How Important is European Integration for Pro-European Voters in Times of Politicization?

Abstract

The increasing success of Eurosceptic parties and multiple European crises have fostered the politicization of European integration. However, we know little about how pro-European voters behave in this politicized environment. How important is European integration for pro-European voters? Do they care more about European integration when their vision of Europe is increasingly contested? The past 15 years have been characterized by persistent politicization of Eurosceptic parties and an increasing willingness of Europhile parties to engage with European topics. Moreover, European crises illustrated potential consequences of political disintegration. I expect these developments to be reflected in an increasing salience of European integration among pro-European voters. I derive salience estimates from spatial vote models encompassing four issue dimensions to track the development of salience. I use data from the European Social Survey, the European Election Studies, and the Chapel Hill Expert Survey, which cover many EU member states between 2008 and 2019. The results show that the salience of the European integration issue is most pronounced among Eurosceptic voters, while it is less pronounced among pro-European voters. However, in countries with a successful Eurosceptic party, it is more likely that the European integration issue also matters to pro-European voters. Moreover, there has been an increasing trend over time that European integration is not only important for anti- but also for pro-European voters across Europe.

Zusammenfassung

Der zunehmende Erfolg euroskeptischer Parteien und Europa bezogene Krisen haben die Politisierung der europäischen Integration vorangetrieben. Wir wissen jedoch wenig darüber, wie sich pro-europäisch eingestellte Wähler:innen in diesem politisierten Umfeld verhalten. Wie wichtig ist das Thema der europäischen Integration für pro-europäische Wähler:innen? Wird das Thema für sie wichtiger, wenn ihre Vorstellung von Europa zunehmend umstritten ist? Die letzten 15 Jahre waren durch eine anhaltende Politisierung durch euroskeptische Parteien und eine zunehmende Bereitschaft europhiler Parteien, sich mit europäischen Themen zu beschäftigen, geprägt. Darüber hinaus haben europäische Krisen die möglichen Folgen einer politischen Disintegration aufgezeigt. Ich erwarte, dass sich diese Entwicklungen in einer zunehmenden Bedeutung der europäischen Integration bei pro-europäischen Wähler:innen widerspiegelt. Ich schätze Salienzmaße mit räumlichen Wahlmodellen, die vier Themenbereiche umfassen, um die Entwicklung der Salienz zu erfassen. Ich verwende Daten aus dem European Social Survey, den European Election Studies und dem Chapel Hill Expert Survey, die eine Vielzahl von EU-Mitgliedstaaten zwischen 2008 und 2019 abdecken. Die Ergebnisse zeigen, dass die Bedeutung des Themas der europäischen Integration bei euroskeptischen Wähler:innen am stärksten ausgeprägt ist, während sie bei pro-europäischen Wähler:innen geringer ausfällt. Jedoch ist es in Ländern mit einer erfolgreichen euroskeptischen Partei wahrscheinlicher, dass europäische Integration auch für pro-europäische Wähler:innen von Bedeutung ist. Darüber hinaus ist ein zunehmender zeitlicher Trend zu beobachten, dass die europäische Integration nicht nur für euroskeptische, sondern auch für pro-europäische Wähler:innen wichtig ist.

1 Introduction

European integration has been considerably politicized over the past 15 years. Eurosceptic parties have become increasingly successful all over Europe (Filip 2021; Treib 2021) and have responded to multiple European crises by becoming even more Eurosceptic (Braun et al. 2019). Moreover, pro-European mainstream parties have reacted to the politicization efforts of Eurosceptic parties. They are no longer trying to depoliticize the European integration issue but have started engaging more intensively with it, especially since 2014 (Carrieri 2021; Adam et al. 2017). Also, multiple European crises have fostered the politicization of European integration (Hutter and Kriesi 2019). However, we know little about how pro-European voters behave in this politicized environment. How important is European integration for pro-European voters? Are pro-European voters more aware of their preferences regarding European integration when their vision of Europe is contested?

Some empirical examples indeed suggest that the pro-European camp of voters care about Europe. For instance, in the 2023 parliamentary election in Poland, the progressive, pro-European camp won the political majority. Also Emmanuel Macron was very successful with his pro-European election campaign in France in 2017 (Bouza García and Oleart 2022). This paper investigates the development of voters’ salience of European integration between 2008 and 2019, a time period which was characterized by increasing success of Eurosceptic parties all over Europe and multiple European crises. I am interested in electoral salience, which captures the relative weight voters give to a specific issue in their vote calculus (Ansolabehere and Puy 2018). Investigating the extent to which voters consider European integration in their vote calculus, is particularly relevant since European countries have recently faced serious transnational challenges, such as the war in Ukraine, the COVID-19 pandemic, climate change, and migration flows. The ability to cope with these crises is also linked to the further course of European integration. The willingness of governments to cooperate with other EU members states relies on public support for further integration and the extent to which parties can succeed with pro-European election campaigns within their countries.

This paper includes two analytical steps. First, as the literature provides mixed results as to whether the European integration issue has an impact on vote choice at the national level at all (De Vries 2007; De Vries and Hobolt 2016; Schoen 2019), I investigate whether it generally matters to voters. To gain a better understanding of the extent to which European integration matters to voters, I compare the salience of this issue to the salience of other issues, thus explicitly considering the multidimensionality of European policy spaces (e.g., Kriesi et al. 2006; De Vries and Hobolt 2012; König et al. 2017).

Second, I investigate how pronounced the salience of the European integration issue is among different subgroups of voters: pro-European, moderate and Eurosceptic voters. I test whether the salience among pro-European voters is more pronounced in contexts of high domestic politicization by presenting salience estimates for countries with and without successful Eurosceptic parties.

To estimate electoral salience, I use spatial vote models that encompass four issue dimensions: European integration, the economy, immigration, and the environment. The models enable a comparison of the electoral salience of European integration with other relevant issues. Measuring issue salience in the electoral context has the advantage of capturing it in a situation in which it also affects party competition. Moreover, this strategy circumvents problems associated with other common measures of issue salience among voters, such as “most important problem” questions (Wlezien 2005). I use data from the European Social Survey (ESS), the Voter Studies of the European Election Studies (EES), and the Chapel Hill Expert Survey (CHES), encompassing a wide range of European Union (EU) member states between 2008 and 2019.

The results show that, in the period studied, the European integration issue mattered to voters, especially in 2019, when the issue was almost as important as economic issues. The salience of the European integration issue is still the most pronounced among Eurosceptic voters. However, in countries with a successful Eurosceptic party, it is more likely that the issue matters to pro-European voters as well. Moreover, in an increasing number of EU member states I find significant salience estimates for the pro-European camp of voters over time. Nevertheless, the results do not indicate a sharp increase of salience of European integration among pro-European voters. Yet, the country-specific results for France and the Netherlands support the notion that even a small increase in the salience of this issue among pro-European voters may open up windows of opportunities for Europhile parties to succeed with strikingly pro-European campaigns.

The paper proceeds as follows: after this brief introduction, I discuss the relevant literature, introduce my argument, and derive my hypotheses. I then present my research design, data, and operationalization. Following that, I report the results. I conclude with a summary and an outlook for future research.

2 Development of European Integration

This paper builds on the literature on the politicization of the European integration process (e.g. Hooghe and Marks 2009; Hoeglinger 2016; Hutter and Kriesi 2019), the literature on EU issue voting (e.g. De Vries 2007; De Vries and Tillman 2010; De Vries and Hobolt 2016), and recently published work on pro-European counter-reactions among parties and voters (e.g. Carrieri 2021; Jones et al. 2021; Malet and Walter 2024).

2.1 Politicization of the European Integration Process

In their seminal work, Hooghe and Marks (2009) present a postfunctionalist theory of European integration. They state that a new political conflict line is emerging, which no longer follows the established economic left–right dimension. Rather, the conflict runs along a cultural dimension ranging from green-alternative-libertarian to traditional-authoritarian-nationalist values. The politicization of the European integration process is at the heart of this conflict. This is characterized by a mobilization of public opinion about European integration, which Hooghe and Marks (2009) describe as a shift from “permissive consensus” to “constraining dissensus”. In the early years of European integration, political elites could more or less rely on tacit public support for further integration steps. However, since European integration increasingly touches areas of state sovereignty and national identity, the domestic salience of the EU increases, thus creating losers of economic and cultural integration (Hooghe and Marks 2009; Kriesi et al. 2006; Schimmelfennig 2018). While center-left and center-right mainstream parties primarily compete on the established economic left–right dimension, left-libertarian and right-nationalist parties compete with extreme positions on new cultural issues, such as immigration, the environment, and European integration. The politicization of the European integration issue is mostly driven by left and right Eurosceptic parties (Hooghe and Marks 2009; Hutter and Kriesi 2019), who act as issue entrepreneurs, while mainstream parties have failed to depoliticize this issue (Hobolt and Rodon 2020). Even if the issue of European integration obviously relates to the European level, the nation states are the decisive arena in which politicization occurs, as party competition still mainly takes place in the national context (Kriesi 2016).

In general terms, politicization encompasses the process of the expansion of the publicly visible conflict regarding European integration (Hutter and Grande 2014; Hutter and Kriesi 2019), which is accompanied by increasing issue salience, actor polarization, and actor expansion (De Wilde et al. 2016; Kriesi 2016). Politicization is not restricted to the party level (Hoeglinger 2016) but also becomes visible at the voter level (De Wilde 2011). Still, the literature on politicization focuses primarily on the party level.

There is widespread agreement that some kind of politicization of the European integration issue has taken place since the mid-1980s (e.g. Schmitter 2009), leading to an increasing mobilization of public opinion about European integration. This politicization is reflected in a new conflict, in which anti-European nationalism and pro-European transnationalism confront each other (e.g. De Vries 2018). However, the extent, intensity, and course of the politicization process are subject to scholarly debate. Some authors show that it has steadily increased over time (Rauh and Zürn 2016), while others point to significant variation in the extent of politicization of the European integration issue over time, across different countries and across political arenas (e.g. De Wilde et al. 2016).

Figure 1 shows some descriptive evidence to illustrate the development of politicization from 1999–2019. Figure 1a shows the average vote shares of Eurosceptic radical right parties in the EU member states, which have risen steadily over time, being most pronounced from 2010 onwards. Figure 1b shows the average salience Eurosceptic radical right parties attach to the European integration issue. Its salience increased from 1999 to 2006 and remained at a stable, high level until 2019. The polarization of the party systems on European integration also increased from 2002 to 2019 (Fig. 1c).¹ Overall, these descriptive results show a politicized environment in the period under study (2008–2019).

Fig. 1

Sources of politicization from 1999–2019 based on CHES Trendfile: a average vote shares of ES RRPs in EU member states; b average salience of European integration of ES RRPs in EU member states; c average polarization of party systems on European integration in EU member states

2.2 The Electoral Salience of European Integration in Multidimensional Policy Spaces

As illustrated above, the conceptualization of politicization focuses mostly on the party level by looking at parties’ issue salience, actor polarization, and actor expansion (e.g. Hutter and Grande 2014; De Wilde et al. 2016). There has been less research linking politicization at the party level with the individual level. If parties polarize their positions and raise the salience of European integration, they form two visible, opposing political camps. Carrieri (2021) argues that a fully fledged politicization of an issue only occurs, if such an altering policy offer by parties matches the voter demand. Thus, this paper focuses on the development of the weight voters attach to the European integration issue at a time of intense politicization at the party level. A party should have even more incentive to politicize an issue if it is salient among the electorate, more precisely among a party’s potential voters. Even though parties have the power to restructure party competition through repositioning and selective emphasis, it is ultimately the voters who decide which issue they consider in their voting decision. Still, causality probably runs in both directions. Parties shape voters’ issue perception by partisan cuing (e.g. Brader et al. 2020; Slothuus and De Vreese 2010), but they also respond to voters’ issue preferences and priorities (e.g. Adams et al. 2004; Klüver and Sagarzazu 2016; Klüver and Spoon 2016). Opinion formation concerning EU issues also runs in both directions, with party elites listening to voters and vice versa (Sanders and Toka 2013; Steenbergen et al. 2007; Malet and Thiebaut 2024).

One approach focusing on the importance of public opinion in the context of the politicization of the European integration issue is presented by the literature on EU issue voting, which investigates whether EU issues have an impact on vote choice in national elections (De Vries 2007). In the EU issue voting literature, there is mixed evidence on whether European integration plays a relevant role in individuals’ voting decisions (e.g. De Vries and Hobolt 2016; Schoen 2019). For example, De Vries (2007) finds that, in the 1990s and early 2000s, the issue was not salient for voters in the Netherlands and Germany, but that it was for the electorate in Denmark and the United Kingdom. Schoen (2019) shows that European integration was not an electoral issue in the 2017 German election. According to De Vries and Tillman (2010), EU issue voting was more pronounced in Central Eastern Europe than in Western Europe in 2004. Furthermore, De Vries and Hobolt (2016) find that the extent of EU issue voting in national elections was greater in 2009; it was present in all EU member states except in Belgium-Wallonia, Bulgaria, Latvia, Poland, and the United Kingdom.

Overall, the results of the politicization, as well as the EU issue voting literature, point to a “punctuated politicization” (Grande and Kriesi 2016) of the European integration issue at the level of parties and voters, indicated by significant variation over time and across countries. This could be due to the conflict between politicization efforts of Eurosceptic parties and (former) depoliticization efforts of pro-European mainstream parties (e.g. Börzel and Risse 2018; Green-Pedersen 2019).

Most of the EU issue voting literature relies on spatial vote models (Downs 1957) that use the policy distance between voters’ and parties’ issue positions on European integration to predict vote choice (e.g. De Vries 2007; Hobolt and Rodon 2020). The resulting spatial coefficients can then be used as unbiased proxies for electoral salience (Ansolabehere and Puy 2018) and capture the relative weight voters attach to a specific policy issue in their vote calculus. Thus, I use the term “electoral salience” to describe issue salience at the level of the voters with respect to their voting decision. To gain a better understanding of the extent to which European integration plays a role in the voting decision, I compare it to the electoral salience of other issues. In doing so, I examine the salience of the European integration issue by explicitly considering the multidimensional character of European policy spaces (e.g., Kriesi et al. 2006; De Vries and Hobolt 2012; Hellwig 2014; König et al. 2017). Voters are confronted not only with the European integration issue, but with many other issues, thus voting in a multidimensional policy space. This has also been brought forward by Toshkov and Krouwel (2022); the European integration issue cannot be regarded as bundled with other cultural issues, so it should be examined how it relates to other issue dimensions.

2.3 Pro-European Counter-Reactions Among Europhile Parties and Voters

So far, we have seen that the European integration process is no longer characterized by tacit support but is being contested by Eurosceptic issue entrepreneurs. But why should we expect Europhile voters to become more aware of their preferences regarding European integration in times of politicization?

The literature on EU issue entrepreneurship shows that Eurosceptic parties have successfully introduced the European integration issue into the policy space by competing with extreme anti-EU standpoints and raising the salience of this issue (e.g. De Vries and Hobolt 2012, 2020; Hobolt and De Vries 2015). It is rather intuitive that once an issue has been successfully politicized by an issue entrepreneurial party, parties with opposing standpoints on this issue cannot ignore it forever but are in a way “forced” to deal with that issue. This is in line with the work of Meguid (2005, 2008), who argues that party competition is not limited to the interaction between ideologically directly neighboring parties. In response to niche party entry, mainstream parties have three options: ignore, move towards, or move away from the niche party’s issue. The results regarding pro-European parties’ reactions are mixed, reflecting the opposing pressures they are confronted with (Rohrschneider and Whitefield 2016). Due to increasing public Euroscepticism (Ejrnæs and Jensen 2019), they have an incentive to adapt more Eurosceptic positions to stay competitive from a spatial viewpoint. However, they potentially face reputational costs, because European integration is largely a product of Europhile mainstream parties’ behavior over the last decades (Vasilopoulou 2023).

Some findings support the notion that Euroscepticism is contagious. For the period between the 1980s and 2010, research shows that Eurosceptic parties have provoked shifts in mainstream parties’ positions on European integration towards more sceptical standpoints (Meijers 2017) and that parties have responded to Eurosceptic attitudes by taking a more Eurosceptic position (Spoon and Williams 2017). However, a study analyzing more recent data from 2006 to 2017 shows that pro-European parties engage more and more in an adversarial strategy by taking stronger Europhile positions and raising the salience of the issue, which was particularly pronounced in 2017 (Carrieri 2021). Also, Adam et al. (2017) show that in 2014 pro-European parties did not differ (any longer) from Eurosceptic parties in their willingness to emphasize the European integration issue, especially when they were internally united on it. However, Malet and Thiebaut (2024) show that Europhile mainstream parties are not generally willing to emphasize European integration more strongly. They raise the salience if the public becomes more supportive for European integration, but decrease the salience if citizens become more sceptical. Literature on party cuing regarding attitudes towards the EU shows that changes in parties’ strategies regarding the salience of and position on European integration indeed affect public opinion (e.g. Down and Wilson 2010; Steenbergen, Edwards and De Vries 2007; Stoeckel and Kuhn 2018). Thus, it is plausible that changes in Eurosceptic and Europhile parties’ strategies should be reflected in changes in voters’ salience.

Moving from the party to the voter level, research shows a revival of public support for the EU (Jones, Daniel Kelemen and Meunier 2021) and a notably high extent of European solidarity among Europeans (Ferrera and Burelli 2019; Gerhards et al. 2019) in recent years. The United Kingdom’s exit from the EU has raised public fears about potential threats of political disintegration (De Vries 2018), not only in the UK but also in other EU member states (Malet and Walter 2024). The importance of looking at the pro-European camp separately is also illustrated by Reinl and Braun (2023), who show that people in favor and against EU cohesion constitute two clearly different classes with distinct characteristics.

Moreover, the increasing politicization of European integration has fostered a pro-European backlash, with pro-European mainstream parties benefiting from EU issue voting, especially in the context of the migration crisis and the Brexit referendum (Carrieri 2021, 2024). This connects to the recent literature looking at the period of multiple European crises (Zeitlin et al. 2019), starting with the Eurozone crisis (starting in December 2009) and followed by the migration crisis (2015/2016), the Brexit crisis (starting in 2016), and the COVID-19 crisis (starting in 2020), to name some of the most far-reaching ones. Ferrara and Kriesi (2022) show that only the migration crisis led to a constraining dissensus, while to some extent the Eurozone crisis, but more clearly the Brexit and the COVID-19 crisis fostered European integration. This suggests that politicization does not only produce negative feedback, as claimed by postfunctionalists (Hooghe and Marks 2009, 2018), but can also lead to deeper integration.

3 Stimulating Dissensus in the Multiple Crises Period

I argue that the postfunctionlist claim that there has been a shift from a “permissive consensus” to a “constraining dissensus” (Hooghe and Marks 2009) has the potential to further develop into a stimulating dissensus. As illustrated above, European integration was backed by a wide-ranging public consensus about further (economic) integration steps in the earlier years of the integration process. This permissive consensus developed into a constraining dissensus because European integration increasingly touched more cultural issues, such as national identity. Increasing Euroscepticism among voters and parties constrained the possibilities for further integration endeavors. I argue that in the recent crises period, this constraining dissensus developed into a stimulating dissensus. But what do I mean by that?

I agree with the existing literature that “Euroscepticism is here to stay” (Treib 2021). However, I argue that this does not necessarily imply constraining consequences for European integration. Rather, persistent politicization should awaken the pro-European side in the long run. I argue that pro-European voters become more aware of their preferences regarding European integration because their concept or vision of Europe is increasingly contested. I am well aware that Euroscepticism is not a phenomenon that is specific to the period of multiple crises of the past 15 years. However, what is characteristic for the multiple crises period is the extent of politicization through Eurosceptic actors, for example due to an increasing polarization of parties’ positions on European integration (Hutter and Kriesi 2019), an increasing success of Eurosceptic parties in elections (Filip 2021), and an increasing scepticism of Eurosceptic parties towards European integration in response to the numerous European crises (Braun et al. 2019). The more the European project is attacked by Eurosceptic parties, the less European unification is a matter of course, and the more pro-European voters should become aware of their preferences for a united Europe.

I expect that Eurosceptic radical right parties (ES RRP) in particular have driven this politicization during the multiple crises period. ES RRPs discuss European integration with regard to identity and national sovereignty. Eurosceptic left parties are more concerned with economic issues, such as the neoliberal character of the EU and its consequences for the national welfare state (Hooghe and Marks 2009; Braun et al. 2019). I expect cultural issues to play a much more important role than economic considerations in this period of multiple crises, especially during the migration and the Brexit crises. Additionally, the distinction between hard and soft Euroscepticism is insightful in this regard (Szczerbiak and Taggart 2008). Euroscepticism of left parties tends to manifest in soft Eurosceptic stances, indicated by the opposition to certain EU policies. Eurosceptic radical right parties are better described as hard Euroscepticists because they oppose the EU as such (Treib 2014). Thus, ES RRPs politicization efforts should lead to a stronger politicization because they criticize the European project much more extensively and fundamentally.

Moreover, there has been an increasing politicization not only through Eurosceptic radical right parties, but also through two other sources in the multiple crises period. First, the literature shows that in recent years Europhile parties have increasingly focused on the European integration issue (Carrieri 2021; Adam et al. 2017). They are no longer trying to depoliticize it but are beginning to compete on it in response to Euroscepticism, which should also cue voters who support European integration. Second, multiple European crises have occurred since 2009. I argue that these crises have led to some sort of exogenous politicization across the EU, which is not specific to national party systems. Especially Brexit and the subsequent negotiations explicitly illustrated the consequences of political disintegration (Malet and Walter 2024; De Vries 2018). Moreover, multiple European crises illustrated the dependency on other European countries to find solutions for transnational problems (e.g., during the migration crisis).

Taken together, I expect that politicization during the multiple crises period has not only produced negative feedback, as claimed by postfunctionalists (Hooghe and Marks 2009, 2018), but could potentially open up (new) opportunities for integration by awakening the pro-European camp of voters. This argument is more in line with the neofunctionalist expectation that politicization in times of crises can lead to positive feedback, fostering deeper integration (Pierson 1996; Schimmelfennig 2018).² If voters in favor of Europe give more weight to European integration in their vote calculus, this public support can open up new ground for pro-European parties to find majorities to pursue further integration steps.

This argument has three implications. First, I expect that, overall, the voters’ salience of the European integration issue has increased in the past 15 years due to increasing politicization, not only by Eurosceptic but also by pro-European actors (especially from 2014) and by European crises.

$H1$::

The electoral salience of the European integration issue has increased over time as compared to the electoral salience of other issues.

Second, I expect that Eurosceptic voters still attach the greatest weight to this issue in their voting decisions. The politicization efforts of Eurosceptic parties are originally targeted at voters opposing European integration, and Eurosceptic parties encountered a demand for more nationally oriented policies. This leads to Hypothesis 2a:

$H2a$::

The electoral salience of the European integration issue is most pronounced among Eurosceptic voters.

Third, I expect that the salience of the European integration issue is more pronounced among pro-European voters in countries with a successful Eurosceptic radical right party because these parties are a primary source of politicization at the national level (Hooghe and Marks 2009). Moreover, I also expect that in countries with a successful Eurosceptic radical right party, it is more likely that also Europhile parties respond to the ES RRPs’ strategies and engage with the European integration issue, which should intensify politicization.

$H2b$::

In countries with a successful Eurosceptic radical right party, the electoral salience of the European integration issue among Europhile voters is more pronounced than in countries without a successful Eurosceptic radical right party.

4 Data, Operationalization, and Research Design

I use spatial vote models to estimate electoral salience in multidimensional policy spaces. For this purpose, voters’ and parties’ issue positions on the same issue dimensions are required. I base my analyses on data from the European Social Survey (ESS), waves 4 (ESS 2018) and 8 (ESS 2020), and the Voter Studies of the European Elections Studies (EES) 2014 (Schmitt et al. 2016) and 2019 (Schmitt et al. 2022) to obtain measures of voters’ attitudes on several issues.³⁴ I combine these data with party positions on the respective issue dimensions from the Chapel Hill Expert Survey (CHES) (Jolly et al. 2022). Concerning the question of which issues to consider, most of the literature agrees in finding an established economic issue dimension and distinct issue dimensions encompassing new cultural issues, such as the environment (Crawley et al. 2022; Knutsen 2018; Kitschelt and McGann 1997), gender (Dalton 2017), or immigration (Green-Pedersen and Krogstrup 2008; Van der Brug and Van Spanje 2009; Henjak 2010). Besides European integration, I include the established economic dimension, as well as immigration and environmental issues in my analyses. This is reasonable since the economic left–right dimension is most visible to voters. Immigration and environmental issues are both picked up by an issue-owning party, radical right and green parties, respectively. By this, I ensure that I compare the electoral salience of European integration to other politicized and visible issues.

The analyses cover EU member states from Central, Southern, Northern, and Eastern Europe.⁵ The fieldwork period of ESS wave 4 took place between 25 August 2008 and 12 March 2010,⁶ ESS wave 8 between 22 August 2016 and 28 December 2017. The 2014 and 2019 Voter Studies were conducted after the European Parliament Elections: between 30 May 2014 and 27 June 2014 and between 14 June 2019 and 7 November 2019, respectively.

4.1 Measuring Voters’ and Parties’ Positions in National, Multidimensional Policy Spaces

To measure voter preferences within multidimensional national policy spaces, I use suitable items addressing four broad policy areas: European integration, the economy, the environment, and immigration. Voters’ preferences regarding European integration are measured with the question of whether European integration has already gone too far or should be pushed further. This item captures voters’ preferences regarding the European project as such, not regarding specific policy outputs. By this, it measures hard Euroscepticism on the one side of the voter distribution and committed pro-Europeanism, which goes beyond preserving the status quo, on the other. Thus, this is a rather strict measure of pro-Europeanism. Voters can indeed be pro-European in the sense that they appreciate the current benefits of European integration but are more reluctant towards deeper integration. Items capturing economic, immigration, and environmental preferences vary between surveys.⁷

I combine these voter preference data with party positions on the same latent policy issue dimensions by using data from the Chapel Hill Expert Survey 2010 to 2019 (Jolly et al. 2022). For party positions on the European integration dimension, I use the question on the general EU position of a party; for the economic dimension, I build an index using the CHES items on deregulation, redistribution, economic intervention, and improving social services; for the immigration dimension, I build an index based on multiculturalism and immigration policy; and for the environmental dimension, I use the single available question on environmental sustainability.

Estimating electoral salience in multidimensional policy spaces to obtain meaningful comparisons of the extent of salience is not an easy task, as it requires voter and party positions on multiple issues. The gold standard would be to rely on voters’ perceptions of parties on different issue dimensions since then individual and party positions have been measured on the same scale. Unfortunately, this type of data is scarce and often only available for the ideological left–right dimension. To the best of my knowledge, there exists no data source covering a wide range of EU member states over time with voters’ own policy positions as well as their perceived positions of parties on various specific issue scales. The second-best option is to use expert survey data, which is generally considered to be of very good quality (Marks et al. 2007; Hooghe et al. 2010) and is also used as an alternative in the literature if perceived party positions are not available (e.g. Carrieri 2024; Rosset and Stecker 2019). In a comparative study, Kurella and Rapp (2024) show that party positions based on expert surveys can serve as proxies if perceived party positions are not available.

The first step to bring party and voter positions measured on different scales onto the same scale is to rescale them to the same interval. I choose the interval of $[-1,1]$, where negative values represent left/libertarian and positive values right/authoritarian standpoints. Still, some other problems persist. One problem with linking voter data with external party positions is that the question wordings in both data sources often are not exactly the same. However, slightly differing question wordings are unlikely to bias the results since party positions on concrete issues capturing the same underlying policy dimension are highly correlated. As an example, this is shown for parties’ positions on the economic dimension in the Appendix in Table 1. All available questions on concrete economic policy issues in the CHES, such as parties’ positions on deregulation, redistribution, and tax reduction, are at least correlated with 0.92. Moreover, this is also illustrated by Fig. 11 in the Appendix, which displays all CHES issue items used for the respective issue indices and shows that they are highly correlated within one policy dimension.

Another challenge is that the labels of the endpoints of the scales may not match. For example, in some cases, I find a strong skewness in the voter distribution, which I attribute to the fact that the labels of the endpoints of the voter scales do not always correspond to the extreme points in the national discourse on the topic. I address this issue by rescaling the voter preferences such that the first and 99th percentiles of the voter distribution mark the end points of the issue scale. This procedure ensures that the end points of the rescaled scale represent the most left and right position among the voters. Assuming that experts also consider the national discourse on a topic by aligning their party placement on the left and right extreme points in the electorate, compatibility is ensured. Another problem is that both ESS waves were not collected in the same year as the CHES. In these cases, I perform robustness checks with an alternative matching based on the second closest CHES wave.⁸ Exemplary for Germany 2019, I present the voter distribution and the CHES party positions in Fig. 11 in the Appendix to see how both scales relate to each other.

4.2 Estimation of Electoral Salience

Issue salience has been operationalized in many different ways in the literature. While many rely on “most important problem” questions, I focus on the extent to which voters take policy proximity into account in their individual vote calculus by estimating issue salience based on a Downsian proximity model of vote choice (Downs 1957). The distance between voters’ issue positions and parties’ issue positions is used to predict vote choice. The resulting spatial coefficients capture the relative weight voters attach to a policy dimension in their vote calculus. The spatial coefficients of these vote models can be used as unbiased proxies for electoral salience (Ansolabehere and Puy 2018). Estimating salience directly based on the voter function has the advantage of circumventing the problems associated with “most important problem” questions⁹ and of capturing the importance voters attach to an issue in their vote calculus in a situation in which it also becomes relevant to party competition.

I set up conditional logit models of vote choice, in which I combine the four policy distance terms with party intercepts to account for non-policy-related factors of vote choice. The dependent variable is vote choice.¹⁰ The model is described by the following formula:

$$\mathit{Pr}(y_{i,j}=1|X_{i,j})=\frac{e^{\beta X_{i,j}}}{1+e^{\beta X_{i,j}}},$$

where the linear predictor

$$\beta X_{i,j}=\lambda_{j}-\beta_{1}(x_{1,i}-z_{1,j})^{2}-\ldots{}-\beta_{k}(x_{k,i}-z_{k,j})^{2}+\epsilon_{i,j}$$

combines policy distance terms $(x_{k,i}-z_{k,j})^{2}$ on multiple policy issues, $k$, with a valence term, $\lambda_{j}$ for each party $j$ and an idiosyncratic error term $\epsilon_{i,j}$. The policy distance terms are calculated using the squared city block distance between the ideal point $x$ of voter $i$ on policy issue $k$ and the corresponding party positions $z$. The $\beta$-parameters indicate the decision weight—or electoral salience—of each policy issue.

To detect heterogeneity in issue salience between extreme pro-European, anti-European, and moderate voters on the various issue dimensions, I follow the approach of Kurella and Rosset (2017) and estimate issue salience for each group separately. I construct these pro- and anti-European voter groups by using a cutoff point of $-0.5$ on the left and 0.5 on the right side of the voter distribution on a $[-1,1]$ scale, resulting in left, moderate and right subgroups on all four issue dimensions.¹¹

In the next sections, I conduct three analytical steps. First, I estimate multilevel conditional logit models with random intercepts on the party level for each data set without separate effects for the subgroups to address the general question of how salient the European integration issue is compared to the other three issue dimensions.¹²

Second, I estimate multilevel conditional logit models with subgroup-specific effects to test whether pro-European voters care more about European integration in a politicized context. To do so, I estimate the models separately for countries with and without successful Eurosceptic radical right parties for each point in time. As suggested in the literature (Meijers 2017; Ray 2007), all parties with a value smaller than 4 (on a seven-point scale) on the CHES eu_position item are classified as Eurosceptic. Since I expect that the politicization of the European integration issue is particularly driven by Eurosceptic radical right parties, I only consider parties that are categorized in the CHES as belonging to the radical right party family. I classify an ES RRP as successful if it received more than 5% of the votes in the last parliamentary election before the respective fieldwork period. In most countries, parties enter parliament if they reach at least 5% of the votes, which should increase their visibility and their leeway to politicize remarkably.¹³ Thus, the existence of a successful ES RRP serves as proxy for the politicization of European integration at the national level.¹⁴

Third, I estimate conditional logit models with party-specific intercepts for each country and point in time separately to obtain a more nuanced picture of electoral salience over time.¹⁵

5 Results

The results section encompasses two analytical steps. First, I focus on how pronounced the electoral salience of the European integration issue is compared to other issues. The aim of the first section is to gain a better understanding of the extent of its salience at the voter level over time. Second, I address the question of how important European integration is for pro-European voters and whether they are more aware of their preferences regarding European integration in politicized environments.

5.1 Does Europe Matter to Voters?

I first present the results based on multilevel conditional logit models for each year separately. Figure 2 shows the results.¹⁶ The dots visualize the point estimates, the vertical lines the 95% confidence intervals.

Fig. 2

Multilevel conditional logit models for ESS4, EES 2014, ESS8, and EES 2019

The larger the effect size of the spatial coefficients, the more weight voters attach to this issue in their voting decision, and the larger the electoral salience of that issue. In more technical terms, a large effect size indicates that a small distance between a voter’s position and a party’s position on an issue is associated with a higher probability to vote for that party.

There is an intermittent pattern with a trend towards a more pronounced salience of the European integration issue. In 2008, 2014, and 2016, European integration ranked among the least salient issues. Unsurprisingly, the established economic issue dimension was always the most salient, but also the immigration (2008, 2014, 2016) and the environmental issue (2014, 2016) exceeded the salience of the European integration issue. However, what is remarkable is the turning point from 2016 to 2019. While the European integration issue was the least salient one in 2016, the distance to the other issues narrowed significantly in 2019. It was almost as salient as the immigration and the environmental issue, and the distance to the economic issue had drastically decreased. This is in line with the literature, according to which the salience of the European integration issue increased after the Brexit referendum in 2016 (De Vries 2018; Carrieri 2021). Overall, European integration was not a major electoral issue right before the multiple crises (the Eurozone crisis started in December 2009 with the financial difficulties of Greece becoming public) and also not in the context of the migration crisis and directly after the Brexit referendum. However, the salience of the European integration issue increased significantly in the aftermath of the Brexit referendum. Thus, the results support $H1$ that the electoral salience of the European integration issue has increased over time as compared to other issues, albeit not in a linear process.

5.2 From Constraining Dissensus to Stimulating Dissensus?—Aggregated Results

Having seen that the European integration issue has become a salient issue in recent years, I now show how its salience varies among anti-European, pro-European, and moderate voters. I present the results of multilevel conditional logit models, in which the effects are estimated for anti-European, moderate, and pro-European voter groups separately for countries with a successful Eurosceptic radical right party (ES RRP) and for those without such a party.¹⁷

Figure 3 shows the results. Again, the points depict the spatial coefficients, the vertical lines the 95% confidence intervals. One consistent finding across all data sources is that European integration has always been most salient among Eurosceptic voters, thus supporting $H2a$. Since this is not only the case in countries with a successful ES RRP, the results support the notion of a reciprocal influence of parties’ and voters’ issue priorities. The success of ES RRPs might also be (partly) driven by demand, since they encountered high issue salience among their target electorate on one of their core issues. This gives them even more incentives to push the salience of the European integration issue, which should have a mutually reinforcing effect.

Fig. 3

Multilevel conditional logit models with subgroup-specific estimates for countries without a successful ES RRP (first column) and countries with a successful ES RRP (second column) in 2008/9, 2014, 2016/7, and 2019

The results concerning $H2b$ are more mixed. The pattern looks quite similar in 2008/2009 (first row) and 2014 (second row). The left panels for 2008/2009 and 2014 show that in countries without a successful ES RRP, the European integration issue was not salient at all for voters who support European integration, indicated by the insignificant effects for the Europhile voter group. The right panels for 2008/2009 and 2014 demonstrate a significant effect for pro-European voters. This supports $H2b$: in 2008/2009 and 2014, the electoral salience of the European integration issue among Europhile voters was more pronounced in countries with a successful ES RRP than in countries without such a party.

The pattern looks different in 2016/2017 and 2019. The third and the fourth rows of Fig. 3 illustrate that, independently of whether there was a successful ES RRP, there is a significant effect for the pro-European camp of voters. The difference between the extent of salience of the European integration issue among pro-Europeans and anti-Europeans remains stable regardless of whether there is a Eurosceptic party actively competing on this issue. This does not support $H2b$. Interestingly, there is no pattern specific to the data source used. The results do not suggest that the issue is generally more salient for Europhile voters in the context of the European Parliament Election 2014 and 2019. On the contrary, there is a comparable pattern in 2008/2009 and 2014, as well as in 2016/2017 and 2019. That makes me quite confident that the results are not driven by the increased visibility of European topics in the context of the European Parliament Elections, but that there is indeed an increasing trend over time towards a more pronounced electoral salience of European integration among Europhile voters in countries without a successful ES RRP. One possible interpretation could be that other sources of politicization, besides those of ES RRPs, play a role in the salience of European integration among pro-European voters. As suggested by the literature, Brexit and the subsequent negotiations on the UK’s withdrawal had the power to affect the public mood regarding European integration not only in the UK but also abroad (Malet and Walter 2024). It could also be that the increasing politicization efforts of Europhile parties played a role in 2016 and 2019.¹⁸

Another interesting finding is that the effect sizes for moderate voters increase remarkably when there was a successful ES RRP in 2016 and 2019 compared to when there is not such a party. These moderate voters are neither committed pro- nor anti-Europeans. It could be that these people prefer the status quo because they neither want deeper integration nor consider the integration process to be too far-reaching. One reading is that voters do not care so much about deeper integration but about the further existence of the EU in its current form.¹⁹ However, these “moderate” attitudes could also reflect ambivalence indicating attitude variability or indifference indicating the absence of attitudes regarding European integration (Stoeckel 2013; De Vries 2013; De Vries and Steenbergen 2013).

To sum up, there is a certain increase in salience of the European integration issue among pro-European voters. However, the results show that this happens at a low level; the question is more about whether European integration matters at all to pro-European voters and not whether there is a pronounced pro-European backlash in terms of electoral salience. The electoral salience of European integration is always considerably higher for Eurosceptics, as compared to Moderates and Europhiles.

5.3 From Constraining Dissensus to Stimulating Dissensus?—Country-Level Results

Since the literature suggests that politicization is punctuated and country specific (Kriesi 2016; Grande and Kriesi 2016), I now present disaggregated results to detect heterogeneity in the electoral salience of the European integration issue among different voter groups across countries and points in time. For this purpose, I present results based on conditional logit models estimated for each country and point in time separately.²⁰ Figure 4 gives a summary of these country level results. Figure 4a shows the share of countries in which I find a significant effect for the electoral salience of European integration for pro-European (dark blue bars) and for anti-European voters (bright blue bars) over time.

Fig. 4

Electoral salience of European integration at the country level: a Share of countries with significant electoral salience effects for pro- and anti-European voters; b share of countries with significant electoral salience effects for pro-European voters in countries with and without Eurosceptic radical party

First of all, the results corroborate the finding that it is much more likely that anti-European voters significantly consider the European integration issue in their vote calculus, as compared to pro-European voters. In at least 62% of the countries, I find significant effects for the anti-European camp of voters. On the contrary, in at most 39% of the countries, there is a significant effect for pro-European voters. However, the share of countries in which I find a significant effect for pro-European voters increases over time. In 2008 in 24% of the countries, European integration was an electoral issue for pro-European voters, this share dropped to 14% in 2014, but then increased from 28% (2016) up to 39% (2019). The share of countries with significant effects for anti-European voters also tends to intensify over time; it increased from 62% (2008) to 89% (2016), and then slightly dropped to 82% (2019).

Figure 4b shows the share of countries with significant effects for pro-European voters for countries with (dark green) and without a successful ES RRP (bright green) over time. The country level results also corroborate the finding that the electoral salience of European integration is more pronounced in countries with a successful ES RRP. It is more likely to detect significant effects for pro-European voters in countries with a ES RRP than in countries without such a party, supporting $H2b$.²¹ However, the extent to which this pattern is observed differs over time. For example, in 2008, in 62% of the countries with a successful ES RRP at this time I find a significant effect for the pro-European camp of voters, and in none of the countries without such a party.²² The difference is less pronounced in the two most recent years, especially in 2016. No causal conclusion can be drawn from the data used. However, one potential interpretation of these results is that in 2016 the European integration issue was not only politicized by ES RRP but also by the Brexit referendum and the resulting societal discussions about European integration all over Europe.

Taken together, the European integration issue is most electorally decisive for Eurosceptic voters. However, there is also an increasing tendency for the issue to gain importance among pro-European voters, even though at a remarkably lower level than among Eurosceptic voters. It is more likely that Europhile voters care about European integration if there is an ES RRP competing on that issue within their country.

In a final analytical step, I take a closer look at some particularly interesting cases with a strong Eurosceptic party and/or an important Europe-related event, as I then expect a high politicization of European integration and thus a higher salience of European integration for pro-European voters. Figure 5 shows the salience estimates for the European integration issue in France from 2008 to 2019. What makes France a particularly interesting case is that there is a well-established Eurosceptic radical right party (Front National/Rassemblement National). Moreover, Emmanuel Macron was elected president in 2017 with a strikingly pro-European election campaign (Bouza García and Oleart 2022). The analyses so far have shown that the European integration issue is usually not highly salient among pro-Europeans and most salient for Eurosceptic voters. The pattern looks somewhat different in France. In 2008/2009, the European integration issue was only salient among Eurosceptic voters. In 2014 and 2016/17, it was salient among the pro- and anti-European camp to a similar extent. In 2019, the issue was still important for pro- and anti-Europeans, but more for the latter group. However, it was still more salient among pro-European voters than among moderate voters. The data do not allow for a causal conclusion, but the increasing success of Front National/Rassemblement National (the party increased its vote share from 4.29 (2007) to 13.6 (2012) and 13.2% (2017) and has gained parliamentary representation since 2012) coincides with significant effects for Europhile voters. The slight increase in salience among voters supporting European integration already seems to open up a window of opportunity for political leaders to succeed with pro-European election campaigns. These results also support the notion that Macron already encountered significant salience among pro-Europeans, as there was already a significant effect for the pro-European camp in 2014. The party En Marche, with which Macron ran in 2017, was founded in April 2016.

Fig. 5

Electoral salience of the European integration issue in France 2008–2019

Another interesting case with strong Eurosceptic forces is England. There are several smaller Eurosceptic parties (British National Party, UKIP, Brexit Party) and also pronounced Euroscepticism among voters (Vasilopoulou 2016), as well as an intense public discourse revolving around the Brexit referendum. The Conservative Party also engaged with the EU issue by initiating the Brexit referendum. Figure 6 shows the salience estimates in England between 2008 and 2019. While the European integration issue was not of importance for pro-Europeans in 2008/2009, 2014, and 2016/17, the picture changed in 2019 in the aftermath of the Brexit referendum. One plausible explanation is that this extraordinary event of politicization of EU issues has shaken up the reluctant pro-Europeans and made them aware of the consequences of political disintegration, finally and led them to care about European integration and consider it in their vote calculus.

Fig. 6

Electoral salience of the European integration issue in England 2008–2019

Figure 7 presents the salience estimates in the Netherlands. There has been a well-established strongly Eurosceptic party (Partij voor de Vrijheid) since 2008. Except for in 2014, the salience of the European integration issue was relatively high among pro-Europeans, even more than among Eurosceptics in 2008/2009. The presence of the ES RRP comes with relatively stable levels of salience of European issues among Europhiles. The Netherlands is also the first country in which the strongly pro- and pan-European party Volt entered a national parliament in 2019. The results support the notion that already small increases in salience could provide opportunities for parties to succeed with strikingly pro-European election campaigns.²³

Fig. 7

Electoral salience of the European integration issue in the Netherlands 2008–2019

Two other interesting cases are Hungary (Fig. 8) and Poland (Fig. 9). In both countries, there are successful Eurosceptic parties. Victor Orbán from the nationalist and Eurosceptic Fidesz has been Hungarian president since 2010, and in Poland the Eurosceptic PiS party was in power from 2015 to 2023. However, there is a pattern of at least punctuated salience among Europhile voters in both countries. In Hungary, the electoral salience among Europhiles was almost as strong as among Eurosceptics in 2014 and 2019. In Poland, the salience of the European integration issue was even more pronounced in the pro- than in the anti-European camp of voters in 2008/2009 and 2016/2017.

Fig. 8

Electoral salience of the European integration issue in Hungary 2008–2019

Fig. 9

Electoral salience of the European integration issue in Poland 2008–2019

Zooming in the country level shows that there are indeed cases in which Europhile voters care about European integration, and that this is particularly the case in politicized environments. However, there are also countries with strong Eurosceptic parties in which the pro-European side does not care about European integration at all. As an example, Fig. 10 shows the salience estimates for Sweden. Although the Swedish party system includes a particularly successful Eurosceptic radical right party (Sverigedemokraterna), there are no significant effects for supporters of European integration throughout the period studied.

Fig. 10

Electoral salience of the European integration issue in Sweden 2008–2019

Thus, the country-level results support the notion of punctuated politicization of the European integration issue reflected in time- and country-specific variance of its electoral salience. Future research should delve deeper into the specific national circumstances to disentangle the concrete mechanisms of the (non-)existence of electoral salience of European integration for pro-European voters.

6 Conclusion

The past 15 years have been characterized by rising Euroscepticism and multiple European crises such as the Eurozone crisis, the migration crisis, and Brexit, fostering the politicization of the European integration process. However, there is little evidence of how these developments have affected voters holding pro-European attitudes. I examine whether Europhile voters have become more aware of their preferences regarding European issues in a time period of increasing politicization.

Overall, the results show that the salience of the European integration issue has increased in recent years. After the European Election in 2019, European integration was almost as salient as immigration and economic issues and as salient as environmental issues. In 2008, 2014, and 2016, the issue played a subordinate role in comparison to economic, immigration, and environmental issues. Moreover, the results demonstrate that the electoral salience of European integration is the most pronounced among voters opposing European integration. The salience of European integration is clearly less visible among Europhile voters. However, it tends to be more pronounced in countries with successful Eurosceptic radical right parties, especially in 2008 and 2014. In 2016 and 2019, there was a “baseline” salience of European integration in the pro-European camp of voters, independently of the existence of a successful Eurosceptic radical right party, indicated by small but significant effects for Europhile voters. This finding suggests that there are also other sources of politicization, which are not specific to the national party system. This is in line with the finding by Malet and Walter (2024) that the Brexit decision and the subsequent negotiations had the power to affect the public mood regarding European integration also abroad.

Looking at countries with particularly strong Eurosceptic radical right parties reveals interesting variation. For example, France and the Netherlands stand out for their relatively high salience of the European integration issue among pro-European voters and for the success of Europhile actors in campaigning on this issue. Even in countries such as Hungary and Poland, which are known for their pronounced anti-EU stances, the salience of the European integration issue is surprisingly pronounced in the pro-European camp of voters. However, salience of European integration among Europhile voters does not always increase in contexts with strong Eurosceptic parties. For example, I do not find any significant effects for the pro-European camp of voters in Sweden. Therefore, future research should further disentangle the mechanism under which the salience of European integration is particularly high or low, beyond the presence of Eurosceptic radical right parties. In particular, the (limited) politicization efforts of Europhile parties and their impact on voters should be addressed. The hesitancy of pro-European voters to raise the salience of European integration could also be due to the opposing pressures Europhile parties face. They are confronted with widespread public Euroscepticism, while at the same time being constrained by their previous ideological commitments, in which they were driving forces of European integration.

One major limitation of the analyses in this paper is that the data used do not allow for making any causal claims because contextual factors are only considered indirectly. With the data available, I can only observe the development of electoral salience of European integration in a period of increasing politicization. This is still an important contribution since it provides an overview of this development in a wide range of EU member states between 2008 and 2019. However, I cannot causally identify whether heterogeneity in salience is indeed due to the politicization efforts of the ES RRPs, whether parties react to voter-induced shifts in issue importance, and which role other factors, such as the strategies of Europhile parties, play. Thus, future research should use panel data to disentangle the interplay of parties’ and voters’ issue salience, as well as other sources of politicization.

Moreover, it is least likely to find pronounced pro-European counter-reactions at the voter level since the politicization literature shows that engagement with the European integration issue at the individual level is still substantially lower than that of actors in parties or the mass media (De Wilde et al. 2016). Again, this stresses the importance that future research should also focus more on the party level, especially on the strategies of Europhile parties, how they interplay with Eurosceptic parties’ strategies, and the public mood regarding European integration.

Supplemantary Information

Operationalization of Voters’ Issue Positions