“The word democratization would then have denoted the creation of new citizenries.”
–Harry Eckstein1. Introduction
Do democracy and democratisation require certain cultural conditions? Before the 1970s, many scholars believed that democracy could settle and thrive only in countries with Western and Protestant cultures. They doubted the possibility of democracy spreading to non-Western countries (Almond and Verba, Reference Almond and Verba1963; Dahl, Reference Dahl1971). As Eckstein pointed out, this is because ‘human beings do not have some natural affinity for democratic political orders; inclusive democracy, even if properly instituted, cannot be expected to develop of itself’ (Huntington, Reference Huntington1991; Linz and Stepan, Reference Linz and Stepan1996). In other words, it was widely assumed that democracy would not survive without democratic citizens (Mattes, Reference Mattes and Thompson2018).
The third wave of democratisation disrupted this widely accepted expectation. As democratic movements occurred and authoritarian governments fell in Southern Europe and Latin America, regions with long-standing Catholic traditions, as well as in Confucian East Asia, communist Eastern Europe, and even Sub-Saharan Africa, the cultural approach to democracy and democratisation substantially declined. Instead, leading scholars focused on the strategic interactions of political elite, institutional engineering, and economic conditions (Huntington, Reference Huntington1991; Linz and Stepan, Reference Linz and Stepan1996).
Over the last three decades, however, an increasing number of scholars have reexamined cultural factors in relation to democracy and democratisation. One common trend is to avoid direct use of the term political culture as a collective entity; instead, they study specific cultural elements that are analytically operationalisable. Although Inglehart declared the renaissance of political culture, as his 1988 APSA article was titled (Inglehart, Reference Inglehart1988), and Putnam published his pioneering book Making Democracy Work (Putnam, Reference Putnam1993), these were just preludes to the resurgence of political culture studies on global democratisation after years of neglect. Thanks to the globalisation of public opinion research (Norris, Reference Norris, Landman and Robinson2009), scholars have made conceptual, analytical, and theoretical innovations expanding the existing knowledge that cultural factors influence democracy. These academic efforts have increased significantly as both old and new democracies have faced democratic crises and backsliding since the early 2010s (Levitsky and Ziblatt, Reference Levitsky and Ziblatt2018; Waldner and Lust, Reference Waldner and Lust2018).
What civic virtues are relevant to democracy and democratisation? Scholars have discussed social capital, post-material values, emancipative values, tolerance, political sophistication, civility, individualism, public attitudes toward democracy, and many other factors regarding this relevance. While these cultural aspects are relevant and desirable for the healthy state of democratic politics, we contend that the most critical for democratic change and resilience at the regime level comes from political support for democracy. There are three valid reasons for this.
Conceptually, other civic virtues are not necessarily associated with a democratic regime because they do not refer to democracy, although they may contribute to the qualities of democratic politics. Practically, such cultural factors respected in the West are not widely present in the countries of the third-wave democratisation. These newer democracies and transitioning countries often lack liberal Western norms and values, whereas authoritarian legacies and influences persist (Pop-Eleches and Tucker, Reference Pop-Eleches and Tucker2017). Therefore, scholars emphasise public support for democracy as the primary cultural resource by which democratisers can drive and persist in democratic progress (Inglehart and Welzel, Reference Inglehart and Welzel2005). Theoretically, regime survival hinges on political support among the mass citizenries (Rose et al., Reference Rose, Mishler and Haerpfer1998; Claassen, Reference Claassen2020; Inglehart and Welzel, Reference Inglehart and Welzel2005). Authoritarianism demises as citizens reject it and advocate for democracy. Similarly, democracy as a regime endures as long as a majority of citizens demonstrate support for and understanding of democracy while maintaining vigilance against authoritarianism.
Building on recent contributions to the literature on democratic support and comparative democratisation, we propose an integrative typology of democratic support. By combining both affective regime preferences and cognitive understanding of democracy, this typology improves measures of democratic citizenship. Democratic citizens are expected to embrace democracy as well as reject authoritarianism, while they are required to hold a proper understanding of democracy by distinguishing liberal democratic elements from authoritarian ones. With this knowledge of democracy, citizens are able to strengthen their commitment toward democracy by appreciating it in relation to authoritarianism.
Applying the integrative typology of democratic support, this study reassesses the cultural legitimacy of democracy among 30 countries since the early 2000s. The study analyses public opinion data culled from those countries participating in both waves 5 (2005–2008) and 7 (2017–2022) of the World Values Survey (hereafter, WVS) to address three research questions: whether democracy is firmly accepted and anchored in the minds of ordinary people worldwide, how public support for democracy has evolved since the mid-2000s, and whether democratic support is associated with both levels of democracy and its change over time.
The findings of this study indicate that the cultural foundation of democracy is fragile and has weakened over the past two decades. It shows that an increasing number of citizens have been less committed to and more uninformed about democracy, particularly in the non-Western context. Moreover, the results reveal that democratic support having both affective and cognitive qualifications is closely related to the level of democracy. These findings suggest significant implications for understanding the third wave of democratisation retrospectively and democratic resilience prospectively.
2. Literature review: Democratic support and the third-wave democratisation
2.1. Defining key concepts
Because this study employs a cultural approach to examine democratic support and democracy after the third-wave democratisation, it is important to clarify the relevant concepts in the context of this study. First, the study defines democracy as a political regime in which leaders are selected and removed by citizens through free and fair elections (Huntington, Reference Huntington1991; Przeworski, Reference Przeworski2019). Although it is a minimalist conception of democracy, it requires such substantial elements as political equality, civil liberties, multi-party competition, and elite consent on election results in the political process. This minimalist conception of democracy enables scholars to investigate its causes and consequences in an analytical manner.
Second, political culture is the sum of the fundamental values, sentiments, and knowledge that give form and substance to the political process (Pye, Reference Pye1991). Because this general concept does not assume democracy, this study follows the civic culture tradition and narrowly defines political culture as a set of attitudes supportive of stable democracy. According to Almond and Verba, these attitudes yield cognitive, affective, and evaluative orientations, and the combinations of these orientations produce such distinguishable types of political culture as parochial, subjective, and civic (Almond and Verba, Reference Almond and Verba1963). To the extent that the democratic political system consists of multi-layered and multi-dimensional parts and citizens’ orientations tend to vary depending on political objects, political culture tends to produce numerous domain-specific attitudes (Norris, Reference Norris2011). Because this research examines recent democratisation and democratic backsliding, democratic support will be examined in the context of regime change.
Finally, prominent scholars of political culture and democracy have emphasised that democratisation is more than a transformation from an old authoritarian regime to democratic institutions (Linz and Stepan, Reference Linz and Stepan1996; Eckstein, Reference Eckstein, Eckstein, Fleron, Hoffmann and Reisinger1998b; Welzel, Reference Welzel2013). It is widely accepted that democratisation unfolds across three interconnected levels: institutional, cultural, and societal (Eckstein, Reference Eckstein, Eckstein, Fleron, Hoffmann and Reisinger1998b). At the institutional level, democratisation entails a transition from authoritarian governance to a political system that allows citizen participation through free and fair elections and multiparty competition. At the cultural level, democratisation is a dynamic process wherein ordinary citizens disengage from the norms and values of authoritarian politics and embrace democracy ‘as the only game in town’. (Linz and Stepan, Reference Linz and Stepan1996) At the final societal level, democratisation necessitates social construction of authority structures that allow democracy to deeply root, fostering a democratic equilibrium of social structures, culture, and politics (Whitehead, Reference Whitehead2002).
Considering this long-term, dynamic, and complex process of democratisation, it is apparent that the survival of a new democracy depends on cultivating democratic culture as well as designing institutions of democratic accountability. Given the lingering effects of old authoritarian regimes and the fact that citizens in new democracies and transitioning countries were not born to be liberal and democratic, it is hardly logical to expect that democratic politics after democratisation would consolidate without encountering crises or facing authoritarian challenges (Lindberg, Reference Lindberg2026). Creating democratic citizens is inherently more challenging than establishing democratic institutions because democratic governments cannot compel democratic citizenship or instil democratic norms and principles through indoctrination. The persistence of authoritarian legacies after democratic transition often outweighs the slow development of democratic citizenship (Pop-Eleches and Tucker, Reference Pop-Eleches and Tucker2017: 203). As a result, many third-wave democracies remain incomplete, illiberal, flawed, defective, or exhibit characteristics of delegative democracies, competitive authoritarianism, electoral autocracy, and hybrid regimes while showing a lack of democratic resilience (Levitsky and Way, Reference Levitsky and Way2024; Merkel, Reference Merkel2026).
2.2. Changing democratic support
Reflecting retrospectively on the 1990s, when the third wave of democratisation became global and expectations for subsequent waves were high, it appears that the early theoretical emphasis on cultural democratisation was not taken seriously. Echoing Francis Fukuyama’s assertion in The End of History and the Last Man (Fukuyama, Reference Fukuyama1993), some scholars documented that public support for democratic change and values was widespread, and the prospects of new democracies were not a concern (Finifter and Mickiewicz, Reference Finifter and Mickiewicz1992; Evans and Whitefield, Reference Evans and Whitefield1995). For example, based on a unique survey of Russians in 1990, Hahn found that ‘political thinking comes closer to what is found in Western industrial democracies’ and concluded that ‘Russian political culture, at least, would appear to be sufficiently hospitable to sustain democratic institutions’. (Hahn, Reference Hahn1991: 393, 421) Similar broad support for democracy was reported in other regions such as Eastern Europe (Dalton, Reference Dalton1994) and China (Nathan and Shi, Reference Nathan and Shi1993). These findings implied that democracy did not require cultural prerequisites and could thrive anywhere. At the time, democratic change was widely perceived as inevitable.
Because these studies challenged the foundational premise of political culture and democracy and redirected academic focus toward socio-economic development, institutional design, and elite interactions, subsequent scholars endeavoured to reexamine the political culture thesis in response to these challenges. In particular, they moved away from viewing political culture as a mega-concept applied to collective entities (Pye, Reference Pye1991); instead, they made three conceptual, analytical, and theoretical innovations.
First of all, a group of scholars conceptualised democratic support as a dynamic phenomenon, stressing both attachment to democracy and detachment from authoritarianism (Rose et al., Reference Rose, Mishler and Haerpfer1998; Shin, Reference Shin1999; Inglehart, Reference Inglehart2003; Mattes, Reference Mattes and Thompson2018). They highlighted a critical issue: Early survey-based research focused heavily on measuring positive preferences for democratic values and reforms, often overlooking enduring affinity for authoritarianism. Methodologically, they argued that survey responses about democracy tended to be influenced by social desirability bias, as few would openly oppose democracy when asked directly. Therefore, they suggested that, despite its utility, solely asking about democracy and democratic change might not provide a valid and reliable measure of democratic support.
The historical trajectory of the first-wave and Western democratisation was gradual, taking more than a century, whereas the third-wave democratisation followed a ‘big-bang’ mode characterised by radical introduction of democratic institutions against a backdrop of entrenched authoritarianism in politics, economy, culture, and society; thus, authoritarian legacies remain more tangible than democratic aspirations in new democracies and transitioning countries (Rose and Shin, Reference Rose and Shin2001). While it is encouraging that democratic desires are strong, as the early survey research indicated, ordinary citizens are unlikely to disconnect from authoritarian politics overnight. Due to economic, cultural, and social inertia, citizens in these contexts often harbour ambivalence toward both authoritarianism and democracy. Raised under authoritarian rule and lacking experiences with democratic politics, they view the presence of democratic façades as insufficient grounds to conclude that regime competition has ended. Therefore, committed support for democracy should require not only acceptance of democracy but also consistent rejection of authoritarian regimes (Rose et al., Reference Rose, Mishler and Haerpfer1998; Mattes, Reference Mattes and Thompson2018).
Following this dynamic concept of democratic support, scholars empirically reported that whereas ‘overt lip service to democracy is almost universal’ (Inglehart, Reference Inglehart2003), the actual detachment from authoritarian regimes is far slower than attachment to democracy. This finding indicates that democratisation remains incomplete and murky in the minds of ordinary citizens, especially in the phase of third-wave democratisation. Moreover, recent studies highlighted that openness to various forms of authoritarianism, such as strongman rule and military dictatorship, had been increasing among younger generations, even within Western democracies (Foa and Mounk, Reference Foa and Mounk2016; Malka et al., Reference Malka, Lelkes, Bakker and Spivack2022; Claassen and Magalhães, Reference Claassen and Magalhães2023). These findings indicate that Western democracies are not unassailable from democratic backsliding and new democracies are experiencing deconsolidation. Accordingly, there has been a resurgence of authoritarian tendencies disguised as democratic façades, as ordinary people have grown tired of democracy and expressed a preference for authoritarianism (Lindberg, Reference Lindberg2026).
The second body of literature contended that existing studies predominantly focused on measuring affective support for democracy over authoritarianism, yet they neglected to explore how ordinary citizens understood democracy (Schaffer, Reference Schaffer1998; Dalton et al., Reference Dalton, Shin and Jou2007; Canache, Reference Canache2012; Welzel, Reference Welzel2013; Cho, Reference Cho2014; Ulbricht, Reference Ulbricht2018; König et al., Reference König, Siewert and Ackermann2022; Lu and Chu, Reference Lu and Chu2022). These scholars suggested that there was a discrepancy between how scholars define democracy and how ordinary people understand it. When survey research asked individual preference for democracy, it was assumed that ordinary citizens and scholars share a similar notion of democracy and that this notion did not vary across countries. As a result, international public opinion research on democratisation omitted survey items examining public understanding about democracy until the early 2000s.Footnote 1
Indeed, the assumption that ordinary citizens understand democracy in a similar way to scholars is incorrect. For example, a significant portion of the population living in many authoritarian countries believes their country is democratic (Kruse et al., Reference Kruse, Ravlik and Welzel2019). This widespread misperception underscores the questionable understanding of democracy among citizens in these contexts. Moreover, authoritarian leaders in the late 20th and early 21st centuries have not rejected the term democracy outright but instead have manipulated its meanings (Hu, Reference Hu2020). This differs from previous waves of democratisation in which dictators openly rejected liberal democracy. These recent authoritarian leaders have driven executive aggrandisement and anti-pluralist policies in the name of democracy and majoritarianism. Thus, ordinary citizens may misunderstand democracy in the context of such non-democratic elements, even while showing literal support for democracy. In such cases, democratic backsliding and authoritarian change become inevitable in the name of democracy (Kirsch and Welzel, Reference Kirsch and Welzel2019). This represents a substantial challenge for democratic resilience in practice and reveals a conceptual flaw in studying public support for democracy.
On the basis of the civic culture tradition (Almond and Verba, Reference Almond and Verba1963), which posits that political orientations encompass affective, cognitive, and evaluative dimensions, scholars have focused on the cognitive dimension of democratic support. This includes meanings (Chapman et al., Reference Chapman, Hanson, Dzutsati and DeBell2024), conceptions (Canache, Reference Canache2012), perceptions (Ulbricht, Reference Ulbricht2018), and understanding of democracy (Cho, Reference Cho2014; Lu and Chu, Reference Lu and Chu2022). These studies investigated whether affective support for democracy was cognitively anchored. Researchers found that popular meanings of democracy often lacked liberal substance, and many citizens misunderstood both democratic and authoritarian characteristics. For example, Kirsch and Welzel demonstrated that authoritarian notions of democracy often overshadowed liberal notions on the global level, suggesting that ‘authoritarian notions reverse the whole meaning of support for democracy, indeed indicating support for autocracy’. (Kirsch and Welzel, Reference Kirsch and Welzel2019: 59).
The final literature stresses political norms as critical for democracy for two reasons. First, some scholars question the methodological validity of measuring democratic support using the ‘D-word’ due to the social desirability bias caused by the term democracy (Schedler and Sarsfield, Reference Schedler and Sarsfield2007; Bratton, Reference Bratton2010; Carlin and Singer, Reference Carlin and Singer2011). Second, theoretically, democracy consists of defining principles such as limited government, contestation, and participation, and the actual practice of democracy relies on informal rules and norms that uphold these principles (Levitsky and Ziblatt, Reference Levitsky and Ziblatt2018). According to this literature, valid measures of democracy should be derived by asking citizens about these principles rather than directly asking about democracy itself because the concept of democracy lacks substance without these defining principles. This is a constructive, indirect, and inductive approach to democratic support.
Academic attention to democratic norms surged when Donald Trump was elected president and began disregarding existing norms of American democracy. Current scholarship has a consensus on the relevance of democratic norms, but there remains disagreement on what constitutes the core norms of democracy (Ahmed, Reference Ahmed2023). For example, Levitsky and Ziblatt argued that institutional forbearance and mutual tolerance among the political elite are the ‘soft guardrails’ of democratic politics and that American democracy was imperilled by the declining presence of these two norms (Levitsky and Ziblatt, Reference Levitsky and Ziblatt2018). Other researchers studied non-democratic norms such as resorting to violence (Bartels, Reference Bartels2020), political interference with investigations (Carey et al., Reference Carey, Clayton, Helmke, Nyhan, Sanders and Stokes2022), and electoral malpractice (Norris, Reference Norris2017; Graham and Svolik, Reference Graham and Svolik2020). These studies indicate that although democratic institutions are present, political norms that sustain democratic governance are often lacking in new democracies and transitioning countries.
This emerging research strand seems to be associated with those comparative scholars of the late 1990s and early 2000s who conceptualised democratic support as a two-level phenomenon: support for democracy involves both a preference for democracy at the regime level and an acceptance of democratic practices (in contrast to authoritarian ones) at the process level (Rose et al., Reference Rose, Mishler and Haerpfer1998; Shin, Reference Shin1999; Rose and Shin, Reference Rose and Shin2001). These scholars distinguished support for democracy as an ideal political system from support for democracy in practice. Given that democratic norms are informal, usually unwritten rules that govern political interactions in a democracy (Ahmed, Reference Ahmed2023), support for democracy-in-practice clearly refers to democratic norms. As democratic backsliding has become symptomatic, violating democratic norms even in advanced democracies, recent scholars have renewed popular support for such political norms as institutional forbearance, tolerance, limited government, and civil liberties (Carey et al., Reference Carey, Clayton, Helmke, Nyhan, Sanders and Stokes2022; Graham and Svolik, Reference Graham and Svolik2020).
2.3. Contributions of existing research
So far, three decades of research on political culture and the third wave of democratisation have shown significant conceptual, analytical, and theoretical contributions. Conceptually, political support for democracy is a complex phenomenon. It is dynamic because it involves both acceptance of democracy and rejection of authoritarianism. Democratic support is multidimensional because it requires not only affective support for democracy over authoritarianism but also a cognitive understanding of it. It is also multilevel because it encompasses endorsement of both the democratic regime and its practical norms.
Analytically, the globalisation of public opinion research has enabled comprehensive cross-national comparisons of democratic support using these innovative approaches (Norris, Reference Norris, Landman and Robinson2009). The WVS, the Global Barometer Surveys, and other international surveys have shown that while public support for democracy and democratic values is high across countries, rejection of authoritarianism, democratic understanding, and support for democratic norms are fragile and significantly vary across regions and over time. These findings suggest that the early thesis positing that political culture is not an obstacle to democratic change in transitioning countries is a hasty generalisation.
Theoretically, the existing literature is based on the congruence theory wherein democratic stability and democratisation tend to coincide with citizens’ support for democracy (Eckstein, Reference Eckstein, Eckstein, Fleron, Hoffmann and Reisinger1998a; Inglehart and Welzel, Reference Inglehart and Welzel2005). However, the political culture approach faces a puzzle in explaining democratic change and lacks empirical tests, because democratisation involves radical change, whereas political culture tends to change slowly. Thus, Przeworski noted that positive or negative preferences toward democracy mean little for its construction and maintenance because they are not likely to be translated into action (Przeworski, Reference Przeworski1991). Grillo and Prato proposed a paradoxical thesis suggesting that the risk of democratic backsliding increases when citizens exhibit high support for democratic values (Grillo and Prato, Reference Grillo and Prato2023). On the other hand, Claassen utilised surveys spanning 135 countries over 29 years since 1988, employing missing data imputations to create time-series data (Claassen, Reference Claassen2020). His findings demonstrate a positive effect of democratic support on subsequent democratic change and suggest that political support for democracy is linked with the survival of democracy rather than its emergence. Moreover, Foa and Mounk provided evidence of democratic deconsolidation, showing that erosion in popular preference for democracy correlates with subsequent decline from 1995 to 2018 in the actual level of democracy across advanced democracies.Footnote 2 Although the democratic debate requires additional time and data for reconciliation, these studies collectively reinforce the long-standing idea that political culture matters with regard to democracy and democratic change.
Overall, these contributions make it clear that the paradigm of political culture has returned to the centre of comparative politics and democracy. Despite these innovative and valuable contributions, the recent literature on political culture is not without limitations. Because it has pursued rigorous analytical methods, it has led to productive but divisive knowledge rather than integration. Given that the civic culture model aimed to provide an inclusive model of democratic culture using affective, cognitive, and evaluative orientations about political authorities (Almond and Verba, Reference Almond and Verba1963) and Pye stressed political culture as a mega-concept rather than specific or partial (Pye, Reference Pye1991), existing studies are limited in integrating diverse measurements and presenting comprehensive stories about political culture and democracy.
This study does not intend to address all the recent contributions or limitations of the political culture literature on democracy. Rather, its aim is modest. The study seeks to integrate the affective support for democracy over authoritarianism and the cognitive understanding of democracy. It aims to reassess democratic legitimacy, often known as the cultural foundation of democracy. Furthermore, the study intends to examine whether the integrative model of democratic support is congruent with the level of democracy and its change after the third wave of democratisation.
3. Analytical framework: Integrative typology of democratic support
The existing literature on political culture and democracy has extensively examined affective support for democracy over authoritarianism and cognitive understandings of democracy in assessing the cultural legitimacy of democracy. For example, Bratton and his colleagues conceptualised democratic support through the demand for democracy and combined both support for democracy and the rejection of authoritarian rule (Bratton et al., Reference Bratton, Mattes and Gyimah-Boadi2005). They also identified awareness of democracy while distinguishing procedural and substantive understanding of democracy, demonstrating that awareness and procedural understanding of democracy reinforce demand for democracy (Mattes and Bratton, Reference Mattes and Bratton2007). Norris separately evaluated both democratic support and the liberal notion of democracy globally by assessing democratic legitimacy (Norris, Reference Norris2011). In this way, prior studies regarded cognitive understandings of democracy as either a distinct part of democratic legitimacy or a source of democratic support.
This study departs from the existing literature by integrating both the affective and cognitive dimensions of democratic support and applying them to assess the cultural foundations of democracy in the post-third-wave democratisation period. This approach is motivated, in part, by the diversity of democratic norms, which makes them difficult to generalise comparatively at this stage (Ahmed, Reference Ahmed2023). Thus, building on the civic culture tradition, we therefore define democratic support broadly as encompassing both affective and cognitive orientations toward democracy and against authoritarianism.
Why should cognitive understanding of democracy be incorporated into affective support for democracy? There are both analytical and theoretical reasons for doing so. Analytically, researchers are unable to verify literal support for democracy without understanding how individuals conceptualise democracy. While many scholars include the rejection of both ideal and realistic forms of authoritarianism as a way to strengthen measures of democratic support, this approach, though necessary, is not sufficient. As Kirsh and Welzel demonstrate, a substantial number of individuals express support for democracy while simultaneously holding authoritarian conceptions, a phenomenon described as regime-specific misunderstanding. In such cases, expressed support for democracy does not necessarily reflect substantive support. Accordingly, when measuring democratic support, it is essential to assess whether individuals can meaningfully distinguish democracy from authoritarianism.
The integration of affective support for democracy and cognitive understanding of democracy is theoretically significant for democratic change and resilience. Individual preference for democracy alone has limited practical relevance unless it is grounded in an informed understanding of what democracy entails, which makes it more credible and durable. Social learning theory (Bandura, Reference Bandura1977) holds that knowledge fosters a positivity bias that can, over time, solidify attitudes. This study does not claim that information alone automatically leads to authentic support for democracy. Rather, political environments marked by democratic change and crisis are likely to attract those who are more aware and possess greater knowledge of democracy, encouraging them to realign their attitudes in favour of democracy and against authoritarianism. Such dynamics are especially salient among those who are informed about democratic principles. Accordingly, affective support for democracy and cognitive understanding of democracy are closely intertwined and mutually indispensable, together forming the integrative components of democratic support.
This study adopts two steps to construct the integrative typology of democratic support while recognising ongoing conceptual and analytical debates (Claassen et al., Reference Claassen, Ackermann, Bertsou, Borba, Carlin, Cavari, Dahlum, Gherghina, Hawkins, Lelkes, Magalhães, Mattes, Meijers, Neundorf, Oross, Öztürk, Sarsfield, Self, Stanley, Tsai, Zaslove and Zechmeister2025; Mattes, Reference Mattes and Thompson2018). First, the study measures affective support for democracy by gauging whether ordinary citizens endorse democracy and reject authoritarianism. In transitioning countries, regime support takes the form of a choice between democracy and its authoritarian alternatives (Rose et al., Reference Rose, Mishler and Haerpfer1998; Shin, Reference Shin1999; Bratton et al., Reference Bratton, Mattes and Gyimah-Boadi2005). Citizens, lacking experience with democratic politics, may be uncertain whether democracy or dictatorship provides better solutions to the problems facing their societies. Consequently, citizens frequently hold ambivalent attitudes toward both democracy and authoritarianism. Recognising this ambivalence toward democracy, scholars argue that democratic support must operate in two dynamic directions: one involving a favourable preference toward democracy and another involving rejection of authoritarianism (Rose et al., Reference Rose, Mishler and Haerpfer1998; Inglehart, Reference Inglehart2003). These two directions can yield four types of regime support, but we dichotomised them as follows: committed democrats supportive of democracy over authoritarianism and the rest, uncommitted (see Figure 1).
Integrative typology of democratic support.
Second, this study examines the cognitive component of democratic support, focusing on citizens’ understanding of democracy at the regime level. In transitional and non-Western countries, most citizens are unfamiliar with an imported concept of democracy from the West and often project traditional political meanings onto the concept (Schaffer, Reference Schaffer1998). Assessing democratic understanding involves two cognitive processes: integration (associating democracy and its defining elements) and differentiation (distinguishing democracy from authoritarian and other opposing elements) (Welzel, Reference Welzel2013; Cho, Reference Cho2014;). In this study, individuals who possess a proper understanding of democracy are termed informed citizens, whereas those who do not are labelled uninformed.
When these affective and cognitive dimensions are combined, they yield four different types of orientations toward democracy: (1) informed and committed, (2) uninformed and committed, (3) informed and uncommitted, and (4) uninformed and uncommitted. These four types provide a more manageable framework for analysing and reassessing democratic legitimacy after the third wave of global democratisation.
4. Data and measurements
To assess democratic legitimacy after the third-wave democratisation, this study utilises wave 5 (2005–2008) and wave 7 (2017–2022) of the WVS and compares them. This study adopts a selective approach in choosing countries. The WVS stands out as the sole international project to systematically investigate various orientations and attitudes toward democracy and authoritarianism across countries. Questions on affective support for democracy have been included since wave 3 (1995–1998), whereas items assessing understanding about democracy were first introduced in wave 5. However, not all countries surveyed in wave 5 continued to participate in wave 7, particularly with these repeating items for both affective and cognitive orientations toward democracy. Therefore, we have selected 30 countries for analysis that participated in both waves and repeated all the necessary questions.Footnote 3 Although those countries do not provide a fully representative picture of democratic legitimacy at the global level, they show a substantial sample across five regions: Europe and North America; East Asia; Eastern Europe and Russia; Latin America; and the Middle East and North Africa.
To measure the affective support for democracy, we focused on acceptance of democracy and rejection of authoritarianism. With regard to acceptance of democracy, we used two questions from the WVS: The first asks respondents whether they consider a democratic political system to be a good way of governance, whereas the second evaluates how important it is for them to live in a democracy (see Appendix for question items, detailed measurements, and sample countries). We dichotomised responses from a 4-point scale for the first question and a 10-point scale for the second. Respondents who indicated agreement or strong agreement that democracy is a good way of governance and highly rated the importance of living in a democratic country were classified as democrats. This represents literal support for democracy.
To combine literal support for democracy with rejection of authoritarianism, we checked whether respondents viewed strongman rule and army rule as bad governance (Foa and Mounk, Reference Foa and Mounk2016; Malka et al., Reference Malka, Lelkes, Bakker and Spivack2022; Claassen and Magalhães, Reference Claassen and Magalhães2023). Given that these two authoritarian regimes remain viable alternatives to democracy worldwide, these questions are critical in determining whether citizens have detached from authoritarianism. Respondents were not fully supportive of democracy when they did not reject both of these authoritarian regimes. Thus, we identified respondents endorsing democracy and rejecting both strongman rule and army rule as committed. Otherwise, we grouped as uncommitted. This measure is labelled committed support for democracy in the existing literature (Rose et al., Reference Rose, Mishler and Haerpfer1998; Bratton et al., Reference Bratton, Mattes and Gyimah-Boadi2005).
To measure the cognitive understanding of democracy, we used a set of four questions from the WVS (Kirsch and Welzel, Reference Kirsch and Welzel2019). As Figure 2 illustrates, the WVS asked respondents about four regime characteristics and had them evaluate how essential each of these characteristics is to democracy. As a closed-ended question, respondents evaluated four propositional statements, representing the four regime characteristics, on a 10-point scale, with 1 referring to ‘not at all an essential characteristic of democracy’ and 10 referring to ‘an essential characteristic of democracy’. The four regime characteristics are popular elections, protecting civil liberties, military takeover, and religious interpretation of law. The first two represent defining characteristics of democracy, whereas the latter two indicate non-democratic characteristics.
The cognitive mapping of informed understanding about democracy.
Analytically, the essence of democratic understanding lies in citizens’ ability to associate democratic characteristics with democracy and distinguish non-democratic ones as opposed to it (Welzel, Reference Welzel2013; Cho, Reference Cho2014). In other words, if citizens comprehend that democracy is different from authoritarianism, they are expected to assess these contrasting regime characteristics differently. To operationalise this, we dichotomised the 10-point scale into two categories: unessential (1–5) and essential (6–10). We identified respondents who correctly sorted the democratic characteristics as essential and the authoritarian characteristics as unessential as informed, whereas those who failed to demonstrate this understanding were classified as uninformed.
When these affective and cognitive dimensions were considered together, they yielded four types of democratic support, as presented in Figure 1. Although the WVS enables us to include the cognitive understanding of democracy and construct the integrative typology, it has two weaknesses in assessing democratic legitimacy across countries. For example, regional and national barometer surveys employ more extensive batteries of support for democracy and rejection of authoritarianism at both ideal and practical levels than the WVS does. They also maintain questions for procedural and substantive understanding of democracy. However, they do not allow researchers to examine an informed understanding of democracy. Moreover, it should be noted in advance that no sub-Saharan countries participated in both waves of the WVS, which indicates caution for generalisation. Thus, this study has a moderate aim in that the sample is instrumental for testing the integrative typology of democratic support across countries and assessing the cultural foundation of democracy over the last two decades. We contend that the integrative typology enhances the assessment of democratic citizens by incorporating necessary qualifications. This typology will offer valuable insights about the cultural foundation of democracy and the future of democratic resilience in the era of the third-wave autocratisation.
5. Reassessing the cultural foundation of democracy after the third-wave democratisation
How solid is the cultural foundation of democracy? Has the cultural foundation changed over the last two decades? If democratic backsliding is evident, is it aligned with public support for democracy? Is the integrative typology useful in studying political culture and democracy? Table 1 presents the distribution of ordinary citizens across 30 countries and five regions on the four types of democratic support between wave 5 and wave 7 of the WVS.
Distributions of four types of democratic support
*** Note: The differences may not be precisely calculated due to rounding percentages to one decimal place.
The general overview shows that both affective support for democracy and cognitive understanding about democracy have declined across countries, consistent with findings from other studies (Foa and Mounk, Reference Foa and Mounk2016; Malka et al., Reference Malka, Lelkes, Bakker and Spivack2022). The predominant group is uncommitted and uninformed, comprising 51% in wave 5 and increasing to 56% in wave 7. Next are those committed and uninformed citizens, who constituted 17% in wave 5 and decreased to 12% in wave 7. Those from the committed and informed groups form a small minority, showing a slight decrease: 18% in wave 5 and 17% in wave 7. The final group of those uncommitted and informed increased from 14% in wave 5 to 15% in wave 7. When we take a close look at affective commitment toward democracy, 6% of citizens moved from the committed group to the uncommitted one between the two waves, showing increasing openness to authoritarianism. Moreover, the numbers indicate a loss of one-fifth of committed democrats, with the uncommitted group twice as large as the committed group. Within the cognitive understanding of democracy, changes are dynamic, but the informed group is stable at 32% in both waves. Considering that regime support is dynamic (i.e., takes the form of a choice between democracy and its alternatives, the uncommitted and uninformed are not passive but might search for authoritarian alternatives to the same degree that the committed and informed seek democratic systems. Therefore, it is reasonable to conclude that the cultural foundation of democracy is very fragile and has weakened, giving strength to authoritarianism over the last two decades.
These results have theoretical implications for the debate of democratic deconsolidation. Foa and Mounk proposed a thesis of democratic deconsolidation that citizens’ support for democracy and openness to authoritarianism have significantly increased and the democracy of the West is in danger (Foa and Mounk, Reference Foa and Mounk2016). In contrast, other scholars argue that the deconsolidation thesis is a myth, reporting that democratic support among global citizens has remained stable since the 2000s (Alexander et al., Reference Alexander, Welzel, Norris, Voeten, Foa and Mounk2017). Employing the integrative typology, our results support the deconsolidation thesis. This debate raises an important puzzle: if the deconsolidation thesis is correct, it aligns with political culture theory, which suggests that democratic backsliding has been accelerating since late 2010. Conversely, if the deconsolidation thesis is a myth, political culture theory struggles to explain the recent backsliding and democratic breakdown. While this is beyond the scope of our research, since our sample is restricted to waves 5 and 7 of the WVS and wave 8 data is not yet available, future research should explore this puzzle.
Additionally, despite the declining state of democratic support at the global level, regional variations are notable. Western Europe and North America stand out as the only regions where informed and committed democrats constitute the largest group, with their proportion increasing from 40% in wave 5 to 43% in wave 7. When combining those committed democrats, informed or uninformed, they form a majority of citizens in these regions: 59% in wave 5 and 57% in wave 7. On the other hand, the uncommitted and uninformed constitute less than one-third of the population. However, specifics vary across countries. A significant decline in committed democrats has been observed in Australia and the United States, consistent with findings from other studies (Foa and Mounk, Reference Foa and Mounk2016; Claassen and Magalhães, Reference Claassen and Magalhães2023). Germany shows a slight decline of the informed and committed democrats from 60% in wave 5 to 56% in wave 7, whereas Canada and Great Britain reveal 12% and 13% growth in the same group between the two waves. Other countries exhibit mixed trends, which do not clearly indicate concrete signs of democratic deconsolidation in the West. It appears that Western advanced countries show democratic resilience in terms of citizen politics in spite of populist and authoritarian challenges, which is consistent with recent empirical research (Wuttke et al., Reference Wuttke, Gavras and Schoen2022).
Informed and committed democrats constitute a small minority, around 10%, and their decline is evident in all other regions. For instance, committed and informed democrats are strikingly low in wave 7 in East Asia, from 0% in Vietnam and Thailand to 5% in India. The only liberal democracies of the region, Japan, Taiwan, and South Korea, show that more than one-fifth constitute committed and informed democrats, but South Korea differs from the other two because committed and informed democrats decreased from 20% in wave 5 to 12% in wave 7, and an increasing number of citizens have become uninformed and uncommitted. It might be related to the democratic crises of South Korea, which experienced two impeachments (2017 and 2025) and the self-coup crisis (2024), showing a substantial number of non-democrats and their protests (Lee, Reference Lee2025). It is notable that an 8% increase in committed and informed democrats and a 20% decrease in the uncommitted and uninformed type were found in China. Overall, those attached to authoritarianism and the uninformed about democracy are prevalent in East Asia, where economic development has been evident under authoritarian rules (Lu and Chu, Reference Lu and Chu2022).
In Eastern Europe and Russia, uncommitted and uninformed individuals make up the largest group, having increased from 57% to 66%. A total of 5 percentage points has shifted away from committed democrats towards the uncommitted group. Serbia had 22% of committed and informed democrats in wave 5 when the country was a liberal democracy, but the group decreased to 13% in wave 7, with this decrease somewhat mirroring the fact that Serbia is not a liberal democracy anymore. This indicates that scepticism and doubt about democracy are becoming more widespread among ordinary citizens in these regions. Democracy appears to be largely symbolic, having weak social roots in the former communist countries; nearly 80% of the citizens are either ambiguous toward democracy or attached to authoritarianism (Pop-Eleches and Tucker, Reference Pop-Eleches and Tucker2017).
In Latin America, the proportion of committed democrats has decreased significantly, dropping from 33% in wave 5 to 22% in wave 7. The largest increase of 12% is observed in the uncommitted and uninformed type. The rise is regionally consistent from Uruguay (11 percentage points) to Argentina (15 percentage points). From the perspective of mass politics, Latin America has experienced the most substantial shift from democracy towards authoritarianism over the last two decades. Democratic backsliding appears to have become a region-wide phenomenon (Mainwaring and Pérez-Liñán, Reference Mainwaring and Pérez-Liñán2023).
In the Middle East and North Africa, differences between the two waves are modest. Notably, Cyprus shows a 20% loss in committed and informed democrats and a 25% growth in the uncommitted and uninformed group. What is interesting, however, is that Iran shows signs of a democratic wave, with committed and informed democrats increasing from 7% in wave 5 to 15% in wave 7. Those committed to democracy, moreover, constitute 31% in wave 7. This trend appears to be linked to recent democracy protests in Iran (Bayat, Reference Bayat2023).
The distribution of the four types of democratic support varies significantly across countries, and Western countries stand out as the only region where those committed and informed democrats form a majority of the adult population. In contrast, these democrats are a small minority in non-Western regions, whereas the most populous group tends to be those of the uncommitted and uninformed type, holding ambivalence toward democracy and affinity toward authoritarianism, while lacking understanding of democracy. This disparity underscores authoritarianism’s deep roots in non-Western countries, where democracy may be perceived as symbolic at best.
If the integrative typology reveals the subjective world of democratic citizenship across countries, is it useful in ascertaining the original thesis of political culture and democracy? Political culture scholars generally accept the congruence theory, positing that political culture and regime interact to seek equilibrium (Eckstein, Reference Eckstein, Eckstein, Fleron, Hoffmann and Reisinger1998a). The concept of democratic congruence is also relevant for studying democratic resilience against authoritarian challenges because both frameworks suggest that a political system tends to pursue an equilibrium when facing external shocks and internal stressors (Merkel, Reference Merkel2026).
If the congruence theory remains valid, does democratic citizenship influence democracy and its change? Although our sample is not large enough to answer this question to a satisfactory degree, our empirical attempt is worthy of testing the comparative utility of the integrative typology of democratic support. As such, we regress the level of democracy and changes in democracy levels between the two waves of the WVS on the three measures of democratic support, including three control variables.
First, we use two representative indices of democracy: The V-dem liberal democracy index and the Freedom House regime rating. The V-dem liberal democracy index consists of two elements: (1) equality before the law and individual liberty; (2) judicial and legislative constraints on the executive. Freedom House regime rating conceptually represents political rights and civil liberties and employs 19 measures of them through an expert survey across countries. We chose these two dependent variables for several reasons. First, because both focus on procedural and institutional aspects of democracy, they differ from mass political culture. Moreover, it enables rigorous tests because two are better than one. Finally, we use a reversed Freedom House rating because the higher the rating is, the more authoritarian the country is.
We sequentially include the three measures of democratic support used in the existing literature: literal support for democracy, committed support for democracy, and committed and informed support for democracy. The key question is whether additional qualifications of democratic support improve empirical models and explain more variance of democracy.
We include three control variables known to be associated with democracy in the existing literature: GDP per capita, economic inequality, and perceived extent of democracy. As the most competitive rival theory of political culture (Putnam, Reference Putnam1993), modernisation theory establishes the relation between economic development and the increasing probability of being democratic. This theory can be summarised by Lipset’s statement: ‘the more well-to-do a nation, the greater the chances that it will sustain democracy’. (Lipset, Reference Lipset1959: 75) We use logged GDP per capita from the International Monetary Fund.Footnote 4 Moreover, economic inequality is known to influence democracy, but the relationship is under debate. Acemoglu and Robinson theoretically contend an inverted U-shaped pattern (Acemoglu and Robinson, Reference Acemoglu and Robinson2006), whereas Debowicz and his colleagues empirically demonstrate a non-linear U-shaped relation (Debowicz et al., Reference Debowicz, Dickson, MacKenzie and Sekeris2025). Regardless of which one is valid, we include both the inequality and its squared term because the expected relation is curved. Although asset inequality is an important type of economic inequality, income inequality is widely used in the literature, and we employ the Gini index from the Standardized World Income Inequality Database.Footnote 5 Finally, performance evaluation theory takes the rational choice approach, positing that democracy thrives by satisfying citizens’ demands (Evans and Whitefield, Reference Evans and Whitefield1995). In other words, democracy becomes stable because it works well. We use the perceived extent of democracy for this theory because it indicates how people evaluate their democracy. In particular, we employ one item from the WVS: How democratically is this country being governed today? Respondents are allowed to rate from 1 (not at all democratic) to 10 (completely democratic). Satisfaction with democracy is a more popular measure than the perceived extent of democracy. However, it was not asked in wave 5, and the perceived extent of democracy is also widely used to evaluate democratic performance (Crow, Reference Crow2010). Finally, we regress the change in democracy levels on the differences of independent and control variables between the two waves of the WVS.
Table 2 reports analytical results of democratic support and the level and change of democracy between the two waves of the WVS. There are several important findings. First, all theoretical models demonstrate explanatory power for the level of democracy, but they fail to do so for the change of democracy level. The explained variance of the level of democracy ranges from 63% in Model 1 to 71% in Model 3 for the V-dem liberal democracy index and from 57% in Model 7 to 60% in Model 9 for the reversed Freedom House rating. On the other hand, the explained variable of Models 4–6 and Models 10–12 is very low, below 15%. This suggests that not only political culture theory but also the other theories are limited in explaining a short-term change in democracy.
Analytical results of democratic support and democracy
Notes: Significance level †p ≤ 0.10, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001; All coefficients are standardised. In the models of change of democracy level, Jordan was dropped because the country experienced negative economic growth and log transformation cannot process the negative value: the results are similar when we use GDP per capita and include Jordan.
Second, among the three measures of democratic support, informed and committed support for democracy stands out as having the strongest effect with statistical significance. The explained variance linearly increases as democratic support sequentially has additional qualifications, such as rejection of authoritarianism and informed understanding of democracy. This has two implications. One is that literal support alone does not matter much for democracy. The other is that citizens’ attitude has a substantial effect on the level of democracy when it has both affective and cognitive orientations for democracy and against authoritarianism. Theoretically, the results are evidence supportive of the congruence theory, although our models failed to explain the short-term change of democracy between the two waves. This aligns with Claassen’s time-series analysis, suggesting that democratic support is ‘more robustly linked with the endurance of democracy than its emergence.’ (Claassen, Reference Claassen2020: 118)
Third, with regard to the control variables, logged GDP per capita is closely associated with the level of democracy. The relationship between income inequality and democracy is a U-shaped pattern rather than an inverted U-shaped one, which is in line with recent empirical research (Debowicz et al., Reference Debowicz, Dickson, MacKenzie and Sekeris2025). Performance evaluation also matters for the level of democracy, but the effect of the perceived extent of democracy is not strong in comparison with the other variables. Finally, the dummy variable of wave 7 indicates that the level of democracy slightly declined, but its effect is not statistically significant except for Model 1.
What implications do these results suggest for survey-based research on political culture and democracy? The integrative typology of democratic support and its empirical results presented in this study have conceptual, theoretical, and substantial implications.
Conceptually, a better definition and operationalisation of democratic support would provide a more accurate picture of political culture and democracy across countries. Given that this study found strong congruence between citizens’ informed and committed support and the level of liberal democracy, early survey-based research against the political culture thesis of democracy has turned out to be rather premature and appears naïve in retrospect. It failed to take the dangerous trajectory of superficial democratic transitions and the persistence of authoritarian legacies in the third wave of democratisation seriously. As a result, scholars and democratic policy-makers overlooked the emergence of non-democratic phenomena such as illiberal democracy, competitive authoritarianism, and democratic backsliding in the context of the third wave of democratisation (Carothers, Reference Carothers2002; Norris, Reference Norris2025).
Therefore, researchers should not solely rely on survey items for democracy and democratic values (Welzel, Reference Welzel2013; De Jonge, Reference De Jonge2016; Lu and Chu, Reference Lu and Chu2022). The terms democracy and democratic are inherently positive and desirable, which can lead both the political elite and citizens to hesitate in rejecting them. This phenomenon is worrisome in survey-based research of democratisation, as the social desirability associated with the term democracy may skew responses. Furthermore, authoritarian leaders manipulate the meanings of democracy to suit their purposes (Kirsch and Welzel, Reference Kirsch and Welzel2019; Hu, Reference Hu2020). Consequently, valid measures of political support for democracy necessitate additional qualifications incorporating multiple items (Claassen et al., Reference Claassen, Ackermann, Bertsou, Borba, Carlin, Cavari, Dahlum, Gherghina, Hawkins, Lelkes, Magalhães, Mattes, Meijers, Neundorf, Oross, Öztürk, Sarsfield, Self, Stanley, Tsai, Zaslove and Zechmeister2025). This study shows one example of an integrative typology that incorporates affective and cognitive dimensions of democratic citizenship.
Theoretically, the empirical findings support the congruence theory, which posits that political culture and democracy are likely to reinforce each other in a political equilibrium. In particular, the theory advances two expectations: first, that citizens’ support for democracy and the level of democracy are aligned; and second, that this support affects both democratic and autocratic change. At this stage of academic knowledge accumulation on this subject, the first expectation has largely been confirmed by various cross-national research (Inglehart, Reference Inglehart2003; Welzel, Reference Welzel2013). However, it is far from conclusive whether citizens’ support for democracy actively affects regime change (Claassen, Reference Claassen2020). This study is also limited due to the small sample size and the short time span of the WVS. This hypothesis requires further conceptual and analytical innovations, along with extensive data, for testing.
Substantially, there has been ongoing debate on whether current democratic backsliding constitutes a global storm and whether current democracies, old and new, have the capacity for resilience against it (Merkel, Reference Merkel2026; Levitsky and Way, Reference Levitsky and Way2024). The empirical results reported in this study suggest that this phenomenon varies significantly across regions, with a marked contrast between the West and other regions. Declining support for democracy among citizenries has not been pronounced in Europe and North America over the last two decades. However, it is both significant and worrisome in other regions. There remains a mile-wide gap in democratic political culture between the former and the latter.
Moreover, the historical trajectories of democratisation differ between the first wave in the West and the third wave in other regions. In the West, liberal contestation preceded electoral participation, and democratisation was a gradual process spanning more than a century (Dahl, Reference Dahl1971; Eckstein, Reference Eckstein, Eckstein, Fleron, Hoffmann and Reisinger1998b). In contrast, non-Western countries experienced abrupt transitions to mass elections against the backdrop of authoritarian legacies and forces, often lacking liberal contestation and its associated norms (Rose and Shin, Reference Rose and Shin2001). In addition, authoritarianism faded away long ago in the West, so citizens do not have direct experience with and a lived memory of authoritarian politics. However, ordinary citizens in new democracies and transitioning countries have real experience with and vivid memories of their past authoritarianism (Bernhard and Kubik, Reference Bernhard and Kubik2014). These experiences and memories tend to persist, weakening democratic resilience and wielding the potential for democratic backsliding, especially when authoritarian successor forces and populist leaders form anti-pluralist parties to rise to power (Loxton and Mainwaring, Reference Loxton and Mainwaring2018). Given these different historical trajectories of democratisation between the West and the rest and the living memory of citizens and leaders with authoritarian politics, new democracies lack democratic resilience capacity and, thus, they are more vulnerable to authoritarian change than Western democracies. Whereas Western countries have faced democratic crises such as the January 6 U.S. Capitol Attack of 2021, non-Western countries have experienced authoritarian regime change, including autocratic surveillance and human rights violations. The substantial fluctuation of citizens’ support in favour of authoritarianism in non-Western countries indicates this trajectory of the third-wave autocratisation (Lindberg, Reference Lindberg2026). This discussion is likely to be valid given that democratic backsliding and authoritarian change have accelerated further around the world since WVS wave 7.
6. Conclusion
This study began with a literature review on political culture and third-wave democratisation. During the third wave of democratisation, early research on democratic support yielded surprising evidence against the political culture thesis, reflecting the optimistic mood of the 1990s regarding the coming global triumph of democracy. Retrospectively, political culture and democracy constituted a declining paradigm until the early 2000s.
In response to the optimistic predictions of further democratisation and growing scepticism about the political culture thesis, subsequent scholars have made conceptual and analytical innovations and revisited this topic. Due to the prevalence of democratic backsliding since the 2010s, they have examined the complex nature of democratic support and have determined that political support for democracy is globally weak and fragile.
Nonetheless, few have attempted to integrate those innovations. Building on prior research and the civic culture tradition, this study proposed an integrative typology of democratic support that encompasses affective support for democracy over authoritarianism and cognitive understanding of democracy. Using this typology, we reassessed the cultural foundation of democracy after the third wave of democratisation. Analysing data from 30 countries in wave 5 (2005–2008) and wave 7 (2017–2022) of the WVS, this study produced two empirical findings. First, democracy is not well anchored effectively and cognitively at the global level, and there is an evident decline in democratic support across countries between these two periods. Regional variation is also evident, with declining support for democracy being more severe in the non-Western countries than in the Western democracies. Second, democratic support conceptualised and operationalised in the typology is congruent with the level of liberal democracy. Additional qualifications, such as affective and cognitive orientations for democracy and against authoritarianism, improve the congruence between democratic citizenship and liberal democracy.
These empirical results carry important implications. Democratic backsliding and authoritarian resurgence have entered the minds of ordinary citizens, particularly in non-Western countries. In established Western democracies, democratic backsliding may be less likely to result in outright authoritarianism, as both citizens and leaders generally lack direct experience or collective memory of authoritarian rule and tend to possess greater democratic resilience (Merkel, Reference Merkel2026). However, the authoritarian past often remains present in the collective memory and lived experience of people in non-Western democracies (Bernhard and Kubik, Reference Bernhard and Kubik2014; Pop-Eleches and Tucker, Reference Pop-Eleches and Tucker2017). In these contexts, autocratisation has already outpaced democratisation, and authoritarianism has reemerged in new and adaptive forms since the 2010s (Lindberg, Reference Lindberg2026). What remains in question is whether a small minority of democratic citizens and institutions of democratic resilience can withstand these authoritarian turns in non-Western countries.
Acknowledgements
This work was supported by the Sogang University Research Grant of 202415019.01. This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF-2024S1A5C2A02046265).
Appendix. Survey Questions, Measurements, and the Sample Countries
A. Support for Democracy
I’m going to describe various types of political systems and ask what you think about each as a way of governing this country. For each one, would you say it is a very good, fairly good, fairly bad or very bad way of governing this country?
▣ Having a democratic political system
How important is it for you to live in a country that is governed democratically? On this scale, where 1 means it is ‘not at all important’ and 10 means ‘absolutely important,’ what position would you choose?
B. Rejection of Authoritarianism
I’m going to describe various types of political systems and ask what you think about each as a way of governing this country. For each one, would you say it is a very good, fairly good, fairly bad or very bad way of governing this country?
▣ Having a strong leader who does not have to bother with parliament and elections
▣ Having the army rule
C. Understanding of Democracy
Many things may be desirable, but not all of them are essential characteristics of democracy. Please tell me for each of the following things how essential you think it is as a characteristic of democracy. Use this scale where 1 means ‘not at all an essential characteristic of democracy’ and 10 means it is definitely ‘an essential characteristic of democracy’
▣ Religious authorities interpret the laws.
▣ People choose their leaders in free elections.
▣ The army takes over when the government is incompetent.
▣ Civil rights protect people’s liberty against oppression.
D. Measurements
We sequentially measure literal support for democracy, committed support for democracy, and committed and informed support for democracy. First, literal support for democracy is binary, having 1 for those answering having democracy as very good or fairly good and living in democracy as important (6–10) and 0 for the rest. Rejection of authoritarianism is binary, having 1 for those answering both strongman rule and army rule as fairly bad or very bad, and 0 for the rest. Informed understanding of democracy is binary, having 1 for the two democratic characteristics as essential (6–10) and the two authoritarian ones as unessential (1–5), and 0 for the rest. Second, committed support for democracy is binary, having 1 for those showing literal support for democracy and rejection of authoritarianism, and 0 for the rest. Finally, committed and informed support for democracy is measured, with 1 for those holding committed support for democracy and an informed understanding of democracy, and 0 for the rest.
F. The Sample Countries
We selected those 30 countries participating in both wave 5 and wave 7 of the WVS and retaining the questions of committed support for democracy and informed understanding of democracy. Those 55 countries participating in wave 5 are as follows: Andorra, Argentina, Australia, Brazil, Bulgaria, Canada, Chile, China, Taiwan, Colombia, Cyprus, Ethiopia, Finland, France, Georgia, Germany, Ghana, Guatemala, Hong Kong, Hungary, India, Indonesia, Iran, Iraq, Italy, Japan, Jordan, South Korea, Malaysia, Mali, Mexico, Moldova, Morocco, Netherlands, New Zealand, Norway, Peru, Poland, Romania, Russia, Rwanda, Serbia, Vietnam, Slovenia, South Africa, Spain, Sweden, Switzerland, Thailand, Trinidad and Tobago, Türkiye, Ukraine, Egypt, Great Britain, United States, Burkina Faso, Uruguay, Zambia. The list of 66 participants in wave 7 is as follows: Andorra, Argentina, Australia, Bangladesh, Armenia, Bolivia, Brazil, Myanmar, Canada, Chile, China, Taiwan, Colombia, Cyprus, Czechia, Ecuador, Ethiopia, Germany, Greece, Guatemala, Hong Kong, India, Indonesia, Iran, Iraq, Japan, Kazakhstan, Jordan, Kenya, South Korea, Kyrgyzstan, Lebanon, Libya, Macau, Malaysia, Maldives, Mexico, Mongolia, Morocco, Netherlands, New Zealand, Nicaragua, Nigeria, Pakistan, Peru, Philippines, Puerto Rico, Romania, Russia, Serbia, Singapore, Slovakia, Vietnam, Zimbabwe, Tajikistan, Thailand, Tunisia, Türkiye, Ukraine, Egypt, Great Britain, United States, Uruguay, Uzbekistan, Venezuela, Northern Ireland.
Open access
