Introduction
The welfare state literature emphasises the importance of solidarity, which is the willingness of community members to share risks with each other. A clear understanding of the territorial boundary, within which individuals can identify with whom they share risks, is a fundamental prerequisite for establishing social solidarity. In established democracies, the construction of the sovereign state predates the development of the welfare state, making the issue of territorial boundaries less prominent. For emerging democracies, however, state building is often ongoing and closely intertwined with the expansion of welfare states. In some extreme cases, statehood continues to be contested by ongoing territorial disputes.
Existing scholarship engages with the implications of territory on solidarity mostly in the context of studying the nexus between (sub-)nationalism and welfare state development (Greer, Reference Greer2010; Gniza and Wrede, Reference Gniza and Wrede2022; Vampa, Reference Vampa2017). Given that a nation is always territory seeking, whether through self-autonomy or independence (Béland and Lecours, Reference Béland and Lecours2005), the literature often conflates these concepts. However, territorial demarcation and the processes of nation-building (or the intensities of nationalism) are theoretically distinct and should not be conflated. For example, in multinational states (e.g. the United Kingdom or Belgium), the state’s territorial boundary and the contours of national identity do not coincide. Therefore, it is precipitous to presume that they exert identical influences on welfare states without empirical evidence. In addition, territorial disputes prevail worldwide, yet the influences of contested territory on the welfare state are underexplored. For instance, both North Korea and South Korea lay claim to the entire Korean peninsula; China and Taiwan do not recognise each other’s sovereignty; India and Pakistan engage in territorial conflicts over the Kashmir region; the annexation of Crimea by the Russian Federation in 2014, and the subsequent escalation of the Russian military invasion in Ukraine starting in 2022, engenders disputed territories; Kosovo unilaterally proclaim independence from Serbia in 2008, which persists without recognition from Serbia. The example of territorial disputes goes on. Although there is a vast body of research on the impact of nationalism on welfare state support (e.g. Galle and Fleischmann, Reference Galle and Fleischmann2020; Johnston et al., Reference Johnston, Banting, Kymlicka and Soroka2010), there is still a dearth of comprehensive understanding regarding how contested statehood influences solidarity and social policy support. In regions characterised by territorial disputes and a lack of clearly defined boundaries, how do divergent understandings of territorial bounds influence the development and administration of welfare state policies?
We study this question in the context of Taiwan, officially the Republic of China (ROC). Due to its complicated relationship with China, officially the People’s Republic of China (PRC), Taiwan has always had issues with its territorial boundaries. Legally speaking, the territory of the ROC (Taiwan) includes both mainland China and Taiwan.Footnote 1 However, since 1949, the ROC government has never officially ruled the mainland area. After Taiwan’s democratisation in the 1990s, the ROC government introduced additional constitutional articles that limit the boundary of effective ruling to only the Taiwan area. As such, there is an ongoing debate in Taiwan about the boundaries of its territory. Some would argue that the ROC’s territory still includes the mainland area, which is only currently ruled by another political entity, and they do not reject unification with mainland China to form a political community that includes mainlanders. They are more open to forming some forms of sovereignty with mainland China in the future. Others, on the contrary, would argue that after democratisation and the additional constitutional articles, ROC’s territory is only limited to the Taiwan area. If given the chance, they would declare Taiwan as separate and independent from mainland China. As the territorial dispute continues, the contested statehood in the Taiwan case provides an excellent opportunity to examine the impact of territorial boundaries on the development of the welfare state.
In regions with territorial disputes, this uncertainty implies that individuals may have divergent perceptions about with whom they are sharing risks not only at present but also going onward to the future. We define individuals who see themselves as sharing common risks and a common future with everyone within the state’s actual territorial boundaries as having a “territorial state identity.” Like national identity, territorial state identity includes a temporal aspect. Contrarily, those who believe the state’s current boundaries are improper or illegitimate– either wrongfully occupying others’ territories or failing to reclaim lands that should belong to it – are not having the territorial state identity.
In this article, we focus on the impacts of people’s contested territory perceptions on social security reform in Taiwan. Social security inherently encompasses a temporal dimension as well: there exists a considerable interval between present contributions and future benefits, and the future is intrinsically uncertain (Yen, Reference Yen2022). As such, risk is redistributed between different generations. Moreover, East Asia is aging faster than any other region, and Taiwan exemplifies this trend. Due to low fertility and longer lifespans, its population aged 65 years and older is expected to reach 24 per cent by 2030, becoming a super-aged society. This rapid demographic shift strains Taiwan’s social security systems, pressuring retrenchment reforms and raising pressing questions about who should bear their financial burden. Using two nationally representative survey data from the 2019 and original survey in 2023, this study finds that individuals who recognise Taiwan’s territorial boundary as strictly limited to its de facto area show stronger support for social security reforms aimed at sustainability and intergenerational justice. This effect remains even after controlling for nationalism.
The central thesis of this paper is that a clear understanding of territorial boundaries can serve as an independent and significant source of solidarity. When examining the foundations of solidarity in welfare states, it is important to consider not only the role of national identity, but also that of territorial state identity. In highlighting this often-overlooked dimension, this paper offers new insight into what constitutes solidarity in welfare states.
In most nation-states – where the boundaries of the nation and the state largely align – national identity and territorial state identity tend to reinforce one another, together forming a robust foundation for social solidarity. However, much of the existing literature on nationalism and the welfare state neglects the independent and positive role of territorial state identity, potentially leading to an overestimation of nationalism’s influence. This paper teases out the influence of territorial state identity from that of nationalism. In contrast, in multinational states where national and state boundaries do not fully coincide, sub-national identity and territorial state identity may diverge. Depending on the scope of a given social policy, either identity may serve as the dominant basis for solidarity. A key implication of this study is that in such multinational contexts, states can adopt inclusive policies that transcend internal national divisions and reflect the full territorial scope of the state. In these cases, territorial state identity can still serve as a powerful source of solidarity, offering a viable pathway to foster cohesion in diverse societies.
Two layers of solidarity in the welfare states
The legitimacy of a welfare state hinges on whether it is genuinely grounded in the concept of solidarity and the sources from which such solidarity arises. Various explanations can be found in the history of late nineteenth- and twentieth-century Europe, including the Judeo-Christian religious roots, the socialist/labour movements and the nationalist or patriotic sentiments derived from reconstruction after the Second World War or other occasions (Béland and Lecours, Reference Béland and Lecours2005; Baldwin, Reference Baldwin1990; Banting and Kymlicka, Reference Banting and Kymlicka2017; Saltman and Dubois, Reference Saltman and Dubois2004). Alongside these factors was the backdrop of industrialisation, which brought about uncontrollable social risks within the feudal system and family based economy. This period also witnessed significant social development, the emergence of a robust middle class, globalisation connecting the world and the formation of nation states as the basic unit of political entity at the global level. These factors collectively contributed to the development of prototype social insurance schemes and welfare arrangements that went beyond simple charity.
The world wars and democratisation transformed conservative welfare arrangements into modern welfare states. Wartime necessities allowed states, now wielding centralised administrative power, to intervene in previously private domains (Titmuss, Reference Titmuss1968), extending coverage once reserved for family responsibility into public provision. Democracy empowered all citizens in policymaking, generally expanding welfare benefits and ultimately converting privileges for selected groups into universal rights of citizenship (Judt, Reference Judt2006).
Within the foundational construct of solidarity, there exist at least two distinct layers pertinent to a state’s welfare configurations. The first layer comprises the solidarity necessary for public welfare arrangements. Solidarity in the first layer is institutional solidarity, referring to individuals’ willingness to support the formal institutional welfare arrangements embedded in the democratic regime of a sovereign state to facilitate the redistribution of risks and income.
In modern states, the state has garnered societal support in intervening in certain social aspects to provide decent living standards for everyone. Welfare states, especially under democracy, are unlikely to survive long-term without citizens’ consent (Brooks and Manza, Reference Brooks and Manza2007). Research shows that wealthier people tend to oppose expansive welfare states, preferring private insurance with risk-based premiums (Romer, Reference Romer1975). When a large segment of society holds these views, sustaining public social insurance becomes difficult. Erosion of institutional solidarity likely occurs when state-sponsored frameworks are deemed illegitimate, prompting people to seek family or market-based alternatives.
In addition to institutional solidarity, the content of the relevant aspects of institutional solidaristic support may vary between welfare states. Some have welfare arrangements that cover services from cradle to grave, whilst others only provide a minimal service package to ensure decent living conditions. The specific content depends on the values that citizens cherish and what they consider relevant to communal life. It also depends on their willingness to bear the costs of mutual assistance through these welfare arrangements. We call solidarity in different relevant respects of mutual assistance in welfare the second-layer solidarity. Solidarity in the second layer is social solidarity, referring to the extent to which people are willing to bear the financial burden of risk and income redistribution to take joint actions to address the common risk (Prainsack and Buyx, Reference Prainsack and Buyx2017). Social solidarity manifests itself differently depending on the social policy at discussion.
Empirically, it is possible to measure whether people are willing to carry the costs to support the welfare arrangements that aim to redistribute risk against specific hazards. Political debates on welfare states often revolve around whether and how much financial cost one should bear for others in the political community. For example, the rising far right’s welfare chauvinism in Europe, advocating limited welfare states excluding immigrants, is a debate about the scope of social solidarity (Harris and Enggist, Reference Harris and Enggist2023).
Notably, both the institutional solidarity and social solidarity in welfare contain a time dimension, that is, the implication that the institutional welfare arrangements in the relevant respects are expected to be maintained and extended for an unspecified time towards the future, crossing through many generations. There exists a pronounced willingness to carry the costs of mutual assistance between present and future generations (Yeh, Reference Yeh2022). As such, intergenerational solidarity, as a form of social solidarity, is of particular importance.Footnote 2 Amongst all welfare sectors, social security involves the highest degree of intergenerational solidarity due to its inherent redistribution of risk across varying generational cohorts. The essence of the territory problem also rests on to what extent towards the future the intergenerational solidarity should be included. Because intergenerational solidarity is deeply intertwined with territorial issues and social security, the next section details their relations and lays out the hypotheses to be tested in this paper.
Territory and solidarity in social security
A modern welfare state is based on a political community with a shared sense of institutional and social solidarity within a defined territory. The boundary, geographical and populational, in almost all cases is identical to the jurisdiction of a sovereign state, and in many cases, is identical to the boundary of a nation. In established democracies, the construction of statehood and nation precedes welfare state development, making the issue of political community boundaries less prominent.
Other than a few well-established democracies, numerous countries in the Global South have built their nascent democratic institutions together with the state-building process (Croissant and Hellmann, Reference Croissant and Hellmann2020). This phenomenon was evident in many democratisation experiments in Southeast Asia after the Second World War. In certain instances, such as those of Kosovo, Cyprus and Taiwan, statehood remains a contentious issue, and the limit of political boundaries continues to be a subject of considerable debate.
A clear understanding of territorial boundaries is crucial for the development and sustainability of a welfare state because it defines the limit of the political community. Citizens of the sovereign state are automatic members of the welfare regime, whereas others (such as migrant workers, international students and undocumented immigrants, etc.) have more variations in their eligibility for membership in the welfare community. The misalignment between the welfare state apparatus and those of statehood can create an ambiguous scope of shared risk (Yeh and Chen, Reference Yeh and Chen2020). In places with disputed statehood, residents may have varying territorial perceptions. An unclear boundary of shared risks can jeopardise the development and sustainability of the welfare state.
To our knowledge, there has been little research that examined the independent effect of territorial boundaries on welfare state development. Discussions about the relationship between territory and welfare state are commonly framed within the context of nationalism. The formation of a nation, distinct from other identities, involves not only shared attributes, whether real or imagined, but also the pursuit of territory, through self-autonomy or independence (Béland and Lecours, Reference Béland and Lecours2005). Consequently, very few studies distinguish the effects of territorial state identity from the broader influences of nationalism.
Nationalism’s relationship with the welfare state is well documented (Béland and Lecours, Reference Béland and Lecours2008). By acting as a social adhesive, nationalism forges a shared destiny amongst community members, generating moral obligations (Johnston et al., Reference Johnston, Banting, Kymlicka and Soroka2010). Miller (Reference Miller, Kymlicka and Banting2006) suggests that a strong national identity benefits welfare states by fostering empathy towards the less fortunate and enhancing interpersonal trust, essential for solidarity. Nation-building and welfare development are also intertwined (Béland and Lecours, Reference Béland and Lecours2005), with sub-state nationalism, as in Quebec and Scotland, using social policy to advance nationalist aims (Béland and Lecours, Reference Béland and Lecours2005).
In this article, we argue that perceiving territorial boundaries consistent with the de facto territorial state boundaries benefits solidarity. Such perception means that people recognise shared risks and a common future within the actual territorial state boundaries that govern the welfare state apparatus. We define those with a clear understanding and consistent perception of territorial boundaries as having territorial state identity. For them, their perceived boundaries of the territorial state, the welfare state and the population for which they are willing to carry costs to mutually assist are identical. Therefore, territorial state identity is forward-looking in nature. They would be willing to pursue a shared future together. To the contrary, those who have a perception of territorial boundaries different from the existing territorial state would think differently. They would be willing to share common risks and pursue a shared future with the population different from the boundary of the existing territorial state. For them, the existing welfare state apparatus is redistributing resources between the wrong persons.
Since nationalism positively affects the sense of solidarity and territory is a component of strong nationalism, it logically follows that territorial state identity should also enhance solidarity. This identity is critical for sustaining welfare states, as clearly defined political boundaries underpin institutional solidarity. Without clear boundaries of a political community or the perception of a shared future within that community, people are more likely to shape their preferences on the basis of self-interest. Self-interest weakens the trust and empathy needed for supporting state interventions that guarantee social rights and citizenship for everyone. Without this shared identity, wealthier individuals may prefer private options over public welfare benefits. In sum, a stable welfare state relies on a distinct, widely accepted sense of boundaries that encourages members of the political community to envision a shared future.
Hypothesis 1: Individuals who possess territorial state identity show greater support for institutional solidarity.
The issue of territorial boundaries magnifies in significance even more when considering the second layer of solidarity, especially within the domain of social security. This is due to the fact that both institutional and social solidarity, by their very nature, extend across an indefinite and boundless future. Varied interpretations of territorial boundaries also lead to different visions and projections of statehood into the future.
The pursuit of intergenerational solidarity is most evident in social security, where risk redistribution spans both social groups and generations. However, perfect cross-generational equity is theoretically unattainable due to demographic shifts. As societies age, intergenerational solidarity becomes harder to achieve. When individuals across generations support the continuation of social security programs, they demonstrate institutional solidarity (the first layer). Willingness to share costs for the other generations reflects intergenerational solidarity (the second layer). Institutional solidarity must come first, laying the groundwork for cost-sharing across generations.
We argue that a clearly defined statehood, congruent with the welfare regime apparatus, enhances the perpetual moral imperative towards mutual support essential for the sustenance of the political community. Social security’s sustainability relies on different generations collectively bearing one another’s burdens. When people recognise a distinct future for their political community – rooted in a well-understood territorial boundary – they are more inclined to share costs with others and uphold social solidarity.
Hypothesis 2: Individuals who possess territorial state identity show greater support for intergenerational solidarity.
The case: social security reform in Taiwan
In Taiwan (ROC), national identity and its stance towards China (PRC), specifically the choice between unification and independence, have emerged as distinct issues that collectively shape voters’ political preferences and behaviours (Achen and Wang, Reference Achen and Wang2017). Both the ROC and PRC constitutions lay claim to mainland China and Taiwan, and although the ROC restricted its effective governance to Taiwan after the 1990s democratisation, the underlying constitutional structure remains unchanged. Public opinion on this issue has shifted over time (Election Study Center, 2024a): in 1994, 16 per cent favoured unification and 20 per cent preferred independence, whilst by 2024 only 7 per cent supported unification and 26 per cent backed independence. The majority still favours the status quo, even as its definition evolves (Wu, Reference Wu2023). Survey data from 2024 also show that when asked to predict the future, about 30 per cent think unification is more likely, whilst 45 per cent expect Taiwanese independence (Taiwan National Security Studies, 2024).
National identity, on the contrary, has become less contentious after three decades of democratisation, with a solidifying Taiwanese identity (Election Study Center, 2024b). In 1992, only 18 per cent identified solely as Taiwanese and 26 per cent solely as Chinese. By 2024, 64 per cent identified only as Taiwanese, 30 per cent as both Taiwanese and Chinese and just 2 per cent solely as Chinese. However, national and territorial state identities differ. Even those with the same national identity can hold different territorial state identities. A nation does not automatically form a state, nor does a state necessarily correspond to a single nation. Thus, territorial state identity is especially relevant for places facing circumstances similar to Taiwan’s.
People pro-independence and pro-unification possess contrasting visions regarding the delineation of statehood. Those who advocate for independence try to maintain the sovereignty of ROC to only the island of Taiwan, along with Penghu, Kinmen and Matsu, territories currently governed by the Taiwanese state. In contrast, proponents who do not reject unification are open to the idea of forming one sovereignty state (e.g. unification, creating a confederation, etc.) with mainland China, now under authoritative control of the PRC. With a growing consensus on national identity, Taiwan’s divergent unification–independence stance and debate on its territorial boundaries offer a unique case to study their impact on welfare state development.
Currently, Taiwan’s old age security regime faces significant challenges. The accelerated aging of the baby boomer cohort, born post-Second World War, coupled with the surging expenditures across all welfare sectors, with the low fertility issue, underscores a critical demographic and economic challenge to the social security system. In 2019, 62 per cent of people were concerned about not receiving pension benefits upon retirement. This concern spans across generations. In Taiwan’s news, there is a recurring theme of the government pension system teetering on the edge of bankruptcy, frequently tied to demographic shifts. Inherent in this demographic shift, intergenerational solidarity, if present, would necessitate fiscal transfers from the younger cohorts to the elder, effectively reallocating resources from future to present generations. Or, the elderly should opt in to cut their benefit, reducing the government’s financial burden reserving resources for future generations. Given the urgency of the issue, reforming government administered social security programs has been a politically salient issue in Taiwan.
Empirical evidence
2019 social image survey
We used the 2019 Social Image Survey to examine the hypotheses. The Social Image Survey conducts two telephone surveys annually since 1999, focussing on tracing the public’s attitudes towards emerging social issues. The Social Image Survey employs stratified multi-stage random digit dialing (RDD) with probability proportional to size (PPS), along with within household sampling. This project includes all Taiwanese residents aged 18 years and older who have a telephone at home as the survey population. Using this sampling method, the results of the survey yield a representative sample of the Taiwanese population.
During 2019, the Social Image Survey (wave 1) asked the Taiwanese population about their preferences regarding pension reform, specifically seeking their opinions on the direction of social security reform, using the following two questions (Yang, Reference Yang2023): ‘Some people say that the retirement pension managed by the government will generate a lot of losses in the future. Do you agree or disagree with [raising premiums or paying more taxes for those currently employed/reducing retirement benefits for those who already receive them?]’ (emphasis added by the author).
This article aims to test the hypothesis of whether territorial state identity serves as the foundation for supporting the sustainability of the welfare state system (H1) and for social solidarity, which is intergenerational solidarity in the case of social security (H2). To this end, the paper constructs two dependent variables. The first dependent variable, institutional solidarity, takes the value of 1 if the respondents support reforming the social security system, either by paying more and/or reducing the benefit as the reform directions. Zero indicates people who prefer the status quo. Institutional solidarity is estimated using a logistic model due to its binary nature.
The second dependent variable, intergenerational solidarity, captures the preferences regarding the distribution of the social security reform burden. It measures the extent to which individuals are willing to have their own generation bear the costs of the reform, without necessarily imposing the same on the other generation. Intergenerational solidarity exists only when the elderly are willing to reduce their benefits or the youth are willing to increase their contributions to sustain the social security system. The age of 50 years is used as the threshold to differentiate between the older and younger generations, as individuals are generally perceived to be approaching retirement after 50 years old. In the United States, for example, one is allowed to make higher contributions to retirement accounts if aged 50 years and older. Intergenerational solidarity assumes a value of 1 when individuals are willing to bear the costs – that is, for the elderly, it means cutting benefits, or for the youth, it means adding contributions – to support the other generation, whilst a value of 0 indicates the contrary.
In the survey, 61 per cent support pension reform, showing a strong demand for social security changes. Opinions are similar across generations, indicating that older people are not more opposed to reform. At the disaggregate level, opinion is divided on the reform directions, as 39 per cent support the intergenerational solidarity approach, whilst 22 per cent prefer a self-interested reform.
The main independent variable of interest is the perception of territorial boundary. We construct territorial state identity combining two survey items about whether a person is willing to defend Taiwan under different circumstances (see appendix for precise wording). We chose this variable because the willingness to fight or defend Taiwan under any circumstances indicates that an individual has a clear sense of statehood and perceives it primarily in terms of its de facto territorial boundaries. The reasonable assumption is that only those with a strong sense of Taiwan being the de facto territorial state would be willing to defend Taiwan. In other words, a clear understanding of territorial boundaries is a prerequisite for the willingness to engage in self-defence. Nevertheless, this measure is also imperfect, as the willingness to engage in self-defence may be influenced by considerations of self-interest (Wu, Reference Wu2023). We address this issue in the original survey in the next section. Territorial state identity takes the value of 1 if respondents are willing to defend Taiwan under any circumstances.
The model also includes age. Older people are expected to show more opposition to reforming the social security system due to the potential risk of jeopardising the benefits they are currently enjoying. The paper uses the 50-year-old cutoff point to create a generational dummy. We also estimate all models using age groups with ranges of 10 years. The results are the same. For simplicity, we present results using the age dummy variable in the empirical analysis.
An individual’s status within the labour market may influence both the likelihood and manner of their support for pension reform. For retirees, whose income is mainly based on savings and pension benefits, it is expected that they may exhibit greater resistance to reform initiatives perceived as detrimental to their interests (Campbell, Reference Campbell2002). The variable retired encapsulates the interests of retirees. Furthermore, to assess the ramifications of the current fragmented social security system, the variable pension refers to the public pension schemes in which an individual participates. A comprehensive set of demographic variables, including male, education and income, are incorporated as standard controls.
The primary political divide in Taiwan centres on its relationship with China (Achen and Wang, Reference Achen and Wang2017). The Kuomintang (KMT) favours closer ties and eventual unification, whilst the Democratic Progressive Party (DPP) supports independence. Consequently, DPP supporters tend to have a clearer sense of the Republic of China’s (ROC) territory as confined to Taiwan, leading them to perceive common risks and a shared future more strongly than KMT supporters. Regression results from the 2019 Social Image Survey show that DPP supporters are significantly more likely to hold a territorial state identity, and in turn, support both institutional and intergenerational solidarity. Since partisanship overlaps with territorial perceptions, we run separate models that include and exclude partisanship to fully assess its effects.
Table 1 presents the results of the regression analysis. Models 1–3 conduct regression analyses with institutional solidarity as the dependent variable, each incorporating different sets of explanatory variables. Model 1 serves as the baseline specification. Model 2 encompasses a comprehensive range of control variables, whilst model 3 incorporates partisanship variables to examine any potential changes in the coefficients. Models 4–6 follow the same structure with intergenerational solidarity as the dependent variable. All results are presented in odds ratio. An odds ratio exceeding 1 indicates that exposure to the specified variable increases the probability of supporting institutional solidarity or intergenerational solidarity, whereas an odds ratio below 1 signifies that such exposure diminishes the likelihood of institutional solidarity or intergenerational solidarity.
Table 1. Regression of support for institutional solidarity and intergenerational solidarity using 2019 social image survey institutional solidarity intergenerational solidarity
Note: Standard errors in parentheses
* p < 0.05;
** p < 0.01;
*** p < 0.00.
The effect of territorial state identity is strong and consistent across model specifications. The territorial state identity emerges as one of the most significant predictors of institutional solidarity (H1 is supported). Individuals who perceive Taiwan as an independent state distinct from China, proxied by willingness to defend it in the event of a military attack, are 98% more likely (model 2) than those without the territorial state identity to support a more sustainable social security system that ensures the wellbeing of the entire society. Territorial state identity also enhances intergenerational solidarity, as indicated by models 4–6 (H2 supported), and the effect is even stronger. Territorial state identity is much more likely to back reforms on the basis of the principle of intergenerational solidarity. In conclusion, it is manifestly clear that possessing a well-defined territorial demarcation that coincides with the welfare state apparatus is critical to the sustainability of social security. By deepening the understanding of territorial boundaries that are congruent with the welfare state apparatus, there is an increase in societal endorsement for both institutional and intergenerational solidarity.
High-income individuals, who have more at stake, strongly favour pension reform and support both the public welfare framework and international solidarity principles. In contrast, age and retirement status show no significant effect on reform preferences or intergenerational solidarity, indicating no generational divides. The elderly do not uniformly oppose reform, suggesting the potential for cross-generational coalitions. Gender and education also do not influence opinions on pension reform. One’s current social security program does not affect one’s support for pension reform. Furthermore, the findings indicate that pension plays an inconsequential role in shaping reform trajectories.
As expected, DPP supporters significantly support pension reform and demonstrate greater endorsement of an intergenerational solidarity approach as a reform direction, which requires increased contributions from the younger population and reduced benefits for the older generation. Being a DPP supporter is associated with a 48 per cent and 62 per cent higher likelihood, respectively, of supporting institutional solidarity and intergenerational solidarity. In contrast, KMT supporters do not show a clear association with institutional solidarity, but it is apparent that KMT supporters are opposed to intergenerational solidarity. The effect of territorial state identity remains after including partisanship.
Using a representative Taiwanese sample from the 2019 Social Image Survey, our findings support the hypothesis that a ‘territorial state identity’ – understanding territory as limited to its de facto boundary and aligned with a welfare state framework – promotes welfare state sustainability. Specifically, individuals who view Taiwan’s boundary as distinct from mainland China and restricted to Taiwan show stronger support for maintaining the public pension system. A clear, shared understanding of Taiwan’s territorial boundary also enhances intergenerational solidarity within social security.
However, there are certain shortcomings in the existing survey. Firstly, a limitation of using existing survey data is that some survey items are not ideally worded. For instance, whilst a respondent’s propensity to defend Taiwan indicates a distinct and unequivocal perception of territorial borders, fundamentally the willingness to fight and potentially sacrifice for one’s country transcends a mere comprehension of territorial limits (Wu, Reference Wu2023). A more ideal survey instrument would limit the question to only an individual’s perception of ROC’s future territorial boundary. Furthermore, the Social Image Survey did not include questions related to nationalism, which, as the literature suggests, can significantly impact attitudes towards the welfare state (Béland and Lecours, Reference Béland and Lecours2005). Given the interrelated nature of state-building and nation-building processes, the beneficial effect of territorial state identity can be more clearly demonstrated by incorporating nationalism as a control variable.
Original survey in 2023
To this end, we collected original survey data through the Rakuten Insight platform in May 2023. This survey received approval from the Institutional Review Board (IRB) of Franklin and Marshall College, ensuring compliance with ethical standards for research involving human subjects. Rakuten Insight panel comprises 230,000 samples of Taiwanese adult citizens, representing approximately 1 per cent of the Taiwanese population. The survey used gender, age and location quotas to mirror the demographic distribution of Taiwan. We collected a total of 1,984 samples.
In our original survey, we incorporated a more explicit question asking individuals’ projections of Taiwan’s territorial boundaries in the future, with particular emphasis on whether they perceive territorial unification with mainland China as a more probable outcome for the ROC. The question explores perceptions of Taiwan’s future statehood by asking about likely outcomes rather than desires, revealing individual views on future territorial boundaries (see appendix for precise wording). It also taps onto the temporal dimension and the forward-looking nature, which constitutes the necessary characteristics of territorial state identity. Respondents are also given the choice of ‘it depends’, which can absorb all respondents who are not firm about the future territorial bounds and/or have considerations beyond territorial issues when forming their opinions.
In our survey, 19 per cent foresee unification with China, 32 per cent predict Taiwan independence and 49 per cent say it depends. With the growing military pressure from China and with the option of choosing ‘it depends’, we consider those who still choose ‘Taiwan independence is more possible’ are firmer than others in viewing the territorial boundary limited to Taiwan now and into the future. We construct territorial state identity, which takes the value of 1 if the respondents select Taiwan independence as the likely future and 0 otherwise. We also include a question specifically on nationalism (see appendix for precise wording). To ensure that empirical results are as comparable across surveys as possible, we follow the 2019 Social Image survey for all pension reform measures and construct the dependent variables in the same way.
Table 2 presents the regression results, the structure of which follows that of Table 1. In the correlational analysis, we regress institutional solidarity on individual perception of the future territorial boundary of Taiwan. Model 1 of Table shows the baseline model and model 2 includes other control variables, including nationalism. The result lends additional support to the argument that the territorial state identity increases both institutional solidarity and intergenerational solidarity. The result shows that when people have a clear idea of, and their perceived future territorial boundary aligns with the boundaries of, the welfare regime, they are significantly more invested in the sustainability of the welfare state and are more likely to support reform rooted in a solidarity approach. Nationalism, as predicted by the existing literature, is also a strong predictor of institutional solidarity. It is crucial to note that both territorial state identity and nationalism exert positive, albeit independent, influences on solidarity. After including nationalism, territorial state identity is still around 30 per cent (model 2 and model 5) more likely to support institutional solidarity and intergenerational solidarity.
Table 2. Regression of support for institutional solidarity and intergenerational solidarity using 2023 original data
Note: Standard errors in parentheses
* p < 0.05;
** p < 0.01;
*** p < 0.01.
However, the effect of territorial state identity is attenuated after including partisanship variables, differing from the findings presented in Table 1. Being a DPP supporter substantially increases the support for institutional and intergenerational solidarity, and the effect is stronger than that of territorial state identity. Because the original 2023 survey’s territorial perception variable specifically asks respondents about their views on the prospective relations between Taiwan and China, asking whether they anticipate unification or independence in the future, this question aligns closer with the concept focussed in this study when compared with the survey instrument used in the 2019 Social Image Survey. However, the belief that Taiwan will eventually move towards independence is also highly overlapped with the DPP supporter base. After all, the DPP party platform is to pursue Taiwan independence. Therefore, it is not surprising that the inclusion of DPP supporters dilutes the effect of territorial state identity.
According to the 2023 survey, the elderly and retired demographics do not exhibit opposition to social security reform; however, they show a pronounced self-interest and a lack of advocacy for reforms on the basis of intergenerational solidarity, contrasting with the outcome of the 2019 Social Image survey. More research is required to clarify the influence of the elderly population on the reform of social security. Furthermore, male and higher income earners are correlated with an increased support for institutional and intergenerational solidarity.Footnote 3
Discussion and conclusions
This article examines whether people’s territorial state identity influences their support for the sustainability of the welfare state and social solidarity, with a particular focus on the social security system in Taiwan. Taiwan presents a unique case in which its territorial boundary is still contested whilst its national identity becomes less so. Using two different survey data, this study demonstrates that territorial state identity not only enhances institutional solidarity, defined as the degree of public endorsement for state intervention in the redistribution of risks and income within designated welfare sectors, but also significantly bolsters the willingness of citizens to incur the financial burdens associated with redistribution, enhancing the level of social solidarity.
The main finding of this paper shows that the effect of territorial state identity on solidarity is distinct from – and not reducible to – the effect of nationalism. Although nationalism and national identity significantly influence solidarity, this article reveals that territorial state identity independently impacts the welfare state, enriching our understanding on the sources of solidarity (Miller, Reference Miller, Banting and Kymlicka2017). The territorial state identity is crucial for fostering intergenerational solidarity, particularly in policies such as pensions and old age security that are inherently time bound.
A promising direction for future research is to test the generalisability of this argument by examining the effects of territorial state identity in other contexts where the boundaries of statehood remain contested. At the more extreme end of the spectrum are regions where statehood is actively disputed, like Taiwan, such as Kosovo, Ukraine and Kashmir, etc. In such cases, future studies could directly assess whether perceptions of territorial boundaries influence levels of solidarity in these contested states. In less extreme but still highly relevant cases are existing multinational states, particularly those in which certain nations have actively pursued independent statehood – including Quebec in Canada, Catalonia in Spain and Scotland in the United Kingdom. In these settings, the jurisdiction of social policy often resides at the sub-national level, and the existing boundaries of state sovereignty and national identity do not overlap. These cases provide valuable empirical opportunities to disentangle the respective contributions of national identity and territorial state identity to solidarity.
Future research could also examine the impact of territorial state identity across a broader range of policy domains. Social security remains a particularly relevant area, as it inherently involves future-oriented commitments and requires a strong foundation of intergenerational solidarity. However, other policy challenges – especially those that demand long-term collective action – may also be influenced by how individuals perceive the territorial boundaries of their political community. One such area is climate change, a global yet deeply national issue. Recent studies have introduced the concept of ‘green nationalism’ and ‘reflexive green nationalism’, which frames environmental protection as a core national interest – a strategy shown to be effective in mobilising public support for climate action (Conversi, Reference Conversi2020). If linking climate action to national identity can generate constructive collective responses, it is plausible that a clear and shared sense of territorial state identity could similarly enhance solidarity and public engagement in addressing future-oriented challenges. This represents a promising avenue for future empirical investigation.
Territorial state identity matters for both institutional and intergenerational solidarity that underpin the social security scheme, as evidenced in this paper, and may also matter for broader welfare systems and the larger climate environment and ecosystem that extends beyond national borders. The findings inform the future welfare reforms and environmental advocacy actions by suggesting the instrumental values of the territorial/local/parochial roots for the universal and cosmopolitan goals.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0047279425101086.
Acknowledgements
An earlier version of this paper was presented at the 2023 Taiwan workshop at Stanford University and the 2023 Taiwan and Asia Program conference at Ramapo College. We appreciate Chelsea Chia-Chen Chou and Jeremy M. Teigen’s suggestions to the previous versions and the two anonymous reviewers. The funding for the original survey was supported by Franklin and Marshall College. All the usual disclaimer applies.
Competing interests
The authors declare none.
Open access
