Over the past two decades, populist radical right parties (PRRPs) (Mudde Reference Mudde2007) have turned from peripheral challengers into key players in Western European party systems (Albertazzi and Vampa Reference Albertazzi and Vampa2021; Mudde Reference Mudde2019), to the extent that their alleged ‘mainstreaming’ has become a leitmotif of recent publications in comparative politics (Akkerman et al. Reference Akkerman, de Lange, Rooduijn, Akkerman, de Lange and Rooduijn2016; Brown et al. Reference Brown, Mondon and Winter2023; Crulli and Albertazzi Reference Crulli and Albertazzi2025). From Marine Le Pen’s National Rally in France (Startin Reference Startin2022) to the Sweden Democrats (Aylott and Bolin Reference Aylott and Bolin2023) and Brothers of Italy (Puleo et al. Reference Puleo, Carteny and Piccolino2025; Puleo and Piccolino Reference Puleo and Piccolino2022; Vampa Reference Vampa2023), these parties have consolidated their positions as influential actors in national and European politics, often holding roles in government. Their sustained electoral growth and inclusion in the establishment through systemic interactions with other parties (Zulianello Reference Zulianello2020) raise crucial questions about what kinds of social conflicts they mobilise and how they resonate with voters across different contexts.
From a cleavage theory perspective (Bornschier et al. Reference Bornschier, Haffert, Häusermann, Steenbergen and Zollinger2024; Hooghe and Marks Reference Hooghe and Marks2018; Hutter and Kriesi Reference Hutter and Kriesi2019; Marks et al. Reference Marks, Attewell, Rovny, Hooghe, Riddervold, Trondal and Newsome2021), the rise of PRRPs in Western Europe is indeed attributed to their ability to politicise specific social conflicts. However, the nature of these conflicts remains a subject of debate. Several scholars emphasise PRRPs’ appeal to the cultural losers of globalisation (Bornschier Reference Bornschier2010; Kriesi et al. Reference Kriesi, Grande, Lachat, Dolezal, Bornschier and Frey2006; Steiner et al. Reference Steiner, Mader and Schoen2024; Teney et al. Reference Teney, Lacewell and De Wilde2014), framing them as one pole of a globalisation-related and value-based cleavage (Bornschier et al. Reference Bornschier, Haffert, Häusermann, Steenbergen and Zollinger2024; Kriesi Reference Kriesi2010).
The literature on these parties has nevertheless not overlooked the role of traditional cleavages, towards which PRRPs have been seen either as a cause of obsolescence or as key players in their revitalisation. For instance, some have highlighted PRRPs’ support base among economically insecure individuals or the working class (Bornschier and Kriesi Reference Bornschier, Kriesi and Rydgren2012). PRRPs have also been associated with peripheral–rural interests or identities (Zollinger Reference Zollinger2024), also due to their overrepresentation outside urban centres (Crulli Reference Crulli2024). Against the proliferation of alternative explanations regarding the linkage between changing cleavage politics and PRRPs, the real impact of old and new social conflicts on the electoral fortunes of these parties, as well as the nature of these conflicts, remains a subject worthy of scholarly investigation.
This article positions itself at the intersection of scholarship on cleavage politics and on the populist radical right (PRR). Drawing on original survey data from seven Western European countries – France, Germany, Italy, the Netherlands, Spain, Sweden and the United Kingdom (UK) – we examine which conflicts PRRPs represent and whether this varies by context. In addition to the traditional Rokkanian cleavages (Lipset and Rokkan Reference Lipset, Rokkan, Lipset and Rokkan1967), we analyse the influence of ‘new’ conflicts that mark our times. Two of these can mostly be linked to the socio-economic and cultural transformations triggered by globalisation: the conflicts between skilled and unskilled workers, and between ethnonationalists and cosmopolitans. Another is mainly related to the cultural value shift in Western societies, i.e. the tension between traditionalists and supporters of LGBTQ+ rights. Finally, a further conflict can be seen as arising at the intersection of both developments: economic growth vs environmental sustainability (Crulli and Emanuele Reference Crulli and Emanuele2026). With specific items, we assess how individuals’ positions on these conflicts, combined with the salience they assign to them, predict voting for PRRPs.
Our findings show that PRRPs most strongly represent conflicts rooted in ethnonationalism, traditionalism and environmental issues. Conversely, their alignment with rural interests and economic or labour-related conflicts is weaker. Interestingly, while social characteristics such as being a production worker or living in rural areas are associated with higher support for PRRPs, this support does not appear to be rooted in strong alignment with pro-worker or pro-rural conflict sides. Instead, our analyses suggest that even working-class and rural voters are motivated more by cultural concerns than by traditional class- or place-based loyalties when supporting PRRPs. Finally, country-level variations underscore the role of national contexts in shaping PRRPs’ resonance with specific conflicts – although the aforementioned picture is fairly consistent across all countries.
The remainder of the article is structured as follows. In the next section, we provide an overview of the intense and still ongoing debate on the link between PRRPs and cleavages or social conflicts. We then describe how we expect the vote for PRRPs to be explained (or not) by the four historical cleavages and the four new conflicts we have investigated. The ‘Research design’ section presents the details of our survey and the processing of the collected data, while the ‘Results’ section describes the results from our hypothesis testing. In the ‘Conclusion’ section, we discuss the implications of our findings.
Theoretical framework and hypotheses
PRR success through the lens of (neo)cleavage theory: a critical assessment
Few theories have had a deeper impact on the study of political parties than Lipset and Rokkan’s (Reference Lipset, Rokkan, Lipset and Rokkan1967) cleavage theory. According to their account, political parties in Western Europe emerged in the nineteenth century from four fundamental social cleavages. Two were linked to the formation of the nation-state: the opposition between the Church and the central state and the conflict between the prevailing national culture and the peripheral identities. The other two were rooted in the industrial revolution: the urban–rural (industrial vs agricultural interests) and the class (owners vs workers) cleavages. Lipset and Rokkan argued that these cleavages led to a ‘freezing’ of party competition for several decades during the twentieth century (a proposition recently challenged by Emanuele and Marino (Reference Emanuele and Marino2026)).
From the 1970s onwards, however, new party families emerged around conflicts over post-materialism, environmentalism and immigration that had previously been absent or latent. The green parties were the first to capitalise on these shifts (Poguntke Reference Poguntke1987), followed by the more electorally successful PRRPs (Ignazi Reference Ignazi2003; Mudde Reference Mudde2007, Reference Mudde2019). These parties have proliferated and become established across Western Europe (Crulli and Albertazzi Reference Crulli and Albertazzi2025), playing a central role in the deinstitutionalisation of its party systems (Chiaramonte and Emanuele Reference Chiaramonte and Emanuele2022).
Given the difficulty in framing PRRPs through the lens of classic cleavage theory, scholars have turned to alternative explanations grounded in post-industrial value changes (Betz Reference Betz1993; Ignazi Reference Ignazi1992; Minkenberg Reference Minkenberg1992). Hanspeter Kriesi (Reference Kriesi1998) hypothesised that globalisation and modernisation created a new social divide between ‘winners’ and ‘losers’ of these processes, with the latter – especially the unskilled working class – forming a potential support base for PRRPs. This was further refined into the integration–demarcation cleavage (Emanuele et al. Reference Emanuele, Marino and Angelucci2020; Kriesi et al. Reference Kriesi, Grande, Lachat, Dolezal, Bornschier and Frey2006, Reference Kriesi, Grande, Lachat, Dolezal, Bornschier and Frey2008), which changed the content of the cultural and economic dimensions of party competition.
This neo-cleavage framework has intersected with debates about the social bases of PRRPs. Research shows that PRRPs overperform among the petty bourgeoisie (craftsmen, shopkeepers, small business owners) and the working class (Arzheimer Reference Arzheimer and Rydgren2012; Ivarsflaten Reference Ivarsflaten2005; Oesch and Rennwald Reference Oesch and Rennwald2018), yet their appeal is mainly driven by cultural concerns, particularly immigration (Norris and Inglehart Reference Norris and Inglehart2019; Oesch Reference Oesch2008). Even within the working class, distinctions matter. While production workers are more inclined to support PRRPs, more vulnerable service workers are not (Bornschier and Kriesi Reference Bornschier, Kriesi and Rydgren2012).
Nonetheless, there have also been scholars questioning the cleavage-based interpretation of PRRP support. For example, Wouter van der Brug and colleagues (Reference van der Brug, Fennema and Tillie2000, Reference van der Brug, Fennema and Tillie2005) found that PRRP support is better explained by attitudinal factors – especially positions on immigration – than by socio-structural characteristics. Subsequently, however, scholars in this research tradition also attributed a role to structural factors (van der Brug et al. Reference van der Brug, Fennema, de Lange, Baller and Rydgren2012).
In sum, the debate over whether PRRP support can be explained through classic and/or neo-cleavage theory remains vibrant. In the following section, we propose a framework that incorporates both perspectives to explain PRRP voting.
Hypotheses on position in social conflicts and PRRP voting
The class cleavage has been the primary dividing line in twentieth-century Western European politics, helping to homogenise party systems in the region (Caramani Reference Caramani2004). The weakening of this cleavage and the growing distance between the working class and its traditional representatives (socialist and communist parties) are well-documented (Emanuele Reference Emanuele2024; Franklin et al. Reference Franklin, Mackie and Valen1992). The concurrent rise of PRRPs has sparked debate on whether they now represent the political expression of class divides. At least since the 1990s, scholars have observed the overrepresentation of PRRPs among working-class voters (Arzheimer Reference Arzheimer and Rydgren2012; Kitschelt Reference Kitschelt1995; Mayer and Perrineau Reference Mayer and Perrineau1992; Oesch and Rennwald Reference Oesch and Rennwald2018). Although the extent of, and the reasons behind, this overrepresentation are debated (Abou-Chadi et al. Reference Abou-Chadi, Mitteregger and Mudde2021), its existence is generally acknowledged.
To make sense of such electoral changes, Herbert Kitschelt (Reference Kitschelt1995) posited a realignment in the economic preferences of some sectors of the working class – particularly those employed in the more competitive sector – although evidence pointing in the opposite direction has also emerged (Ivarsflaten Reference Ivarsflaten2005). Furthermore, over time, most PRRPs have moderated or even blurred their originally neoliberal economic stances (Rovny and Polk Reference Rovny and Polk2020), especially when compared to their much clearer positions on cultural issues such as immigration/multiculturalism, minority rights and climate change (Crulli and Albertazzi Reference Crulli and Albertazzi2025; Schwörer and Fernández-García Reference Schwörer and Fernández-García2024). Many of these parties now typically embrace welfare-chauvinist positions (de Lange Reference de Lange2007; Ennser-Jedenastik Reference Ennser-Jedenastik2018), consistent with their nativism.
Nevertheless, some arguments call for caution in drawing a simplistic link between PRRPs and economic conflict. PRRPs do have a broad, interclass appeal. Moreover, they do not politicise the class conflict in the classic, twentieth-century fashion, by pitting employers against workers, and calling for greater equity in the distribution of wealth from the former to the latter. Furthermore, radical left parties may retain their appeal among this segment of the population (Ramiro Reference Ramiro2016), and recent research suggests that while perceived job insecurity is linked to PRRP voting, actual job insecurity better predicts support for the radical left instead (Sipma et al. Reference Sipma, Lubbers and Spierings2023). It is therefore legitimate to assume that left-wing actors, especially the radical ones, may still be able to capitalise on this conflict, making the association with radical right-wing parties more nuanced.
Still, criticism of capitalists can be part of PRRPs’ agenda – for example, by contrasting the vices of financial elites with the virtues of the petty bourgeoisie (e.g. Mény and Surel Reference Mény and Surel2000), or by accusing the former of enabling mass immigration that worsens the living conditions of the working class (e.g. Piccolino and Puleo Reference Piccolino and Puleo2023).
Despite some caveats, the link between the working class and PRRPs, and the evolution of their economic agenda, lead us to hypothesise that being positioned with the workers against the owners in the traditional class conflict will increase the likelihood of voting for a PRRP.
Hypothesis 1: Being pro-workers in the owners–workers conflict increases the probability of voting for PRRPs.
As for the other traditional cleavage stemming from the industrial revolution, the electoral geography of PRRPs is characterised by a certain weakness in urban areas and strongholds typically located in rural areas. This has led to the hypothesis that the differing impact of globalisation on various territories plays a role in explaining the vote for PRRPs (Broz et al. Reference Broz, Frieden and Weymouth2021; Rodríguez-Pose Reference Rodríguez-Pose2018). Affluent cities, receptive to the economic benefits of globalisation and open to cosmopolitan values (Maxwell Reference Maxwell2019), would be less prone to embracing the exclusionary positions of PRRPs. On the other hand, the nativist and authoritarian recipe of these parties aligns more closely with the value system of rural areas (Zollinger Reference Zollinger2024), which have not benefited from globalisation in the same way. Although recent analyses paint a complex picture, urging us not to overlook the considerable variance in support for PRRPs across different geographical categories (Arzheimer et al. Reference Arzheimer, Berning, de Lange, Dutozia, Evans, Gould, Harteveld, Hood, Ivaldi, Norman, van der Brug and van der Meer2024; Harteveld et al. Reference Harteveld, van der Brug, de Lange and van der Meer2022), we can nonetheless hypothesise that aligning with the interests of rural areas will be positively associated with voting for PRRPs.
Hypothesis 2: Being pro-rural areas in the urban–rural conflict increases the probability of voting for PRRPs.
Cleavages arising from the birth of the nation-state are more difficult to frame. Starting with the state-church cleavage, while many PRRPs present themselves as defenders of traditional religious values, they cannot be described as confessional parties. Moreover, tensions between these parties and religious institutions have not been uncommon, for example, on immigration. Indeed, with some exceptions, religion has generally been seen as a factor which dampens the vote for PRRPs, as religious voters generally remain loyal to historically confessional parties (Arzheimer and Carter Reference Arzheimer and Carter2009; Immerzeel et al. Reference Immerzeel, Jaspers and Lubbers2013).
Similarly, regarding the contraposition between the national culture and ethno-linguistic peripheries within a nation, one might expect that the defence of the former would be a key explanatory factor for PRRP voting, given their nationalist stances. However, several aspects call for caution. Although there may be PRRPs that politicise this cleavage in favour of greater centralism – e.g. Vox in Spain (Vampa Reference Vampa2020) – it is entirely possible for radical right parties to emerge in direct opposition to the country’s central power, advocating for greater autonomy or even independence of some regions of the country (Martínez-Cantó and Tudó-Cisquella Reference Martínez-Cantó and Tudó-Cisquella2025). This is the case, among the parties included in our study, of the (Northern) League, even though under Salvini’s leadership it largely blurred this ethno-regionalist stance (Albertazzi et al. Reference Albertazzi, Giovannini and Seddone2018). In short, the centre–periphery conflict does not seem structurally associated with PRRPs and thus may not necessarily be a valid explanatory factor for their vote. Such considerations lead us to formulate the following third hypothesis.
Hypothesis 3: Individuals’ positions in the Rokkanian conflicts stemmed from the National Revolution – (Hypothesis 3a) centre–periphery and (Hypothesis 3b) state–church – are not related to voting for PRRPs.
The idea that the socio-economic restructuring triggered by globalisation fuelled the success of PRRPs has been present since the beginning of the debate on their voting determinants. Therefore, we should expect a certain relevance of a social conflict between the ‘winners’ (professionals and middle class operating in sectors that have benefited from increased international trade, but also skilled manual workers) and the ‘losers’ of these processes, i.e. those employed in protected economic sectors exposed to competition and/or unskilled workers (Kriesi Reference Kriesi2007: 85). Consequently, we hypothesise that the vote for PRRPs is also driven by alignment with unskilled workers when considering the economic dimension of the ‘transnational cleavage’ (Polk and Rosén Reference Polk and Rosén2024).
Hypothesis 4: Being pro-unskilled workers in the conflict between economic winners and losers from globalisation increases the probability of voting for PRRPs.
Switching to the cultural conflicts of the globalisation era, the connection between PRRPs and immigration-related issues is an enduring one. Suffice it to say that in the 1990s there was a debate over whether the label of ‘anti-immigration’ parties was suitable for them (Fennema Reference Fennema1997; Mudde Reference Mudde1999). Although this label has gradually been sidelined, nativism remains one of the core elements of their ideological foundation (Mudde Reference Mudde2007). Moreover, the literature analysing PRRP voting highlights immigration as one of the main explanations for this electoral choice (Arzheimer Reference Arzheimer and Rydgren2018). For this reason, in the conflict between ethnonationalism and cosmopolitanism, alignment with the former is a strong candidate as a predictor of voting for PRRPs.
Hypothesis 5: Being pro-ethnonationalism in the conflict between ethnonationalism and cosmopolitanism increases the probability of voting for PRRPs.
Although social conservatism is considered an ideological tenet of radical right parties (Kitschelt Reference Kitschelt1995; Mudde Reference Mudde2007), over the years there has been some evolution on this issue. Support for (some) LGBTQ+ rights has become part of the agenda of certain of these parties in the name of the defence of Western values threatened by Islam (Akkerman Reference Akkerman2015; Bracke Reference Bracke2012). However, voters who combine nativist and liberal values in a homonationalist fashion (Turnbull-Dugarte and Ortega Reference Turnbull-Dugarte and Ortega2024) remain a minority, and their support for PRRPs varies considerably across countries (Spierings et al. Reference Spierings, Lubbers and Zaslove2017). Consequently, when faced with the conflict between traditionalism and LGBTQ+ rights, we consider it likely that siding with the former will be a positive predictor in explaining the vote for PRRPs.
Hypothesis 6: Being pro-traditionalism in the traditionalism–LGBTQ+ conflict increases the probability of voting for PRRPs.
Finally, on environmental issues, the stances of PRRPs are generally clear (Lockwood Reference Lockwood2018). They usually display scepticism towards climate change mitigation measures or, at least, a more conservative stance compared to other party families. This attitude resonates with that of the voters of these parties, who also exhibit scepticism on these issues (Crulli and Zulianello Reference Crulli and Zulianello2026; Kulin et al. Reference Kulin, Johansson Sevä and Dunlap2021). Therefore, it is reasonable to assume that being in favour of those who prioritise economic growth vis-à-vis environmental sustainability will be a positive predictor of voting for PRRPs.
Hypothesis 7: Being pro-economic growth in the economy–environment conflict increases the probability of voting for PRRPs.
Our seven research hypotheses are summarised in Table 1.
Summary of Research Hypotheses
Table 1 Long description
Research hypotheses connect major political conflicts to expected voting for populist radical right parties, listing each conflict, its proposed linkage, and the hypothesis label. For workers versus owners, PRR support is expected to be higher among manual workers, and a pro-workers stance is hypothesised to raise PRR voting. For urban versus rural areas, PRR strength is linked to peripheral rural areas, and a pro-rural stance is expected to increase PRR voting. Centre versus periphery and state versus church are described as difficult to frame for PRR parties, and positions on these two conflicts are hypothesised to have no relationship with PRR voting. For skilled versus unskilled workers, PRR parties are framed as appealing to economic losers of globalisation, and a pro-unskilled stance is expected to increase PRR voting. For cosmopolitanism versus ethnonationalism and for traditionalism versus LGBTQ plus rights, nativism and social conservatism are treated as core PRR elements, so pro-ethnonationalism and pro-traditionalism are expected to increase PRR voting. For economic growth versus environment, PRR parties are said to usually oppose green policies, so a pro-growth stance is expected to increase PRR voting. These are theoretical expectations rather than observed results, and several linkages are noted as contested or hard to operationalise.
Research design
Data and operationalisation
To test our hypotheses, we relied on survey data collected as part of the NEWCOMING Project (Emanuele et al. Reference Emanuele, Guidi, Chiaramonte, Moschella, Bosco, Crulli, Improta, Lupi, Mannoni, Marino, Paparo, Piccolino, Sesti Osseo, Soare, Trastulli and Verzichelli2025). Between March and May 2024, an original, identical CAWI survey was fielded in seven countries: France, Germany, Italy, the Netherlands, Spain, Sweden and the UK. Approximately 1500 people were surveyed in each country, for a total sample size of 10,701 respondents. Beyond being representative of different Western European sub-regions and political systems, all these countries have been affected by the surge of PRRPs over the last decade, albeit to different extents. Respondents in each country were selected through a quota sample to ensure representativeness in terms of education, gender, age, geographical area and municipality size.
These survey data are suitable for this article’s purposes, as they feature questions designed to determine both individuals’ positions in social conflicts and the importance they assign to such conflicts – in addition to variables on voting intentions. Having these types of questions available is a crucial advantage for our study. Indeed, analyses on how social conflicts or cleavages structure voting behaviour are usually conducted through survey items measuring socio-demographic characteristics or attitudes on key issues. Despite some recent efforts (Bornschier et al. Reference Bornschier, Haffert, Häusermann, Steenbergen and Zollinger2024; Crulli and Emanuele Reference Crulli and Emanuele2026), there is still a lack of demand-side studies that – in line with classic cleavage theory (Bartolini and Mair Reference Bartolini and Mair1990; Knutsen and Scarbrough Reference Knutsen, Scarbrough, van Deth and Scarbrough1995) – rely on measures of individuals’ subjective alignment with the social groups involved in conflicts to better understand the role of these conflicts in explaining voter behaviour.
Our study aims to be loyal to classic cleavage theory by adopting an operationalisation that faithfully reflects its core tenets. To the best of our knowledge, this survey is the first effort to include specific items tailored to capture respondents’ positions on, and the emphasis they assign to, both classic and potentially emerging social cleavages. To give just one example, Table 2 reports the question used to assess the respondent’s stance on the traditionalism–LGBTQ+ conflict. The corresponding questions on the other conflicts are reported in Table A1 in the Supplementary material.
Survey Items Used to Assess the Individual’s Position in the Traditionalism–LGBTQ+ Conflict and the Importance Assigned to the Same Conflict
Table 2 Long description
Two survey questions assess attitudes towards a perceived conflict between traditional family values and LGBTQ plus rights, and how salient that conflict is to the respondent. The first item asks respondents to place their views on a six-point scale anchored at one end by complete support for traditional family values and at the other end by complete support for LGBTQ plus rights. The second item asks how important the conflict is personally, using an eleven-point scale from not at all important to very important. The table provides the exact question wording and the labeled endpoints for each scale. No results or distributions are included, so the table documents measurement design rather than reporting findings.
Dependent variable
The outcome variable is a dummy indicating the intention to vote for a PRRP (0 = no; 1 = yes). PRRPs were identified through a crosscheck of the PopuList (Rooduijn et al. Reference Rooduijn, Pirro, Halikiopoulou, Froio, Van Kessel, de Lange, Mudde and Taggart2023, Reference Rooduijn, Pirro, Halikiopoulou, Froio, Van Kessel, de Lange, Mudde and Taggart2024) and the Populism and Political Parties Expert Survey (POPPA) (Meijers and Zaslove Reference Meijers and Zaslove2021). Specifically, we classified as PRRPs those that the PopuList categorises as ‘populist’ and ‘far right’. Our selection differs from the PopuList classification in only two respects. First, based on select news articles, we included a small Italian party, Italexit per l’Italia, which is not mentioned in the latest versions of POPPA and the PopuList, likely due to its very limited electoral support, which falls below the threshold of relevance used by both. Second, we excluded the Dutch Farmer-Citizen Movement, as PopuList considers it a borderline case, and POPPA categorises it as an agrarian party. Table 3 reports the country-by-country list of PRRPs in the survey data we use, together with the percentage distribution of voting intentions for these parties.
List of PRRPs in the Survey and Distribution of Voting Intention for Them
Table 3 Long description
Survey results list country-level voting intention for selected populist radical right parties, giving a frequency count and a percentage for each country. The Netherlands has the highest share at 33.5 percent with 468 responses for the Party for Freedom, followed closely by France at 33.2 percent with 436 responses for National Rally. Italy and Sweden are also high, at 28.6 percent with 395 responses for Brothers of Italy and 28.5 percent with 416 responses for the Swedish Democrats. Mid-range values include Spain at 16.3 percent with 222 responses for Vox and Germany at 14.8 percent with 213 responses for Alternative for Germany. The United Kingdom is lowest at 12.2 percent with 171 responses for Reform UK. Additional party names are listed under France, Italy, and the Netherlands, but only one frequency and percentage is provided per country, so the figures should be read as country totals tied to the named lead party rather than separate results for each listed party.
Focal predictors
For each hypothesis, the focal predictor is a measure of the individual’s position in the conflict, combined with the salience they assign to that conflict (see, e.g. Table 2). Specifically, each item on positioning first describes each conflict as a contraposition between two groups and then asks the respondent to indicate – on a 6-point scale – which group they are in favour of. Asking respondents to position themselves about two competing poles of a conflict – rather than with respect to issues – allows us to capture the trade-offs involved and foster a dimensional understanding of conflicts among respondents. This wording also distinguishes our approach from related studies (e.g. Bornschier et al. Reference Bornschier, Haffert, Häusermann, Steenbergen and Zollinger2024, Reference Bornschier, Häusermann, Zollinger and Colombo2021), which relied on questions about ‘feeling close to’. Our items, instead, ask the respondents to place themselves on one of the two polarities of a social conflict.
In addition, we deemed looking at the mere position too simplistic. Indeed, it is likely that the conflict becomes a factor pushing people to vote for a specific alternative when it is also considered salient. Therefore, for each conflict, using the variables on positioning and salience, we created a new variable such that the lowest value corresponds to the strongest positioning towards one side (e.g. LGBTQ+) combined with the highest importance (i.e. 10); the highest value corresponds to the strongest positioning towards the opposite side (e.g. traditional family values) combined with the highest importance (i.e. 10). After constructing these eight conflict-specific variables, we normalised them so that they range between 0 and 1 (see notes to Table A1 in the Supplementary material for a more comprehensive explanation of how variables on position and salience were combined).
Conflicts-related socio-structural predictors and other covariates
Some social characteristics are clearly related to individuals’ positions within old and new social conflicts: ethnic group, religious denomination, type of residence (urban vs rural), type of profession, level of education and sexual orientation (Bornschier et al. Reference Bornschier, Häusermann, Zollinger and Colombo2021; Crulli and Emanuele Reference Crulli and Emanuele2026; Hooghe and Marks Reference Hooghe and Marks2026). Therefore, our models control for all these ‘objective socio-structural characteristics’ (to say it with Simon Bornschier et al. Reference Bornschier, Haffert, Häusermann, Steenbergen and Zollinger2024) that underpin ‘subjective’ conflict positions.
Beyond these, we deem it necessary to also control for other socio-demographic variables commonly employed in analyses of voting behaviour – and PRRP voting in particular – namely gender, age and household living standards. Finally, as customary in this type of analysis, we also control for the left–right self-placement of individuals.
Methods
Since our dependent variable is dichotomous, we ran separate logistic regression models, one for each hypothesis. First, we tested the hypothesis at an aggregate level by running eight logistic regression models on the pooled European sample using country-fixed effects. The results of these first eight models are reported in Figure 1 and Table A2 in the Supplementary material. Second, we replicated the analysis on a country-by-country basis to identify which hypotheses hold across all countries under study. The results of these additional 56 logistic regression models are summarised in Figure 2 and fully reported in Tables A3–A9 in the Supplementary material. Post-stratification weights, based on socio-demographic factors (education level, gender and age, geographical area and municipality size) and political variables (vote in the last general election), were applied to all analyses. In Figures 1 and 2, we show the average marginal effects (AMEs) instead of logistic regression coefficients to offer a more intuitive and direct interpretation to the reader.
Conflict Positions and Intention to Vote for PRRPs in Western Europe
Figure 1 Long description
Dot plot illustrating average marginal effects for various socio-political categories. The x-axis is labeled ′Average Marginal Effects′ with no units, ranging from minus 0.2 to 0.3. The y-axis lists categories such as pro-periphery, pro-church, pro-rural, workers, pro-unskilled workers, pro-ethnonationalism, pro-traditionalism, pro-economic growth, left-wing, centre, right-wing, woman, young, old, graduate, manager, professional, qualified worker, production worker, armed forces, student, housewife, unemployed, rich household, periphery of big city, small town, rural area, catholic, protestant, other religion, ethnic minority and non-heterosexual. Each category is represented by a dot with horizontal lines indicating confidence intervals, likely 95 percent. Positive effects are observed for categories like pro-economic growth and right-wing, while negative effects are seen for pro-periphery and pro-church. Categories such as centre and woman show effects near zero. The plot provides insights into socio-political positions and their impact on the studied outcome, with notable positive effects around 0.2 and negative effects near minus 0.1.
Conflict Positions and Intention to Vote for PRRPs: Country-by-Country Models
Figure 2 Long description
Seven forest plots are presented, one for each country: UK, Netherlands, France, Germany, Spain, Sweden and Italy. Each plot displays average marginal effects on the horizontal axis, ranging from minus 0.4 to plus 0.4 in increments of 0.2, with a vertical reference line at 0. A value to the right of 0 indicates a positive marginal effect and a value to the left indicates a negative marginal effect on intention to vote for populist radical right parties. Each plot shares the same eight category labels on the vertical axis: pro-periphery, pro-urban, pro-rural, pro-workers, pro-unskilled workers, pro-internationalism, pro-traditionalism and pro-economic growth. Each category is represented by a dot indicating the point estimate and a horizontal line extending in both directions indicating the confidence interval around that estimate. Categories whose intervals lie entirely to the left of 0 show a negative association, those entirely to the right show a positive association and those whose intervals cross 0 are not distinguishable from zero. Across most country plots, pro-traditionalism and pro-economic growth tend to show point estimates positioned closer to or beyond plus 0.2, with relatively narrow intervals. Pro-internationalism shows point estimates consistently positioned to the left of 0 across several countries, with intervals that in some panels lie entirely below 0. Pro-periphery, pro-urban, pro-rural and pro-workers show estimates clustered near 0 with wider intervals crossing the reference line in most panels. Pro-unskilled workers shows variation across countries, with some panels placing the estimate to the left and others near 0. The most precise estimates, indicated by the shortest horizontal intervals, appear for pro-traditionalism and pro-economic growth across the UK, France and Germany panels.
Results
We start by commenting upon the results of the eight models run on the pooled sample. The first interesting remark is that Hypothesis 1 is rejected: the sign of the relationship between being pro-workers and voting for PRRPs goes in the opposite direction compared to our expectation. The AME of −0.06 means that being pro-workers instead of pro-employers decreases the probability of voting for a PRRP by 6 percentage points. Although relatively small in size, this effect suggests problematising the notion that those upholding the working class may constitute a preserve of PRRPs. Interestingly, Figure 1 also tells us that being a production worker increases the probability of choosing to vote for a PRRP by slightly more than 5 percentage points. Hence, the takeaway here is that PRRPs may well represent some segments of the working class, but not individuals who side with workers in the class conflict.
Similar arguments apply to our second hypothesis, regarding the urban–rural conflict, which is also rejected. Being pro-rural areas does not alter the probability of voting for PRRPs. This result is even more unexpected, given previous research showing that PRRPs and their supporters are more prevalent in the countryside. However, our model also suggests a disconnect between objective residence and position in the conflict. While supporting rural areas in the urban–rural conflict does not influence PRRP voting, residing outside big cities enhances the likelihood of choosing PRRPs by approximately 5 percentage points. Hence, rural residence seems to matter more than pro-rural positioning for PRRP support.
Regarding the two Rokkanian conflicts originating from the National Revolution, we find confirmation of Hypothesis 3b but not Hypothesis 3a. Position in the state–church conflict does not play a relevant role in predicting PRRP voting, as the effect of being pro-church is not statistically significant. Unlike the two previously analysed conflicts, in this case, the objective measure – i.e. religious denomination instead of position in the religious conflict – is also generally non-significant. Indeed, being Catholic or Protestant – instead of non-religious – does not alter voting intentions for PRRPs. Only practising another, non-Christian religion decreases the probability of PRRP voting.
On the other hand, the effect of being pro-periphery is statistically significant and rather strong. Supporting the periphery in the centre–periphery conflict decreases the probability of choosing a PRRP in an election by 16 percentage points. While unexpected, this finding may well resonate with PRRPs’ manifest nationalism and ethnocentrism. Hence, it is likely that this survey item has served as a proxy for the respondent’s cultural nationalism. Moreover, no PRRP in our sample is associated with ethno-regionalists and/or separatist tendencies nowadays.
We can now move on to the four more recently emerged conflicts. Hypothesis 4 – economic winners vs losers from globalisation – is rejected. Following the argument that PRRPs may well represent economically insecure people, we hypothesised that being pro-unskilled workers would enhance the probability of choosing PRRPs in an election. However, our results go in the opposite direction, as the AME of this variable is −0.04. After all, this finding seems to corroborate the argument – typical of neo-cleavage theory – that PRRPs thrive on the new cultural conflicts rather than the economic ones.
Therefore, it comes as no surprise that our three final hypotheses (Hypotheses 5–7) receive unequivocal support from the models summarised in Figure 1. The effects of being pro-ethnonationalism (instead of cosmopolitanism), pro-traditionalism (instead of LGBTQ+ rights) and pro-economic growth (instead of environmental protection) are all remarkably strong. Ceteris paribus, these variables increase the likelihood of choosing a PRRP by 22, 21 and 19 percentage points, respectively. In fact, their magnitudes are equivalent to that of being right-wing (instead of non-classified) on the left–right divide, which is typically one of the most – if not the most – important predictors of PRRP voting.
Interestingly, all these effects are also much stronger than the effect of education level, i.e. being a graduate. This challenges the idea, advanced by recent contributions in neo-cleavage theory, that education may constitute a cleavage in itself (Hooghe and Marks Reference Hooghe and Marks2026) – although it remains true that highly educated people are significantly less likely to tilt towards PRRPs.
To sum up, our answer to the question of which conflicts PRRPs represent in Western Europe today is clear and in line with recent developments in comparative politics: they mostly represent new conflicts, especially those cultural in nature. Still, while most studies have stressed their nativist/ethno-nationalist focus, our results suggest that conflicts over traditional family values and the environment are now as relevant to their political representation (Table 4).
Outcome of Research Hypotheses
Table 4 Long description
Seven hypotheses test whether taking specific sides in major social and cultural conflicts is linked to a higher probability of voting for PRRPs. Three hypotheses are confirmed: being pro-ethnonationalism in the cultural winners versus losers from globalisation conflict, being pro-traditionalism in the traditionalism versus LGBTQ plus conflict, and being pro-economic growth in the economy versus environment conflict. Three hypotheses are rejected: being pro-workers in the owners versus workers conflict, being pro-rural areas in the urban versus rural conflict, and being pro-unskilled workers in the economic winners versus losers from globalisation conflict. The Rokkanian conflicts hypothesis is split: the centre versus periphery component is rejected, while the state versus church component is confirmed. Overall, confirmed relationships cluster around cultural and values-based positions rather than class or place-based economic divides. Outcomes indicate support or rejection of each proposed relationship but do not convey effect size, statistical uncertainty, or causality.
Country-by-country overview
We replicated the previously run logistic regression models for each country individually. From these models, we observe that the effect of being pro-workers (Hypothesis 1) is statistically significant and negative – as in the pooled sample – in three countries only: Spain, Sweden and Italy. In all other countries, the effect is not significant, thus reconfirming that siding with the pro-worker camp is not a meaningful factor for PRRP voting.
The pro-rural effect (Hypothesis 2) is confirmed not to be statistically significant, with the partial exception of the UK, where it is significant at the p < 0.1 level. This suggests that a pro-rural position enhances the likelihood of voting for Reform UK, which might indeed reflect Farage’s and UKIP’s historical attention to – and strength in – rural communities (Jennings and Stoker Reference Jennings and Stoker2016).
The previous observations on Hypothesis 3a are also consistent across countries, with the only partial exception of France, where the pro-periphery position maintains a negative effect but is only significant at the p < 0.1 level. As for the state–church conflict (Hypothesis 3b), this appears to be substantially relevant for PRRP voting in the Netherlands only. Here, a pro-church stance decreases the likelihood of choosing PRRPs, arguably reflecting the presence of prominent confessional parties in the country.
Moving on to the new conflicts, the negative effect of being pro-unskilled workers we found at the aggregate European level (Hypothesis 4) seems to be driven by the UK, Spain and, to a lesser extent, Sweden. Finally, and most importantly, Hypotheses 5–7 are robustly confirmed across all Western European countries included in the analysis. Of course, there are some country-by-country nuances. For example, the strongest effect among pro-ethnonationalism, pro-traditionalism and pro-economic growth varies across countries. However, the takeaway is that the picture remains consistent across different Western European polities.
Making sense of the decoupling of socio-structural characteristics from class and place attachments
Beyond the association between new conflicts and PRRP voting, our pooled analyses revealed another interesting finding: the decoupling of socio-structural position and subjective alignment in both class and urban–rural conflicts. Taken together, these two findings seem to suggest the following narrative: when voting for PRRPs, working-class individuals and rural dwellers – like other socio-demographic groups – are motivated more by cultural grievances (concerning immigration, family values) or by others that mix cultural and economic elements (such as the environment) than by class- or place-based solidarities.
However, to determine whether this is indeed the case, we must employ more appropriate statistical analyses. To this end, we replicated our previous logistic regressions, this time incorporating interaction effects. First, we ran four models in which occupation was interacted with positions in the owners–workers, cosmopolitanism–ethnonationalism, LGBTQ+–traditional values and economy–environment conflicts. This allowed us to test whether, among production workers, it is cultural conflicts – rather than class conflict – that drive PRRP voting. Second, following the same logic, we ran another four models in which residential category was interacted with positions in the urban–rural, cosmopolitanism–ethnonationalism, LGBTQ+–traditionalism and economy–environment conflicts. The results of these additional analyses are visually summarised in Figures 3 and 4, which plot the predicted probabilities of PRRP voting among production workers and rural dwellers, respectively, conditional on their positions in the various social conflicts.
Predicted Probabilities of PRRP Voting for Production Workers across Class, Ethnonationalism, Traditionalism and Environmental Conflict Dimensions
Figure 3 Long description
Four subplots arranged in a two by two grid. Each subplot contains a black line with point markers and vertical error bars, overlaid on a background of vertical bars. Top-left subplot: The left y-axis label reads Predicted probabilities of PRRP voting, ranging from 0 to 4. The right y-axis label reads PRRP vote, ranging from 0 to 15. The x-axis label reads pro-workers, with tick labels 0, 2, 4, 6, 8. The plotted line starts near 3.0 at 0, then is near 2.9 at 2, near 2.8 at 4, near 2.7 at 6 and near 2.6 at 8. Top-right subplot: The left y-axis label reads Predicted probabilities of PRRP voting, ranging from 0 to 4. The right y-axis label reads PRRP vote, ranging from 0 to 15. The x-axis label reads pro-ethnonationalism, with tick labels 0, 2, 4, 6, 8. The plotted line starts near 2.0 at 0, then is near 2.3 at 2, near 2.6 at 4, near 2.9 at 6 and near 3.2 at 8. Bottom-left subplot: The left y-axis label reads Predicted probabilities of PRRP voting, ranging from 0 to 4. The right y-axis label reads PRRP vote, ranging from 0 to 10. The x-axis label reads pro-traditionalism, with tick labels 0, 2, 4, 6, 8. The plotted line starts near 1.0 at 0, then is near 1.6 at 2, near 2.2 at 4, near 2.8 at 6 and near 3.4 at 8. Bottom-right subplot: The left y-axis label reads Predicted probabilities of PRRP voting, ranging from 0 to 4. The right y-axis label reads PRRP vote, ranging from 0 to 15. The x-axis label reads pro-economic growth, with tick labels 0 and 1. The plotted line is near 2.0 at 0 and near 2.6 at 1.
Predicted Probabilities of PRRP Voting for Rural Dwellers across Urban–Rural, Ethnonationalism, Traditionalism and Environmental Conflict Dimensions
Figure 4 Long description
Four small-multiple graphs display predicted probabilities of PRRP voting for rural dwellers across four attitude dimensions. Each graph contains a single line with circular markers and vertical error bars representing uncertainty, plotted against the left vertical axis labeled predicted probabilities of PRRP voting for rural dwellers, ranging from 0 to approximately 4. A separate series of vertical bars representing case counts is overlaid in each panel, corresponding to the right vertical axis labeled N, ranging from 0 to approximately 15. The horizontal axis in each panel runs from 0 to 1. Image A shows pro-rural on the horizontal axis. The line with markers begins at approximately 2 at the low end of the scale and rises gradually to approximately 3 at the high end. Error bars are narrow throughout. The case-count bars are distributed across the scale with moderate height. Image B shows pro-ethnonationalism on the horizontal axis. The line begins at approximately 1 at the low end and rises steeply to approximately 10 at the high end, representing the steepest increase across all four panels. Error bars widen noticeably at the high end of the scale. Case-count bars are concentrated towards the lower end of the scale. Image C shows pro-traditionalism on the horizontal axis. The line starts at approximately 2 at the low end and rises to approximately 7 at the high end. Error bars widen at the upper end. Case-count bars are spread across the scale. Image D shows pro-economic growth on the horizontal axis. The line starts at approximately 2 at the low end and rises to approximately 8 at the high end. Error bars widen towards the upper end. Case-count bars are concentrated towards the middle and lower end of the scale. Across all four panels, predicted probabilities increase from the low to the high end of each attitude dimension. The steepest increase is visible in the pro-ethnonationalism panel.
Figure 3 confirms that, among production workers, alignment with the workers’ side in the class conflict does not affect the intention to vote for a PRRP. In contrast, support for ethnonationalism, traditionalism and anti-environmental positions unequivocally increases the probability of voting for a PRRP. In the model interacting occupation with class conflict positioning, the predicted probability of voting for a PRRP among production workers is around 0.25, regardless of alignment with either employers or workers (with differences that are neither statistically nor substantively significant). Conversely, for the three more recently emerged conflicts, strong alignment of production workers with ethnonationalism, traditionalism and anti-environmental positions predicts probabilities of voting for a PRRP of 0.34, 0.36 and 0.31, respectively.
Thus, while production workers may serve as an electoral reservoir for PRRPs, the reasons differ markedly from those that once made this group the natural constituency of socialist and communist parties. In the past, a close alignment of class disadvantage and class identity fostered support for such parties among production workers. Today, it is the misalignment of class disadvantage and pro-workers positions that creates opportunities for PRRPs to attract votes from this sector based on cultural conflicts.
Similarly, Figure 4 supports the idea that, among rural dwellers, alignment with rural interests vis-à-vis urban interests does not affect PRRP voting intentions, whereas support for the ‘conservative’ side of new conflicts substantially increases such intentions. In the model interacting residential category with urban–rural conflict positioning, the predicted probability of voting for a PRRP among rural residents is around 0.22, regardless of alignment with either urban or rural interests (again, with differences that are neither statistically nor substantively significant). In contrast, for the three more recently emerged conflicts, strong alignment of rural residents with ethnonationalism, traditionalism and anti-environmental positions predicts probabilities of voting for a PRRP of 0.38, 0.36 and 0.37, respectively. This suggests that PRRPs are successful in rural areas not because of a particular appeal to pro-rural positions, but because their message of cultural closure resonates strongly among rural dwellers.
Robustness checks and further evidence
We performed a series of additional analyses to test the robustness of our arguments and provide further evidence on our most interesting results. First, regarding model specifications, the inclusion of left–right self-placement as a control may cast some doubts. One may see this variable and PRRP voting as simultaneously influencing each other, creating a two-way causal relationship, or endogeneity problems. Aligning with the established tradition in studies of voting behaviour, we regarded left–right ideology as a clear, fundamental antecedent of vote choice, being a sort of ‘super-issue’ (Fuchs and Klingemann Reference Fuchs, Klingemann, Jennings and van Deth1990; Inglehart Reference Inglehart, Dalton, Flanagan and Beck1984) guiding voters on more specific matters. Therefore, we expect that, by removing this important control, the effects of social conflicts would become larger.
We checked this, and this is indeed the case: by replicating the models in Figure 1 without left–right self-placement, all significant predictors exhibit increased effect sizes, as part of their effect that was previously absorbed by left–right ideology is now fully visible (see Table A12 in the Supplementary material). Additionally, positioning with the church becomes significant and positive, although the effect is still weak. This is because pro-church respondents also tend to be more right-wing. The fact that the state–church conflict only becomes significant when left–right ideology is excluded suggests that the latter is more influential.
Second, we conducted additional tests to reinforce our argument that salience improves our understanding of conflict positions, increasing their effects. Specifically, we (a) replicated the analysis in Figure 1 using positions alone as the main predictors (operationalised as dummy variables by recoding the original 1–6 scales; e.g. 0 = pro-cosmopolitanism, 1 = pro-ethnonationalism). We also (b) estimated interaction effects between positions and salience. Both tests reinforce our argument. Regarding (a), the results are consistent with the main analysis, but position alone has much weaker AMEs compared to the measure combining position and salience (compare Tables A2 and A13 in the Supplementary material). Regarding (b), we find that, for all the conflicts clearly related to PRRP voting – centre–periphery, cosmopolitanism–ethnonationalism, LGBTQ+ rights–traditionalism and environment–economic growth – the effect of position increases with salience (Figure 5). At low levels of salience, i.e. when the conflict is not important to the respondent, position alone is either non-significant or has a negligible effect. However, as the conflict becomes more important to the respondent, the effect of position becomes stronger. Finally, these interactions also provide new insights into the state–church conflict, which showed no effect in the main analysis. Being pro-church has a positive effect among respondents who consider this conflict unimportant, but this effect fades and eventually reverses as salience increases. This suggests that PRRPs’ instrumental use of religion resonates more with superficially pro-church voters than with those for whom the church is central.
Interactions between Conflict Positions and Salience in Predicting Intention to Vote for PRRPs
Figure 5 Long description
Nine line graphs arranged in a four-row, two-column grid, each showing the marginal effect of a different conflict position across salience levels. All graphs share the same structure: the horizontal axis runs from ′Not salient at all′ to ′Very salient′ with discrete tick points in between. The left vertical axis shows marginal effect values ranging from negative 15 to positive 15. The right vertical axis shows percent values ranging from 0 to 20. Each graph contains a line with circular data points and error bars, a dashed horizontal reference line at zero and background bars showing the distribution of salience levels. Graph 1, marginal effect of pro-periphery: The line starts near zero at ′Not salient at all′ and declines steadily, reaching approximately negative 10 at ′Very salient,′ crossing below zero midway through the salience range. Graph 2, marginal effect of pro-church: The line starts at approximately positive 5 at ′Not salient at all′ and declines across the salience range, ending near negative 5 at ′Very salient,′ crossing zero roughly in the middle. Graph 3, marginal effect of pro-workers: The line remains close to zero throughout, showing a near-flat trend with slight fluctuation between approximately negative 2 and positive 2 across the full salience range. Graph 4, marginal effect of pro-workers (second panel): Similar flat trend, staying near zero with marginal variation across all salience levels. Graph 5, marginal effect of pro-ethnonationalism: The line starts near zero and rises gradually, reaching approximately positive 5 at ′Very salient.′ Graph 6, marginal effect of pro-ethnonationalism (second panel): Shows a stronger positive trend, starting near zero and rising to approximately positive 10 at ′Very salient.′ Graph 7, marginal effect of pro-economic growth: The line shows the strongest positive trend, starting near negative 2 at ′Not salient at all′ and rising to approximately positive 12 at ′Very salient,′ crossing zero at a mid-range salience level. Graph 8, marginal effect of pro-economic growth (second panel): Shows a moderate positive trend, starting near zero and rising to approximately positive 8 at ′Very salient.′ Graph 9, marginal effect of pro-periphery or related factor: The line shows a slight negative trend, declining from near zero to approximately negative 5 at ′Very salient.′ Across all panels, graphs showing pro-ethnonationalism and pro-economic growth positions display the strongest positive increases with salience, while pro-church and pro-periphery positions show declining or negative trends. Pro-workers panels remain near flat throughout.
Finally, we wanted to further verify the robustness of our main finding: the strength and country-by-country consistency in how new value-based conflicts are associated with PRRP voting. We performed two tests. First, we ran another logistic regression model with country-fixed effects, with all the conflict variables simultaneously included as predictors. This was to provide additional insight into which of the conflicts appears to be ultimately most important for explaining PRRP voting. As the reader can see by inspecting Table A15 in the Supplementary material, the three globalisation- and modernisation-related socio-cultural conflicts (i.e. being pro-ethnonationalism, pro-traditionalism and pro-economic growth) show strong effects even when included in a single model – with pro-ethnonationalism being, by a small margin, the strongest predictor. Hence, our key conclusion is intact.
Second, we checked whether more variation emerges when there are several PRRPs in one country – an increasingly relevant phenomenon in Europe (Piccolino and Puleo Reference Piccolino and Puleo2026). Among the three countries with more than one PRRP – France, Italy and the Netherlands – this type of disaggregated analysis is only feasible in Italy. Only there does the second most-voted PRRP, Lega, have enough declared voters to make such an analysis statistically meaningful (107 respondents, equivalent to almost 8% of the Italian sample). Therefore, we proceeded as follows. First, we created a new nominal variable: 0 = intention to vote for other parties; 1 = intention to vote for FdI; 2 = intention to vote for Lega; 3 = intention to vote for Italexit (a very small PRRP with only 19 declared voters in our sample). We then conducted three separate multinomial regression analyses, where this nominal variable served as the dependent variable and the focal predictors were the three new conflict dimensions, mostly concerning the cultural sphere. The results are reported in Table A16 in the Supplementary material. They show that positions on the three conflicts predict the intention to vote for both FdI and Lega (as compared to non-PRR Italian parties) in an extremely similar manner. Hence, there appears to be very little variation between PRRPs within the same country.
Conclusion
This article examined which social conflicts the populist radical right represents in Western Europe today. Drawing on original survey data from seven countries and grounded in cleavage theory, our analysis sought to empirically test whether the appeal of PRRPs is rooted in traditional Rokkanian cleavages, new conflicts stemming from globalisation and cultural value change, or both.
The evidence is unequivocal in one respect: PRRPs primarily thrive on new conflicts, especially those that are culturally rooted. Across all countries analysed, alignment with ethnonationalist, traditionalist and pro-growth (vs environmentalist) positions significantly increases the probability of PRRP voting. This not only confirms the literature that identifies PRRPs as champions of the cultural backlash against cosmopolitanism, liberalism and environmental concerns (Bornschier et al. Reference Bornschier, Haffert, Häusermann, Steenbergen and Zollinger2024; Norris and Inglehart Reference Norris and Inglehart2019), but also suggests that these parties now represent a broader ideological bundle of cultural demarcation that extends well beyond immigration alone. The strength and consistency of these effects reconfirm that the new cultural dimension has fully become the dominant terrain for PRRP contestation and mobilisation.
Yet, beyond this expected finding, one of the most revealing insights from our analysis concerns the disjuncture between objective socio-structural characteristics and subjective positions in the class and urban–rural cleavages. While PRRPs are indeed overrepresented among production workers and rural residents – aligning with long-standing patterns in the geography and sociology of radical right support (Crulli Reference Crulli2024; Oesch and Rennwald Reference Oesch and Rennwald2018) – our models show that individuals who support the interests of workers (vs owners) or rural areas (vs cities) are not more likely to vote for PRRPs. On the contrary, the alignment with the working class or rural interests either has no significant effect or is negatively associated with PRRP voting.
This decoupling of structural position and subjective alignment has important implications for cleavage theory and the study of political representation, as it suggests a weakening of the traditional mechanisms of class and territorial cleavage politics. In the case of class, PRRPs appear to attract working-class voters not because they represent labour interests, but because they recast working-class belonging in cultural, rather than economic, terms (as already suggested by Bornschier and Kriesi Reference Bornschier, Kriesi and Rydgren2012, for example). In other words, the appeal of these parties – which, differently from the left, does not focus on the traditional economic divide between workers and employers – resonates most with production workers whose political preferences are influenced more by cultural concerns than by the class-based conflict. Similarly, rural residence predicts PRRP voting, but this effect does not translate into a broader pro-rural stance vis-à-vis urban areas among PRRP voters. In both cases, PRRPs arguably convert structurally rooted grievances into culturally framed demands, bypassing traditional collective solidarities.
All this supports the idea that PRRPs do not necessarily ‘represent’ the traditional working class or rural communities by defending their interests vis-à-vis those of their classic social opponents; rather, they capitalise on the perceived cultural displacement and anxieties associated with these social locations. This dynamic aligns with the idea of a new ‘culturalised’ working class that demands cultural protectionism (Oesch Reference Oesch2008), as material grievances are readily translated into cultural resentment.
In short, our findings suggest that even where socio-structural conditions conducive to cleavage politics persist, the mechanisms behind the politicisation of these conditions have become more fragmented and less predictable. This raises questions about the contemporary relevance of identity as a mediating mechanism between structural characteristics and political behaviour (Bornschier et al. Reference Bornschier, Haffert, Häusermann, Steenbergen and Zollinger2024). Therefore, future studies may further analyse the role of subjective positioning within conflicts as a mediator of socio-structural characteristics on voting for the PRR to gain further insights into the changing face of cleavage voting.
Finally, this article also contributes methodologically by operationalising individuals’ positions within social conflicts as combinations of conflict position and conflict salience, rather than as attitudinal stances alone, consistent with the logic of classic cleavage theory. This approach allows for a more refined understanding of the extent to which conflictual alignments are politically activated. The finding that positions in new cultural conflicts have strong predictive power for PRRP voting only when these conflicts are also perceived as salient underscores the importance of such methodological innovation.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/gov.2026.10054
Data availability
Replication data for the article are available upon request.
Acknowledgements
An earlier version of this manuscript was presented at the ENGAGE.EU Early-stage researchers workshop (UT Capitole, Toulouse, May 2025); at the 2025 EPSA Conference in Madrid; and at the 2025 SISP Conference in Naples. We thank the participants in all these events, especially Sophie Borwein and Andrej Zaslove, for their insightful suggestions.
Financial support
The research and the original cross-national survey data presented in the manuscript are part of the 2022 PRIN Project ‘New Conflicts’ Mapping in the Age of Globalization (NEWCOMING)’, funded by the European Union – Next Generation EU, Mission 4 Component 1 CUP I53D23003130001.
Disclosure statement
The authors report that there are no competing interests to declare.
Open access
