1. Introduction
Under what conditions does foreign policy become partisan? Can partisan framings change support for an otherwise bipartisan foreign policy issue? In the United States, there are almost no policy issues left, including foreign policy, that have not become subject to deep polarization (Freidrichs and Tama, Reference Friedrichs and Tama2022; Smeltz, Reference Smeltz2022). There is an enigmatic exception, however: Taiwan. Taiwan is one of the last bipartisan issues in the United States that both Republicans and Democrats support. As Senator Lindsey Graham said in his speech during his bipartisan delegation trip to Taipei in April 2022, ‘America has been politically divided, but I want you to know when it comes to Taiwan, we are united’. House and Senate voting records, along with US public opinion data, show that Graham’s words are true; bothoth Democratic and Republican politicians vote unanimously in support of any US-Taiwan policy put before them. Public opinion polling of American voters also shows that Republican and Democratic partisans increasingly stand together on Taiwan, contrary to almost every other political issue today (Smeltz and Kafura, Reference Smeltz and Kafura2021).
Over the last decade, however, political scientists have closely followed how partisanship has broken down a number of critical US political issues into partisan camps (Schultz, Reference Schultz2017). Foreign policy has not been immune to this polarization trend (Myrick, Reference Myrick2021; Friedrichs and Tama, Reference Friedrichs and Tama2022; Saunders, Reference Saunders2022). Taiwan offers a particularly important puzzle because of its sustained bipartisan nature. Unlike other critical foreign policy issues like the Iraq War or the Israeli-Palestinian Conflict that have seen polarization along partisan lines, Taiwan has maintained bipartisan support for decades (Cavari and Freedman, Reference Cavari and Freedman2019). As US-China and US-Taiwan relations become increasingly salient for American voters, subsequent support for foreign policy issues has the potential to become polarized. Although Democrats and Republicans are seemingly aligned when it comes to an increasingly hawkish stance towards China, some partisan politicians have already begun to frame US-China foreign policy and US-Taiwan foreign policy in partisan terms by accusing one party of being ‘stronger’ or ‘weaker’ when it comes to standing up to China or defending Taiwan (Reference CathyCathy, 2023, ; Friedman, Reference Friedman2023). It would be a mistake to assume the bipartisan support that Taiwan enjoys is inevitable, regardless of the current state of the US’s domestic politics.
How do we know Taiwan is bipartisan? Every major piece of US foreign policy legislation towards Taiwan has been supported across party lines in the United States. The Taiwan Relations Act Affirmation and Naval Vessel Transfer Act of 2014, Taiwan Travel Act of 2018, and the Taiwan Allies International Protection and Enhancement Initiative Act of 2020 were all unanimously supported by both Democrats and Republicans. Since 2018, there have been over 15 bipartisan congressional delegations to Taiwan featuring partisan heavyweights from Nancy Pelosi to Mike Pompeo. Even at the state level, more than a dozen states in the US have pro-Taiwan caucuses, ranging from blue-leaning states like Hawaii and Connecticut to red-leaning states like West Virginia and Kentucky. Although the way Democrats and Republicans debate Taiwan within their party can vary from advocating for more support for Taiwan to questioning support for Taiwan, when it comes to voting on policy, both parties remain bipartisan. Public opinion confirms that elites are not alone in their support for Taiwan, but that across party lines, Americans have favourable views towards US-Taiwan policy (Smeltz and Kafura, Reference Smeltz and Kafura2021).
This paper tests whether or not Taiwan can stay bipartisan in a world that is becoming increasingly partisan. Using data from two preregistered original survey experiments – one sampling partisan Democrats and the other partisan Republicans – this paper tests and identifies under what conditions and on which policy issues bipartisanship holds or does not hold. I argue that Taiwan can deeply inform our understanding of under what conditions foreign policy becomes partisan. Contrary to theoretical expectations, partisans neither fully ‘follow the leader’ nor use partisan-motivated reasoning for all Taiwan-related issues. Instead, only certain issues are affected by partisan framing. Specifically, I find that Democrats are more likely to lower their support for diplomatic policies when Taiwan is framed as a Republican issue. Republicans, however, are less likely to change their support for Taiwan’s diplomatic policy, even if it is framed as a Democratic issue. Contrary to existing theories of foreign policy and public opinion, the case of Taiwan shows that not all foreign policy issues are polarized in the same way.
Why does Taiwan matter? Beyond the contemporary need to understand the US-Taiwan relationship, its connection to geopolitics, and its potential as a military flashpoint, Taiwan is a useful case for building and testing theories of foreign policy because of how multifaceted and wide-reaching Taiwan is as a global foreign policy issue. The US-Taiwan relationship is itself complex, reaching deep into economic, military, and diplomatic policy realms. This paper considers the single case of Taiwan, but the basket of highly salient issues related to foreign policy issues related to Taiwan vary widely, and are all integral to the US-Taiwan relationship and global peace more broadly. As one recent study of Taiwan politics put it, ‘the next world war may be fought over Taiwan’, (Sullivan and Nachman, 2024). Understanding this case is not just of scholarly interest, but has implications that are meaningful in and outside of the academy.
This study makes an important contribution, in particular, to studies of US support for different partners around the world. Although studies often only measure support for a country in general terms or as a single measure of defence, this study shows that different foreign policy issues related to Taiwan vary, especially when primed with a partisan frame. By asking questions for support on a wide variety of concrete foreign policy issues rather than more abstract support for defending Taiwan, I am able to build on previous studies to show how and where we see varying levels of support for specific aspects of US-Taiwan relations (Rich, Reference Rich2023). This study adds to the literature on the relationship between foreign policy and public opinion by theorizing and providing empirical evidence about the effects of partisanship on US foreign policy towards Taiwan, an outlying yet critical case that has been understudied in the existing literature, and has real-world implications for today’s turbulent US-China relationship.
2. Theories of foreign policy and public opinion
What is the relationship between foreign policy and public opinion? A large body of academic literature argues that voters form opinions from top-down, elite-driven cues (Agadjanian, Reference Agadjanian2021; Cavari and Freedman, Reference Cavari and Freedman2019; Baum and Groeling, Reference Baum and Groeling2009; Jentleson and Britton, Reference Jentleson and Britton1998). Since voters typically do not have well or deeply informed opinions about foreign policy, they often look to their party elites for signals on how and what they should support when it comes to US foreign policy, especially on issues they are unfamiliar with (Foyle, Reference Foyle1997, Zaller, Reference Zaller1992). Voters often simply follow political elites, even if elite stances differ from how individuals feel (Barber and Pope, Reference Barber and Pope2019; Broockman and Butler, Reference Broockman and Butler2017). These elite cues are then promoted and disseminated through the news or social media (Brody, Reference Brody1991; Orhan, Reference Orhan2023).
Alternatively, bottom-up models suggest that voters take partisan cues from other voters, affiliated organizations, or they may simply rely on their own moral values and beliefs in the absence of elite or other cues (Bullock, Reference Bullock2011; Rathburn et al., Reference Rathbun, Kertzer, Reifler, Goren and Scotto2016; Ketzer and Zeitoff, Reference Kertzer and Zeitzoff2017 Tomz et al., Reference Tomz and Weeks2020). Bottom-up models challenge the disputed common wisdom dictates that public opinion on foreign policy does not matter, because on certain issues, politicians will take cues from public opinion and diverge from party lines, such as Democrats diverging from party lines over the Israel-Palestine conflict (Cavari and Freedman, Reference Cavari and Freedman2020). Under certain conditions, elected officials do bend to the will of their bases of support (Baum and Potter, Reference Baum and Potter2015; Saeki, Reference Saeki2013). Depending on the issue and context, it is a combination of both top-down elite cues and bottom-up social cues that drives political opinion formation on foreign policy.
Foreign policy itself is often partisan. Whether it is support for military intervention, the use of force, or humanitarian aid, Democrats and Republicans often disagree on certain issues (Jacobson, Reference Jacobson2010; Koch and Sullivan, Reference Koch and Sullivan2010; Fordham, Reference Fordham1998). Foreign policy in the United States has also not been immune to the growing trend of affective polarization (Iyengar et al., Reference Iyengar, Lelkes, Levendusky, Malhotra and Westwood2019; Schutz, Reference Schultz2019). Affective polarization in the United States has made unity over foreign policy even more difficult, an effect that the United States has seen increase over time (Myrick, Reference Myrick2021). The polarization of foreign policy, especially since the Trump presidency, has not gone unnoticed by political scientists, who note how Trump’s divergent attitude towards the US’s role in global order has only caused more division within the United States’ political elites and public (Tarzi, Reference Tarzi2019; Smeltz et al., Reference Smeltz, Daalder, Friedhoff and Kafura2017). Partisan motivated reasoning has also pushed voters to agree with platforms when their party supports them and disagree with platforms when the opposing party supports them. For example, Bolsen et al. (Reference Bolsen, Druckman and Lomax Cook2014) find that Democrats and Republicans were more supportive of the 2007 Energy Act when endorsed by their own partisan elites, but significantly less supportive of the same policy when endorsed by the other party’s elites. Guisinger and Saunders (Reference Guisinger and Saunders2017) importantly note, however, that partisanship over foreign policy will vary not just by issue but also by the timing and context in which certain issues are introduced.
US-China relations as a foreign policy issue are also not immune to partisan effects. Although Republican and Democrat elected officials appear united over countering the People’s Republic of China’s (PRC) rising challenge to the United States (Carothers and Sun, Reference Carothers and Sun2023), this unity is largely contingent upon partisan messaging and framing. For example, Myrick (Reference Myrick2021) found that when China is presented as a threat from a non-partisan source, both Democrats and Republicans converge on their perceptions of China as a threat. But, when Chinese threats are framed as coming from Donald Trump, Democrat and Republican threat perception of China becomes drastically partisan (946).
Why does it matter if foreign policy becomes increasingly partisan? Kenneth Schultz specifically outlines four ways in which polarized foreign policy can lead to precarious outcomes: (1) it is more difficult to get bipartisan support for ambitious or risky undertakings, particularly the use of military force and the conclusion of treaties; (2) it is hard to agree across parties on the lessons of foreign policy failure, complicating efforts to learn and adapt; (3) the risk of dramatic policy swings from one administration to one of the opposite party complicates long-term commitments to allies and adversaries; and (4) the vulnerability of our political system to foreign intervention is heightened (Reference Schultz2017). Subsequently, it is not only interesting for scholars to understand under what conditions foreign policy becomes bipartisan from a social science perspective. Such research is also empirically important to politicians, analysts, and foreign policy makers who want to ensure peaceful continuity.
Despite affective polarization and partisan motivated reasoning creating more divisions across the foreign policy realm, Taiwan is still somehow supported across bipartisan lines. What do theories of foreign policy and public opinion combined with empirical studies of US-China relations thus suggest we should see in the United States when it comes to testing support for Taiwan foreign policy? We should theoretically expect that under the condition that Taiwan is framed as a partisan issue that partisans from the opposing party should support it less. In other words, when Taiwan is framed as a Democratic issue, Republicans should support it less than when Republicans see it as a Republican issue. Subsequently, when framed as a Republican issue, Democrats should support Taiwan policy less than when framed as a Democratic issue.
3. US-Taiwan relations
Since the normalization of US-China relations in 1979, Taiwan has continuously defined and refined the two powers’ contentious relationship. Even when relations between the US and the PRC were established, Taiwan was one of the fundamental questions on which both sides eventually decided to agree to disagree (Rigger, Reference Rigger2019). One of the key demands from the PRC was that the United States end formal diplomatic relations with Taiwan, to which the United States agreed. The US, along with much of the international community, accepted the PRC’s demand to diplomatically and internationally isolate Taiwan. Despite the PRC’s objections, however, the US established informal relations with Taiwan. These relations are defined through a series of policies, declarations, and communiques, including the Three Communiques, the Six Assurances, and the Taiwan Relations Act.
Based on the above policies and acts, US policy towards Taiwan is summarized as follows: the United States does not take a formal position on the status of Taiwan. Although the US acknowledges China’s claims over Taiwan, it does not recognize those claims (Drun and Glaser, Reference Drun and Glaser2022; Liff, Reference Liff2022). The United States wants a peaceful resolution to the issue of Taiwan, but only through peaceful negotiation between both sides. The United States does not endorse either side unilaterally changing the current status quo, meaning the US does not want China to unilaterally attack Taiwan and bring it into its territory, and it does not want Taiwan to unilaterally declare formal independence. US policy towards Taiwan is designed with this dual-deterrence in mind. The United States sells weapons to Taiwan ‘of a defensive nature’ with the intent of deterring China from attacking. At the same time, the US does not formally use ‘Taiwan’ or ‘Republic of China’ when creating institutions that govern US-Taiwan relations as to not infer Taiwan’s formal independence or status as a recognized country (Dickey, Reference Dickey2019).
US-Taiwan and US-China policy, however, has qualitatively changed since the Trump presidency, especially since the COVID-19 pandemic. Before the COVID-19 pandemic, his track record was mixed; he took a controversial phone call with then-Taiwanese president Tsai Ing-wen, leading many to perceive him to be pro-Taiwan. He also, on multiple occasions, openly stated his admiration and reverence for Xi Jinping (Shepardson, Reference Shepardson2018; Wong, Reference Wong2019). After the COVID-19 pandemic began, Trump’s attitude towards China and Taiwan changed. Anti-China rhetoric combined with a new-found love for pro-Taiwan policy became the new standard approach for the Trump administration (Aurora and Kim, Reference Arora and Kim2020).
As the Biden administration took office, hawkish attitudes towards China and overt support for Taiwan remained consistent. One major change, however, is the potential shift in the bipartisan nature of China policy. Despite President Joe Biden’s continued support for Taiwan, Republicans are increasingly using China and Taiwan as a way to attack Democrats in partisan ways (Lima, Reference Lima2023 Cathey, Reference Cathy2023; Friedman, Reference Friedman2023). Republicans have begun to use China’s foreign policy to frame Democrats as ‘weak’, despite the seemingly partisan agreement on how to approach US-China relations and US-Taiwan policy. Republican claims that Democrats are weak on Taiwan are also not necessarily new. As far back as the 1996 Third Taiwan Strait Crisis, Republicans were critical of Democrats for a perceived weak response against China (Kulacki, Reference Kulacki2020).
Public opinion from American voters has also consistently supported Taiwan for decades. Chicago Council on Global Affairs surveys of Americans have asked questions about support for Taiwan and varying US-Taiwan policies since 1978. It is from their data that we see bipartisan support for Taiwan in Congress is mirrored by the American public. For example, a 2022 survey showed nearly identical levels of support from Republican and Democratic respondents across a whole array of economic, diplomatic, and military policies in support of Taiwan (Smeltz and Kafura, Reference Smeltz and Kafura2022). When asked if China were to invade Taiwan, 76% and 77% of Democrats and Republicans, respectively, support the US imposing economic sanctions. 65% and 62% of Democrats and Republicans support sending additional military supplies, and 62% by both parties support using the US Navy to stop a blockade around Taiwan. The only clear varying issue was whether the US should accept Taiwanese refugees, which saw a 10% difference between Democrats in favour at 67% versus Republicans at 57%. When asked using a feeling thermometer about sentiment towards Taiwan from 0 to 100, Democrats were at 60% and Republicans 59%. Just like how US politicians advocate for military, diplomatic, and economic support for Taiwan, all three foreign policy issues are also supported across party lines by American partisans.
Regardless of the political drama from Trump or Pelosi over Taiwan, and despite increasing affective polarization in the United States, Taiwan has enigmatically remained one of the last bipartisan issues in the United States, all while becoming increasingly salient in US domestic politics. Whether or not bipartisanship can hold – especially as parties begin to frame Taiwan in partisan terms – will become a critical question for social scientists and policymakers.
4. Hypotheses
Data indicate that both political elites and civil society feel strong, bipartisan support for Taiwan, but how durable is that support? If framed in a partisan way, does support for Taiwan hold or waver? Based on previous literature on foreign policy and public opinion, combined with studies of US-Taiwan foreign policy, I hypothesize that partisans will support US-Taiwan policy when framed as something their own party supports. However, when framed as an issue from the opposing party, support will be significantly lower. In other words, we should see support for Taiwan foreign policy decline when a partisan sees the opposing party support it. I test this across the three key sets of US-Taiwan foreign policy issues: military support, diplomatic support, and economic support.
I propose three sets of hypotheses to test support for US-Taiwan policy. Each set is broken down into three key areas of US-Taiwan foreign policy: military support, diplomatic support, and economic support. These three areas of focus and the questions created for these surveys are based both on the above-mentioned scholarly literature on foreign policy and public opinion and on the decades of public opinion data gathered by the Chicago Council on Global Affairs. Military support questions ask varying levels of commitment from the United States in the event of a conflict between the US and Taiwan. Diplomatic support is measured by asking questions on supporting the US-Taiwan’s relationship at the expense of the US-China relationship, along with the consistently polled question of whether the United States should formally recognize Taiwan. Finally, economic support asks about the US’s reliance on Taiwan’s economy, and specifically Taiwan’s semiconductor industry.
4.1. H1: Military support
Democrats/Republicans who see US support for Taiwan primed as the opposite party’s issue are more likely to have lower support for US military policy for Taiwan than when Taiwan is presented as their party’s issue.
4.2. H2: Diplomatic support
Democrats/Republicans who see US support for Taiwan primed as the opposite party’s issue are more likely to have lower support for US diplomatic policy for Taiwan than when Taiwan is presented as their party’s issue.
4.3. H3: Economic support
Democrats/Republicans who see US support for Taiwan primed as the opposite party’s issue are more likely to have lower support for US economic policy for Taiwan than when Taiwan is presented as their party’s issue.
5. Methods
To test this question, I conducted two preregistered survey experiments of American partisans to see if partisan framing causes support for US-Taiwan policy to vary. Prior to conducting this study, the research questions, hypotheses, survey design, and analysis plan were preregistered (anonymized pre-registration available here: https://osf.io/bz9r3/?view_only=8cbbdc165f39435085de3495dcfb80bf). Survey experiments are particularly apropos for investigating this question because they allow for the specific manipulation of which party is expressing support for Taiwan. I focused explicitly on partisans instead of a nationally representative random sample. The scope of my sample was selected because the fundamental question is about testing whether bipartisanship holds, and doing so requires that partisans – those who already identify with a political party – be tested, rather than a random sample that includes independents. I accessed my sample through Prolific, an online survey company that is well established in the survey and survey experiment world as a high-quality source for academic survey data (Douglas et al., Reference Douglas, Ewell and Brauer2023; Eyal et al., Reference Eyal, Rothschild, Gordon, Zak and Ekaterina2021). I then designed the survey on Unipark, a survey design platform, before fielding the survey through Prolific’s panel of respondents. Prolific provides options to sample randomly from their large pool of pre-screened self-identified Republicans and Democrats. In each experiment, respondents were assigned to one of two groups: a Republican treatment group or a Democratic treatment group. The two survey experiments were identical in design. I conducted two separate surveys that sampled only Democrats and Republicans in each in order to ensure that we are able to test effects on partisans explicitly. The research question here is particularly focused on whether or not partisans have a particular response to supporting Taiwan when presented as their party versus the opposing party. Conducting two identical surveys of Democrats and Republicans enabled a focused study of this exact question. Independents and non-partisans were purposefully not included. Although the question of how partisan frames may affect or change attitudes for non-partisans is meaningful, it is beyond the scope of this paper’s focus on partisans.
A power analysis showed that a mean–difference of 0.35 and an SD of 1.3 (Cohen’s D of 0–23) require a sample size (N) of 400 to achieve a power of 0.9. I set a higher sample size of 550 per group to enhance the robustness of my results. In other words, I would need a sample of 400 per group to identify effects in 90% of the studies. I set a higher sample size of 550 per group to enhance the robustness of my results. The first was a sample of Democrats (n = 1171) and the second a sample of Republicans (n = 1156).
The Republican and Democratic treatments were designed using both traditional and social media to frame party support for Taiwan. Twitter was the main platform highlighted to show party support. For decades, Twitter has been a key platform through which parties and elites try to sway public opinion, including foreign policy (Pal, Reference Pal2017; Pain and Chen, Reference Pain and Masullo Chen2019). Twitter can be used to add or disperse political pressure, signal transparency, and as a means for ideas or policy diffusion (Hassanpour, Reference Hassanpour2014; Simmons and Elkins, Reference Simmons and Elkins2004). Recent studies continue to show how Twitter increases partisanship and creates divides within partisan camps on foreign policy issues (Nachman et al., Reference Nachman, Rauchfleisch and Hioe2022). Although Elon Musk’s new management of Twitter (now X, however, this study was done prior to the rebrand) has called into question the role of the platform and if it will continue to serve the function it once did as a source for news and political posturing, the platform continues to be a relevant source of political information (Rohlinger et al., Reference Rohlinger, Rose, Warren and Shulman2023).
The treatments subsequently displayed a tweet backdropped onto a mass-media headline news breaking story. The headline describes how each party is expressing support for Taiwan on Twitter. The tweet then has the official Democratic Party handle (@TheDemocrats) or official Republican Party handle (@GOP), along with a modest number of comments, reposts, likes, and overall engagement. The language within the tweet reads, ‘America’s solidarity with the 23 million people of Taiwan is more important than ever, as the world faces a choice between autocracy and democracy’. The language within this tweet comes from a tweet Nancy Pelosi posted during her trip to Taiwan in August 2023. The reason for using the language from her tweet was twofold: first, it does not designate a specific aspect of Taiwan policy, but rather expresses vague support for Taiwan in general. Second, language from an actual political elite makes the treatment more realistic and akin to something respondents would see in real life. Unlike previous studies, these two survey experiments do not frame partisanship through a single politician, but instead by each major party’s name. This allows certainty that respondents are primed by the party itself rather than an individual politician. I used the two parties rather than specific political elites like Nancy Pelosi or Mike Pompeo – two outspoken advocates for Taiwan from each party – because it would be difficult to disentangle if respondents were making their choices based on party affiliation or some other factor, such as Nancy Pelosi’s gender. Since the goal of this survey is to test partisanship rather than particular political elites, it is also more valuable to use the party as a prime rather than an individual because of how partisans evaluate individual politicians even within the party differently from the party itself (Bartels, Reference Bartels2018). Using parties rather than individual elites or grassroots organizations also attempted to strike a middle-ground between top-down and bottom-up theories of foreign policy and public opinion. The survey design was also done so in a way that made the issue explicitly about supporting Taiwan. Rather than include China or anything related to the PRC in the design, respondents were primed only to see support for Taiwan, rather than any sort of anti-PRC or opposition to China rhetoric. This design choice also ensures that those taking the survey are responding to support for Taiwan, rather than opposing China (Figure 1), (Figure 2).
Figure 1. Democratic treatment.
Figure 2. Republican treatment.
The survey was designed as follows: first, all respondents were asked about their political knowledge of a handful of countries, including Taiwan. This allowed for control over how much political knowledge respondents felt they had about Taiwan. Next, respondents were told they would be given a recent headline and instructed to read it carefully. Each respondent was randomly assigned to either the Republican treatment or the Democratic treatment. Next, participants were asked a battery of questions along three main policy areas: military support, diplomatic support, and economic support. The questions were based on both previous surveys of American attitudes towards Taiwan and academic research investigating foreign policy attitudes (Jahni et al., Reference Jahani, Gallagher, Merhout, Cavalli, Guilbeault, Leng and Bail2022; Sambanis et al., Reference Sambanis, Tsakonas and Lee2023). Attention checks were also included in the survey. Respondents who did not pass did not have their data recorded in the final dataset, which is why there is a slight variation in the size of the two survey experiments.
All items were asked on a seven-point scale. There are three key outcome variables: military support, diplomatic support, and economic support. Military and diplomacy items were combined to mean indices. Military and diplomacy support are each based on three items. Economic support is based on two items that will not be combined into an index. Initial factor analysis based on the pre-test data showed that military items and diplomacy items could be combined, while economic questions will be analysed as single items, which was subsequently included in the pre-registration.
For data analysis, I ran linear regressions with the adjustment. The Lin estimator is more suitable for experimental data because it better accounts for covariate balance and treatment effects, offering a more robust finding than typical ordinary least squares regression analysis. I also conducted a difference in means test for each of the four groups (Democrats who see Democratic prime, Democrats who see Republican prime, Republicans who see Democratic prime, Republicans who see Republican prime). The full models using the (Lin, Reference Lin2013) adjusted regression for each hypothesis can be found in the appendix.
6. Results:
Both Democrats and Republicans reported a similar level of knowledge about Taiwan. Among the six countries about which respondents were asked to report how much they know (Germany, Canada, Russia, Ukraine, China, and Taiwan), Taiwan had the lowest average score for both sets of respondents, with 2.79 for Democrats and 2.8 for Republicans, respectively. All six countries had an average knowledge below four, indicating that all respondents had a relatively low level of knowledge across cases, even if Taiwan was the lowest. Average knowledge of China was almost one point higher than Taiwan at 3.62 for Republicans and 3.57 for Democrats (Figure 3), (Figure 4).
Figure 3. Reported knowledge of countries (0–7).
Figure 4. Difference in means test.
(Figure 3. Visualizing differences between Democrat (on the left) and Republican (on the right) survey respondents. Point estimates represent the difference of means between those shown a Democratic versus a Republican treatment. Red colour indicates a significant negative effect with 95% confidence intervals displayed.)
7. Diplomatic support:
Following exposure to either the Republican or Democratic treatment, respondents were asked their support for three items revolving around US diplomacy towards Taiwan: ‘The US should increase ties with Taiwan, even if it hurts its relationship with China’, ‘It is important for the United States to help advocate for Taiwan because it is a democracy’, and ‘The United States should formally recognize its relationship with Taiwan’. Based on initial factor analysis, these three questions were then indexed together into a single diplomacy score per respondent (α = .82).
According to H1a and H1b, Democrats who are exposed to the Republican prime should have lower support for diplomatic foreign policy than when exposed to the Democratic prime. Conversely, H1b predicts that Republicans exposed to the Democratic prime should have lower support for diplomatic foreign policy than when exposed to the Republican prime. These two hypotheses, however, saw varying results depending on whether respondents were Democrats or Republicans. The treatment had a very weak effect on Republicans, who were neither more nor less likely to support diplomatic policies for Taiwan if primed by either the Democratic or Republican treatment (b = 0.047, p = 0.535, 95% CI [−0.103, 0.19]). Democrats, however, were statistically significantly less likely to support diplomatic policy when Taiwan was primed as a Republican issue (b = −0.1893, p = 0.0045, 95% CI [−0.32, −0.058]). This finding demonstrates that, under certain conditions, partisanship for Taiwan does not hold, and support for diplomatic policies towards Taiwan can be lowered if primed along partisan lines. The effect, however, appears to only be significant with Democrats, not Republicans.
8. Military support:
Like the diplomatic support questions, H2 examining military support consisted of three questions that were later combined together into a single index, giving each respondent a military support score (α = .93). The three questions for military support were focused on different levels of US involvement in the event of a military conflict in the Taiwan Strait. The three items were as follows: ‘If there is a war over Taiwan, America should be involved’, ‘If China creates a blockade around Taiwan and cuts off its access from the world, then the United States should send troops to help Taiwan’, ‘If China declares war and invades Taiwan, then the United States should send troops to help Taiwan’. The questions were presented in this fashion to account for scale and intensity of different US intervention scenarios, ranging from military support no matter what, to military support under certain conditions.
Neither Republican (b = 0.191, p = 0.0583, 95%CI [−0.006,0.39]) nor Democratic (b = −0.139, p = 0.117, 95%CI [−3.14,0.035]) respondents’ support for military support from the United States was significantly affected by either partisan treatment. Republicans, however, came close (0.058), hinting that there could be some correlation between Republican support for military policy and partisan framing. These results should not be over-interpreted, since they still did not reach the lowest significance threshold.
9. Economic support:
Initial factor analysis showed that economic support (H3) questions needed to be analyzed separately. Subsequently, there is no combined economic index, but instead each item is analysed individually. The two items were as follows: ‘The United States should do more to separate our economy from China’, and ‘The United States should reduce our reliance on Taiwan for semiconductors’. Neither the question regarding economic decoupling for Republicans (p = 0.042, p = 0.616, 95% CI[−0.123, 0.207]) or Democrats (b = −0.147, p = 0.084, 95% CI=[−0.314, 0.019]), nor the question regarding semiconductor reliance for Republicans (b = −0.0127, p = 0.178, 95% CI=[−0.313, 0.058]) or Democrats (b = 0.099, p = 0.253, 95% CI[−0.071, 0.271] yielded significant results, nor was there any weak correlation.
10. Discussion:
Under what circumstances does bipartisan support for Taiwan break? The findings in this study indicate that the most significant change in support for Taiwan occurs among Democrats when Taiwan is framed as a Republican issue. Specifically, there is a notable decline in diplomatic support. While Republicans also exhibit a correlation between Taiwan being framed as a Democratic issue and decreased military support, these results fall short of statistical significance. Economic policy, irrespective of partisan framing, does not seem influenced by support for either party. This survey experiment demonstrates that partisan framing does not equally polarize both parties. While some aspects of US-Taiwan policy are affected by partisanship, not all partisans are influenced by partisan framing. The study’s findings suggest that bipartisan support for Taiwan isn’t entirely immune to the growing polarization in the United States. However, not all issues related to US-Taiwan foreign policy are vulnerable to polarization.
Studying Taiwan contributes to existing theories of foreign policy and public opinion in two primary ways. First, established theories propose that partisans tend to align with the cues provided by their party’s elites (Agadjanian, Reference Agadjanian2021; Cavari and Freedman, Reference Cavari and Freedman2019; Baum and Groeling, Reference Baum and Groeling2009; Jentleson and Britton, Reference Jentleson and Britton1998). Findings, however, suggest that, while voters generally follow elite cues, this behaviour occurs selectively, contingent upon specific conditions and policy issues. Notably, certain policy domains – specifically diplomatic and military policies – appear more susceptible to partisan framing compared to economic policy. This finding implies that, while the conventional understanding of voters following elite guidance holds true in many instances, there are instances where this pattern breaks. Future theoretical developments in foreign policy and public opinion should aim for more targeted theories tailored to distinct issues. Different aspects of foreign policy issue areas may cause varying kinds of partisan effects. This study underscores the necessity of refining broad theories regarding the conditions under which foreign policy becomes partisan, emphasizing the importance of accounting for variations between different policy issues and context. Issues that are particularly contentious or prominent in contemporary discourse, such as recognizing Taiwan or expressing explicit support for Taiwan in a conflict with China, are more likely to become polarized along partisan lines. Conversely, relatively less contentious foreign policy matters – like discussions concerning Taiwanese semiconductors – might be less susceptible to partisan polarization. This study also shows that support for Taiwan is not simply a reaction to China-related foreign policy or US-China relations (Rich, Reference Rich2023). Instead, we see that partisans respond to different Taiwan-related policy realms and specific policy issues. Rather than showing how voters support foreign policy issues in the abstract, this study is able to show that different concrete policy actions solicit different reactions across party lines.
Second, affective polarization and partisan motivated reasoning can still have an effect on Taiwan foreign policy. Findings partially align with theoretical expectations of individuals using cues from the elites of the opposite party to decide what issues not to support (Peterson and Iyengar, Reference Peterson and Iyengar2021). In practical terms, this means that Democrats observing Republican support for Taiwan may be less inclined to support US-Taiwan policy, and conversely, Republicans perceiving it as a Democratic issue may also show diminished support. While previous studies, such as Agadjanian (Reference Agadjanian2021), indicate that voters may follow leaders even when their stance is counterintuitive to the party’s typical positions, my findings highlight that voters are equally attentive to cues from the opposite party, even when counterintuitive to their own party’s stance. Theories centred around the concept of ‘follow the leader’ should further consider the theoretical possibility that voters ‘run away from the other leader’. In alignment with what theories of affective polarization suggest (Jenke, Reference Jenke2023), partisan voters might be shaping their foreign policy stances as much in opposition to the other party as they do in alignment with their own party.
Given the results of these two survey experiments, why then would Democratic support for diplomatic policy specifically be lowered when framed as a Republican issue? One potential explanation lies in current discourse surrounding US-China relations in the United States. Despite this survey experiment showing relatively low levels of knowledge on Taiwan and China by both Democratic and Republican participants, US-China relations and US-Taiwan relations have become mainstream in US news and politics. Specifically, discourse around a ‘new cold war with China’ has become a serious topic of discussion both in mainstream media and academic discourse (Brands and Gellis, Reference Brands and Gaddis2021; Zhao, Reference Zhao2019).
The diplomatic questions asked in this survey experiment, revolving around how the US interacts with China and Taiwan, could be affected by how much the Republican and Democratic parties lean into more aggressive ‘New Cold War’ rhetoric, versus more cautionary and careful rhetoric. Democratic respondents who see Taiwan as a Republican issue could be interpreting diplomatic issues as overly inflammatory towards the US-China relationship, and subsequently lower their support for issues such as improving US-Taiwan relations at the expense of US-China relations, or formally recognizing Taiwan. When Taiwan is framed as a Republican issue within the broader context of escalating tensions with China and ongoing affective polarization, Democrats might recalibrate their support for diplomatic policies related to Taiwan. The framing of Taiwan within the context of US-China relations could trigger shifts in Democratic support, potentially due to considerations about how these policies might impact or align with broader stances on managing relations with China.
Although the experiment’s partisan primes on military support did not yield significant results, the weak correlation between Republicans and the Democratic prime still merits some discussion. Why would Republicans, who have traditionally been more outspoken on hawkish policy towards China, be even somewhat less supportive of military policy towards Taiwan? One potential point of context is the changing support from the Republican Party towards Ukraine. At the beginning of Russia’s invasion of Ukraine, military support for Ukraine was bipartisan across both parties (Kull et al., Reference Kull, Gallagher, Fehsenfeld, Lewitus, Bunn and Sapp2023). In the last year, however, a dramatic change has occurred within the Republican Party, where now more conservatives are less likely to support additional military aid for Ukraine (Galston, 2023). Although Republican rhetoric on Taiwan has not dwindled despite changes in support for Ukraine, one potential explanation for any changes in Republican attitudes towards military aid for Taiwan could be explained by changes in elite signals on Ukraine. Given the stark shift in Republican military support for Ukraine and the subtle findings in this survey experiment, it is reasonable to hypothesize that, under certain conditions, Republicans could still see a significant decline in military support for Taiwan in the future.
Finally, why did neither economic question yield results, or even weak correlations? This null finding is particularly novel, especially because discourse around Taiwan, its ‘Silicon Shield’, and US-China economic decoupling have played no small role in policy discussions over the last five years. One potential explanation is that economic policy is truly bipartisan, and that framing economic decoupling or attitudes towards semiconductors as a Republican issue or Democratic issue has no resonance with partisans. Ever since the semiconductor shortage during the COVID-19 pandemic, the need for Taiwanese semiconductors has become known around the world (Sullivan and Nachman, Reference Sullivan and Nachman2023). Such economic demands and need for supply chains are perhaps truly bipartisan, and do not seem to be affected by partisanship. Similarly, the question of whether or not the US should economically decouple from China is also equally immune to partisan effects. This finding is consistent with other public opinion attitudes towards China and voting records by Democratic and Republican voters and how politicians have seen China over the last number of years (Silver et al., Reference Silver, Huang, Clancy, Lam, Greenwood, Mandapat and Baronavski2023). Despite earlier discussed variations in rhetoric, when it comes to voting on policy, Republicans and Democrats have both agreed on the need for a stronger economic stance on China.
11. Conclusion
This study highlights a critical implication for contemporary US-Taiwan foreign policy. Even though Taiwan remains a bipartisan issue in the United States, such support has the potential to be destabilized by growing polarization in the United States. Since Joe Biden has come into office, a number of Republican politicians have tried to frame him as weak on China and challenged his commitment to resisting encroaching PRC influence around the world (Lima, Reference Lima2023 Cathey, Reference Cathy2023; Friedman, Reference Friedman2023). Framing China and Taiwan as Republican issues that Democrats do not care about is not only counterproductive for the US-Taiwan relationship but it also has the potential to lower public support for US-Taiwan foreign policy. Framing Taiwan as a partisan issue is harmful for US-Taiwan relations, and political elites from all parties should be cognizant of this potential effect of their rhetoric and framing of US-Taiwan relations.
Taiwan is not a singular issue. Instead, there are a multitude of critical policy areas that involve Taiwan, from economic, military, and diplomatic policy. Studies of US foreign policy ought to continue to evolve the way we study these cases in order to account for variation not just in foreign policy issue, but in how the public and elites support different issues. This study shows that certain foreign policy issues related to Taiwan are supported differently, and that partisan frames can change support for some foreign policy issues, but not others. Accounting for these within-case variations will help us better understand not just partisan effects but also where we still see bipartisan unity.
There are also limitations to this study. It only focuses on one foreign policy issue – the US-Taiwan relationship. If Taiwan were swapped out for, for example, Ukraine or Palestine, we may see varying results, especially in which issues are vulnerable to partisan effects. Although this study is explicitly about foreign policy, the finding that partisan framing under certain conditions affects support for policy ought to also be further tested beyond foreign policy, especially given the abundance of domestic issues in America that are particularly prone to partisanship. This study is also specifically a case study of partisans – those who identify explicitly with the Democratic or Republican party. The argument made here can therefore not be fully generalized to the more general US voting public, which includes independents and other party supporters. Future research ought to also further interrogate the varying demographics of who supports which foreign policy issues. Beyond party identification, other demographic features such as ethnicity, race, gender, etc. may be useful measures for understanding under what conditions foreign policy issues are supported or not. Taiwan as a foreign policy issue is also deeply tied to broader discussions of US-China relations and the US’s role as a global leader. Although connected, the findings of this study should not be extrapolated to all realms of these complex discussions involving China, Taiwan, and the US’s role in international order. In this study, I am unable to account for whether or not anti-China sentiment is affecting support for Taiwan. However, previous research shows that anti-China sentiment is correlated with support for Taiwan; however, support for Taiwan and anti-China sentiment also grow independently from each other (Rich, Reference Rich2023). Finally, while (Guisinger and Saunders, Reference Guisinger and Saunders2017) rightfully point out the shortcoming of single-country survey experiments studying support for foreign policy, this study still contributes an answer to an important question of Taiwan policy at a time when the specific issue of US-Taiwan and US-China relations is contextually salient to both elites and the US public.
Taiwan remains one of the world’s greatest flashpoints, and may be the location of the next geopolitical conflict of our time (Sullivan and Nachman, Reference Sullivan and Nachman2023). As Taiwan’s most critical ally, it is imperative we further study and analyse the US-Taiwan relationship, especially under what conditions US voters and elites support US-Taiwan foreign policy. Bipartisanship in the United States is difficult to achieve, and even harder to maintain. From environmental issues to gun control, bipartisanship in the US continues to dwindle because of partisanship (Ehret et al., Reference Ehret, Van Boven and Sherman2018). Issues like Taiwan also present risk, where politicians could strategically use partisan framing to increase or limit support for Taiwan while maintaining a guise of bipartisanship in other policy areas. The necessity for bipartisanship on issues like Taiwan also extends beyond just politicians, but must also be taken into consideration by lobbyists, interest groups, and even foreign governments who offer statements on Taiwan and the US-Taiwan relationship.
While Donald Trump prepares to enter his second administration, there has never been more uncertainty about US foreign policy towards East Asia. Donald Trump’s support for Taiwan has waxed and waned over the last election cycle, and the foreign policy voices surrounding him offer radically different opinions on how the US ought to approach China, Taiwan, and the entire East Asia region (Nachman, Reference Nachman2024). Some Republicans are advocating for continued bipartisan support for Taiwan, while other key voices in the Republican establishment offer a much cooler tone on Taiwan. While we do not know what version of Donald Trump’s US-Taiwan policy we will see in the coming years, how said policy is framed and presented to the United States, Taiwan, and the world will have serious ramifications. As findings from this study suggest, it has never been a more important time for US-Taiwan policy to be presented in a bipartisan way. At a time when US-China relations are struggling to find a floor to stand on, clear-eyed analysis of the complex triangular relationship between the three states has never been more imperative. This study is meant to contribute to that insight, with the hope that those who study and are responsible for US-Taiwan policy can do so as effectively and productively as possible.
Data availability statement
Pre-registration research plan can be found at: https://osf.io/bz9r3/?view_only=8cbbdc165f39435085de3495dcfb80bf
Acknowledgements
The author would like to thank the anonymous reviewers for their time and effort, Adrian Rauchfleisch for his guidance throughout this project, and Shane Lin for his feedback on early drafts of the surveys.
Funding statement
This project was generously funded by the Republic of China (Taiwan) National Science and Research Council Grant #: 111-2410-H-004 -227 -MY3
Project was approved under National Chengchi University Internal Review Board #: NCCU-REC-202209-I108
Competing interests
I have no conflicts of interest to disclose.
Appendices:
Table 1. Support for US-Taiwan policy: mean (SD) and sample size by party
Note: Difference in means estimated using Lin (2013) covariate-adjusted estimator.
Table 2. Difference in means analysis: b and p values by party
Table 3. Democratic diplomatic support
Table 4. Republican diplomatic support
Table 5. Democratic military support
Table 6. Republican military support
Table 7. Democratic decoupling with China
Table 8. Republican decoupling with China
Table 9. Democratic support for Taiwanese semiconductors
Table 10. Republican support for Taiwanese semiconductors
Open access
