether, when, and how experience and learning from coercion and resistance can be transferred to other governments and publics may be worth further exploration. The degree to which decisionmakers account for public preferences when facing coercive threats and when coercers decide to escalate despite rising costs and risks may present pathways for further investigation. Conditions that affect learning by decisionmakers and publics in third-party states when observing coercive action and responses to such coercion, either in general or when facing pressure from the same coercer, are another avenue for examination. In this regard, we hope to spur more research into coercive diplomacy.
Coercion and coercive diplomacy are common phenomena in international politics that consistently receive significant scholarly attention. More recent literature has highlighted limitations facing such action as well as efforts to study and explain them (Christensen, Reference Christensen2011; Kastner and Pearson, Reference Kastner and Pearson2021). The longstanding US inability to elicit compliance from North Korea, Iran, and Cuba, as well as the large literature discussing the ineffectiveness of sanctions indicate that coercive moves do not always succeed, even for ostensibly capable and powerful actors (Baldwin, Reference Baldwin1999). Appreciating why is important given the prevalence of coercive attempts and efforts to resist them. By highlighting the complications that public opinion in a democracy can pose for coercive diplomacy, our piece follows in and further develops one aspect in this vein of scholarship.
Examining coercion in the context of the cross-Strait relationship as well helps shed light on the question of resistance to coercion among democracies. Power asymmetry between Taiwan and China is nothing if not extremely stark despite Taiwan's many positive attributes, given China's status as the world's most populous country, second largest economy, and one with a large and highly capable military force. In any other context, Taiwan, with a population, wealth, and level of development comparable to Australia, could easily count as a middle power. This makes Taiwan a ‘hard case’ for whether a liberal democracy can stand up to longstanding coercion by an entity whose material advantages are overwhelming. That Taiwan's population displays persistent resilience and growing defiance in face of expanding and intensifying coercive threats from Beijing suggests that citizens in a liberal democracy can pushback against pressure despite material limitations.
As China's wealth and coercive capabilities develop, there have been increasing doubts in some quarters over Taiwan's willingness to defend itself and questions about its possible capitulation. Our findings suggest that at least among the Taiwan public, PRC pressure seems to engender resistance rather than submission to compellence so long as options for security partnerships with other major powers remain. Threats can drive resolve as well as a desire to maintain the status quo while also deterring adventurism as well as risk-taking behavior among targets of coercion. Future studies can investigate how ideational or psychological factors lead to individual perceptions on threat and how such perceptions result in different preferences to strategic responses. Should Taiwan's population and government succumb to Beijing's demands in some way, shape, or form, it is likely to be for reasons other than threats from across the Taiwan Strait. These findings should prompt greater confidence in Taiwan's resolve and prudence among its citizens and partners, while encouraging Beijing to find less menacing and less costly means of reaching out to Taiwan if it is committed to peaceful engagement.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1468109923000014 and https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/AGNOU0
Acknowledgments
The authors would like to thank Hsin-Hsin Pan, Simona Grano, the editors and anonymous reviewers of JJPS for helpful comments, and Zhu-Ling Tseng for excellent research assistance.
Financial support
This study was supported by the Social Science Research Council of Singapore (grant no. MOE2026-SSRTG-20) and National Science and Technology Council of Taiwan (grant no. NSTC111-2628-H-001-005).
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix
Table A1. Operationalization of variables
