The concept of globalization, once widely embraced in the West as a solution to many issues, has now become a source of concern. In the past, many believed that the increased interconnectedness of the world would lead to progress. This sentiment was reflected in Francis Fukuyama’s thesis (Reference Fukuyama2006) on the triumph of liberal democracy and the end of history. The attacks on democracy that we are witnessing today, however, might not be a temporary setback; rather, they seem more analogous to a reversal of progress. An increasing number of individuals in Western democracies have grown sceptical of the legitimacy of democratic regimes and have developed more favourable attitudes towards authoritarian alternatives (Foa and Mounk Reference Foa and Mounk2016). While elections in the immediate post-1989 era were seen as a means of inclusion and empowerment for religious, racial and sexual minorities, they now serve to strengthen majority groups who feel threatened by the possibility of foreigners taking over their country and altering its culture and way of life (Krastev Reference Krastev and Geiselberger2017).
The recent increase in the electoral popularity of authoritarian parties and leaders in the old democracies has triggered numerous scholarly endeavours aimed at understanding the underlying factors and mechanisms of this development. Much of this research has focused on the electoral dynamics and voting patterns and elucidated the determinants of the electoral success of populist authoritarian political actors. However, these studies do not necessarily indicate any alignment between increasing support for authoritarian actors and the regime preferences of the citizenry. This is because electoral systems and demographic characteristics vary widely and thus are not quite comparable between different countries, making it difficult to identify general patterns. Additionally, voting decisions are influenced by a wide range of factors and may not accurately reflect the alignment of voters’ ideas and motivations with those of the candidates. Therefore, instead of relying on electoral measures, this study investigates individuals’ support for authoritarian regime alternatives as indications of people’s attitudes towards governance. We avoid using support for democracy as a measure due to its post-1945 normative dominance, varying definitions and interpretations, its intersubjective construction and the risk of social desirability bias skewing survey results.
We focus on 17 old consolidated democracies since the citizenry in an established democratic regime is expected to see democracy as the ‘only game in town’ (Linz and Stepan Reference Linz and Stepan1996: 5). We expect that increasing immigration, interacting with individuals’ authoritarian predisposition, is associated with a greater perception of immigrants as a threat to social order. This perception, which manifests as anti-immigrant attitudes in individuals, in turn correlates with higher support for authoritarian regimes. Our analysis supports this hypothesis, indicating that rising immigration rates interact with individuals’ authoritarian predisposition, generating patterns of anti-immigrant sentiment and potential alignment with support for authoritarian regimes. This effect is particularly pronounced among individuals with low authoritarianism, who are markedly more responsive to demographic changes. These individuals display attitudes that increasingly converge with those of high authoritarians, who were already predisposed to view authoritarian regime alternatives as legitimate governance solutions even before the emergence of immigration-related demographic shifts. While it is premature to claim that these established democracies are at immediate risk of backsliding, our findings serve as a warning signal for Western democracies.
This article proceeds as follows. We begin by examining the rise of authoritarian politics in Western democracies and highlighting recent developments. We then discuss the limitations of focusing solely on electoral studies to understand the potential undermining of democracy, arguing for the importance of analysing people’s perceptions of different regime types. Next, we explore the connection between demographic change driven by immigration and anti-immigrant attitudes and support for authoritarian alternatives, presenting our theoretical framework and hypotheses. We then detail our data sources, variable measurements and methodology. Finally, we present our results, discussing their implications for the stability of Western democracies, and conclude with suggestions for future research.
Authoritarian politics in the West
In 1995, 24% of American respondents to the World Values Survey (WVS) stated that it would be better to have a strong leader who is not constrained by the legislature and elections (Inglehart et al. Reference Inglehart2014). This percentage rose to 35.1% in the 2017 wave of the study (Haerpfer et al. Reference Haerpfer2022). Additionally, the proportion of respondents holding favourable views of army rule climbed to 19% in 2017 from 5.3% in 1995. In a recent study, Matthew Graham and Milan Svolik (Reference Graham and Svolik2020) evaluated the robustness of Americans’ adherence to democratic principles. They found that a substantial portion of the electorate is willing to prioritize partisan interests over democratic principles when the two are in direct conflict.
The increasing prevalence of authoritarian attitudes among the US public has led to significant developments in the country’s political landscape, such as the deepening divide between the Republican and Democratic parties (Hetherington and Weiler Reference Hetherington and Weiler2009), growing support for anti-Muslim policies in the name of the ‘War on Terror’ post-9/11 (Hetherington and Suhay Reference Hetherington and Suhay2011) and the election of Donald Trump as president of the United States despite his strong authoritarian tendencies (MacWilliams Reference MacWilliams2016). Trump’s allegations of electoral fraud in the 2020 elections further contributed to the erosion of democratic norms, as evidenced by a significant proportion of Republican voters believing these claims and refusing to accept the outcome of the election (Rose Reference Rose2020), leading to the events of 6 January 2021, where Trump supporters attempted to occupy the US Capitol in an effort to overturn the election results. Several members of this group reportedly planned to hang Vice President Mike Pence for not overturning the election results (NPR 2021). Yet, a survey conducted among Republican voters following the event revealed that 68% of the surveyed did not view the actions as a threat to democracy, 45% actively supported the actions of the Capitol rioters, while 52% believed that President-elect Joe Biden was the real culprit of the attack (Sanders et al. Reference Sanders, Smith and Ballard2021).
One of the most recent and notable instances of authoritarianism in Europe can be observed in the United Kingdom. In August 2019, the Queen gave her consent for Prime Minister Boris Johnson to prorogue parliament for five weeks (Bowcott et al. Reference Bowcott2019). This unusual decision, aimed at avoiding parliamentary scrutiny of the government’s plans to leave the European Union (EU), was deemed unconstitutional by the UK Supreme Court. In September 2019, Johnson expelled 21 members of parliament from his party for joining opposition parties to block a no-deal Brexit (Bienkov Reference Bienkov(2019)). Before their expulsion, anonymous sources alleged these MPs were under investigation for foreign collusion or treason. Johnson labelled objections to a no-deal Brexit as ‘surrender’ to the enemy (Applebaum Reference Applebaum2020: 102). Boris Johnson was eventually forced to resign as leader of the Conservative Party due to mounting pressure from his party. Despite his controversial and sometimes authoritarian tendencies, it was unusual for British democracy that the public had given him a comfortable majority in the 2019 general election, a level of support not seen for any Conservative leader in over 30 years.
A brief look at the Democratic Erosion Event Dataset (DEED) reveals a similar trend of increased incidents that threaten democratic rule and order in other established Western democracies (Gottlieb et al. Reference Gottlieb2022). Examination of established democracies reveals a trend of restriction on civil liberties in France, Australia and Belgium, specifically the right to protest. In some of the established democracies, including Germany, Belgium and France, the power of law enforcement agencies has been expanded beyond judicial oversight, leading to an increased potential for arbitrary use of force against the public. Additionally, in countries such as France, Belgium and the Netherlands, minority groups, particularly Muslim minorities, have increasingly been targeted as a perceived threat to security or national culture.
As democratization scholars have noted, the stability of a democratic regime depends on whether its citizens view it as the ‘only game in town’ (Linz and Stepan Reference Linz and Stepan1996: 5). This expectation is typically met in long-established democracies, where populations have been socialized into democratic norms in the post-World War II era when democracy held an unprecedented positive normative value (Denemark et al. Reference Denemark, Mattes and Niemi2016). However, as shown in Figure 1, data from the 2008 and 2017 waves of the European Values Study (EVS) for countries in Europe and data from the 2017 wave of the WVS for countries outside Europe reveal considerable levels of support for authoritarian alternatives such as dictatorship and military rule in many of these countries (EVS 2015, 2020; Haerpfer et al. Reference Haerpfer2022). For instance, around one-third of the population in the Netherlands and Great Britain, nearly 40% in the US and Ireland and over 40% in Northern Ireland and Belgium consider authoritarian regimes as viable alternatives. In France, Canada and Australia, support for authoritarian regimes is at least 25%. We argue that this phenomenon poses a potential threat to democracy, underscoring the need to examine the determinants of authoritarian regime support to better understand the current political landscape. Consequently, this phenomenon should be studied within the context of democratization literature rather than solely within electoral studies.
Figure 1. Percentage of Population Supporting Dictatorship or Army Rule by Country, Using Data from Latest Available Survey Wave
Authoritarian alternatives to democracy
The recent rise of populist authoritarian parties and leaders has generated considerable interest among scholars, with a growing number of studies seeking to understand this unexpected trend. As Anne Marthe van der Bles et al. (Reference Van der Bles2018) note, societal dissatisfaction – despite contentment with personal circumstances – drives support for extreme parties in Western countries, reflecting protest against perceived societal decline (Bles et al. Reference Van der Bles2018: 384). Additionally, Ilse Cornelis and Alain van Hiel (Reference Cornelis and Van Hiel2015) argue that individuals with anti-egalitarian attitudes, such as economic conservatism, social dominance and belief in hierarchy, are more likely to support far-right, extreme parties in Western Europe. Likewise, Van Hiel et al. (Reference Van Hiel2007) show that various authoritarianism scales account for the support for radical-right parties in the region.
In addition to research that examines individual-level factors, other studies have investigated the role of environmental factors or their interplay with individual dynamics in shaping political preferences. For example, Robert Ford and Matthew Goodwin (Reference Ford and Goodwin2010) have demonstrated that middle-aged working-class white men with low levels of formal education residing in declining industrial towns with a significant population of Muslim immigrants constitute the core supporter base of the far-right British National Party. Similarly, Eva Green et al. (Reference Green2016) have found that the presence of stigmatized immigrants from former Yugoslavia and Albania leads to an increase in radical right-wing votes for the Swiss People’s Party, a major far-right party in Switzerland. Additionally, Italo Colantone and Piero Stanig have identified exposure to Chinese imports as a key factor in explaining support both for isolationist and radical parties in Western Europe between 1988 and 2007 (Reference Colantone and Stanig2018b) and for the Brexit vote for the UK to leave the EU (Reference Colantone and Stanig2018a). Despite the diversity of explanations offered, these studies commonly focus on the factors that contribute to the electoral success of a particular candidate or political party with authoritarian leanings without any assessment of individuals’ preferences for authoritarian regimes as a legitimate way to govern their country; the assessment cannot be fully captured through the examination of electoral patterns for a few reasons.
First, it is important to note that these countries exhibit variations in key areas such as religious makeup, state–religion relations, economic development, electoral institutions, government types and sociocultural heterogeneity, and therefore electoral outcomes are not comparable in a way that connects authoritarian developments in different countries. Populist authoritarian parties and candidates in different countries employ different campaign strategies and appeal to different demographic groups. Matthijs Rooduijn (Reference Rooduijn2018) has found that there is no single type of populist voter across different contexts, highlighting the need for an in-depth analysis of individual cases to understand the local manifestations of this broader phenomenon. In other words, an exclusive focus on specific electoral processes and outcomes fails to provide the necessary analytical basis for connecting these cases to one another.
Additionally, it is important to note that there may be a disconnect between the level of authoritarian tendencies within society and the level of public support for authoritarian parties or leaders. Regardless of what party they are voting for, people’s decision to go out to vote is shaped by several factors, such as declining party identification (Abramson and Aldrich Reference Abramson and Aldrich1982), income or education level (Reiter Reference Reiter1979), party system and registration laws (Powell Reference Powell1986) and so on. Thus, many potential supporters of exclusionary parties may not end up voting for them. Furthermore, some voters of these parties may be motivated by a desire to protest and punish mainstream parties for their poor performance (Schumacher and Rooduijn Reference Schumacher and Rooduijn2013), rather than because they are seeking detachment from democracy itself. As such, it may be difficult to fully gauge the level of attitudinal support for authoritarianism within the public solely by analysing vote shares for specific parties in elections.
Figure 2 illustrates how the levels of public support for authoritarian regime alternatives (as shown in Figure 1) compare to the aggregate vote share of far-right parties in the closest election by calendar year. The findings reveal nuanced patterns across different countries. Northern Ireland, the United States, Ireland, the United Kingdom, Australia and Canada display high levels of public support for authoritarian regimes despite negligible or non-existent far-right vote shares, highlighting a disconnect between authoritarian inclinations and far-right electoral success. Countries such as the Netherlands, Germany, France, Denmark and New Zealand show moderate far-right electoral support, yet their public support for authoritarian regimes remains substantially higher. Conversely, Switzerland and Austria demonstrate substantial far-right vote shares with comparatively lower public support for authoritarian alternatives. Only three Scandinavian democracies – Sweden, Finland and Norway – exhibit a closer alignment between public support for authoritarianism and far-right vote shares. The broader trend, however, reveals pronounced discrepancies in most countries, with public support for authoritarian regimes markedly exceeding far-right vote shares. This suggests that far-right parties do not serve as the primary drivers of authoritarian inclinations within populations, and their electoral performance should not be interpreted as a reliable gauge of public support for authoritarianism.
Figure 2. Far-Right Aggregate Vote Share from the Closest Election by Calendar Year Added to Figure 1
To overcome the limitations of electoral frameworks, we propose to investigate individuals’ perceptions of different regime alternatives. While previous studies have employed standard indicators to gauge support for democracy, such as enquiring about individuals’ perceptions of democracy as an effective system of governance for their country (Cordero and Simón Reference Cordero and Simón2016), we argue that assessing individuals’ attitudes towards authoritarian regime alternatives may provide a more comprehensive understanding of the current democratic malaise. We think that it is unlikely that a substantive understanding of the normative legitimacy of democracy would be achieved by asking individuals about their views of democracy alone for three reasons.
First, as Guillermo O’Donnell and Philippe Schmitter (Reference O’Donnell and Schmitter2013) note, the term ‘democratic’ has been widely used as a legitimizing trope in post-1945 politics, regardless of genuine adherence to democratic norms. This hegemonic value of democracy is especially pronounced in established Western democracies, where individuals have been born and socialized into democratic values and institutions since World War II (Denemark et al. Reference Denemark, Mattes and Niemi2016). Thus, declarations of allegiance to democracy may be nothing more than lip service (Inglehart Reference Inglehart2003: 52). Additionally, as Wendy Brown (Reference Brown2010) highlights, democracy is an empty signifier and every individual fills it with their hopes and dreams. While most individuals seem to support democracy, how they make sense of democracy changes from person to person and from culture to culture (Schaffer 2000). This issue can lead to non-equivalence in cross-national studies and calls into question the reliability of using support for democracy as an indicator (Stegmueller Reference Stegmueller2011).
Finally, social desirability bias may also skew interpretations of democratic legitimacy by indicating higher levels of support for democracy than there are (Phillips and Clancy Reference Phillips and Clancy1972). While one might contend that support for autocratic regimes in democratic contexts could be distorted by respondents’ reluctance to express socially undesirable views, this measure nonetheless offers a more accurate assessment of people’s regime preferences. The potential bias is likely less impactful than the pervasive tendency to overstate support for democracy, especially in an era where democracy has become a hollow, hegemonic normative concept. Moreover, the survey’s methodological design strategically circumvents this bias by framing questions about dictatorship and military regimes through descriptive language that avoids explicit terminology, thereby minimizing the social desirability effect on participants’ responses. In light of these considerations, in a similar – but not identical – fashion to several other studies (Dalton Reference Dalton2004; Inglehart and Welzel Reference Inglehart and Welzel2005; Mattes and Bratton Reference Mattes and Bratton2007; Welzel Reference Welzel2007), we suggest using survey questions that ask respondents about their attitudes towards dictatorship and military rule. To elicit accurate responses, we use survey questions that define dictatorship without explicitly mentioning the term and that enquire about army rule without priming its anti-democratic nature.
Immigration, anti-immigrant attitudes and authoritarian alternatives
After presenting the rationale for our selection of dependent variables and their operationalization, we proceed to argue for the importance of examining the relationship between immigration-related demographic change, anti-immigrant attitudes and support for authoritarian regime alternatives.
While the literature on the consequences of anti-immigrant attitudes and determinants of regime preferences are both extensive, the link between these two indicators is underexplored. In the pertinent literature, anti-immigrant attitudes are linked to the rise of far-right political parties (Arzheimer and Berning Reference Arzheimer and Berning2019), decreased support for welfare and redistribution policies (Fox Reference Fox2004), decreased trust in politics (McLaren Reference McLaren2012), changes in party affiliation (Abrajano and Hajnal Reference Abrajano and Hajnal2017) and reduced support for European integration (de Vreese and Boomgaarden Reference de Vreese and Boomgaarden2005). Previous research has examined various individual-level determinants of regime preferences such as membership in voluntary associations (Verba Reference Verba1965), trust in political institutions (Mishler and Rose Reference Mishler and Rose1997), socialization into democratic values in early childhood (Denemark et al. Reference Denemark, Mattes and Niemi2016), partisan attachment (Huang et al. Reference Huang, Chang and Chu2008), participation in electoral processes (Finkel et al. Reference Finkel, Sabatini and Bevis2000), perceptions of the economy (Cordero and Simón Reference Cordero and Simón2016), adoption of traditional/authoritarian values (Miller Reference Miller2017b) and evaluations of democratic performance (Magalhães 2014).
As Christopher Claassen succinctly noted, ‘no research has directly analyzed the link between immigration flows and public support for political systems’ prior to his recent article (Claassen Reference Claassen2024: 153). Although he does not directly analyse the association between anti-immigrant attitudes and regime preferences, Claassen’s recent study is the closest research to ours. Our study differs from Claassen’s work in several crucial respects. First, Claassen’s study adopts a regional concentration, focusing on 30 European countries. In contrast, we examine a specific cluster of 17 long-established democracies in Europe and British settler colonies as these established democracies are expected to exhibit resilient and stable support for democracy (Huang et al. Reference Huang, Chang and Chu2008). Thus we exclude recently democratized polities, such as those in Eastern Europe, which are more susceptible to democratic backsliding (Greskovits Reference Greskovits2015). Second, whereas Claassen considers support for democratic regimes (along with satisfaction with democracy and trust in political institutions) as his dependent variables, we focus on support for authoritarian regimes, such as military rule and dictatorship. We view these measures as more reliable indicators of citizens’ openness to authoritarian alternatives, as elaborated earlier. Third, while Claassen employs a national-level analysis, we conduct an individual-level analysis that enables us to work with more granular micro-level data, where we can control for individual-level confounders and examine the association between rate of immigrant population, anti-immigrant attitudes and authoritarian regime preferences more directly. Last, while Claassen uses the annual rate of immigration, our study focuses on the rate of immigrant population. This approach captures the cumulative effect of immigration over time, reflecting the long-term demographic changes that shape societal attitudes and political preferences. The overall presence of immigrants in a population better represents the extent of social and cultural changes in a society, which develop gradually rather than instantaneously.
In recent decades, Western democracies have undergone significant demographic changes, with substantial migration reshaping their social composition. This shift has also transformed the political landscape, giving rise to a new societal divide centred on immigration. As Liesbet Hooghe and Gary Marks (Reference Hooghe and Marks2018) argued, this immigration-centred cleavage was notably intensified especially after the 2008 euro crisis and the 2015 migration crisis, superseding previous societal divisions such as centre versus periphery, rural versus urban, owner versus worker and religious versus secular. The increasing ethnic and cultural diversity in Western societies thrust immigration to the forefront of public discourse, reshaping party dynamics, facilitating the rise of far-right parties critical of established norms, and catalysing anti-immigrant political movements which intensified political and social tensions surrounding immigration (Kriesi et al. Reference Kriesi2012; Schain et al. Reference Schain, Zolberg, Hossay, Schain, Zolberg and Hossay2002). This new cleavage surrounding the increasing demographic diversity alters people’s political attitudes by interacting with their underlying value orientation, such as a predisposition for authoritarianism (Bornschier Reference Bornschier2010; Teney et al. Reference Teney, Lacewell and Wilde2014).
Departing from earlier conceptualizations of authoritarianism as a fixed personal trait associated with far-right ideology (Adorno Reference Adorno2019), scholars now understand authoritarianism as a predisposition activated by situational threats to social fabric, which translates into amplified prejudice and support for coercive practices against outgroups. The interaction between stable authoritarian predisposition and contextual threats provides an analytical framework for understanding the dynamic processes of authoritarian activation. Two competing theoretical mechanisms explain this phenomenon: one model proposes that when individuals perceive a threat to established order, those with high authoritarian predisposition become more ardent supporters of authoritarian policies, while those with low authoritarian predisposition embrace more liberal positions (Feldman and Stenner Reference Feldman and Stenner1997; Stenner Reference Stenner2005). An alternative account suggests that even without explicit threats, individuals with high authoritarian predisposition already display greater support for authoritarianism, with low authoritarian individuals converging towards similar positions when they perceive social order is at risk (Hetherington and Suhay Reference Hetherington and Suhay2011; Hetherington and Weiler Reference Hetherington and Weiler2009).
Authoritarians generally tend to view the world as a dangerous place (Sibley et al. Reference Sibley2013), crave social conformity (Duckitt et al. Reference Duckitt2002) and desire order (Cornelis and Van Hiel Reference Cornelis and Van Hiel2006) to feel safe. Moreover, individuals with authoritarian tendencies demonstrate a strong inclination towards maintaining existing social structures, viewing those who challenge conventional norms as social deviants (Butler Reference Butler2009). When experiencing perceived threats to their worldview, they engage in cognitive strategies that process information selectively, deliberately seeking out sources that validate their existing beliefs and alleviate their underlying anxieties, which ultimately serves to entrench their original perspectives more deeply (Lavine et al. Reference Lavine, Lodge and Freitas2005).
Authoritarianism reflects a fundamental orientation towards societal preservation, making individuals more sensitive to maintaining group norms and social stability (Feldman Reference Feldman2003a: 46–47). As demographic diversity increases, the authoritarian predisposition of individuals is triggered, leading to a heightened sense of threat to their idealized social structure, which intensifies racial, ethnic and political intolerance (Van Assche et al. Reference Van Assche2019; Velez and Lavine Reference Velez and Lavine2017: 523–524). Immigration, as a primary agent of social diversification, triggers broader sociotropic anxieties that extend beyond individual self-interest (McLaren and Johnson Reference McLaren and Johnson2007). Immigrants are critically evaluated as potential sources of material and symbolic conflict with ingroup members, perceived not only as competitors for resources (Citrin et al. Reference Citrin1997) but also as destabilizing forces that could increase crime, social chaos and potentially erode national values and unity (Ivarsflaten Reference Ivarsflaten2005; Newman et al. Reference Newman, Hartman and Taber2012; Peresman et al. Reference Peresman, Carroll and Bäck2023). Consequently, rising immigration can trigger profound feelings of displacement, with some individuals perceiving themselves as strangers in their own lands or relegated to outsider positions (Nachtwey Reference Nachtwey and Geiselberger2017).
As diversity increases, those who feel displaced experience a profound erosion of political trust and a deepening sense of societal disintegration, perceiving established institutions as failing to maintain social cohesion (Van Assche et al. Reference Van Assche2018). The resulting feelings of social marginalization and anxiety can drive these individuals to support aggressive policies and leadership that promise to restore a broken social order (Peresman et al. Reference Peresman, Carroll and Bäck2023). People who are apprehensive of uncertainty and diversity tend to legitimize power abuse towards outgroup members and are drawn to autocratic regimes that can suppress social differences and enforce normative homogeneity (Dambrun and Vatiné Reference Dambrun and Vatiné2010; Feldman Reference Feldman, Sears, Huddy and Jervis2003b). As succinctly expressed by Steven Levitsky and Daniel Ziblatt (Reference Levitsky and Ziblatt2018: 208), ‘it is difficult to find examples of societies in which shrinking ethnic majorities gave up their dominant status without a fight’. They perceive autocratic systems as potent mechanisms to halt societal changes, restrict immigration and restore traditional status quo which undermines the public legitimacy of liberal democracy.
We posit that the influx of immigrants into Western societies serves as a perceived threat to the established order. This threat perception leads to a greater degree of anti-immigrant attitudes by interacting with individuals’ authoritarian predisposition subsequently manifesting as support for authoritarian regime alternatives. The chain reaction from the interaction between demographic change and the level of authoritarian predisposition among individuals to authoritarian regime support is mediated by anti-immigrant attitudes. This process not only explains the intensification of anti-immigrant sentiments but also elucidates the considerable appeal of authoritarian governance models in traditionally democratic societies. By highlighting the interplay between demographic shifts, psychological predispositions and political attitudes, this framework provides a comprehensive explanation for the complex relationship between immigration, authoritarianism and contemporary political landscapes in Western democracies. Based on this theoretical discussion, we propose the following hypotheses.
Hypothesis 1: Anti-immigrant attitudes mediate the relationship between the rate of immigrant population and support for dictatorship and army rule. Specifically, higher levels of immigrant population lead to stronger anti-immigrant attitudes, which in turn increase support for authoritarian regimes.
Hypothesis 2: Stronger anti-immigrant attitudes are associated with higher support for dictatorship and army rule.
Hypothesis 3: Authoritarian predisposition has a direct positive effect on both anti-immigrant attitudes and support for dictatorship and army rule, independent of the rate of immigrant population.
Hypothesis 4.1: The impact of the rate of immigrant population on anti-immigrant attitudes increases as an individual’s level of authoritarian predisposition increases.
Hypothesis 4.2: The impact of the rate of immigrant population on anti-immigrant attitudes decreases as an individual’s level of authoritarian predisposition increases.
Data and measurement
This study examines support for authoritarian regime alternatives in 17 Western European and Anglo-Saxon countries. We exclusively focus on these old Western democracies as they are characterized by a strong and stable embrace of democratic rule (Huang et al. Reference Huang, Chang and Chu2008; Linz and Stepan Reference Linz and Stepan1996), and thus their citizen are expected to be less receptive to authoritarian regime alternatives. We rely on survey data drawn from the EVS and the WVS. Following EVS methodology, we analyse Northern Ireland and Great Britain separately, treating Northern Ireland as a distinct entity in our research design regardless of its lack of sovereignty. Although both survey series provide data dating back to the 1980s, our analysis is limited to the fourth wave (2008–2009) and fifth wave (2017–2020) of the EVS and the seventh wave (2017–2020) of the WVS, as these are the waves that contain the questions relevant to our research design (EVS 2015, 2020; Haerpfer et al. Reference Haerpfer2022).
Our dependent variables measure the degree of support for authoritarian alternatives to democracy, using two subquestions from the surveys on support for different political systems. Respondents were asked to indicate their level of support for ideas such as ‘Having a strong leader who does not have to bother with parliament and elections would be …’ and ‘Having the army rule would be …’ on a scale of very bad, fairly bad, fairly good or very good. These responses were recorded as 0 to 3 respectively for our two dependent variables: dictatorship and army rule.
While each survey wave that covers the attitudes towards immigrants includes a different set of questions, we identified three survey questions for our key independent variable common to the fourth and fifth waves of the EVS and the seventh wave of the WVS. These questions ask respondents to share their opinions on the following statements: ‘When jobs are scarce, employers should give priority to natives over immigrants’, ‘Immigrants take jobs away from natives in a country’ and ‘Immigrants make crime problems worse’. Our key independent variable is an additive index of anti-immigrant attitudes which is derived from responses to these survey questions. We recoded these items to ensure consistency between the questions and across different waves of surveys and produced a scale ranging from 0 to 6, with higher scores indicating stronger anti-immigrant attitudes.
We assessed the reliability and unidimensionality of the index through Cronbach’s alpha analysis and principal component factor analysis. The Cronbach’s alpha was 0.653, slightly below the 0.7 threshold but acceptable for an exploratory social attitude measure. This indicates moderate internal consistency, supported by an average inter-item covariance of 0.248. Factor analysis revealed a single factor with an eigenvalue of 1.784, explaining 59.47% of total variance. Item loadings were strong: 0.762 (Job Scarce), 0.804 (Take Away Jobs), and 0.747 (Increase Crime), with low uniqueness values (all below 0.5). These results confirm a single underlying dimension of anti-immigrant attitudes, demonstrating the construct’s reliability and unidimensionality.
We include a scale of authoritarian predisposition based on child-rearing preferences, which is used as a standard measure in the literature on individual authoritarianism (Hetherington and Suhay Reference Hetherington and Suhay2011; Stenner Reference Stenner2005). We used a three-item index that consists of ‘independence’, ‘imagination’ and ‘obedience’. We coded responses to these items, with a score of 0 indicating that the respondent chose the item as a type of quality to encourage in children and 1 indicating that they did not. For the item ‘obedience’, we reversed the coding so that a score of 1 indicates that the respondent chose it as a desirable quality and 0 indicates they did not. We created an authoritarian predisposition index by summing the scores of these items, with higher scores indicating stronger authoritarian tendencies within individuals.
We calculated the percentage of the immigrant populations based on international immigration stock and population data from the United Nations (United Nations Department of Economic and Social Affairs, Population Division 2020, 2022). Since immigration data is available only every five years, we interpolated the data for the intervening years to provide annual estimates. We then calculated the percentage of the immigrant population for each year. As per the EVS methodology, we disentangled the immigration data for Northern Ireland (Northern Ireland Statistics and Research Agency 2023) and Great Britain from the United Kingdom. We utilized immigration data from two years prior to the survey calendar year. This approach ensures that sufficient time is allowed for the effects of demographic changes resulting from immigration trends to manifest. To address the right-skewness in the distribution of the share of immigrants, we applied a logarithmic transformation. This transformation helps to normalize the distribution, reducing the impact of extreme values and making the data more suitable for regression analysis. Additionally, the log-transformed variable resulted in lower Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) values, indicating an improved model fit.
There are two alternative indicators to actual immigrant population rates that reflect migrant presence and the extent of demographic shifts: perceived immigration rates and variation in local immigration rates, both widely used in relevant literature (Blinder and Schaffner Reference Blinder and Schaffner2020; Hoxhaj and Zuccotti Reference Hoxhaj and Zuccotti2021). While these indicators are unavailable in existing datasets in a form that allows us to test our model, we believe the actual national immigration rate aligns better with the scope, research question and design of our study for the following reasons.
With respect to perceived immigration rates, the transient and volatile nature of perceptions means that this measure lacks the analytical consistency required for robust comparative research across different time points and geographic contexts. Additionally, by utilizing actual immigration rates, we mitigate the methodological risk of artificially amplifying psychological interpretation. High authoritarians tend to perceive external changes more intensely, and their predisposition fundamentally shapes how they process stimuli. Using perceived immigration rates would overemphasize this internal psychological filter, whereas actual immigration rates offer a more objective lens for understanding how demographic shifts interact with political attitudes.
Local immigration rates prove inadequate for our dependent variables, which examine individuals’ preferences for national political community governance. These localized measurements fall short of capturing the broader dynamics shaped by national discourse, broad public sentiment and overarching societal changes that fundamentally inform our research focus. Furthermore, local variation in immigration rates is limited in capturing attitudes towards immigrants independently, as residential self-selection affects exposure to outgroups (Oliver Reference Oliver2010). Moreover, negative attitudes towards local migrants may remain depoliticized unless immigration is a salient national issue, which is best represented by the nationwide immigration rate (Hopkins Reference Hopkins2010). When immigration is nationally politicized, on the other hand, anti-immigrant politics often emerge more strongly in the regions marked by relatively less immigration concentration (Alba and Foner Reference Alba and Foner2017).
We control for alternative explanations and possible confounding effects by including additional independent variables. We included fundamental demographic characteristics, such as income, education and age, which are commonly used as control variables in survey research and are relevant factors in shaping individuals’ opinions about different regime types (Hofmann Reference Hofmann2004; Huang et al. Reference Huang, Chang and Chu2008). We also included a time variable (wave) to differentiate between surveys conducted in the 2008 wave and those conducted in the 2017 wave. The sample size for both models is 37,792 observations. Descriptive statistics for all variables employed in our analysis, including means, standard deviations, and minimum and maximum values, are presented in Table 1.
Table 1. Descriptive Statistics
Note: DV = dependent variable.
Methodology
Our theoretical model explores the potential associations between immigrant population rates, authoritarian predisposition, attitudes towards immigrants and support for authoritarian regimes. While we hypothesize potential relationships, we acknowledge the cross-sectional nature of our data limits definitive causal claims. We used a structural equation modelling (SEM) approach to model this relationship, allowing us to simultaneously estimate multiple interrelated dependence relationships. Our two-step SEM model ensures that the relationship between anti-immigrant attitudes and support for authoritarian regimes captures the variance explained by the predictors in the first step and decomposes the total effect of variables into direct and indirect effects. The first step examines how authoritarian predisposition, immigrant populations and their interaction influence anti-immigrant attitudes. The second step explores how these anti-immigrant attitudes and authoritarian predisposition influence support for authoritarian regimes. Below are the first and second steps of our model presented in statistical notation.
First step: Anti-immigrant attitudes (antiimm)
\begin{align*}antiim{m_i} &= {\beta _{{0_1}}} + {\beta _{{1_1}}}autdi{s_i} + {\beta _{{2_1}}}limmpo{p_i} + {\beta _{{3_1}}}\left( {autdi{s_i} \times limmpo{p_i}} \right)\\
&\quad + {\beta _{{4_1}}}incom{e_i} + {\beta _{{5_1}}}edu{c_i} + {\beta _{{6_1}}}ag{e_i} + {\beta _{{7_1}}}wav{e_i} + { \in _{{1_i}}},\end{align*}Second step: Support for authoritarian regime types (dictator, army)
\begin{align*}dictato{r_i} &= {\beta _{{0_2}}} + {\beta _{{1_2}}}antiim{m_i} + {\beta _{{2_2}}}autdi{s_i} + {\beta _{{3_2}}}incom{e_i} + {\beta _{{4_2}}}edu{c_i}\\
&\quad + {\beta _{{5_2}}}ag{e_i} + {\beta _{{6_2}}}wav{e_i} + { \in _{{2_i}}},\end{align*}
\begin{align*}arm{y_i} & = {\beta _{{0_2}}} + {\beta _{{1_2}}}antiim{m_i} + {\beta _{{2_2}}}autdi{s_i} + {\beta _{{3_2}}}incom{e_i} + {\beta _{{4_2}}}edu{c_i}\\
&\quad + {\beta _{{5_2}}}ag{e_i} + {\beta _{{6_2}}}wav{e_i} + { \in _{{2_i}}}.\end{align*}Results
Table 2 shows the direct, indirect and total effects of the variables in our theoretical mechanism along with the total effects of our additional controls for each dependent variable. Additionally, it includes coefficients, standard errors and standardized coefficients for each independent variable.
Table 2. Mediation Analysis
* Note: p < 0.1, ** p < 0.05, *** p < 0.01.
The results of our analysis provide support for Hypothesis 1, which posits that anti-immigrant attitudes mediate the relationship between the proportion of immigrants in the population and support for authoritarian governance. Our findings indicate that higher levels of immigrant population have modest but significant positive association with anti-immigrant attitudes, which in turn are linked to increased support for authoritarian regimes. This indirect effect translates to a small but significant increase in support for both dictatorship and army rule. The significance of this indirect effect underscores the crucial role that anti-immigrant attitudes play in translating demographic changes into regime preferences. Notably, our findings contrast with those of Claassen (Reference Claassen2024), who found no statistically significant effect of the annual rate of immigration on aggregated support for democratic regimes.
Our analysis confirms Hypothesis 2, establishing a direct connection between heightened anti-immigrant attitudes and an increased propensity to endorse authoritarian regimes. The standardized coefficients show a moderate positive relationship between the anti-immigrant attitudes and support for both dictatorship and army rule. These findings indicate that negative attitudes towards immigrants can influence political preferences for authoritarian governance both as a mediator of demographic change and as an independent factor. Alexander Kustov (Reference Kustov2023) argues that anti-immigrant groups, viewing immigration issues as crucial to personal and national well-being, are more likely to base their political opinions on these concerns. This tendency potentially creates a base for authoritarian politics, with significant implications for democratic stability. Given Harry Eckstein’s (Reference Eckstein1961) and Seymour Lipset’s (Reference Lipset1959) assertions that widespread public support for democratic rule stabilizes regimes over time, the stronger embrace of authoritarian alternatives by committed anti-immigrant individuals may erode this stabilizing factor.
Our analysis corroborates Hypothesis 3, illustrating that individuals with a higher level of authoritarian predisposition are more likely to harbour negative sentiments towards immigrants and exhibit stronger support for authoritarian governance. The standardized coefficients reveal that having authoritarian predisposition has a moderate direct effect on support for dictatorship and army rule. They also show that authoritarian predisposition is the strongest predictor for anti-immigrant attitudes in the first step. Moreover, the indirect path from authoritarian predisposition through anti-immigrant attitudes provides a modest and significant contribution to support for both dictatorship and army rule. These findings align with previous research (Miller Reference Miller2017b; Stenner Reference Stenner2005), suggesting that individuals with higher levels of authoritarian predisposition are prone to supporting authoritarian measures in response to perceived threats. These results underscore the multifaceted role of authoritarian predisposition in shaping both anti-immigrant attitudes and regime preferences. Additionally, the support for Hypotheses 1 and 2 persists despite controlling for authoritarian predisposition, highlighting the robustness of the relationships between immigrant population rates, anti-immigrant attitudes and support for authoritarian regimes.
Disconfirming Hypothesis 4.1, informed by Stanley Feldman and Karen Stenner (Reference Feldman and Stenner1997) and Stenner (Reference Stenner2005), and corroborating Hypothesis 4.2, inspired by Marc Hetherington and Jonathan Weiler (Reference Hetherington and Weiler2009) and Marc Hetherington and Elizabeth Suhay (Reference Hetherington and Suhay2011), our findings reveal a nuanced relationship between authoritarian predisposition and anti-immigrant attitudes. Specifically, higher levels of authoritarian predisposition appear to attenuate the impact of immigrant population rates on anti-immigrant sentiment, as evidenced by a significant negative interaction between authoritarian predisposition and immigrant population size. While both factors independently contribute to increased anti-immigrant attitudes, their combined effect is less than the sum of their individual impacts, suggesting a dampening interaction. This pattern implies that the perceived threat from increased immigration may be so pronounced that even individuals with low authoritarian predisposition are prone to adopting anti-immigrant views. The negative interaction coefficient thus points to a broader, more pervasive influence of immigration on attitudes across the spectrum of authoritarian predisposition.
Our models yield nuanced insights into the impact of various demographic variables. Education level emerges as the most potent factor, exerting the strongest negative effect on support for dictatorship, with a slightly less pronounced but still substantial impact on army rule. Income demonstrates a smaller yet statistically significant negative correlation with both dictatorship and army rule support. These findings resonate with research by Steven Miller (Reference Miller2017a) and Pedro Magalhães (2014), confirming that individuals with higher educational attainment and improved economic standing are less receptive to authoritarian regime alternatives. Age reveals a more complex pattern of influence. While it exhibits a moderate negative effect on support for dictatorship, the standardized coefficient for army rule is more than double, indicating an even more pronounced aversion to military governance among older generations in Western societies. Consistent with the observations of Roberto Foa and Yascha Mounk (Reference Foa and Mounk2016) and Eva Fernández Guzmán Grassi et al. (Reference Fernández Guzmán Grassi, Portos and Felicetti2024), our analysis suggests that younger individuals display a relatively higher propensity to support authoritarian regimes compared to their older counterparts. The survey wave shows a positive effect exclusively for army rule. However, this increase is not representative of a broader trend but is instead confined to localized surges in only three specific cases: Great Britain, France and Denmark, limiting its generalizability to wider populations.
Conclusion
The legitimacy of democratic systems is crucial for their stability. As Juan Linz and Alfred Stepan argue, a democracy is only consolidated when citizens view it as the ‘only game in town’ (Linz and Stepan Reference Linz and Stepan1996: 5). Widespread public support for democratic rule indeed plays a stabilizing role across time and space (Claassen Reference Claassen2020). While established Western democracies still enjoy high levels of public support (Huang et al. Reference Huang, Chang and Chu2008), recent developments such as declining trust in political institutions, weakening democratic norms, rising anti-system movements and growing disaffection with democratic processes have prompted scholars to reassess the stability of democratic rule in these countries (Levitsky and Ziblatt Reference Levitsky and Ziblatt2018; Mounk Reference Mounk2018; Norris and Inglehart Reference Norris and Inglehart2019). Our study contributes to this ongoing discussion by investigating the factors influencing people’s support for authoritarian alternatives. This approach serves as an indicator of citizens’ willingness to consider ‘other games in town’, thereby attempting to provide additional insights into understanding the current political landscape.
This article examines the relationship between immigrant population rates, anti-immigrant attitudes and preferences for authoritarian regimes in Western societies. Our results imply that increases in immigrant population rates are often perceived as a threat to the established social order and interact with individuals’ authoritarian predisposition. While high authoritarians consistently exhibit stronger anti-immigrant attitudes regardless of immigration rates, the mobilization effect is even more pronounced among low authoritarians, who seem more sensitive to rising immigration levels. Consequently, while both immigration rates and authoritarian predisposition are positively correlated with anti-immigrant attitudes, the strength of the association between each variable and anti-immigrant attitudes diminishes as the influence of the other variable increases. Our study also reveals that individuals holding strong anti-immigrant views are more likely to view authoritarian regimes as legitimate, suggesting they may desire a more authoritarian system to address their immigration-related anxieties. While our findings do not predict imminent democratic backsliding in Western democracies, they imply that the spread of anti-immigrant attitudes may challenge the stability of democratic rule. The correlation between support for anti-immigrant attitudes and openness to authoritarian alternatives increases the potential for mobilization by political actors promising a more homogeneous society.
Our results highlight the risks associated with mainstream political parties’ strategies to emulate the agendas of anti-immigrant parties in an attempt to regain lost votes. Far from preventing the rise of ethnonationalist, anti-immigrant parties (Down and Han Reference Down and Han2020), this strategy may legitimize anti-immigration attitudes in society and lead to the deterioration of political trust in politicians, parties and the country’s parliament (Geese Reference Geese2024). This electoral approach not only enables anti-immigrant parties to shape the course of politics without winning elections (Van Spanje Reference Van Spanje2010) but also has the unintended consequence of undermining the legitimacy of the democratic process itself by expanding the pool of anti-immigrant citizenry.
This study highlights several promising avenues for future research. First, our cross-sectional data design limits definitive causal inferences about the relationships between immigrant population rates, anti-immigrant attitudes and authoritarian regime preferences. Our inability to observe attitudinal changes over extended periods suggests a critical need for longitudinal studies. Systematically collecting data on immigrant-related attitudes over longer timeframes would allow researchers to track attitudinal shifts more accurately, potentially reveal long-term trends, disentangle the complex causal mechanisms at play and address potential endogeneity and bidirectional relationships more rigorously. Second, the constraints imposed by finding common survey questions across different waves limited our ability to devise more sensitive measurements. Future research could benefit from developing and implementing more comprehensive and nuanced survey instruments specifically designed to capture the complexities of anti-immigrant attitudes. Last, data limitations prevented us from examining variations in the link between anti-immigrant attitudes and support for authoritarian regime types at the subnational level. This is particularly significant given that identity formation and political developments are heavily influenced by distinct authoritarian institutions and their legacies in subnational units, as demonstrated in countries like the United States (Acharya et al. Reference Acharya, Blackwell and Sen2018) and Germany (Hildebrandt and Trüdinger Reference Hildebrandt and Trüdinger2021). Future studies could focus on capturing these subnational variations, potentially revealing more pronounced effects in certain regions and providing a more granular understanding of how local contexts shape the relationship between anti-immigrant attitudes and authoritarian preferences. Addressing these research gaps would not only build upon our findings but also provide a more comprehensive and nuanced understanding of the complex interplay between immigration, public attitudes and regime preferences in democratic societies.
Acknowledgements
The authors made use of artificial intelligence (AI) assistants in the preparation of this manuscript. Specifically, we utilized OpenAI’s ChatGPT 3.5 and GPT-4o and Anthropic’s Claude 3.5 Sonnet to assist with various aspects of drafting, editing and refining the article. These AI tools were employed to enhance clarity, improve phrasing and ensure consistency throughout the document. While these AI assistants provided valuable support, all substantive content, arguments and conclusions presented in this article are the original work and responsibility of the human authors. The final version of the manuscript was thoroughly reviewed and approved by the authors to ensure accuracy and adherence to academic standards.
Open access
