*Edited* 

<sup>b</sup>*y*

Marina Costa Lobo

The Impact of EU Politicisation on Voting Behaviour in Europe

#### PALGRAVE STUDIES IN EUROPEAN UNION POLITICS *SERIES EDITORS:* MICHELLE EGAN · WILLIAM E. PATERSON · KOLJA RAUBE

# **The Impact of EU Politicisation on Voting Behaviour in Europe**

*Edited by* Marina Costa Lobo

# Palgrave Studies in European Union Politics

Series Editors

Michelle Egan, American University, Washington, USA William E. Paterson, Aston University, Birmingham, UK Kolja Raube, KU Leuven, Leuven, Belgium

Following on the sustained success of the acclaimed European Union Series, which essentially publishes research-based textbooks, Palgrave Studies in European Union Politics publishes cutting edge research-driven monographs. The remit of the series is broadly defined, both in terms of subject and academic discipline. All topics of significance concerning the nature and operation of the European Union potentially fall within the scope of the series. The series is multidisciplinary to reflect the growing importance of the EU as a political, economic and social phenomenon. To submit a proposal, please contact Senior Editor Ambra Finotello ambra. finotello@palgrave.com. This series is indexed by Scopus.

#### **Editorial Board:**

Laurie Buonanno (SUNY Buffalo State, USA) Kenneth Dyson (Cardiff University, UK) Brigid Laffan (European University Institute, Italy) Claudio Radaelli (University College London, UK) Mark Rhinard (Stockholm University, Sweden) Ariadna Ripoll Servent (University of Bamberg, Germany) Frank Schimmelfennig (ETH Zurich, Switzerland) Claudia Sternberg (University College London, UK) Nathalie Tocci (Istituto Affari Internazionali, Italy)

Marina Costa Lobo Editor

# The Impact of EU Politicisation on Voting Behaviour in Europe

*Editor*  Marina Costa Lobo Institute of Social Sciences University of Lisbon Lisbon, Portugal

ISSN 2662-5873 ISSN 2662-5881 (electronic) Palgrave Studies in European Union Politics ISBN 978-3-031-29186-9 ISBN 978-3-031-29187-6 (eBook) https://doi.org/10.1007/978-3-031-29187-6

© The Editor(s) (if applicable) and The Author(s) 2023. This book is an open access publication.

**Open Access** This book is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this book are included in the book's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the book's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.

The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This research was supported by the MAPLE project (European Research Council Grant no. 682125)

Cover illustration: Magic Lens/Shutterstock

This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG

The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

# Acknowledgements

This book is the culmination of five years of collaborative work in the ERC MAPLE Project- Mapping and Analysing the Politicisation of Europe, before and after the Eurozone Crisis. Most of the contributors to this volume were or still are members of the MAPLE Project. I am therefore extremely grateful to Lea Heyne, Roberto Pannico, Tiago Silva, Yani Kartalis, Susana Rogeiro Nina, and Nelson Santos. I have been lucky enough to hire researchers who became my friends and each of them has made an invaluable contribution to the Project's achievements.

Besides the project's members, the book benefited enormously from the contributions of Marc Debus, Hugo MarcosMarne, Roula Nezi, Rosa Navarrete, and Dieter Stiers, who agreed to join this endeavour even in the difficult pandemic times.

This book would not be published without the invaluable help of all the graduate students who manually coded thousands of media articles and parliamentary debates and they deserve a special thanks.

During the book's preparation, the authors were able to present earlier versions of the Chapters in various conferences, namely EPSA, ECPR workshops, ECPR General Conference, Standing Group on European Union, and IPSA. We received useful feedback from several colleagues, including Katjanna Gatterman, Vin Arceneaux, Romain Lachat, Ignacio Lago, Christian Rauh, Liisa Talving, Catherine Moury, Ignacio Jurado, Hermann Schmitt, Eftichia Teperoglu, Wouter van der Brug. Most Chapters have been presented in the ICS weekly SPARC Seminar at the Institute of Social Sciences, and we wish to thank all the members who participate in it regularly for their feedback, namely Pedro Magalhães, Edalina Sanches, Rui Costa Lopes, as well as the graduate students who attend SPARC.

Thanks to the insights provided by the Advisory Board, namely, Michael Lewis-Beck, José Ramon Montero, Dimitri Sotiropolous, Lieven de Winter, Virginie van Ingelgom, and Martin Rosema.

Institutionally, I wish to acknowledge the ERC Grant no. 682125 which generously financed this volume, as well as the MAPLE Project. The Institute of Social Sciences in Lisbon always supported the MAPLE Project and our goals, in what were, objectively, very difficult circumstances. In particular, I would like to thank the Director Karin Wall, António Martinho, Eugénia Rodrigues, and José Pedro Cardeiro for their support along the way.

The authors are grateful to our reviewers at Palgrave Macmillan, and to the editorial team, Ambra Finotello and Hemapriya Eswanth, for their support throughout the entire process.

My final and sincere thanks go to my family, my husband Rui and my daughters, Inês and Vera, who over the past years have had to hear me endlessly on such strange terms as politicisation, salience, or supranationalisation, and put up with much absent-mindedness from me.

# Contents






# Notes on Contributors

**Marina Costa Lobo** is Research Professor in Political Science at the Institute of Social Sciences of the University of Lisbon where she was PI of ERC Project MAPLE. She is Director of the Portuguese Election Studies, a member of the Planning Committee of CSES, and Editor of the *European Political Science Review*. Her research interests are mainly in political behaviour, especially short-term factors, as well as Europe's influence on legislative elections.

**Marc Debus** is a professor of political science, comparative government, at the University of Mannheim, Germany. His research interests include political institutions, in particular in multilevel systems, and their effects on the political behaviour of voters and legislators, as well as party competition, coalition politics, and decision-making within parliaments and governments.

**Lea Heyne** is a postdoctoral researcher at the Institute of Social Sciences, University of Lisbon. Her work focuses on the relationship between democratic quality, political attitudes, and electoral behaviour.

**Yani Kartalis** is a postdoctoral researcher at the Copenhagen Center for Social Data Science (SODAS) and a Data Steward at the Center's Social Science Data Lab. He is also a Research Fellow at project HOME, at the Center for International Studies at Lisbon University Institute (CEI-IUL).

**Hugo MarcosMarne** is an Assistant Professor at the Department of Political Science of the University of Salamanca, where he is also a member of the Democracy Research Unit (DRU). His research focuses on the intersection between political attitudes and electoral behaviour, national identities, and populism.

**Rosa M. Navarrete** is a postdoctoral researcher and teaching fellow at the Saarland University. She previously worked as a researcher at the Mannheimer Zentrum für Europäische Sozialforschung (MZES), as a research associate at the University of Mannheim, and as a researcher at the University of York and at the Autonomous University of Madrid. Her work has been published in *Party Politics, Electoral Studies, Comparative European Politics* and *Journal of Elections, Public Opinion and Parties*  among others.

**Roula Nezi** is Senior Lecturer (Associate Professor) in Political Science at the University of Surrey and a Senior Fellow of the Higher Education Academy. Her research examines developments in public opinion and political attitudes, political parties and party systems, and electoral behaviour. Her research has been published in the *British Journal of Political Science*, *Party Politics*, the *Journal of European Social Policy*, and *Electoral Studies* among others.

**Susana Rogeiro Nina** is Assistant Professor in Political Science, European Studies, and International Relations at Lusófona University. She completed a Ph.D. in Political Science at the Institute of Social Sciences of the University of Lisbon within the ERC MAPLE Project. Her research interests include the EU, political communication, and media effects.

**Roberto Pannico** is Ramón y Cajal Fellow at the Department of Political Science of the Autonomous University of Barcelona. Previously, he has been Beatriu de Pinós Fellow at the University of Barcelona and postdoctoral researcher for the ERC MAPLE Project at the Social Sciences Institute of the University of Lisbon. He completed a Ph.D. in Political Science at the Autonomous University of Barcelona in 2017. His research interests include attitudes towards the EU, economic voting, information processing, party cues, and experimental methodology.

**Nelson Santos** is a Ph.D. student in Comparative Politics at ICS-Ulisboa. He holds a Master in Economics and Public Policies from ISCTE-Lisbon University Institute, and completed his B.A. in Political Science at the Nova University, Lisbon. His research interests include European integration and politics, parliaments, and computational social science methods.

**Tiago Silva** is a researcher at the Institute of Social Sciences (University of Lisbon), previously working for the ERC project MAPLE. He holds a M.A. in Political Science, from University of Aveiro and a Ph.D. in Political and Social Sciences from the European University Institute.

**Dieter Stiers** is a postdoctoral researcher of the Research Foundation Flanders (FWO) at the Centre for Political Science Research of KU Leuven (Belgium). His research focuses on elections and voting behaviour, with a special interest in mechanisms of electoral accountability and the congruence of the vote.

# Abbreviations




# List of Figures






# List of Tables


#### xxviii LIST OF TABLES


# Holding Europe Accountable Nationally: Media, Parliaments and Voting in Europe

*Marina Costa Lobo* 

# Introduction

This book answers a puzzle which has emerged within the European Union (EU) in the twenty-first century: the EU's importance has unequivocally grown for decision-making, both in politics and policies following a decade of concatenated crises, from the Eurozone in 2009, the refugee crisis in 2013, Brexit in 2016 to the covid crisis in 2019. At the same time, accountability mechanisms at the EU level have remained largely unchanged since the signature of the Treaty of Lisbon in 2007, even though support for the integration process has varied and currently seems to have largely returned to pre-crises levels. Most of the studies who attempt to understand the way in which the EU is being legitimised focus on the European Parliament elections. Indeed, the issue of EU accountability has led to a large literature around the concept of secondorder elections, and how the preferences about the EU are channelled

© The Author(s) 2023

1

M. Costa Lobo (B)

Institute of Social Sciences of the University of Lisbon, Lisbon, Portugal e-mail: marina.costalobo@ics.ulisboa.pt

M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_1

through these Europe-wide elections (Gattermann et al., 2021; Hix & Marsh, 2011; Reif & Schmitt, 1980).

Instead, we argue that to fully understand how EU accountability works, it is necessary to focus on the national political environment and national elections. This channel of accountability has been established (de Vries, 2007), but is still poorly understood. In this book, our goal is to determine the EU's national channel of accountability in two steps: first, through an extensive analysis of politicisation of the EU in media and parliamentary debates, to show the media and party context which provides cues for voters to formulate preferences about the EU. Second, establishing the importance of the EU for voting in legislative elections through both experimental and observational analysis, and examining, in detail, the contexts which condition it.

There have been a number of studies which focus on EU politicisation, as well as its importance for EU issue voting. Yet, we believe our study is timely due to two new developments in the nature of the EU that arguably warrant more focus on this topic. First, the onset of the Eurozone crisis has been seen as an important moment for EU politicisation at the domestic level (De Vries, 2018; Hoeglinger, 2016; Hooghe & Marks, 2018; Hutter & Kriesi, 2019; Ruiz-Rufino & Alonso, 2017; Schäfer & Gross, 2020), and it has given rise to a large debate on the consequences of "Europe" for electoral behaviour (De Vries & Hobolt, 2016; Lobo & Lewis-Beck, 2012, Lobo & Pannico, 2020). Second, the nature of the EU has evolved from a regulatory to a distributional power (Börzel & Risse, 2018). Whereas previously it could be argued that the pursuit of electoral legitimacy by the EU was unnecessary, since it consisted of a regulatory power, the monetary union and the Eurozone crisis have made clear that the decisions taken at the EU level actually have important distributional consequences, not only between member-states, but also within countries (Börzel, 2016). Both these trends intersect with national politics, more precisely through the media, parliamentary, and electoral arenas.

Our goals determined our choice of country cases: namely Belgium, Germany, Greece, Ireland, Portugal, and Spain. These countries were selected on the basis of their different trajectories within the EU both in political and economic terms. Whereas Belgium and Germany are founding members, the others joined at later moments. Economically, the countries differ in performance and underwent different paths following the onset of the Eurozone crisis, with Germany and Belgium performing relatively well and the other countries having to implement bailouts.

The findings have implications for the ongoing debates on EU democracy, since they underline the importance that national institutions media, parliamentary debates, and national elections—have in enabling citizens to hold politicians accountable for their EU positions. This suggests that when discussing EU quality of democracy and how it can be improved, it is necessary to take into account national political institutions.

# The Book's Argument

The EU's importance is undeniable in member-states' policy-making. From the euro to agriculture, from competition to the single-market, European-level decision-making influences the design of member-states' policies. The reality of the EU is everywhere, and it has gained in importance in the last decades, both in level and in scope (Börzel, 2005), as a result of the successive crises (Genschel & Jachtenfuchs, 2016), and may even have undergone a paradigmatic shift in policies and processes in the health and economic domains following the covid-19 crisis (Schmidt, 2020).

Concerning policy-making, the Maastricht Treaty and the implementation of European Monetary Union was a turning point for European integration, as it opened the way for further deepening. The introduction of the euro as well as free movement of citizens made it clear that the EU had become a part of citizens' everyday life. Hooghe and Marks' (2009) seminal contribution posited that, since Maastricht, the EU had shifted from a "permissive consensus" to a "constraining dissensus", where citizens and their views on Europe had to be taken into account when considering further steps in European integration.

In the last 15 years, marked by successive crises, the EU has become more political, as its institutions have often intervened in a discretionary way to respond to the multiple challenges that have arisen (Middelaar, 2019). During the Eurozone crisis, in the response to covid-19, as well as following Russia's invasion of Ukraine, the EU has responded in a purposeful way, rather than simply following rules and norms set out by the Treaties.

Yet, the deepening of EU integration through the effective supranationalisation of policies and politics has not been accompanied by a change in the Treaties, which would formally increase the level of accountability of the EU, as would perhaps have been expected by neo-functionalists and federalists alike. This, therefore, leads us to pose the following question: Where are then, the deepened channels of accountability to accompany the renewed importance in decision-making which the EU has acquired? While many scholars continue to monitor the functioning of EP elections to detect a Europeanisation of political behaviour at that level, this book argues that accountability is occurring at the national level. It follows this argument through an examination of politicisation at the media and parliamentary levels, and then on the importance of EU issue voting at the national level and how it is shaped by the informational and party contexts.

# The Book's Contribution

Given that the successive crises have increased EU politicisation, and that these have asymmetric distributional consequences, it is necessary, beyond establishing its existence, to understand whether EU accountability works similarly in countries at opposite ends of these consequences. In this study, we characterise EU politicisation in a multidimensional fashion. Then we establish EU issue voting and the factors which shape it, focusing on the relation of EU issue voting with left–right, the media and the parliamentary debates context in each of the countries included in the study.

Our multidimensional and multimethod analysis of politicisation across two arenas (media and parliamentary debates) shows that politicisation is present in them, with the share of negative tone in media being considerably higher than in parliamentary debates. Also, in both arenas, the main focus of EU articles and debates is "policies" rather than the "polity" dimension. These findings are common to all countries, and suggest that both arenas are contributing to enable citizens to form opinions on the EU. Yet, differences emerge in the types of policies being politicised in each country both in the media and parliamentary debates, suggesting EU differentiation between countries. Namely, in some countries domesticated EU policies are being discussed, rather than EU policies *strictu*  sensu or any institutional EU issues.

Turning to our voting behaviour analysis we find that European citizens do take into account their attitudes towards the EU when making electoral choices. This finding is robust to different methods, namely using experimental and observational data. Moreover, EU issue voting is asymmetric, with parties that are more Eurosceptic than the respondent being punished on average more than parties that are more pro-EU. While the importance of EU issue voting is established in very clear terms, and we find it does cross-cut the left–right issue, it is not replacing the left–right issue in the vote calculus.

As regards the contexts which shape EU issue voting, we find that the media plays a role in shaping the relative electoral chances of pro-EU and Eurosceptic parties, with low salience and high negativity contributing to undoing the asymmetry detected, on average, in EU issue voting. Regarding parliamentary debates, they are a source of information especially for citizens who vote for Eurosceptic parties. This suggests that in a post-crisis context, the politicisation in the media and parliamentary debates contributes to the strengthening of EU issue voting, thus underlining the national channels of European accountability which exist in Europe today.

We turn now, in this introduction to the book to a brief survey of the concept of EU politicisation and how it was measured both for media and parliamentary debates. Next, we explain the two main channels of EU accountability, at the EU and national level, and how we have focused on the characterisation of national channels of EU accountability in this book. After describing the various datasets which were collected in preparation of the book, the final section outlines the book chapters.

#### EU Politicisation

Politicisation refers to "an increase in polarization of opinions, interests or values and the extent to which they are publicly advanced towards the process of policy formulation" (de Wilde, 2011, p. 559). Scholars of EU politicisation have focused on how contestation over regional integration connects to domestic conflict, and on how this contestation influences the speed and direction of regional integration (Hooghe & Marks, 2009). By stating that since Maastricht, public opinion had evolved from a "permissive consensus" to a "constraining dissensus", the post-funcionalist theory, proposed by Hooghe and Marks (2009), was fundamental to turn the focus on public opinion towards the EU.

Measuring EU politicisation has not been systematically done in similar ways, and thus, the evidence which has been accumulating is not totally comparable (Hutter & Kriesi, 2019; Rauh, 2015; Risse, 2015). Studies differ in the way they measure politicisation, in the way they collect the data, and in the forum where the data is collected. More concretely, some studies focus mostly on salience, while others include not only salience but also polarisation (Lobo & Karremans, 2018). Yet, there is an issue of whether it would have been important to include actors too (Hutter & Kriesi, 2019). Namely, for any issue to be politicised, it needs to be salient, it has to be divisive, and it needs to be divisive among a large number of political actors.

Considering media studies, research is accumulating in favour of an increase in EU politicisation, even though there are some sceptics (De Wilde & Zürn, 2012; Green-Pedersen, 2012; Hutter & Grande, 2014; Kriesi et al., 2013; Statham & Trenz, 2013). Green-Pedersen (2012) concludes that there has been little EU politicisation, by looking at both media and party programmes in Denmark. Similarly, Hoeglinger argues that the EU politicisation has been limited (2016, p. 146). On the other hand, other studies find that EU politicisation has been increasing (Hutter & Grande, 2014; Statham & Trenz, 2013, p. 169).

In their comprehensive comparative media study, Hutter and Kriesi (2019) find evidence that the Eurozone and refugees crises contributed to the party-in-the-media politicisation of the EU on a regional basis. These authors do not find a linear increase in EU politicisation. On the contrary, they point to a punctuated politicisation, which increases both due to anticipated (e.g. EP elections) or unexpected (e.g. crisis) events. More recently, Silva et al. (2022), using automated text analyses, show that the Eurozone crisis increased the media politicisation of the EU, in Greece, Portugal, Spain, and Ireland vis-à-vis Germany and Belgium.

Other literature on politicisation has moved beyond the salience and polarisation issue. Silva et al. (2022) highlight that the crisis brought a greater emphasis by the media to supranational institutions vis-àvis intergovernmental ones and an increasingly negative tone of articles. Hurrelmann et al. (2015) distinguish between EU dimensions in all arenas. Namely, *membership*—costs and benefits of being in the EU; *constitutional structure*—institutions, decision-making processes; EU *policy issues*—European policies on the agenda; and *domesticated issues* national issues deriving from one's own country membership. Whereas the first two dimensions pertain to polity, the last two concern policy issues. This differentiation highlights the need to evaluate if the EU polity itself is being contested, or whether only policies which would suggest a Europeanisation of political debates.

In relation to EU politicisation in parliamentary debates, there has been considerably fewer studies. When considering parliaments, the literature on EU politicisation has focused mainly on parliaments' legislative and oversight roles, rather than its communicative function. In order to study the latter, it is necessary to concentrate on plenary debates (Auel & Raunio, 2014a, 2014b). Compared to media, parliamentary debates offer an unmediated access to party positions on issues. Certainly, there are issues of comparison, as parliamentary speeches are constrained by different institutional rules across countries. Still, they remain an important and relatively untapped resource to understand party positions.

The existing studies point to divergences concerning the importance of parliamentary debates. On the one hand, it has been perceived that the EU has been politicised in the parliamentary arena (Hurrelmann et al., 2015). On the other hand, some studies have found that there is a lack of debates on the EU (Rauh & Wilde, 2018) and that the level of politicisation has been generally low (Auel & Raunio, 2014b). Nevertheless, the effect of the crisis seems to have contributed to an increase in EU politicisation, which tends to depend on periods around important EU events. Several studies have pointed to the fact that, beyond institutional differences between parliaments that may affect comparability, the party system and the existence of Eurosceptic parties in Parliament are the most important predictors of EU politicisation in plenary debates (Auel & Raunio, 2014a, 2014b; Wendler, 2016).

In this book, politicisation was measured both in automated fashion, to measure salience and polarisation, as well as through manual content analysis to probe into the dimensions of EU topics discussed in the media and parliamentary debates. Each of the measurements is explained in detail in Chapters 2 and 3, and here we detail how these measures were conceptualised and how they complement each other to form a complex perspective of EU politicisation.

Concerning the measurement of EU politicisation in the media, the focus was not, as previous studies had done, on political actors' statements covered by the media (e.g. Hutter & Kriesi, 2019; Hutter et al., 2016). Instead, the book considers all articles from the relevant (political, economic, and business) sections of each newspaper 30 days before each election date. As in Schmidtke (2016), EU salience is operationalised as the percentage of articles, given our sample, that were about the EU topic, based on a list of EU terms. When a newspaper article contained more than one EU-related term, or if one term appeared in the title, the article was coded as being about the EU. The same principle applied to parliamentary speeches, using the same list of terms. Yet, for parliamentary speeches, the data corpus included all years from 2002 to 2017.

The concept of polarisation or contestation in the media was examined by looking at the proportion of articles, in a year/newspaper, with a negative tone towards the EU. This was done using sentiment analysis after translating all the corpus to the English language, using the English language sentiment dictionary. The same method was applied to the parliamentary debates. These measures of salience and tone were then used to compare trends in politicisation across arenas in Chapter 4, as independent exogenous variables to explain EU issue voting in Chapters 5 and 6, and as media and party contexts to the in-depth case studies covered in Chapters 7–12.

Further, manual coding was employed to probe deeper the EU dimensions which were being politicised in the media, following Hurrelmann et al.'s (2015) typology. Native language researchers were recruited to code, for each article about the EU, if the EU issue mainly dealt with in the article concerned (1) membership, (2) constitutional structure, (3) EU policies, or (4) domesticated policies as explained above, as well as the general topic of the policies discussed. The same frame and method was applied to a sample of parliamentary debates held in the year prior to each national election, to understand the nature of the EU topics which are being debated.

Our analysis of media and parliamentary debates politicisation shows that it is present in both arenas, and increased on average following the crisis in the media but not to the same extent in parliamentary debates. Second, when we distinguish between politicisation in news articles vs. op-eds in the media, differences in EU stances between left and right in the mainstream media can be observed, and these increased following 2009. While politicisation increased in all countries' media following the Eurozone crisis (especially if we consider op-eds), it did so focusing on different EU topics. In Spain, *domesticated policies* have always been the most salient dimension in articles about the EU, with Portugal following its neighbour, from 2011 onwards. Ireland stands out since, in all of the four elections analysed, the EU was mostly debated in terms of *EU policies*, until 2016 where the *membership* dimension became dominant. Finally, in the Greek case, we find the only instance where, at the peak of the crisis, the *membership* dimension became the most important dimension. Finally, Germany tends to discuss *European policies* to a greater extent than the other countries in the sample.

When comparing the two arenas, we find that policy-related issues are, in general, more addressed than polity-related ones, both in the media and parliamentary debates, which is positive from the perspective of EU accountability. Within policies, it is economic- and financialrelated matters ("Economy and Work" and "Finances and Taxes") which dominate the European debate, both in the media and parliamentary debates.

A major important difference emerges between parliaments and media, however. Namely, the share of negative tone in media is considerably higher than in parliaments, and the Eurozone crisis reinforced that trend. This multidimensional and multimethod analysis of politicisation across two arenas (media and parliamentary debates) suggests the following: politicisation is present, and therefore fosters EU accountability, despite working differently depending on the arena. Yet, when we deepen the analysis, differences emerge in the dimensions of the EU being politicised in each country, suggesting EU differentiation, with Germany debating more EU policies than Southern European countries.

#### Holding the EU Accountable?

The chain of responsiveness—what occurs when the democratic process induces the government to form and implement policies that the citizens want—is perhaps one of the most important elements of the quality of democracy (Powell, 2004). Powell established the main elements of this chain, which include the establishment of voters' preferences, the polarisation of parties on key issues, the ability of voters to select parties based on different policies, as well as the ability of politicians to implement the preferred policies. In each step of this causal chain, there are numerous factors which may break or strengthen it. While it is exceedingly difficult to verify how the chain of responsiveness is working, Powell argues that vertical accountability, i.e. the ability to hold politicians responsible for their preferred policies at elections, plays a big role in ensuring the chain's strength and thus the quality of democracy.

Indeed, traditionally, the main focus of studies which considered EU chain of responsiveness and quality of democracy, have focused on *vertical*  accountability, processed through European Parliament (EP) elections. Many studies have criticised the EU for lacking democratic legitimacy (Follesdal & Hix, 2006; Hix, 2013). At the centre of the "EU democratic deficit" critique has been the second-order nature of the European Parliament elections. As the only EU institution which is directly elected, scholars have established that these elections are not really about European issues, but rather national ones (Reif & Schmitt, 1980). The concept of European elections being second-order follows from a number of principles about the EU. While EP competences in decision-making have increased substantially, there is still no direct link between voting for the EP and the formation of the European Commission, which is largely still decided by national governments. Also, despite the increased powers of the EP, the European Commission and Council still wield dominant power, and the successive crises since 2009 have only served to reinforce this trend. Due to the nature of the electoral system, at present, voters in each country select national parties to represent them. As such, they tend to consider the national political context and use the EP election to punish or reward the national incumbent and mainstream parties. Following from these principles, second-order elections such as the EP elections are supposed to have higher abstention levels, higher votes for extreme or protest parties, and fewer votes for incumbent or mainstream parties, vis-à-vis the national legislative elections (Schmitt, 2005). Indeed, a lot of research on European Parliament elections has found that these remain second-order national elections (Nielsen et al., 2017).

Nevertheless, the European Parliament elections in 2019 did not follow strictly the expectations of a second-order election (Gattermann et al., 2021). For the first time since the 2004 enlargement, EP elections saw an increase in turnout, which surpassed the symbolic 50% threshold. Yet, overall, in terms of individual vote choice, Ehin and Talving (2021) have shown that the second-order model is still supported for those elections.

Thus, by and large, EP elections continue to be second-order, and this makes it even more obvious that we need to consider the way European issues are processed through national channels of accountability during national elections. Indeed, EU accountability is a multi-level system of governance that enjoys dual electoral legitimacy. On the one hand, citizens can express preferences through a direct vote at European Parliament elections. On the other hand, in national legislative elections, they also select representatives who express EU preferences that are congruent with their own (De Vries & Hobolt, 2016). These electoral choices, at the national level, then contribute to the formation of governments who make decisions at EU-level bodies, namely the Council of Ministers and the European Council (Schneider, 2019). This dual legitimacy system can only work, however, if we establish the existence of EU issue voting at the national level. Beyond establishing its importance, in a comparative perspective, this book explores the national contexts which foster or discourage the expression of EU preferences at the national ballot box. Thus, the chain of responsiveness and ensuing quality of EU democracy will depend on the existence of EU issue voting at the national level.

When considering EU issue voting, it has been studied at the national level mainly in three ways. Namely, identifying its existence, explaining its determinants, and, to a lesser extent, examining its consequences. De Vries (2007) was the first to establish the evidence of EU issue voting at the national level, while more recently, Hobolt and De Vries (2016) confirmed its importance in national as well as in EP elections, using observational data collected after the European Parliament elections in 2009. Further, there is research on the determinants of EU issue voting, and in particular the media environment. The degree to which EU-related information is available is also important to determine EU issue voting (De Vries et al., 2011) and dependent on the degree of EU politicisation (Hobolt & Spoon, 2012). Previous studies have focused on the characteristics of the political competition on EU issues, namely when the EU is salient and the parties have different positions on the process of European integration (De Vries, 2007).

There is an emerging debate on the degree to which European issue voting relates to other factors of individual voting behaviour. Indeed, most of the research concerning the importance of European integration for national politics have focused on the party system and its changes (Hooghe et al., 2002; Hutter & Kriesi, 2019; Marks & Steenbergen, 2002). Hooghe et al. (2002) were path-breaking in explaining the way Europe relates to the left–right dimension, through the introduction of the concept of GAL-TAN. GAL parties are in favour of European integration, have a cosmopolitan outlook, and also share pro-Green policies. The TAN category aggregates those parties who are Eurosceptic, have a nationalist outlook, and share traditional values. TAN parties tend to be radical right parties, and this pole of the GAL-TAN axis was responsible for the politicisation of EU integration. These authors posited that the GAL-TAN axis was orthogonal to the left–right axis, and that the party systems tended to form a curvilinear pattern, with parties on the extremes of the left and of the right tending to be more Eurosceptic, and those on the centre tending to be more pro-EU.

Other authors which have focused on party systems and voting choices, place parties' positioning on the EU, as well as citizens' attitudes, as part of a "globalization cleavage" (Kriesi et al., 2008), which distinguishes between citizens who tend to be pro-EU, pro-migration, and cosmopolitan, vs. those who are Eurosceptic, anti-immigration, and nationalist (Bornschier, 2010; Dalton, 2018; Norris & Inglehart, 2019). In certain elections, notably in France, this cleavage has gained prominence in relation to the left–right dimension in structuring vote choices (Lachat & Michel, 2020; Schön-Quinlivan, 2017). The eurozone crisis also leads to studies which analyse the importance of the EU issue for voting in Southern Europe. Extant research shows that challenger parties continue to be the ones which politicise the EU issue to a greater extent and, therefore, the EU issue may still cross-cut the left–right axis in these countries (Santana & Rama, 2018). Others state that there has not been an increase in the importance of EU for voting, since austerity has served to reinforce the left–right dimension, rather than replace it (Hooghe & Marks, 2018; Otjes & Katsanidou, 2017).

In this book, we show that European citizens do take into account their attitudes towards the EU when making electoral choices. Through an experimental analysis (Chapter 5) fielded in our six countries, we are able to state with confidence that the proximity between parties' and voters' positions on the EU is the cause, rather than the consequence, of respondents' electoral choice. Further, we show that EU issue voting is asymmetric, with parties that are more Eurosceptic than the respondent being punished on average more than parties that are more pro-EU.

In addition, through observational data collected in Greece, Ireland, Portugal, and Spain, we confirm that EU issue voting cross-cuts the left– right issue, and in every country analysed (with the exception of Ireland), left–right proximity has a higher impact on likelihood to vote for a party than proximity on EU issues. Even when we distinguish between mainstream and challenger parties, left–right is still more important than EU issue proximity for the vote. In our country case studies, where EU issue voting is assessed in comparison to other political issues in each of the six countries, we find that EU issue voting tends to matter more for right wing challenger parties, but in certain countries, namely Portugal and Greece, it works principally on the left, and also serves to distinguish between the two principal mainstream parties.

Concerning the way media and parliamentary debates impact the importance of the proximity between parties' and voters' positions on the EU for the vote, a distinction was made between the EU salience and tone. As regards the media context, in the experimental study we find there is a connection between the availability of EU information and the strength of EU issues voting, while the tone of EU news appears less relevant. Turning to the impact of parliamentary context, we found that, for parties with a more negative tone—or Eurosceptic discourse—EU proximity matters more in determining the voters' choice than for parties with a positive or neutral EU tone in parliament.

### Data Collection

Our substantial data collection underpins and strengthens our research goals and scope. The final media dataset includes 165,341 print newspaper articles, of two mainstream newspapers per country, of all relevant articles published one month before each national legislative election between 2002 and 2017. Each of these articles were also manually coded to consider the EU topics discussed. In addition, online media in each of the mainstream newspapers was scraped one month before each 2019–2021 webpanel (115,876 articles).

As regards parliamentary debates, the dataset includes all plenary debates in the six countries (2002–2019) (Kartalis & Lobo, 2021). All speeches "about the EU" in the year preceding the legislative elections were identified in automated fashion. In addition, a representative sample of these have been manually coded on EU content. A total of 10,514 speeches were randomly selected from all speeches delivered in parliament in the year preceding each of the 26 national legislative elections that took place in our countries of choice.

Further to the parliamentary and media data collection, 22 representative online surveys were carried out in the six countries concerned. We ran one simultaneous two-wave panel in January and June 2019, which functions as an EP election panel, and where the conjoint experiment was embedded. Next, we ran panel surveys when national legislative elections were held (Greece, July 2019; Portugal, October 2019; Spain, April 2019, Ireland, January 2020, Germany, September 2021).1 This array of different types of data strengthens our research findings.

<sup>1</sup> Belgium held its legislative election simultaneously with the 2019 EP election, and therefore no separate survey was needed.

We now turn to outlining in detail the book chapters, which are organised in two parts. In the first part, the importance of the EU for mainstream media and parliamentary debates in Belgium, Germany, Greece, Ireland, Portugal, and Spain is established. Thus, the first part includes three chapters devoted to setting up the longitudinal trends of EU politicisation in each country. In the second part, there are eight chapters, delving on the importance of the EU for electoral behaviour in 2019/2020 national elections in these countries, both from a comparative and single case-study perspective.

### The Book Chapters

The first three chapters chart the level of EU politicisation in the media and parliamentary arenas, from a longitudinal perspective and across the six case studies. Besides revealing temporal and spatial variance in how the EU penetrated the national arenas, they investigate the factors that account for such variation.

In Chapter 2, Tiago Silva and Yani Kartalis measure EU politicisation in mainstream media, between 2002 and 2017, in Belgium, Germany, Greece, Ireland, Portugal, and Spain. The focus is on the distinction between news and opinion articles, as well as on the dimensions of EU in the media. The chapter shows that, after the crisis, the EU media politicisation diverged more in tone than salience, even if these two indicators are quite correlated. When news and opinion articles are compared, it is clear that the latter diverge more across newspapers in the same country. When considering the EU dimensions portrayed in the media, while all countries have increasingly focused on policies, Spanish and Portuguese media focus on domesticated policies, whereas the rest of the countries tend to debate European policies to a larger degree.

In Chapter 3, by Yani Kartalis and Tiago Silva, automated measures of salience and tone are employed to examine EU politicisation in parliamentary debates. The Chapter also analyses the four distinct dimensions of EU, and the extent to which these are discussed in each Parliament by different categories of parties. The findings show that the EU has not been very politicised in national parliaments, and this did not change considerably with the Eurozone crisis. As regards the determinants of parliamentary politicisation, the EU position of parties was the strongest predictor of EU politicisation, being statistically significant for both EU salience and contestation. Also, in the parliamentary speeches of the six countries concerned, the EU was likelier to be contested by the parties on the left and the parties not in the government.

Chapter 4 complements the previous analysis of EU politicisation through a comparison of trends in media and parliamentary debates, between 2002 and 2017. While acknowledging the inherent differences between the two arenas, Nelson Santos and Susana Rogeiro Nina compare salience and tone within EU dimensions, as well as the policies which are discussed. Politicisation is present in both arenas, albeit negative tone is much more prevalent in the media than in parliaments. In both arenas, (economic) policy-related issues are predominant, which may foster EU accountability. A more fine-grained dimensional analysis shows differences between countries, suggesting EU differentiation, with Germany debating more EU policies than Southern European countries.

The second section of the book is composed of two comparative chapters and six single case studies. It establishes EU issue voting, and develops fully the analysis of the contexts which may help or hinder the expression of EU preferences by citizens, thus holding national politicians accountable.

In Chapter 5, Roberto Pannico and Marina Costa Lobo approach the issue of the importance of the EU for voting in an experimental setting. This model enables a truly exogenous estimation of the importance of EU attitudes for voting in the six countries. The results confirm that EU preferences are the cause not the consequence of electoral choices. It also shows that EU issue voting is asymmetric, with parties that are more Eurosceptic than respondents being on average more punished than parties which are more pro-EU than the respondent. However, this asymmetry only seems to hold in contexts where EU information is more available and positive.

In the following chapter, Lea Heyne, Marina Costa Lobo, and Roberto Pannico make use of post-election surveys carried out in 2019–2020 in Greece, Ireland, Portugal, and Spain, to consider the importance of the EU vis-à-vis left–right issue proximity on the likelihood to vote for different parties. Our results show that, in all countries surveyed, EU issue voting does cross-cut the left–right dimension. Yet, left–right proximity has a higher impact on likelihood to vote for a party than EU issue proximity, in all countries, both for mainstream and challenger parties. The interaction of parliamentary tone employed by parties in EU debates and EU proximity shows that, for parties which have a more negative tone, EU proximity determines vote choice to a larger extent than for other parties.

Having established EU issue voting and its determinants, in a comparative perspective, using experimental and observational data, the rest of the book is composed of six single case studies. In each of the country chapters, media, parliamentary, and survey data are combined to explain EU politicisation and EU issue voting in comparison to competing relevant political issues.

In Chapter 7, Dieter Stiers examines the case of Belgium, which is, as he demonstrates, a singular case on different levels. As the author explains, the legislative and European elections in Belgium are held simultaneously, and voting is mandatory. Combined, these factors make it unlikely to find EU issue voting in Belgium since voters can express their European preferences on the same day using the EP ballot, and a lot of voters show very little interest in politics. Further, the data on media shows that, overall, the European issue is not very salient in Belgium, since most parties share a positive consensus towards the EU. Yet, EU issue voting is detected at the national level. The author distinguishes between Wallonia and Flanders, as well as between electors with different levels of political sophistication to nuance the results.

In Chapter 8, Marc Debus and Rosa Navarrete explain the way in which key economic issues, as well as immigration, compare to EU issue voting in Germany. In this case, there is a sort of paradox: while Germany is decisively important for decision-making processes in the EU, the German media and parliamentary debates tend not to politicise the EU, with the exception of the (AfD—*Alternative für Deutschland*). The authors find that parties behave strategically regarding the EU in the plenary debates. There is also evidence of EU issue voting in the last federal elections, and the authors further show how that it interacts with individuals' perceptions on immigration policy.

In Chapter 9, Roula Nezi examines the case of Greece which is the country in our sample where EU salience and polarisation increased most dramatically, both in the media and in parliamentary debates, following the onset of the Eurozone crisis and the successive bailouts which the country endured. The chapter analyses the importance of EU issue voting in two models, one where that variable is included alongside ideological self-placement and economic perceptions, and another contrasting the EU issue with two other relevant political issues. Those relevant issues were the Prespa agreement which had been signed by the Tsipras government to resolve the "Macedonia" question and proved highly controversial, as well as the asylum law, with both issues serving to differentiate between the two main parties—Syriza and ND (*Nea Demokratia*). Results show that the EU issue, as well as perceptions of the economy, are the two most important issues differentiating the vote between Syriza and all other parties in the party system, with those holding Eurosceptic perceptions voting for Syriza, and pro-EU attitudes reinforcing the vote for the centre-right party of ND.

The Irish case is analysed by Lea Heyne in Chapter 10. She begins by highlighting the importance of the 2020 Dáil elections, where, for the first time ever, Sinn Féin was the most voted party. The analysis of EU politicisation shows that it has increased in both media and parliamentary debates in the last decades, and the media has become more negative since the Eurozone crisis and Brexit. In parliamentary debates no such negativity is found, with parties adopting a more positive tone. The multivariate analysis of electoral behaviour shows that there is no strong impact of EU issue voting, with the exception of voting for Sinn Fein.

In the Portuguese case, covered in Chapter 11, Marina Costa Lobo compares EU issue voting to socio-economic attitudes during the 2019 legislative elections. As in the Irish case, there are increasing trends of EU issue politicisation from 2002 to 2019, both in the media and parliamentary debates, particularly noticeable after the onset of the Eurozone crisis in 2009. Yet, in a post-bailout stage, and especially since the left coalition government took office in 2015, a degree of depoliticisation of the EU issue is detected in parliament. Concerning EU issue voting the data show that it continues to be relevant among Communist voters, but also for the mainstream centre-right PSD (*Partido Social Democrata*), signalling that in Portugal it matters not only for voters of challenger but also mainstream parties.

In Chapter 12, Hugo MarcosMarne examines the politicisation of the EU in media and by different political parties in Spain as well as the dynamics of EU issue voting in the May 2019 legislative elections. Results show that there was an increase in the salience of EU issues after 2008 that subsides after 2012, and an unclear pattern concerning tone, even if, overall, it seems that salience has not been accompanied by more contestation of the EU. Concerning EU issue voting, it predicts only votes for the radical-right VOX both in general and within-ideological-blocks comparisons.

In the Conclusion, we examine the evidence provided in favour of holding Europe accountable at the national level. The comparison of media and parliamentary debates trends longitudinally make clear that politicisation is higher in the former than the latter. In the media, there is overall growth in salience and tone turning negative. In Parliaments, parties tend to behave strategically, depending mainly on their EU stance. As regards, the consequences of this politicisation for voting behaviour, it is demonstrated through the experimental analysis as well as observationally. Furthermore, both the EU media as well as parliamentary politicisation appear to be associated with the strength of EU voting. Indeed, both voters of mainstream and challenger parties use the vote to express EU preferences in national legislative elections.

Our analysis has consequences for the debates about EU legitimacy today. Most of the efforts for democratising Europe involve supranational reforms, at the level of the European Parliament, or citizens' initiatives at the EU level. Yet, the important vertical accountability which is occurring in national legislative elections regarding the EU is being largely ignored. The national channel of EU accountability, and namely the domestic institutions which contribute to it, must become a central part of the debate on EU legitimacy, for this issue to be properly addressed.

#### References


*of Political Research, 8*(1), 3–44. https://doi.org/10.1111/j.1475-6765. 1980.tb00737.x

Risse, T. (Ed.). (2015). *European public spheres*. Cambridge University Press.


**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# News vs. Opinion Articles on the European Union: The Politicisation of the EU in the Mainstream Press

*Tiago Silva and Yani Kartalis* 

# Introduction

The traditional media play an undoubtedly important role in the process of politicisation of the European Union. They remain, first and foremost, societies' most important source of political information and the main link between institutional and public arenas. This role is particularly important

T. Silva (B)

Y. Kartalis Center for International Studies, Instituto Universitário de Lisboa (ISCTE-IUL), Lisbon, Portugal e-mail: kartalisie@sodas.ku.dk

Copenhagen Center for Social Data Science, University of Copenhagen, Copenhagen, Denmark

© The Author(s) 2023 M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_2

25

Institute of Social Sciences, University of Lisbon, Lisbon, Portugal e-mail: tiago.silva@ics.ulisboa.pt

in the context of the European integration project since the main political institutions and actors of the EU, compared to the national ones, are more distant from the citizens. Therefore, a thorough study of EU politicisation in the media, combined with the analysis of the parliamentary debates conducted in the third chapter of this book, are crucial steps to understand how the EU politicisation shapes the EU issue voting in national elections, which is the core research question of this book.

Studies have shown, by analysing political actors' statements in media, that the EU politicisation is a punctuated phenomenon, varying across countries and regions. Moreover, when it comes to the overall media coverage, Silva et al. (2022) demonstrated that the Eurozone crisis not only increased the levels of EU politicisation in the bailout countries but it also led to a higher awareness, or salience, of the different institutions of the EU. These existing studies on the magnitude of politicisation, however, do not consider important differences within the media and in the dimensions of the EU that can be the object of contestation. By looking at the media coverage of the six MAPLE countries,1 that experienced the Eurozone crisis in different ways and to different degrees, this chapter investigates two important and untapped dimensions of the news coverage of the European integration topic that can shape voting behaviour.

The first dimension concerns the role of mainstream media as agents of EU politicisation. In this chapter, we explore whether the ideological leaning of mainstream press (left vs. right) or the type of newspaper items (news vs. opinion articles) have an impact on the salience and tone of the news coverage of the EU. We are particularly interested in investigating whether the Eurozone crisis was associated with higher convergence, or divergence, in the news coverage of the EU, within each country. The second dimension concerns the longitudinal changes in the way that the EU is reported, or debated, in the media. This chapter goes beyond the magnitude of EU politicisation to also investigate which EU dimensions (*membership*, *constitutional structure*, *EU policies* or *domesticated policies*) were more salient in the media.

The politicisation of the EU is analysed, in this chapter, using a unique dataset of 165,341 newspaper articles, from 12 quality newspapers, collected for 29 legislative elections that occurred between 2002

<sup>1</sup> The six countries are Belgium, Germany, Greece, Ireland, Portugal and Spain.

and 2017. We analyse the data using an innovative approach, combining manual and automated content analysis, which takes advantage of the strengths of both methods. On the one hand, the automated approach maximises the reliability and comparability of the results across time, newspapers, type of article and countries. On the other hand, the manual coding allows us to offer a more detailed, and differentiated, analysis of EU politicisation that is missing in the literature, in a longitudinal comparative perspective.

Our results indeed confirm that even though the crisis comparably increased, in terms of magnitude, the politicisation of the EU in the bailout countries, that politicisation differed significantly in their nature. Moreover, they found that the similarities between left and right newspapers, in their coverage of the EU, decrease significantly when we only look at opinion articles. This important finding tells us that the role of the media, in shaping EU voting, is not limited to their strict coverage of the electoral campaigns.

The remaining of this chapter is structured in four sections. The next section offers a brief review of the literature focusing on the politicisation of the EU. It discusses how the magnitude of EU politicisation has been assessed in the media and what are the important aspects of that concept that remain unexplored in a comparative perspective. The section after that explains the data and methods used to analyse this phenomenon. The third section presents and discusses the results and main findings of our analysis. Finally, in the last section, we summarise our main findings and discuss their contribution to this book's objective and research question.

# Literature Review

The concept of politicisation, or the politicisation hypothesis, originally referred to the anticipated expansion of actors interested in the regional integration process. According to Schmitter (1969), the European integration should become, with the gradual deepening of its process, a far more salient and contested dimension within its member-states. More recently, this concept has been divided into three core dimensions regarding EU debates: *salience, polarization, and actors' expansion* (de Wilde et al., 2016, p. 4). The politicisation of the EU can therefore be understood as 'an increase in polarisation of opinions, interests or values and the extent to which they are publicly advanced towards the process of policy formulation within the EU' (de Wilde, 2011, p. 560).

The importance and interest in this concept increased considerably after the claim, by Hooghe and Marks (2009), that changes in the level of European integration, resulting from the Maastricht treaty, finally shifted public opinion, from a 'permissive consensus', to a 'constraining dissensus' towards the EU. This premise suggested that the creation of the 'European Union' increased the salience and polarisation of the regional integration topic, making it, for the first time, an important and divisive dimension in national politics. Since then, several studies have tested that claim and measured, over time, the magnitude of EU politicisation in different EU countries (Grande & Hutter, 2016; Höeglinger, 2016; Hutter & Grande, 2014; Hutter & Kriesi, 2019).

Despite some inconclusive, or even contradictory, results (e.g., Grande & Hutter, 2016, p. 87; Green-Pedersen, 2012; Höeglinger, 2016, p. 146; Statham & Trenz, 2013), the most exhaustive comparative studies suggest that, rather than being a linear phenomenon, or similarly shared across all EU member-states, EU politicisation has been a punctuated phenomenon with clearly identifiable drivers (e.g., EU treaties or crisis) and circumscribed to certain geographical regions of Europe (Hutter & Kriesi, 2019). In this regard, the 2009 Eurozone/financial crisis was a key event, bolstering considerably the salience and contestation of the EU in south European countries, particularly in Portugal and Greece (Hutter & Kriesi, 2019). However, most of these studies on EU politicisation rely on the media only as a source of data for measuring the direction and amount of political actors' statements concerning the EU.

There are very few comparative assessments of EU politicisation, across countries and across time, that investigate the media as a primary source, or arena, of politicisation. Nevertheless, we do know that the traditional media have the capacity, or mechanisms, to influence and shape citizens' perceptions and attitudes towards politics (Weaver, 2007). Moreover, certain media characteristics seem to influence the visibility of certain topics. In terms of both volume and content, aspects such as media ownership, media type and newspaper style can determine how the EU is portrayed in the news coverage of certain political events (Nord & Strömbäck, 2006; Peter & de Vreese, 2004; Pfetsch, 1996). More concretely, the EU, often perceived as a more complex/technical topic (Kevin, 2003; Statham, 2007), has higher visibility in the mainstream press and public broadcasting news, in comparison to commercial television and tabloid newspapers (de Vreese et al., 2006).

Despite those observed differences, the coverage of EU in the media, particularly outside major EU events, has been scarce in volume (Peter & de Vreese, 2004; Pfetsch, 1996) and predominantly neutral in tone (de Vreese, 2003; Norris, 2000). Differently than most of those studies, that look at the news coverage of the EU during major EU events, this chapter focuses, instead, on national election periods, which are more 'demanding' settings for EU to 'matter'. However, in line with the goal of the book, the focus on legislative elections allows us to access the magnitude of politicisation in periods when it is likelier to impact national politics or being advanced towards policy formulation.

Overall, the existing assessments of EU politicisation on traditional media, either as a data source or as an actor, have predominantly focused on two dimensions: salience and contestation. The exploration of a multifaceted contestation of the EU (Braun et al., 2019) has been largely absent from longitudinal and comparative assessments of EU politicisation. However, according to the literature, not only there are important distinctions in how European integration has been, or could be, contested, but those differences can also have considerably different implications.

When it comes to the differentiation of EU politicisation, an extremely influential contribution was made by Peter Mair (2004, 2007), which identified two distinct dimensions of conflict about the EU. The first one, a *Europeanisation* dimension, dealt with the creation, consolidation and the geographical reach of the EU institutions. The other one, to some extent connected with the first, related to the conflict concerning the *penetration*, or *reach*, of EU legislation into domestic spheres. Building on this differentiation, de Wilde (2011) identified three groups, or dimensions, of EU politicisation: institutions, decision-making processes and the politicisation of issues. Finally, Hurrelmann et al. (2015) also contribute to this discussion by distinguishing policies emanated from EU and domesticated policies, which are the policies emanated from national institutions, as a consequence of EU membership. This distinction, between European and domesticated policies, is particularly important in the context of the Eurozone crisis, where domesticated policies likely became more salient.

Overall, despite its importance and theoretical development, this differentiation of politicisation remains largely untapped in comparative, longitudinal, studies of EU politicisation. Moreover, the different types of EU politicisation seem to entail different implications. More concretely, there is the perception that conflict directed to the 'constitutive' issues (i.e., Europeanisation, institutions or membership dimension), compared to the politicisation of European policies, have a more negative impact on the European public sphere (Risse, 2010). It is fair to assume that the differentiation in the debate/contestation of EU may lead to either a deepening of Euroscepticism, or further Europeanisation.

A differentiation of EU politicisation has been observed in the news coverage of the EU. Dutceac Segesten and Bossetta (2019) found that 'euroscepticism' or 'eurosceptic' media articles are predominantly framed in a European context, while in the UK, differently, they are primarily framed in a domestic context. More recently, Silva et al. (2022) show that the salience of different EU institutions, in the media, has shifted overtime. This chapter contributes to this topic by exploring, in addition to the salience and contestation of the EU in the media, what type of EU dimensions were discussed. We focus our analysis on the period before, during and after the Eurozone crisis. While we know, from the literature, that the crisis bolstered the politicisation of the EU, it remains unknown whether this crisis also changed the salience of the different dimensions of EU politicisation.

### Research Questions, Data and Methods

This chapter offers a detailed and original analysis of EU politicisation in the mainstream media of the six MAPLE countries, from 2002 until 2017. This assessment is an essential step to understand the impact of EU politicisation on EU issue voting, which is the primary goal of this book. Moreover, this chapter answers in itself three main research questions: Can we observe differences between mainstream (left and right) newspapers when it comes to EU politicisation? Is there a difference between news articles and opinion articles? What dimensions of EU politicisation have been more discussed in the press? By answering those questions, this chapter makes two important contributions to the study of EU politicisation.

The first contribution concerns the role of media, as an actor itself, in the process of politicisation. Different from the existing studies, which mainly use mainstream media as a source of political actors' statements, this chapter examines how mainstream media diverge in their politicisation of the EU. The second contribution is about the media content. While most of the existing longitudinal and comparative assessments of the magnitude of EU politicisation have focused on the overall levels of salience and conflict, we know that the multidimensionality of the news coverage of the EU has been increasing over time (Silva et al., 2022). Building on this perspective, we analyse EU politicisation in a more nuanced way and investigate which of its dimensions is more debated in the media. This differentiation might have, as we discussed, different implications for the European integration and EU issue voting.

Based on the existing literature, considering the extensive period that we analyse, we can bring forward two expectations regarding the salience of the different EU dimensions. The first one is that, due to the austerity measures imposed in some countries, the debates about domesticated policies became considerably more salient, after the Eurozone crisis, in the group of bailout countries. The second expectation is that membership debates became more prominent, after that crisis, in all countries. The reasoning for that is that the Eurozone crisis also gave an opportunity for Eurosceptic actors to contest the institutional foundations of the EU polity. Alternatively, a competing hypothesis is that the crisis increased the Europeanisation and, consequently, also led to more discussions regarding its dimension of penetration (i.e., policies related to the EU).

In order to answer those four questions, this chapter uses a unique dataset of newspaper articles collected from six Eurozone countries: Belgium, Germany, Greece, Ireland, Portugal and Spain. The articles were collected from 12 newspapers, two per country, during the 30 days that preceded all legislative elections that occurred between 2002 and 2017. The final dataset includes, from 29 elections, a total 165,341 newspaper items (including journalistic and opinion articles). We analyse the politicisation during electoral campaigns which are key periods when political contestation is at its most intense form and likelier to lead to policy outcomes.

We limit our analysis to the mainstream/broadsheet press, which is usually a reference to other media and a great proxy for a country's media coverage of political events (Boomgaarden et al., 2010). Moreover, broadsheet newspapers, compared to tabloids, are much likelier to discuss the EU topic (de Vreese et al., 2006). The selection of the two newspapers, per country, followed the criteria of both being quality/broadsheet papers, having comparatively high number of readers and diverging in their ideological leaning.

An expert survey on European media systems (Popescu et al., 2011) was used to select, for each one of the six countries,2 one newspaper from the left and one from the right. From the 12 selected newspapers (Appendix: Table 2.2), we extracted, from their printed versions (except when content was available in a machine-readable format3 ), all news and opinion articles published during the 30 days4 before a legislative election. The articles published in secondary sections (e.g., sports, culture or classified advertisement) were not included in the analysis. For newspapers only available in printed version, Optical Character Recognition (OCR) software was used to manually extract the articles' titles and main text. In those cases, additional steps were taken to assure the quality of the final dataset.5 During the extraction process, relevant information about the article such as section name and page number was also collected.

Concerning the methods used, our analysis was made in two steps that combined automated and manual content-analysis techniques. While, in the first stage, automated methods were used to access the salience and contestation of the EU dimension, in the second stage, the main EU dimension present in the articles was identified by a team of trained coders. This multi-method approach allowed us to not only maximise the comparability of our results but also move beyond that and offer a more detailed, and diversified, picture of how the EU dimension has been discussed in media.

For the first stage of the analysis, we adopt a multidimensional concept of politicisation focused on two main attributes of the news coverage of a topic: its *salience* and the levels of *contestation* surrounding it. Here, as we discussed, we depart from existing studies on the politicisation of the

2 Belgium was an exception since it presents a more complex/unique media landscape. In the case of this country, instead, one French and a one Dutch-speaking mainstream newspaper were selected.

3 The data for Germany was received directly from the two newspapers in a data frame format.

4 This period was chosen since it usually corresponds to the period established for the official campaign in the EU countries (Swanson & Mancini, 1996, pp. 259–260).

5 Due to the use of OCR, some misspelling errors were occasionally produced. These errors were investigated by counting the percentage of unrecognized words in the articles, using a well-document/popular and free open-source spell-checking package/library named hunspell for the automatic spellchecking. The results confirmed an insignificant proportion of unrecognized words.

EU (e.g., Grande & Hutter, 2016; Hutter & Kriesi, 2019; Hutter et al., 2016), that focus on the statements of political actors covered by the media. In this sense, we offer an alternative, and more encompassing, way of assessing EU politicisation that takes into account a more diversified set of actors represented in the mass media. We therefore follow in the tradition of Pfetsch et al. (2008) that perceives the media itself as an important political actor, and a non-neutral debate setting (Wilde & Lord, 2016), that is able to (re)shape the debates surrounding the European integration process. Consequently, our analysis pays special attention to the contrasts between newspapers and types of newspaper items.

Similar to Schmidtke (2016), we operationalise EU salience as the percentage of articles, out of all articles in a particular election/newspaper, that is about the EU. The articles about the EU were identified using an extensive list of EU-related terms (Appendix: Table 2.3). This list, translated into seven languages, was adapted from the codebook of Maier et al. (2014). When a newspaper article included one EU-related term in the title or more than one in the remaining text, the article was coded as being about the EU. Overall, 13.8% (22,769) of articles in our dataset mentioned the EU and 7.7% (12,716) were coded as being about the EU. The minimum value of EU salience was 2.8% and the maximum was 19.6%. The average value of EU salience in our dataset was 7.96%.

The dimension of contestation was examined by looking at the proportion of articles, in a year/newspaper, with a negative tone towards the EU. The articles' tone was measured using sentiment analysis. The sentiment analysis is a well-developed automated text-analysis technique that extracts the valence/tone of a text by means of natural language processing (Pang & Lee, 2008). It has become a widely used method in political communication studies and is used to examine a diversified array of media effects/mechanisms such as agenda-setting (Ceron et al., 2016), media framing (Burscher et al., 2014), conflict (Proksch et al., 2019) and, as in the case of this chapter, also the media tone (Soroka et al., 2015; Young & Soroka, 2012).

Since our dataset includes articles written in seven different languages, we translated all articles into a single 'pivot' language (Lucas et al., 2015). This approach, of applying existing methods designed for English in the translated corpus, rather than translating the lexicons into each language of the articles, has been shown to yield the best results (Araújo et al., 2016). Furthermore, this allowed us to use more sophisticated/valid sentiment measures that use valence shifters,6 which also gives us higher confidence for our measurements of tone.

The tone of each article mentioning the EU was therefore calculated in the following way. We first translated to English all sentences with an EU-related term.7 The translations were done using the R package 'googleLanguageR', which accesses the Google's translation API service. Using the sentiment algorithm provided in 'sentimentr'8 (Rinker, 2019), we calculated two sentiment variables for each article. The first one was the mean value of the sentiment scores for each sentence mentioning the EU in the body of the articles. The second variable corresponded to the sentiment score of the titles that mentioned the EU. Finally, the mean of those two scores was used as the overall tone towards the EU of each article.

Based on the signal of the final sentiment score, each article was then coded as being either positive or negative towards the EU. We did so because the direction of the sentiment scores (i.e., being negative/positive towards the EU) is a far more objective and meaningful factor than the variation of its degree (i.e., how negative/positive is the article towards the EU). Ultimately, our contestation measure is the ratio between negative and positive articles towards the EU.

Our contestation measure relies on the assumption that contestation towards the EU increases when the proportion of negative articles towards the EU, in a particular newspaper and year, also increases. This assumption is valid if the proportion/percentage of positive articles is usually higher, and the percentage of negative articles does not exceed considerably the 50 per cent threshold.9 This was indeed confirmed in the data, where the percentage of negative articles has a mean of 41%, a minimum of 25.3% and a maximum value of 50.3%. Furthermore,

6 Valence shifters are words in the text that alter or intensify the valence of sentences, largely improving the measurement of the tone expressed in the text.

7 We measure the sentiment only in the articles' sentences about the EU. This allows us to capture the tone towards the EU in each article, rather than the overall tone of articles where EU is mentioned.

8 This algorithm takes into consideration valence shifters, or augmented sentiment dictionary lookup. This guarantees that expressions like 'not good' do not receive a positive score, or that 'extremely bad' has a more negative score than 'a little bad'.

9 The idea is that the contestation of a certain topic is higher when the proportions of negative and positive articles are identical.

our contestation measure is more accurate/meaningful if there is a low proportion of 'neutral' EU articles. This fact was also confirmed with the data, where only 6.6% of the EU articles had neutral valence.

The second step of our analysis concerned the different EU dimensions salient in the media. Those dimensions were manually coded by a team of trained coders that identified, for each article about the EU, if the EU article was mainly about (1) *membership*, (2) *constitutional structure*, (3) *EU policies* or (4) *domesticated policies*. 10 The first category concerns debates surrounding EU *membership* (e.g., statements questioning whether a country should be a member of the EU, its costs and benefits, debates surrounding the geographical reach of the EU). The second category, *constitutional structure*, concerns the politicisation of the decision-making process and debates surrounding the procedures and responsibilities of the different institutions. The *EU policies* category deals with policies emanated from, or in the agenda of, the EU's legislative, executive and judiciary institutions. Finally, *domesticated policies* refer to the debate of policies,11 originated from national institutions, that result from the EU membership (e.g., cuts mandated by budgetary requirements of the Eurozone). This distinction between EU and domesticated policies, which is the main departure from the typology of de Wilde (2011), is particularly important in the context of the Eurozone crisis and, consequently, to the analysis made in this chapter.

Finally, to increase the comparability of the results across newspapers and over time, the coders involved in the coding process analysed a randomised set of articles from all years and different newspapers. This process, made possible by having the data in a machine-readable format, further allows us to avoid a coder bias in our results and maximise the comparability within the countries.

# Results

The first aspect analysed in this chapter is the level of EU salience and contestation in the news media coverage before legislative elections. Similar to previous assessments of the magnitude of EU politicisation,

<sup>10</sup> We adopt the typology in Hurrelmann et al. (2015, p. 45) that further differentiates debates related to EU policies.

<sup>11</sup> Also the absence/impossibility of certain policies/decisions.

looking at the overall trends of the two dimensions over time, our results also confirm that the Eurozone crisis of 2009 bolstered the salience and contestation of the EU (Fig. 2.1). This increase is evident in the group of bailout countries, particularly in Greece and Portugal, cases where the salience of the EU doubled in the 'critical' elections.12 In the case of the creditor countries (Belgium and Germany), the increase of EU salience in the media seems to be gradual, less steep and not necessarily related to that crisis.

When it comes to contestation, it is also evident, in the bailout group,13 that the proportion of articles with a negative tone towards the EU saw an increase after 2009. In the case of Greece, the EU was contested the most in the 201214 election, becoming gradually less contested in the two-following elections. Differently, in the case of Ireland and Portugal, the proportion of negative EU articles has been increasing since 2009, having their highest values in the last elections analysed. The case of Spain is somehow different. While the salience and contestation increased in the first election after the crisis (2011), these two elements decreased considerably in the following election, in 2015, to increase again in the election held in the following year. These results suggest that the EU was, somewhat, depoliticised in the 2015 Spanish election.

Regarding the levels of EU contestation, we can identify some differences between the two creditor countries. In the case of Germany, the proportion of negative EU articles has been relatively stable and comparatively low in all elections. Differently, the Belgium case shows in the 2010 election, right after the eurozone crisis, an increase in terms of both EU salience and contestation. These increases observed for Belgium, however, especially in salience, are not as steep as the ones observed in Greece, Portugal and Spain.

Regarding differences between media, our results suggest a strong parallelism between the left and the right-wing mainstream newspapers in their politicisation of the EU. This is evident in Fig. 2.1, with the newspapers exhibiting similar diachronic patterns in all six countries. This

12 We consider critical elections the ones that occurred immediately after the bailout agreements (2011 for Portugal and 2012 for Greece).

13 I.e., Greece, Ireland, Portugal and Spain.

14 We only consider in the analysis the second election of 2012, in June.

**Fig. 2.1** Salience (lines) and Contestation towards the EU in the news coverage of legislative elections in two mainstream newspapers

idea is further confirmed, as we can see in Table 2.1, with a strong positive correlation between the pairs of newspapers, both in terms of EU salience (0.739) and in terms of contestation (0.629). The ideological leaning does not seem, even after the eurozone crisis, an important factor for the changes in EU politicisation. Furthermore, the data also shows a moderate and positive correlation between EU salience and EU contestation. These results go in line with the idea that news media attribute *news-value* to conflict, suggesting that the observed increase in EU salience resulted from its increasing contestation. Albeit not completely


**Table 2.1** Pearson's correlation results between different groups/variables for different types of articles

straightforward, our results, particularly in the cases of Greece, Portugal and Spain, indeed suggest that an increase in contestation *preceded* the increase in salience.

The second question investigated concerns the difference between types of articles (news articles vs. opinion articles) in the politicisation of the EU. The results, in Fig. 2.2, show that, when it comes to salience, the EU has been, in general, more salient in the opinion articles. Moreover, in most countries, the Eurozone crisis seems to have increased those differences. In the case of Ireland, the clearest example, the Eurozone crisis seemingly led to the politicisation of the EU exclusively in the opinion items of the mainstream press. A clear difference of EU salience between types of articles is also observed in Greece, Portugal and Belgium. Contrarily, in the case of Spain, when it comes to the salience of the EU, we cannot observe any noticeable difference between the two types of newspaper items.

When it comes to EU contestation, we also observe some differences and patterns, between the two types of articles. In general, with the exception of Belgium, in the majority of the elections analysed, the proportion of negative articles has been higher in the opinion items. Moreover, in the bailout countries, the results show that the contestation towards the EU after the crisis was also more prominent in the opinion articles, in comparison with the traditional news pieces, with the only exception to this pattern being, again, the Spanish election of 2015.

Overall, the results show that, in general, the salience and contestation towards the EU have been higher in opinion articles, in comparison to journalistic news pieces. Moreover, the correlation between newspapers, when it comes to both EU salience and contestation, becomes the weakest

**Fig. 2.2** Salience (lines) and Contestation towards the EU in the news coverage of legislative elections, a comparison between news and opinion articles

when we only consider opinion pieces (Table 2.1). These results suggest that the journalistic coverage has been truthful to the salience of the EU during legislative campaigns, with the mainstream media being, therefore, good sources to assess the magnitude of politicisation during electoral campaigns. However, if we consider the media as a political actor capable of (re)shaping politicisation, one cannot ignore the opinion pieces. We find that not only opinion articles are likelier to politicise the EU, but they also increase the differences in mainstream media, particularly in terms of EU tone.

Indeed, as Fig. 2.3 shows us, the divergence between mainstream press increases considerably when we only look at opinion articles. While the differences between newspapers are more prominent in the tone, we can highlight two findings concerning the dimension of salience. The first one is that, during the German elections of 2005, 2009 and 2013, the EU topic was practically absent from the opinion pieces published in the Frankfurter Allgemeine Zeitung. Interestingly, in the other two elections, the salience of the EU was identical in the opinion articles of the two German newspapers. The second interesting result is that, in the case of Belgium, Greece, Portugal and Spain, the two mainstream newspapers diverged considerably in their first election after the Eurozone crisis, a pattern that, when we considered both opinion and news pieces, was either not as evident or, in the case of Belgium, not even present. Moreover, in Greece, Portugal and Spain, it was the left-leaning newspapers that gave considerably higher salience to the EU in their opinion pieces.

When it comes to EU contestation, not only the divergence is higher between mainstream press but it is also a lot more difficult to identify consistent patterns in the data. Differently from the salience dimension, there isn't a single case where, for all elections, the EU contestation was consistently higher in one of the newspapers' opinion items. However, it is in Spain that we can see the clearest and most consistent divergence between left- and right-leaning newspapers. With the exception of the first election, where both newspapers had prevalently positive tone towards the EU, the situation shifted considerably and El Mundo, closer to the right, became considerably more negative towards the EU in its option pieces. In the case of Portugal, the Eurozone crisis seems to have also shifted the tone of the two main broadsheet newspapers, with the Diário de Notícias, closer to the right, becoming the newspaper with a higher proportion of opinion articles that were negative towards the EU.

The last question explored in this chapter concerns the different dimensions of EU politicisation that have been salient in the media (Fig. 2.4). Regarding this aspect, we can observe differences between the two groups of countries (bailout and creditor) and between pre and post bailout periods. With the exception of Spain, before the Eurozone crisis, the most salient dimension of the EU politicisation was the *EU policies*' dimension. Furthermore, the way the EU was politicised in Belgium, Germany and Ireland has been very similar, with a much higher visibility of debates on *EU policies* (on average higher than 60%) and a relatively low salience of the remaining dimensions. Differently, in Greece

**Fig. 2.3** Only for opinion articles: Salience (lines) and Contestation towards the EU in the news coverage of legislative elections in two different mainstream newspapers

and Portugal, besides EU policies, there was before the crisis, to a certain extent, also some prominence of *domesticated issues* in the news coverage of legislative elections. Finally, Spain differed considerably from the other countries in the pre-crisis period. More concretely, the *domesticated issues*  have been the most prevalent dimension of EU politicisation discussed in media. Overall, these results show that, in the pre-crisis period, despite the relatively low salience of the EU in all countries, there were considerable differences in the way that EU was politicised in media. More concretely, the *domesticated policies*, which refer to the domestic policy implications of being an EU member, were always a more salient concern in the South, particularly in Spain.

Regarding how the crisis, and subsequent bailouts, might have affected the way that EU was politicised, we can, to some extent, confirm our first expectation. More concretely, the results denote an increase in the salience of *domesticated policies* immediately after the Eurozone crisis in Ireland, Portugal and Spain. The increase is particularly evident in the case of Portugal where, in 2011, about 70% of the articles about the EU dealt with *domesticated policy* aspects. Contrarily, Greece was the only bailout

**Fig. 2.4** Salience of different EU dimensions during legislative elections' periods (year/election values are the average of the two newspapers)

country where the salience of *domesticated policies* decreased in their first critical election post-bailout, in 2012. Instead, in Greek media, the EU became more politicised in terms of *membership* (constitutive) issues. Very interesting, our analysis shows that the way the EU was politicised in Greece after the bailout was very different from what happened in Ireland, Portugal and Spain, where *membership* aspects were never at the forefront of the debate.

Our second expectation anticipated a higher salience of the *membership*  dimension after the Eurozone crisis. The data, however, did not confirm this hypothesis. Overall, the EU membership aspect has never been very salient in the media, a situation that the Eurozone crisis also did not change. Greece seems to be the exception to that trend. In this country, the proportion of debates about EU membership saw a sharp increase in the post-bailout election of 2012. Despite the increase in the magnitude of EU politicisation, with the exception of Greece, the crisis did not contribute, in the bailout countries, to a higher contestation of the institutional foundations of the EU polity. Instead, the debate remained, with some changes in 'framing' (i.e., European versus domesticated), in the realm of policies. Contrarily, we also did not find evidence that this crisis led to more Europeanisation, in the form of a higher discussion of policies. The discussion of policies, either EU or domesticated, was already a predominant dimension before the crisis.

Finally, it is also interesting to notice a similar pattern in the last three elections analysed (Spain and Ireland 2016 and Germany 2017). These three elections show, in their results, a steep increase in the salience of the *membership* dimension. The refugee's crisis might have contributed to this increased salience of EU constitutive aspects, as well as Brexit. In this sense, it would be very interesting, in the future, to see whether other countries observed similar patterns and whether this change has persisted over time. This change of focus from 'EU penetration' to 'EU institutionalisation' can have, as theory suggests, and as it is discussed in the introductory chapter of this book, important implications for both the European integration process and the accountability of national political systems. While, on the one hand, our results somehow confirmed, as Bartolini (2005, p. 349) suggested, the parties' lack of capacity to debate the constitutive issues of the newly created political community, on the other hand, that inability seems to have waned considerably in the last national elections, after the refugee crisis and Brexit.

# Conclusion

This chapter analysed the overtime politicisation of the EU in the mainstream press of six eurozone countries (Belgium, Germany, Greece, Ireland, Portugal and Spain). Using automated text analysis, we were able to clearly show that the Eurozone crisis was associated with an increase in EU salience and contestation in the news coverage of legislative elections in the group of bailout countries. These results go in line with previous studies that found that the Eurozone crisis increased the politicisation of the EU in the south of Europe (Hutter & Kriesi, 2019).

Our results also showed a positive and significant correlation between salience and contestation, which has two important implications for the future study of this topic. The first one is that, in the case of media, salience can be a satisfactory indicator of politicisation. This is relevant since measuring salience is far easier than contestation, therefore opening the door, in the future, to even more encompassing studies of the magnitude of EU politicisation, which is crucial for better understanding its national and regional implications. The second one is that the recurrent calls for a higher EU visibility in media, as a crucial step to deepen the European integration, might not be as straightforward. In line with what Boomgaarden et al. (2010, p. 518) also found, when it comes to the EU, more news is not necessarily 'good news'. The implications of increasing visibility and negativity are not yet clear. While negative evaluations could indeed reduce public support and trust, a higher salience of the EU topic can promote more engaging debates and increase the public's interest and involvement in the European integration process.

We also found differences between news and opinion articles, when it comes to the magnitude of EU politicisation. More concretely, the salience and contestation of the EU have been, in general, higher in the opinion articles with the crisis accentuating those differences. Furthermore, our analysis also shows that, in terms of EU salience and contestation, the differences between newspapers become more noticeable when we consider only opinion articles, in comparison to more traditional journalistic pieces. This suggests that journalistic/editorial norms contribute to a certain standardisation of the extent to which the EU is politicised in the media, or that the strict news coverage made by mainstream newspapers depicts very accurately the overall levels of EU politicisation in a certain country.

Our analysis shows that it is essential to consider opinion items, not only in attempts to explore differences within mainstream media in the politicisation of the EU, but also to better understand the role played by the media in (re)shaping that politicisation. In cases like Portugal and Spain, we observed interesting patterns in the politicisation of the EU in opinion pieces. For instance, the newspaper with higher EU salience is also the one showing lower levels of EU contestation. Moreover, after the eurozone crisis, it was the right-leaning newspapers that became the most negative towards the EU. It is important to stress that these differences were observed in mainstream newspaper that are very close to the centre of the ideological spectrum. To put in other words, the differences in the politicisation of the EU should be higher when comparing more ideologically distant media outlets.

In this chapter, we went beyond the magnitude of EU politicisation to focus on how differentiated that politicisation has been in the traditional news coverage of legislative elections. Our results show that relying exclusively on the magnitude of politicisation offers an extremely incomplete picture of how the traditional media politicised the EU before and after the eurozone crisis. In fact, the politicisation of the EU, after the crisis, varied considerably within the group of bailout countries analysed in this book. In Spain, *domesticated policies* have always been the most salient dimension in articles about the EU. In Portugal, that dimension only became the most salient aspect in the elections after the Portuguese bailout. In the case of Ireland, in all of the four elections analysed, the EU was primarily debated in terms of *EU policies*, even though this trend was finally challenged, in the last election of 2016, by the *membership*  dimension. Finally, in the Greek case, we found that the crisis changed considerably the way in which the EU was politicised in mainstream media, with *membership* becoming for the first time, in the critical election of 2012, the most frequently debated dimension.

Overall, our analysis highlighted substantive differences within the group of debtor countries that can be important to make sense of the seemingly different implications, or effects, that the politicisation of the EU might have on EU issue voting, which is the central aspect of this book. Indeed, the study of the implications of EU politicisation should consider not only how much but also in which ways has the EU been politicised in each country. As Baglioni and Hurrelmann (2016) suggest, part of the difficulties to understand the implications of EU politicisation might arise from an undifferentiated treatment of that phenomenon.

While this chapter focused on the media, the role of different political actors is an equally important dimension in the study of EU politicisation. The next chapter focuses on the institutional arena to analyse how the EU has been (de)politicised by the different parties during parliamentary debates. Using a unique dataset of parliamentary speeches, the next chapter analyses how the main political parties have debated the EU before and after the eurozone crisis. Moreover, it also explores, at the party level, what are the main determinants of EU politicisation in the parliamentary debates of the six MAPLE countries.

#### Appendix

See Tables 2.2 and 2.3.

**Table 2.2** List of elections, newspapers, observation periods and the number of articles analysed


(continued)


**Table 2.2** (continued)

(continued)


**Table 2.2** (continued)


**Table 2.3** List of the expressions used to identify EU articles

### References


the European Union. *European Journal of Political Research, 58*(3), 797–819. https://doi.org/10.1111/1475-6765.12321


**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# Parties Acting Strategically: National Parliaments' Role in Holding the EU Accountable

*Yani Kartalis and Tiago Silva* 

# Introduction

The existing literature assessing the magnitude of EU politicization has focused predominantly on the intermediary arenas of political communication. In line with some of the main findings of the previous chapter,

Y. Kartalis (B)

© The Author(s) 2023 M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_3

Center for International Studies, Instituto Universitário de Lisboa (ISCTE-IUL), Lisbon, Portugal e-mail: kartalisie@sodas.ku.dk

Copenhagen Center for Social Data Science, University of Copenhagen, Copenhagen, Denmark

T. Silva

Institute of Social Sciences, University of Lisbon, Lisbon, Portugal e-mail: tiago.silva@ics.ulisboa.pt

those studies have shown us that EU politicization occurs on a punctuated basis in certain countries/regions, amid important EU-related events. Differently, however, the degree to what EU issues has been politicized in the institutional arena remains largely untapped in a comparative and longitudinal basis. This chapter addresses this lack of research by offering an extensive and comparative analysis of EU politicization in the parliamentary arenas of six EU member-states.

Using an original dataset of parliamentary plenary speeches from the six MAPLE countries, over a period of 20 years, this chapter explores the following three research questions: (1) To what degree has the EU been politicized, in terms of its salience and contestation, in the parliamentary debates of different EU countries? (2) Did the Eurozone crisis impact the frequency and the way in which the European integration dimension is discussed at the institutional level? (3) what party characteristics, namely, left–right ideology, incumbency and EU position, are associated with higher levels of EU politicization? We explore these questions by combining automated and manual methods of content analysis, applied to a corpus of 724,963 speeches.

Our research questions matter for at least three main reasons. The first one concerns, as we mentioned, the lack of longitudinal and crossnational research on how the EU has been debated in this arena, particularly covering the period following the onset of the Eurozone crisis. The existing studies on EU politicization not only have mainly focused on other channels but also, when looking at the institutional arena, the literature mainly deals with the parliament's legislative and oversight roles, rather than its communication function. While the usefulness of mainstream media in understanding societal interests, and their levels of contestation, is indisputable, national parliaments remain the most relevant and consequential platform to debate complex political issues such as the regional integration process.

The second reason concerns the anticipated differences, in terms of political communication, between institutional and intermediary arenas. Even though the main political actors in media and parliaments are largely the same, the two arenas have different operating logics, and rules, that can distinctly shape the EU's salience and contestation. This is a crucial aspect since the media logic influences the content of the news coverage of political events. Therefore, a key aspect of the parliaments' communication function is offering an unmediated measurement of EU politicization and an undistorted picture of how the EU has been discussed by the citizens' elected representatives. An analysis that, therefore, complements the existing studies of media content, offering us a more encompassing understanding of EU politicization.

Finally, the third reason concerns the increasing importance attributed to parliamentary debates for the democratic legitimacy of the European polity. When it comes to the different modes of representation in the EU, the national parliaments, and the deliberation that occurs in them, are the most important foundations for the emergence of a European '*demoi-cracy*'. In this sense, the comparative analysis made in this chapter is important not only to offer a more nuanced, or alternative, assessment of the magnitude of EU politicization in different European countries but also to corroborate envisioned expectations regarding the role of national parliaments in furthering the European integration process. To put it in other words, as it is argued in the introduction of this book, the focus on national political institutions is crucial to understand how the growth in Europeanization is being legitimized. The analysis in this chapter not only serves that purpose but is also a fundamental contribute to the main goal of this book, which is to understand the effect of EU politicization on national elections.

This chapter is structured into five sections. After this first introductory part, the second section of the chapter offers a brief literature review of EU politicization in national parliamentary debates, focusing on its magnitude and determinants. The third section of the chapter presents the research questions and expectations of this study, as well as the data and methods used to investigate them. After that, the main results of our analysis are presented and discussed in the fourth section of the chapter. Finally, in the last section, we summarize our main findings and discuss some of their potential implications.

# Literature Review

National parliaments are, unquestionably, the cornerstones of representative democracy and political contention in the parliamentary systems of the different EU countries. Yet, the two main conceptions of the European integration process (i.e., intergovernmental and supranational) have always attributed a secondary/indirect role to national parliaments regarding their involvement in the EU policy-making (Wendler, 2016). These two main 'schools of thought', or lenses to study regional integration, have focused either on the role of national governments or the supranational institutions and elites of the EU (Schimmelfennig & Rittberger, 2006; Tsebelis & Garrett, 2001). Moreover, the perceived decline of parliaments' capacity to carry out their functions in Western democracies (Döring, 1995a) might have contributed as well to the lack of attention given, by the literature, to the national parliaments' involvement in the European integration process. However, this situation started to change in the early 2000s with the Laeken declaration and the respective commitment of the EU to a 'greater democracy, transparency and efficacy'.<sup>1</sup>

The increasing debates, and concerns, surrounding the lack of democratic legitimacy led to a distinct formulation of the European polity, as a demoi-cracy (Besson, 2006; Nicolaïdis, 2004), that placed considerably more importance in the relationship between national parliaments and the EU.2 This became particularly evident with the Lisbon Treaty, which gave national parliaments a more important role in EU affairs by strengthening their scrutiny and participation rights (Auel et al., 2018; Nicolaïdis, 2013). The public debates are an essential aspect of this emerging conception of Europe, which demands, as well, a more important communicative role of parliaments. In this view, the national parliament became a central arena for public reflection and debate about the EU (Wendler, 2016), which, in a way, has sparked the interest of EU scholars on the, somehow overlooked, communication function of those institutions. To put it in other words, the communication and information role of parliaments has become a key dimension for the study of EU politicization.

While acknowledging the multifaceted and resourceful nature of the parliaments, this chapter focuses on the parliaments' role as arenas of political communication. In that context, the public debates in the plenary are, notwithstanding the question of whether citizens actually follow

<sup>1</sup> Opinion of the Committee of the Regions on the Treaty establishing a Constitution for Europe (2005 C 71/01), *Official Journal of the European Union*.

<sup>2</sup> The ability of national parliaments to use the EU politicization and debates to progress with the European integration, however, remains uncertain or challenged (Bellamy & Kröger, 2016; Riekmann & Wydra, 2013).

them,3 the most important means to fulfil the information and communication functions of parliaments (Auel & Raunio, 2014a, 2014b). As Auel and Raunio (2014a, p. 4) put it, the plenary debates offer a concrete arena for both 'articulating and representing societal interests and informing the electorate about issues on the political agenda'. These debates are therefore important for citizens to hold political parties accountable for their positions on the EU, even though the parties have some strategies available to them, such as 'blame avoidance' and 'credit claiming', that can still blur that accountability (Lord & Pollack, 2010; Hobolt & Tilley, 2014). To put it differently, the plenary debates are both an important and unique setting to access the magnitude of EU politicization since they offer, in the utmost consequential setting, an unmediated discussion of the most pressing societal issues, as well as the stance of the key political parties on them. For those reasons, most of the existing studies assessing the politicization of the EU at the institutional level have focused on plenary debates (e.g., Auel & Raunio, 2014b; Rauh, 2015; Rauh & de Wilde, 2018).

When it comes to the magnitude of EU politicization in national parliaments, we can find some opposing views in the literature. For some authors, on the one hand, it seems indisputable that European integration has always been politicized at the institutional arena, even though those high levels of contestation might not be reflected in the political actors' communication addressed to the citizens (Hurrelmann et al., 2015, p. 46). Somewhat in line with this idea, Wendler (2016) found that normative EU claims are both more frequent and more contentious in parliamentary debates, compared to more pragmatic ones.

On the other hand, the strong incentives for parties to strategically *depoliticize* the European integration process are also very well known (see Mair, 2000) and, to a great extent, also more empirically substantiated. In fact, some studies have shown that most national parliaments do not 'live up to their task of bringing "Europe" closer to the citizens' (Auel & Raunio, 2014a, p. 10). Using an automated content analysis of parliamentary speeches in four EU countries, from 1991 to 2015, Rauh and de Wilde (2018) also showed that, despite being somehow responsive to supranational decisions and events, there is a lack of balanced debates on European integration and a limited supply of electoral choice. Looking

<sup>3</sup> It is however shown that an increasing involvement of national parliaments in EU affairs is reflected in the news coverage of traditional media (Auel et al., 2018).

at four countries (Finland, France, Germany, and UK), between 2002 and 2012, Auel and Raunio (2014b) find that, despite some cross-country variation, EU politicization has been generally low in parliamentary debates. However, these authors observe an increase in EU politicization in the last two years of their analysis (2011 and 2012), particularly in the Finnish *Eduskunta*.

Overall, the perceptions of high levels of EU politicization at the institutional level seem to be based more on conventional wisdom than thorough comparative research. As we saw, despite considerable variation between countries and parties (Lauwers et al., 2021), the existing longitudinal and cross-country research suggests that the levels of EU politicization in parliaments have been, overall, relatively low, particularly when it comes to the sheer volume of debates. Moreover, similar to what we see in the media (Hutter et al., 2016), the EU becomes more salient in parliamentary debates during periods around important EU events (Rauh & de Wilde, 2018). In this sense, in line with what research found for media (Hutter & Kriesi, 2019, Silva et al., 2022), the eurozone crisis might have contributed to higher levels of EU politicization also in the parliaments. Some studies have indeed pointed in this direction, suggesting that EU politicization increased in the parliaments' debates (Auel & Raunio, 2014b; Riekmann & Wydra, 2013) and activities (Auel & Höing, 2014) with the eurozone crisis. Other authors, with more encompassing assessments, have suggested a less significant impact of the crisis in the parliaments' politicization of the EU (e.g., Rauh & de Wilde, 2018). Overall, in the institutional arena, the impact of the eurozone crisis in the magnitude of EU politicization, across countries and political parties, remains a puzzling and inconclusive subject.

Regarding the comparative assessment of EU politicization in the institutional arena, and the possible impact that the eurozone crisis had on it, the literature has taken into consideration two important dimensions or categories of factors. The first one concerns the existence of considerable differences between parliaments, arguably much higher than in the media, that make cross-national comparisons not only important but far more challenging (Döring, 1995a). When it comes to the parliamentary debates, institutional factors such as the governments' level of agenda control (Döring, 1995b), type of legislature, being a 'debating' or 'working' parliament (Dann, 2003) and degree of EU affairs delegated to committees (Auel & Raunio, 2014b), can potentially shape how much, and in which way, the EU is debated. The literature has found, however, a rather negligible impact of institutional factors on the levels of EU politicization (Auel & Raunio, 2014b). Instead, actors seem to be the most important factor in explaining the levels of EU politicization in parliaments.

The second important dimension in the assessment of EU politicization, at the institutional level, concerns the main political actors and the differences between parties in terms of their characteristics and respective roles in the legislature. More concretely, strategic considerations, and therefore government/opposition, left/right and pro-/anti-EU differences seem crucial for the (de)politicization of the EU issues in national parliaments (Kaniok & Brusenbauch Meislova, 2021; Lupato, 2014; Navarro & Brouard, 2014). Similarly, Wendler (2016) also found that EU contestation in parliamentary debates is mainly determined by the existence of Eurosceptic parties, being institutional arrangements, and public opinion on the EU, rather insignificant dimensions.

When it comes to the discussion of EU topics, some studies also found that incumbent parties outperform the challenging ones (Rauh & de Wilde, 2018). Moreover, a far more puzzling finding from Rauh and de Wilde (2018) is that opposition parties seem to debate the EU less when citizens' Euroscepticism increases. In line with that, Auel and Raunio (2014b) found that, overall, the EU salience is higher in contexts where party conflict and public Euroscepticism are both low. In this sense, differently from what we saw for the media in the previous chapter, the correlation between EU salience and contestation seems to be negative in the parliaments, which contributes to a lack of EU accountability in this arena.

Finally, moving beyond the magnitude of politicization, the aspects of the European Integration discussed in the parliaments can also have important implications for the EU issue voting. According to the literature, there are different dimensions of the EU that can be the object of political contestation (Mair, 2007). In this regard, Hurrelmann et al. (2015) offer a useful typology that categorizes four types of EU debates. The first dimension, *membership*, concerns the creation, consolidation, and geographical reach of the EU institutions. The second one, *constitutional structure*, deals with debates surrounding the procedures and responsibilities of the different EU institutions, as well as the EU's decision-making process in general. Finally, the last two dimensions deal both with the reach and penetration of EU legislation in the member-states. While one, *EU policies*, focus on policies emanated from EU institutions, the other, *domesticated policies*, concerns national-level policies that are somehow a consequence of the European integration process.

Overall, regarding the magnitude of EU politicization in parliamentary debates, the existing literature has echoed two key aspects. The first one is that existing differences between parliaments seem to be determined mainly by party-level characteristics rather than institutional factors or levels of EU politicization in the other two arenas (i.e., citizens and media). In fact, looking at three EU countries (Denmark, Ireland, and the Netherlands), de Wilde (2014) finds that, when it comes to the politicization of Europe, differences between arenas (i.e., media and parliaments) are more substantive than cross-national and diachronic variation. The second aspect is that the Eurozone crisis, differently from the media, might not have contributed to a higher politicization of the EU in the institutional arena. In this sense, if EU politicization did not increase, the crisis might have contributed to a deficit of democratic accountability in the EU member-states. These two aspects are explored in this chapter with a longitudinal analysis of EU politicization in the parliamentary debates of the six MAPLE countries.

### Expectations, Data, and Methods

To answer our research questions, we develop and test five expectations, based on existing literature, regarding the results of our analysis of parliamentary debates. First of all, we expect that the magnitude of EU politicization in parliamentary debates will vary more across countries/parliaments than across time. The literature has consistently showed that the overall levels of EU politicization in parliaments can differ considerably between countries, being particularly higher in places where, intriguingly, the EU is a less contested issue. This is justified by the fact that it is less risky for parties to debate the EU when the electorate has a predominantly favourable view of the European Integration process. In this regard, concerning the MAPLE countries, literature has found the salience of the EU to be comparatively higher in the German parliament (Auel & Raunio, 2014a, 2014b, Rauh & de Wilde, 2018) and lower in the parliaments of Spain (Lupato, 2014) and Greece (Bélanger & Schimmelfennig, 2021). The situation in the remaining countries is much less known from a comparative perspective, making, therefore, this chapter's analysis even more important.

The second expectation concerns the impact of the crisis. Similar to what was found for the media in the previous chapter, we also anticipate, in the parliaments, an increase in EU salience and contestation after the Eurozone crisis. As we mentioned, there are some hints of evidence in the literature to the fact that the EU politicization increased in parliamentary debates after that crisis (Auel & Raunio, 2014b; Riekmann & Wydra, 2013), even despite none of those studies includes one of the *debtor countries* (i.e., Greece, Ireland, Portugal, and Spain). However, at the same time, the salience of EU in parliaments also seems to be negatively associated with its contestation levels. These seemingly contradictory findings suggest, as we anticipate for this chapter, that the eurozone crisis had a more noticeable impact on the levels of *contestation* rather than on the *salience* of the EU dimension.

The third expectation concerns the dimensions of the EU that are more salient in the parliamentary debates. On the one hand, the multifaceted nature of parliaments and its diversity of actors should stimulate a higher variety of debates about the EU, which is also important for the legitimization of the European integration process. On the other hand, as Bartolini (2005, p. 349) discusses, because political parties lack the capacity to debate constitutive issues of the EU, they seem to have adopted a strategy of 'depoliticization' of that specific dimension (Mair, 2007). Therefore, similarly to what we observed for the media in the previous chapter of this book, we also anticipate that the EU debates, in the national parliaments, have remained predominantly in the realm of policies, instead of dealing with aspects related to EU membership and constitutive issues.

The fourth expectation concerns the anticipated differences, between types of parties, in the levels of EU politicization. More concretely, we explore three different distinctions. The first one is government versus opposition. In this regard, we anticipate that government parties mention the EU more frequently in their interventions in the parliament (Lupato, 2014; Rauh & de Wilde, 2018). The second is the position of the party towards the EU. Here we expect that Eurosceptic parties are likelier to politicize the EU in their parliamentary speeches (Navarro & Brouard, 2014; Wendler, 2016). Finally, the third dimension considered is the ideological leaning (left–right) of the parties. While we do not have a concrete expectation regarding this dimension, we do expect to observe distinct patterns across the six countries analysed regarding this distinction.


**Table 3.1** Total number of speeches per country and period

The fifth, and final, expectation concerns the relationship between salience and contestation. Some studies have found, at the country level, that the salience of EU is higher in countries that show lower levels of EU contestation (e.g., Auel & Raunio, 2014b). In line with that, we also expect that the countries with higher levels of EU salience are also the ones where the proportion of articles with a negative tone towards the EU is lower.

Our research question and expectations are examined using an original dataset of plenary sessions' speeches from six different European parliaments: Chambre des Représentants (Belgium), Bundestag (Germany), Vouli (Greece), Oireachtas (Ireland), Assembleia da República (Portugal), and Congreso de los Diputados (Spain). The dataset includes all the speeches with more than 40 words4 from a period of 20 years (see Table 3.1). The dataset includes, for each speech, relevant information such as its date, speaker's name, and the party that he or she belongs to. In total, the final dataset used for this analysis includes 724,963 unique speeches.

Our assessment of EU politicization in parliaments focuses on two dimensions of the concept: *salience* and *contestation*. 5 We operationalize EU salience as the percentage of speeches, in a particular year or legislative term, that mention the EU. We identify the EU speeches using an

4 We cleaned the original dataset by removing speeches that, by having less than 40 words, are likelier to correspond to interruptions or other non-meaningful interventions.

<sup>5</sup> We mirror the approach used in the previous chapter and in Silva et al. (2021). Nevertheless, it is also far less meaningful to assess the third core dimension of EU politicization, the expansion of actors (de Wilde, 2011; Schmitter, 1969), at the institutional level/arena.

extensive list of EU-related terms (Appendix: Table 3.2). This list, translated into seven languages, was adapted from the codebook of Maier et al. (2014). Overall, about 12% (86,751) of the speeches in our dataset mentioned the EU.

The dimension of EU contestation was operationalized as the proportion of speeches in a year, or term, that had a negative tone towards the EU. The process to calculate this score was the following. The first step was extracting, for each speech, all sentences mentioning the EU. Since our dataset includes speeches in seven different languages, we then translated those EU sentences into a single 'pivot' language (Lucas et al., 2015), using the R package googleLanguageR(), which gives users access to Google's translation Application Programming Interface (API) service. After that, the EU sentences' tone was measured using the algorithm provided by the package sentimentr(),6 by Rinker (2019). A speech was considered negative if the average of all its EU sentences' tone values/scores was lower than zero. Similar to the previous chapter, we assume that the contestation of EU increases, in a particular year or term, when the share of negative articles also increases.

The different manifestations of the EU discussed in the parliamentary speeches were manually coded by a team of trained graduate students that identified, from a random sample of speeches mentioning the EU, if the speech dealt with (1) membership; (2) constitutional structure, (3) EU policies, or (4) domesticated policies (Hurrelmann et al., 2015, p. 45).7 Additionally, each party with parliamentary representation was also coded, for the three categories of interest, using Chapel Hill expert surveys. These categories were whether the party was in government, the CHES general left/right variable, and, lastly, the parties' stance towards the EU.

Finally, to better understand the determinants of EU politicization at the party level, and further test our main expectations, we have also conducted a series of linear regression analyses with our dataset. The two dependent variables used in the models were EU Salience and EU Contestation. The unit of analysis in the models was the party score for each plenary session and includes all political parties with parliamentary

6 This algorithm takes into consideration valence shifters, or augmented sentiment dictionary lookup. This guarantees that expressions like 'not good' do not receive a positive score, or that 'extremely bad' has a more negative score than 'a little bad'.

7 Read Chapter 2 or Hurrelmann et al. (2015) for a clarification of each dimension.

seats. In addition to our variables of interest (EU stance, Ideology, and Incumbency) we also include, as controls, the parties' seat share, if the party is from a debtor country and if the plenary session occurred after the Eurozone crisis (2009). Additionally, the third and fourth model of each table also includes, respectively, country dummies and an election variable, controlling for the last year of each plenary session, to analyse a time trend. All variables included in the models were normalized to vary between 0 and 1.

#### Results and Discussion

Regarding the magnitude of EU politicization in the six MAPLE countries (Fig. 3.1), we can observe substantial cross-national variation in terms of EU salience. On the one hand, as we anticipated, the EU has been a relatively salient dimension in the German parliament. The same can also be said for Greece and Ireland. In fact, it was in Greece that, on average, the EU was salient the most (20.2% of the speeches), followed by Germany (16.6%) and Ireland (14.3%). Moreover, it seems that the Eurozone crisis might have indeed contributed to a higher visibility of the EU in the national parliaments of Greece and Ireland. On the other hand, the salience of the EU has been comparatively low in the other three parliaments, particularly in the case of Portugal where, despite a slight increase after 2009, only 2.9% of the speeches, on average, mentioned an EU-related term. In this particular dimension of EU politicization, Belgium and Spain had on average 6.2 and 6.8% of EU speeches, respectively. Moreover, in the case of Spain, we can also observe a slight increase of EU salience after 2011.

When it comes to contestation, different from the EU salience, the variation within countries seems more relevant than the differences between them. Nevertheless, on average, the EU has been a more contested topic in Greece (35.3), Belgium (31.8), and Portugal (28.3). Contrarily, the German parliament showed the lowest levels of EU contestation. On average, only 22.4% of the Bundestag speeches about the EU had a negative tone. It is very interesting to see that, as we anticipated, and in line with previous studies (e.g., Auel & Raunio, 2014b), with the exception of Greece, the EU contestation is higher in the parliaments with low EU salience and vice-versa. To put it in other words, apart from Greece, EU contestation has hardly been a prominent aspect in

**Fig. 3.1** The Politicization of the EU per legislative term in six parliaments (EU salience indicated in the lines)

the plenary debates of the six countries. Still, we can observe an overtime increase in EU contestation in some of the countries, particularly in Portugal, Spain, and Greece.

The second aspect analysed in this chapter concerns the different EU dimensions debated in the parliaments. Once again, we can observe noticeable differences between the six countries (Fig. 3.2). In some cases, like Spain and Portugal, *domesticated policies* have been consistently the most salient dimension of the EU debated in the parliaments. On the contrary, in the other four countries, the debates about *EU policies* were consistently more frequent. The remaining dimensions of the EU have been much less debated in the plenary sessions. The only exceptions are the comparatively higher salience of *constitutional structure* debates in Portugal and, to a lesser extent, also in Germany.

As we anticipated, the analysis confirmed the low salience of the *membership* dimension in the parliamentary debates of the six countries.

**Fig. 3.2** The salience of different EU dimensions in legislative debates

This dimension was only relatively salient in some of the periods analysed in the Greek and German parliaments. In fact, in Greece, the results suggested that the high salience of the membership dimension (before 2003, 2012, and 2015) might be related to important EU events, namely the 5th EU enlargement and the Eurozone crisis. Overall, the low salience of this dimension, which is undoubtedly the most polarizing manifestation of the European integration process, might help us understand not only why the overall EU contestation is low in the institutional arena but also why the EU contestation is comparatively high in the Greek parliament.

When it comes to this dimension of *membership*, we can also observe some interesting differences between political parties (Figs. 3.3 and 3.4). Firstly, we can see that, particularly when looking at Greece, but also Belgium and Germany, the parties from the right were likelier, compared to the ones from the left, to discuss the membership dimension of the EU. Secondly, the discussion of the membership dimension, in Belgium and Germany, seems to be mainly 'driven' by the more Eurosceptic parties. Somehow differently, in Greece, the situation seems to have changed over time. While the Eurosceptic parties focused more on the membership dimension in the first legislative term analysed, the pro-EU parties started to emphasize more this dimension of the EU in the following plenary debates, making this change particularly evident after the beginning of the eurozone crisis. Finally, we can also observe a clear difference between pro- and anti-EU parties in Belgium when it comes to the type of policies discussed. More concretely, while the pro-EU parties in Belgium focus predominantly on debates about EU policies, their Eurosceptic counterparts prioritize the domesticated policies in their interventions.

The remaining analyses made in this chapter focus on party differences. More concretely, we examine the politicization of the EU in national parliaments by making three distinctions, or categories, of parties (ideology, incumbency, and EU stance). The first category is whether a party is from the left or right (Fig. 3.5). When it comes to EU contestation, the only clear pattern that we can observe between left and right

**Fig. 3.3** The salience of different EU dimensions in legislative debates by left and right parties

**Fig. 3.4** The salience of different EU dimensions in legislative debates by Eurosceptic and Pro-EU parties

is in Ireland. More precisely, in the case of Ireland, the parties from the left have been consistently more negative towards the EU. Interestingly, at the same time, Irish parties from the right were the ones more likely to mention the EU in their speeches.

In addition to that, it seems that, overall, the EU is indeed likelier to be more contested by left-wing parties, particularly after the Eurozone crisis. This pattern is particularly evident in the group of creditor countries. In Belgium and Germany, while the EU had been a more contested issue for the right in the first two legislative terms, the situation inverted in the remaining period of analysis. There are some possible explanations for the higher EU contestation levels from leftist parties. One is that, at least in this group of six EU countries, the extreme parties from the left are likelier to have better electoral success and, therefore, more visibility in national parliaments. Nevertheless, it is interesting to see that, overtime, our data shows that the EU has become increasingly more contested on the left side of the political spectrum.

**Fig. 3.5** The politicization of the EU in legislative debates by left and right parties (EU salience indicated in the lines)

The second distinction that we explore is whether a party is in government or not. As we saw, the literature suggests that government parties are likelier to discuss the EU in their plenary interventions. This was clearly not the case for Greece and Belgium, where the opposition has been, in a very consistent way, likelier to mention the EU in their speeches (Fig. 3.6). In other countries, government parties indeed seem likelier to talk about the EU, even though those differences seem relatively small and inconsistent. Once again, the most consistent patterns were found for EU contestation. The parties in the opposition have been more negative towards the EU in practically all the periods analysed. The only exceptions were a single legislative term in Belgium, Greece, and Spain.

The third characteristic analysed is the parties' position towards the EU (Fig. 3.7). In this case, we anticipated that the more Eurosceptic actors were also more likely to politicize the EU during parliamentary debates. With the exception of Spain, this expectation was confirmed in all countries. The difference is particularly clear in terms of EU contestation,

#### 72 Y. KARTALIS AND T. SILVA

**Fig. 3.6** The politicization of the EU in legislative debates by government and opposition parties (EU salience indicated in the lines)

with the Eurosceptic parties being much likelier to discuss the EU in a negative way. In the cases of Greece and Belgium, the Eurosceptic parties were also more likely to mention the EU in their speeches in comparison with the parties with a more pro-European stance.

As we mentioned, the Spanish parliament is the only exception to this pattern of higher EU contestation from Eurosceptic parties. Interestingly, this was only the case in the three legislative terms after the Eurozone crisis. In this sense, it seems that this crisis had some impact on how—or by whom, to be more precise—the EU was politicized in the Spanish parliament. Nevertheless, in the last term, once again the Eurosceptic actors had a more negative tone when debating the EU.

Overall, when it comes to differences between parties, the clearest pattern identified in our dataset is that Eurosceptic parties were consistently likelier to contest the EU in their parliamentary speeches. This result, while not surprising, is an important validation of our measure

**Fig. 3.7** The politicization of the EU in legislative debates by Eurosceptic and Pro-EU parties

of EU contestation. Still, the differences between parties were not as straightforward when it comes to EU salience. In fact, when looking at the institutional arena, salience, and contestation, differently from the media, seem to follow different logics. More concretely, the media logic of giving more salience to contested topics does not seem to apply to the plenary speeches (Table 3.3).

Finally, we investigate in this chapter the relationship between EU politicization and key party-level characteristics using linear regression analysis. The coefficients of the regressions, using EU salience and EU contestation as dependent variables, are plotted in Fig. 3.8. These plots refer to models8 3 and 6, with the country controls omitted and 95% confidence intervals. As we can see, when it comes to the salience of the

<sup>8</sup> The results of the regression models are reported in the Table 3.4, in the Appendix. Some observations, with an extremely low number of EU speeches, were dropped to significantly improve the goodness-of-fit of the models.

EU in the parliamentary speeches, the parties' position on the EU issue is the only variable of interest with a statistically significant effect. Moreover, as we anticipated, the more Eurosceptic a party is, the likelier it is for the EU to be mentioned in that party's parliamentary interventions. This variable was statistically significant (at the 0.001 level) in the three models.

Concerning the remaining explanatory variables, we did not find a statistically significant effect of ideology (left–right positioning) and incumbency (being in government) on the salience of the EU. The only other aspect in the models, besides EU position, with a significant effect on EU salience, concerns the period analysed. The models show that the salience of the EU has increased in the parliaments over time. Moreover, as we anticipated, the salience of the EU in parliamentary debates post-2009 was also higher, being the difference statistically significant at the 0.05 level.

When it comes to EU contestation, which refers to the proportion of EU speeches that were negative, we have found negative and statistically significant effects for all three main independent variables (EU position, Left/Right, and Government). The strongest predictor of EU contestation in parliamentary speeches was the party's position on the EU, with the models showing that, unsurprisingly, the more Eurosceptic parties were also much likelier to contest the EU in their speeches. This is

**Fig. 3.8** Coefficient plots from linear regression analyses on EU Salience and Contestation (models 3 and 6 from Table 3.4, in the Appendix), with omitted country dummies

an anticipated finding that, nevertheless, further validates our automated measure of EU contestation.

The parties' ideology, on the left–right scale, has also a statistically significant effect on the proportion of negative speeches towards the EU. More concretely, the levels of EU contestation increase significantly when the parties' ideological positioning moves to the left. Contrarily, the parties on the right seem to debate the EU dimension with a more positive tone. When it comes to the effect of incumbency, as we anticipated, the EU dimension is more likely to be contested by the parties in the opposition.

Overall, in line with the existing literature, we found that parties' position on the EU, and being in opposition, are both important predictors of EU contestation in the parliaments. Moreover, the contestation of the EU also seems to have been increasing over time, having been significantly higher in the period post Eurozone crisis. Considering that time is a good predictor of both EU salience and contestation, our analysis indeed suggests that the democratic legitimacy of the European integration has slightly increased over time, in the past two decades.

### Conclusion

In this chapter, we offered an extensive analysis of the EU politicization in the institutional arena. Similar to previous assessments, our analysis confirmed the relatively low politicization of the EU in the institutional arena, a situation that even the eurozone crisis was not able to noticeably change. Despite important cross-national differences, the overall salience of the EU in the parliaments has been relatively stable and low. When it comes to the contestation dimension, while its variation seems more pronounced, the general tone towards the EU has only swung from very positive to less positive. In this sense, despite the intentions expressed in the Lisbon treaty, we cannot say that, perhaps with the exception of Greece, national parliaments have made a very noticeable contribution to reduce the democratic legitimacy of the European Union.

Considering the discussed determinants of EU politicization, and the reason why parties might strategically avoid politicizing the EU, it remains puzzling the reason why the Eurozone crisis did not lead to a more noticeable politicization of the EU in national parliaments, particularly when compared to the media. Factors such as the lower number/variety of active actors (compared to the media), the institutional obligation to a more 'responsible' (and less 'responsive') action, as well as the need to maintain a sense of legislative capacity (to avoid suggesting that their hands are tied), might have all contributed to this outcome. Furthermore, it is also possible that the existence of EU-specific committees has also contributed to this low salience of the EU in the plenary debates. Regardless of the reasons, this chapter highlights the importance of looking at each country individually to better understand the different nuances and implications of the EU politicization.

Nevertheless, our statistical analysis of EU politicization confirmed two important aspects regarding the party-level determinants of EU salience and contestation. On the one hand, the position of the parties in the EU issue was the strongest predictor of EU politicization, being statistically significant for both EU salience and contestation. On the other hand, the parties' position on the left–right scale and being in the opposition had a significant effect only on contestation. More concretely, our analysis showed that, in the parliamentary speeches of the six countries concerned, the EU was likelier to be contested by the parties on the left and the parties in opposition.

Overall, this chapter confirmed the existence of strong cross-national variation at the institutional level. Our results highlight the importance of a less extensive and more in-depth analysis of the EU politicization, as well as suggesting the importance of the arena where politicization is the measure. Using a different methodological approach, the next chapter offers a more detailed picture of how the EU was politicized in the six countries and the similarities between the institutional and the intermediary arena of political communication.

### Appendix

See (Table 3.2, Table 3.3, and Table 3.4)


**Table 3.2** List of the expressions used to identify EU articles

**Table 3.3** Summary of descriptive statistics of the variables analyzed



Robust standard errors in parentheses \*\*\* *p* < 0.01, \*\* *p* < 0.05, \* *p* < 0.1

#### Bibliography


**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# Polity or Policies? The European Union in Parliamentary Debates and the Media

*Nelson Santos and Susana Rogeiro Nina* 

# Introduction

In the two previous chapters, the salience and contestation of the EU in parliaments and newspapers over time were measured using a similar framework and automated methods. While the novelty of a similar framework is important in itself, it also allows for comparisons. Although both arenas relate to and interact with each other (e.g. Auel et al., 2018), they have very different institutional characteristics and purposes and are composed of different actors. As is well known, the media is not a neutral transmitter of messages, and its commercialization process has led to an increase in commercial programming, implying a decrease in useful and compelling political content (McChesney, 2000). On the other hand,

N. Santos (B)

S. R. Nina Universidade Lusófona, Lisbon, Portugal e-mail: susana.rogeiro.nina@ulusofona.pt

© The Author(s) 2023

83

Instituto de Ciências Sociais da Universidade de Lisboa, Lisbon, Portugal e-mail: nlrsantos@edu.ulisboa.pt

M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_4

parliamentary debates follow distinct procedures that are designed around the law-making process and its outputs. Acknowledging those differences and their potential effects on the politicization of the EU turns this comparison into a valuable enterprise, as one could reach different conclusions and anticipate distinct outcomes depending on the arena considered. Therefore, here we address both arenas comparatively over time and in different countries.

To achieve this, the chapter builds on previous literature that considers the EU in its multiple dimensions (Hurrelmann et al., 2015), assessing four dimensions: (i) membership; (ii) constitutional structure; (iii) European policy issues; and (iv) domesticated issues. This typology allows us to better grasp the scope and possible consequences of the EU politicization process. While politicization of policy-related issues might suggest the EU is in the realm of "normal politics" (Bartolini, 2005, pp. 347–362), if the polity is still strongly salient in the political debate, this signals that its foundations are still contested. Since this chapter shows policy issues, both European and domesticated, are the most debated topics in the media and parliaments in respect of the EU, we take a step further by establishing what policies are the most salient in each arena.

As in the previous chapters, the empirical analysis here will focus on two elements of politicization—salience and tone—and will make use of the data that has been collected by the MAPLE project, focusing on the same countries (Belgium; Germany; Greece; Ireland; Portugal and Spain) over the same time period (2002–2017).

The chapter is structured as follows. First, we offer an overview of the state of the art on the phenomenon of EU politicization, on the media and the role of parliaments in it, and how these two arenas interact with each other in the process. Second, we set out down our expectations on how the politicization of the EU has been evolving. Third, we explain our empirical approach, detailing the data collection process as well as the research techniques employed. Fourth, we present the results and analysis. Finally, we summarize our main findings, reflecting on their potential implications for the future study of EU politicization.

#### Literature Review

#### *The Politicization of the European Union*

The concept of politicization has gained greater traction in European integration studies since the Maastricht Treaty, with different authors reaching conflicting conclusions. Hooghe and Marks' landmark study asserted that the "permissive consensus" that ruled relations between EU elites and citizens has gradually given way to a "constraining dissensus" (Hooghe & Marks, 2009). This has been verified subsequently, with the politicization of the EU at the national level confirmed by Hutter and Grande (2014) and Hutter et al. (2016b).

Research on the politicization of the EU has focused on three main dimensions: salience; polarization; and the expansion of actors (Hutter & Kriesi, 2019a; Rauh, 2015; Risse, 2015). Accordingly, the EU is considered to be politicized if there is (i) an increase in the importance attributed to European integration, (ii) a growing importance of extreme positions, either in favour of or against different aspects of the EU and (iii) a growing number of actors interested and engaged with EU topics (Kauppi & Wiesner, 2018). Moreover, past research has focused on three questions: (1) how has politicization developed over time (Hutter & Grande, 2014; Hutter et al., 2016b; Van der Eijk & Franklin, 2004); (2) what are its causes and (3) what are its consequences (Hutter et al., 2016b; Risse, 2010; Statham & Trenz, 2013). Yet, despite all the efforts, we are still "far from reaching an agreement on how and at what pace politicization has occurred" (Lobo & Karremans, 2018, p. 52).

#### *Where is the EU Politicized?*

Nonetheless, the research on the politicization of the EU has often disregarded the potential differences between distinct arenas of political discourse (i.e. institutional, intermediary and citizen arenas). According to Hurrelmann et al. (2015), politicization can be observed in different arenas: (a) institutional, which encompasses full-time politicians, such as in the European or national parliaments; (b) intermediary, constituted by political parties or the national media, the actors with the ability to link the EU decision-making process with European citizens and (c) citizen, in which ordinary citizens discuss and debate EU politics.

The interaction between the former two—intermediary (national media) and institutional (national parliaments)—is particularly interesting in the context of EU politicization. Indeed, some authors have shown, in the context of EU debates, the existence of important links between political actors and media (Auel et al., 2018) since both arenas share the same audience, which might lead to similar trends in EU politicization (de Wilde, 2014). However, the multidimensional nature of the EU (polity and policy-related issues) and the different features of the media and parliaments might lead to distinct patterns of European politicization. On one hand, parliamentary debates are intimately linked with the lawmaking process and its outputs (Fernandes et al., 2021), thus focusing on policy-related issues. On the other hand, polity-related issues are more salient in media than policy-related ones (Hutter et al., 2016b). Therefore, there are reasons to believe the multiple dimensions of the EU might be addressed in similar or distinct ways in both parliaments and the media.

Extant literature has shown the media play a pivotal role in informing citizens about political events and activities that might affect their lives while helping organizations communicate their values and interests (Bennett & Entman, 2000). As such, the media is often used as a vehicle for parliaments to connect with citizens (Bennett & Entman, 2000). Furthermore, political actors and their statements tend to be the main focus of the media when EU issues are covered (Koopmans, 2007).

Research focusing on the interaction between the media and political actors has shown it might influence the relationship between politics and the media (Van der Pas et al., 2017). For instance, in the context of elections to the European Parliament (EP), Jansen et al. (2019) found parties in each Member State are the main agenda setters in the national media in respect of the attention devoted to the EU. Also, the type of frameworks used in news coverage influences party incentives in responding to the media, i.e. the parties tend to respond to media attention if the news issues are framed in the party's terms (Van der Pas, 2014). Moreover, the type of issues discussed in the media might determine the government's and parliament's political agenda (Walgrave et al., 2008). Likewise, the media tone used will define a party's reaction to news depending on the extent to which it favours the government or the opposition (Thesen, 2013).

This connection between both arenas was categorized as political parallelism by Hallin and Mancini (2004) and developed further by Brüggemann et al. (2014). Political parallelism assesses the existing links between political actors and media and to what extent the national media reflects political divisions (Hallin & Mancini, 2004): in other words, political parallelism refers to the ties between a medium and a political actor. Nevertheless, the typology proposed by Hallin and Mancini (2004) presupposes a considerable degree of stability in political cleavages and in the behaviour of media organizations to identify consistent alignments between them. Besides that, the media organization is used as a dependent variable that evaluates the degree to which the media system mirrors party lines. More recent approaches to the concept of political parallelism suggest the concept only can be used productively if there is a competitive political system and if the institutional relationship between the media and political actors is sufficiently stable to identify interaction patterns (de Albuquerque, 2013).

Recent studies of political parallelism in Europe have found political agenda-setting is particularly strong in newspapers and parties with similar political orientations (Vliegenthart & Montes, 2014) and that those political ties are also important in countries with low levels of political parallelism where there is no such partisan bias (van der Pas et al., 2017). Analysis carried out in the Netherlands shows political parties tend to respond to the media's agenda-setting only when the issues are addressed by those newspapers read by the party's voters. Likewise, newspapers tend to respond to the agenda of parties associated with their political preferences. Brüggemann et al. (2014) discovered political parallelism is higher in countries in which the media system is more partisan, such as in Greece and Spain, and less so in the other countries included in the study.

Another important element for understanding how the EU has been politicized in the media and parliaments is the concept of newsworthiness, which is the set of attributes that may make a story worth reporting (Bednarek & Caple, 2014; Eilders, 2006). In this sense, and depending on the issues discussed and the quality of parliamentary activity, a certain topic might be newsworthy to the media (de Vreese, 2003), which in turn might change the relevance, salience and nature of certain events (O'Neil & Harcup, 2009). When the EU is considered, there is evidence parliamentary activity and parliamentary news supply on EU topics are relevant to the national media (Auel et al., 2018). Additionally, the newsworthiness of the activities of parliaments in relation to EU affairs tends to be influenced by the institutional power and EU salience in public opinion (Auel et al., 2018).

However, there are very few studies about the relationship between parliaments and media. de Wilde (2014) is a noteworthy exception, distinguishing and emphasizing the potential role of the distinct arenas in politicizing the EU and assessing both simultaneously. Parliaments and the media are two distinct arenas with different characteristics: they have different purposes; different working logic; and a distinct variety of actors that can express their political stances. Moreover, as Zürn (2016, p. 166) argues: "the public sphere as reflected in mass media is not necessarily identical with the political". The mass media's agenda-setting function often results in the prominence of issues that are "utterly apolitical", dismissing important political events in their coverage (Zürn, 2016). In this sense, it is plausible the EU has been differently politicized across arenas over time.

There is a growing body of literature looking at EU politicization in national parliaments (Auel & Raunio, 2014; de Wilde, 2011; Winzen, 2010). Most of these studies have focused on legislative debates since they are crucial moments for "electoral competition as they provide for a public articulation of societal interests and the discussion of policies, thus informing citizens about complex political issues" (Auel & Raunio, 2014, p. 13). In comparison with other parliamentary activities, such as committee work and meetings, plenary debates are more important due to their communicative function, because anything an MP says on the floor of parliament can be heard by the wider electorate and more easily conveyed by the media. Additionally, legislative debates offer parties the opportunity to represent their constituents (Fernandes et al., 2021), without risking the legislature's survival, and even represent an opportunity to produce arguments in an attempt to make the other parties change their stance (Fernandes et al., 2021).

#### *The Multidimensional Nature of the EU*

The debate surrounding European integration can assume different natures. According to Mair (2004), there are two distinct—but intertwined—dimensions of contestation about the EU. The first is the Europeanization of national public spheres, which deals with the creation, consolidation and territorial reach of European institutions; the second is focused on the penetration of EU legislation into domestic arenas. de Wilde (2011) has further developed these distinctions and identified three manifestations of EU politicization: (i) institutions; (ii) decision-making processes and (iii) policies. More recently, Hurrelmann et al. (2015) proposed another differentiation along four dimensions: (a) membership—one's own and other countries' membership as well as its costs and benefits; (b) constitutional structure—institutions, decision-making processes; (c) EU policy issues—European policies on the agenda and (d) domestic issues—national issues stemming from one's own country's membership. The former two dimensions are related to polity issues and the latter two to policy issues. It may of course be difficult to make a clearcut distinction between policy and polity issues. For instance, some of the debates on EU decision-making concerning economic and financial policies, namely, the collectivization of debt and the creation of Eurobonds, may be considered simultaneously as policy and polity debates. While acknowledging these difficulties, we still believe these categories are useful to inform our analysis.

Indeed, more recent research has considered the EU in its multidimensional role, focusing on the distinct forms EU politicization can take (Braun et al., 2016; de Wilde et al., 2016). These distinctions can have consequences: with the polarization between parties in relation to the EU being greater on polity-related than it is on policy-related issues and with polity-related issues tending to be more salient in European debates than are its policies (Hutter et al., 2016b).

Assessing the phenomenon of its multidimensionality is key, since different patterns of EU politicization might lead to very different consequences for the future of the EU. The focus on the distinction between policy- and polity-related issues can be linked to earlier research that distinguished between soft and hard Euroscepticism as different types of opposition to the EU (Szczerbiak & Taggart, 2008). Hard Euroscepticism is a fundamental opposition to European integration and preference for withdrawal from EU membership, while soft Euroscepticism stands for a general criticism of European policies over a specific period. Since negativity and conflict are two of the most important elements determining what makes the news (Galtung & Ruge, 1965; Harcup & O'Neill, 2017), principled opposition to the EU is likely to receive media attention. However, as parliaments generally address legislative matters, ideological discussions about the European polity might fall short.

Despite the valuable insights the distinction between policy- and polityrelated issues might bring, this chapter will take a step farther and distinguish the different European policies that are addressed. Presently, the EU has a wide array of competencies across multiple areas. In fact, according to the Treaty on the Functioning of the European Union (TFEU), three main types of competencies might be established: exclusive; shared and supporting. The number of exclusive and shared competencies has increased over the years, so disentangling the different policy fields might provide us with valuable insights into the process of European integration.

### Expectations

As previously shown, both institutional and intermediary arenas can influence each other and the strength of their connections differs across countries and media systems, according to the levels of political parallelism. Thus, the salience and tone of parliamentary speeches and newspaper articles tend to be more similar the greater the political parallelism in each country. Following the revisited model proposed by Brüggemann et al. (2014), we expect a higher resemblance between parliamentary speeches and media coverage in Greece and Spain followed by the remaining countries.

The institutional and intermediary arenas of communication have important differences, but they also can influence each other mutually. Not only are they constituted by distinct actors, but they encompass unique working logics. Parliaments, and more specifically parliamentary debates, answer both representative and deliberative democracy concerns. As Auel and Raunio note (2014, p. 13): "debates are vital elements of electoral competition as they provide for a public articulation of societal interests and the discussions of policies, thus informing citizens about complex political issues". Furthermore, debates are intimately linked with the legislative process and its outputs (Fernandes et al., 2021). Therefore, while both polity and policy-related issues might be addressed in the parliamentary arena, we expect a higher proportion of parliamentary speeches about the EU to be about policy-related issues.

In turn, the media is the most important source of information about politics (Bennett & Entman, 2000; Koopmans & Statham, 2010), becoming "the central means by which individuals are connected to the wider social and political world" (Hallin & Mancini, 2004, pp. 33–34), with the extent to which the EU or any other topic is addressed in the media depending on its newsworthiness (Eilders, 2006; Zoch & Supa, 2014). The existence of conflict is one of the most important factors determining the newsworthiness of each event (Galtung & Ruge, 1965; Harcup & O'Neill, 2017; Lippmann, 1922). As Hutter et al. (2016b) note, the conflict (polarization) between parties in respect of the EU is greater on polity-related issues than it is on those that are policy-related. Moreover, by examining media statements, they show that polity-related issues are in general twice as salient in general debates about Europe than are debates on European policies, leading to a media "polity-bias" (Galpin & Trenz, 2018). Therefore, and contrary to our expectations for parliamentary debates, we expect polity-related issues to lead the debate in the print media.

Empirical research has also shown there to be a negative bias in political news coverage (Farnsworth & Lichter, 2011; Haselmayer et al., 2017; Soroka, 2014). In other words, negativity determines what events are worthy of media coverage. In fact, during electoral campaigns, candidates using a more negative tone receive greater media coverage (Haselmayer et al., 2017; Maier & Nai, 2020). This bias also applies to the specific case of elections to the EP in some countries (de Vreese et al., 2006), including in the UK in 2014 (Galpin & Trenz, 2018), where the success of UKIP and the 2016 Brexit referendum were due to both polity and negative media bias (Galpin & Trenz, 2018). Taken together, this research leads us to expect that, generally speaking, a greater share of newspaper articles will have a negative tone when compared with parliamentary speeches. Nonetheless, there is another important dynamic at play. As Wendler (2016) notes, the tone with which the EU is addressed in national parliaments is determined mainly by the existence of Eurosceptic parties. Therefore, in countries in which Eurosceptic parties have a large parliamentary presence, such as Greece, we might not see a clear difference in the negative tone between arenas since we would also observe a higher level of EU contestation in parliament.

Finally, in addition to the distinction between European polity and policy, we must consider the many areas within which the EU has the competence to legislate. Despite the numerous issues in which the EU has exclusive, shared or supporting competencies, economic and financialrelated issues have always been the EU's central theme. In fact, the EU began as a project for economic integration that sought to bring the benefits of scale, internal efficiency and robustness to the EU economy as a whole and to the economies of each individual Member State. The later decision to form the Economic and Monetary Union was a major and further step in the development of the integration process. Therefore, we expect those issues to be more salient in the political debates whenever policies are addressed, regardless of whether we consider the media or parliaments.

### Methods and Data

To test our expectations, we considered both media content (newspapers) and parliamentary activity (parliamentary debates) in Belgium, Germany, Greece, Ireland, Portugal and Spain between 2002 and 2017.

The content of newspaper articles and parliamentary debates was analysed using a combination of automated methods—a dictionary-based approach—and manual coding. In contrast with the previous chapters that relied on automated analyses to assess politicization patterns in both the mainstream media and parliamentary debates (see Chapters 2 and 3), here we resort mainly to in-depth manual content analysis. This approach allows us to delve into the content of both newspaper articles and parliamentary debates to disentangle the different European dimensions of interest.

This chapter makes use of the MAPLE dataset collection for media and parliamentary debates, as described in the previous two chapters. From the collected dataset, 10,516 parliamentary debates and 22,618 newspaper articles were manually coded. Considering the comparative purpose of this chapter, all parliamentary speeches in the year preceding each legislative election were included, as were all newspaper articles during the 30 days before the same election. This approach allowed us to address two important aspects: (1) to capture a period of heightened party competition and (2) to have a meaningful number of speeches and articles to analyse. Electoral campaign periods offer a "window of opportunity" into a period of conflict between political parties (Hutter & Grande, 2014; Hutter & Kriesi, 2019b) that heightens as the election draws nearer. Simultaneously, non-political actors might try to put forward the issues they most care about in an attempt to influence electoral choices or the political positions of the parties. However, while the campaign period is suitable for media analyses, it does not provide enough parliamentary speeches for our purposes. Consequently, we also analysed all parliamentary speeches made during the 12 months prior to each legislative election.

The methodological strategy followed a two-step process. First, we proceeded with a keyword approach for identifying newspaper articles/parliamentary speeches mentioning the EU. The extensive list of EU-related strings is the same as that used by Silva et al. (2021),1 which was adapted from the codebook developed by Maier et al. (2014) and translated into seven languages. Second, a random sample of 10,516 parliamentary speeches and 22,616 newspaper articles mentioning the EU were coded manually by a team of native speakers. Our unit of analysis was each newspaper article and parliamentary speech. Coders were asked to identify several features within each article/debate, such as whether the EU was the main topic or if it was simply mentioned, the European dimensions addressed, the main topic, the tone and the actor/organization the addressee is affiliated with, among others.2 All coders received proper training, were monitored throughout the process and only began coding after achieving high levels of inter-coder reliability.

We start by comparing the politicization of the EU across arenas by focusing on its most crucial element: salience. Then, as previously discussed, different European dimensions are considered. Following the typology proposed by Hurrelmann et al. (2015), we distinguish four EU dimensions: (1) membership; (2) constitutional structure; (3) EU policies and (4) domestic policies (see Table 4.1). While the former two are EU polity-related, the latter two are policy-related. While this typology is undoubtedly useful, we are aware that certain issues may fall within constitutional and policy domains simultaneously. For instance, the debate on the collectivization of debt and the creation of Eurobonds which occurred during the Eurozone crisis period can be seen as a combination of the two types of dimensions.

In addition to the distinction made between the European polity and its policies, we identify the category of policies debated in the speeches and discussed in media articles about the EU. Since some policies are mentioned only very residually, we present only those that are most salient in each arena.

Finally, since "Economy and Work" and "Financial and Taxes" policies are the most salient, we look to the tone associated with them in both arenas. In fact, assessing the tone is as important as the salience of the different dimensions of the EU, and is needed to test our expectation that tone will be more negative in the media than it is in parliamentary

<sup>1</sup> The list of EU-related terms used to identify EU speeches can be found in Appendix 4.1.

<sup>2</sup> The codebook for the parliamentary debates and media articles is available in Appendix 4.2.

**Table 4.1** Operationalization of the different dimensions of the European Union


debates. Indeed, existing literature has focused on the tone (see de Vreese et al., 2006; Hobolt & Tilley, 2014; Silva et al., 2021; Nina, 2022 for other studies assessing tone). Depending on its overall valence towards the EU integration process, each article/speech about the EU was coded as either positive, negative, neutral or mixed/balanced, when it included both positive and negative claims.

### Results

Assessing the salience of the EU's multiple dimensions, results suggest policy-related issues are the most salient topics in both the media and parliaments. Yet, despite the overall residual importance of polity-related issues, it seems that in the media of some countries (Greece, Spain, Germany and Ireland), matters related to membership received a great deal of attention during some elections. While in Greece this might reveal a discussion around its membership in the June 2012 election as it nearly

<sup>3</sup> Leaving the eurozone has been coded as "membership" rather than policy because the treaties have not yet stipulated a formal way to opt-out of eurozone membership that does not involve leaving the EU.

**Fig. 4.1** Salience of different European dimensions in the media4

left the EU around 2015, in Spain, Ireland and Germany (elections in 2016 and 2017) the debate on membership was highly influenced by Brexit. Despite polity-related issues being of minor importance, whenever they became salient they invariably occurred in the media and never in parliaments (Figs. 4.1 and 4.2).

When it comes to the policies discussed in media and parliaments, the results confirm issues related to economic and financial matters are those most addressed in both. We opted to group European and domestic policies to show the trends in policy salience across arenas, as results do not differ dramatically if we disaggregate them. Since the EU has exclusive competence in the customs union, the establishment of competition rules, monetary policy, common commercial policy and increasingly shared competencies in related economic and financial areas, we expected economic and financial issues would lead the debate when policies are addressed, regardless of the arena considered. Moreover, the eurozone crisis has drawn the EU's attention to economic and financial-related matters, with further economic integration occurring during this period. Parliaments and the media have reflected this trend, so we can confirm the

<sup>4</sup> All graphs presented in this chapter pertain to the following legislative elections:

Greece—2004, 2007, 2009, 2012, 2015 (both elections).

Spain—2004, 2008, 2011, 2015, 2016.

Portugal—2002, 2005, 2009, 2011, 2015.

Germany—2002, 2005, 2009, 2013, 2017.

Belgium—2003, 2007, 2010, 2014.

Ireland—2002, 2007, 2011, 2016.

**Fig. 4.2** Salience of different European dimensions in parliaments

**Fig. 4.3** Salience of European and domestic policies in the media

salience of those issues has also increased in both arenas in every country, but particularly in Greece, Spain and Ireland.

Besides economic and financial-related issues, there is no other single topic with similar prominence in either parliaments or the media. However, the German Bundestag emerges as a unique case since "Foreign policy" is the most addressed policy up until 2014. If we consider both "Economy and Work" and "Finances and Taxes" as related matters, except in Germany, there are no significant differences between either when it comes to the most addressed policies (Figs. 4.3 and 4.4).

**Fig. 4.4** Salience of European and domesticated policies in parliaments

As established above, assessing the tone is as important as assessing the salience of the different dimensions of the EU. Hence, we consider how European and domestic policies were addressed in both arenas (i.e. the tone). In parliaments, apart from Greece, the share of negative parliamentary speeches on policies becomes more residual over time. In the Greek parliament, we discern a very negative stance on European and domestic policies, which might be explained by two different factors. On the one hand, Eurosceptic parties have an important share of seats in the Greek parliament from 2012 on with the success of Syriza. As Wendler (2016) noted, the stronger Eurosceptic parties are electorally the greater is the contestation towards the EU in parliament. On the other hand, since the dominant topic in the Greek parliament was related to "Economy and Work" related issues, and considering how severe the eurozone crisis was felt in Greece, Greek parties would certainly be the most critical of some of the European policies after 2009.

A slightly different picture emerges in the media. Overall, and with the exception of Greece, the share of negative articles is considerably higher than the share of negative parliamentary speeches in both policies. Unlike those in the other countries, the Greek parliament is much more critical of European policies than its media. However, important differences emerge between countries and over time. This is in line with our expectations and with previous research, which has shown political news coverage to have a negative bias (Farnsworth & Lichter, 2011; Haselmayer et al., 2017; Soroka, 2014), leading to more negative reports in comparison with party communication (Walter & Vliegenthart, 2010) (Figs. 4.5 and 4.6).

Despite parliaments and the media having very different characteristics and working logics, they have a mutual influence over each other. Therefore, the way the EU is addressed in both arenas might be more or less similar depending on strength of their connections.

Considering the salience of both polity and policy-related issues, a similar picture emerges in the two arenas. More concretely, policy-related

**Fig. 4.5** Tone on European and domestic policies in the media

**Fig. 4.6** Tone on European and domestic policies in parliaments

issues (European or domestic) are the most salient topics in both the media and parliaments. A different scenario emerges when we delve into what policies are addressed. Indeed, the prominence of economic and financial-related issues is much greater in the media than it is in parliaments. Nevertheless, it seems the gap between the media and parliaments on these matters may be narrowing as parliaments have also been paying more attention to economic and financial-related matters in the wake of the eurozone crisis. In sum, the more specific we get on what European issues are addressed in parliaments and the media, the greater the difference between both arenas.

Nonetheless, it is when we assess the tone that we discern the greatest differences between parliaments and the media. With the exception of Greece, the share of negative utterances is much greater in the media than it is in parliaments, confirming the negative bias in political news coverage previously established in the literature. However, there are important differences between countries over time and even between the policies considered. No other clear differences between arenas can be identified beside the important negative bias in the media.

Finally, while we expected greater resemblance between parliamentary speeches and newspaper articles in Spain and Greece, as past literature classifies both have high levels of political parallelism, we do not have any evidence supporting that expectation in respect of the EU. Instead, our results suggest there are important differences in how parliaments and the media politicise the EU, particularly when considering the salience of the different European dimensions. In fact, assessing Tables 4.2 and 4.3, we see a higher number of negative correlations when we consider the salience vis-à-vis tone: in other words, regarding salience, we see that in several dimensions, parliaments and the media are incongruous. The only country in which both arenas are congruent across every dimension is Ireland. Focusing on tone, we also observe some negative correlations, even if they are weaker. Moreover, in relation to tone, parliaments and the media are congruent in three different countries—Portugal, Spain and Ireland. Therefore, taking everything into account, Ireland emerges as the country in which both arenas seem to go hand-in-hand on most occasions, at least when the EU is the issue in question.


**Table 4.2** Correlation between the salience in parliaments and media by dimension

**Table 4.3** Correlation between the tone in parliaments and media


# Conclusion

Current research has assessed the politicization of the EU in national parliaments and the media; however, since each arena has a distinct purpose, audience and actors conveying their political positions, different patterns of politicization might emerge. Up until this point, the literature has not addressed this phenomenon by comparing the arenas in which the EU has been mostly politicized. In this chapter, we sought to fill this lacuna with a multidimensional study on the politicization of the EU, one that uncovers the salience and tone in both the media and parliamentary arenas by exploring the impact the eurozone crisis had on the European debate. This comparative endeavour is guided by the expectation that both arenas are considerably different, and acknowledging those differences is crucial to improving our knowledge of the EU's politicization patterns.

First, we expected policy-related issues to be more salient in the parliamentary setting, while polity-related ones would emerge as the main European topic in the media. This chapter's results only support the first half of this proposition, since policy-related issues were in general addressed more in both parliaments and the media. This cross-arena focus on policies may suggest the increasing role played by Europe in national policy-making, and also that the country's EU membership is not contested. To that extent, the focus on policies signals a deepening of European integration through national parliaments and media. Nevertheless, a more grim interpretation might also follow: if the European electorate falls short of knowing and having a meaningful say about the membership, competencies and institutional design of the EU, then its democratic character might not be entirely fulfilled.

Second, as economic and financial-related issues have been the EU's central theme since its foundation, we expected those policies to rate higher in political debate. By disentangling what policies were being addressed over time, the results confirm initial expectations since "Economy and Work" and "Finances and Taxes" are the most addressed policies in both the media and in parliaments. Moreover, the prominence of one of the two policies increased significantly in some countries immediately after the eurozone crisis and regardless of the arena considered, although this impact has faded over time. Besides economic and financialrelated issues, there is not any other single topic with similar prominence in either parliaments or the media. The only exception is the German parliament, where the "Foreign policy" issue was the most addressed up until 2013, where it was considered much more often than anywhere else.

Third, political parties have been addressing economic and financialrelated policies in a rather mixed or neutral tone. Greece emerges as an exception, since the share of negative parliamentary speeches is in general quite high, as was to be expected given the importance Eurosceptic parties have in the Greek parliament. In turn, the share of negative mentions in the media is considerably higher compared with mentions in parliaments and was even the dominant tone in some periods. These results support the expected and previously established negative bias in political news coverage. The eurozone crisis also impacted the way these policies were addressed, but only in the media. While the share of negative mentions does not seem to have changed in parliaments, it has definitely changed within the media, although with cross-regional differences.

Finally, the main goal of this chapter was to compare EU politicization patterns in the media and in parliaments. The results are mixed and suggest different pictures emerge depending on the element being focused on. In general, the deeper and more specific our focus gets, the greater the differences. While the balance between polity and policyrelated issues is quite similar in both arenas, it gets a little different when we focus on which—and to what extent—different policies are addressed. Considerable differences also emerge when tone is assessed: the share of negative mentions is much higher in the media, confirming a negativity bias in media reporting. Moreover, we did not find the expected differences between countries based on different levels of political parallelism. In fact, considering both salience and tone, it is in Ireland where the media and parliament seem most congruent, followed by Spain and Portugal.

The results of this chapter demonstrate that considering the politicization of the EU in its multidimensional aspects is crucial to better understand this phenomenon and its potential consequences. Much more focus should be placed on the arena in which the politicization process occurs. Since the media and parliaments are very different institutions with singular features and their own working logics, and while they emphasize the same European dimensions to a similar extent, they address policies quite differently. Future research should try to address the reasons for this. Are the same political actors conveying different messages in different arenas? Or is the media drifting away from the prevailing consensus among Europe's political elites? Responses to these questions will help us better understand the politicization of and public attitudes towards the EU.

# Appendix 4.1

See (Table 4.4).


**Table 4.4** List of the base/root words used to identify the articles that mention EU

# Appendix 4.2

#### MAPLE's Parliament Codebook:

Unit of analysis.

. Speeches by budget plenary session ordered

#### **ELIGIBILITY Is the Speech a substantial intervention to the debate?**


If the Speech is Too short, or it's an interpellation/intervention/interruption, or it's the president's speech (concerning: Procedural issues, Rules, Calls to order).

Filter: If **ELIGIBILITY = 0** Stop coding here

#### **EUSALIENCE—How salient/important is the EU (EU as a topic or EU-related actors/institutions/organizations) in the speech?**

0. EU is not mentioned in the speech 1. EU, EU institutions or EU actors are mentioned but the speech is mainly about something else. 2. EU, EU institutions or EU actors are a central topic/aspect in the speech.

#### **TONEEU—Overall valence towards the EU**

Code from the perspective of the EU, i.e., would those who believe in the EU and European integration consider that the EU is evaluated positively or negatively?


Is the EU itself evaluated/portrayed as something positive or negative? The coding decision should be based on the manifest positive or negative wording on the overall impression of the speech—if the speech presents the EU in a positive or negative way.

Some Examples might be:

Positive—Someone saying we should remain in the EU.

Balanced—Despite the EU causing a lot of problems to our Economy, it would be worse to leave.

Negative—The EU is responsible for the bad economic situation.

#### **EUDIMENS\_MAIN—What is THE MAIN dimension of the European integration that is discussed in the speech?** (Hurrelmann, et al., 2015)


#### **Code 1**—Membership;

This includes discussions regarding the geographical reach of the EU, whether a particular country should be in the EU, the benefits and costs of being an EU/Eurozone member).

**Code 2**—Constitutional structure;

Discussions regarding the objectives and responsibilities of the EU, its institutional arrangement, its institutions and its decisionmaking processes, and the functioning of the EU in general (e.g. stories about how EU institutions work, how their members are chosen/elected, about EU's *democratic deficit*, etc.).

**Code 3**—EU policies;

Articles discussing issues that emerge from EU-level institutions legislative, executive and judiciary—that have implications for all EU member states. These are issues/policies that are currently on the agenda of the EU's legislative, executive or judiciary institutions (e.g. EU data protection law).

**Code 4**—Domesticated policies;

Articles discussing issues in national politics that emerge as an implication of membership, such as budget cuts mandated by Eurozone requirements, etc. (e.g. austerity measures in order to comply with EU deficit or debt limits).

#### **EUDIMENS\_ADD**—**If any, what alternate dimension of the European integration is CLEARLY DISCUSSED/REFERENCED in the speech?**


Filter: Code **MAINTOPIC** only if **EUDIMENS = 3 or 4** 

#### **MAINTOPIC—What is THE MAIN policy area (around the EU, as per your selection earlier) discussed in the speech?**


#### **TOPIC\_ADD**—**If any, what alternate policy areas CLEARLY DISCUSSED/REFERENCED (around the EU, as per your selection earlier) in the speech?**


MAPLE's Media Codebook.

#### **EUSALIENCE**—**How salient/important is the EU (EU as a topic or EU-related actors/institutions/organizations) in the article?**


**FILTER:** 

#### **TONEEU—Overall valence towards the EU**

*Media:* All subsequent variables are only coded if **EUSALIENCE** = 2

Code from the perspective of the EU, i.e., would those who believe in the EU and European integration consider that the EU is evaluated positively or negatively?


Is the EU itself evaluated/portrayed as something positive or negative? The coding decision should be based on the manifest positive or negative wording on the overall impression of the article—if the article presents the EU in a positive or negative way.

Some Examples might be:

Positive—Someone saying we should remain in the EU. Balanced—Despite the EU causing a lot of problems to our Economy, it would be worse to leave.

Negative—The EU is responsible for the bad economic situation.

#### **EUDIMENS—What is the main dimension of the European integration that is discussed in the item?** (Hurrelmann et al., 2015).


**Code 1**—Membership;

This includes discussions regarding the geographical reach of the EU, whether a particular country should be in the EU, and the benefits and costs of being an EU/Eurozone member.

**Code 2**—Constitutional structure;

Discussions regarding the objectives and responsibilities of the EU, its institutional arrangement, its institutions, and its decisionmaking processes, and the functioning of the EU in general (e.g. stories about how EU institutions work, how their members are chosen/elected, about EU's *democratic deficit*, etc.).

**Code 3**—EU policies;

Articles discussing issues that emerge from EU-level institutions legislative, executive and judiciary—that have implications for all EU member states). These are issues/policies that are currently on the agenda of the EU's legislative, executive or judiciary institutions (e.g. EU data protection law).

**Code 4**—domesticated policies;

Articles discussing issues in national politics that emerge as an implication of membership, such as budget cuts mandated by Eurozone requirements, etc. (e.g. austerity measures in order to comply with EU deficit or debt limits).

**FILTER:** All subsequent variables are only coded if **EUDIMENS =3 or EUDIMENS =4** 

#### **MAINTOPIC—What is the main policy area discussed in the article:**


#### References


in Germany and the UK. *European Politics and Society, 20*(3), 260–276. https://doi.org/10.1080/23745118.2018.1501899


**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# Experimental Evidence of EU Issue Voting

#### *Roberto Pannico and Marina Costa Lobo*

In the previous chapters, the politicisation of the EU at both the media and parliamentary levels was established in the six EU countries which form part of the study, for both debtor (Greece, Ireland, Portugal and Spain) and creditor countries (Belgium and Germany) during the Eurozone crisis. Having established the degree to which the EU has emerged as a growing issue both at the level of mainstream media and parliamentary debates across Europe since the onset of the crisis, in the remainder of the book we turn to the *consequences* which politicisation may have for voting in legislative elections.

In this chapter, we combine data from different sources. First, we test the relevance of EU issue voting in six EU countries using an online survey experiment. This design allows us to cope with possible endogeneity problems. Second, we investigate how the politicization of the

M. Costa Lobo

© The Author(s) 2023

115

R. Pannico (B)

Universitat Autònoma de Barcelona, Barcelona, Spain e-mail: roberto.pannico@uab.cat

Institute of Social Sciences of the University of Lisbon, Lisbon, Portugal e-mail: marina.costalobo@ics.ulisboa.pt

M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_5

EU affects electoral behaviour. To this aim, we integrate measurements of EU politicisation in the analyses of the experiment's results. Differently from previous chapters, where the data was longitudinal and covered the whole period since the creation of EMU (2002–2017), we perform a synchronous analysis using data from 2019. Moreover, in this chapter, we consider the two components of politicisation (salience and tone) separately, since they may have different effects on EU issue voting.

The chapter is organized in the following way. We first present succinctly the extant literature on EU issue voting and its conditionality. Second, we justify the choice of the experimental design. Third, we present the data and the conjoint design. Fourth, we analyse the results and discuss the implications of the findings for the importance of EU issue voting in national elections.

#### EU Issue Voting in Legislative Elections

Citizens can use elections to select political representatives that share their political views. Through this selection mechanism, voters provide democratic mandates to politicians and parties for implementing specific policy platforms. Such a mandate is fundamental for democratic representation because it creates a connection between policy outputs and voters' policy positions. In other words, when citizens choose representatives with whom they share policy preferences, the latter can legitimately pursue their policy agenda (De Vries & Hobolt, 2016). The extent to which this legitimacy mechanism works for EU preferences and EU policies seems to have changed over time.

The classical debate about the democratic deficit in the European Union has often stressed the lack of such a link between voters' and parties' EU preferences (Follesdal & Hix, 2006; Hix, 2008). According to this view, EU citizens do not have the possibility to select representatives that share their EU positions, given that no political contest incorporates EU issues. Indeed, the elections for the EU Parliament are considered "second-order" elections that are treated by parties as a test for their activity at the national level (Reif & Schmitt, 1980). Similarly, national elections fail to create a link between voters' and parties EU preferences because EU issues are rarely discussed during electoral campaigns. Consequently, EU integration remains the "sleeping giant" of the national political debate (Van der Eijk & Franklin, 2004).

In more recent times, however, this perspective has been partially challenged by studies showing that citizens may consider their EU preferences when deciding their vote at the national level. At least under certain circumstances, there is evidence of what has been labelled "EU issue voting" (De Vries, 2007).1 Using data from the European Elections Study in 2009, De Vries and Hobolt (2016) have shown that in the vast majority of EU member states citizens' likelihood to vote for a specific party in national elections decreases when the distance between them and the party on the EU issue increases. Similarly, De Vries and Hobolt (2012) have found that voters' likelihood to vote for a challenger party in the national elections is strongly and significantly affected by their attitudes towards European integration. Moreover, according to De Vries and Tillman (2011), EU attitudes have an effect on the national vote in both East-Central and Western Europe, even though EU issue voting seems to be larger in the former case.

However, the literature has also highlighted that the presence of EU issue voting cannot be taken for granted and seems to depend on contextual factors. In particular, EU issue voting seems to depend on the political competition over EU issues and on the availability of EU information. De Vries (2007), for example, has shown that voter-party proximity on the EU dimension affects electoral choices at the national level only when citizens perceive EU issues as important and political parties have different positions on them. Similarly, De Vries (2009) found that in the Netherlands the 2005 EU referendum increased both the salience of EU issues and the inter-party conflict over the EU, resulting in a more likely EU issue voting in the subsequent national elections. Focusing on European elections, De Vries, Van der Brug et al. (2011) showed that EU issue voting is stronger in those contexts where EUrelated information is more available, while Hobolt and Spoon (2012) highlighted that citizens are more likely to base their vote on EU considerations when the degree of pollicization of the EU in the national

<sup>1</sup> The literature provides evidence of EU issue voting also in the elections for the European Parliament (e.g. De Vries, Van der Brug et al., 2011; Hobolt & Spoon, 2012). However, given the object of this book, we focus on vote in national elections. The mandate given by citizens in national elections represents, as stressed by De Vries and Hobolt (2016), the indirect path of representation provided to voters to select representatives that share their EU positions. Through the selection of national parliaments and, in turn, national governments, citizens select their representatives in intergovernmental institutions of the EU (e.g. the Council and the European Council).

political arena is higher. Similarly, De Vries and Hobolt (2016) found that EU issue voting in European elections is stronger when EU issues receive more attention in the media. Finally, Hobolt (2005) and Beach et al. (2018) showed that when the campaign is more intense voters are more likely to rely on EU attitudes to vote in EU referendums or European elections.

In sum, previous literature has highlighted that, at least under certain circumstances, party positions on EU issues are congruent with the positions of their voters. This congruence is good news for the legitimacy of European Union activity, to the extent, it is due to EU issue voting. If the proximity between voters' and parties' positions depends on citizens voting for parties that share their view on EU issues, EU policies are more likely to match voters' preferences. However, as we explain in the next section, party-voter congruence on EU issues might also be due to alternative mechanisms.

#### Experimental Evidence of EU Issue Voting

The present chapter contributes to the literature that analyses EU issue voting. Differently from the most part of the previous works, our study uses experimental data to assess the effect of EU attitudes on the vote. The need for experimental studies derives from the fact that the literature on party cues has established that voters' positions on EU issues are partially shaped by the positions of their preferred party (Steenbergen et al., 2007; Stoeckel & Kuhn, 2018; Torcal et al., 2018). Given the complexity of the EU political system and their general lack of information about it, voters rely on cues from more informed political actors to develop opinions on EU issues. In other words, voters tend to align their positions on EU issues with those of their party (Pannico, 2017, 2020). This means that the congruence between voters' and parties' positions on the EU can be both the cause and the consequence of citizens' vote preferences. It can be due to voters selecting parties based on their EU positions or to citizens aligning to party stances to cope with the complexity of EU politics. Only the former represents a selection mechanism that provides a democratic mandate to politicians for their EU activity. However, observational data are not suited to cope with these endogeneity problems. The experimental manipulation of parties' positions, on the contrary, allows us to analyse EU issue voting net of the effect of party cues.

To the best of our knowledge, there are only two studies that use an experimental design to analyse EU issue voting. Hobolt and Rodon (2020) run a conjoint experiment in the UK in 2017 to test the relative impact of the EU dimension and the left–right dimension on vote choice. They present the respondents with pairs of candidate profiles formed by eight attributes and ask them to vote for one of the candidates. Their results show that participants consider the proximity between their own positions on the EU and the candidates' ones when making their choice. Moreover, in this experimental context, the impact of the EU dimension on vote appears to be greater than the impact of the left–right dimension. Using a similar design, Schneider (2019) runs a conjoint experiment in Germany in 2016 to look at the effect of attitudes towards the EU and attitudes towards specific EU policies (i.e. EU refugee policy and EU financial aid to Greece) on the vote. She presents respondents with pairs of hypothetical politicians' profiles composed of six attributes. The results of this experiment show that the congruence between candidates' and voters' positions on the EU integration matters only for the vote of Eurosceptic respondents, while party-voter congruence on EU specific policies matters for both Eurosceptic and pro-EU participants.

The experiment we present in this chapter follows the same approach as previous experimental studies on EU issue voting, but at the same time aims to overcome some of their limitations. First of all, it goes beyond single-country cases, given it was run simultaneously in six EU countries: Belgium, Germany, Greece, Ireland, Portugal and Spain. This multi-country design allows us to validate our results beyond context (Bruter, 2009), increasing the external validity of the experiment. In this regard, it is particularly relevant the fact that our case selection includes countries with different economic paths within the Eurozone.

Secondly, different from previous experimental works, our design exposes respondents to a clear and comprehensive range of party positions on the EU. The profiles used by Hobolt and Rodon (2020) in their experiment in the UK include the candidates' position on the 2016 Brexit referendum (Leaving the EU/Remaining in the EU). This operationalization of the candidate position on the EU is strongly related to the British context and is hardly replicable in countries where a referendum on the EU membership has never been held. Moreover, being framed as a dichotomy between the status quo and Brexit, the operationalization disregards the cases in which the candidate and/or the respondent would prefer a deeper integration of the country in the EU. At the same time,

in the experiment run by Schneider (2019) candidates' positions on EU integration are not part of the profiles (i.e. it is not indicated if they are pro or against the EU). The author assumes that their position on EUspecific policies also indicate to the respondents the candidates' position on the EU integration (i.e. the author assumes that candidates that favour more immigration or an additional bailout of Greece are seen as pro-EU candidates by respondents). On the contrary, our experimental design incorporates all possible EU party positions and clearly assigns them to the parties presented to participants.

#### Data and Experimental Design

The conjoint experiment was embedded in an online survey run in February–March 2019 in six European Union member states: Belgium (*n*  = 3076), Germany (*n* = 2568), Greece (*n* = 1507), Ireland (*n* = 1514), Portugal (*n* = 2049) and Spain (*n* = 2026). For each of the countries, the sample was recruited through a crossed-quota design based on the 2011 census data.2 The survey included different experimental manipulations. For this chapter, we only rely on respondents that participated in the conjoint experiment without being exposed to any other previous manipulation. Their number corresponds to approximately one-sixth of the original national samples.3 Given that the respondents' membership in these sub-groups has been randomly assigned, the final sub-samples are in principle as representative as the original national samples.

Participants were presented with the profiles of two hypothetical parties ("Party A" and "Party B") that were defined by a set of six party attributes with independently randomly assigned categories. Respondents were asked to indicate for which of the two parties they would vote in a hypothetical general election. The experiment had two rounds, meaning that after making their first vote choice, respondents were presented with

2 The crossed-quota design was defined by gender (two categories), age (three categories: 18–34; 35–55; 55+) and education (three categories: less than secondary, secondary and more than secondary). The perfect fulfilment of all crossed quotas was not possible in all countries. However, in most cases the final national samples were representative in terms of gender and age, while higher educated population were over-represented at the expense of lower educated population.

3 Belgium <sup>=</sup> 495, Germany <sup>=</sup> 411, Greece <sup>=</sup> 242, Ireland <sup>=</sup> 252, Portugal <sup>=</sup> 338, Spain = 347.

other two randomly defined party profiles and again asked to vote for one of them.

The party profiles were formed by the attributes shown in Table 5.1. The table also provides, for each attribute, the different categories it could assume within the party profiles, and the corresponding text (in English).4 Given the aim of this chapter, the most relevant attribute is the one that indicates the party's "EU position". It was randomized over three categories: pro-integration, neutral, and against integration. The table shows that these categories refer to the fact that the party supports, respectively, a deepening of the EU integration, the status quo, or a reversal of the EU integration. As shown in the next section, these party positions on the EU will be matched with the positions of the respondents to assess the effect of the EU issue on the vote.

The profiles were composed of five additional party attributes. The party leader attribute was randomized over three categories that indicate different levels of leader's experience in public governance. The ideology attribute refers to the party position on the left–right axis, while the "Economic performance" attribute informs the respondent about the trend of the economy during the party's last term in office. Finally, the immigration and corruption position attributes refer to the party's stance on these issues and were included in the party profiles because they were relevant political issues in several EU countries at the time of the fieldwork.

Within each party profile, the categories of the attributes have been completely randomized. This means that any combination was possible. Respondents may have been exposed to party profiles with unusual combinations of characteristics (e.g. a party from the right that supports a more open immigration policy). However, we opted for full randomization to break the correlation that exists in the real world between party characteristics and isolate the effect of each of them on the vote (Hainmueller et al., 2014). Finally, the order of the attributes in the party profiles have been randomized between respondents, but kept constant between rounds for the same respondent.

The use of this conjoint design was motivated by the need to maximize both the internal and the external validity of our study. The random assignment of party EU positions allows us to estimate the effect of respondents' EU opinions on the vote accounting for potential sources

<sup>4</sup> The Appendix provides an example of conjoint round, as well as the specific formulation for the vote choice question.

#### 122 R. PANNICO AND M. COSTA LOBO


**Table 5.1** Attributes, categories and text for the conjoint experiment

of endogeneity (i.e. the effect of party cues on voters' opinions). At the same time, the complexity of the party profiles and the manipulation of several attributes provide respondents with different elements on which to base their vote, approximating the task voters face in the real world (Sigelman et al., 1991).5 Moreover, the experiment presents participants with two party profiles and asks them to vote for one of them. This design simulates the choice respondents face on the ballot paper, where they have to choose one of the candidates or parties that compete in the election.

#### Results

To assess the effect of the different attributes on respondents' vote, we created a stacked dataset consisting of four observations for each respondent, one for each party profile s/he saw in the two rounds of the conjoint. Using this dataset, we estimate the following model:

$$\begin{split} V\_{ijk} &= \beta 1\* Leader\_{ijk} + \beta 2\* Ideology\_{ijk} \\ &+ \beta 3\* Econ.Performance\_{ijk} + \beta 4\* EUDistanence\_{ijk} \\ &+ \beta 5\* Imm.Position\_{ijk} + \beta 6\* Corr.Position\_{ijk} + e\_{ijk} \end{split} \tag{5.1}$$

In other words, we use the party's attributes to predict the vote of participant *i* for party *k* in round *j*. Given that the categories of the party's attributes have been randomly assigned, the average marginal component effects (AMCEs) can be estimated using linear regression with standard errors clustered by the respondent (Hainmueller et al., 2014). The pooled model also includes country dummies.

As shown in Eq. (5.1), the model does not include the "EU position" attribute in its natural form. We are not interested in whether voters are more likely to vote for Eurosceptic, pro-integration or neutral parties. To provide evidence of EU issue voting, the model should show that

5 The cost of this approach, however, is that we are forced to rely on hypothetical parties for the experiment. If we had used real party labels, the random manipulation of party characteristics and policy positions, including EU positions, would have probably been ineffective. Because of party reputation, presenting respondents with real parties with positions different from the ones they have in the real world, would have jeopardized the credibility of our treatment. Given that at least part of the effect of partisanship is likely to be captured by the ideology attribute of the party profiles, we considered that the experiment's loss in terms of external validity caused by the use of hypothetical parties was lower than the gain generated by providing respondents with multiple decision elements. the congruence between parties' and voters' positions on the EU has an effect on citizens' electoral choices. In other words, respondents should be more likely to vote for a party that has an EU position similar to theirs than for a party that does not. To test this possibility, we recoded the "EU position" attribute so that it would indicate the proximity between the respondent's and the party's position on the EU. To collect respondents' positions on the integration process, we used the following question included in the survey before the experiment: "Some people believe that the process of European integration should move forward to the creation of the United States of Europe. Others believe that the European Union should be dissolved in order to return to a situation in which states are fully sovereign. In which point of the following scale would you place yourself?". To provide an answer, participants used a scale ranging from 0 "The EU should be dissolved" to 10 "The EU should move towards the United States of Europe". We coded as "Pro-integration" respondents the participants that placed themselves in the 7–10 range, while participants in the 4–6 and 0–3 ranges were coded as "Neutral" and "Against integration", respectively. Finally, matching the party position with the respondent one, we created the "EU distance" party attribute, consisting of the following three categories: "More pro-EU than the respondent", "Same position", "More Eurosceptic than the respondent". Capturing party-voter congruence, this party attribute is more suitable for the estimation of EU issue voting than the "EU position" attribute. For this reason, the former rather than the latter is included in the model.

Figure 5.1 plots the results from the model for the pooled sample.<sup>6</sup> The coefficients in the graph can be interpreted as the average change in the probability that the party profile will be chosen by the respondent when it includes the listed characteristic value instead of the baseline characteristic value. For example, the graph shows that, keeping all other attributes constant, advocating for a more open immigration policy decreases by approximately 7 percentage points the probability of the party profile being chosen by the respondent, compared to the case in which it advocates for a more restrictive policy.

6 The first column of Table 5.2 in the Appendix provides the detailed results. We excluded from all the analyses the 5 percent of respondents that spent the least amount of time on each round of the conjoint. Our aim is to restrict the analysis to respondents that actually took the time to accomplish the task, rather than simply randomly choose one of the party profiles.

**Fig. 5.1** The predictors of the vote in the pooled sample

The coefficients that refer to the "EU distance" attribute reveal two interesting characteristics of respondents' votes. First of all, the figure shows that participants do take into account their attitudes towards the EU when making electoral choices. When the EU position of the party does not match the position of the respondent, either because it is more Eurosceptic or because it is more pro-EU, the participant is less likely to vote for the party, compared with the situation where the respondent and the party have the same EU position. In other words, the graph shows evidence of EU issue voting. Given that parties' EU positions were randomly assigned, our experiment excludes by design potential sources of endogeneity. Differently from the case of observational results, we can confidently state that the proximity between parties' and voters' positions is the cause, not the consequence, of respondents' electoral choice.

Secondly, the results also show that EU issue voting is somehow asymmetric. Parties that are more Eurosceptic than the respondent are punished more than parties that are more pro-EU. The decrease in the probability to be voted is 8 percent points in the former case and 5 percentage points in the latter one. A Wald test for the equality of the two coefficient produces a *F* statistic of 4.21 and a *p* value of 0.04, confirming that the two effects are actually different from each other. Therefore, not all types of incongruence between parties' and voters' positions are punished to the same extent.

The disaggregation of the results by country provides a finer-grained picture of EU issue voting in our sample. Figure 5.2 plots the "EU distance" coefficients for each country, while columns 2 to 7 of Table 5.2 in the Appendix provide the full models. In four of the countries, EU issue voting seems to follow the same general pattern we identified in the whole sample. In Germany, Greece, Ireland and Portugal, participants take into account EU positions when making their electoral choices. Moreover, the electoral punishment for parties that are more Eurosceptic than the respondent seems to be larger than the punishment for parties that are more pro-EU than the participant. This is clearer in Greece, Ireland and Portugal, where there is no significant punishment in the latter case. In Germany, the results also point towards the same asymmetry in EU issue voting. However, in this case, both coefficients are significant, and the Wald test does not allow us to conclude that they are different from each other.

Finally, Belgium and Spain are, to different degrees and for different reasons, particular cases. In Belgium, the results show both the existence of EU issue voting and its asymmetric nature. However, in this country, voters do not significantly distinguish between parties that match their

**Fig. 5.2** EU issue voting in six EU countries

EU preferences and parties that are more Eurosceptic than them. Their electoral punishment only targets parties with a more positive perspective on the EU than their own. Therefore, the asymmetry in Belgium is reversed compared with the other countries. In Spain, voters do not seem to take into account EU positions when choosing a party. Both coefficients are not statistically significant. For Spanish respondents, the congruence between their EU positions and the position of the party did not matter when making their vote choice. This is, therefore, the only country in which we found no evidence of EU issue voting.

These results show that voters partially base their electoral choices on EU attitudes. They are more likely to vote for a party if they share with it a common vision of EU integration. This is true in almost all the countries analysed and when accounting for possible endogeneity problems, as in the case of our design. Therefore, the indirect path of representation in the EU looks effective. In national elections voters tend to select politicians who represent their EU positions within the intergovernmental institutions of the EU. This selection mechanism makes it more likely the implementation of EU policies that match citizens' preferences.<sup>7</sup>

The second step of our analysis is aimed to investigate the effect of contextual factors on EU issue voting. In particular, following previous literature, we are interested in the availability of EU information and in the level of political competition over EU issues. In order to capture part of this contextual variation, we use two measures of EU politicization in the national political debate: the salience and the tone of EU news. In each country, for the 30 days before the fielding of the experiment, the content of all the online articles of two mainstream newspapers has been recorded (see Table 5.3 in the Appendix). For the salience measure, we calculated for each newspaper the percentage of articles "about the EU" on the total published during the 30-day period. An article was considered "about the EU" if its title contains at least one of the keywords listed in Table 5.4, or if its body contains at least two of them (or only one, but repeated at least twice). We then averaged the salience value of the two

<sup>7</sup> We performed two different robustness checks. First, we replicated the pooled and the country models matching respondents' positions not only with the "EU position" party attribute, but also with the "ideology", "immigration" and "corruption" ones. Second, we replicated the pooled and the country models using a different operationalization of the "EU distance" attribute. The results (see the Appendix) show that our conclusions remain valid.

mainstream newspapers at the country level to proxy the EU salience in the general national media environment.

The tone measure refers to the direction of EU discourse. For each article, all the sentences mentioning the words listed in Table 5.4 have been translated into English. For each sentence, a sentiment score has been calculated using the R package "sentimentR". The sentiment score has a negative sign if the sentence has an overall negative tone and a positive sign if the tone is positive. For each article, an average sentiment score has been calculated using all the sentences "about the EU". For each newspaper, an overall tone for the 30 days has been calculated as the percentage of articles with an average negative tone on the total of articles "about the EU". The tone at the national level has been calculated by averaging the values of the two newspapers. Table 5.5 in the Appendix provides the average country values of both EU salience and EU tone.<sup>8</sup>

To explore the relationship between EU politicization and EU issue voting, we plot the national coefficients from Fig. 5.2 against the national values of EU salience and EU tone. The left-hand panel of Fig. 5.3 suggests that the availability of EU information in the national media context is relevant for the role that EU attitudes play in citizens' voting. The figure shows an overall negative relationship, meaning that the higher the salience of the EU, the greater the electoral punishment suffered by the party for having a different position in the EU than the respondent. As might be expected, the relationship is stronger in the case where the party is more Eurosceptic than the respondent than in the opposite situation, but in both cases, the correlation coefficients are significant (–0.61 vs. –0.28). The results are less clear in the case of EU tone (right-hand panel). The coefficients are more dispersed in the figure and it is more difficult to identify a clear relationship. Indeed, the correlation between the EU tone and EU issue voting is strong in the case where the party is more pro-EU than the respondent (–0.79), but very weak in the opposite case (–0.03).

<sup>8</sup> It is worth noting that we assess the moderating role of EU salience and tone for the whole population, and not for mainstream newspapers' readership only. In this sense, as stressed, the content of mainstream newspapers is used as a proxy for the national information environment. This strategy minimizes possible problems of endogeneity that stem from the fact that citizens might choose to read a specific newspaper because of its (EU) content.

In sum, the results of our cross-country experiment support previous observational findings. On the one hand, with the notable exception of Spain, we found that in all countries considered voters take their EU positions into account when deciding which party to vote for. On the other hand, our analysis suggests that EU politicization matters for EU issue voting. This is the case first of all when we consider the salience of EU information, while the relevance of EU tone is more doubtful.

### Conclusions

In this chapter, we considered the importance of EU issue voting. Using a conjoint experiment design, we investigated whether citizens use national elections to select representatives who share their views on the EU. Such a selection mechanism would be important for the democratic quality of the EU, because it would create a link between policy output and voters' policy positions.

Our study provided two main findings. First, EU citizens do take into account their attitudes towards the EU when making electoral choices. The fact that these findings were produced using experimental data shows that party-voter congruence on EU issues can be a cause, and not only a consequence of voters' preference for a specific party. We also found evidence of asymmetry in the EU issue voting. Parties that are more Eurosceptic than the respondent are punished more than parties that are more pro-EU. This may suggest that not all types of incongruence between parties' and voters' EU positions are punished to the same extent. These patterns are present in different countries, very heterogenous in terms of trajectories within the EU and Eurozone. With the relevant exception of Spain, we found evidence of EU issue voting in all the countries analysed, being them creditor countries like Germany and Belgium, or bailout ones, like Portugal or Greece. Taken together, these results present a decisive confirmation of EU attitudes as an exogenous factor of voting behaviour.

Following previous literature, we also linked the media context to EU issue voting. The EU salience and EU tone of two mainstream newspapers were used as proxies for the information environment of the different countries. The highlighted relationships suggest a connection between the availability of EU information and the strength of EU issues voting, while the tone of EU news appears less relevant. However, the low number of countries considered does not allow us to draw statistically valid conclusions from this part of the analysis. Further studies considering a larger number of news contexts are needed.

#### Appendix

#### *Table of Contents*

1. Experiment Material

2. Analysis

3. Robustness Checks.

### Experiment Material

#### *Example of Conjoint Round*

On the following screens, you will be presented with two pairs of **hypothetical** party profiles with different characteristics. For each pair of profiles, you will be asked some questions.

[On the following screen] FIRST PAIR OF PROFILES

Please read the characteristics of each party profile (leader, ideology, economic performance, EU position, immigration position, corruption position) carefully in order to answer with precision the following questions.


#### First pair of profile

Taking into account the characteristics of these two parties, if we had general elections in [COUNTRY], which party would you vote for?

1. Party A 2. Party B.

# Analysis


**Table 5.2** The EU issue voting in different countries

5 EXPERIMENTAL EVIDENCE OF EU ISSUE VOTING 133


**Table 5.2** Standard errors in parentheses \**p* < 0.05, \*\**p* < 0.01, \*\*\**p* < 0.001


**Table 5.4** Keywords for newspaper articles about the EU



#### Robustness Checks

#### *Additional Matches*

As first robustness check, we replicated the pooled and the country models matching respondents' positions with the "EU position", "Ideology", "Immigration position" and "Corruption position" party attributes. However, it is worth noting that differently from the "EU position" attribute, for the other three we could not rely on survey items that would allow us to adequately match respondents' and parties' positions.

For the ideology attribute we proceeded in the following way. First, we used respondents' ideology self-placement (0–10 scale) to create five ideology categories: "Left" (0–2); "Centre-left" (3 and 4); "Centre" (5); "Centre-right" (6 and 7); "Right" (8–10). Second, respondents in the "Centre" category were reassigned to either the "Centre-left" or the "Centre-right" category, depending on whether, during the survey, they expressed a vote intention for a leftist or a rightist party. Respondents in the "Centre" category that did not express a vote intention were coded as missing. Finally, the ideology of the party was recoded as "matching the respondent's one" or as being "1", "2" or "3" ideological categories distant from it.

To match the "Immigration position" party attribute to respondents' positions, we used an item in the survey that asked participants "To what extent do you agree or disagree with the following statements: [COUNTRY] should implement a more restrictive immigration policy". The answer options were "Strongly disagree", "Somewhat disagree", "Neither agree nor disagree", "Somewhat agree", "Strongly agree". Respondents who chose the central category could not be matched to the party position.

Finally, to match parties' and respondents' positions on corruption, we used the answers to the following question in the survey: "To what extent do you think corruption is widespread in the following institutions? To answer, use a scale from 0 to 10, where 0 means 'not at all widespread' and 10 means 'extremely widespread'". Participants were asked about corruption in the national government, the national parliament, and the political parties. First, we created an additive index with the answers the respondents gave on the three items. Secondly, we split the sample based on the median value of the additive index.

Figures 5.4 and 5.5 show that our conclusions remain valid also when including these additional matches in the model. The strength of the EU issue voting remains unaltered in the pooled model (Fig. 5.4). The analysis by country (Fig. 5.5) shows significant evidence of EU issue voting (*p* < 0.05) in Germany, Ireland, Portugal and, to a lesser extent, in Belgium (*p* < 0.1). In Greece and Spain, the coefficients are statistically insignificant.

**Fig. 5.4** The predictors of the vote in the pooled sample (with matched attributes)

**Fig. 5.5** EU issue voting in six EU countries (with matched attributes)

#### *Alternative "EU Distance"*

As a second robustness check, we replicated the main analysis using a different operationalization of "EU distance". As in the main analysis, we recoded the survey item on respondents EU attitudes to create the three categories "Pro-integration", "Neutral" and "Against integration". However, this time we operationalized the "EU distance" as the absolute distance of the party position from the respondent's one (from 0 to 2).

Using this operationalization of "EU distance" in the model does not change our conclusions. As it is possible to see in Fig. 5.6, we found strong evidence of EU issue voting in the whole sample. Voters punish the party that is distant from their positions on the EU, whether the distance is considerable or small. Figure 5.7 confirms that at least a considerable distance is consequential in terms of votes in all the countries, with the usual exception of Spain.

**Fig. 5.6** The predictors of the vote in the pooled sample (with alternative EU distance)

**Fig. 5.7** EU issue voting in six EU countries (with alternative EU distance)

#### Bibliography


**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# The Left–Right Dimension, Europe and Voting in Bailout Europe

*Lea Heyne, Marina Costa Lobo, and Roberto Pannico* 

# Introduction

In the previous chapter, an experimental analysis of EU issue voting in the six countries included in this volume, namely Belgium, Germany, Greece, Ireland, Portugal and Spain was implemented. Now, we revisit the topic, but using observational data collected in countries which underwent bailouts during the Eurozone crisis: Greece, Ireland, Portugal and Spain.

M. Costa Lobo e-mail: marina.costalobo@ics.ulisboa.pt

© The Author(s) 2023 M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_6

143

L. Heyne (B) · M. Costa Lobo

Institute of Social Sciences of the University of Lisbon, Lisbon, Portugal e-mail: lea.heyne@ics.ulisboa.pt

R. Pannico Universitat Autònoma de Barcelona, Barcelona, Spain e-mail: roberto.pannico@uab.cat

Despite the fact that there were evident consequences from the onset of the crisis in terms of government incumbency and party system change (Hutter & Kriesi, 2019), less is known about the importance of the EU in relation to the left–right dimension of competition in these countries (for exceptions, see Hobolt & Tilley, 2016; Santana & Rama, 2018), which is the focus of this chapter. Our goals are the following: to establish the importance of the left–right dimension of competition for the vote in Greece, Ireland, Portugal and Spain, and to understand whether the EU issue reinforces or undermines this left–right divide in bailed-out Europe. Further, we test whether the relative importance of this issue is significantly different comparing mainstream and challenger parties. In the final section, we consider the importance of EU integration in the context of parliamentary debates: namely, we integrate the parliamentary debates data analysed in Chapter 3, to test whether EU issue voting impact on likelihood to vote for a party increases when the party is more negative about the EU in parliamentary debates.

The questions we pose will help us understand how the EU relates to the left–right dimension of competition in each of these countries, allowing us to make some generalizations about the contexts which shape voting in post-bailout Europe in a post-crisis scenario. In particular, we aim to contribute to two ongoing debates in the literature.

Firstly, extant research has shown that challenger parties tend to politicize novel issues, and this is then reflected in their importance in relation to voting for mainstream parties (Hobolt & Tilley, 2016; Santana & Rama, 2018). Yet, there are interesting changes in bailed-out European countries which are worth considering. In the case of Greece, where the main challenger party, namely SYRIZA, became incumbent since 2015. In Portugal, the PCP and Bloco de Esquerda, while remaining outside government, had a coalition of parliamentary incidence with the minority PS government between 2015 and 2019.

Second, we wish to test whether parliamentary debates can shape EU issue voting. Politicization of public debates around the EU has been extensively studied in the media, but less so in parliamentary debates (See Chapter 3). Analysis of EU parliamentary debates has focused on level (Auel et al., 2015; De Wilde, 2011) and causes (Auel & Raunio, 2014; Rauh, 2015). Here we consider its consequences for voting behaviour.

Our analysis is innovative for several reasons. First, by focusing on electoral behaviour, we look at what drives electors, rather than making assumptions about them, by looking at parties and their manifestos. Secondly, we measure the position of parties through the tone employed during parliamentary debates, just before the elections, to understand whether debates may influence voting behaviour.

In relation to data availability, each of the four countries considered— Greece, Ireland, Portugal and Spain—held legislative elections in 2019 and 2020. We were able to run online panels in each of these elections and, therefore, can measure the importance of EU issue voting, in a "real" legislative setting that is almost simultaneous across the four countries.1 2019/2020 constituted a post-bailout election for all countries involved, including the first for Greece, and thus an opportunity to measure EU issue voting when the relative salience of Europe due to the Eurozone crisis has subsided. In that sense, 2019–2020 is a good test to measure the lasting consequences of the crisis, and thus constitutes a "hard test" for EU issue voting, especially since it focuses on legislative rather than the 2019 EP elections.

The chapter will proceed as follows: first, we present the research which has focused on EU issue voting and how it interacts with the left–right dimension of competition, both conceptually, and in particular, in the countries we are concerned with. Next, we explain how challenger parties are supposed to contribute to the politicization of the EU issue. Our literature review leads us to a number of expectations, followed by a presentation of the data, which have hitherto not been presented in the book. We use country by country as well as pooled regression models to understand how different issues matter for voters in post-Eurozone bailout countries.

#### European Integration Issue Voting Before and After the Crisis

The process of European integration has been a focus of analysis as a driver of party system change (Marks & Steenbergen, 2004), and a factor for vote influence (De Vries, 2007, 2010). The way in which Europe relates to the left–right dimension of competition was

<sup>1</sup> Spain held two legislative elections in 2019, in April and November—our panel pertains to the April election. We also ran panels during the Greek legislative election in July 2019, Portuguese election in October 2019, and Ireland elections held in February 2020.

particularly developed by, Hooghe et al. (2002), when they introduced the concept of GAL-TAN, "a new politics dimension […] conceive(d) as ranging from Green/alternative/libertarian (GAL) to traditional/authoritarian/nationalist (TAN)", and found this dimension is the most general and powerful predictor of party positioning on the issues that arise from European integration, including immigration. GAL parties are in favour of European integration and immigration and TAN parties tendentially are against both issues. TAN parties tend to be radical right parties, and this pole of the GAL-TAN axis was responsible for the politicization of EU integration and anti-immigration, much more than GAL parties which did not initially tend to be so homogenously pro-EU.

Ideological and strategic considerations combine to form a u-shaped curve in terms of support for the EU if we consider parties on the left– right axis (Hooghe et al., 2002). Ideologically, the most left-wing parties, namely communist parties, opposed the EU for its alleged neoliberal policies. At the other end of the ideological spectrum parties of the populist right, the extreme right, or the radical right, also opposed the EU. In contrast to the communist parties, these parties' opposition to Europe is driven by concerns about national identity, as well as alleged political sovereignty erosion. It has also been shown that the fact that the EU project changed over time, means that attitudes towards the EU also evolved. A longitudinal analysis showed that whereas Euroscepticism was located on the left originally, and thus had a linear relationship with left–right position, from Maastricht onwards it assumes a curvilinear relationship with left–right, as right-wing Euroscepticism grows (Van Elsas & Brug, 2015).

According to Kriesi et al. (2008) globalization altered both the left– right and the EU issues. Whereas *economic* globalization reinforced the importance of the left–right continuum, the *cultural* issues arising from globalization revolve around national identity, and reinforce the GAL-TAN continuum. On one side of this continuum, conservatives believe that in order to preserve national identity, it is necessary to reduce immigration, and reverse the process of EU integration, whereas liberals are in favour of a multicultural and European identity (Otjes & Katsanidou, 2017). Thus, as the nature of Euroscepticism changed to become more about cultural issues, it was increasingly associated with attitudes towards immigration too (Boomgaarden et al., 2011). Boomgaarden et al. (2011), go further and find anti-immigration attitudes are strongly correlated with opposition to the EU (for example see Boomgaarden et al., 2011), while Toshkov and Kortenska (2015) show that real-world levels of immigration from Central and Eastern Europe also lead to higher levels of euroscepticism in European countries.

Thus, even before the multiple crises from 2009, European integration as a political issue had been the focus of extensive studies on its nature, its relationship to the left–right axis, and its consequences for party systems. Yet, the Great Recession brought even more attention to these issues. Some authors saw the period as a "critical juncture" since the economic crisis not only led voters to punish incumbents, but to withdraw support from all mainstream parties (Hernández & Kriesi, 2016). Hobolt and Tilley (2016) present challenger parties as alternatives, to which many voters turned in the fallout from the crisis. This connects well to the idea that party system change originates from new parties, as existing parties tend to have difficulty adjusting to changes in voters' preferences. According to Hobolt and Tilley (2016), the choice of specific challenger parties across Europe was determined by voters' preferences on three issues that directly flow from the euro crisis: EU integration, austerity and immigration.

### Southern Europe, the Left–Right Dimension and EU Issue Voting

Focusing on Southern Europe, and more precisely bailed-out Europe, studies have shown that challenger parties emerged first and foremost on the left (Hooghe & Marks, 2018). Concurrently, Santana and Rama (2018), show that EU issue voting in bailed-out Europe depended mostly on radical left parties. As the authors have shown, using different measures from the EES 2014 data, comparing voters from left radical parties to left mainstream parties in Europe, the former tend to be more Eurosceptic.

Yet, this view is perhaps not completely consensual. Jurado and Navarrete (2021), using EES data from 2014 to 2019, find that following the crisis EU issue voting did not increase in bailed-out countries. On the contrary, according to these authors, it is in countries with more influence in the EU, measured in terms of the number of MEPs or budget contributions to the EU, where this did occur. They posit that this may be due to a downplaying of the role of Europe by governments which underwent bailouts (Turnbull-Dugarte, 2020). In a similar vein, Hooghe and Marks explain that in Southern Europe "largely as a consequence of austerity, the euro crisis reinforced rather than challenged economic left– right conflict centered on distribution and welfare" (2018, p. 125). Also, Otjes and Katsanidou (2017) find strong evidence that "in the Southern European debtor states economic and European issues are merging as a result of strong European interference in their economic policy".

Thus, on the one hand, some studies suggest that the EU integration issue is mobilized by parties on the extreme of the left–right spectrum, while others suggest that, at least in bailed-out Europe, austerity reinforced the left–right dimension of competition, and indeed EU issue voting does not seem to have increased.

#### EU Issue Voting in Context: The Tone of Parliamentary Debates

In the final part of this chapter, we examine the way in which the tone of parliamentary debates shapes the importance of EU issue voting in bailed-out Europe. Thus, our chapter goes further than just capturing the relative importance of left–right vis-à-vis EU issue voting, among challenger and mainstream parties using observational data. We focus on a specific context—the Parliaments' plenary debates—to understand the effects of voting behaviour of parties' EU politicization. As explained in Chapter 3, the plenary debates are both an important and unique setting to assess the magnitude of EU politicization since they constitute an unmediated discussion of societal issues, and thus can be assumed to capture parties' "pure" positioning on these issues. Extant literature on the politicization of the EU at the institutional level has focused on the plenary debates (e.g., Auel & Raunio, 2014; Rauh, 2015; Rauh & De Wilde, 2018).

While there seems to be a rather negligible impact of institutional factors on the levels of EU politicization (Auel & Raunio, 2014), actors and especially the presence of Eurosceptic parties in parliament—seem determinant to explain the levels of EU politicization in parliaments (Wendler, 2016). Further, in Chapter 3, while considering the trends in the politicization of the EU in the bailout countries Kartalis and Silva concluded that "the overall salience of the EU in the parliaments has been relatively stable and low. When it comes to the contestation dimension, while its variations seem more pronounced, the general tone towards the EU has only swung from very positive to less positive".

Having reviewed the literature, there are a number of findings which have been put forward concerning bailed-out Europe, or Southern Europe, which we will analyse in this chapter. First, it is important to note that since the multiple EU crises occurred, there has also been much more research on consequences for party or party system change than individual-level studies. It is important to revisit the issue of the importance of the left–right dimension for voting in bailed-out Europe, in relation to EU integration issue voting, especially given the post-bailout scenario, where the EU is less salient. The focus on the importance of parliamentary debates in determining the strength of EU issue voting will also contribute to our understanding of the ways in which national institutions contribute to the legitimacy of EU integration.

We saw that the literature which focused on party system change explained how existing parties had difficulty in representing new issues. Instead, it is challenger parties that find it easier to voice these new concerns and indeed do so strategically in order to win more votes. This has also been confirmed by studies which consider EU issue voting. In addition, there is also evidence that this happened, especially during the crisis, by left challenger parties in bailed-out Europe. Thus, following a first broad evaluation of the importance of EU issue voting, we focus on its relative importance for mainstream vs. challenger parties.

Next, we consider the relationship between left–right and EU issue voting in each country. Existing research about party systems, and their alignment on left–right and Europe has found that the parties' positions on Europe are curvilinear in relation to their left–right position. Namely, parties on the extremes of the left–right spectrum tend to have more Eurosceptic positions, vis-à-vis the centrist parties. We will show the degree to which this occurs in each country considered.

Finally, following the rising importance that national parliaments have gained in the EU we investigate whether the tone of EU parliamentary debates, positive neutral or negative, moderate the importance which EU issue voting has on vote. Indeed, we expect that the party's EU tone, measured through parliamentary debates data may magnify the impact of congruence between respondents' and parties' perceived position on the EU in explaining likelihood to vote for that party.

#### Data and Methods

We make use of a unique dataset to investigate several of these claims. Fortuitously, Spain, Greece, Portugal and Ireland all held legislative elections in a twelve-month period, between 2019 and 2020. We were able to field panel surveys in all of these elections and thus are well positioned to analyse if the changes which were detected in the aftermath of the Eurozone crisis remain structural to electoral behaviour and political attitudes, ten years after the onset of the crisis. Our choice of countries is relevant since we did not opt to perform an analysis of Southern Europe, per se. A lot of the research which has been undertaken on the effects of the Eurozone crisis on political support, or parties, is aggregated. Analysis of individual data is lacking, and that is what we attempt in this chapter.

What unites our case studies is the fact that all of them faced bailouts in the years following 2009. Thus, beyond austerity, which was a common recipe applied in many countries in the EU, implementing a bailout had political connotations which are not strictly present in austerity policies. Namely, they involve the agreement between the government of the day and external authorities for a program which may be accepted by the government but is externally produced. This meant the salience of the Eurozone crisis was very high in each of these countries.

The survey is a representative two-wave panel online survey with a sample of 800–3000 respondents per country (see Table 6.1 for details) and fulfils a crossed quota of gender (2 categories), age (3 categories) and education (3 categories). While nonprobability online surveys are less established than probabilistic face-to-face surveys and tend to differ in their marginal distributions, they have been shown to yield very reliable results especially when it comes to causal inferences and explanatory models such as vote choice (Dassonneville et al., 2020), which is what we do in this chapter.

For our analysis, we use data from the first survey wave in each country. Our main dependent variable is the probability to vote (PTV). This item asks respondents to indicate the probability that they would vote for a party on a scale from 0 (definitely would not vote for this party) to 10 (definitely would vote for this party). We then created a stacked dataset in which each respondent is turned into several observations, one observation for each party that is included in the PTV item (10 parties in Portugal, 9 parties in Ireland and Greece, and 6 parties in Spain). This stacked dataset is used to predict probabilities to vote


**Table 6.1** Survey characteristics

in linear regression models, separately for each country and with robust standard errors clustered by the respondent. Our main independent variables are issue proximities. These issue proximities are calculated by using the respondents' position *r* on an issue *j* (measured from 0–10) as well as the party's position *p* on the same issue (equally measured from 0– 10) in two issue dimensions (left–right and Euroscepticism), as follows:

$$\text{tissue proximity}\_{f} = -\frac{\left(\eta\_{j} - p\_{j}\right)^{2}}{10}$$

This measure results in values from −10 (maximum issue distance from party) to 0 (same issue position as party). On the respondents' side, we use the standard question on left–right self-placement ("In politics people sometimes talk of left and right. Where would you place yourself on a scale from 0 to 10 where 0 means the left and 10 means the right?"), as well as an item on European integration ("Some people believe that the process of European integration should move forward to the creation of the United States of Europe. Others believe that the European Union should be dissolved in order to return to a situation in which states are fully sovereign. At which point of the following scale would you place yourself? 0—The EU should be dissolved; 10—The EU should move towards the United States of Europe"). These items on respondents' policy positions are also used as dependent and independent variables in the second part of our analysis. On the party side, we use Chapel Hill Expert Survey (CHES), (Bakker et al., 2020) data from 2019, which estimates parties' positions on a wide range of issues, including left–right (lrgen), and EU integration (eu\_position).

As a further independent variable, we use an item to distinguish types of parties according to their left–right position, as well as whether they are challenger or mainstream. Table 6.2 shows the coding of all parties.

In addition, we also use new data from the Maple project on the EU tone of parties in parliamentary debates before the elections. This data is the result of an automated content analysis of parliamentary debates in the year previous to the national elections of 2019/2020 in Greece, Ireland, Portugal and Spain. As described in Chapter 3 of this book, we used automated techniques to identify EU speeches and to measure their valence towards the EU. The tone of the speeches was measured the following way. Firstly, by translating all of the sentences mentioning the EU (using Google Translate API) in each speech. Afterwards, we attributed sentiment scores to each EU sentence using the R package "sentiment". We use those sentiment scores to measure the levels of EU contestation/negativity in the parliamentary debates. In this analysis, we chose two different variables: the strength of the tone (hence, the absolute difference from 0, without distinguishing between positive and negative), and the direction of the tone (negative vs. positive tones). Both measures


**Table 6.2** Party groups

Definition mainstream: Parties that have been incumbent in the past 30 years, \*party not included in the CHES data, +BE and PCP were not considered "incumbent", as they have never formally participated in the Portuguese government. From 2015 to 2019, these two parties agreed with the minority PS government to form a coalition of parliamentary incidence

are on the party level (See Table 6.3 in the Appendix). To increase the variance in the sample, we use pooled models across all four countries for this part of the analysis, given that otherwise, the number of parties is very low. To ensure that we control for the country-level differences, we centre the tone variables by country mean before using them.

Lastly, all our models control for socio-demographic characteristics (gender, age and education), and the models on vote choice further control for trade union membership, religiosity, and party ID (0 "No party ID", −1 "Left party ID", 1 "Right party ID").

#### Results

First, we want to understand if the left–right dimension of competition still matters most for citizens' vote choice, or if European integration has become more relevant. To do that, we regress issue proximity to the two dimensions of competition (left–right, European integration) on probabilities to vote (PTV) and interact this effect with party type. Figure 6.1 shows the marginal effects plots for each country. We can see that left– right proximity remains the most important driver of likelihood to vote in all bailout countries except Ireland, which clearly is a special case when it comes to dimensions of competition. EU proximity matters significantly in determining vote choice as well, but to a lesser degree than left–right. This confirms our assumptions. Moreover, there are no clear differences between challenger and mainstream parties when it comes to the relevance of the Euroscepticism. Contrary to what we initially assumed, left–right proximity is the most important driver of the vote not just for mainstream, but also for new and challenger parties.

Next, we look at the relationship between the left–right dimension of competition, and EU attitudes, to better understand how they interact in their effect on the vote. Figure 6.2 shows predicted effects from regressing respondents left–right self-placement—direct and in interaction with itself—on their EU position. As expected, we can see that Euroscepticism crosscuts the left–right dimension in all four countries, showing a curvilinear relationship—although in Ireland, the relationship is least strong and inverted.

Lastly, we introduce data on parties' tone in parliament about the EU, to understand how discourses about the EU affect the strength of

**Fig. 6.1** Marginal effects of issue proximity on PTV across different party types. *Marginal effects of interaction terms in country-wise linear regression, with 95% confidence intervals. Full models in* Tables 6.4 and 6.5 *in the Appendix*

EU proximity on respondents' likelihood to vote (see Table 6.3 in the Appendix). We assumed that a party's EU tone, measured through parliamentary debates data, may magnify the impact of congruence between respondents' and parties perceived position on the EU in explaining likelihood to vote for that party. Figure 6.3 shows the results of interacting with each parties EU tone in parliament with party-citizen EU proximity to predict PTV. As we can see, there is no significant effect for the strength of the EU tone (neutral vs positive or negative), but a significantly negative effect of the direction of the EU tone (positive vs negative). This means that for parties that show a more negative tone on the EU (hence, have a more Eurosceptic discourse), the EU proximity matters more in determining the voters' choice than for parties with a positive or neutral EU tone in parliament.

**Fig. 6.2** Relationship between left–right self-placement and EU position. *Predicted effects of interaction terms in country-wise linear regression, with 95% confidence intervals. Full models in* Table 6.6 *in the Appendix*

#### Conclusions

In this chapter we assessed the relative importance of left–right vis-àvis attitudes towards the Europe Union for vote choices in bailed-out Europe. Whereas there has been a lot of research devoted to party system changes following Europe's multiple crises, less is known about the consequences of the crisis for voting behaviour. Our goals were the following: to establish the importance of left–right for the vote in Greece, Ireland, Portugal and Spain, and its relation to the EU issue. Then, we analysed the (relative) importance of these issues contrasting mainstream and challenger parties. Finally, we consider whether EU parliamentary debates, and in particular their tone, moderate the importance of EU issue voting.

This contribution is important for at least two reasons: firstly, due to the fact that in bailed-out European countries key challenger parties

**Fig. 6.3** Marginal effects of issue proximity on PTV according to the EU tone of parties in parliamentary debates. *Marginal Effects of Interaction Terms in Pooled Linear Regression, with 95% Confidence Intervals. EU Tone is Centred by Country Means. Full Models in* Table 6.7 *in the Appendix*

have become full or partial incumbents. In January 2015, SYRIZA which was hitherto a challenger party became incumbent in Greece, while the PCP and Bloco de Esquerda, while remaining outside of government, entered a coalition of parliamentary incidence with the Socialist party, which formed a minority government following the 2015 elections in Portugal. We are able to test to which degree in 2019, these parties continue to politicize the EU. Second, we wished to understand EU issue voting in in relation to the tone employed in national debates.

Our results show that in Greece, Portugal and Spain, for each of the categories we tested, left–right proximity has a higher impact on likelihood to vote for a party than proximity on EU issues. Ireland is a partial exception to this trend since neither of these issues seems to matter very much for the challenger left, and the mainstream left is quite different from the mainstream right. Whereas in the former the EU issue is more important than left–right, the situation is reversed for the mainstream right parties.

When we consider the difference between challenger and mainstream parties, we do not see that European issue proximity to parties' positions are more important determinants of likelihood to vote than left–right for challenger parties. There does not seem to be any relevant difference between the importance of European issue proximities across party types. The reason for this similarity between mainstream and challenger parties, and the resilience of left–right, at least in Greece and Portugal, maybe due to the fact that in these countries, existing challenger parties on the left became either incumbent (Greece) or formally supported the government from parliament (Portugal), thus contributing to downplay the EU proximity issue, vis-à-vis mainstream parties.

Third, we explored the relationship between left–right placement and the EU proximity, by regressing respondents left–right self-placement direct and in interaction with itself—on EU position. As expected, we find that the EU issue proximity has a curvilinear relationship with left–right. Thus, indeed, EU issue voting does cross-cut left–right in all countries concerned.

Thus, in this chapter, we have been able to show the resilience of the left–right self-placement, both overall, and also for challenger parties across bailed-out Europe. EU issue voting matters for voters across party types, and it does not reinforce left–right placement, in bailed-out Europe. Also, it is interesting to note that Portugal is not an exception in terms of voters while for a long time it has been considered as an exception for parties and party system perspective as an ultra-stable system. Even though these trends are unequivocal and identified, across the chapter it became clear that there are important country differences. For instance, neither of these issues seems particularly relevant in 2019, in Ireland. The next chapters which consider each of these countries as case studies, combining media, parliamentary and voter data will be able to explain some of these distinctive features.

Finally, we showed how politicization, in this case measured in automated fashion and exogenously in parliamentary debates is associated with an increase in EU issue voting, at least for Eurosceptic parties. Contrary to media, which has its own agenda, we can say that in plenary debates, parties have control of the message. Hence, EU politicization which is developed by parties in the plenary debates has an impact on voting behaviour. This is an important signal for the relevance of national institutions, and in particular parliaments, in contributing to legitimizing the EU.

#### Appendix

See Tables 6.3, 6.4, 6.5, 6.6 and 6.7.



**Table 6.4** The effects of left-right proximity on PTV across four party groups


Country-wise linear regression. Standard errors clustered by respondent. *t* statistics in parentheses. +*p* < 0.10, \**p* < 0.05, \*\**p* < 0.01, \*\*\**p* < 0.001




162 L. HEYNE ET AL.



Country-wise linear regression. Standard errors clustered by respondent. *t* statistics in parentheses. +*p* < 0.10, \**p* < 0.05, \*\**p* < 0.01, \*\*\**p* < 0.001


**Table 6.7** The effects of EU proximity on PTV according to parties' EU tone in parliament

Pooled linear regression. Standard errors clustered by respondent, country-level predictors are centered by country mean. *t* statistics in parentheses. +*p* < 0.10, \**p* < 0.05, \*\**p* < 0.01, \*\*\**p* < 0.001

#### References


**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# EU Issue Voting in Simultaneous Elections: The Case of Belgium

*Dieter Stiers* 

# Introduction

On 24 May 2019, Belgium held a major Election Day in which representatives for three different governmental levels were chosen. In this "mother of all elections", voters could cast three different votes: for the regional, the national, and finally the European level—in that order. With this system of simultaneous elections, the European level could be expected to be omnipresent in the campaign and therefore also influence the vote at the regional and national levels. On the other hand, this also means that the European issue faced fierce competition for attention during the campaign, which focused mostly on regional and national issues (Pilet, 2020). The electoral results showed substantial losses by governing parties, which were gained by the opposition (Pilet, 2020). Especially radical (populist) parties performed particularly well (Gallina et al., 2020; Goovaerts et al., 2020). In general, the country appeared

© The Author(s) 2023

169

D. Stiers (B)

Centre for Political Science Research, KU Leuven, Leuven, Belgium e-mail: dieter.stiers@kuleuven.be

M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_7

divided, with the most support for (centre-)right parties in Flanders, and (centre-)left in Wallonia (Hooghe & Stiers, 2022).

The purpose of this chapter is to investigate whether people voted for the national level with European issues in mind. As explained in the theoretical framework of the book (Chapter 1), there are important implications of this kind of EU issue voting in national elections for the legitimacy of the European institutions. As the national governments largely decide on their representation on the European level, from a normative point of view it is important to investigate whether people vote for the national level with the European issue in mind. As a net contributor, the investigation of the Belgian case moreover provides a useful point of comparison for the other countries discussed in the subsequent chapters to test whether the extent of EU issue voting diverges or converges.

Even though it is a central part of the EU, Belgium very much presents a least-likely case to find evidence of EU issue voting, as the European issue is likely to be overshadowed by the issues that are important for the "first-order" regional and national elections (Reif & Schmitt, 1980), and as the compulsory voting system in place is likely to draw non-interested voters to the ballot box, engaging less in ideological issue voting (Dassonneville et al., 2019; Selb & Lachat, 2009). The findings in Chapter 5 already revealed Belgium to be an odd case out. While there was evidence for EU issue voting and its expected asymmetric nature, voters only seem to punish parties that are more positive on the EU than they are themselves—i.e., the reversed asymmetry than what was found in the other countries. This chapter will use individual-level survey data to contrast the EU issue with other determinants of the vote to test whether and which voters vote for the national level with the European issue in mind.

### EU Issue Voting in Context: Belgium as a Least-Likely Case Study

The European Union (EU) is an important issue in Belgium, one of its founding members. Not in the least, Brussels, Belgium's capital, houses the most important institutions of the European Union and is the location for its summits. This means that Belgians feel the presence of the EU very concretely. Media report regularly on European decision-making, as it is happening in the centre of the country, and reporting that "Brussels has decided to…", means something more concrete in the minds of the people than when the same news is brought to people in countries further away. However, many people also feel the impact of the EU institutions on their daily lives, e.g., when demonstrations obstruct the traffic in Brussels. Belgian politicians have also held important positions at the European level, with two of the three permanent Presidents of the Council so far. Clearly, Belgians have good reasons to pay attention to the EU, and to take considerations about this level into account when they take part in the political process, for instance when they turn out to vote. There are indeed good reasons to believe that Belgian voters should be likely to take the EU issue into account, as in general Belgians have been found to be quite positive and supportive of the EU (McLaren, 2004). Furthermore, while the economic crisis of 2009 decreased public support for the EU in many countries, there was only a limited loss in support for the EU between 2007 and 2011 in Belgium compared to other countries (Armingeon & Ceka, 2014; Braun & Tausendpfund, 2014).

However, even though the European Union is very much present in the Belgian political landscape, when it comes to explaining voting behaviour, several characteristics of the Belgian national elections make it a least-likely case to find the effects of the European issue on the national vote choice (de Vries, 2007). More specifically, the system of simultaneous elections together with compulsory voting, makes it a difficult context to find the effects of opinions on the EU on the vote at the national level.

First, as mentioned above, one peculiar characteristic of the Belgian elections of 2019 is that elections for three different levels of government were held on the same day: the regional, national, and European elections. This system has been in place since the election of 2014 and was introduced to increase the stability of the federation by making it more likely that the coalitions at different levels of government are the same (Deschouwer & Reuchamps, 2013). The only elections that follow a different schedule are the local elections, held every six years. Simultaneous elections also make for simultaneous campaigns, which means that the attention of parties, media, and voters, is divided over issues of the different electoral levels. In principle, it could be thought that these simultaneous elections make it more likely to find EU issue voting in national elections: as the European elections are held together with the national elections, this means that the EU issue should also be discussed at length during the (national) campaign. Especially compared to national elections in other countries, where European elections follow a different schedule, this means that the European level should be at least somewhat salient in the months before Election Day.

However, there is reason to believe that simultaneous elections actually *decrease* the attention on the EU issue in the campaign overall. As previous chapters already discussed, European elections are often considered "second-order national elections", as they are decided mainly on issues that are at play in the first-order (i.e., national) electoral arena (Reif & Schmitt, 1980; Schmitt, 2005). This means that other issues on the national level might overshadow the EU issue. This effect can be expected to be even stronger when the elections for these levels are held on the same day. Furthermore, and importantly, given the Belgian federal structure, the regional elections can also be considered first-order elections (Deschouwer, 2012; Schakel & Jeffery, 2013), so regional issues also compete with the EU issue for the necessary attention. Instead of the EU issue being more prominent overall because of the European elections taking place, it thus competes for attention against two "firstorder" electoral arenas, which further limits the attention for this level specifically when voters turn out to vote for all three levels. Therefore, it can be expected that the simultaneous elections decrease the importance of the EU issue overall, and the fact that the European elections are held together with the national elections does not increase the salience of the EU issue for the electoral contest on the national level.

Besides the fact that the EU issue likely receives less attention, simultaneous elections can also decrease the importance of the EU issue in the vote for the national level specifically: voters who wish to express their opinions on the EU issue can do so in the election for that level while expressing their interests on regional and national issues with the votes for those levels respectively. This might decrease EU issue voting in the national election. A finding that further corroborates this argument is that a substantial amount of voters split their ticket (i.e., vote for different parties for the different levels). In their investigation of the simultaneous elections of 2014, Willocq and Kelbel (2018), for instance, find that about a third of the voters voted for at least two different parties on the three levels, and the largest group of the ticket-splitters (14.8%) votes for the same party on the regional and national level while supporting a different party on the European level. Van Aelst and Lefevere (2012) confirm for the 2009 elections that people split their tickets based on Euro-specific considerations. While these studies thus imply that voters do take the European issue into account when casting a vote at the European level, their findings further indicate that Belgium is a least-likely case to find the EU to influence voting behaviour for the national level specifically. If people split their ticket between the national and European levels with European issues in mind, they are also more likely to cast a vote for the national level with specific national issues in mind instead of using the same—possibly European—considerations for the vote on every level.

Finally, an important characteristic of the Belgian electoral system that makes it a hard test for EU issue voting is that all voters are obligated to turn out to vote on Election Day. This system of compulsory voting has ensured that turnout rates are stable at around 90% for each electoral level. This is a stark contrast with the lowering turnout rates in other countries (Hooghe & Kern, 2017). This decline has also been observed for the European elections: even though the 2019 election saw an increase in turnout compared to the previous elections, in general, there is a downward trend in participation that is expected to continue in the decades ahead (Bhatti & Hansen, 2012). As a consequence of these high rates, however, Belgium draws voters to the ballot box that would otherwise not turn out because they are not interested in politics (Hooghe & Pelleriaux, 1998; Selb & Lachat, 2009). While there is some evidence that voters who are forced to turn out inform themselves at least to some extent (Shineman, 2018), other studies have found that these "reluctant" voters cast votes that are less in line with their own ideological opinions and interests because they are less interested in politics (Dassonneville et al., 2019; Hooghe & Stiers, 2017; Selb & Lachat, 2009). In terms of the research question under investigation, if the compulsion to turn out to vote draws a larger proportion of voters casting less congruent or even random votes to the ballot box, this makes it harder to find evidence for EU issue voting. Therefore, as will be further explained below, voters' levels of political information will be taken into account to empirically investigate EU issue voting.

### Politicisation of EU in Media and Parliamentary Debates

Before testing EU issue voting at the individual level using survey data, it is useful to examine how salient the European level is in general in Belgian politics, and in what tone the EU is usually discussed. To do so, I look at salience levels and tone both in the country's media and in parliamentary debates.

**Fig. 7.1** EU salience and tone in Belgian media

First, I investigate the salience and tone in two main newspapers one of each main region respectively: De Standaard (Flanders) and Le Soir (Wallonia). The data span the last five national elections, between 2003 and 2019.1 Salience was measured as the percentage of articles that referred to the EU. The tone is measured as a combination of a measure of the sentiment of the title of the article and a measure of the average sentiments of the EU sentences in the article itself. Positive values represent a positive tone, and values below zero represent a negative perception (see also Chapter 3). The results are displayed in Fig. 7.1.

The results in Fig. 7.1 show that, in all election years, there is more attention for the EU in the Flemish newspaper (i.e., De Standaard) than in the Walloon newspaper (Le Soir). However, there is substantial variation in EU salience between the years. In Flanders, EU salience is highest in 2014, while it was highest in Wallonia in 2010. In general, there was an upward trend in salience starting from 2003 on. 2019 breaks this trend, however, with the lowest EU salience in the series. This could partly be

<sup>1</sup> Note that the data for 2019 are based on the online webpage only, while the previous years are based on the printed versions.

explained by the fact that this year the regional and national elections were held on the same day as the European election and received the most media attention. However, interestingly, also in 2014 these elections were held on the same day, and this year we note the highest salience numbers. Looking at the tone, a downward trend can be observed. While the worst tone was observed in 2010 in both regions, the last two elections under investigation scored lower than the first two. Furthermore, while the Flemish newspaper scores highest in the whole series, in 2007, in all other years, the Walloon paper is somewhat more positive overall.

Second, I look at the salience and tone of the interventions by politicians of different parties during parliamentary debates. Here as well the salience measure refers to the percentage of interventions referring to the EU, while tone reflects the average tone of all sentences about the EU in a given speech. The results are displayed in Fig. 7.2. The top part of Fig. 7.2 shows the salience for Flemish parties and the bottom part of the Walloon parties.2

Looking at the salience of the EU in Flanders, what stands out is the very high salience for the nationalist party N-VA and liberal party Open VLD in 2007. This year, national elections were held in Belgium, and the outgoing prime minister, Guy Verhofstadt, would hold the first position on the European list and won more than half a million preference votes in the European election two years later. It is likely that his European ambitions, which he pursued ever since, also led him to attach importance to the European issue—something that the ideologically close N-VA had to follow in reaction to this increased attention. Do note, however, that these numbers are based on a rather low number of speeches, so they might be somewhat inflated. What also stands out is the recent high attention for the EU of the two radical parties. First, the radical-left PTB-PVDA scores are constantly high. Second, since 2015, the radical-right Vlaams Belang (VB) has the most attention for the European issues of all Flemish parties—although this is mostly because of its criticism of the European project (see below).

<sup>2</sup> PTB-PVDA is a unitary, national, party and is therefore included in both graphs; ECOLO-Groen consists of the green parties of the two main regions respectively; they act as one fraction in the national Parliament and are therefore coded together and included in both graphs.

**Fig. 7.2** Salience in parliamentary debates in Flanders (top figure) and Wallonia (bottom figure)

The bottom part of Fig. 7.2 depicts the same information but for the Walloon parties. A first observation is that the EU is generally less salient—as can be seen in the lower range of the y-axis. Among those parties, however, it is especially the ecologists that talk most about the EU, although this decreases somewhat in the last 10 years. The liberal MR, on the other hand, strongly increased in attention to the EU in the last years. PTB-PVDA, which scored lowest among the Flemish parties, scores relatively high among the Walloon parties, which again shows the large difference in salience for the EU overall.

Next, I look at the tone with which the EU is discussed during these parliamentary interventions. The results are displayed in Fig. 7.3.

The top graph in Fig. 7.3 shows that Flemish parties are generally positive about the EU. The highest score (i.e., most positive) is reached by the green party in 2004, although this party scores rather average in the other years. Again, the radical parties stand out to some extent, with Vlaams Belang scoring the lowest in some recent years, and PTB-PVDA being very critical in 2017, although they score just above zero in 2019. The Walloon parties started out very positive two decades ago, but more recently they converged as well around a slightly positive tone. Especially the socialist PS stabilised on positive scores after a more volatile period in which they were also critical. Mostly the MR seems to be positive in recent years, while also here the low scores of the PTB-PVDA stand out. Overall, however, Belgian political parties are generally positive about the EU, and only the radical parties have been most strongly negative in the last years.

### Modelling EU Issue Voting in 2019

The results above imply that, overall, there is a positive consensus about the European level among most Belgian political parties, but generally the European level is not very salient in national elections. Does this mean individual voters do not take into account the EU issue when they vote? As the main test of EU issue voting in the Belgian national elections of 2019, individual-level survey data are used. More specifically, I use the data of the post-electoral wave of the MAPLE project, gathered in an online survey among Belgian respondents after the elections of 24 May 2019. Sampling was based on gender, age, educational level, and NUTS1 regions.

The dependent variable in the models indicates the party voted for in the election at the national level. Respondents could indicate their vote choice, and the vote for the major parties will be modelled.3 The main

<sup>3</sup> Parti Populaire and DéFI were excluded because of a low number of respondents indicating having voted for these parties.

**Fig. 7.3** Tone in parliamentary debates in Flanders (top figure) and Wallonia (bottom figure)

independent variable measures the respondent's opinion about further European integration: 'Some people believe that the process of European integration should move forward to the creation of the United States of Europe. Others believe that the European Union should be dissolved in order to return to a situation in which states are fully sovereign. At which point of the following scale would you place yourself?' Respondents could indicate their opinion on a scale ranging from 0 ("The EU should be dissolved") to 10 ("The EU should move towards the United States of Europe").

Other possible determinants of the vote are included as well. First, a question asking the respondent's opinion about immigration is included, as this was a highly salient issue in Belgian politics in the years before the elections—and even led the largest Flemish party, N-VA, to quit the government some months before Election Day (Pilet, 2020). In the survey, respondents could indicate their opinion on the statement that Belgium should implement a more restrictive immigration policy on a scale from 1 ("strongly disagree") to 5 ("strongly agree). The values on this scale have been reversed so that higher values indicate a more proimmigration opinion. As a more traditional issue, people's opinions on the statement that "women should have the right to access to an abortion in all circumstances", measured on the same scale, is included as well. As radical parties performed well in the 2019 elections, two measures of support for the political system are included as well. The first is a measure of political trust, calculated as the mean indicated trust in the country's National Government and National Parliament on a scale from 0 ("no trust at all") to 10 ("complete trust"). The answers to these two questions are taken together as they correlate substantially (Pearson's *r* = 0.87). A second measure of system support is satisfaction with the way democracy works in Belgium.

Besides these more specific issues, a measure of ideology is included as well. To this end, I include the respondent's self-placement on the general ideological continuum in which 0 refers to a left position and 10 to a right position. Previous research has shown that people can interpret this scale meaningfully and that it leads to a useful summary of the ideological stances of voters vis-à-vis parties (Dalton et al., 2011; van der Eijk et al., 2005). Finally, in the first model, two more strong valence measures will be included: a retrospective sociotropic evaluation of the economy, which has been shown to influence voting for an incumbent party (Lewis-Beck & Stegmaier, 2000), and a rating of how much the voter likes the leader of the party they voted for—on a scale from 0 ("strongly dislike") to 10 ("strongly like").

Besides these main variables of interest, some control variables are included in the models as well. First, I control for the standard sociodemographics sex, age, and educational level. Sex distinguishes male and female respondents, with female respondents as the reference category. Age denotes the age of the respondent in the election year. Educational level is measured on a 9-point scale starting from early childhood education to doctoral or equivalent education. Besides these characteristics, I control for trade union membership status (0 = not a member; 1 = member) and religiosity (self-declared religiosity on a scale from 1 ("I have no religious believes") to 4 ("I am very religious")).

As the dependent variable consists of several categories without a logical ordering (i.e., party choice), I estimate multinomial logistic regression models to investigate the impact of the different issues and control variables on the vote. To take into account the different scales on which the variables are measured, all continuous measures have been rescaled to range between 0 and 1. Given that almost all parties operate exclusively in one of the two main regions, there are basically two different party systems, split along the language border (Deschouwer, 2012). Therefore, the analyses will be presented for Flanders and Wallonia separately. In Flanders, the Christian-Democratic CD&V will be included as a baseline category as it was one of the two Flemish incumbent parties (together with the liberal party Open VLD) and it has the most centrist ideological position. For the analysis in Wallonia, the single Walloon incumbent party MR (liberals) will be included as the baseline category.

In the first set of analyses, I will look at the direct effects of the variables of interest on the vote. In a second analysis, I take into account the challenging setting Belgium offers for voters who want to express their issue stances with their vote. As explained above, the compulsory voting system in Belgium also draws the less politically interested to the ballot box. Previous research has shown that less sophisticated voters tend to be less consistent in their attitudes and are less able to connect their political opinions with a vote that expresses these interests (Converse, 1964; Lau et al., 2014; Zaller, 1992), and this has also been found to matter in the elections under investigation specifically (Gallina et al., 2020). Therefore, I include an interaction of the measure on the EU issue with political sophistication, to test whether the more sophisticated voters take into account the EU issue to a larger extent when they vote. Sophistication is measured by an indicator of media use, measuring how many times per week the respondent uses the newspaper for political information (ranging from 0 "never" to 8 "7 days per week").

Before turning to the results, one important note that needs to be made is that the number of observations in the dataset is low. While there is some attrition between the first survey wave—held before the election and the second wave after the election (of main interest here), the number of observations for some parties drops more given that the respondents are divided over the two main regions, and then the many parties within those regions. This results in some cells that are too small to make reliable estimates. Therefore, two sets of results will be discussed: one focusing on the vote choice as reported after the election, and one focusing on the vote intention as reported in the first survey wave, gathered some months before the elections on a larger sample. With this dual strategy, the first wave is used in its advantage of containing more observations, and the second as it is a post-electoral study. In this way, the conclusions from both sets of analyses reinforce each other. In terms of reporting the results in more detail, the focus here will be on the post-electoral study, while the results using the vote intention are reported in Appendix 2.

#### Results

Was the EU an important issue when voters cast their votes for the national level in 2019? To investigate this question, I estimate multinomial models explaining the vote. As the coefficients of multinomial models are complex to interpret and depend on the chosen baseline outcome, I present the results by plotting the average marginal effects of the main variables of interest in Fig. 7.4 (Flanders) and Fig. 7.5 (Wallonia). The full tables with results are included in Appendix 1.

The results in Fig. 7.4 show that the different variables of interest help to explain the vote choice between Flemish parties. First, looking at the European issue, this indicator only shows modest effects on the vote. When examining the recalled vote after the election, there is a significant positive effect for the Christian-Democrat CD&V, where voters are 14 percentage points more likely to vote for this party when they believe that European integration should go further than when they oppose to the EU. This result is not confirmed when looking at the intention to vote before the campaign, but in this model, there is a small positive effect of the liberal Open VLD, and a negative effect of almost 11 percentage points for radical-right Vlaams Belang. This result could be expected, as Open VLD is strongly present at the European level by means of Guy Verhofstadt, while Vlaams Belang is openly opposed to further European integration (Almeida, 2010; Marks et al., 2007).

**Fig. 7.4** Explaining the vote in Flanders (*Note* Average marginal effects and 95% confidence intervals, based on the results reported in Model 1 in Table 7.1)

The other variables of interest also show some interesting patterns. Positive attitudes about immigration positively predict support for the radical-left PVDA, and negative for the Christian Democrats (vote) and nationalist party N-VA and Vlaams Belang (vote intention). Being in favour of abortion is negatively associated with support for the Christian Democrats, which is in line with its customary call for traditional family values. As was expected, the measures of support for the system—political trust and satisfaction with democracy—are strongly negatively related to supporting Vlaams Belang, while they are positively related to supporting the traditional parties. Finally, the general left–right continuum performs well in the model, with most left-wing voters supporting the radical left and social democrats, and the most right-wing voters finding their way to the nationalists and radical right. While the Christian Democrats are commonly assumed to hold the centre position, in this case, their voters score somewhat more on the left side of the political spectrum.

In Fig. 7.5, the same analysis is repeated, but this time for Walloon parties. Also here, we find some significant effects of the EU issue on the vote. The largest effect is the negative association between favouring

**Fig. 7.5** Explaining the vote in Wallonia (*Note* Average marginal effects and 95% confidence intervals, based on the results reported in Model 1 in Table 7.2)

more European integration and the vote for the socialist party PS. Voters supporting a fully integrated "United States of Europe" are 33 percentage points less likely to vote for the PS than voters who want to dissolve the EU. This is rather surprising, as the party does not have a strong anti-EU narrative. Possibly, voters remembered the PS's strong stand against the CETA agreement between the EU and Canada from some years before. Contrarily, there is a sizeable positive effect for supporting Christian Democrats cdH (vote) and the green party ECOLO (vote intention).

The other issues, abortion and immigration, add less to explaining the vote than they did in Flanders. The only significant association is the positive effect of positive perceptions of immigration on the intention to vote for ECOLO. Political trust and satisfaction with democracy also only help explaining the intention to vote—the former positively for liberal incumbent MR and negatively for its main challenger PS; the latter negatively for radical-left PTB. Finally, as in Flanders, the general left–right continuum is an important determinant of the vote, with the expected negative associations with the radical left and socialists, positive for the liberals, with a moderate position for the Christian Democrats. Somewhat contrary to its commonly attributed left-wing position, there is no significant association with supporting the ecologist party.

The results above show mixed results regarding the EU issue: while there are some significant effects for some parties, these are not consistent and substantially rather small. However, it is possible that there is individual-level heterogeneity. More specifically, I test whether more sophisticated voters are better able to express their issue opinions with their vote, by including an interaction between the EU issue and media use. The results are summarised in Table 7.3 in Appendix 1 for Flanders and Table 7.4 for Wallonia. The replication using the vote intention from the first survey wave, is reported in Appendix 2, and leads to the same conclusions as the analyses reported here.

The results do not show strong support for a moderation effect to be at play. Only in Flanders, there is a significant negative interaction between the EU issue and media use for voting for Vlaams Belang compared to the baseline CD&V. This indicates that voters who are critical of the EU become less likely to vote for Vlaams Belang at higher levels of political sophistication—and indeed supports the assertion that higher sophisticates express their issue opinions more strongly with their vote. To get a full overview of the interaction effects, Figs. 7.6 and 7.7 display the average marginal effects of the EU issue at different levels of media use for the different parties respectively.

The results in Fig. 7.6 show a more nuanced picture of the effects of the EU issue on the vote. For Christian Democrat voters, there is a positive effect of favouring further European integration on the vote, but only from a certain level of political sophistication. The same counts for Vlaams Belang, where the negative effect manifests itself among its more sophisticated voters. These results are in line with those presented in Fig. 7.4 but show that the main mechanism is strongest for voters at higher levels of sophistication. These results provide further evidence that the Belgian system of compulsory voting might inhibit EU issue voting in national elections.

Figure 7.7 shows the same results for the Walloon parties. Also here there is a positive effect of the EU issue on voting for the Christian Democrats, although it only manifests itself at lower levels of sophistication—where the number of observations is higher and the uncertainty around the estimates is generally lower, also for the other parties. The

**Fig. 7.6** Average marginal effects of the EU issue at different levels of media use: Flanders (*Note* Average marginal effects and 95% confidence intervals, based on the results reported in Table 7.3)

**Fig. 7.7** Average marginal effects of the EU issue at different levels of media use: Wallonia (*Note* Average marginal effects and 95% confidence intervals, based on the results reported in Table 7.4)

results also show that the surprising negative association between the EU issue and voting for the PS is limited to its least sophisticated voters. At higher levels of sophistication, the association is not significant.

### Conclusion

At first sight, Belgium seems to provide an ideal setting to find strong effects of the European issue on political behaviour. With the main European institutions in the centre of the country, the European Union is omnipresent in Belgium. Yet, however, when looking at media and Parliamentary data, while media and parties are generally positive about the EU, it has a very low salience. Furthermore, in 2019 specifically, its salience was substantially lower compared to the previous elections. Low partisan conflict and low salience are considered factors that strongly inhibit EU issue voting (de Vries, 2007). Furthermore, Belgium's simultaneous elections create a challenging setting to find evidence for EU issue voting in national elections, as voters can express their specific European concerns with another vote almost instantaneously. Previous research has also indicated that those voters who split their ticket between the different levels mostly do so between the national and European levels, casting the latter vote with a European issue in mind. This context, together with the Belgian system of compulsory voting also drawing the least politically interested voters to the voting booth, makes Belgium a least-likely case to find evidence of EU issue voting in national elections. This would lead to a rather pessimistic normative evaluation of the state of democracy for the European representation in Belgium: as the representatives on the national level play an important role on the European level, it is important that people choose which party to support in the national election with the European issue in mind.

To some extent, the results corroborate that Belgium is a hard case to test the EU issue voting on the national level, as there is only limited evidence for the EU issue to be at play in voting behaviour for the national level. In Flanders, the strongest and most consistent finding is the negative association between pro-European opinions and support for the radical-right party Vlaams Belang. This result could be expected, given the party's open objection against further European integration (Marks et al., 2007), which increases the politicisation and thus salience of this issue for this party (Hutter & Kriesi, 2019). This association is especially prominent among voters with higher levels of political sophistication, who are better able to bring their vote in line with their political opinions. In Wallonia, pro-European voters seem to be more likely to vote for the Christian Democrats.

Importantly, it needs to be noted that there are no consistently strong effects of the other issues included in the models either. Besides the general left–right ideological continuum and, to some extent, measures of support for the system, the models have low explanatory power when it comes to explaining the vote choice. This shows the complexity of the Belgian party system, which is structured on many different cleavages and has a complex history. Hence, when putting the results of the EU issue in the context of the findings of the other issues included in the models, the conclusion needs to be that its effects are comparable to those of more traditional issues on the national level: it helps to explain the vote for some but not all parties, varying in size. This provides further evidence for the claim that the politicisation of the European issue is indeed growing to be a new structuring conflict next to existing cleavages (Hutter & Kriesi, 2019). There is most consistent evidence for a positive association with support for the Christian-Democratic parties. These parties traditionally hold the most moderate ideological position overall in Belgium, and it is possible that this pushes voters to consider a wider range of issues than for other parties campaigning strongly on one or a small set of traditional national issues. Overall, however, it seems like Belgian voters support different parties for a variety of reasons, and there is no one specific issue that primarily drives voters' decisions.

#### Appendix 1: Full Tables of Models Presented in the Text

See Tables 7.1, 7.2, 7.3, and 7.4.



(continued)


*Note* Entries are log-odds coefficients; standard errors in parentheses, average marginal effects in brackets. Significance levels: +*p* < 0.1; \*\**p* < 0.01; \*\*\**p* < 0.001

#### 190 D. STIERS



(continued)


192 D. STIERS

**Table 7.2**

(continued)


*Note* Entries are log-odds coefficients; standard errors in parentheses, average marginal effects in brackets. Significance levels: +*p* < 0.1; \*\**p* < 0.01; \*\*\**p* < 0.001




\*\**p* < 0.01; \*\*\**p* < 0.001

**Table 7.3**

(continued)


**Table 7.4** Multinomial regression models explaining the vote in Wallonia including interaction effect

(continued)


#### **Table 7.4** (continued)

*Note* Entries are log-odds coefficients; standard errors in parentheses, average marginal effects in brackets. Significance levels: +*p* < 0.1; \**p* < 0.05; \*\**p* < 0.01; \*\*\**p* < 0.001

# Appendix 2: Replication of the Results Using Vote Intentions (Survey Wave 1)

See Figs. 7.8, 7.9, 7.10 and 7.11.

**Fig. 7.8** Explaining the vote intention in Flanders (*Note* Average marginal effects and 95% confidence intervals)

**Fig. 7.9** Explaining the vote intention in Wallonia (*Note* Average marginal effects and 95% confidence intervals)

**Fig. 7.10** Average marginal effects of the EU issue at different levels of media use: Flanders—vote intention (*Note* Average marginal effects and 95% confidence intervals)

**Fig. 7.11** Average marginal effects of the EU issue at different levels of media use: Wallonia—vote intention (*Note* Average marginal effects and 95% confidence intervals)

#### References


*Journal of Political Research, 48*(5), 573–597. https://doi.org/10.1111/j. 1475-6765.2009.01834.x


**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# The Importance of EU Issues in German Elections

*Rosa M. Navarrete and Marc Debus* 

# Introduction

Germany is the largest economic power in the European Union (EU) and exercises the most influence on EU policy (Busse et al., 2020; Krotz & Schramm, 2021). Nevertheless, for years Germany's European vocation and commitment to European integration made it cautious and reluctant to impose its national preferences in the EU, resulting in most of its EU policy initiatives being presented in tandem with France (Krotz & Schramm, 2021; Paterson, 2011). This way, German Europeanism could be summarized in the belief that what is good for Europe is also good

R. M. Navarrete (B) Department of European Social Research, Saarland University, Saarbrücken, Germany e-mail: rosa.navarrete@uni-saarland.de

M. Debus

School of Social Sciences & Mannheim Centre for European Social Research, University of Mannheim, Mannheim, Germany e-mail: marc.debus@uni-mannheim.de

© The Author(s) 2023

205

M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_8

for Germany (Paterson, 2011). For decades, the EU was not a polarizing issue in Germany, and until the turmoil of the European financial crisis, the stance of the various political parties on European integration were broadly similar (De Vries & Hobolt, 2016; Debus, 2023). Following the crisis, Germany assumed a more prominent role as the leading country defending austerity and, as in all the other creditor countries, German public opinion discussed the pros and cons of assisting EU member states that had been severely affected by the crisis, which forced the political parties to rethink their positions on European integration (Gross & Schäfer, 2020).

Despite a significant share of citizens becoming increasingly sceptical about European integration after "the Maastricht blues" (Teschner, 2000), and particularly during and after the sovereign debt crisis within the Eurozone, political parties, with the exception of Alternative for Germany (AfD—Alternative für Deutschland), tended to downplay the European issue in their campaigns and adopt more moderate positions towards the EU in their manifestos (Debus, 2023; Schmitt-Beck, 2017). AfD was founded in 2013 in reaction to the economic shock and public discontent towards the measures agreed upon in Brussels to alleviate the impact of the crisis in the Eurozone. In fact, its name appealed to an alternative for the apparent consensus of the other parties in defending the monetary union; the origin of the new party's name was a statement by Chancellor Angela Merkel in which she said there "is not alternative" to saving the common European currency (Bebnowski, 2016: 32; Prantl, 2013), indicating the saliency of European integration policy for the AfD on the one hand and for economic and financial issues on the other. Thus, the AfD would initially have intended to fill a vacant space on the political spectrum by being the first German party with a clearly stated negative view of European integration (Arzheimer, 2015; Debus, 2023). Yet, the AfD could be considered as a "soft Eurosceptic" party because most of its critique of the EU focused on its monetary policy and the financial assistance provided to other EU states (Arzheimer, 2015; Schmitt-Beck, 2017). Moreover, the party's subsequent electoral success during the first five years following its formation is not be so clearly tied to its contestation of EU policy, which is now mixed in with a range of other more disruptive radical right-wing propositions (Conrad, 2020; Lees, 2018; Schmitt-Beck, 2017). Furthermore, as Schmitt-Beck (2017) notes, most AfD voters in the 2013 federal elections, the 2014 European Parliament election and the subsequent regional elections chose AfD for largely xenophobic motives, with only a minority supporting it for its position on the European currency union. In subsequent years, especially after the refugee crisis of 2015, the AfD focused almost entirely on an anti-immigrant agenda that was defended by means of a nativist rhetoric (Bräuninger et al., 2020; Schmitt-Beck et al., 2022a; Stecker & Debus, 2019).

Nevertheless, the stance on European policy is important for German voters, even if it is not a polarizing issue for political parties. While the parties may agree on their positive view of the EU—again, with the exception of the AfD—citizens pay attention to the nuances of support for the European project when deciding whom to vote for (see the chapter by Pannico and Lobo in this volume). In the federal elections of 2002 and 2005, citizen opinions on EU issues and on Turkey's possible accession to the EU influenced voter choice in Germany (Debus, 2007: 286; Schoen, 2008). This suggests citizen concerns on matters decided at the European level are relevant in terms of their voting behaviour at the national level. Similarly, De Vries and Hobolt (2016) argue that EU issue voting in Germany is more pronounced in national elections than it is in European elections, with a plausible explanation for this Europeanization of national elections being Germany's leading role in the EU, which leads citizens to believe their vote and the resulting governing parties will not only determine domestic policy, but that they will also be decisive—or "pivotal", as Torcal and Rodón (2021) put it—in determining what will be implemented at the European level (Jurado & Navarrete, 2021). In this respect, because Germany is a net contributor to the EU budget, voters pay more attention to the positions of parties on EU issues simply because there are more economic costs at stake with EU policies (Jurado & Navarrete, 2021). Hence, from an economic voting perspective, voters in Germany have incentives to gather information about how decisions in Brussels affect them and, given that a significant proportion of policy is decided at the European level, will also take issues related to EU integration in general into account. For this reason, the voting behaviour of German citizens is influenced by their EU attitudes as well as by the positions parties take on EU integration, despite the lack of politicization of the European issue. Because national politicians in such an important and influential EU member state as Germany are in a better position to affect EU policy, German voters are likely to understand federal elections are also an instrument of EU accountability (Clark & Rohrschneider, 2009; Torcal & Rodón, 2021).

#### Politicization of the EU in the German Media and Parliamentary Debates

Recent studies analysing public debates and press coverage suggest that, despite the prominent role of German politics for the EU and decisionmaking processes in Brussels, the EU is not a source of polarization in the German media (Grande & Hutter, 2016; Silva et al., 2022), which is supported by our analysis using MAPLE data.

Figure 8.1 shows the average share of news the German press devoted to the EU and the tone of those articles during the four weeks leading up to the federal elections.1 Regarding the prominence given to the EU by German newspapers, it can be seen that the EU became significantly more important in the 2017 elections and then again in 2021. It is unclear whether the increase in the last elections was due to the use of different data sources or to the greater media attention to the EU while the Covid-19 pandemic meant a number of policies were being coordinated from Brussels. What can be said is that the data shows that during the electoral campaign the press did not devote much space to the EU.

Regarding the tone in the media reports, the stability observed in the Bundestag is very different to what we can observe when analysing the newspaper articles. In the elections after the European debt crisis, the way newspapers spoke about the EU was less positive compared to 2002 and 2005. In particular, in 2013, the first election after the formation of the AfD, the average tone used in reports about the EU was more negative than positive, which could also be related to the effects and handling of the European financial crisis. Nevertheless, these results must be treated cautiously given the values of the average sentiment expressed in these reports are consistently close to zero, meaning the share of positive and negative words in articles about the EU is very similar. However, this does not tell us much about the degree of polarization on the EU, because the apparent impartiality of the press could be the result of the neutrality of German newspapers in reporting the EU, or it could be a consequence of the mutual cancellation of the messages of a polarized press. To better estimate how the German press depicted the EU, Fig. 8.2 shows the share

<sup>1</sup> The data corresponds to news from the *Frankfurter Allgemeine Zeitung* and the *Süddeutsche Zeitung* for the elections running from 2002 to 2017. For the last federal elections of 2021, the news had to be collected from a different source and the newspapers considered were the *Süddeutsche Zeitung*, *Handelsblatt* and *Die Welt*.

**Fig. 8.1** Politicization of the EU in major German newspapers (*Note* The bars represent the percentage of news about the EU in the four weeks prior to the federal elections. The line indicates the average tone of the news referring to the EU. The tone was calculated using Rauh's [2018] sentiment dictionary for the German language. Data from 2002 to 2017 was collected from the *Frankfurter Allgemeine* and the *Süddeutsche Zeitung*. Data for 2021 is from the *Süddeutsche Zeitung*, *Handelsblatt* and *Die Welt*)

of news about the EU by newspaper, distinguishing between positive news (black) and negative news (grey). At this point, it has to be noted that the press outlets were selected to ensure coverage of a broad ideological spectrum: from the most widely read newspapers on the centre-right and the centre-left (Schmitt-Beck & Staudt, 2022). We see that the centreright leaning press (*Frankfurter Allgemeine Zeitung*, *Handelsblatt* and *Die Welt*) pay more attention to EU issues during the campaign compared to the centre-left newspaper *Süddeutsche Zeitung*. However, the share of negative articles about the EU is not associated with the ideological leanings of the newspaper, and in all six election campaigns the number of positive and negative news reports are well balanced. This evidence supports previous research that contends the national press in Germany is not polarized on the EU issue.

**Fig. 8.2** Percentage of news about the EU and share of negative news about the EU by newspaper (*Note* The black bars represent the percentage of articles about the EU that were more positive than negative. The grey bars represent the percentage of articles about the EU that were more negative than positive. The tone was calculated using Rauh's [2018] sentiment dictionary for the German language)

We have seen that the EU is not a polarizing issue in Germany and that political parties do not make it a major issue in their election campaigns. According to Schmitt-Beck (2017: 126), even in the 2013 elections the first involving the AfD—party strategists considered the Eurocrisis to be a "toxic" topic, so the EU issue was avoided. It might look as if the parties decided to hide the EU topic as a way to minimize the eventual controversy around it. But what about the speeches in the Bundestag? Did the parties refer to the EU when speaking in parliament? Do they also employ moderate language when speaking about the EU in parliament?

Figure 8.3 shows the average share of speeches mentioning the EU out of all the speeches delivered between 1998 and 2017. Also, the line in Fig. 8.3 represents the average tone used in parliamentary speeches referring to the EU, ranging from a negative to a positive tone in the respective speeches. The first finding to highlight is that the EU is much more salient in the Bundestag than it is in the German press. While the average share of parliamentary speeches about the EU never falls below 10 per cent after 1998 and was steadily above 15 per cent after the introduction of the Euro in 2002, the share of media news about the EU is significantly lower. This also means that with the monetary union, the EU is more important to Germany's economy and, consequently, representatives address this issue more often in their contributions to debates in the Bundestag. Since 2002, the years in which the EU was more prominent in parliamentary debates were 2004, when ten countries, including Germany's neighbours Poland and the Czech Republic, joined the EU; the election year of 2005 when Angela Merkel was first elected German Chancellor; and 2014, the year after the elections in which AfD almost achieved the electoral threshold for representation in the Bundestag.

To make their positions clearer to the electorate, parties emphasize their thematic priorities during election campaigns (Baumann et al., 2021; Jurado & Navarrete, 2021; Wagner & Meyer, 2014). The increasing salience given to one or more issue domains by party representatives in their campaign statements helps citizens make reasonable decisions by appealing to those issues that are important to them. Given that

**Fig. 8.3** Politicization of the EU in German Parliamentary debates (*Note* The bars represent the salience of the EU among the total number of parliamentary speeches. The line indicates the average tone of the total of the speeches mentioning the EU)

scholars have found the EU issue is important to German voters, despite most parties agreeing on EU policy, one might expect political actors in Germany could also behave strategically in the emphasis they give to questions relating to the EU and in the way they talk about it. Figures 8.4 and 8.5 show that parties indeed behave strategically in respect of the prominence they give European integration issues and in the way they talk about them in their discourses.

The first finding from Fig. 8.4 is that governing parties attach more importance than opposition parties to the EU in their contributions to parliamentary debates. In the period between 1998 and 2017, the Christian Democratic Union (CDU/CSU—Christlich-Demokratische Union) is the party that has been the longest in government and is, on average, the party with the highest share of discourses about the EU. Between 1998 and 2005, the Social Democratic Party (SPD—Sozialdemokratische

**Fig. 8.4** Salience given to the EU in parliamentary speeches, by party

**Fig. 8.5** Tone used in parliamentary speeches mentioning the EU, by party

Partei Deutschlands) and the Greens (Bündnis 90/Die Grünen) formed a coalition government under Gerhard Schröder. During those years, the Foreign Minister was the Europeanist Joschka Fischer of the Greens, who called for further European integration. It should come as no surprise, therefore, that during Schröder's two governments, the Greens made the EU more of an issue in the Bundestag (which tended to adopt less EUfriendly positions during the 1980s) (Debus, 2023). With the election of Angela Merkel in 2005, the Greens cut back on their parliamentary interventions about the EU and the CDU/CSU's coalition partner, the SPD, took over the foreign affairs portfolio to become the party that spoke most about the EU in parliament from 2005 to 2009, at which point the liberal Free Democratic Party (FDP—Freie Demokratische Partei) replaced the SPD as the CDU/CSU's coalition partner. Perhaps because of Chancellor Merkel's domination of EU issues, the FDP did not focus on EU issues in parliamentary debates, and with the beginning of the financial crisis Angela Merkel took the lead on EU relations, which is when the CDU became the party that gave more prominence to EU issues in its parliamentary contributions.

As for the tone used in parliamentary speeches that referred to the EU, it is possible to identify patterns in the behaviour of party representatives that can be summed up in three findings. First, while there is no negative tone towards the EU in the Bundestag between 1998 and 2017, there are clear differences between government and opposition parties. As Fig. 8.5 shows, governing parties speak more positively about the EU. Second, the distance in tone between those in government and those in opposition widens after 2009, which could be the result of the European financial crisis. Finally, between 2005 and 2008, the governing parties, the CDU/CSU and SPD, used almost the same tone when talking about the EU, then became increasingly distant in the election year of 2009. While the increasingly less positive tone used when talking about the EU could be related to the outbreak of the Eurozone debt crisis, the distance between the coalition partners in 2009 could just be a part of their electoral strategy. This view is supported after the 2013 elections, following which the CDU/CSU and SPD exhibited a broadly similar tone towards the EU until the federal election year of 2017, when they both began to diverge in their tone in respect of the EU.

The evidence presented here suggests German parties make the EU more or less salient in their speeches and adopt a more or less positive framing that seems to depend both on the ideological background of the parties and on strategical reasons. However, the differences are not that large and it could be difficult for uninformed citizens to be affected by the nuances of what was discussed about the EU in the Bundestag. While parties are consistently more positive than negative when talking about the EU in parliament, the tone used to address the EU is far from enthusiastic and actually mostly neutral. This could be related to the fact that in Germany, unlike what happens in other European countries, work experience in the EU is often a stepping stone to a career in domestic politics (Edinger, 2015) and the role of the "EU policy expert" has become significant within parties represented in the Bundestag (Kropp, 2010). Consequently, some parliamentary contributions about the EU tend to be informative and lack the terminology that contributes towards polarizing the debate around it. The EU is not—at least until 2017—an issue that sparks a great deal of dissent between the different groups in parliament.

In this respect, Schröder and Stecker (2018) note that if the public is not inclined to pay attention, voters have only a limited amount of time to follow the public debate while the media can address only a finite range of topics that are of interest to the public. Therefore, emphasising an issue on which the parliamentary parties are in broad agreement might be inefficient as it is strategically more difficult to get the media to pay attention, and without media coverage the issue can pass the electorate unnoticed. So the outstanding question is: do EU issues affect voting behaviour in national elections in Germany?

#### The 2021 Elections

The 2017 federal elections seemed to inaugurate a new era of political turmoil. The main parties, the CDU/CSU and SPD, attracted only 53 per cent of the votes, and—six months after the election—both parties agreed to renew the incumbent Grand Coalition (Bräuninger et al., 2019). In one of the most fragmented parliamentary party systems in Germany's post-war history, the shadow of a premature end to the coalition threatened the stability of Angela Merkel's final cabinet (Schmitt-Beck et al., 2022b). Representatives of six different parties were elected to the Bundestag, with the new radical-right AfD, which had representatives elected to the European Parliament and all state parliaments, emerging as the main opposition. In this difficult political context, Angela Merkel announced her intention not to seek re-election as Chancellor after the state election in October 2018. This announcement set the end date to the "Merkel Era", a period of sixteen years of only apparent stability during which the Chancellor faced several challenges at the European and national levels, as well as from within her own party, the CDU/CSU. For the first time in the history of post-war Germany and during the extraordinary circumstances imposed by the Covid-19 pandemic, citizens went to the polls without the incumbent Chancellor running for re-election.

The federal elections of 26 September 2021 resulted in both Grand Coalition parties hitting a new low in electoral support, with the CDU/CSU and SPD receiving less than half of all votes cast. The SPD and its candidate, Olaf Scholz, won the elections with only 25.7 per cent of the vote, seven points fewer than the CDU/CSU received in 2017 and more than fifteen less than Angela Merkel's best result in 2013. Scholz's experience as Finance Minister during Merkel's last government may have contributed to the SPD's victory, because he was the "actual incumbent" as he was the only candidate for the position of Chancellor who had government responsibilities at the national level. This is not a minor issue. Given Merkel's popularity during her mandate and the important effects of candidate evaluations on the vote in Germany (Hansen & Olsen, 2020), the CDU/CSU result suggests that its candidate, Armin Laschet, had not successfully convinced voters that he was Merkel's heir, with the result that he did not receive the traditional advantage of representing the party that occupies the office of the Chancellor.

In a highly fragmented parliament, the SPD formed a coalition with the third- and fourth-strongest parties: the Greens and the FDP. This coalition can be seen as a logical consequence of the expressed will of the SPD to avoid a new Grand Coalition; however, it could also be seen as an agreement between the three parties that presented some of the most important issues during the electoral campaign: overcoming the Covid-19 pandemic; managing the post-pandemic economic recovery; tackling climate change and the country's digitalization deficit. That being said, however, all three parties differ on issues related to finance and the economy, ensuring conflicts between them are likely to appear during the legislative period from 2021 until 2025 (Debus, 2022).

The EU was not a salient issue during the 2021 campaign, and even less so when the Recovery and Resilience Plan (RRP) set out by Scholz's ministry was approved by Brussels. The emphasis of the RRP lies in themes that were to be key issues during the election campaign: climate action; digitalization; growth; and jobs. So can we say that the position of citizens in the EU had an impact on the 2021 elections? In the sections below, we address this question and discuss the determinants of the votes cast in the last German elections.

### Determinants of German Voting Behaviour in the 2021 Elections

#### *Data and Methods*

To identify the factors that affected vote choice in the last federal elections, we rely on data from the post-electoral Maple online survey. This data includes 2002 respondents and was collected between 11 October and 21 December 2021. Using this dataset, we are able to analyse whether positions on European integration affected the vote for each of the parties represented in the Bundestag by considering both their sociodemographic characteristics and their positions on the main campaign issues. In doing so, we refer to voter recall as our dependent variable, which is divided into six categories that identify each of the parties elected to the German parliament in 2021, which include the CDU/CSU, SPD, Bündnis 90/Die Grünen, FDP and AfD.

As our interest lies on the impact of EU attitudes on voter choice, our main independent variable is "support for the EU" which in this case is conceptualized as a desire for EU integration. This is measured on an 11-point scale ranging from "The EU should be dissolved" to "The EU should move towards a United States of Europe".

We also include explanatory variables better associated with the German context and being central during the election campaign. These contain a variable to indicate respondents' views on whether immigration policy should be more restricted because of the significance of the antiimmigration agenda in the vote for the AfD. This variable is measured on a five-point scale, ranging from whether the respondents "strongly agree" or "strongly disagree" with the statement "Germany should implement a more restrictive immigration policy",2 with higher values on this scale indicating a more pro-immigration position. The climate crisis was another key issue in the campaign, so we include a variable designed to capture attitudes towards climate change on an 11-point scale ranging from "We should prioritize economic growth, even if it makes it more difficult to combat climate change" to "We should prioritize combating climate change, even if it hurts economic growth".<sup>3</sup>

Also, since the assessment of candidates is an important factor in terms of voting behaviour in Germany (as elsewhere), we also consider the popularity of the four major candidates, three of whom—Armin Laschet (CDU/CSU), Olaf Scholz (SPD) and Annalena Baerbock (Bündnis 90/Die Grünen)—were running as their parties' official candidates for

2 For the sake of an easier interpretation, we reversed the scale of the variable (Q24\_4\_w4) which originally runs from "strongly disagree" to "strongly agree".

3 The original variable in the dataset (Q75\_ger\_w4) runs from "We should prioritize combating climate change, even if it damages economic growth" to "We should prioritize economic growth, even if it makes it more difficult to combat climate change". We decided to reverse the scale to ease interpretation of results when assessing the impact of pro-climate attitudes.

the office of Chancellor (the FDP did not nominate a candidate for Chancellor, with FDP chairman Christian Lindner being the party's leading candidate [*Spitzenkandidat* ]). The popularity of these leaders is measured on an 11-point scale ranging from "strongly dislike" to "strongly like".

Because the German media is not divided on the EU issue, it could be that citizens using other sources for their information during the campaign are exposed to more polarizing content. In the case of social media, in the absence of a gatekeeper, misinformation is more easily spread. Even more, citizens often choose to engage only with content that reaffirms their pre-established beliefs or which is even more radical (Engesser et al., 2017; Nir, 2017). For this reason, we include a variable to measure the use of social media as a source of political information during the campaign as a way to consider how some polarizing issues on social media affect citizens' decisions without generating contestation by the main political actors nor attracting the attention of the traditional media. This variable captures how frequently respondents use social media networks, such as Facebook and Twitter, to obtain political information and is measured on a four-point scale ranging from "never" to "frequently".4 The higher the score on this variable, the greater the use of social media for campaign information.

Finally, as control variables we include a set of socio-demographic variables and one that refers to general political attitudes. These are gender (dichotomous), age, education (from low to high), household trade union membership (dichotomous), religiosity (four-point scale from "not at all religious" to "very religious") and left–right self-placement (11-point scale). To better interpret the magnitude of their effects, these and the previous independent variables have been re-coded on a scale of 0 to 1.

Given the nominal nature of our dependent variable, we use multilevel logistic regression models with the incumbent CDU/CSU as the baseline category. We run two models: the first includes all the variables described above; while the second contains an interaction on the EU issue with pro-immigration opinion as a way of capturing the actual net effect of our main independent variable and to test whether less favourable opinions towards immigrants made voters take the EU issue into greater consideration while voting.

<sup>4</sup> The original variable in the dataset (Q93\_5\_w4) had the following options: (1) Frequently, (2) Occasionally, (3) Rarely, (4) Never. The order of these items was reversed to ease the interpretation of results.

#### *Findings*

Did the EU issue affect voter choice in the 2021 German elections? Based on the results presented in Fig. 8.6, German citizens take European integration preferences into account when deciding how to cast their vote. The average marginal effects of the main explanatory variables included in the first multinomial model (see Table 8.1 in the Appendix) show that support for European integration is the variable that affects the likelihood of voting for three out of the six parties being studied. As Fig. 8.6 shows, a pro-EU position increases the probability of supporting the CDU almost as much as supporting the environment over economic growth raises the likelihood of voting for the Greens. However, being in favour of greater European integration reduces the probability of voting for the FDP or AfD. While these results are not surprising in respect of the AfD, the negative association between pro-Eu opinions and voting for the German liberals is striking, because the FDP's election manifesto advocated a new constitution that promoted a move towards a more federal EU.5 According to our results, this ambitious goal for the EU clashes with FDP voter preferences.

Going more into detail with regard to the AfD, we see that the effect of being against the EU on the probability of voting AfD is similar to that of being in favour of a more restrictive immigration policy. Moreover, opinion on immigration is a predictor of vote only for the AfD, because the likelihood of voting for any of the other five parties is not affected by an individual's stance on immigration policy. Our analysis also offers valuable information for creating a profile of the typical AfD voter, since all the main explanatory variables are in this case statistically significant. According to our results, AfD voters are significantly more right-wing, more anti-EU integration, more supportive of more restrictive immigration policies, more likely to prioritize the economy over protecting the environment and tend to use social media for their political information. In respect of this latter conclusion, we explained above that the inclusion of the variable measuring social media use was motivated because citizens could be exposed to more polarizing and engaging content on social

<sup>5</sup> As explicitly mentioned in the FDP's manifesto: "*Wir Freie Demokraten wollen nach Abschluss der Konferenz zur Zukunft Europas einen Verfassungskonvent einberufen. Dieser Konvent sollte einer dezentral und föderal verfassten Union eine rechtsverbindliche Verfassung mit einem Grundrechtekatalog und starken Institutionen geben*". https://www.fdp. de/nie-gab-es-mehr-zu-tun.

**Fig. 8.6** Explaining the vote in Germany (*Note* Average marginal effects and 95 per cent confidence intervals, based on the results reported in Table 8.1)

media, given that EU issues are not particularly salient in the German media which is not itself polarized on the EU.

Holding a pro-EU opinion increases the probability of voting for the CDU which was the incumbent party in the 2021 elections and the one that made the EU more salient in its parliamentary speeches. Nevertheless, being informed about politics via social media has a negative effect on the likelihood of voting for the CDU. Of course, this does not allow us to establish a clear link between Eurosceptic leanings and the use of social media, but indicates that the average AfD voter is less willing to support further advances on European integration and uses social media for campaign information more often than the average CDU voter.<sup>6</sup>

6 We replicated our analysis, including the interaction between the use of social media to gather political information and the support for the EU (see Fig. 8.8). For the CDU and AfD, the use of social media slightly affects the magnitude of the effect of the EU issue in the likelihood of voting for any of these parties, but it does not change the direction of the effect. This means the negative association between support for the EU

Concerning opinion on protecting the environment, as expected this is an issue that significantly affects the Green vote because more proenvironment positions increase the probability of casting a vote for Bündnis 90/Die Grünen. However, it is interesting to see that the vote for the FDP is also positively affected by positions in favour of prioritizing the environment over economic growth. This is an important finding, because having an electorate that coincides with that of the Greens in the issue the latter emphasizes most may have contributed to easing negotiations between the two minority partners in the governing coalition.

Finally, left–right self-placement is an important factor in explaining the vote for the parties on the opposite ends of the ideological scale: Die Linke on the left and the AfD on the right. For all the other parties with more mainstream views with regard to ideology, the left–right selfplacement of individuals does not affect the voting probability.

The analysis, including the interaction between opinion on immigration policy and support for the EU, does not present significant changes but offers a more fine-grained assessment of the effect of opinions on EU integration on vote choice. As shown in Fig. 8.7, opinions about immigration moderate the effect of issue voting for the CDU, FDP and AfD. The positive effect of supporting EU integration on the vote for the CDU is statistically significant only for those voters with a less favourable opinion of immigration. Surprisingly, the opposite is true for the FDP, where the negative effect of support for the EU on their vote affects only those individuals who support more restricted immigration policies. The same can be observed with the AfD whose voters are more likely to be critical of the EU, with the exception of those with more pro-immigration views.

# Conclusion

The new government coalition that emerged from the September 2021 election brought together three parties—SPD, Bündnis 90/Die Grünen and the FDP—with different political views and issue priorities. However, they share an ambitious view of the future of Europe, which was reflected

and vote for the AfD is stable for all individuals, irrespective of how much they depend on social media for their political information. Similarly, the positive association between EU support and voting for the CDU does not change as a result of the use of social media.

**Fig. 8.7** Average marginal effects of the EU issue at different opinions about immigration (*Note* Average marginal effects and 95 per cent confidence intervals, based on the results reported in Table 8.2)

in the coalition agreement and welcomed by Europhiles. This public political consensus about the EU does not make the EU issue irrelevant in voter choice at the national level and, as shown here, the position of citizens in respect of European integration is an important predictor of the vote in federal elections, despite the issue not being one that polarizes opinion.

We have shown that parties behave strategically when talking about the EU in parliament. While they are in government, German political actors make the EU more salient and depict it more positively than they do while they are in the opposition. This suggests that parties understand that, while there are no major differences between them on European policy, the EU is an important issue for citizens. When German voters select a party, they are choosing it by considering who will represent their national interests in Brussels and who will influence policy-making on the EU level. It is therefore important for citizens to envision a political party negotiating at different levels of the European political system. Thus, even as the media is not divided on the EU issue and this topic is often "de-thematized", even during election campaigns, citizens take their preferences with regard to EU integration into account when casting their votes in national elections. This supports what other scholars have found when analysing the impact of EU issue voting in German elections (De Vries & Hobolt, 2016; Jurado & Navarrete, 2021).

However, our analysis presents some limitations. The exceptional circumstances of the 2021 elections deserve more consideration and research. First, voters went to the polls after 18 months of the Covid-19 pandemic during which several of the government measures to control the spread of the virus faced popular opposition. In this context, the EU vaccination strategy and the coordinated purchase of medical equipment might have affected the way in which citizens viewed the EU. Unfortunately, we have no data on the parliamentary interventions during the last legislative period so we are unable to determine if these controversies were evident in the parliamentary debates.

Also, the speed at which German politics was changing may have had an impact on the answers respondents gave to the survey. The data used in this research was collected between 11 October and 21 December 2021, which means respondents were completing the survey as Armin Laschet resigned as leader of the CDU/CSU, the coalition deal between the SPD, Greens and Liberals was agreed and the new federal government took office. It would be disingenuous to state that respondents' perceptions, especially those concerning the popularity of candidates, could not have been influenced by subsequent events when the respondents completed the survey. Future research, undertaken in a less hectic political context and using more sophisticated panel data, should allow for testing the extent to which our findings on EU issue voting in Germany are correct.

#### Appendix

See Tables 8.1 and 8.2. See Fig. 8.8.


**Table 8.1** Determinants of vote choice in the 2021 elections

Multinomial regression models (Base category CDU/CSU); Standard errors in parentheses

*Data source* Maple online survey, wave 4 +*p* < 0.10, \**p* < 0.05, \*\**p* < 0.01, \*\*\**p* < 0.001


**Table 8.2** Determinants of vote choice in the 2021 elections including the interaction between opinion on immigration and support for the EU

Multinomial regression models (Base category CDU/CSU); Standard errors in parentheses *Data source* Maple online survey, wave 4 +*p* < 0.10, \**p* < 0.05, \*\**p* < 0.01, \*\*\**p* < 0.001

**Fig. 8.8** Average marginal effects of the EU issue at different levels of media use

#### References


B. Weßels, & C. Wolf (Eds.), *The changing German voter* (pp. 313–336). Oxford University Press.


**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# After the Crisis: EU Issue Voting in Greece

#### *Roula Nezi*

### Introduction

On the face of it, Greek politics has finally "returned to normal" as in 2019, after more than ten years of austerity politics, populism and legitimacy crises, the country elected a strong executive that enjoyed a sizable parliamentary majority. After a long period of polarization and division, it seems the political system has finally reached some kind of resolution.

Nevertheless, this apparent return to normalcy remains fragile. While everyday politics appear calm, public opinion is far less settled. There is still a lot of mistrust in the political system and democracy, while political apathy and cynicism remain widespread. There is a consensus among scholars that the 2015 referendum accentuated deep divisions in Greek politics, in a manner similar to the 2016 EU membership referendum in the United Kingdom and the 2014 independence referendum in Scotland (Hobolt, 2016). As was the case during the Brexit debate (Hobolt et al., 2020), a new identity divide emerged in the wake of the Greek

231

R. Nezi (B)

University of Surrey, Guildford, UK e-mail: s.nezi@surrey.ac.uk

<sup>©</sup> The Author(s) 2023

M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_9

referendum that has split society into opposing camps—those supporting the memorandum and those against it.

This cleavage between the supporters and the opponents of the memorandum has become the primary division around which other social cleavages are realigning. Research suggests that, along with cultural and national issues such as views on the European Union, social polarization and party preferences can reinforce ideological distinctions in society (Alwin & Tufi¸s, 2016; Druckman et al., 2021; Jacoby, 2014). In the Greek case, this polarization, which is based on national and cultural issues, is being gradually absorbed by partisanship and has become the primary cleavage diving citizens much more than any other salient issues (Robison & Moskowitz, 2019; Westwood et al., 2018).

The aim of this chapter is to examine the extent to which the vote choice of Greek citizens is driven by European considerations. In pursuing this question, we carry out a comprehensive examination of the role of EU issues on voter choice. We study voter choice in Greece during a critical period for the EU: during the economic crisis, the 2016 referendum on the UK's membership of the EU and the rise of populist and nationalistic sentiments. The Greek case provides a useful example of a country in which EU issues surrounding the economic crisis were responsible for polarizing public opinion.

### EU Issue Voting in Context

At the beginning of the European sovereign debt crisis, eurozone governments and the International Monetary Fund (IMF) agreed to provide temporary financial assistance to the three Member States hardest hit by the crisis: Greece, Ireland and Portugal. Soon after the first rescue package, and amid fears the debt crisis would spread to other indebted EU Member States, EU finance ministers funded the European Stability Mechanism (ESM).1 From a political perspective, allowing Greece, which was the first country to receive financial support, to become bankrupt would have been interpreted as the EU being unable to protect one of its oldest Member States, which could have serious repercussions on the European integration process (Ozturk & Sozdemir, 2015). The main aim of the bailout agreements between European institutions (European

<sup>1</sup> The ESM is a financial institution funded by contributions from other euro area Member States to support other eurozone countries in severe financial distress.

Commission [EC], European Central Bank [ECB] and the IMF, which became known as the troika) and national governments was to ensure a reduction in the debt of those countries hit by the crisis. This was to be achieved primarily through the promotion of a series of austerity measures, which included curbs on government spending, increasing direct and indirect taxes and property taxation (Nezi & Katsanidou, 2014). At the European level, the memorandum sought to prevent the eurozone from collapsing when the European sovereign debt crisis broke out in 2008.

From the beginning of the economic crisis, the political discourse was dominated by accusations of responsibility for the economic crisis and for blaming this or that party for signing yet another bailout agreement. During that period two antagonistic groups emerged: one that believed the memorandum was necessary to overcome the crisis; another that believed the memorandum was the reason why the crisis was so deep and prolonged. At the time, it was clear the austerity measures introduced created deep divisions within Greek society, creating a social polarization that was so profound it challenged national and European unity.

During that period, polls recorded strong sentiments of animosity towards Europe's leading powers that were seen to be supporting tough austerity measures as the solution to the crisis (Michailidou, 2017). For example, in Greece there was a wide popular belief that Germany was profiting from the Greek crisis (Allen & Chazan, 2018), beliefs that were being fuelled by existing experiences of victimization and a perception of victimhood among the general public (Antoniou et al., 2020). The economic crisis reinforced these beliefs, with issues of national identity competing with issues related to national self-image (Lialiouti & Bithymitris, 2017).

Yet, the crisis also fuelled stereotypes among the creditor countries. The mainstream media and radical right-wing parties in central Europe, such as the Austrian Freedom Party (FPÖ), painted a portrait of the "*lazy*" Greeks in contrast to "hard working" West Europeans. As a result, solidarity in the EU was mainly expressed within national borders (Kohut et al., 2012). Public protests and growing political discontent emerged to challenge the European orientation of the "debtor" countries as a response to the introduction of strict austerity policies (Karyotis & Rüdig, 2015).

Traditionally, Greek citizens supported Greek membership of the EU and tended to be positive about European integration. In the wake of the economic crisis in 2008, however, attitudes towards the EU changed dramatically, and from 2009 onwards the trend was clear: the vast majority of the Greek citizens have negative views of the EU, much more negative than the EU average. The EU's involvement in the implementation of austerity measures meant the issue of the economy and the EU became interlinked (Katsanidou & Otjes, 2016). However, this was not the only issue linked with the EU. The economic crisis also coincided with a refugee crisis, with thousands of refugees arriving on the Greek islands of Chios, Kos, Lesvos and Samos every day. Most important of all, this multifaceted crisis reinforced existing perceptions of the EU and feelings of resentment among Greeks, almost 80% of whom said they distrusted the EU.

Associating the state of the country's economy with developments at the EU level was commonplace among the general public, largely since the austerity measures were negotiated with European institutions. As a result, and due to the new divisions within the party system, new parties emerged and old parties collapsed. Between 2009 and 2019, the Greek party system changed drastically. Before the crisis, two parties alternated in government: left-wing PASOK and right-wing ND. Alongside them, smaller left- and right-wing parties would also gain some parliamentary representation, including the Communist Party and Syriza on the left, and the newly formed LAOS on the right. Traditionally parties competed along left–right ideological grounds; however, the modernization of the early 2000s caused old divisions to fade and new political issues to emerge. During the economic crisis, the traditional left–right dimension disappeared to be replaced by a schism around economic policies and the EU (Nezi & Katsanidou, 2014).

Greece historically was among those countries with high levels of support for the European Union and the process of European integration. From the beginning of the economic crisis, only the Communist Party (KKE) held a strong anti-EU position and openly supported Greece's exit from the European Union. All mainstream and governmental parties, such as PASOK and ND, as well as the radical right-wing party LAOS, supported the country's EU membership and the process of European integration.

The pro-European parties also supported economic reforms, while anti-European parties opposed austerity measures introduced at the behest of the troika (Nezi, 2012). This can also help us understand the coalition government that involved a party of radical left—Syriza—and the radical right-wing Independent Greeks. For the former, opposition to the austerity, and for the latter, a populist discourse on the EU, played a central role in their election campaigns.

The association of austerity measures with the EU has a number of implications for Greek politics. First, if the public did not become more Eurosceptic, they certainly became more critical of the EU. Second, the economic crisis led parties that otherwise supported European integration towards a more Eurosceptic position, which was certainly the case with ND while it was in opposition (Gemenis & Nezi, 2015; Lefkofridi & Nezi, 2020). It was this new divide that explains the coalition between Syriza and the Independent Greeks.

In June 2019, a month before the elections, it was clear ND was ahead in voting intentions and heading for victory at the polls. The party had a substantial lead over the other parties on all salient political issues, including austerity and the agreement between Greece and the Republic of Macedonia. The 2019 parliamentary elections were the first since the outbreak of the 2008 economic crisis in which the government did not have to implement a new set of austerity policies after Athens had become the final country to exit from the financial bailouts in August 2018. This resulted in economic news during the election campaign being generally more positive than it had been in previous years.

There is a broad agreement in the literature that the economy is an important driver of shifts in party support between elections. However, there remains some controversy over the specifics of the economic voting models during the economic crisis and the mechanism behind the economic conditions as a consequence of policies set out by international institutions in collaboration with national governments and the EU. Economic voting had become a positional issue and no longer a valance issue (Nezi & Katsanidou, 2014), with every debate on economic issues structured around the question of whether or not the Greek government will honour its agreements with the other eurozone and if Greece would leave the eurozone (Gemenis, 2013; Gemenis and Nezi, 2012, 2015).

Freire and Lobo (2005) and Nezi (2012) conducted early analyses of how macroeconomic conditions and individual preferences affect the vote in Greece. Data from the MAPLE pre-election study suggest exiting the bailout agreements has not translated into more general optimism. The vast majority of respondents believe the economy had deteriorated and three out of four believed the same about their personal economic circumstances. The balance of opinion regarding the performance of the incumbent in government has shifted sharply towards negative evaluations around the time of the July 2019 elections. Despite the efforts made by Syriza's leader Alexis Tsipras to convince the electorate the end of "memorandums" had been reached, voters continued to believe his coalition with the Independent Greeks had played a leading role in ensuring the extension of austerity measures and the implementation of additional taxes and austerity policies after his first term in power.

#### *Politicization of the EU in the Media and Parliamentary Debates*

Since the beginning of the economic crisis, there seems to have been a transformation in the political culture of the Greek public, one that has altered their values as a result of the changing economic conditions. This transformation seems to affect the position citizens take on political issues and has created a long-term tendency to alter existing patterns of political competition. This suggests political conflicts that are visible at the individual level are also likely to be present at the party level. We would expect to find evidence of these conflicts in the media and parliamentary debates, which implies the EU may have become an extension of domestic politics as a consequence of its involvement in the bailout agreement, domestic and European issues have come together on the issue of austerity politics.

Though the overall expectation would have been that the polarization on the EU issue would have been reflected or even driven by the media, existing data as shown in Table 9.1 suggests the opposite.

In our analysis, we examined two national daily newspapers, the centreright *Kathimerini* and the centre-left *Ta Nea*. Table 9.1 reports the salience of the EU issue, the tone and the percentage of negative attitudes during all elections from 2004 until 2019.

As expected, during the economic crisis years from 2012 onwards the salience of the EU issue is greater compared to the period prior to the crisis. While the salience of the issue is increasing, which can be explained by the austerity measures and the role of European institutions, the tone on average remains positive, while the percentage of negative articles is not significantly different compared to the period before the economic crisis (Fig. 9.1).

Typically, the EU issue has not been politicized by the mainstream parties in Greece, largely because both PASOK and ND are pro-EU.

**Fig. 9.1** EU politicization, 2004–2019

However, in recent decades, mainly as a result of the implementation of austerity measures, opposition parties have successfully employed the European issue in their appeals to voters who are against the austerity measures (Lefkofridi & Nezi, 2020).

Because of the MAPLE project, it was possible to measure the level of contestation on the EU issue in the Greek parliament from 2000—a period of prosperity—until 2019, the end of the economic crisis. Contestation is determined by whether a sentence discusses EU-related issues in a negative way.2 The trend is clear, the economic crisis sparked a wave of high levels of politicization of the EU issue, and Grexit (2015) was the peak. This result comes as no surprise: the public was divided on the issue, as were the parties. The divisions that emerged at the individual level are also visible in the party system.

In the literature, Europe is described as an issue that cuts across the traditional left–right dimension (Fieldhouse et al., 2021), and this became apparent during the economic crisis. The literature also suggests

2 For our purposes, we opted to present the tone of the articles and speeches mentioning the EU. In the case of the media, this refers to an average of two measures obtained in an automated fashion, through the combination of a measure of the sentiment of the article title and a measure of the average of sentiments in the sentences within the article mentioning the EU. In the case of parliamentary debates, this measure refers to the average sentiment in all sentences mentioning the EU in a given speech. Thus, once the EU sentences have been identified and translated into English, a sentiment score is calculated for each with the tone measure being the average sentiment score within those speeches.

mainstream parties will have fewer incentives to compete on the EU issue (Hooghe et al., 2002) while challenger parties will emphasize the extremes of the EU issue (De Vries & Hobolt, 2012). Figure 9.2 supports this hypothesis: higher levels of contestation are observed among challenger parties, particularly among Golden Dawn, the Union of Centrists and the Communist Party (KKE). Among the mainstream parties, the level of contestation is significantly lower, with the main party in the coalition government (Syriza) reporting the lowest score. This pattern can also be observed in those parties positioned to the left and the right of the ideological spectrum, with Golden Dawn and KKE scoring equally high.

Figure 9.3 provides a concise view of the evidence reported in Fig. 9.2. This is done by examining the tone of the parliamentary debates on the EU issue. The tone of the speech is the mean of the sentiment scores obtained for each sentence that mentioned the EU. The results indicate that on average both right-wing and left-wing parties debate EU-related issues in a positive tone. However, the factor that differentiates them is whether or not the party is in government or opposition. Syriza and

**Fig. 9.2** EU Contestation in parliamentary debates

Antonis Samaras's ND provide a clear illustration of this. During the economic crisis, when the EU issue was politicized, these two parties in opposition discussed EU-related issues in a less positive way compared to when they were in government. Parties in government needed to negotiate with European institutions on austerity policies and, following Peter Mair's argument, they had to be responsible and respectful towards international institutions such as the EU. This finding is consistent with existing literature on contestation on the EU issue. Radical parties, and particularly radical right-wing parties, have become associated with negative views of the EU in an attempt to gain electoral support (Down & Han, 2021).

By 2019, the majority of Greek citizens (59%) once again had a positive view of the EU. While the party system had been substantially polarized around the EU issue, by 2019 there was a general feeling the worst had passed. Consequently, the issues dominating the political campaign were not only about austerity measures or Greece's membership of the EU.

**Fig. 9.3** Tone in parliamentary debates

### Modelling EU Issue Voting in 2019

The following section provides a series of competing models aiming to examine the impact of perceptions about the EU on vote choice while controlling for a series of issues monopolizing the political debate. In 2019, for the first time in almost a decade, the issue of austerity politics is no longer the sole issue of the election campaign. In contrast, national issues, such as the Prespa Agreement, are at the forefront. The models presented below examine the explanatory power of attitudes towards the EU in contrast to issues related to the economy and two national issues the agreement with Skopje and the law banning police from entering university campuses, which is commonly known as the asylum law.

Towards the end of the Syriza-led coalition's time in office, government austerity no longer monopolized the political discourse. In August 2018, European leaders heralded Greece's exit from the international bailouts, marking the end of the eurozone's financial crisis. Greece was the final eurozone country to conclude the bailout agreement. While similar help was given to Portugal, Ireland and Cyprus, Greece's crisis was deeper and longer lasting. Compared to the events and disputes marking the vote for the first bailout agreement, the end of the bailout agreement went almost unnoticed.

Why did the Greeks vote in favour of the conservative Mitsotakis over Tsipras? The 2019 national elections were Greece's sixth since the beginning of the financial crisis in 2008. The elections ended the premiership of the left-wing populist leader who had promised to end austerity in 2015. Instead, Tsipra's party implemented the toughest bailout package in the most iconic year of the crisis that was marked by a referendum, bank closures and two elections. If the general proposition that economic conditions affect party choice is accepted, then it follows that ND was not necessarily in an enviable position in 2019. The main economic indicator in ND's favour was the unemployment rate. During the months preceding the elections, unemployment was relatively high (18.5%) compared to the EU average, and it remained high during throughout Syriza's time in office.

#### *Dispute Between Greece and North Macedonia*

The economic crisis was no longer making front-page new; however, an old issue returned to prominence. In 1991, following the collapse of Yugoslavia, the Republic of Macedonia declared its independence. Since then, Greece has objected to it using Macedonia in the country's name, arguing that doing so implied it had territorial claims to the neighbouring northern Greek province of Macedonia, which was associated with the legacy of Alexander the Great and his empire. Greece blocked Skopje's accession to NATO and prevented it from integrating more formally with Europe, because it was using the name "Macedonia" without having any clear historical identity and cultural heritage to associate it with it.

However, after almost 27 years of mutual distrust, the Prespa Agreement was reached between Athens and Skopje, that saw the latter agreeing to call itself the Republic of North Macedonia in exchange for Greece dropping its opposition to its neighbour from joining NATO and strengthening its relations with the EU. In the summer of 2018, it was not clear how the electorate would react to the Prespa Agreement; however, one year later it was evident that one out of two respondents believed the agreement did not serve Greece's interests at all. Admittedly, the months that followed were not easy for the government. As expected, there were massive demonstrations in both countries, while the conservative nationalist party ANEL abandoned the coalition government in opposition to the agreement.

The asylum law was initially implemented to protect free speech and academic freedom following the collapse of the military dictatorship. In recent years, however, many came to believe the law no longer serves its purpose, and that it had helped create a culture of violence in Greek universities. Rescinding this law formed a part of Kyriakos Mitsotakis's election manifesto.

#### *Explaining Voting in Greece's 2019 General Elections*

Multinomial regression analysis is employed where the dependent variable measures vote choice for the 2019 national election. The dependent variable is a categorical variable with voting choices being Syriza, the party in government, controlling all other parties in the parliament, namely ND, KINAL, KKE, Greek Solution and MERA25.

Traditionally, vote choice in Greece was the outcome of left–right orientations and attitudes towards the economy. As was the case in many other European countries, until recently party attachment was aligned with pre-existing historical divides, which created two main camps: those on the left and those on the right of the ideological spectrum. When this divide began to fade with the emergence of new issues in the political landscape, attitudes towards the economy and matters of accountability on economic-related policies began shaping vote choice. The first model controls for these two dimensions of Greek politics while at the same time controlling for a series of socio-demographic characteristics, such as age, gender, education, trade union membership and religiosity. Political ideology is measured as self-placement on the left–right (L–R) spectrum, where zero (0) corresponds to left and ten (10) to right. Individual perceptions about the economy were measured by asking participants in the survey to evaluate the state of the economy compared to a year ago—with higher values indicating positive evaluations. This model also includes individual attitudes towards the EU. Since the beginning of the crisis, the economy and the EU issues were closely linked, and Greek citizens become strongly critical of the EU and European institutions. Attitudes towards the EU are measured using a ten-point scale variable, where zero (0) indicates opposition to the concept of European integration and ten (10) strong support for it.

Figures 9.4 and 9.5 present the results of the analysis. As expected L–R ideology, economic perceptions and attitudes towards the EU did determine vote choice. The same holds for the national issues of the Prespa Agreement.

The issue of Europe differentiates vote choice between Syriza and its main competitors—ND and the centre-left KINAL. Both ND and KINAL—formerly known as PASOK—supported Greece's EU membership and campaigned for a "Yes" vote during the referendum, while Tsipras's government was campaigning for a "No". The question posed to Greek voters during the referendum was whether or not to accept the

**Fig. 9.4** Explaining the vote in Greece: EU and the economy (*Note* Average marginal effects and 95% confidence intervals, based on the results reported in Model 1 in Table 9.1)

economic proposals put forward by the ECB, EC and IMF. For Syriza, the referendum was a powerful weapon during the negotiation period and an opportunity for direct democracy, while for the Independent Greeks and the Golden Dawn, it was an opportunity to exert national sovereignty (Rori, 2016). The complicated wording of the question reflected and endorsed the divisions that emerged during the crisis. While the supporters of "Yes" framed the referendum as a vote for Greece's EU membership and as a vote against Grexit, for the "No" supporters, it represented an opportunity for Greeks to express their opposition to austerity and regain their national pride and dignity (Crespy & Ladi, 2019). To this end, the issue of Europe is reflected in the two camps formed during the referendum.

**Fig. 9.5** Explaining the vote in Greece: competing issues (*Note* Average marginal effects and 95% confidence intervals, based on the results reported in Model 2 in Table 9.1)

Ideology has always been the most important predictor of vote choice in Greece (Nezi, 2012), which this analysis confirms. Voters identifying as right-wing will not support Syriza and will instead vote for either ND or the new radical right-party Greek Solution. Ideology will not differentiate the vote between Syriza and the parties that are close to it, such as KINAL and MERA25, this latter being the left-wing party formed by Syriza's former Finance Minister, Yanis Varoufakis.

This analysis confirms previous studies that have examined the importance of the economy in voting behaviour in southern Europe, and in Greece in particular (Freire & Lobo, 2005; Kosmidis, 2014; Lewis-Beck & Nadeau, 2012; Nezi & Katsanidou, 2014). What all studies confirm is that Greek voters will "throw the rascals out" when they believe the economy is deteriorating. From this perspective, Syriza's victory in the September 2015 elections was paradoxical, as just a few months earlier Tsipras had signed another bailout agreement that introduced a range of harsh austerity measures. The explanation lies in the fact there was not enough time between elections for the economic policies to be felt by citizens in their everyday life (Tsatsanis & Teperoglou, 2016).

The second model (Table 9.1) includes a series of salient domestic political issues that monopolized the 2019 election campaign, including the Prespa Agreement and the asylum law. The asylum issue differentiates the vote between Syriza and ND, which has traditionally been in favour of abolishing the law.

Of the two national issues examined, the issue of the name of the country's northern neighbour had a strong impact on Syriza's electoral fortune. With the vast majority of people opposing the agreement between Athens and Skopje, it is unsurprising that those who oppose the agreement will not support Syriza.

All scenarios tested above reinforced the hypothesis that Europe and the economy are the two most important issues differentiating the vote for Syriza and for all other parties. Negative attitudes towards the EU strengthened the vote for Tsipras's party, while positive attitudes tended to enhance the vote for the centre-right ND. The same holds true for the agreement that ended the stand-off between Athens and Skopje over the name of Greece's northern neighbour. Negative attitudes towards the Prespa Agreement almost certainly drove some citizens into the arms of ND.

### Conclusion

The aim of this chapter was to examine how, and to what extent, the issue of Europe determined vote choice in Greece in 2019. Historically, Greek public opinion has been generally supportive of both European integration and Greece's membership in the EU. However, the political and financial implications of the economic crisis after 2008 "aroused" the sleeping giant and, gradually, after a sequence of events created by the introduction of harsh austerity measures, the public eventually turned against Europe and its leaders, with the issue of Europe becoming bound to the issue of austerity, thereby creating a super issue that divided both society and the political parties into two camps: those that supported the memorandums and those that opposed them. From this perspective, policy positions were attached to strong feelings of animosity, with evidence suggesting the emergence of affective polarization.

This chapter also contributes to the systematic understanding of polarization on the EU issue in parliament and the media. The trend is clear, closer to the referendum the debate around the EU issue that generated high levels of polarization. This comes as no surprise to the analysts who followed developments in southern Europe during the economic crisis. By analysing two widely read newspapers, one representing centre-left opinion and the other centre-right, we see the salience of the EU issue is greater during the crisis, and that negative articles frequently appeared in even moderate newspapers.

This chapter also adds new insights into the determinants of party choice for the 2019 elections. Has Alexis Tsipras's party been punished for its ambivalence towards the EU and for signing the bailout agreement? Or does the average voter perceive the agreement with the Republic of North Macedonia as a threat? In 2019, Syriza successfully closed the circle of the memoranda that was opened by the Papandreou government in 2009 and which could only have provided major benefits for Tsipras and Syriza. However, there was also a considerable potential downside with it seeming the average Greek voter believed Tsipras was responsible for Greece's economic plight. In addition to the economic situation, at the same time ND mobilized citizens and public emotions by calling on people to "rally around the flag" in support of its leader, Kyriakos Mitsotakis. As the analysis suggests this strategy was ultimately successful.

### Appendix

See Table 9.1.



#### References


**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# Salient But Not Polarized: The Role of the EU in the Irish Electoral Arena

#### *Lea Heyne*

# Introduction

The February 2020 Dáil elections in Ireland have not been an ordinary election: First, they are the first real post-crisis election—in the Irish case, even a post "dual crisis" election (the Eurozone crisis and Brexit). And second, for the first time ever, Sinn Féin won the most votes, while the two formerly dominant parties, Fianna Fáil and Fine Gael, shrank to a fraction of their former strengths, and the government to emerge was a coalition between these two previously irreconcilable enemies (Cunningham & Marsh, 2021). For these reasons, the election marks the end of an era in Irish politics (Gallagher et al., 2021). Thus, it is especially interesting to understand the relevance of the EU and Euroscepticism in this election.

This chapter will begin by discussing Ireland's relationship with the EU as well as the role of Euroscepticism in Irish politics, to then explain the

L. Heyne (B)

© The Author(s) 2023

251

Institute of Social Sciences, University of Lisbon, Lisbon, Portugal e-mail: lea.heyne@ics.ulisboa.pt

M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_10

general political context of the February 2020 Dáil elections, discussing the relevance of social issues, the EU and Brexit. Then, we will present data on the salience and the tone of the EU in media and parliamentary debates in Ireland in 2020, in order to understand how important the EU was during electoral campaigns, and during the previous parliamentary year for parties. Next, we move to an analysis of voting behaviour in the February 2020 elections, using survey data to determine which factors mattered most in driving vote choice amongst the Irish electorate, and end the chapter with conclusions.

### Ireland's Relationship with the EU: Stable Support Despite the Euro Crisis and Brexit

Irish politics are traditionally marked by high levels of public support for the EU, based on a broad appreciation of the benefits of EU membership. Economically, EU membership has been a key factor in the modernization, diversification and growth of the Irish economy (Murphy & Hayward, 2009) Moreover, being equal partners in the EU has helped to "disrupt and dilute the historically asymmetrical relationship between Ireland and the UK" (Murphy, 2021, p. 105). Indeed, public support for the EU in Ireland has been consistently above the EU average since the 1980s, and did not suffer significantly since the onset of the Eurozone crisis (Simpson, 2019a). Although there has been a drop in trust towards the EU since the onset of the financial crisis, the Irish are still amongst the countries that rate the EU most positively and report a higher-thanaverage attachment to the EU (Galpin, 2017). While many Irish were disappointed and angered by the performance of the EU during the economic crisis, in particular, those who experienced increased economic instability, overall, most Irish remain enthusiastic supporters of the EU project (Simpson, 2019a, 2019b). A possible explanation for this fact is that the financial crisis was understood most widely as a domestic crisis in Ireland, and attributed to longstanding problems in Irish society, particularly amongst the political and economic elite: "Having revealed widespread corruption in the country's banking, economic, and political system, the crisis was seen as one brought about primarily by the Irish elite themselves" (Galpin, 2017, p. 140).

But apart from a generally positive view amongst the electorate, issues having to do with the European Union are "a minority interest in Ireland" (Murphy, 2021, p. 100). In other words, unlike in the UK, Europe is not an issue that mobilizes Irish public opinion in any significant way. As Murphy (2021, p. 101) remarks, there is also a certain disconnect amongst Irish between "positive perceptions of the EU and low levels of knowledge," with voters typically "taking their cues from political actors including political leaders, parties, interest groups and movements" (*ibid.)*. Both main parties, Fianna Fáil and Fine Gael, have been consistently pro-EU since its beginnings (Benoit, 2009; Hayward & Fallon, 2009). The Labour party was initially rather opposed to the EU, but has become clear, if sometimes critical, EU supporters during the 1980s (Murphy & O'Brennan, 2019). The Green Party shows a similar development, but remained eurocritical until well into the 2000s. Sinn Féin's position on the EU has been more complex—after a very strong rejection of European integration in the name of national sovereignty in the 1970s and 1980s, the party has started to become more nuanced since the 1990s (Maillot, 2009). They now combine a support for continued EU membership with a core element of soft Euroscepticism, including opposition to the EU's economic and social agenda, which has led them to campaign against all Irish EU treaty referendums (Murphy, 2021; Murphy & Hayward, 2009). While Sinn Féin, together with a variety of civil society anti-EU movements, has clearly channelled an anti-EU narrative during those referendums in 2001 (Treaty of Nice) and 2008 (treaty of Lisbon), none of them ever advocated for an outright Irish exit from the EU (Hobolt, 2005). And while the initial No-vote in both referendums shows that there is a certain level of Euroscepticism amongst the Irish electorate, researchers have concluded that "the depth and intensity of opposition to the EU in Ireland is not deep-rooted" (Murphy, 2021, p. 103).

Despite the severity with which the financial crisis hit Ireland in the years following 2008, and the "earthquake election" of 2011 (Marsh & Mikhaylov, 2012) that swept away the long-standing incumbent Fianna Fáil, the impact of the Eurocrisis and the bailout agreement on public sentiment towards the EU was similarly minimal (Simpson, 2019a). In contrast to the two previous EU referendums, and fearful of the consequences of rejecting another EU treaty, the Irish electorate supported the Fiscal Treaty in a 2012 referendum (Murphy & O'Brennan, 2019). Moreover, public protest against austerity and the EU was limited, and largely concentrated around internal policy proposals such as water rights and housing (Fitzgibbon, 2013). In sum, the Irish political system has accommodated public displeasure with the EU and there remains a strong pro-EU consensus within the Irish political system and across society more broadly (Murphy & O'Brennan, 2019). Interestingly, while a growing Euroscepticism since the Eurozone crisis has been observed in media and elite discourses (Gora, 2018), there are no indications of a similar lasting trend in the Irish electorate (Simpson, 2019a).

Even Brexit could not fundamentally change this dynamic, despite having a very direct and profound impact on Ireland, and sparking concerns about the stability of the Northern Ireland peace process over a re-opening of the inner-Irish border issue. But all major political parties in Ireland made it clear since the UK's 2016 Brexit referendum that they were hoping for a Remain decision, and subsequently shared a similar perspective on how the Irish government should approach the UK−EU negotiations (Murphy, 2021; Murphy & O'Brennan, 2019). In essence, "Brexit has been quite remarkable for having generated so little division and disagreement in Ireland" (Murphy, 2021, p. 95). As a result, Brexit—and the EU—has been generally considered as a topic of minor importance for the Irish election campaign of 2020 (Cunningham & Marsh, 2021).

### The Elections of 2020---An End to Irish Exceptionalism?

Traditionally, the Irish party system is strongly weighted towards the centre-right and structured around a competition between Fine Gael and Fianna Fáil to lead government coalitions with smaller parties, recently mostly the Labour party (Little, 2021). The party system was "defined and structured by the integrity of the quarrel between Fianna Fáil and Fine Gael, even if its origins lay in events of the early 1920s and policy differences between the two parties were increasingly harder to discern" (Gallagher et al., 2021, p. 5). The economic crash of 2008 set in motion what seems to be the end of this system: the 2011 elections brought a clear punishment of the incumbent, Fianna Fáil, mostly as a result of public dissatisfaction with the economic crisis and the Troika-led bailout programme that the government signed in 2010. However, the 2016 elections partly normalized the situation again, with Fine Gael's results reverting back to their normal range, and Fianna Fáil recovering some ground (Gallagher & Marsh, 2016). Tolerated by Fianna Fáil, Fine Gael formed a minority government in coalition with independent members of the Dáil. A version of the traditional two-and-a-half-party system seemed to persist, with Sinn Féin taking Labour's traditional position as the half-party (Field, 2020).

Under the pressure of the unfolding Brexit process which threatened Ireland's fragile political equilibrium, this minority government remained stable until late 2019 (Little, 2021), despite having been rocked by scandals for much of its duration (Field, 2020). The focus on Brexit dictated the pattern and focus of Irish politics after 2016 and "distracted attention from other domestic policy priorities and consequential wider global developments" (Murphy, 2021), despite being more important amongst political elites than voters. Yet, with the finalization of the Withdrawal Agreement between the UK and the EU in late 2019, support for the government soon decreased, and social issues moved back into focus. While the economy had recovered between 2016 and 2019, healthcare, retirement and housing took the place of unemployment and the economy as the public's main concerns (Little, 2021). When a series of scandals culminated in motions of no-confidence against the housing minster as well as the health minister, the Government decided to call for new elections instead of risking a defeat in early 2020. In this climate, Sinn Féin managed to run a very successful campaign strongly based on social media (Park & Suiter, 2021). The party opposed the planned increase in retirement age, an issue that became central to the electoral campaign, to the surprise of the two centre-right parties. Moreover, Sinn Féin also focused on state intervention, welfare spending and public housing, taking a clear opposition to Fianna Fáil and Fine Gael, and successfully drew an "anti-elitist populist narrative" (Park & Suiter, 2021). While Fine Gael pursued a broadly traditional centre-right platform of tax cuts and economic liberalism, Fianna Fáil positioned itself slightly to their left, combining tax cuts with increased cash transfers in social welfare and promising to address the housing crisis through measures to increase private home ownership (Field, 2020).

As a result of this focus on social issues, Brexit was much less salient during the 2020 elections than it had been in the previous years—only 1% of exit poll respondents indicated it as the most important issue in the elections (Field, 2020). The two most important issues were clearly health (32%) and housing (26%), followed by pension age (8%), jobs and climate change (6% each). The fact that Brexit was not an influential issue at the polls was certainly to the disadvantage of Fine Gael, which was perceived to have handled it competently (Little, 2021), and was the only political party that sought to mobilize electoral support around its record on Brexit and its capacity to steer the country through the next phase of negotiations (Murphy, 2021).

At a low turnout of 62%, the 8th of February elections resulted in vote shares of 24.5, 22.2 and 20.9% for Sinn Féin, Fianna Fáil, and Fine Gael, respectively. Independent candidates—traditionally a very important factor in Irish elections—gathered 12.2%, while the smaller parties came out at 7.1 (Green Party), 4.4 (Labour Party), 2.9 (Social Democrats), 2.6 (People Before Profit), 1.9 (Aontú) and 0.4 (Independents 4 Change) %. These elections have thus clearly shown that the Irish political landscape has changed, continuing a pattern that began in the two preceding elections: Sinn Féin became the largest party by vote share for the first time, only failing to become the largest party in the Dáil (lower house) because it underestimated its own electoral potential and therefore selected too few candidates (Field, 2020; Little, 2021). As Fianna Fáil and Fine Gael both refused to collaborate with Sinn Féin, they were left with only one option—a coalition with each other, which, including the Green Party, finally secured a majority in the Dáil. This formal coalition is another novelty in the Irish party system, although the two major parties did have a history of cooperation, like Fianna Fáil supporting the Fine Gael minority government under a confidence and supply agreement since 2016. Overshadowed by the sudden onset of the COVID-19 crisis shortly after the elections in February, government formation took a record 20 weeks, and a minority caretaker government remained in power until June 2020, when the new coalition government finally took over. Interestingly, government and opposition are now divided along right−left lines more than at any other moment, suggesting that "Irish exceptionalism in the structure of its party politics may be coming to an end" (Little, 2021).

#### Politicization of the EU in the Irish Media and Parliamentary Debates

#### *The Media*

To understand better how the EU is portrayed in the Irish media, we analyzed the two main Irish newspapers: The Irish times, a liberal mainstream newspaper, as well as the Irish independent, a rather conservative and anti-elitist Tabloid newspaper. For both newspapers, we have data from 2002 to 2020, which covers the three months before each election. We look at both EU salience (measured as the % of articles that mention the EU) as well as EU tone (the tone of those articles mentioning the EU). The tone measure refers to an average of two measures obtained in an automated fashion. Namely, it combines a measure of the sentiment of the title of the article, and a measure of the average sentiments of the EU sentences in the article itself. Thus, once the EU sentences have been identified, a sentiment score is calculated for each of them, and the tone measure is the average sentiment score within those speeches.

Looking at the two newspapers, we can see a similar development over time: While the salience of the EU has steadily increased since 2002, the tone has become more negative. This is true for both newspapers, although the salience of the EU is slightly higher in the Irish Times, while the tone is slightly more negative for the Irish Independent. Both newspapers have evolved from a rather low salience of the EU combined with a positive tone in the 2002 and 2007 elections, to an increasing salience and gradually more negative tone since the 2011 elections. Clearly, the Eurocrisis with the following bailout and the austerity politics have caused Irish media—both mainstream and tabloid—to talk more about the EU, and to be more critical. The 2016 and 2020 elections see even higher levels of salience while the tone remains neutral to negative. This effect is certainly strongly influenced by the debate about Brexit, which started in 2016 with the British referendum, and was very salient again in the months previous to the February 2020 election, given that a Brexit agreement became finalized after long and difficult negotiations in December 2019 (Fig. 10.1).

#### *Parliamentary Debates*

Next, we look at the role of the EU in parties' discourses in parliament. Here, we have data from 1997 to 2019, which covers each year that a party has been present in the Dáil, the lower house of the Irish parliament. Again, we are comparing the salience of the EU—as a percentage of speeches that mention the EU out of all speeches made by members of parliament from that party—as well as the tone of those speeches, ranging from positive to negative. In the case of parliamentary debates, this measure refers to the average sentiment in all the EU sentences uttered in a given speech. Looking at the right-wing parties first, we can see a similar development when it comes to salience, which has slowly increased in both parties' speeches since the late 2000s, again showing

**Fig. 10.1** Politicization of the EU in Irish media

a clear effect of the Eurozone crisis. However, the EU tone shows a different development: Fianna Fáil started off with a very positive tone about the EU, but shows a clear trend to the negative since 2007, which intensifies from 2011. This is in line with Fianna Fáil being in government until 2011, when they lost the elections in a landslide, mostly due to their role in implementing the Troika-led bailout and austerity policies. Clearly, the party has since turned away from their strongly pro-European stance, and used their role as an opposition party to show a more critical discourse on EU issues. Fine Gael, at the same time, had a more neutral EU tone during the 1990s and 2000s, and then shows a development to a more positive tone since 2011—the year they won the elections and became incumbent. In general, Fine Gael is considered to be the most pro-EU political party, and it tends to outperform Fianna Fáil at European Parliament elections. This comparison of the two major right-bloc parties shows again that the Irish bailout and the following elections in 2011 changed the party dynamics as well as the way the EU is discussed in parliament substantially.

Turning to the parties of the left bloc now, the situation is overall similar, pointing to the fact that the EU is not an issue that aligns strongly with the left−right divide. Overall, almost all left-wing parties follow the trend of an increasing salience of EU topics in parliamentary debates, especially since the Eurozone crisis. When it comes to the EU tone, the development is different again. Looking at the Green Party and the Labour party—two traditionally rather pro-European parties—we can see that after an increasingly positive EU tone during the 2000s, the tone turned more negative in recent years, following the Eurozone crisis. In the case of the Green party, the negative trend in tone started in 2008, while the Labour party only turned more critical of the EU from 2015. Both parties show the most negative tone on the EU in 2019. Sinn Féin, the most Eurosceptic amongst the mainstream parties in Ireland, shows a steadier trend, with a mostly neutral tone on the EU throughout the 1990s, 2000s and 2010s, pointing to the fact that that the Eurozone bailout has not changed Sinn Féin's position on the EU—which makes sense given that it was rather critical in the first place. Since the 2016 elections however, we can see a trend to a more positive discourse on the EU, which is in line with the party adopting a more pro-European approach and clearly opposing Brexit. Lastly, People Before Profit (PBP) and the Social Democrats, two challengers left parties that grew strongly after—and partly in reaction to—the Eurozone crisis, show a similar trend of relatively high levels of EU salience, together with a rather negative tone on the EU since their entry into the Oireachtas in 2011 and 2016, respectively (Fig. 10.2).

### Determinants of Irish Voting Behaviour in the 2020 Elections

#### *Data and Methodology*

As the other country-case chapters, the analysis of voting behaviour in the 2020 Irish elections is based on data from the two-wave Maple online survey fielded before and after the election, and relies on the second, post-electoral, wave. In Ireland, this wave contains 998 respondents and was fielded between February 17 and April 05, 2020. This section will focus on the major factors explaining voting behaviour for each of the main parties. To do this, we create a model of voting behaviour, including sociodemographic variables, ideology, economic perceptions and leader effects. Our goal is to contrast different issues and their importance for the vote choices.

**Fig. 10.2** Politicization of the EU in Irish parliamentary debates

Our dependent variable is vote recall, capturing respondents' vote choice in the previous elections. We use a categorical variable that captures the vote choice for each major party.1 As independent variables, we use gender, age (in years, from young to old), education (8 categories, from low to high), being trade union membership (respondent and household), and religiosity (4 categories). When it comes to political attitudes, we test for ideology (left−right-placement, 11-point scale), assessment of the national economic situation (5-point scale), and opposition to immigration (support for a more restrictive immigration policy, 5-point scale). Support for the EU is measured with the following indicator:

Some people believe that the process of European integration should move forward to the creation of the United States of Europe. Others believe that the European Union should be dissolved in order to return to a situation in which states are fully sovereign. In which point of the following scale would you place yourself? (0. The EU should be dissolved—10. The EU should move towards the United States of Europe)

Moreover, we also test the effect of an Ireland-specific question we included in the survey about retirement age, which was one of the most salient issues in Ireland at the time of the elections:

To what extent do you agree with the following statement? "Workers should be able to retire at 65, reverting past legislation which extended the pension age to 68, even if that could have future financial implications for the sustainability of the Irish welfare state."

Lastly, we test for leader effects using a battery of questions that ask respondents how much they like a political leader (on an 11-point scale from "dislike strongly" to "like strongly"), we use the party leader(s) of the respective party in each analysis (apart from PBP who have a centralized leadership). All independent variables are recoded to a 0 to 1 scale to ease comparison of the effect sizes. Given that our dependent variable,

<sup>1</sup> We only analyze the vote for parties that have at least 10 respondents indicating that they voted for them. Due to too few cases, we had to exclude Aontú and Independents4Change from the analysis. The remaining parties are Fianna Fáil, Fine Gael, Labour, Sinn Féin, the Social Democrats, the Green Party, People Before Profit, as well as independent candidates.

vote choice, is categorical, we use multinomial logistic regression models, with the incumbent party, Fine Gael, as a baseline category.

#### *Results*

Table 10.1 in the appendix shows the results when predicting the vote for the seven parties we analyze (Fine Gael, Fianna Fáil, Labour, Sinn Féin, the Social Democrats, the Green Party and People Before Profit) as well as for independent candidates, comparing voters of each opposition party to the baseline, the incumbent Fine Gael. Model 1 includes sociodemographics, ideology and political issues, and Model 2 adds economic perceptions as well as leader effects.

When it comes to sociodemographic characteristics, some effects stand out: namely, compared to Fine Geal voters, voters from most left parties (Sinn Féin, Labour, Green Party as well as PBP) are younger, while Independent voters are slightly older. Sinn Féin and Labour voters are also less educated. Trade Union membership is not significant. A higher religiosity clearly distinguishes Fiánna Fail voters from Fine Gael voters. Gender, just as trade union membership, does not significantly affect vote choice for any party in our model.

Moving to political attitudes and issues, we can see a clear effect of ideology on the vote for almost all parties compared to Fine Gael: Fiánna Fail voters are slightly more left-leaning, and voters for the leftbloc parties (Sinn Fein, Labour, the Green Party, and especially the Social Democrats and PBP) as well as the Independents are significantly more left-leaning. Sinn Féin is the only party that benefits from EU issue voting, by attracting more Eurosceptic voters compared to Fine Gael, but also Independent voters tend to be more Eurosceptic. While immigration is not a highly polarizing issue, both Sinn Féin, and, to a lesser degree, Fianna Fáil and Independent voters tend to be slightly more antiimmigration than Fine Gael supporters. The Ireland-specific issue—early retirement age—matters for Sinn Féin voters, who tend to support it, and Social Democrats, who surprisingly reject early retirement more than Fine Gael.

Looking at Model 2, negative economic evaluations clearly increase the likelihood of voting for Fianna Fáil over incumbent Fine Gael. Lastly, several parties profit from strong leader effects, especially Fianna Fáil, Sinn Féin and the Green party. Compared to Fine Gael, voters of all opposition parties tend to reject the current PM and Fine Gael candidate Varadkar.

Next, we look at the average marginal effects of ideology and political issues, comparing all parties. We can see that all three political issues (retirement age, immigration, and the EU), most strongly drive the vote for Sinn Féin on the one hand, and for the incumbent Fine Gael on the other hand. Clearly, amongst the opposition parties Sinn Féin managed to politicize all those issues most, and to attract voters in favour of an early retirement age, but critical of immigration and the EU. The two parties are also on the ideological extremes when it comes to left−right—Sinn Féin voters are strongly driven by a left-leaning ideology, and Fine Gael voters by a right-wing ideology (Fig. 10.3).

Average marginal effects (AME) and 95% confidence intervals based on multinomial regression models. For the full table of AMEs see Table 10.1 in the appendix (Model 1). Data source: Maple online survey, wave 4.

**Fig. 10.3** Average marginal effects of selected variables

### Conclusions

This chapter has attempted to understand the degree of politicization of the EU in Irish politics and media, and the implications for the 2020 national elections. Generally, the literature agrees that Ireland is a traditionally pro-European country, in which Euroscepticism has never been a strong driver of political discourses, electoral campaigns or party competition: "What opposition there is to the EU in Ireland is largely confined to periods when the issue is publicly salient, namely during EU treaty referendum campaigns. Outside of these episodes, Euroscepticism is a marginal force in Irish politics, a trend confirmed by the results of the 2020 election" (Murphy, 2021, p. 94). The data we analyzed can, at least partly, confirm this statement: looking at the politicization of the EU in Irish media and parliamentary debates, two trends have become obvious: first, the salience of the EU has very clearly increased across both newspapers and in all parties' speeches over the past 20 years. Given that Ireland has witnessed three referenda on EU issues (2001, 2008, 2012) as well as an EU-led bailout (2010) and a long debate about the Brexit (2016−2019), this is hardly surprising. At the same time, Irish print media has also developed a more critical tone on the EU since the Eurozone crisis, and reinforced this trend during the Brexit debate. Parliamentary debates show a more nuanced picture, with parties somewhat converging in their tone on the EU—while the two centre-right parties have either become more critical in their tone (Fianna Fáil) or remained the same (Fine Gael), the formerly Eurosceptic left-wing Sinn Féin has gradually become more positive in their EU tone. Given that party competition is not taking place over EU issues, it is unsurprising that our analysis does not reveal a very strong impact of EU issue voting. Yet, we can see that Eurosceptic attitudes are still driving the vote for Sinn Féin, despite the softer Eurocriticism the party has recently adopted. Pro-European attitudes, at the same time, significantly affect the vote for the incumbent Fine Gael. Interestingly, in the EU issue as well as other salient issues such as immigration and social welfare, Sinn Féin and Fine Gael are at the two extremes of the spectrum, and seem to attract voters with opposite convictions. They are also the two parties that profit most from left-wing (Sinn Féin) and right-wing (Fine Gael) ideology amongst the voters. The fact that Sinn Féin, of all the opposition parties, managed to position itself most clearly against the incumbent when it comes to the EU as well as other issues is certainly an explanation for their electoral success in 2020.

Lastly, it is important to remember that the COVID-19 crisis which hit just after the Irish elections has clearly shifted the political landscape profoundly, with the potential to strengthen Ireland's focus on economic and social issues over the EU issue. While Sinn Féin is the only party that does, at present, profit from Eurosceptic voting, it seems unlikely that they will use this issue in the near future and return to a more critical agenda. After all, the current political conditions already present an opportunity for Sinn Féin to consolidate its new place as a major party, given that they have a clear electoral profile that distinguishes them from the centre-right parties, but also within their own political bloc. As Little (2021) has noted, the economic and social costs of the Covid pandemic will inevitably be a major political issue, and existing problems of housing and healthcare provision have not gone away either. With the Brexit issue finally off the table, it seems unlikely that Irish electoral competition will focus on European over internal issues any time soon, despite its potential for mobilization due to an increased salience.

#### Appendix

See Table 10.1.





**Table 10.1** (continued)



**Table 10.1**

(continued)




+*p* < 0.10, \**p* < 0.05, \*\**p* < 0.01, \*\*\**p* < 0.001

*Data source* Maple online survey, wave

#### References


**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# Portugal: EU Issue Voting in Mainstream and Challenger Parties

*Marina Costa Lobo* 

# Introduction

In this chapter, we bring together the media, parliamentary and voting data to provide a comprehensive picture of the role of EU politicisation for Portuguese voting. The Portuguese case-study will contribute to the overall goal of the book, namely to understand whether EU issue voting is occurring and national channels of representation are serving as accountability mechanisms for the process of EU integration. The volume begins by setting the stage, namely by providing the trends on EU politicisation in the media and parliamentary debates in Europe, both before and after the Eurozone crisis. Then, Chapters 5 and 6 establish the existence of EU issue voting in all countries, demonstrate that it is magnified by increases in media salience and parliamentary debates' negative tone, and show the prominence of left–right positioning over other issues in the four bailout countries considered, namely Greece, Ireland, Portugal and Spain.

© The Author(s) 2023

275

M. Costa Lobo (B)

Institute of Social Sciences of the University of Lisbon, Lisbon, Portugal e-mail: marina.costalobo@ics.ulisboa.pt

M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_11

Yet, there are still unanswered questions which only an in-depth analysis of the Portuguese case-study can provide, and that constitute the focus of this chapter. First, we present the salience and tone of the EU issue in media and parliamentary debates in Portugal in a longitudinal perspective. This will allow us to understand how present the EU was in the media, compared to past elections, and thus give a proper sense of how important it was in the 2019 election context. Then, comparing the EU issue's relative importance, with other political issues, which is the goal of this chapter, we are in effect benchmarking our findings on EU issue voting.

Portugal is an interesting case, to analyse the phenomenon of EU politicisation and its consequences, for several reasons. First, because it has been seen as a country of euroenthusiasts, both in terms of its political elites, as well as its citizens (Llamazares & Gramacho, 2007; Jerez-Mir et al., 2009; Verney, 2011). While this relative consensus may caution against politicisation, research has demonstrated that, in fact, not only the EU was politicised, especially on the Left of the party spectrum, but it was also an explanatory factor of voting behaviour for extreme-Left positioned citizens (Lobo, 2003, 2021). Second, because Portugal was one of the countries at the epicentre of the Eurozone crisis. Portugal had to ask for a bailout in 2011, which lasted until 2014. The bailout brought with it stringent fiscal policies, with governments agreeing to harsh cuts both in public sector wages, in pensions and other welfare subsidies, as well as tax increases. The austerity which ensued from the bailout raised the profile of the EU in the country, and may have made it a more relevant issue for voting. Third, because the data we collected pertain to the 2019 legislative election, held on the 6th October, which can be considered a post-bailout election, and where the exacerbated role the EU may have played during the bailout may have subsided. Thus, any effects which are found now may be considered more long-lasting, independent of the crisis.

The Chapter is organised in the following way: first we start by presenting a brief overview of the literature on parties, voting and the EU in Portugal. Then, we describe briefly the context of media and parliamentary debates to understand the data we present. Next, we explain the general political context of the 2019 elections and the data collected in relation to Público and Diário de Noticias, two mainstream newspapers, both in terms of salience and tone, from 2002 to 2019 as well as the data on parliamentary parties' salience and tone, during the same time-frame. It is relevant to discuss both these arenas as different forums where information about the EU may have been communicated from parties to citizens. Also, as shown in Chapter 4, these arenas differ in communicating the EU.

Then, we present the analysis of EU issue voting in the 2019 legislative elections. We consider the degree to which the EU issue explains the vote for each major party on the Left and on the Right, using multinominal regression and presenting Average Marginal Effects of EU issue voting compared to other issues, in two different models of voting behaviour. The first includes socio-demographic controls, ideology and different issues, including the EU issue. The second includes all the variables in Model 1, as well as the short-term variables of attitudes towards the leader of the party voted for and economic perceptions. In effect, we will be able to determine the relative importance of the EU in the context of other issues, which have been deemed important for political debate in Portugal.

#### Parties, Voting and the EU in Portugal

The Portuguese party system has been characterised by "limited EU contestation" (Taggart & Szczerbiak, 2008) with Euroscepticism being politicised mainly on the Left (Freire & Teperoglu, 2007; Sanches & Santana-Pereira, 2010; Santana-Pereira & Fernandes, 2014). Indeed, the European cleavage was, since democratisation, contained in a larger "regime cleavage" and served to distinguish the Communist Party from the other parties with parliamentary seats. Namely, the Socialists, PS, the centre-right PSD, as well as the conservative CDS considered, that European integration would be useful for the consolidation of a liberal democracy in Portugal (Pinto, 2011). Contrarily, the Communists considered that EU membership would be a way of ensuring the diffusion of right-wing neoliberal policies in Portugal. Then, in 1999, another left-wing party gained access to Parliament: the *Bloco de Esquerda* (Left Block), which joined the Communists in adopting an Eurosceptic stance (Lobo & Magalhães, 2011). The BE concurred on the negative economic consequences for Portugal of the EU, but had a pro-EU stance in other dimensions of the European integration, namely on the benefits of the EU project itself. These are the only two parties which have been systematically Eurosceptic in the Portuguese party system. On the Right, the CDS-PP flirted with Euroscepticism in the late 1990s as a way to distinguish itself from the centre-right governing party, the PSD. Yet, since the early 2000s, it became a steady coalition partner for that party and abandoned its Eurosceptic agenda (Lobo, 2003; Sanches & Santana-Pereira, 2010).

Since the onset of the Eurozone crisis there have been notable changes in the party system. Firstly, there has been a tendency towards a decline of the strength of the two main parties, the PS and the PSD. While between 1987 and 2005 the sum of their votes was on average 76%, from 2009 to 2022 it has decreased to 68%. This value is slightly inflated considering that in 2015, we also added the votes of the CDS, taking into account that PSD and CDS formed a pre-electoral coalition in that year.

Secondly, there was a change in the dynamics of the party system following the crisis (Lobo, 2021), with greater bipolarisation. In 2015, the Socialist party, which came second in the elections, decided to form a majority alliance with the parties to its left, the Left Block (BE) and the Communists (PCP). This was the first time such an alliance was held in the history of Portuguese democracy, and it took the form of a coalition of parliamentary incidence, with the smaller parties not taking any seats in government. In part, it had not happened until then due to ideological differences between the parties, including their position on Europe, detailed above. The coalition lasted its full mandate, until 2019, but the parties presented themselves independently to the election. This is an important election to study from the perspective of media, parliamentary debates and votes, since the unprecedented alliance of these parties, which lasted the full mandate 2015–2019, may have mitigated the importance of the EU issue across the different forums and even for voting behaviour.

Thirdly, since 2019, there has been fragmentation on the Right. In that election, there were three new entrants into Parliament, two of them on the Right: one MP each from an extreme-right party Chega, and a liberal party, Iniciativa Liberal. The entry of the far-right Chega into Parliament constituted an important watershed, as Portugal was one of the few remaining countries of Europe without an extreme-right populist party (Mendes & Dennison, 2021). The third party to enter Parliament was Livre, a left-libertarian party, which also elected one MP. In the 2022 elections, both Chega and IL increased their vote substantially, with Chega becoming the third most-voted party in Parliament, winning 7% of the votes and 12 MPs. These parties on the Right have different positions regarding the EU. While Chega adopted in 2019 an "Europe of Nations" Eurosceptic position, IL was clearly pro-EU. Taken together, it seems that the Eurozone crisis did have some important consequences for the party system in terms of dynamics, as well as its format.

In contrast with research on party positions, there has been relatively less research on the importance of the EU for individual attitudes and political behaviour, in particular. In terms of attitudes, the Portuguese electorate initially combined a very positive outlook on the EU with a relative lack of knowledge and interest. Moreover, largely positive attitudes did not translate into electoral participation for EP elections (Lobo, 2011). Indeed, Portugal has one of the lowest levels of participation in EP elections, even when we consider EP elections from 2004 onwards, which already include the Central European countries. Moreover, in 2019, the first time the EP elections reached an overall 50% turnout, in Portugal only 37% of voters participated in the elections. When it comes to European attitudes, it was systematically found that support for the EU in Portugal was rather instrumental, dependent more on economic benefits than on political values of membership (Lobo, 2011).

Concerning what explains support for the EU in Portugal, satisfaction with democracy was the most important variable in explaining support for the EU, followed by voting for the Communist Party (Lobo, 2003). More recently, Freire et al. (2014) show that the onset of the Great Recession led to a strong growth in Euroscepticism at the voter level in Portugal, as occurred in other bailout countries. The authors found that having an extreme-left or an extreme-right positioning was predictive of Eurosceptic attitudes, in line with the findings by Santana and Rama (2018). In addition, even after controlling for all the major factors of Euroscepticism, attitudes towards the Troika agreement and debt renegotiation had a significant impact on voter's support for the EU.

Teperoglu and Belchior (2020) find that in Portugal, at the peak of the crisis, self-placement on the centre-left and, to a lesser extent, on the extreme left was a significant determinant of Eurosceptic stances, but this effect had lost significance by 2018. On his part, Lisi (2020) shows that extreme-Left voters are the most Eurosceptic; negative economic perceptions fuel Euroscepticism and those who tend to trust national institutions also trust the EU to a greater extent. In what concerns voting, there is a difference between the two main parties (PS and PSD), whose sympathisers are clearly pro-European, and challenger parties. As expected, both BE and PCP sympathisers are more Eurosceptic, in a significant way. Moreover, for explaining the vote, the EU issue was significant in voting in the 2014 EP elections (Freire & Santana-Pereira, 2015). Overall, the studies have shown that Euroscepticism is located on the left of the electorate, and that the crisis sharpened the existing differences. We turn now to the analysis of the EU across the two different contexts: media and parliamentary debates.

#### Studying the EU Across Different Forums: The Case of Portugal

As has been explained above, in this Chapter, we examine different kinds of data, namely media articles and parliamentary debates to contextualise voting. In this section we seek to present a brief literature review on each of these different political arenas research in Portugal to better understand the data included in the chapter. Concerning the media landscape, Hallin and Mancini categorised Europe's media systems (2004, 2012), following a number of structural criteria. According to them, Portugal belongs to the polarised-pluralist type of media system, alongside Spain, Greece and Italy. This model is defined by a weak, underfinanced media market, political control of the media, and state intervention in the media as owner, regulator and financial backer (Santana-Pereira, 2016).

Indeed, Portugal may be considered a case where campaigning is *permanent*, and it occurs through the various mass media, especially television and newspapers (Santana-Pereira, 2016). The way that "permanent" campaigning happens in Portugal is through the role that party politicians taken on as pundits both in television and the main newspapers (Figueiras, 2011, 2019). Figueiras has shown that the time/space dedicated to punditry has increased since the 1980s in Portuguese mass media (2011). Yet, according to Silva et al. (2017), the instrumentalization of the media has been mitigated in Portugal, relative to other Southern European countries. This is due to the fact that there is less differentiation between mainstream parties, as well as the political professionalism of journalists in general. Indeed, repeated surveys show that the media, both television and newspapers are highly trusted in Portugal (Newman et al., 2019).

Within this context of high interpenetration between politicians and mass media, there is little information about the kinds of topics which are discussed during campaigns in a systematic fashion. Concerning the EU topic, it was found, considering the 2009 EP elections, that the salience of the EU during EP elections was the highest in Portugal, of the 13 countries considered. In terms of tone, whereas the majority of countries' media had a positive tone towards the EU, in Portugal it was slightly negative (Stromback et al., 2011). In the same volume, Jalali and Silva (2011), find that there are differences between the way in which parties and media politicise the EU, as well as differences between parties. They find that the media is significantly less preoccupied with Europe—in terms of issues and themes, if not actors—than the parties. Further, they find that it is government parties that focus on the EU issue, whereas opposition parties, of the left and the right, focus mainly on national issues. We now turn to the analysis of the 2019 election, and analyse the media and parliamentary data in a longitudinal fashion.

### The 2019 Election Campaign: Media and Parliamentary Data

The 2019 election followed the end of the full mandate of a minority Socialist government, which had the support from the Left Block and the Communist Party. Following the 2015 elections, which were the first post-bailout elections, the right-wing party, PSD, won the election but without a majority. However, rather than supporting the PSD's executive, the Socialists decided to form a coalition of parliamentary incidence with the two small parties on its left. This was the first time that these parties were able to form a coalition, which was labelled "*geringonça*". The mandate between 2015 and 2019 proved to be politically stable. During that period, the indicators of support for democracy as well as trust in government improved quite substantially. With improving economic indicators, the major beneficiary of this government mandate was the Socialist party, which saw its vote increase from 32.31% to 36.34%, while both the PCP and the BE saw their vote decrease slightly (Fernandes & Magalhães, 2019). Yet, despite this result, following the 2019 elections, the Socialists decided not to re-enact their coalition of parliamentary incidence. Instead, they formed a minority government without any parliamentary agreements (Jalali et al., 2020). This election also saw the entry into Parliament of three new parties: on the far-right, Chega, on the liberal right, Iniciativa Liberal, and, on the Left, Livre, each elected one MP to the Assembleia da República, winning seats in the Lisbon electoral district, the largest in the country. Thus, while confirming the resilience of votes for the centrist parties, the PS and the PSD, 2019 also saw the entry of three new parties to Parliament, which signalled future party system fragmentation.

Next, we present data on the politicisation of the EU in media and parliamentary debates for the period 2002–2019. It is relevant to discuss both these arenas as different forums where information about the EU may have been communicated from parties to citizens. "Politicisation" has been defined as a process whereby a collective decision generates disputes, and wherein the audiences of those disputes gradually expand (Schmitter, 1969). It refers to "an increase in polarization of opinions, interests or values and the extent to which they are publicly advanced towards the process of policy formulation" (De Wilde, 2011, 559). In our research, politicisation has been operationalised in two dimensions: salience and polarisation (Silva et al., 2022). Salience is measured through the number of articles/speeches which mention the EU in a significant way (in the title or in the article body for media articles, and in speeches for parliamentary debates) as a proportion of the total number of articles/speeches. Polarisation is harder to measure. For our purposes, we opted for presenting the tone of the articles/speeches which mention the EU. In the case of media, this measure refers to an average of two measures obtained in an automated fashion. Namely, it combines a measure of the sentiment of the title of the article, and a measure of the average sentiments of the EU sentences in the article itself. In the case of parliamentary debates, this measure refers to the average sentiment in all the EU sentences uttered in a given speech. Thus, once the EU sentences have been identified, a sentiment score is calculated for each of them, after having been translated into English, and the tone measure is the average sentiment score within those speeches (see also Chapter 3).

In Fig. 11.1, we present data from the two main daily newspapers, Público and Diário de Notícias, which are traditionally associated respectively with the centre-left and the centre-right of the Portuguese party spectrum. For each time point, the average of salience and tone is presented for the relevant newspaper articles in the thirty days before the election. Rather than present only the 2019 data, we opted to present data longitudinally from 2002 to 2019, which helps us to understand the specificities of our election of interest. Thus, the data below present EU salience and tone in Portuguese mainstream media from 2002 to 2019.

Firstly, as has been noted elsewhere (Silva et al., 2022), there was a sharp increase in media salience to the EU following the onset of the

**Fig. 11.1** EU politicisation in the Portuguese media, 2002–2019 (*Source*  MAPLE data)

Eurozone crisis, which has not abated. Thus, for both newspapers, the period post-2011 shows greater EU salience than pre-2011. Despite this trend, there are some differences between newspapers. Namely, Público consistently mentioned the EU to a greater extent than Diário de Noticias, and this was especially the case in 2011, the election which followed the bailout agreement in Portugal.

When we consider tone, the 2002–2015 trend is the following: there has been a decline in the positive tone of articles, signalling an increase in negativity of articles, from 2005 onwards in both newspapers. When we consider tone, we also detect differences between the newspapers. Whereas centre-left Público has a steady decline in tone from 2002 onwards, in centre-right Diário de Noticias, the tone becomes more positive from 2002 to 2009, and then drops precipitously until 2015.

When we focus on tone in 2019, we see that contrary to past trends in both newspapers, tone improves in 2019, to pre-crisis levels (2009) both in Público and Diário de Notícias. Thus, it seems that although attention to the EU suffered a dramatic shift post-2009 for both newspapers, which has not been undone in 2019, in terms of tone, there was an improvement in the latest election. There could be several reasons for this improvement in tone. First, the simple fact that the bailout has ended, and that Portugal was able to meet its public finances commitments in the EU after 2014, eased relations with the EU. Second, the government's Finance Minister, Mário Centeno, became President of Ecofin, which may have contributed to more positive-toned articles mentioning the EU. Thirdly, as explained above, the main Eurosceptic parties in Portugal, PCP and BE were supporting the government, and thus less likely to effectively politicise the EU. The data on parliamentary debates, that we discuss next, indeed confirm this.

Unlike the media data, where each time point represents data collected one month before the election, we were able to collect all speeches for the entire year for parliamentary debates. Thus, the parliamentary debates dataset includes all plenary speeches which were uttered from 2002 to 2019.

Considering salience, we note that EU salience in plenary speeches is low, and does not increase dramatically since 2009 (Fig. 11.2). Yet, the differences between parties increase following the crisis. Moreover, from 2011 onwards, the mainstream parties tend to distinguish themselves. The salience given by the Socialist party has a higher increase, vis-à-vis the salience attributed by the PSD, independent of the former being in opposition (2011–2015) or in government (2015–2019). The Communist party exhibits a stable pattern, with higher salience attributed at election times, while the Left Block gave EU issues more salience during the bailout period (2011–2014), than since they supported the PS minority government (2015–2019). On the Right, the CDS-PP seems to follow closely the PSD in the salience attributed to the EU. They were coalition partners from 2011–2015, and incumbency increased the salience attributed to the EU.

**Fig. 11.2** EU salience in parliamentary debates per Portuguese Party (*Source*  MAPLE data)

Figure 11.3 presents the tone employed by parliamentary parties from 2002 to 2019. Contrary to salience, where parties hardly distinguished themselves from each other until the onset of the crisis, in the case of tone we see differences for the whole period. Concerning the mainstream incumbent parties, the Socialist party tends to be less positive when in opposition (2002–2004; 2011–2015) than when in government (1999– 2001; 2005–2009; 2015–2019). In particular, the most positive tone coincides with the year in which the PS won its first absolute majority (2005), and on average remaining until 2011 with a more positive tone than the mainstream opposition party, the PSD. When the latter party formed government in 2011 to administer the bailout, and governed until 2015, its tone was more positive than the Socialists, which were at its lowest for the period analysed. Once the PS returned to power in 2015, its tone towards the EU becomes more positive. On the right, the PSD tone was most positive when the party was in government (2002– 2004; 2011–2014), and less so when in opposition, in 2014–15, only to recover during the more recent period. Considering the smaller parties, the CDS-PP seems to emulate the PSD trends. On the Left, as we would expect, there are more differences. Since joining the euro, in 2001, the Communist party as well as the Left Block have had a rather more negative tone regarding the EU than the Socialists. For the Communists, the tone decreases from 2006 to 2015, where it reaches its lowest point. The Left Block follows this trend too, with low and declining levels of tone from 2006 to 2012. Forming the left alliance with the Socialists seems to have had some impact for these two parties. Both parties, especially the Communists, saw increases in tone, from 2015 onwards. In that period, the Communists' tone was almost identical to the PS, something that had not happened since 2001.

Therefore, taken together, we observe that, in terms of EU politicisation, 2019 saw the relative salience in the media, but slightly less polarisation in the newspapers during the campaign. A similar trend was also observed in the legislative term in Parliament (2015–2019), where the tone of opposition parties was on average much more positive than during the previous years. All in all, we can consider that there was a depoliticisation of the EU in these two political forums in 2019, compared to the previous period in the *Assembleia da República*. Thus, this context is one where, despite more awareness of the EU, due to its salience, it is a time of less contestation, especially due to the fact that the

**Fig. 11.3** EU tone in parliamentary debates per Portuguese Party (*Source*  MAPLE data)

Communists and the Left Block were supporting the government. This leads us to expect that EU issue voting may not be very significant in these elections. Interestingly, this, however, doesn't seem to be confirmed in our analysis as we will see in the next section.

### Benchmarking EU Issue Voting in Portugal

The survey employed is a representative two-wave panel online survey with a sample of 1540 in the first wave and 1608 respondents in the second one. We are using the second wave, post-election data, which was collected between 7th October and 30th November 2019. The panel provider was able to fulfil a crossed quota of gender (2 categories), age (3 categories) and education (3 categories), using the 2011 census as the matrix to build the sample. A model of voting behaviour was built which includes socio-demographic controls, ideological self-placement, political issues, including the EU, leader barometers and perceptions of the economy to explain the vote for each main party that won seats in 2019. We only included in the analysis parties where at least 40 respondents stated they had voted for it. Thus, our analysis includes the Left Block (BE), Communists (PCP), Animal Party (PAN) and Socialists on the Left, and PSD, CDS-PP, and Chega on the Right.

Our dependent variable is vote recall, which is a categorical variable. We employ multinomial regression, with the Socialists as the baseline reference. As the coefficients of multinomial models are complex to interpret and depend on the chosen baseline outcome, the results are here presented, in Fig. 11.4, by plotting average marginal effects of the main variables of interest. The regression results are presented in the Appendix: Table 11.1. Average Marginal Effects of the key issue variables are presented, to contrast the importance of EU issue voting among all parties, and also to contrast the relative importance of EU and other political issues.

Our main independent variable to measure position on the EU issue is the following variable: "Some people believe that the process of European integration should move forward to the creation of the United States of Europe. Others believe that the European Union should be dissolved in order to return to a situation in which states are fully sovereign. In which point of the following scale would you place yourself?" (10-point scale from completely agree with dissolution to completely agree with a United States of Europe).

**Fig. 11.4** AMEs of voting for each of the main Portuguese parties (*Note*  Average marginal effects and 95% confidence intervals, based on the results reported in Model 1 in Appendix: Table 11.1)

Other issue variables, salient due to their relevance for the 2019 campaign, were included to benchmark the importance of the EU issue. Firstly, attitudes towards public services and social protection, measured with the following question: "Using a scale from 0 to 10, where 0 means "We should improve public service and social protection even if it means a tax increase" and 10 means "we should reduce taxes even if it means reducing public service and benefits, where would you stand?" Secondly, attitudes towards equality in redistribution, using the following item: "And now using the same scale from 0 to 10, where 0 means "there should be more equality in income distribution" and 10 means "there should be more incentives for individual initiative" where would you stand? Thirdly, we measure attitudes towards the national health system in the following way: "Finally, using the same scale from 0 to 10, where 0 means "the national health system should be privately controlled" and 10 means "the national health system should be state controlled", where would you stand?

Simple leader barometers, which we have included in the survey, and have been shown to be highly correlated with party choice, were also included in the regression analysis. Also, sociotropic retrospective economic perceptions were included in a different model. Therefore, we will show the results of two models, first without leaders and economy, and then including these two variables, to understand if the significance of EU issue voting resists the inclusion of these important short-term variables. Another reason to have two models is due to the fact that we did not include a question about attitudes towards André Ventura, the Chega leader. For this party, its results are only presented in the first model.

Socio-political control variables were also included to understand the relative importance of each in explaining the vote for each party. Namely, age (3 categories), gender (2 categories), education (3 categories), religiosity (4-point scale) and ideology (11-point scale). All independent variables were standardised to vary between 0 and 1, for the sake of results interpretation.

The regression results presented in the Appendix: Table 11.1 show that concerning socio-political controls, on the left, BE voters tend to be younger than the Socialists, whereas the Communists are more unionised. Both BE and Communists are significantly less religious than the Socialists. On the Right, men and individuals with higher education are likelier to vote for CDS and PSD, compared to PS.

The general left–right continuum performs well in the model, with those who voted BE and Communists being significantly more to the left of the Socialists, and PSD and CDS voters to the Right. PAN voters do not distinguish themselves ideologically from PS. *Chega* voters in this sample are not significantly different from the socialists in terms of ideological self-placement, but caution should be had relative to this party as the number of respondents in the sample which stated they voted for that party is very low.

The results in Fig. 11.4 present the Average Marginal Effects for the EU issue as well as the other socio-economic issues and left–right. When examining the recalled vote after the election, there is a significant negative effect for the Communist party with voters 10 percentage points more likely to vote for this party when they oppose the EU than when they believe the EU should move forward to the United States of Europe. This effect is significant, as can be seen in Appendix: Table 11.1 in both models, with and without short-term variables. BE voters are also more likely to choose this party if they oppose the EU, while PS and PSD voters are, on the contrary, more likely to choose those parties if they support the furthering of EU integration. Yet, the relationships between the latter three parties are not significant, with the exception of the PSD.

Concerning the other political issues, being in favour of state control of the national health system has a significant positive effect in voting for Communists, while the opposite occurs for the CDS-PP and the PSD. The impact of the other two socio-political issues, namely improving public services and defending income redistribution is not significant. The political issue effects on voting for the Communists, the PSD and BE, with PS as the baseline, remain significant even after the inclusion of short-term variables, leader barometers and economic perceptions.

# Conclusions

In this Chapter, we analysed trends in media and parliamentary politicisation of the EU in Portugal, as well as its consequences for voting behaviour. In order to do so, we employed unique data collected longitudinally, from 2002 to 2019, as well as panel survey data collected after the October 2019 legislative elections.

We showed that, since the onset of the crisis, two periods can be distinguished in terms of EU issue salience and tone both in the media, and in parliamentary debates. From 2009 to 2015, the whole crisis/bailout period, salience has increased and its tone has tended to become more negative in the mainstream news media. Then, from 2015, salience does not decline to pre-2009 levels, both in the media and parliamentary debates. Yet, the tone in the mainstream media, as well as in parliamentary debates improved in that period.

Thus, the parliamentary parties, in the mandate 2015–2019, did not really differentiate themselves in Europe. This may be a reflection of the fact that the main Eurosceptic parties, the Left Block and the Communists, had formally agreed to support the minority PS government, for the first time in 40 years, from 2015 to 2019.

We then turned to the analysis of the survey data. We saw that the EU issue explains the vote for the Communists and the PSD in 2019, relative to other factors. With the exception of Health, we also saw that most of the other political issue variables included, namely those relating to the welfare state, do not increase the likelihood to vote either for left or right parties. We also ran the models that included leader barometers and the economy and the results remained the same.

Overall, there is a continuity of the eurosceptic nature of the PCP electorate despite the fact that during the 2015–2019 left-wing alliance government, the party distinguished itself less from the Socialists in parliamentary tone regarding the EU. The results also show that the PSD voters are more euroenthusiastic about advancing EU integration than the PS. This is important as the PSD is the main opposition party, and it shows that the EU issue is not only important for voters on the extremes of the party system (Communists), but also for mainstream voters, such as those who vote PSD. Overall, the chapter shows that EU issue voting matters and its impact is larger than that of other comparably salient issues, which signals the importance that national institutions can have for the legitimization of the European Union.

# Appendix

See Table 11.1




**Table 11.1**

(continued)



Standard errors in parentheses; AMEs in [] +*p* < 0.10, \**p* < 0.05, \*\**p* < 0.01, \*\*\* *p* < 0.001

#### References


**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# A Broken National Consensus? EU Issue Voting and the Radical Right in Spain

#### *Hugo MarcosMarne*

# Introduction

Previous chapters in this book have considered from a comparative perspective the extent to which EU politicization is taking place in media and parliaments and, adopting an experimental standpoint, whether EU issue voting exists. Evidence contained in these chapters indicates that Spain has witnessed an increase in the salience of EU issues over the past years that is not necessarily accompanied by more contestation in parliaments and media. It also suggests that we must be careful when considering the existence of EU issue voting, as heuristics may be operating. This chapter delves into the political dynamics of Spain to shed light on the effects of EU issues for voting in the November 2019 elections. For that, it presents an overview of the perceptions and positions towards the EU among Spanish citizens and parties, disaggregates indicators of politicization in both media and parliamentary debates, and

© The Author(s) 2023

299

H. MarcosMarne (B)

University of Salamanca, Salamanca, Spain e-mail: marcosmarne@usal.es

M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_12

conducts an innovative electoral analysis using a complete set of variables that includes key issues for voting in 2019, such as general preferences towards the territorial organization of the State, perceptions of the degree of autonomy and the state of civil and political rights in Catalonia, immigration, and feminism (the last one with a focus on the perception of violence against women). The rest of this section revolves around the evolution of EU perceptions among parties and citizens in Spain.

Spain was for a long time an exceptional case due to the positive consensus existing around the European Union (EU), at least among mainstream state-wide political parties (Benedetto & Quaglia, 2007; Elias, 2008; Gomez-Reino et al., 2008; Jerez Mir et al., 2008). EU membership appeared linked to ideas of democratization and modernization (Farrell, 2004; Powell, 1986; Royo & Manuel, 2007) and putting an end to the international isolation experienced under the dictator Franco (Powell, 2003), which made contestation around the issue almost nonexistent in the years immediately before and after the accession (Vázquez García et al., 2010; Verney, 2011). Although Spain joined the EU amid negotiations conducive to the adoption of the Single European Act (SEA), concerns about its economic impact were balanced out by the existence of European funds that would be particularly beneficial for less developed countries to adapt to the single market. Therefore, under the presidencies of Felipe González (1982–1996) and José María Aznar (1996–2004) that happened before and after the effective accession, both public opinion and a large majority of political actors in Spain maintained a markedly Euro-enthusiast position (Powell, 2003). In fact, less than 10% of the Spanish population declared to oppose EU membership until 2008 (Real-Dato & Sojka, 2020). However, the economic and political earthquakes that started with the economic and financial crisis of 2008 caused some tremor in the Euro-enthusiasm of Spain.

Spain had the dubious honor of being among the European countries most affected by the crisis, which hit hard the labor market. Unemployment rose by 37% in just one year, and the situation only got worse afterwards (Royo, 2009). By the beginning of 2013, the unemployment rate had more than tripled 2007 figures. Besides rising poverty and economic inequality (Zamora-Kapoor & Coller, 2014), the negative effects of the crisis were (and still are) visible in fundamental aspects such as public health, or gender inequality (Gili et al., 2013; Karamessini & Rubery, 2013). It thus stands to reason that the political sphere would not be immune to the situation.

While mainstream political parties of the left (Partido Socialista Obrero Español, PSOE) and the right (Partido Popular, PP) changed little their discourses towards the EU (Bakker et al., 2012; Volkens et al., 2020), public opinion became much more critical of the EU after 2008. As a textbook example of this trend, the percentage of people who declared to trust the European Parliament in Spain fell from 69 to 44% between 2007 and 2011, the second biggest decline of the EU after Greece (Armingeon & Ceka, 2013), and trust towards the EU shrunk from 66 to 16% between 2008 and 2014 (Cordero & Montero, 2015). In fact, the 15 M-indignados mobilization that spread after 2011 was a paradigmatic example of a social movement challenging EU austerity while asking for pro-democratic reforms (Flesher Fominaya, 2017).

Connected with this public mobilization, the ongoing reconfiguration of the Spanish party system that started with the 2014 European elections affected the general consensus around the EU. Podemos, a new party characterized by its radical left-wing ideology and populist discourse was born with a soft Eurosceptic discourse directed towards neoliberal and adjustment policies (Plaza-Colodro et al., 2018). The radical-right VOX also displayed a soft Eurosceptic discourse that, unlike that of Podemos, focused much more on sovereignty and national identity issues (MarcosMarne et al., 2021; Real-Dato & Sojka, 2020). On the contrary, the third new party that gained importance in this period, the center-right Ciudadanos (Cs), never questioned the mainstream consensus around the EU (Polk et al., 2017). Overall, by the turn of the decade a new scenario emerged in Spain that, far from representing a Eurosceptic turn, looked more prone to EU politicization along the lines of increased issue salience, actor expansion, and actor polarization (Hutter & Kriesi, 2019). On the one hand, public opinion towards the EU slightly changed from its monolithic positive view. On the other hand, new political actors emerged that were willing to compete on EU issues, a strategy situated in sharp contrast with the no-contestation approach largely followed util then (Vázquez García et al., 2010).

# Politicization of the EU in Media and Parliamentary Debates

To further delve into the indicators of EU politicization in media and parliamentary debates, this chapter presents two types of evidence. First, a graphical representation of the salience and tone of the coverage of the EU in the two most important newspapers in the country, El Mundo, and El País, in the period between 2004 and 2016 (Fig. 12.1). Second, a party-disaggregated representation of EU salience and tone for the five most important state-wide parties (PSOE, PP, Cs, Podemos, VOX), in the period between 2002 and 2019 (Figs. 12.2 and 12.3). The salience variable represents the percentage of news/interventions that refer to the EU. Meanwhile, the tone variable represents the overall perception of the EU in each unit of analysis (be it parliamentary intervention or newspaper news). Values over zero represent positive perceptions, and values below zero are negative ones. More detailed information about the coding procedure can be found in Chapter 3.

Regarding salience, data from Fig. 12.1 reflects the increasing coverage of EU issues in media between 2004 and 2016, with a moderation in 2015, and an ending point in 2019 somewhat higher than 2004. This trend can be seen in both El Mundo and El País. Data from parliamentary debates gathered in Spain evidence that political parties speak about the EU, although the percentage of EU interventions was rarely above 10% (Fig. 12.2). For the two most important political parties, PSOE and PP, mentions about the EU increased after 2008 (and especially after 2011, coinciding with the most evident effects of the Eurocrisis) (Braun & Tausendpfund, 2014). Mentions of the EU reached its maximum in 2012 both for the PP (11%), and the PSOE (9.3%), coinciding with the official request of help from the Spanish government for the financial system.

As for the tone variable, the indicator for media reached the lowest point in 2011 both in El Mundo and El País, thus coinciding again with the Eurocrisis. EU tone in political parties is positive for both the PP and the PSOE except in 2014 (PSOE), with no evident evolution trend (Fig. 12.3). Cs and Podemos' interventions, conversely, show a negative trend (but there are only four points in time available for these parties). Data on VOX is lacking since they only accessed the national Parliament after the April 2019 elections (Turnbull-Dugarte et al., 2020).

Overall, data from both newspapers and parliamentary discourses evidence an increase in the salience of EU issues after 2008 that declines again after 2012, and a not-so-clear pattern regarding negative mentions of the EU. However, it is evident that Euroscepticism never dominated public debates in Spain. Considering this analysis of media and parliamentary data, this chapter now asks, is there evidence of EU issue voting in the general elections that took place in November 2019?

**Fig. 12.1** EU salience and tone in Spanish media (*Note* self-elaboration with data from MAPLE)

**Fig. 12.2** EU salience in parliamentary debates by party in Spain (*Note* selfelaboration with data from MAPLE)

**Fig. 12.3** EU tone in parliamentary debates by party in Spain (*Note* selfelaboration with data from MAPLE)

#### Did EU Preferences Matter for Voting Decisions?

An increasing bulk of comparative empirical research suggests that EU issues matter for voting decisions, both at the European and national level (Beach et al., 2018; De Vries & Tillman, 2011; Hobolt & Rodon, 2020). In a nutshell, these studies find that whenever information about the EU is available, individuals may ponder EU issues when deciding whom to vote. Overall, the framing and intensity of the information about the EU are expected to influence both voting decisions and the comparative strength of EU issues vis-à-vis other variables (van Elsas et al., 2019). Basically, theory predicts that EU politicization will correlate with an increased importance of EU issues in voting, which explains for example the relatively low salience of EU issues in times of the "general consensus" before the 90s (Hooghe & Marks, 2009). However, it is also suggested that not all parties will benefit the same from this situation. The issue salience and polarization trends that define politicization (Hutter & Grande, 2014) imply that parties adopting extreme positions and emphasizing EU issues can benefit the most from it. In this sense, the general picture of parties' position towards the EU in Spain in 2019 largely corresponded to the inverted U-shaped proposed by Hooghe et al. (2002), with Izquierda Unida (IU), Podemos, and VOX displaying comparatively more critical ideas about the EU (Bakker et al., 2020). However, Euroscepticism was not particularly strong in either of the extremes (Carrasco et al., 2021; Fitzgibbon, 2013). In fact, even in 2018 an overwhelming majority of Members of Parliament in Spain (91% of the interviewed) expressed a positive view of the EU (Carrasco et al., 2021).

To unravel if EU issues influenced voting decisions in the November 2019 national elections, I used data from an online survey conducted in Spain in May 2019. Data was gathered by the company Netquest using a quota sampling strategy that took into account gender, age, education, and NUTS1 regions to replicate as much as possible the composition of the census of the country (*N* = 3,006). Respondents were selected from a large sample of individuals available to the panel provider (opt-in panel). While opt-in panel data has proven to be potentially problematic to calculate population estimates, due to sampling issues, these are far less troublesome when studies focus on relationships between variables, as it is the case in this chapter (Baker et al., 2010).

Multinomial logistic regression models were conducted using vote recall in the past elections as a dependent variable (*Which party or coalition did you vote for in the last legislative elections?*), and binomial regression ones were also run to see the relevance of EU issues in voting for new parties within ideological blocks (Cs vs PP; VOX vs PP; Unidas Podemos vs PSOE). The key independent variable in all models is attitudes towards the EU, for which I used a question that asks respondents about the future of the European integration process (*Some people believe that the process of European integration should move forward to the creation of the United States of Europe. Others believe that the European Union should be dissolved in order to return to a situation in which states are fully sovereign. In which point of the following scale would you place yourself?*). This question taps into the strengthening dimension identified by Lubbers and Scheepers (2010).

The models include a rigorous battery of controls that allow considering socio-demographic features and important predictors of voting in Spain. I controlled for age (continuous), gender (female as reference), education (up to secondary, secondary, more than secondary), religiosity (from 1, I have no religious beliefs to 4, I am very religious), trade union membership (dichotomous), self-location in the left–right ideological scale (from 0, extreme left, to 10, extreme right), evaluation of the leader of the party voted (from 0, I strongly dislike, to 10, I strongly like), assessment of the economic situation (when compared to the last 12 months, the economy is now… from 1, much better, to 5, much worse), preferences about the territorial organization of the state (from 1, a state without autonomies, to 5, a state in which autonomies can become independent), evaluations of the political situation in Catalonia (as far as relations between Catalonia and Spain are concerned, do you think Catalonia has reached… from 1, too much autonomy, to 3, an insufficient level of autonomy), assessment of rights in Catalonia using a dichotomous question (do you think that during the last year the citizens of Catalonia have seen a reduction of their individual and collective freedoms as well as their fundamental rights?), and perceptions of violence against women (many women exaggerate the problem of male violence, responses ranging from 1, strongly disagree, to 5, strongly agree). Listwise deletion was applied to deal with missing cases, which results in a final sample of 1512 cases. The reduction of the cases from the general sample is mainly caused by the number of people who said they did not vote or did not declare the party they voted for.

Descriptive data is shown in Table 12.1, and an analysis of preferences on EU strengthening considering vote recall is further included in Table 12.2. Data in Table 12.2 indicate that voters of VOX rank the lowest in the scale of EU strengthening, even if they also show the highest Standard Deviation (SD). Voters of Unidas Podemos (UP, a coalition lead by Podemos and IU) come in second place, followed by voters of PSOE. Voters of Cs and PP rank the highest in the EU strengthening scale. This lends additional support for the expectation that EU issues may influence voting for new radical parties that question, at least to a certain extent, the general positive consensus around the EU.

To see if EU issue voting was operating in the 2019 elections, I first conducted two multinomial regression models using vote choice as dependent variable (using voters of the incumbent PSOE as reference). The first model is the most stringent one, as it included all control variables referred above (Appendix: Table 12.3). The second model maintains the same structure, but two powerful predictors of vote (evaluation of the leader of the party voted in the past elections, and assessment of the economic situation) were removed to see if the electoral effect of EU issues increased (only minor differences were found) (Appendix: Table 12.4).1 A graphical representation of the Average Marginal Effects (AMEs)2 of main covariates were included in Fig. 12.4. Results show that preferring a more integrated EU is negatively related with having voted for one of the two new radical parties, VOX, and positively related with having voted for PP and Cs. This is partially connected to the

<sup>1</sup> In all models run Variance Inflation Factors (VIFs) were consistently below 2. Therefore, no collinearity issues are expected.

<sup>2</sup> The average effect of a variable (*X* 1) on the outcome being 1 (*P*(*y* = 1)) (Mood, 2010, p. 75). Predicted AMEs are based on 'actual values' contained in the dataset (Mize, 2019), allow for a more accurate group comparison (Long & Mustillo, 2018), and across models too (Mood, 2010).


*Source* MAPLE panel data


*Source* MAPLE panel data

U-shape relationship existing between EU positions and ideological radicalism among voters (Hooghe et al., 2002; Rooduijn et al., 2017), clearer in Spain for voters of the radical right (the AME of EU integration on having voted for UP is negative but it does not reach statistical significance at 0.05). It also indicates that left-wing voters, even if they are not radical, may have become more critical of the EU in Southern Europe because of the orthodox economic policies adopted after 2008 (Hutter & Kriesi, 2019).

To further delve into these results, three binomial logistic models were run that separated voters of new parties from voters of mainstream ones within the same ideological space. Accordingly, voters for VOX and Cs were compared to voters of PP, and voters of UP with voters of PSOE (see Fig. 12.5). While choosing between UP/PSOE and Cs/PP seems not influenced by preferences on European integration, results confirm that voting for VOX, and not for PP, is more likely the less individuals favor European federalism (*B* = −0.14, *p* < 0.01). Furthermore, the within-blocks comparison is helpful to shed additional light on the reasons that trigger voting for new parties in Spain. Using PSOE voters as reference, individuals are more likely to vote for UP if they are younger, less religious, more critical of the economic situation, located more to the left, more in favor of more competences for Catalonia, more likely to understand there was a reduction of rights in Catalonia, and more favorable towards immigration (see Appendix: Table 12.5, model 1). Using

**Fig. 12.4** AMEs visual representation. Vote choice in the past national elections in Spain (*Note* Average marginal effects with confidence intervals (95%). The visual representation is based on results included in Appendix: Table 12.3)

**Fig. 12.5** AMEs visual representation. Voting for new/mainstream parties within ideological blocks in Spain (*Note* Average marginal effects with confidence intervals (95%). The visual representation is based on results included in Table A.3)

voters of PP as reference, individuals are more likely to vote for VOX if they are younger, less educated, value more their leader, perceive that violence against women is an overstated issue, favor a more centralized state and crucial for this chapter, oppose more the EU integration process (see Appendix: Table 12.5, model 2). Meanwhile, individuals are more likely to choose Cs over PP if they are younger, less religious, assess the economic situation more negatively, value more their leader, are located more to the left, and understand that violence against women is not an overstated issue (see Appendix: Table 12.5, model 3).

# Conclusion: European Integration Discourses and EU Issue Voting in Spain

The results shown in the previous section support that EU issue voting was not absent from the November 2019 general elections in Spain. While not the most powerful explanatory variable, which makes total sense considering the stringency of the models proposed, preferences about the supranational character of the EU influenced voting for parties in a way that is coherent with their electoral platforms. Preferring less integration increases the likelihood of voting for the radical-right VOX. On the contrary, favoring a federal EU is positively related with having voted for PP and Cs, and has no effect on having voted for PSOE and UP. Although more studies are needed to confirm it, I suggest these results can be linked to left-wing voters being comparatively more critical of the EU after the Eurocrisis and the measures implemented to tackle it, particularly visible in Southern Europe.

Considering electoral competition within ideological blocks, preferences about the EU integration process play a role only when choosing between right-wing parties (VOX over PP), which may have important consequences for electoral competition. Crucially, all these electoral effects can be observed in 2019, years after debates around the EU peaked in both newspapers and parliamentary debates. This seems to suggest that preferences about the EU can be a (modest) contributor to voting decisions in Spain, especially among voters of right-wing parties, even when politicization remains at low levels.

A fundamental implication of these results is the apparent persistence of a nuanced general consensus around the EU in Spain. The two most voted parties in the past national elections, likely to repeat according to the polls, still hold positive views of the EU, and nothing in current trends make us expect a radical change in that regard (even less with Covid funds being mobilized under the program NextGeneration EU and defense challenges evidenced after the invasion of Ukraine). Combined with the no-contestation strategy followed by mainstream parties, it is likely that EU issue voting maintains a modest role vis-à-vis other variables such as self-positioning in the left–right scale (Lancaster & Lewis-Beck, 1986). This is even more the case in light of recent studies that emphasize the importance of heuristics for the formation of attitudes towards the EU (Armingeon & Ceka, 2013; Pannico, 2017; Torcal & Christmann, 2018). Evidence presented in Chapter 5 points in this direction. Once party cues are removed from the calculus preferences about the EU did not explain voting. Accordingly, it cannot be ruled out that part of the strength of the EU coefficients found in voting models is a consequence of which is the preferred party, rather than a prior attitude that explains voting for it.

Importantly, and in line with the overall general consensus mentioned above (in fact, it should be explicitly mentioned that no strong Eurosceptic party emerged in Spain even amid the worst of the crisis), discourses of UP and VOX do not question the EU in its entirety. However, they are clearly more critical towards the EU, and such discourses are likely to maintain at least some relevance in the future. For example, UP may still criticize neoliberal policies of the EU in the implementation of neoliberal policies to tackle the effects of the COVID-19 crisis (e.g., benefiting large companies over affected sectors of the population). From a different perspective, VOX could easily combine anti-immigration and anti-EU rhetoric to reinforce its positioning in the cosmopolitan-parochial divide that is increasingly salient in West European Politics (Ford & Jennings, 2020). These strategies could in turn influence the effect of EU preferences on voting, provided that partycueing is operating (Pannico, 2017; Hobolt, 2007; Steenbergen et al., 2007). In that sense, a combination of salient and critical discourses towards the EU among radical parties of the left and the right could translate into a more critical public opinion, especially among their voters, which could be reflected again in voting patterns. At the moment of writing these lines the anti-European strategy seems much more likely in the case of VOX, which has recently started to campaign along the lines of the debate between globalism or motherland (*globalismo o patria*). If successful, this strategy could incorporate relevant changes to the electoral scenario in Spain, a country where was historically feasible to combine a strong national identity with a positive perception of the EU (Carey, 2002; McLaren, 2004).

#### Appendix

See Tables 12.3, 12.4, and 12.5



**Table 12.3** *Note* Entries are coefficients with standard error in parentheses and average marginal effects (AMEs) in brackets \**p* < 0.10, \*\**p* < 0.05, \*\*\**p* < 0.01


*Note* Entries are coefficients with standard error in parentheses and average marginal effects (AMEs) in brackets

\**p* < 0.10, \*\**p* < 0.05, \*\*\**p* < 0.01



*Note* Entries are coefficients with standard error in parentheses and Average Marginal Effects (AMEs) in brackets \**p* < 0.10, \*\**p* < 0.05, \*\*\**p* < 0.01

#### References


**Hugo MarcosMarne** is an Assistant Professor at the Department of Political Science of the University of Salamanca, where he is also a member of the Democracy Research Unit (DRU). His research focuses on the intersection between political attitudes and electoral behavior, national identities, and populism.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# Consequences of EU Politicisation for Voting in National Elections

#### *Marina Costa Lobo*

This book started from a puzzle that has recently emerged in Europe. Since 2009, the EU has undergone a series of crises, which posed challenges in different policy areas, from monetary and fiscal, to foreign, migration and health (Genschel & Jachtenfuchs, 2016; Schmidt, 2020). The response to these challenges has generally been to deepen EU integration, with a greater supranationalisation of policies (Laffan & Schlosser, 2016). In doing so, the EU has become more political, acting in a purposeful way, rather than simply following the Treaties, as its institutions have often intervened in a discretionary way to respond to the Eurozone crisis, COVID-19, or even Russia's invasion of Ukraine (Middelaar, 2019).

Still, there has been no institutional change since the Treaty of Lisbon to match these political and policy trends. In addition, there is evidence to show that the European Parliament elections remain second-order

M. C. Lobo (B)

© The Author(s) 2023

323

Institute of Social Sciences, University of Lisbon, Lisbon, Portugal e-mail: marina.costalobo@ics.ulisboa.pt

M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6\_13

(Nielsen & Franklin, 2017). How, then, is the EU being held accountable for these increased competencies, given the lack of institutional change, and the unchanged "second order" character of EP elections? In this book, we argue that it is fundamental to shift our academic gaze to the domestic level of politics in order to understand how Europe is being held accountable by national institutions. There has been research on the way in which European legitimacy is achieved through national institutions (De Vries, 2007; De Vries & Hobolt, 2016), yet this book contributes to that discussion by offering novel/unique perspectives and evidence on how national institutions are holding the EU accountable.

Following Powell's classification (Powell, 2004), our objective was to examine the chain of EU responsiveness which can be established through national institutions, namely media, parliamentary debates and legislative elections, in the following countries: Belgium, Germany, Greece, Ireland, Portugal and Spain. We show that the EU politicisation has an impact at the individual voting level, with citizens using their EU preferences to express their party choices at the national level. This means that national politicians, MPs and governments are being (s)elected, and held accountable, for their stance on European integration. What this volume concludes is that the European Union is being held accountable nationally.

We now discuss the conclusions of our research to engage and assess the empirical results from a comparative perspective. We will consider first our findings about the context within which EU politicisation occurs across Europe in the media and parliamentary arenas, and second our findings on the consequences of EU politicisation for the vote at the national level.

Chapters 2, 3 and 4 allowed us to understand the differentiated EU information context within which electors have been placed over time. Research on EU politicisation has evolved, with current studies focusing on the "differentiated" forms it may assume (Braun et al., 2019; Hutter & Kriesi, 2019). Among theorists of EU integration, there was a sense that Euroscepticism fed on "constitutive" issues of membership and institutional design (Mair, 2000), whereas, "politicizing European policies" was likely to lead to a European public sphere (Risse, 2015). We took these perspectives on board to evaluate salience, tone, as well as the types of issues which are being politicised over time, by mainstream media and by parliamentary parties in plenary debates in each country in the twenty-first century.

Our analysis was driven by a media-centric approach which considered all articles (news and op-eds) that focused on the EU. Indeed, other studies of EU politicisation (Hutter & Grande, 2014) have measured the concept in terms of "how parties discuss the EU in the media", which we have shown conflates measures of parties and media which are not shaped similarly. In each country's mainstream media, salience and contestation are substantially correlated in each newspaper analysed, independent of their ideological leaning. Distinguishing between news articles and opeds, it becomes clear that the EU has been, in general, more salient, as well as more contested, in opinion articles, and that the differences between left leaning and right leaning newspapers increase when the focus is exclusively on op-eds. These results suggest that, while the journalistic coverage of the EU may be similar during legislative campaigns, op-eds may be a better object of enquiry when researching the way media shape EU politicisation. Further, focusing on salience and contestation provides a relatively incomplete picture of how the traditional media politicised the EU before and after the eurozone crisis, since it ignores the topics being covered. Our analysis shows that all countries' media are focusing on *policies* rather than *membership*, signalling an increasing tendency for the EU to be discussed in terms of the consequences of its policies, rather than questioned on membership per se. Yet, a more fine-grained analysis shows that while Germany and Ireland discuss EU policies per se to a larger extent, the Southern European countries discuss the EU from the perspective of domesticated policies. Indeed, even within the domain of "policies", it is possible to discuss their design from a European perspective, and only a few countries are doing that. In the Southern European countries, debates on Europe tend to take the EU policy for granted, and the news focuses on the domestic consequences of EU policies.

Turning to the longitudinal analysis of EU politicisation in parliamentary plenary debates, Chapter 3 by Kartalis and Silva shows that Parliamentary parties behave rather strategically in that arena. Namely, salience tends to be higher when EU contestation is low. Ie the larger mainstream parties will discuss the EU less when they share the plenary with an Eurosceptic party which devotes part of their speeches to contesting the EU. Thus, in Parliamentary debates, salience tends to be higher at lower levels of contestation. In the countries which we analyse, and with the partial exception of Greece, national parliaments have not made a very noticeable contribution to the politicisation of the EU, nor did this change with the Eurozone crisis. Concerning the determinants of the politicisation of the EU in parliament, a Eurosceptic position is the strongest determinant of both salience and, especially, contestation.

Assessing the salience of the EU's multiple dimensions, results suggest policy-related issues are the most salient topics in both the media and parliaments. Within the realm of policies, Chapter 4's findings by Santos and Rogeiro Nina confirm that these mostly concern economic and financial matters. Importantly, the greatest differences between parliaments and the media are found in levels of tone. With the exception of Greece, the share of texts (articles or speeches) about the EU that have a negative tone is much greater in the media than it is in parliaments, confirming the negative bias in political news coverage, and the strategic behaviour of parties in Parliament. The picture which emerges then from this in-depth longitudinal analysis of mainstream media and parliamentary debates EU politicisation in Europe is one of differentiation between media and parliaments, rather than parallelism between the two arenas in each country.

Overall, the chapters suggest that EU politicisation is established broadly in a similar fashion across Europe. Yet, it does not function similarly in the media and parliamentary arenas. While salience and contestation are substantially correlated in the mainstream media, this does not tend to be the case in Parliaments. This suggests that parties in Parliament act strategically to reduce EU contestation, whereas this does not occur to the same extent in the media since it operates under a different logic. Thus, the media are contributing to a greater extent to EU politicisation than Parliamentary debates, while the latter are the ideal arena to observe parties' preferences unfiltered, but in dialogue with other parliamentary parties. This matters for the overall nature of politicisation, as citizens may be receiving mixed messages from different sources, which becomes obvious when both arenas are considered simultaneously.

In the second part of our book, we established the relative importance of EU issue voting using a variety of methods and focuses on the *consequences* which EU politicisation may have for that factor of voting behaviour.

Chapter 5's experimental results by Pannico and Lobo showed unequivocally that there is EU issue voting, with parties being punished on average if their EU position is different from the voter's. Analysing the data per country produces similar findings, with the exception of Spain where no EU issue voting is detected. Overall, this constitutes a major finding that confirms the indirect path of EU representation and accountability via the national elections. We have therefore unequivocally demonstrated, in the experimental analysis, that EU attitudes are a cause rather than a consequence of voting behaviour. Chapter 5's analysis was then confirmed in Chapter 6, by Heyne, Lobo and Pannico. In it, observational data from post-election surveys fielded in the four debtor countries is employed to understand the relative importance of the EU issue. Results show that respondents' EU position proximity visà-vis the party they voted for matters significantly in determining vote choice, but to a lesser degree than left–right proximity. Thus, while the exogenous impact of the EU issue is clear, it is not a predominant factor of voting behaviour. Moreover, there are no clear differences between voters of challenger and mainstream parties when it comes to the relevance of the EU position proximity in the vote calculus. Indeed, in the bailout countries not only is left–right proximity always a more important explanatory factor than EU issue proximity, generally for voting behaviour this hierarchy is maintained both for mainstream and challenger parties.

When the importance of the EU has been considered for national politics, the perspective has been mostly on parties and party systems (Hooghe and Marks, Hutter and Kriesi). Further, these analyses consider the importance of the EU not on its own, but as part of a larger cleavage which has been variously called the "globalisation", "integrationdemarcation" or "GAL-TAN" cleavage, their goal is to understand the degree to which this new cleavage is supplanting the left–right cleavages in terms of party competition. There is a view that this cleavage is replacing the left–right cleavage as a determinant of voting behaviour. While we are not evaluating the importance of this (variously denominated) cleavage per se, vis-à-vis the left–right dimension, we do evaluate, in the most thorough way possible the importance of European preferences for voting. As we explained in the previous paragraphs, we have established EU issue voting in such a robust way as to be sure that it matters, but we can also qualify its importance vis-à-vis other factors which are determinant of voting behaviour, in particular in relation to the left–right dimension. We establish that the EU issue cross-cuts the left–right dimension, as has been determined for the party systems. Yet, we do not find that the EU issue is replacing the left–right dimension in terms of its importance for voters, whether they vote for mainstream or more extreme parties. This simply is not happening at the voter level.

Whereas EU issue voting has been established comparatively in Chapters 5 and 6, some differences between countries emerged. Therefore, the rest of the book was dedicated to country case chapters. The country case chapters are very rich in detail on the intricacies of EU issue voting across Europe. Taken together, the message is similar to the one emerging from the comparative chapters. Using full multivariate models that include other relevant political variables, we show that EU issue voting is occurring not only for voters of Eurosceptic parties, that politicise the issue in Parliament, but also for mainstream parties, in every single country considered. Namely, Vlams Beland (Flanders) and the Christian Democrats (Wallonia) in Belgium, as shown in Chapter 7 by Stiers; the AfD, CDU and FDP in Germany as shown in Chapter 8 by Navarrete and Debus; Syriza, Kinal and ND in Greece, as shown by Nezi in Chapter 9; Sinn Fein and Fine Gael in Ireland as shown by Heyne in Chapter 10; PCP and PSD in Portugal as shown by Lobo in Chapter 11; and finally Vox, PP and Cs in Spain as shown by Marne in Chapter 12. Thus, we have managed to congregate very strong multimethod evidence in favour of EU issue voting in each of these countries, in the post-crisis period.

While Chapter 3 showed that the parties' Eurosceptic position explains both salience and tone, leading to strategic behaviour, what is interesting is that, as both Chapter 5 and the country chapters show, EU issue voting is not circumscribed to voters of Eurosceptic parties. Instead, citizens voting for different types of parties, namely mainstream and challenger parties, use the vote to express EU preferences. This was shown experimentally in Chapter 5, with the exception of Spain. It was also shown observationally, in Chapter 6, and in the country case chapters.

Both the comparative Chapters, as well as the Country case studies also analyse in different ways how the EU politicisation context matters for voting. In Chapter 5, the association between EU media and salience and EU issue voting is tested. Results suggest that there is an association between the availability of EU issues and the strength of EU issue voting, while the tone appears less relevant. Chapter 6 interacts the parties' tone in plenary debates and the strength of EU issue voting, showing that for parties that have a more negative tone towards the EU (hence, have a more Eurosceptic discourse), the EU proximity matters more in determining the voters' choice than for parties with a positive or neutral EU tone in parliament.

Overall, the book's findings concerning the consequences of politicisation for voting behaviour matter for different reasons. First and foremost, they have consequences for the debates about EU legitimacy which are ongoing today. Most of the efforts for democratising Europe involve supranational reforms, at the level of the European Parliament, or citizens' initiatives at the EU level (Hennette et al., 2019). Yet, what our research suggests is that there is vertical accountability in national legislative elections regarding the EU. Namely, voters of both mainstream and challenger parties are using their positions on the EU to determine the vote. Therefore, the national channel of EU accountability, and namely the domestic institutions which contribute to it, need to become a central part of the debate on EU legitimacy.

Secondly, our findings have consequences for the study of EU politicisation. We have demonstrated that EU politicisation media and parliamentary debates suggest that they do not work in tandem, they follow different trends and that citizens are taking cues from both arenas. Importantly, the media has a negative bias as issues become more salient. Yet, even when parties actively tried to depoliticize the issue, such as in the cases of Ireland, Germany or Portugal, the EU issue still appears as relevant for vote choices, not only for challenger but also for mainstream parties. This suggests that politicisation works through multiple channels, depending on the arena being considered, and it impacts voters differently, depending on whether we consider the media or parliaments.

Finally, our findings suggest that while the chains of accountability are similar from Greece to Germany, in many respects, there are important differences too. Namely, whereas all countries mostly discuss "policies" rather than "membership" in the media and parliamentary debates, suggesting the EU's legitimacy, they do so differently. Whereas Spain and Portugal mostly discuss "domesticated EU policies", they hardly discuss "EU policies", while the latter tend to dominate debates in Germany. These findings suggest again that not all politicisation is alike, and it is necessary to go beyond salience and tone, to see the dimensions which are being debated as we do in this volume. Also, it is the case that in the experimental study, EU issue voting is not significant in Spain. Further, while EU issue voting is significant using observational data in all countries, both for mainstream parties and extreme ones, whether these belong to the left or right side of the ideological spectrum varies between countries.

#### Further Research

Our research findings notwithstanding, they also raise issues that call for more investigation. Namely, we should focus on deepening the patterns of differentiation in the different arenas and between countries. Thus, even when salience and tone are converging across Europe, there may be fundamentally different perspectives on Europe originating from different countries that can only be understood from a closer, qualitative analysis of discourse. These deeper insights into the topics and frames of EU politicisation may give important insights on the nature of politicisation and whether it helps or hinders EU legitimacy. Further, the way we analyse the relationship between the EU media salience and tone and the vote is at the aggregate level, and should be explored further at the individual level. In addition, studies should take into account other arenas of politicisation such as social networks, given their enormous and ever-growing importance. While the comparison of the media and parliamentary debates' patterns of politicisation opened up new perspectives on how voters are receiving multiple and different cues, we still did not explore these alongside social networks patterns of politicisation. We expect that the latter may follow a similar pattern to the mainstream media, but this needs to be confirmed. Finally, other countries should be included in the analysis in order to consolidate the overall finding concerning the importance of EU issues for voting in national elections.

#### References


**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

# Index

#### **A**

Anti-EU, 69, 183, 219, 234, 253, 311 Anti-immigration, 12, 122, 146, 217, 262, 267, 270, 311 Austerity, 12, 147, 148, 150, 206, 231, 233, 235–237, 239, 240, 243, 245, 253, 257, 258, 276, 301 Austerity measure, 31, 94, 105, 109, 233–237, 239, 244, 245 Automated text analysis, 44

#### **B**

Bailout, 2, 16, 36, 40, 42, 43, 45, 120, 143, 147, 150, 232, 233, 235, 236, 240, 244, 246, 253, 254, 257–259, 264, 276, 283–285, 289 Bailout countries, 26, 27, 31, 36, 38, 43–45, 145, 148, 153, 275, 279, 327 Belgium, 2, 6, 13, 14, 16, 26, 31, 32, 36, 38, 40, 44, 48, 64, 66,

68–72, 84, 92, 100, 115, 119, 120, 126, 127, 130, 133–137, 143, 169–171, 173, 175, 179, 180, 186, 187, 324, 328 Brexit, 1, 17, 43, 49, 77, 91, 95, 103, 119, 135, 231, 251, 252, 254–257, 259, 264, 265 Brussels, 170, 171, 206–208, 216, 222

#### **C**

Centre-right, 17, 122, 132, 136, 209, 236, 245, 246, 254, 255, 264, 265, 277, 278, 282, 283 Challenger parties, 12, 15, 18, 144, 145, 147, 149, 153, 155, 157, 238, 279, 327–329 Cleavage, 12, 87, 187, 232, 277, 327 Communism, communist, 17, 146, 277, 284–286, 288–290 Conservatism, conservative, 146, 240, 241, 256 Constitutional arrangements, 94 Constitutive issue, 43, 63

333

© The Editor(s) (if applicable) and The Author(s) 2023 M. Costa Lobo (ed.), *The Impact of EU Politicisation on Voting Behaviour in Europe*, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-031-29187-6

Constraining dissensus, 3, 5, 28, 85 Content analysis, 7, 27, 56, 59, 92, 152 Council of the European Union, 49, 77, 135 Covid-19 pandemic, 215, 216, 223 Creditor countries, 36, 70, 115, 130, 206, 233 Critical election, 36, 43, 45

#### **D**

Debtor countries, 45, 63, 327 Deliberation, 57 Demarcation-integration, 327 Democracy, 3, 9, 11, 57, 90, 179, 182, 183, 186, 189, 192, 195, 197, 231, 243, 277–279, 281 Democratic deficit, 94, 116 *Domesticated issues*, 6, 41, 84

#### **E**

Economic voting, 207, 235 Electoral behaviour, 2, 116, 144 Electoral campaign, 27, 31, 39, 91, 92, 208, 216, 252, 255, 264 Electoral preference, 10, 116 Enlargement, 10, 68 Environment, environmental issues, 2, 11, 128, 131, 219, 221 EU contestation, 36–38, 40, 45, 61, 64–75, 77, 91, 152, 325, 326 EU issues, 4, 11–13, 17, 56, 61, 86, 116–118, 127, 130, 131, 146, 156, 207, 209, 214, 215, 220, 232, 242, 258, 264, 284, 299, 301, 302, 305–307, 330 EU issue voting, 2, 4, 5, 8, 11, 12, 15–17, 26, 30, 31, 45, 61, 115–119, 123–128, 130, 131, 137, 138, 143–145, 147–149, 155–157, 170, 171, 173, 177,

184, 186, 207, 223, 262, 264, 275–277, 286–288, 290, 299, 302, 307, 310, 311, 326–329 Euro, 3, 211, 232 European Commission (EC), 10, 49, 77, 103, 135, 233, 243 European Council, 10, 49, 77, 103, 117, 135 European integration, 3, 11, 26–29, 31, 33, 43, 44, 49, 56–59, 61–63, 68, 75, 77, 85, 88–90, 101, 103, 104, 107, 117, 124, 135, 145–147, 151, 178, 181, 183, 184, 186, 205, 206, 212, 213, 217, 219, 220, 222, 232–235, 242, 245, 253, 277, 287, 306, 309, 324 European Monetary Union (EMU), 3, 116 European Parliament elections, 1, 10, 11, 206, 258, 323 European Parliament (EP), 4, 6, 9–11, 13, 16, 18, 49, 64, 77, 86, 91, 103, 117, 135, 145, 215, 279, 280, 301, 324, 329 Eurosceptic, 5, 7, 11–13, 15, 17, 30, 31, 61, 63, 68, 69, 71, 72, 74, 91, 97, 101, 119, 123, 125–128, 130, 133, 147–149, 154, 157, 220, 235, 259, 262, 264, 265, 277–279, 284, 290, 301, 311, 325, 326, 328 Euroscepticism, 30, 61, 89, 146, 147, 151, 153, 251, 253, 254, 264, 277–280, 302, 305, 324 Eurozone, 1, 6, 28, 31, 35, 44, 49, 77, 94, 103, 105, 108, 109, 130, 135, 206, 214, 232, 233, 235, 240, 259

Eurozone crisis, Euro crisis, Economic crisis, 2, 3, 6, 8, 9, 12, 14, 16, 17, 26, 29–31, 35–38, 40,

42–46, 56, 60, 62, 63, 66, 68–70, 72, 75, 93, 95, 97, 99–101, 115, 143, 145, 147, 150, 171, 232–237, 239, 241, 245, 246, 251, 252, 254, 258, 259, 264, 275, 276, 278, 279, 283, 323, 325


#### **F**

Far-right party, 278, 281 Finances and Taxes, 9, 106, 109 Fiscal policy, 276

#### **G**

Gender, 120, 242, 247, 261, 262, 266, 269, 286, 288, 291, 300, 305, 306 Germany, 2, 6, 8, 26, 32, 324, 325, 328, 329 Globalisation, 327 Great recession, 279 Greece, 2, 6, 26, 36, 239–246, 275, 280, 301, 324–326, 328, 329 Greek bailout, 240, 246 Greek crisis, 240 Green, 175

#### **H**

Hard Euroscepticism, 264

#### **I**

Immigration, 106, 109, 127, 261–264, 300, 308, 309, 314, 315, 317 Incumbent party, 262 Institutionalisation, 4, 7, 64, 323, 324 Ireland, 2, 6, 8, 26, 36, 145, 240, 251–255, 259, 261, 264, 265, 275, 324, 325, 328, 329 Issue, 1, 6–9, 237–243, 245 Issue evolution, 300 Issue ownership, 255 Issue position, 7, 151 Issue salience, 289, 301, 305

#### **L**

Left-right dimension, 237, 327 Left-right divide, 259 Left-right issues, 5 Left-right proximity, 327 Left-right scale, 311

#### **M**

Maastricht, 3, 5, 146 Mainstream media, 8, 282, 290, 324–326, 330 Mainstream newspaper, 32, 128, 256, 276 Mainstream parties, 238, 259, 280, 284, 311, 325, 327–329 Media coverage, 26, 31, 35, 90, 91, 215 Membership dimension, 8 Member states, 3, 86, 91, 94, 105, 108, 117, 120, 206, 207, 232 Migration, 106, 109, 323

#### **O**

Opposition, 61, 63, 71, 75, 76, 86, 89, 146, 169, 212, 214, 215,

222, 223, 235, 237–239, 241–243, 253, 255, 256, 258, 261–264, 281, 284, 285, 290

#### **P**

Parliamentary debates, 2–5, 7–9, 12–18, 26, 46, 56, 57, 59–63, 67, 71, 74, 84, 86, 90–94, 115, 144, 145, 148, 149, 152, 154, 155, 157, 173, 175, 211, 212, 214, 223, 236, 237, 330 Parliamentary speeches, 7, 8, 14, 46, 59, 63, 65, 72, 74, 76, 90–93, 97, 99, 101, 210, 214, 220 Party competition, 92, 264, 327 Party support, 121, 122, 132, 235 Party system, 7, 11, 12, 17, 144, 145, 147, 149, 155, 157, 180, 187, 215, 234, 237, 239, 254, 256, 277–279, 282, 290, 301, 327 Permissive consensus, 3, 5, 28, 85 Plenary debates, 7, 13, 16, 59, 67, 69, 76, 88, 148, 157, 158, 324, 325, 328 Polarisation, 6–9, 282 Policy formation, 29 Political change, 101, 147 Political participation, 58, 173, 279 Political system, 43, 87, 118, 179, 223, 231, 252–254 Politicisation, 2, 4–9, 11, 14–16, 18, 26–33, 35–41, 43–46, 57, 61, 65, 66, 73, 76, 85, 115, 116, 128, 144, 148, 186, 187, 264, 275, 276, 282, 285, 289, 324–326, 328–330 Populism, Populist, 146, 169, 231, 232, 235, 240, 278, 301, 322 Portugal, 2, 6, 8, 12–15, 17, 26, 28, 31, 36, 38, 40–45, 63, 64, 66, 67, 78, 84, 92, 99, 100, 102, 115, 119, 120, 126, 130,

135–137, 143–145, 150, 152, 155–157, 232, 240, 275–281, 283, 284, 289, 324, 328, 329 Pro-EU, 5, 11, 12, 15, 17, 69, 119, 120, 125, 126, 128, 130, 133, 146, 219, 220, 236, 253, 254, 258, 277, 279

Protest party, 10

#### **R**

Radical left, 147, 175, 182, 183, 234 Radical right, 11, 146, 182, 215, 308, 310 Refugee crisis, 1, 43, 207 Regression analysis, 73, 242, 288 Rescue package, 232 Responsibility, 233 Responsiveness, 9, 11, 324

#### **S**

Sentiment analysis, 8, 33 Sleeping giant, 116, 245 Social democrats, socialists, socialism, 256, 259, 261, 262, 277, 281, 285–290 Social issues, 252, 255, 265 Social policy, xiv Soft Euroscepticism, 89, 253 South of Europe, south European, 28, 44 Sovereignty, 146, 243, 253, 301 Spain, 2, 6, 8, 12–15, 17, 26, 31, 36, 38, 40–45, 62–64, 66, 67, 71, 78, 84, 87, 90, 92, 94–96, 99, 100, 102, 115, 119, 120, 126, 127, 130, 135–138, 143–145, 150, 152, 155, 156, 275, 280, 299–302, 305, 306, 308–312, 324, 326, 328, 329

#### **T**

Traditional-Authoritarian-Nationalist (TAN), 11, 146 Transparency, 58 Treaty of Lisbon, 1, 49, 77, 103, 135, 253, 323 Treaty of Maastricht, 3, 28, 85 Treaty of Nice, 253 Treaty on the Functioning of the European Union (TFEU), 90 Troika, 49, 77, 103, 135, 233, 234, 279

#### **V**

Voter, 2, 9, 10, 12, 13, 16–18, 87, 116–119, 123–127, 130, 138, 145, 147, 154, 157, 169–173,

177, 179–184, 186, 187, 206, 207, 212, 215–223, 232, 236, 237, 242, 244, 246, 253, 255, 262–264, 279, 288–290, 307–312, 326–330 Voter–party congruence, 118, 119, 124, 130 Voter responses, 307 Voting behaviour, 4, 11, 18, 26, 130, 144, 145, 148, 155, 158, 171, 173, 186, 207, 215, 217, 244, 252, 259, 276–278, 286, 289, 326–328 Voting determinant, 16, 327 Voting predictor, 125, 137, 139, 219, 222, 244, 306, 307

Voting preference, 117, 118, 327