Anti-Establishment Political Parties: Conception, Measurement, and Consequences Teresa Lee Cornacchione

Home , Great South (Italy), Italian Democratic Party, Liberal Party (Iceland), Moderates (Italy), Political spectrum, The Establishment

Florida State University Libraries

Electronic Theses, Treatises and Dissertations The Graduate School

2018 Anti-Establishment Political Parties: Conception, Measurement, and Consequences Teresa Lee Cornacchione

Follow this and additional works at the DigiNole: FSU's Digital Repository. For more information, please contact [email protected] FLORIDA STATE UNIVERSITY

COLLEGE OF SOCIAL SCIENCES AND PUBLIC POLICY

ANTI-ESTABLISHMENT POLITICAL PARTIES:

CONCEPTION, MEASUREMENT, AND CONSEQUENCES

TERESA CORNACCHIONE

A Dissertation submitted to the Department of Political Science in partial fulﬁllment of the requirements for the degree of Doctor of Philosophy

2018

Sean Ehrlich Professor Directing Dissertation

Jonathan Grant University Representative

Carol S. Weissert Committee Member

Brad Gomez Committee Member

Quintin Beazer Committee Member

The Graduate School has veriﬁed and approved the above-named committee members, and certiﬁes that the dissertation has been approved in accordance with university requirements.

ii I dedicate this dissertation to my daughter, Sadie. I hope she always remembers that a girl should be two things: who and what she wants. -Coco Chanel

iii ACKNOWLEDGMENTS

I would like to thank several people for their assistance, professional and personal, in the completion of this dissertation. Thank you to several research assistants, Kevin Monnell, Sara Moustafa, and Linnea Blackmore for assisting in the collection and coding of data. Thank you to my dissertation chair, Dr. Sean Ehrlich, for not only reading multiple versions of everything I have written, and holding to me a high standard, but also making sure my sanity was preserved in this process. I would also like to thank my committee members, Dr. Carol Weissert, Dr. Brad Gomez, Dr. Quintin Beazer, and Dr. Jonathan Grant, for agreeing to serve on my committee. Also, thank you to the political science faculty at Florida State University, particularly, Dr. Matthew Pietryka, and the comparative subfield faculty members, for offering feedback on various drafts of dissertation chapters. Personally, I would also like to thank my parents, Connie and Frank, for their unending support and encouragement. They have always taken a great interest in what I do, and never cease to support me in my endeavors. I also would like to thank my partner, David, for not only providing me with excellent personal support, but also for always being willing to read my work and offer feedback. His love and support not only improved my work, but it made me better (and more efficient) too. Thanks for not letting me float away. Last, but never least, I would like to thank my daughter, who at the time of my completing this dissertation is a little too young to fully grasp what her mother was working toward. She is the light of my life and her smile keeps me going every day.

iv TABLE OF CONTENTS

List of Tables ...... vii List of Figures ...... x Abstract ...... xiii

1 Deﬁning the Anti-Establishment 1 1.1 Deﬁning the Anti-Establishment ...... 5 1.1.1 Distrust of Political Elite ...... 7 1.1.2 Departure from the Status Quo to Alleviate a Crisis ...... 10 1.1.3 Putting the Nation First ...... 11 1.2 From the Right, the Left, and the Center ...... 13 1.2.1 From the Right: The PVV, AfD, and National Front ...... 13 1.2.2 From the Left: The Five Star Movement, Podemos, and Pirates ...... 15 1.2.3 From the Center? En Marche-and the Curious Case of Emmanuel Macron . . 17 1.3 Discussion ...... 18

2 Measuring the Anti-Establishment 20 2.1 From Conceptualization to Measurement ...... 20 2.2 Where are the Parties? ...... 25 2.2.1 United Kingdom ...... 27 2.2.2 Italy ...... 30 2.2.3 Spain ...... 31 2.2.4 Germany ...... 33 2.2.5 France ...... 35 2.2.6 Netherlands ...... 37 2.2.7 Austria ...... 39 2.3 Validity of the Measure ...... 41 2.4 Conclusion ...... 47

3 Letting the Outsiders In: Trust and Electoral Support of the Anti-Establishment 49 3.1 Vote Choice for the Anti-Establishment ...... 51 3.1.1 The European Elections ...... 55 3.1.2 From the Right and Left and Center: Anti-Establishment Parties in the Netherlands, Italy, France, and Germany ...... 56 3.1.3 The Netherlands 2010 ...... 57 3.1.4 Italy 2013 ...... 58 3.1.5 France 2017 ...... 58 3.1.6 Germany 2013 ...... 59 3.2 Data and Methods ...... 60 3.2.1 Dependent Variables ...... 60 3.2.2 Independent Variables ...... 62 3.2.3 Control Variables ...... 63

v 3.3 Results ...... 65 3.3.1 Cross-National Results ...... 65 3.3.2 Country-Level Results ...... 72 3.4 Discussion ...... 97

4 From Making Noise to Making Policy: Parliamentary Activities of the Anti- Establishment 100 4.1 Parliamentary Activities, MPs, and Party Dynamics ...... 102 4.2 The Anti-Establishment Agenda and Legislative Activities ...... 104 4.3 Data and Methods ...... 106 4.3.1 Dependent Variables ...... 107 4.3.2 Explanatory and Control Variables ...... 108 4.4 Results ...... 109 4.4.1 European Parliament ...... 109 4.4.2 Italian Parliament ...... 122 4.5 Discussion ...... 135

5 Left, Right, and Everything in Between: Final Thoughts on the Anti-Establishment Movement 138

Appendices A Anti-Establishment Scores 145 A.1 All European Countries ...... 145 A.2 Greek Parties ...... 153

B Additional Models 155 B.1 Netherlands ...... 155 B.2 Italy ...... 157 B.3 France ...... 160

C Issue Attention of Non-Core Issues 163 C.1 European Parliament ...... 163 C.2 Italy ...... 163

Bibliography ...... 164 Biographical Sketch ...... 175

vi LIST OF TABLES

2.1 Concepts and Operationalization ...... 23

2.2 Correlation Among Anti-Establishment Variables ...... 24

2.3 Anti-Establishment Sentiment Components ...... 24

2.4 Alternative Anti-Establishment Sentiment Components ...... 25

2.5 Predicting Corruption Salience. Anti-Estabishment Sentiment and CHES Anti-Elite Salience Measure ...... 43

2.6 Comparing the Measure to Abedi’s Classiﬁcation ...... 45

2.7 Predicting Sentiment Toward the EU ...... 46

3.1 Probit Model: Voting for an Anti-Establishment Party for National Parliament . . . . 66

3.2 Model 2, Multinomial Results: Voting for an AE Party on the Domestic Level . . . . 69

3.3 Probability of Voting for an Anti-Establishment Party: European Parliament . . . . . 71

3.4 Main Results-Netherlands ...... 73

3.5 Model 5, Netherlands-Multinomial Results ...... 76

3.6 Probability of Voting for the PVV over the VVD ...... 77

3.7 Italian Probit Model: Predicting Support for the Five Star Movement ...... 79

3.8 Italy-Multinomial Results ...... 81

3.9 Probability of Voting for the M5S over the Left Alliance ...... 82

3.10 Probability of Voting for Macron or Le Pen in the 1st Round ...... 85

3.11 Multinomial Results: Votes in the First Round ...... 87

3.12 Probability of Voting for Macron or Le Pen in the Second Round ...... 88

3.13 Democratic Satisfaction and Voting for Anti-Establishment Parties-Germany 2013 . . 91

3.14 Multinomial Results. Eﬀects on Probability of Selecting Parties Over the CDU/CSU- Party List Voting ...... 93

3.15 Multinomial Results. Eﬀects on Probability of Selecting Parties Over the CDU/CSU- Constituency Level Voting ...... 94

vii 3.16 Eﬀects on Probability of Voting for the AfD over Die Linke ...... 95

4.1 Data Availability by Case ...... 107

4.2 European Parliament. Eﬀect of Anti-Establishment Score on the Number of Questions Asked ...... 110

4.3 European Parliament. Number of Parliamentary Questions Anti-Establishment Par- ties versus Establishment Opposition ...... 111

4.4 European Parliament. Eﬀect on Probability of Asking a Question Devoted to a Core Issue ...... 114

4.5 European Parliament. Eﬀect on Probability of Asking a Question Devoted to a Core Issue Anti-Establishment Parties versus Establishment Opposition ...... 115

4.6 European Parliament. Proportion of Questions Devoted to Core Issues Across Anti- Establishment Scores ...... 117

4.7 European Parliament. Proportion of Questions Devoted to Core Issues Anti-Establishment Parties versus Establishment Opposition ...... 119

4.8 Core Issue Attention in the European Parliament ...... 120

4.9 Non-Core Issue Attention in the European Parliament ...... 121

4.10 Italian Parliament. Inﬂuence of Anti-Establishment Score on Number of Parliament Questions Asked ...... 123

4.11 Italian Parliament. Anti-Establishment Members versus Establishment Opposition Members ...... 124

4.12 Italian Parliament.Eﬀect on Probability of Asking a Question Devoted to a Core Topic127

4.13 Italian Parliament. Issue Prioritization. Eﬀects on the Probability of Asking a Ques- tion Devoted to a Core Issue: Anti-Establishment versus Establishment Opposition . 128

4.14 Italian Parliament. Issue Prioritization. Proportion of Questions Devoted to Core Issues Across Anti-Establishment Scores ...... 130

4.15 Italian Parliament. Issue Prioritization. Proportion of Questions Devoted to Core Issues. Anti-Establishment Parties versus Establishment Opposition ...... 132

4.16 Core Issue Attention in the Italian Parliament ...... 134

4.17 Non-Core Issue Attention in the Italian Parliament ...... 135

A.1 Full List of European Parties and Anti-Establishment Scores for the Most Current Manifesto Year-European Union Countries Plus Norway, Switzerland, and Iceland . . 145

viii A.2 Predicting Corruption Salience. Anti-Establishment Sentiment and CHES Anti-Elite Salience Measure - Nationalism Excluded ...... 152

B.1 Netherlands. Probability of Voting for the PVV-Disaggregated Trust Measure . . . . 155

B.2 Netherlands. Probability of Voting for the PVV over the VVD-Disaggregated Trust Measure ...... 156

B.3 Italy. Probability of Voting for M5S-Disaggregated Trust Measure ...... 157

B.4 Italy. Probability of Voting for M5S over the Left Alliance-Disaggregated Trust Measure158

B.5 Italy. Multinomial Model with Lega Nord Voters Excluded ...... 159

B.6 France. Probability of Voting for Macron in the First Round-Disaggregated Trust Measure ...... 160

B.7 France. Probability of Voting for Le Pen in the First Round. Disaggregated Trust Measure ...... 161

B.8 France. Probability of Voting for Macron over Le Pen in the Second Round. Disag- gregated Trust Measure ...... 162

ix LIST OF FIGURES

2.1 Cross-National Anti-Establishment Sentiment ...... 26

2.2 British Parties’ Anti-Establishment Sentiment 2015 ...... 28

2.3 British Parties’ Anti-Establishment Over Time ...... 29

2.4 Italian Parties’ Anti-Establishment Sentiment 2013 ...... 31

2.5 Italian Parties’ Anti-Establishment Sentiment Over Time ...... 31

2.6 Spanish Parties’ Anti-Establishment Sentiment 2016 ...... 32

2.7 Spanish Parties’ Anti-Establishment Sentiment Over Time ...... 33

2.8 German Parties’ Anti-Establishment Sentiment 2017 ...... 33

2.9 German Parties’ Anti-Establishment Sentiment Over Time ...... 34

2.10 French Parties’ Anti-Establishment Sentiment 2017 ...... 36

2.11 French Parties’ Anti-Establishment Sentiment Over Time ...... 37

2.12 Dutch Parties’ Anti-Establishment Sentiment 2012 ...... 37

2.13 Dutch Parties’ Anti-Establishment Sentiment Over Time ...... 38

2.14 Austrian Parties’ Anti-Establishment Sentiment 2013 ...... 39

2.15 Austrian Parties’ Anti-Establishment Sentiment Over Time ...... 40

2.16 Anti-Establishment Sentiment v. CHES Measure ...... 42

2.17 Predicting Corruption Salience:Anti-Establishment Sentiment v. CHES Anti-Elite Salience ...... 44

3.1 Probability of Voting for an AE Party: Domestic Level ...... 67

3.2 Probability of Voting for an AE Party: European Parliament ...... 72

3.3 Predicted Probability of Voting for the PVV: Probit Model ...... 74

3.4 Predicted Probability of Voting for the PVV over the VVD ...... 78

3.5 Predicted Probability of Voting for the M5S: Probit Model ...... 80

3.6 Predicted Probability of Voting for the M5S over the Left Alliance ...... 83

x 3.7 France 2017. Predicted Probabilities of Voting for Macron or Le Pen in the First Round 86

3.8 France 2017. Predicted Probabilities of Voting for Macron v. Le Pen in the Second Round ...... 89

3.9 Germany 2013. Predicted Probability of Voting for AfD ...... 92

3.10 Germany 2013. Predicted Probability of Voting for Die Linke ...... 92

3.11 Predicted Probability of Voting for AfD over Die Linke-Party List ...... 96

4.1 Predicted Number of Questions-European Parliament May 2014-May 2017 ...... 110

4.2 Predicted Number of Questions Anti-Establishment Parties versus Establishment Op- position ...... 112

4.3 European Parliament. Predicted Number of Questions By Anti-Establishment Status and Ideology ...... 113

4.4 European Parliament. Predicted Probability of Asking a Question Pertaining to a Core Issue Across Anti-Establishment Scores ...... 114

4.5 European Parliament. Predicted Probability of Asking a Question Pertaining to a Core Issue Anti-Establishment Parties versus Establishment Opposition ...... 116

4.6 European Parliament. Predicted Proportion of Agenda Pertaining to a Core Issue Across Anti-Establishment Scores ...... 118

4.7 European Parliament. Predicted Proportion of Agenda Pertaining to a Core Issue Across Anti-Establishment Parties versus Establishment Opposition Scores ...... 119

4.8 European Parliament. Disaggregated Core Issue Attention of Left-Wing and Right- Wing Anti-Establishment Parties ...... 121

4.9 Italian Parliament. Predicted Number of Questions Across Anti-Establishment Scores and Ideology ...... 124

4.10 Italian Parliament. Predicted Number of Questions Anti-Establishment Parties versus Establishment Opposition ...... 125

4.11 Italian Parliament. Predicted Number of Questions by Anti-Establishment Status and Ideology ...... 126

4.12 Italian Parliament. Probability of Asking Question Pertaining to a Core Issue Across Anti-Establishment Scores ...... 127

4.13 Italian Parliament. Marginal Eﬀect of Ideology on the Probability of Asking a Ques- tion Pertaining to a Core Issue Anti-Establishment Parties versus Establishment Op- position ...... 129

xi 4.14 Italian Parliament. Predicted Probability of Anti-Establishment Members Asking a Question on a Core Issue-Across Ideology ...... 129

4.15 Marginal Eﬀect of Ideology on the Proportion of Agenda Devoted to Core Issues Across Anti-Establishment Scores ...... 131

4.16 Proportion of Agenda Devoted to Core Issues. Anti-Establishment Parties versus Establishment Opposition ...... 132

4.17 Proportion of Anti-Establishment Members’ Agendas Devoted to Core Issues - Across Ideology ...... 133

4.18 Italian Parliament. Disaggregated Core Issue Attention Left-Wing and Right-Wing Anti-Establishment Parties ...... 134

A.1 Alternative Anti-Establishment Measure-Cross National Placement ...... 151

A.2 Alternative Measure of Anti-Establishment Sentiment-Nationalism Excluded . . . . . 151

A.3 CHES v. Anti-Establishment Sentiment, Nationalism Excluded ...... 152

A.4 Predicting Corruption Salience ...... 153

A.5 Cross National Party Placement Including Greece ...... 153

A.6 Greek Parties’ Anti-Establishment Placement 2015 ...... 154

A.7 Greek Parties Over Time ...... 154

C.1 European Parliament. Attention to Non-Core Issues Among Anti-Establishment Par- ties By Ideology ...... 163

C.2 Italian Parliament. Attention to Non-Core Issues Among Anti-Establishment Parties By Ideology ...... 163

xii ABSTRACT

The incredible rise of so-called “anti-establishment” parties in Europe has left scholars scram- bling to define and classify the movement. Much scholarly attention has been paid to radical right wing parties, and the sources of their electoral support. While important and intriguing, the current literature has yet to develop a cohesive definition of the anti-establishment, and has too heavily used terms such as “populist,” “anti-establishment,” and “radical right wing” interchangeably. Further, extant research has based theories of these parties’ electoral support largely with the radical right wing in mind, potentially ignoring theories that could explain support for these parties from the left, right, and center of the political spectrum. Finally, current research has not substantially explored how these parties, traditionally excluded from policy-making, behave once they are seated in parliaments. This dissertation aims to remedy these three shortcomings. First, I develop a conceptual definition and measurement scheme that encapsulates both ideological positioning and anti-establishment sentiment. Then, I explore how political trust influences electoral support for anti-establishment parties positioned at all areas of the classic left-right spectrum. Finally, I analyze their parliamentary behavior, assessing their level of activity and their preferred policy domains. My findings underscore the importance of conceiving anti-establishment parties as existing along a unique dimension, separate from ideology, whose electoral viability can be explained via a unified theory, and who behave uniquely in parliament.

xiii CHAPTER 1

DEFINING THE ANTI-ESTABLISHMENT

Across Europe, an interesting and potentially alarming trend has taken place: the rise of outsider parties. Often, the leaders of these parties come from non-political backgrounds, with little to no governing experience. The policy consequences of these parties’ rise could be tremendous and damaging, as they aim to deviate wildly from the status quo, and are comprised of actors who often hold radical or extreme views. Take for example, the Italian Five-Star Movement. Founded by stand-up comic, Beppe Grillo, the party has skyrocketed from a movement largely based on social media, to a fully-fledged political party attaining the largest share of votes of any one single party in the 2018 Italian General Election. The current leader of the party, Luigi Di Maio, was a waiter before serving in parliament in 2013. His new coalition partner, Matteo Salvini, is hardly from the political mainstream either. As leader of the xenophobic and radically right, Lega Nord, Salvini also stands to dramatically influence the trajectory of Italian politics. Italy is hardly unique. Other non-traditional parties from across the continent are increasingly viable. Recently, Alternative for Deutschland (Germany) attained the position of third largest party, and the largest opposition group, in the German Bundestag. It is the first time since World War II that a radical right wing party is seated in the German parliament. The party’s rhetoric invokes German nationalism and is decidedly anti-immigrant. Its leadership, like that of the Five-Star Movement, largely consists of non-politicians, such as Alicia Weidel, one of the two parliamentary leaders, who came to parliament from the banking sector. This pattern is repeated throughout Europe. Alternative for Deutschland, the Five-Star Move- ment, the Austrian Freedom Party, The Dutch Party for Freedom, Podemos in Spain, Pirate parties across Scandinavia, the French National Front, the Danish People’s Party, and the United Kingdom Independence Party, are only but a few of the popular and potentially influential non-traditional parties that are shaping European politics. Many scholars and pundits alike have termed these parties as “anti-establishment.” While some researchers have identified certain common characteristics of these parties, such as anti-elite sentiment (Polk, Rovny, Bakker, Edwards, Hooghe, Jolly,

1 Koedam, Kostelka, Marks, Schumacher et al. 2017), challenging the status quo (Abedi 2009), and claiming to fight corruption as an alternative to the mainstream (Hartleb 2015), the discipline still lacks a single definition and a systematic classification scheme of the movement. The closest the discipline has to a measure of a party’s anti-establishment rhetoric comes from the 2014 Chapel Hill Expert Survey (CHES) (Polk et al. 2017).1 However, this measure only classifies parties in one year: 2014, missing newer parties that have emerged, such as France’s En Marche!, and currently does not account for how parties may shift from one election to the next, precluding over-time analyses. Aside from the CHES measure, the typical treatment of the anti-establishment movement suffers from three shortcomings. First, it is often conflated with the radical right. Second, it is blankly termed populism. Third, it is assumed to be dichotomous: a party is either mainstream or anti-establishment. The problem with the first approach is that it neglects parties whose rhetoric is anti-elite, anti-mainstream, and thus, anti-establishment, from the left and the center. The issue with the problem with interchangeably using the terms “populism” and “anti-establishment,” is that it lacks conceptual clarity, neglecting to adequately identify and define the key features of populism espoused by these parties. The third problem, treating parties dichotomously, fails to consider the importance of “anti-establishment” sentiment as its own separate dimension, apart from ideology, and assumes that all such parties possess the same level of antipathy toward the mainstream, and ignores how parties evolve in their anti-establishment sentiment over time. I seek to remedy these three issues. I argue that being an anti-establishment party amounts to expressing antipathy toward the mainstream. This antipathy is, like ideology, a spectrum, along which all parties can and do exist. As scholarship and media have chosen to focus on the radical right, other important and influential parties from the left and center have been ignored (Müller2017). Antipathy toward the mainstream is not owned by parties on the right. Parties from everywhere on the political spectrum can espouse rhetoric that is distrustful of the status quo and mainstream political elites. Scholarship risks mistakenly assuming that the motivations of the radical right are the motivations of the entire anti-establishment phenomenon. For instance, it is assumed that supporters of these parties are motivated by xenophobic attitudes, and that these parties’ agendas are focused on an exclusive sense of nationalism. Yet, while the PVV in the Netherlands is notoriously xenophobic, the Five Star Movement (M5S) in Italy is not particularly

1Available from https://www.chesdata.eu/2014-chapel-hill-expert-survey.

2 hostile to immigration for cultural reasons.2 Neglecting treatment of the anti-establishment as a distinct movement has also led scholars to ignore similarities of such parties from across the ideological spectrum. For example, the United Kingdom Independence Party (UKIP) and the M5S are coalition partners in the European Parliament, despite their marked difference with respect to ideology. By focusing on the radical right, scholars risk making both type 1 and type 2 errors: falsely categorizing ideologically extreme parties as anti-establishment, and overlooking parties from elsewhere on the ideological spectrum that display an extreme affinity for anti-establishment ideals. Recent work has even espoused that populism in Europe is exclusively owned by the radical right (Müller2017). The argument is that these parties are also “anti-pluralist” - meaning that these parties do not espouse inclusive rhetoric, and this is the property of the radical right. However, I argue that this is fundamentally false. Claims about the elite, are not sufficient to consider a party anti-establishment or populist, but additional sentiments such as claiming to be to true protectors of the nation, and rescuing the country from an unspecified crisis exist all along the political spectrum.3 Commonly, scholars and those in the media have termed the anti-establishment movement as “populist” (Taggart 2004, Betz 1994, Inglehart and Norris 2017, Müller2017). Populism as a concept is convoluted and can take on a variety of meanings, resulting in a catch-all definition of the movement: if they are not establishment, they are populist (Taggart 2004, Canovan 1981, Mudde 2010). While it may be the case that these parties espouse populist rhetoric, the current trend of referring to the movement as “populist” has resulted in a sloppy categorization of these parties with little regard to which features of these parties’ rhetoric and behavior is exactly populist (Bale, Van Kessel, and Taggart 2011). Ultimately, using the term has resulted in yet another dichotomous

2The Five Star Movement is highly unconstrained with regard to its policies and ideology (Mosca 2014). Many have argued recently that the party is moving to the right on the issue of immigration (See https://www.theatlantic.com/international/archive/2018/03/italys-populist-youth/554408/ and https://www.washingtonpost.com/news/worldviews/wp/2018/03/05/italys-election-is-another-blow-to-the- european-establishment/?noredirect=on&utm term=.f87d2236d4ce). Certainly, the party’s alliance with UKIP in the European Parliament could be viewed as evidence of the party’s right turn. However, scholarship has acknowledged that transnational links are not adequate to classify parties, see Mair and Mudde (1998). Further, I argue the party’s unconstrained approach to politics is provides an example of precisely why a classiﬁcation separate from ideology is needed to address such parties.

3For a more detailed explanation of this argument, see Müller2017. He argues that attention to the grass-roots nature of populist movements is misleading. While I do agree with Müllerthat critiques of the elite are a necessary but not a sufficient condition to classify parties as anti-establishment, I depart from Müllerin his assertion that these attributes are somehow owned by the radical right-wing in Europe.

3 classification: mainstream and populist. I argue that there are specific features of populism that these parties espouse, such as “vertical populism,” the perceived detachment of the political elite from the people (Hartleb 2013). Further, these parties promote the use of tools of direct democracy, such as referenda, to enact changes to the status quo. The recent electoral success and prominence of these non-traditional parties requires a con- crete conceptualization of the nebulous “anti-establishment.” In the end, the term does not matter as much as the definition. Whether scholars choose to call these parties “outsiders,” ”anti- establishment,” or “populist,” is irrelevant, as a rose by any other name would still smell as sweet. The term we use is not as consequential as precision in identifying the attributes that define these parties and the development of a systematic way to classify them. To my knowledge, the only other scholarly work to attempt this comes from recent work by Polk, et al. (2017) which presents a classification of European parties in 2014 by expert political scientists. The measure considers anti-elite sentiment in such parties, and importantly, does acknowledge that this sentiment exists on the extreme left and extreme right, but does not consider parties over time, nor does it consider the emergence of this sentiment in the ideological center. Parties can and should be distinctly identified, and systematically classified according to their level of anti-establishment sentiment. Here, I develop a theoretical conceptualization of the anti- establishment movement. Primarily, I argue that “anti-establishment” sentiment should be con- ceived as spectrum, similar to, but distinct from, ideology. While it is not incorrect to divide parties into two categories: mainstream and outsiders, doing so misses important attributes of these parties. Rather, scholars should recognize the continuum along which these parties exist. Just as ideology can vary over time, the degree to which parties espouse antipathy toward the mainstream or status quo is also fluid. This conceptual definition permits the classification of parties along two dimensions: anti-establishment sentiment and ideology. This contributes to our understanding of parties’ composition and belief, and offers new avenues of future research for assessing these parties’ sources of support and institutional activities. I demonstrate the importance of conceptual clarity and the careful assessment of these parties that is distinct from ideology.

4 1.1 Deﬁning the Anti-Establishment

Defining the anti-establishment movement, and consequently classifying parties, is a complex endeavor. Scholars have devoted much attention to the proper grouping of parties (Mair and Mudde 1998). Parties’ self classification, and the use of party names have been used as sources to properly assign parties. Other methods have focused on sociodemographic characteristics of party members (McGann and Kitschelt 1995), cleavages (Lipset and Rokkan 1967), or the makeup and preferences of parties’ supporters (Ignazi 2003). Despite all these various approaches, the most agreed upon method is the exclusive use of ideology (Mair and Mudde 1998, Mudde 2007). Even when parties are ideologically similar, assignment to a “party family” can be difficult. For example, even classifying radical right wing parties, both populist and non-populist, can be met with confusion (Mudde 2007). Additionally, scholars have concentrated on institutional reasons for the rise of radical right wing parties. When establishment parties are situated closely to one another on the ideological spectrum, radical parties then have space to emerge (Kitschelt 1994). Sartori’s (1976) work marries the number of parties with ideological distance, or polarization. Future research could question this link between polarization and the emergence of anti-establishment parties (Abedi 2002). While often discussed, the concept of an “anti-establishment” party is seldom clearly defined. I argue that the anti-establishment “movement” actually represents a continuum of beliefs about the political status quo, elites, the neglect of the people in political decision making, and nationalism. I term this continuum “anti-establishment sentiment.” I contend that all parties express these sentiments and vocalize this rhetoric to varying degrees. It is those parties expressing this antipathy loudly that are often termed “anti-establishment.” Just as a party’s ideology can shift over time, so too can its expressed antipathy toward the mainstream, or anti-establishment sentiment. I argue it is just as important to devote attention to the anti-mainstream rhetoric expressed by these parties as it is to where on the ideological spectrum they reside. In addition to attempts to define the movement, scholars have spent considerable time trying to determine the causes of these parties’ success and the consequences of their electoral viability (Golder 2016, Bustikova 2014, Agerberg 2017, McDonnell and Newell 2011). Most commonly, researchers and pundits have focused heavily on a certain type of party: the radical-right wing, using terms “radical-right” and “anti-establishment” interchangeably. The conflation of these two

5 movements has led to scholars either ignoring center and left anti-establishment parties, or neglect to acknowledge the same theories can explain the movement holistically. Additionally, classification of these parties has been largely dichotomous: a party is either establishment or an outsider. Further, scholars have focused on the rise of the radical right and neo-Nazi parties such as the Golden Dawn in Greece or Jobbik in Hungary. These groups only represent part of the anti- establishment phenomenon. This focus on the radical right, while worthy of scholars’ attention, fails to consider the importance of the anti-establishment left or center, or even acknowledge that a separate dimension apart from ideology exists. The left-right ideological spectrum is insufficient to understanding what anti-establishment parties are, how they behave, and who supports these political actors, as anti-establishment parties and actors can arise from anywhere on the political spectrum. Several scholars have acknowledged that radicalism and extremism can exist on both the ideological left and the right (Golder 2016, Abedi 2009). Additionally, few researchers have offered a conceptualization of the anti-establishment that encompasses both ends of the ideological spectrum (Polk et al. 2017). In several works, Abedi defines an anti-establishment party as one that challenges the status quo, presents itself as alternative to all mainstream parties, and asserts a fundamental divide between the people and the elite (Abedi 2009, Abedi 2002, Abedi and Lundberg 2009). While the definition does not preclude the existence of the anti-establishment left or center, it does imply a dichotomous classification of parties. This approach misses the nuances of parties’ messaging, and ignores that both traditional and outsider parties can share negative sentiments about mainstream politics. Similarly, anti-establishment parties are cited as sharing several criteria. Attributes such as labeling themselves as an opposition party, asserting corruption on the part of political elite, and demanding forms of direct democracy (Hartleb 2015). While helpful in identifying the movement’s common attributes, criteria is numerous, with little to no indication as to which criteria represent necessary and/or sufficient conditions. In short, the qualitative definitions of the anti- establishment movement provide a helpful starting point for identifying characteristics of such parties. However, they often imply a dichotomous classification of parties (mainstream or outsider), lack clarity on which attributes are important, and remain subjective in their assignment of the “anti-establishment” moniker.

6 I argue that the anti-establishment can be defined along a unique dimension that is separate and distinct from ideology: anti-establishment sentiment expresses a fundamental antipathy toward the mainstream. I claim that this antipathy toward the mainstream, and thus an anti-establishment party, is defined by three core tenants. First, anti-establishment parties express a great deal of distrust in political elites, who they allege have launched the country into a crisis. Second, they advocate the use of tools of direct democracy to depart from the political status quo, in order to save the country from an existing or a looming crisis. Finally, they express a “nation first” attitude. Further, antipathy toward the mainstream exists as a continuum. Parties can and do shift over time in their antipathy rhetoric, just as they do ideologically. All parties may express this antipathy, but the degree to which they emphasize this sentiment determines the extremity of their anti-establishment position. Also, as parties evolve they may soften their rhetoric, or change their behavior as they achieve elected office and join in coalitions (Albertazzi, McDonnell, and Newell 2011, Albertazzi and McDonnell 2008).

1.1.1 Distrust of Political Elite

Populism and the Anti-Establishment. Research in this area has focused on the movement’s use of populism (Betz 1994, Taggart 2004, Golder 2016, Inglehart and Norris 2017, Müller 2017). Many have used the terms “populist,” “anti-establishment,” and “radical right wing” interchangeably, or even stacked together. As such, these parties are referred to as a movement with only vague references to what these parties actually espouse. Also, by focusing on the radical right, this approach fails to consider the left and center’s use of populism, or even spec- ifying what is distinctly populist about the anti-establishment movement. The problem with populism is that it tends to be a “chameleon concept” that can take on a variety of meanings (Taggart 2004, Canovan 1981, Mudde 2010, Müller2017), and apply to any group that appears to react to a national crisis (Taggart 2004). While the concept of populism can explain or provide context to the modern anti-establishment movement, it is too often used vaguely or conflated with other concepts, such as the radical right. The crucial question is whether the current crop of anti- establishment parties can be classified as populist. The short answer to this question is a qualified “yes.” While populist rhetoric is used extensively by the radical right, it is largely “horizontal populist” rhetoric that aims to draw a distinction between the people and others, such as immigrants

7 (Hartleb 2013, 359). However, “vertical” populist rhetoric aims to draw a distinction between the people and the elite (Hartleb 2013, 359) and is present in all of what many would term “anti- establishment” parties. I argue that all anti-establishment parties espouse rhetoric that represents vertical populism, to draw distinctions between the people and the political elite. Horizontal populism may be present in the right wing variety of anti-establishment parties. But all anti- establishment parties will incorporate vertical populism into their rhetoric, regardless of where they are situated on the ideological spectrum, as demonizing the elite is the primary objective of anti-establishment parties. This hearkens to Mudde’s (2004) conception that populism draws a distinction between the “pure people and the corrupt elite”(p.543). Commonly, these parties call for the removal of corrupt elites who have failed the people. For example, page three of the Italian Five-Star Movement’s manifesto states, “The Parliament no longer respects the people.”4 Further, non-oﬃcial rhetoric from these parties implicates establishment political elites as incompetant and violating the people’s trust. Beppe Grillo, founder of the Five-Star Movement once said, “The amateurs are the ones conquering the world and I’m rejoicing in it because the professionals are the ones who have reduced the world to this state.”5 So while it is certainly not incorrect to characterize anti-establishment parties as populist, doing so without attention to the type of populist rhetoric invoked does little to enlighten our understanding of these parties, or their motivations and behavior. At their core, anti-establishment parties express a fundamental distrust of traditional mainstream parties, both in government and in opposition (Abedi 2009, Abedi and Lundberg 2009). This distrust takes a populist face, and signals the ineptitude of the political elite. The mainstream is corrupt, and the existing elites have mishandled public resources. This has created a political crisis in the state, where traditional politicians and parties cannot be trusted to to handle the serious challenges the country faces. Anti-establishment parties are often vague about the nature of the crisis. However, certain European parties, such as the Dutch PVV, have been able to politi- cize events, such as the recent refugee crisis to label their current governments as incompetent (Van Kessel 2011).

4Available from https://manifesto-project.wzb.eu/ 5See http://www.euronews.com/2016/11/14/political-amateurs-are-conquering-the-world-beppe-grillo-tells- euronews

8 There is a claim among anti-establishment actors that there is distinction between interests of the people and the actions of the elite (Mudde 2007, Abedi 2009). Messages from theses parties often take the form of empathizing with the “common man.” For example, Beppe Grillo, the founder of Italy’s left-wing Five-Star Movement (M5S), has declared, “A lesson to everyone: you can’t lie forever to the people without suffering consequences.”6 Capitalizing on Italian governments’ propensity to collapse, Grillo and his “movement” position themselves between the people and the political elite. Further, these parties tend to relish their role as an underdog (Hartleb 2015), often touting their non-political experiences as a way to garner support and connect with the common voter, who is also outside the political elite class (Mazzoleni, Stewart, and Horsfield 2003). Implicit in this claim is the idea that the political elite are corrupt and incapable of handling the diverse problems of the nation. Corruption is widespread and systemic. They do not constrain themselves to critiquing the governing party. Rather, they indict all members of the establishment, even those occupying the opposition (Abedi 2009). Their claim is that any ill, including economic woes or other crises (Albertazzi and McDonnell 2008), in the current political environment is the fault of corruption on behalf of the political elite. This emotional assignment of blame to elites (Hameleers, Bos, and de Vreese 2016) underscores the core argument of the anti-establishment that the people are at odds with the elite (Mudde 2007, Aalberg and de Vreese 2016, Albertazzi and McDonnell 2008). In this, the anti-establishment messaging emphasizes a negative tone, claiming that the political system is fundamentally dysfunctional due to widespread corruption on the part of elites (Müller2017). They emphasize statistics that paint a pessimistic picture of the state of the nation, amplifying concerns about the system and manipulating fears threading through the public and adopting them as their platform (Williams 2006). In doing so, they can claim to have issue ownership over “the will of the people.” Anti-establishment parties speak to wanting true democracy and adherence to what the everyday citizen desires, even if those desires are only reflective of a fringe segment of the population. They then attribute the blame for these negativities on the current corruption propagated by existing political elites. In this they provide “solutions and scapegoats” (Pasquino 2008). They proclaim that electing someone outside the system, i.e. their

6Spoken after the defeat of Matteo Renzi’s constituitonal referendum, December 2016, https://www.washingtonpost.com/news/worldviews/wp/2016/12/04/trying-to-imitate-trumps-victory-italys-beppe- grillo-could-help-to-throw-europe-into-crisis/?utm term=.593f9e581bb2.

9 anti-establishment party, is the only way to ﬁx systemic problems. They assert that because the anti-establishment operates outside of the political apparatus it is more representative of the people’s interests, and thus, more qualiﬁed to govern than establishment parties.

1.1.2 Departure from the Status Quo to Alleviate a Crisis

In addition to, and perhaps as a consequence of, their fundamental distrust of elite, anti- establishment parties reject the political status quo. Further, these parties claim that a departure from the status quo is the only to save the country from a current or a looming national crisis. For example, UKIP, in advocating for Brexit, claimed detaching from the European Union was the only way to protect the country from the impending crisis of immigration and the dangers it posed (Goodwin and Milazzo 2017). Sartori’s (1976) concept of the “anti-system” party may be the earliest iteration of this aspect of the modern anti-establishment movement. These parties exist within the context of a polarized political system. Anti-system parties share the common feature of a “delegitimizing impact,” where they question the legitimacy of the regime they oppose, and provide an “opposition of principle” (Sartori 1976, 133). The principled stand that these parties take against the status quo is observed throughout anti-establishment movements. This includes claiming that the system will beneﬁt from more “democracy,” as the current problems in the country can be attributed to establishment elites failing to represent the population, and address critical problems. As a unique form of opposition, these parties assert that dramatic systemic change is needed to purge the political system of the damage done by existing political elite, and rescue the country from a crisis. To achieve this end, anti-establishment parties tout a populist message claiming that the best way to solve the country’s problems is through dramatic change to the status quo, subject to the people’s direct approval. In this they often advocate instruments of direct democracy. An example of this comes from the United Kingdom, and the push for Brexit, the UK’s self-removal from the European Union (EU). Not only does exiting the EU represent a massive departure from the status quo, it was largely based on a populist message. Those advocating leaving the EU claimed that the institutional arrangement hurt the common British citizen. They argued that a distant, and questionably democratic, institution, could not possibly understand or represent the basic needs of Britons. Britain, they championed, would largely be better oﬀ deciding its own future. The message of Brexit was populist, as was its implementation. Those promoting

10 leaving the EU, pushed the British government into a popular referendum to decide the issue. This complicated process of leaving an international institution was decided via a process of direct democracy, and represents a populist strategy. The status quo of British politics is fundamentally disrupted, creating rift within the Conservative Party, and forever altering its political relationship with other European states. This pattern of behavior was also repeated in the recent March 2018 Italian election. Both Lega and the M5S campaigned on holding a referendum on Italy’s use of the Euro. This would be a major policy shift, and both parties advocated changing this policy via a populist tool: referendum voting. However, departing from the Euro would require a constitutional amendment (Groppi 2012). These cases are just two examples of how the anti-establishment advocates for placing proliﬁc policy changes into the hands of the people, and away from a perceived corrupt and inept political elite class.

1.1.3 Putting the Nation First

Finally, anti-establishment includes rhetoric claiming that the nation must come first. This attitude can be expressed in two ways. First, it may be expressed as xenophobic rhetoric, or an exclusive sense of nationalism. Scholars often pay disproportionate attention to this attribute of the anti-establishment. Certainly, the hyper-patriotic rhetoric of the far-right is both noteworthy and alarming. This concept is well-illustrated by the anti-establishment and hyper-nationalist party, National Front (FN) in France. Marine Le Pen, the current leader of the party, is unashamedly anti-immigrant, not only toward irregular immigration, but any immigration. Quoted as saying, “a multicultural society is a society that has multiple conflicts,”7 Le Pen’s party has made immigration central to its campaign, fueling fears about increases in immigration and its impact on French culture and security. The combination of a sharp disdain for outsiders, and a sense of restoring “French” values underscores the party’s resolve to restore national sovereignty from what it considers an illegitimate EU government. The second way a nation-first attitude manifests is through inclusive nationalism, which is often ignored at the expense of attention devoted to exclusive nationalism. The “nation first” attitude can be expressed via patriotism, including rejecting membership in any organization that chips away at

7As quoted on the French Presidential campaign trail in 2017. https://www.thetimes.co.uk/edition/world/i-ll- curb-ﬂood-of-migrants-let-in-by-eu-vows-le-pen-qzgqtlcbh.

11 national sovereignty, or even promoting a larger or leading role of the nation within organizations. The rhetoric does not have to be xenophobic, as not all anti-establishment parties express racism. For example, the Italian Five-Star Movement is not anti-immigration in the same way Lega is. However, the party seeks to remove Italy from the Euro, supports Britain’s effort to leave the EU, and operates in a coalition with the United Kingdom Independence Party (UKIP) in the European Parliament, based on a shared eurosceptic sentiment. Other parties, that are neither radically right nor xenophobic, also express a level of patriotism. France’s newest president, Emmanuel Macron, founder of the new party En Marche! is a devoted centrist. However, even Macron expresses the idea of restoring national pride to an implicitly de- jected nation. In his inaugural address he prioritized “giving back to the French people confidence,” and providing the world and Europe a “strong France.”8 While hardly an ideological extremist, his rhetoric underscores the sentiment expressed in many anti-establishment parties that the nation, on a decline, needs uplifting and a sense of purpose, one that has been damaged by status quo politics. The “nation-first” attitude therefore, is not necessarily nationalist in the sense of promoting national superiority at the expense of “others.” Rather, the most important aspect of the anti- establishment version of nationalism, is the desire of the party to restore a sense of nostalgic glory. The recent re-introduction of mandatory military service in France, put forth by Emmanuel Macron, was established to re institute a sense of national pride and patriotism among the youth of France. This is an example of attempting to restore national glory, that has been supposedly dampened by status quo politicians.9 In essence, placing the nation first, may have little to do with dividing the population between the pure nationals and outsiders. Instead, these parties charge that under the current political elite, the population has lost a sense of their nationality. To these parties, placing the nation first means restoring a deeper sense of patriotism, and selling themselves as the true protectors of the nation’s identity, legacy, and history. In short, anti-establishment sentiment is defined by a distrust of established political elite who have hurled the country into crisis. Further, anti-establishment rhetoric proposes remedying this crisis by rejection of the status quo, achieved through the tools of direct democracy, and a desire to

8Available from http://www.businessinsider.com/full-text-emmanuel-macron-ﬁrst-speech-president-2017-5.

9https://www.economist.com/europe/2018/03/01/emmanuel-macron-wants-to-bring-back-national-service-in- france.

12 put the nation ﬁrst, which can stem from an inclusive or an exclusive sense of nationalism. These sentiments exist along a continuum, and, to a degree, all parties express this rhetoric. However, it is those parties high in these sentiments that qualify as truly “anti-establishment.”

1.2 From the Right, the Left, and the Center

To illustrate the variety of parties coming from all areas of the political spectrum, I highlight a few cases of interest from Europe. The parties below are all instrumental players in their respective countries’ domestic governments. Many have seats in domestic parliament, with some even controlling or supporting the government. Most have considerable representation in the European Parliament, placing pressure on mainstream parties, and European integration.

1.2.1 From the Right: The PVV, AfD, and National Front

The most commonly acknowledged parties that belong to the anti-establishment are those who also hail from the radical right. Below are three examples of radical right wing parties who also espouse anti-establishment rhetoric.

The Dutch Party for Freedom. The Dutch Party for Freedom (PVV), made its rise out of the ashes of List Pim Fortuyn, a xenophobic party. The party has steadily made gains over the past few Dutch elections, first gaining a handful of seats (six) in the 2006 general elections. In 2010, the party attained the rank of third largest party in Dutch parliament, and increased its seat share even more in 2012. In 2017, the party was in legitimate contention to steal away a majority, and become the governing party. They lost out to another right-wing, but mainstream, party, the VVD. In this wake of the most recent election, the party pledged to support the government. The party’s leader, Geert Wilders is known for his flamboyant rhetoric surrounding Muslims and immigrants. The party’s manifesto explicitly cites Islam as a threat to Dutch lifestyle and security, calling it a “constant source of unrest” in the their 2012 election program.10 Due to inflammatory remarks about the religion and his production of a controversial documentary, he is banned from the UK. Beyond the party’s anti-Islam rhetoric, the party also asserts a need to drastically change

10From page 46 of the party’s manifesto, available from https://manifesto-project.wzb.eu/

13 the status quo in the Netherlands by advocating for the abolition of the Senate, and the increased use of referenda.11

Alternative for Deutschland. Another radical right wing party is the German Alternative for Deutschland (AfD). The party is relatively new, and has recently attained the status of third largest party in the German Bundestag, making them largest opposition party, and imparting upon them important roles, such as managing the budget committee. Formed from disgruntled members of the CDU, the party has made incredible strides in German politics in a relatively short time (Arzheimer 2015), largely by challenging the government of Angela Merkel and her pro-EU and immigration policies. Interestingly, the party’s supporters are noted to be male, young, well-educated, and interested in politics, though dissatisﬁed with democracy (Jankowski, Schneider, and Tepe 2017). The party promotes a shockingly xenophobic agenda whereby they claim that the mishandling of issues such as immigration and the refugee crisis on the party of the mainstream parties (i.e. CDU/CSU and SPD) have propelled the country into a crisis, and potentially a “culture war.”12

National Front. Perhaps one of the most famous, if not infamous, right wing European parties is the French National Front. It is not a new party, in fact, it has been influencing French politics for several decades. Its founder, Jean Marie Le Pen is noted Holocaust denier and xeno- phobe. His daughter, Marine Le Pen, has attempted to alter the image of the party, but xenophobia is still a mainstay of the party’s platform. The party claims that France is being undercut by the influx of immigrants into the country, and making it distinctly non-French. The party’s manifesto calls for preserving French culture, French language, and French national identification.13 The National Front is also anti-EU, claiming that giving French sovereignty over to the organization has damaged French identity and culture. While the National Front has long been involved in French politics, its recent electoral popularity is due in large part to shift from xenophobic discourse to populist rhetoric (Stockemer and Barisione 2017). Marine Le Pen is credited with increasing the party’s membership by portraying herself as a

11From page 29, of the party’s 2012 election program, available from https://manifesto-project.wzb.eu/

12From page 33 of the party’s 2017 manifesto, available from https://manifesto-project.wzb.eu/.

13From the 2017 party manifesto, available from https://manifesto-project.wzb.eu/.

14 “crisis manager” who could reconstruct France (Stockemer and Barisione 2017, 104). The party’s most recent manifesto repeatedly refers to restoring France by returning the the country to the French people. Further, the party has historically relished its role as the restorer of law and order, economic well-being, and French culture (Lewis-Beck and Mitchell II 1993).

1.2.2 From the Left: The Five Star Movement, Podemos, and Pirates

While scholarship has devoted much attention to right-wing parties, they are not the only parties that display anti-establishment sentiment. Anti-establishment parties on the left have become increasingly popular, both within and outside of parliaments. Three parties below, the Italian Five- Star Movement, Spanish Podemos, and numerous Pirate parties have enjoyed success in domestic parliaments, European Parliament, and in drawing attention to their agendas.

The Five Star Movement. In 2012, the Five-Star Movement (M5S), a left wing anti- establishment party, founded by Beppe Grillo, a stand-up comedian, and Gianroberto Casaleggio, an internet strategist, achieved victories across Italian government. In March 2018, they won the most votes of any single party in the general election. Now, under the leadership of Luigi Di Maio, the party has joined the right-wing Lega Nord to form the “Italy First” coalition. In 2013, the Five-Star Movement was not just a new party, constructed from the political elite. Rather, it presented candidates outside the political “caste,” heavily lambasted by Grillo himself (Natale 2014, Biorcio 2014, Franzosi, Marone, and Salvati 2015). Online campaigning combined with candidates from outside the political elite produced a winning strategy. They became the third largest parliamentary group in 2013, and in 2018, the party with the largest vote shares 14 The party now stands to dramatically alter Italian politics, and Italy’s place in the EU. While it is my argument here that the party is a left-wing anti-establishment party, I must address that this is a newly contentious point in the discipline and the study of anti-establishment parties. The party is highly unconstrained with regard to policy and ideological leaning, and its ﬁve page manifesto does little to enlighten their viewpoints. However, their 2013 program does call for some classically left-leaning policies, such as government subsidized higher education, and increases in the social safety net. In 2013, the manifesto year used in this dissertation (the most recently

14Italian electoral laws changed prior to the 2018 election, and Italian parties were allowed to run as a coalition. The Five-Star Movement did not run in a coalition with any other party, but they attained more votes than any other party. The right coalition, de facto led Silvio Berlusconi, and formally by Matteo Salvini and Lega Nord, achieved the most votes as a coalition.

15 available manifesto year provided by the CMP), the RILE score of -49.03 places it as a left-wing party. Additionally, the CHES measure also places it to the left of center, though not quite as far left as the RILE score from the CMP. There can be little doubt that the party (at least at its inception) is a left-wing party. Scholars today point to the party’s supposed shift to the right on the issue of immigration. However, being anti-immigrant is not strictly the under the purview of right-wing parties. Labor unions have espoused anti-immigrant views in an eﬀort to protect domestic workers, and are also distinctly left wing supporters (Citrin, Green, Muste, and Wong 1997). Others point to the partnership between M5S and UKIP in the European Parliament. I argue that this red herring represents a marriage of convenience. Parties in the European Parliament divide along ideological lines, and outsider parties such as UKIP, M5S, National Front, AfD, etc. are unlikely to join groups based around mainstream parties. Further, being a member of a group in European Parliament is essential to parties’ functionality within the body (Hix, Noury, and Roland 2006). Scholars have noted that transnational ties are not ideal at best, and inappropriate at worst, for classifying parties into families (e.g. Mair and Mudde 1998). Therefore, it is my contention that the shift on immigration and M5S’s alliance with UKIP and other right wing parties is insuﬃcient for claiming that the party is on the right. If anything, the confusion surrounding the party’s ideological placement underscores the dire need to perceive a second dimension separate from ideology.

Podemos. Formed in January 2014 from university researchers in Madrid, Podemos (We Can), escalated quickly in its popularity and appeal (Ramiro and Gomez 2017), effectively ending the Spanish two-party system (Sola and Rendueles 2018, Orriols and Cordero 2016). Podemos is heavily influenced by the mainstream left opposition (Ramiro and Gomez 2017). While frequently characterized as “populist,” scholars often also claim that their supporters are not classic populists themselves (Ramiro and Gomez 2017). Podemos supporters are largely cited as being anti-mainstream, disaffected, and disengaged (Orriols and Cordero 2016). The party itself reflects this dissatisfaction. It rails against the classic left-right political divide, and considers itself a representative of those individuals cut out from traditional politics (Ramiro and Gomez 2017). The Spanish political environment appeared prime ground for the rise of such a party. Economic and political crises in the country left many Spanish citizens disillusioned with the traditional two party system. The recession, a banking crisis, and political corruption scandals all led to the rise of multiple outsider parties, including Podemos (Ramiro and Gomez 2017, Orriols and

16 Cordero 2016, Sola and Rendueles 2018). Not only did the party perform well in the European Parliamentary elections, mere months after its formation, it grew in popularity in 2015, gaining ground in many local elections. The party’s group currently occupies 67 seats in the lower house of the Spanish parliament.

Pirates. Another crop of left-leaning anti-establishment parties are the pirate parties. These groups are largely anti-elite; they use the cyber age to promote an increasingly “vertical” populist message (Hartleb 2013). The most recent pirate party to gain attention is the Icelandic Pirate Party. The party formed in response to the 2008 Icelandic banking crisis. While it enjoyed only meager vote shares for the ﬁrst few years (Indridason 2014), in 2016, it surged in popularity, and was invited to form the government. The collapse of the banking sector, propelled its anti-elite message, and its promotion of direct democracy. This is the latest in a string of pirate parties, the most noteworthy coming from Sweden and Germany. The Swedish Pirate Party is the earliest example of a successful pirate party, founded in 2006 (Erlingsson and Persson 2011). Mostly drawing its support from the youth, the Swedish Pirate Party advocates personal integrity and anti-commercialism. The German Pirate Party also enjoyed a non-trivial amount of support. In 2011, the party made gains in local elections, making it akin to other “ﬂash” parties, that seemingly arise from nowhere (Hartleb 2013). While their momentum has waned since 2013, their message is distinctly anti-establishment, promoting that they are the party for everyone, and that the use of technology facilitates mass participation, which in turn, is a check agains mainstream elites (Hartleb 2013). They have even been termed “good populists” in the German media (Hartleb 2013, 364).

1.2.3 From the Center? En Marche-and the Curious Case of Emmanuel Macron

Perhaps one of the more perplexing cases is that of a centrist anti-establishment movement. Emmanuel Macron does not appear to have much in common with other anti-establishment actors. He is not a political newcomer, as he served as the Minister of Economy under the Hollande administration. Contrary to other European anti-establishment actors, Macron is not eurosceptic. Rather, Macron is a globalist. However, his new party, En Marche!, is nonetheless representative of a populist, anti-status quo politics. The party is rooted in recruiting common citizens for politics,

17 increasing the status of France in the international community, and even is committed to changing domestic institutions, such as moving France to a proportional representation system. While their rhetoric may not be as vile as that of the National Front, it is an anti-establishment message nonetheless. The party aims to “reconnect with the spirit of French conquest,” reduce the number government oﬃcials and elites in political oﬃce, and further develop citizen participation in the day to day running of the French government. The party also swept up voters disillusioned by the mainstream, which in France, had become increasingly polarized (Evans and Ivaldi 2017).

1.3 Discussion

The current left-right ideological spectrum is insufficient to explain and account for the variety of parties currently termed “anti-establishment.” The movement is defined by rhetoric that asserts a fundamental distrust of political elite, who have hurled the country into a crisis, a push to change the status quo via populist tools of direct democracy, and a nation-first attitude that can be expressed as inclusive or exclusive nationalism. Parties from the right, such as AfD, and the PVV, express these sentiments, as do parties from the left, such as Podemos, and the M5S. Importantly, we also see these behaviors from parties in the political center (i.e. En Marche!). While scholars and pundits have indeed taken notice of the increasing popularity of these anti- establishment parties and candidates, the extant literature lacks both a comprehensive way to classify such parties, and a unified explanation for their increasing electoral gains. The world’s attention is on rise of the anti-establishment and its consequences: from Brexit to multiple domestic elections. However, the focus has largely been on explaining the sources of support for these parties while lacking a clear definition, or systematic classification scheme of “anti-establishment parties.” In the chapters that follow, I introduce a measure by which scholars can systematically conceive and classify parties according to their anti-establishment sentiment over time, explore the sources of their support, and assess their parliamentary activities. Chapter Two introduces the measure of anti-establishment sentiment, and details its construction. The measurement is constructed using the Comparative Manifesto Project (CMP) Data. Using examples from several European countries, I demonstrate the location of parties along two dimensions: ideology and anti-establishment sentiment. I show that while some parties may reside on opposite sides of the ideological spectrum,

18 they may indeed share common anti-establishment sentiments. I also demonstrate how parties may change of over time with regard to their expressed level of anti-establishment rhetoric. Chapter Three explores the sources of these parties’ support. I depart from traditional explanations of radical right wing support, and instead focus on how institutional trust (diffuse support) influences vote choice for anti-establishment parties. Using data from the 2014 Eurobarometer, I demonstrate this relationship cross-nationally. Leveraging the 2010 Dutch election, the 2013 Italian election, the 2017 French Presidential election, and the 2013 German Bundestag election, I show how this relationship works in the context of domestic elections. These elections each represent a case where an anti-establishment party made electoral gains. In some instances, these parties achieved substantial growth in parliament, and even attained political leadership. Those people who express low institutional trust (diffuse support) are more likely to vote for anti-establishment parties over mainstream parties on the left and right. This relationship holds when controlling for ideology, and other typical explanations for radical right wing support. The findings underscore the importance of considering anti-establishment sentiment as a separate dimension apart from ideology, with unique explanations for their support. In Chapter Four, I assess anti-establishment parties’ parliamentary activity. While these parties are traditionally “noise-makers” outside of parliament, their recent electoral viability has created a need to more deeply understand their role as policy makers. I explore their level and content of their parliamentary activities in the European parliament and in the Italian Chamber of Deputies. I find that anti-establishment actors are more generally more active than establishment actors, both in government and in opposition. However, the policy focus of these parties varies. Finally, in Chapter Five I conclude with a discussion of the anti-establishment phenomenon, with suggestions for future research.

19 CHAPTER 2

MEASURING THE ANTI-ESTABLISHMENT

2.1 From Conceptualization to Measurement

I argue that anti-establishment sentiment is deﬁned by three core beliefs. First, anti-establishment parties believe political elite are distrustful, corrupt, and have hurled the country into a crisis. Sec- ond, they believe that in order to rescue the country from a crisis, the country needs to depart from the political status quo, using the people and tools of direct democracy. Third, these parties espouse that the country should be focused on a “nation-ﬁrst” approach in its policies. These attributes can manifest in multiple ways. Largely, anti-establishment parties levy charges of en- trenched corruption within the political elite. This corruption, they charge, has led the country into a crisis that the current mainstream political elite are ill-equipped to remedy. They promote change, primarily suggesting the use of populist tools, such as referenda. These parties also invoke nationalist rhetoric, either inclusive, such as restoring the nation’s reputation, or exclusive - by either calling for the removal of outsiders who muddy national identity, or restoring the nation’s glory on the international stage. I use the Comparative Manifesto Project Dataset (CMP) to develop a measure of anti-establishment sentiment in European countries. I use a select group of European countries, where anti-establishment parties have enjoyed a measure of electoral success: United Kingdom, Italy, Spain, Germany, France, the Netherlands, and Austria,1 to illustrate party positions. Manifesto data is the most comprehensive and formal source of party platform information. The party publishes a manifesto and prioritizes its agenda, while claiming ownership over key issues. The CMP calculates a percentage of the manifesto devoted to a particular agenda issue. These appear as quasi-statements. I leverage these quasi-statements determine a party’s level of anti-establishment sentiment. 1Because of the nature of the CMP data, I can and have calculated this measure for all parties in the manifesto dataset. I simply choose to present these countries and parties here for ease of illustration.

20 First, as part of their message, anti-establishment parties hurl accusations of widespread corruption on the part of the political elite. I operationalize this concept using the variable Political Corruption. This variable measures the percentage of the manifesto devoted to mentioning political corruption in the political system. Specifically, it mentions the need to eliminate abuses of power. These abuses have led the country into a crisis that the mainstream is unable to solve. This gives a sense of urgency to the anti-establishment message. To capture this, I include a measure from the CMP that assess quasi-statements about there being a general crisis in the country, using the variable General Crisis. Often, anti-establishment parties are unclear about the nature of the crisis. These parties tend to frame the crisis as the consequence of the political elite’s abuse of power, which has irreparably damaged the country, and the anti-establishment party is the only party that can remedy the crisis and save the country from the misdeeds of mainstream politicians. For example, multiple sections of the German AfD’s 2017 manifesto begin with a statement of crisis, such “Our Health System is in Danger,” or ”Islam’s Conflict with our Values.”2 Both statements indicate a sense of urgency in the system. I also argue that to fix this crisis, anti-establishment parties promote departing from the political status quo. Further, they advocate that these changes are enacted via the tools of direct democracy. To capture this belief, I assess favorability of the democratic regime, including support for citizens to be involved in democratic decision making. I utilize the variable Direct Democracy: Positive. This particular variable evaluates the amount of the manifesto dedicated to calling for referenda, and other forms of direct democracy. Referendum voting is a favorite tool of the anti-establishment. In their 2013 manifesto, the Italian M5S calls for the direct participation of citizens (via the internet) in every public meeting. The AfD demands that citizens have referendum power over legislation to save the people from the “oligarchy who has the lever of state power.”3. Even the centrist anti-establishment party, France’s En Marche!, outlines a plan for “citizen juries” to evaluate parliamentary work.4 Finally to capture the “nation first” attitude, I use the variable Fabric of Society: General. This variable not only assesses how much the manifesto discusses patriotism, but it also captures how

2From the 2017 manifesto, available from https://manifesto-project.wzb.eu/.

3From page 7 of the 2017 manifesto, available from https://manifesto-project.wzb.eu/.

4From page 29 of the 2017 manifesto, available from https://manifesto-project.wzb.eu/.

21 much the manifesto calls for the suspension of freedoms in order to prevent subversion of the state. Anti-establishment parties from everywhere on the ideological spectrum, express a nation-first sentiment. Right wing parties are most noted for their nationalism language, such as the French National Front’s promise to defend French culture, French language, and French sovereignty.5 Sim- ilarly, AfD identifies Islam as a threat, and calls for the “German people to be preserved,” even encouraging increases in population growth from white Germans.6 However, left anti-establishment parties also espouse nationalism (though it may be more inclusive). From the left, Podemos (Spain), calls for special offices to ensure the teaching of Spanish language and culture, and emigration offices to help assist Spanish emigrants to return to Spain.7 Centrist parties, such as En Marche! in France also call for national cohesion and to “reconnect with the spirit of French conquest.”8 Theoretically, I argue that political corruption, crisis, and changing the status quo via tools of direct democracy jointly capture the essence of “vertical populism” (Hartleb 2015) that manifests in anti-establishment parties. A nation-first attitude, along with these three concepts, further adds nuance to anti-establishment parties’ messaging. Table 2.1 summarizes the operationalization of the concepts. I employ principal component analysis to create a new variable to capture anti-establishment sentiment, Anti-Establishment Sentiment. Principal component analysis is used to linearly trans- form a larger set of correlated variables into a smaller set of uncorrelated variables (Dunteman 1989). Analysis confirms that these concepts load onto a single component, therefore collectively representing the same latent concept. I then re-scale the variable so that it ranges between 1-10, where “10” represents the highest possible Anti-Establishment Score, indicating the most extreme levels of antipathy toward the mainstream, and “1” indicates little to no antipathy.9 Tables 2.2 and 2.3 summarize the correlations between these variables, and their weights according to principal components analysis, respectively. Principal component analysis displays the relative weights of

5From the 2017 manifesto, available from https://manifesto-project.wzb.eu/

6From page 27 of the 2017 manifesto, available from https://manifesto-project.wzb.eu/.

7From pages 168-170 of the 2016 manifesto, available from https://manifesto-project.wzb.eu/.

8From page 3 of the 2017 manifesto, available from https://manifesto-project.wzb.eu/.

9I also rescaled this variable for each country individually, as I argue that anti-establishment sentiment can also be thought of as a relative concept. For example, a party may be anti-establishment compared to parties within the country, but may not be particularly anti-establishment compared to parties from other countries

22 each variable, corresponding to each variable’s contribution to the component, and their individual role in explaining overall variance in the latent concept, represented by the component estimated (Dunteman 1989). Components with eigenvalues over “1” are candidates to be used in analysis. The eigenvalue presented below, 1.08505-indicates that the variable explains approximately 27.13% of the variance of the latent concept.10 The two variables that have the highest loadings are political corruption and general crisis. Principal component analysis on these two variables alone, generates an eigenvalue of 1.082, with each variable loading on the component at 0.701. While using this combination would explain roughly 50% of the variance in the latent concept, I believe it theoretically misses the nuances of messaging anti-establishment parties espouse. However, a graphic of party placement using only these two variables is listed in the Appendix.11 There are numerous quasi-statements on a variety of topics in the CMP dataset. Several other variables were also considered in the construction of the anti-establishment sentiment measure. Table 2.4 summarizes the various candidate variables and their factor loadings.

Table 2.1: Concepts and Operationalization

Concept Operationalization

Distrust of Political Elite Political Corruption

Crisis Status of the State General Crisis

Using Tools of Direct Democracy to Direct Democracy Depart from the Status Quo

Nation First Fabric of Society:Positive National Way of Life

10(100*1.080505/4), see Dunteman 1989.

11Additional iterations of the measure are also available in the Appendix. I have included a measure without the nationalism variable to demonstrate that the placement of these parties is stable even without accounting for nationalism, thus demonstrating that the measure is not simply a measure of nationalism.

23 Table 2.2: Correlation Among Anti-Establishment Variables

Political Corruption Direct Democracy General Crisis Nation First

Political Corruption 1.000

Direct Democracy 0.032 1.000

General Crisis 0.077 -0.007 1.000

Nation First 0.028 -0.014 0.007 1.000

Table 2.3: Anti-Establishment Sentiment Components

Component Loading

Political Corruption 0.71404

Direct Democracy 0.1755

General Crisis 0.6478

Nation First 0.2118

Eigenvalue 1.08505

Explained Variation 27.13%

Note: Component loadings represent each variable’s contribution to the component, and their individual role in explaining the latent concept. The eigenvalue represents the explained variation in the latent concept.

While many of these variables could adequately represent the concept of anti-establishment sentiment, I still maintain that political corruption, promoting direct democracy, alleging the country is in a state of crisis, and promoting a nation-ﬁrst attitude are the best empirical matches to my conceptual deﬁnition of anti-establishment sentiment.

24 Table 2.4: Alternative Anti-Establishment Sentiment Components

Variable Loading Loading Loading Component 1 Component 2 Component 3

Political Corruption 0.5358 0.4438 0.0737

Direct Democracy -0.1872 0.2539 0.7319

General Crisis 0.1420 0.4821 -0.2400

Nation First -0.3601 0.3650 -0.4528

Negativity About the Status Quo -0.0877 0.0774 0.4399

Promotion of Party’s Political Authority -0.0460 0.5918 0.0508

Critique of Government Eﬃciency 0.7198 -0.1295 0.0132

Eigenvalue 1.17924 1.1557 1.01964 Explained Variation 17% 17% 14%

2.2 Where are the Parties?

Figure 2.1 demonstrates the placement of parties cross-nationally according to their level of anti- establishment sentiment. It displays the parties according to their Anti-Establishment Sentiment on the y-axis, and their RILE (ideology) score on the x-axis using the manifesto from each country’s most recent election available from the CMP. I display mainstream and popularly-termed “anti- establishment” parties for the following countries: United Kingdom (2015), Italy (2013), Spain (2016), Germany (2017), France (2017), Netherlands (2012), and Austria (2013). 12 These countries are selected for display based on their popularity in the news, and recency of elections.13

12In the Appendix, I display the party positions for Greece. Broadly, Greek parties display higher levels of anti- establishment sentiment than other European parties. The political environment in Greece is highly eurosceptic, as the country has been the target of many economic reforms. Even mainstream parties in Greece score higher in anti-establishment sentiment then even anti-establishment parties in other countries, making Greece an outlier in anti-establishment sentiment.

13A list of recent anti-establishment (from the most current manifesto year available) for all European Union countries, and Norway, Switzerland, and Iceland, is available in the Appendix.

25 Figure 2.1: Cross-National Anti-Establishment Sentiment

Note: Parties placed according to the quasi-statements from the CMP pertaining to the four variables listed above.

Several observations are noteworthy. First, there is a degree of face validity for the measure. Parties that are popularly termed “anti-establishment” also appear to be higher in rhetoric expressing anti-establishment sentiment, as compared to mainstream parties in their own countries. For example, Alternative for Deutschland (AfD) is positioned higher than the German CDU, or the SPD. This is also true for the relative position of UKIP, compared to the British Labour and Conservative parties. The trend also holds for the National Front in France, Podemos in Spain, the Five-Star Movement (M5S) in Italy, and the Party for Freedom (PVV) in the Netherlands. Second, parties that have a higher score on the anti-establishment scale, can and do exist all along the x-axis (ideology). On the right, are parties such as UKIP, the National Front, and the

26 PVV; but, parties that express this antipathy toward the mainstream also exist on the left, such as the M5S and Podemos. Interestingly, the new French party, En Marche!, a self-proclaimed and objectively centrist party, scores higher in anti-establishment sentiment than many of its domestic counterparts, including the National Front. Third, parties that are popularly termed “anti-establishment” are not necessarily as high in their anti-elite rhetoric as compared to other parties that are considered ideologically moderate. The French case illustrates this point. The ideologically moderate En Marche! scores higher in anti-establishment rhetoric than even the National Front. This is not to say that En Marche! is more radical than the xenophobic party of Le Pen. Rather, this underscores the reality that radical right-wing parties are often termed “anti-establishment” simply because they occupy the far right section of the ideological spectrum. Parties are that are ideologically moderate, yet express hostility to status quo politics, are often ignored in favor of the more ﬂamboyant far right parties. Parties exist in all four-quadrants, there are establishment left and right parties, and anti-establishment left and right parties. While Figure 2.1 is informative of where parties currently reside on the two dimensions cross- nationally, further exploration is necessary to see how these parties rate according to the other parties in their countries, and how parties shift from one election to the next. Below, I present the current placement of parties within each country of interest, and how select parties in each country have shifted in their anti-establishment sentiment over time.

2.2.1 United Kingdom

Figure 2.2 demonstrates the placement of British parties according to their anti-establishment sentiment for the country’s most current manifesto year available, 2015. I rescaled the anti- establishment sentiment variable from “1” to “10” to be case-speciﬁc, and for the reader to be able to more clearly discern diﬀerences among domestic parties.

27 Figure 2.2: British Parties’ Anti-Establishment Sentiment 2015

Note: . Anti-Establishment score is scaled 1-10 for Britain in 2015.

As expected, UKIP scores higher in anti-establishment sentiment than most of its domestic counterparts. It should also be noted that the Democratic Unionist Party (DUP), a Northern Irish party, rates the highest in terms of anti-establishment sentiment. This is not necessarily surprising. However, the DUP most often made the news because of its noted right-wing orientation, not because of its anti-establishment rhetoric. This underscores the importance considering more than just a party’s ideology. Theresa May and the UK Conservative Party recently relied on the DUP for support in creating the government after the 2017 snap election. The Conservatives have made ties with not only a radical right-wing party, but also one that is decidedly anti-establishment. DUP’s position in supporting the May government could allow it the opportunity to influence policy. Understanding both its anti-establishment position along with its ideological position facilitates a more holistic understanding of this party’s behavior and potential influence. This may be especially true regarding the 2018 abortion issue in Ireland. Because the Republic of Ireland voted to repeal its ban on abortion, a similar call has been echoed in Northern Ireland. The DUP is staunchly anti-abortion, but because Northern Ireland currently lacks a regional government, the decision to ease abortion restrictions there will be left up to the government at Westminster. May relies on DUP support in parliament to govern, so the DUP, normally a political outlier, may be able to wield unprecedented influence on policy.

28 Figure 2.2 provides an initial picture of British parties’ anti-establishment position in a single year. The nature of my measure allows scholars to track anti-establishment over time. Figure 2.3 displays British parties’ anti-establishment over time.

Figure 2.3: British Parties’ Anti-Establishment Over Time

Note: Anti-Establishment Sentiment is scaled 1-10 for Britain over all manifesto years available.

Several interesting patterns should be noted. First, the Conservatives, Labour, and the Liberal Democrats all spiked in their anti-establishment sentiment from the late 1990s through the early 2000s. Around 2010, it appears as though these mainstream parties’ anti-establishment sentiment reacted to the introduction of UKIP, as their anti-establishment scores increased around this time. Second, while UKIP and the DUP remain the highest in anti-establishment sentiment, relative to other UK parties, their rhetoric has been on a decline since their initial manifesto years. The anticipation of Brexit could be an explanation for the changes in anti-establishment sentiment. The CMP does not yet have post-Brexit manifesto data for the UK. The ﬁnality of the UK’s exit from the European Union may be most salient for UKIP, whose main focus in the 2015 electoral campaign was leaving the EU. However, euroscepticism is not the only aspect of UKIP rhetoric. UKIP also expresses a desire to enact citizens’ initiatives every two years and transition to a proportional electoral system.14

14From page 53 of the 2015 manifesto, available from https://manifesto-project.wzb.eu/.

29 2.2.2 Italy

Italian politics are no stranger to volatility, and its parties present an interesting visual. Several new parties emerged before the 2013 elections (the most recent manifesto year). Most noteworthy is the formation of the Five-Star Movement (M5S), a left of center anti-establishment party. Figure 2.4 presents the positioning of Italian parties in 2013. Figure 2.5 displays how they have shifted over time. In 2013, the M5S rates particularly high in anti-establishment sentiment, as expected. The Civil Revolution Party (RC) rates the highest in anti-establishment sentiment. However, the M5S is in a greater position to influence Italian politics and policy in the long term. It holds influential local posts, such as the mayorship of Rome. More consequential though, is its ability to attain the most votes of any single party in the March 2018 Italian election, and create the governing “Italy First” Coalition along with Lega Nord. Also noteworthy in 2013, is the clustering of Italian parties in the middle of the anti-establishment spectrum. However, the Democratic Party (PD) remains the most mainstream of the major Italian political parties. Italian parties are volatile. Many survive a few election cycles, and either dissolve or join with other parties. I display PD, Lega, and Forza Italia because of their critical roles in the most recent 2018 election. The incumbent PD has been escalating its anti-establishment rhetoric since its earliest manifesto year, 2008. Lega Nord, led by Matteo Salvini is currently the M5S’s partner in the new Italian government, and is also in coalition with Forza Italia, led by the infamous Silvio Berlusconi. Both parties entered Italian politics higher in anti-establishment sentiment. This is expected since both were born out of the Mani Pulite (clean hands) scandal, where multiple Italian politicians, representing most mainstream parties, were indicted on corruption charges. However, when these parties were part of governing coalitions in the early 2000s, their anti-establishment sentiment moderated (Albertazzi and McDonnell 2008). On the eve of the current electoral cycle, Lega Nord has intensified this rhetoric, mostly likely due to the increases salience of immigration and the economic crisis. Beyond these issues, Lega also calls for the direct election of the president, reflecting a message advocating direct democracy.15

15From page 3 of the 2013 manifesto, available from https://manifesto-project.wzb.eu/.

30 Figure 2.4: Italian Parties’ Anti-Establishment Sentiment 2013

Note: Anti-Establishment Sentiment rescaled 1-10 speciﬁcally for Italy in 2013.

Figure 2.5: Italian Parties’ Anti-Establishment Sentiment Over Time

Note: Anti-Establishment Sentiment rescaled 1-10 speciﬁcally for Italian parties across all available manifesto years.

2.2.3 Spain

Like Italy, Spain has also seen the emergence of a left-wing anti-establishment party. Figures 2.6 and 2.7 display 2016 positions for Spanish parties, and their over time positions, respectively. With

31 its entrance into Spanish parliament in 2015, Podemos effectively altered Spain’s long standing two party system, and has risen in popularity to become the third largest party in parliament (Orriols and Cordero 2016, Ramiro and Gomez 2017). Interestingly the EHB, a Basque nationalist party, that is sometimes classified as left-wing, is positioned to the right, and is high in anti-establishment sentiment. This is potential evidence that the popular conversation about such parties is too focused on ideological position, and conflates anti-establishment sentiment with right-wing orientation. Over time, mainstream Spanish parties, Popular Party (PP) and the Spanish Socialist Workers party (PSOE) have remained fairly static in their anti-establishment sentiment. Podemos has as well, though it still remains slightly higher than the mainstream. Interestingly, EHB has made a significant jump in its anti-establishment sentiment from 2015 to 2016. The party critiques the parliament, specifically the PP, as being plagued with scandals and corruption. It also claims that Spain is going through both an economic and political crisis in an attempt to legitimize a potential EHB-led government.16 Given Spain’s recent issues with Catalonia, economics, and a recent no-confidence vote (the first time in Spanish history), anti-political elite rhetoric is unsurprising.

Figure 2.6: Spanish Parties’ Anti-Establishment Sentiment 2016

Note: Anti-establishment rescaled 1-10 for Spanish parties in 2016.

16From page 8 of the party’s 2016 manifesto, available from https://manifesto-project.wzb.eu/.

32 Figure 2.7: Spanish Parties’ Anti-Establishment Sentiment Over Time

Note: Anti-establishment score rescaled 1-10 for Spanish parties for all years in manifesto dataset.

2.2.4 Germany

Figure 2.8: German Parties’ Anti-Establishment Sentiment 2017

Note: Anti-establishment score rescaled 1-10 for German parties for 2017.

The most recent election in Germany saw a far right political party taking seats in German parliament since World War II. Alternative for Deutschland (AfD) has been characterized by its

33 xenophobic rhetoric that is hostile toward immigrants. The recent refugee crisis has further fueled the AfD’s nationalist rhetoric. In 2017, AfD attained the third largest number of seats in the Bundestag. Undoubtedly, the AfD is a party that builds its reputation around xenophobic, anti- Islam, anti-immigrant rhetoric. What gets ignored in the discussion about the AfD, is that they are the most extreme party with respect to their anti-establishment sentiment. While popularly characterized as an anti-establishment party, Figure 2.8 here demonstrates just how distant the party is in terms of their anti-establishment sentiment. The AfD was founded by disgruntled CDU members, who were frustrated with Chancellor Angela Merkel’s approach to the eurozone and refugee crises (Jankowski, Schneider, and Tepe 2017). While their xenophobic rhetoric (e.g. calling for higher birthrates among white Germans) is alarming, and contributes to its incredibly high anti-establishment score, its rhetoric about the political elite, crisis, and direct democracy also place it distant from other German parties. They look for radical departures to the status quo to help alleviate what they perceive as a political crisis. For example, on page seven of the party’s 2017 manifesto, AfD calls for the people to have the opportunity to own legislative initiatives themselves and to decide national policy via referenda.

Figure 2.9: German Parties’ Anti-Establishment Sentiment Over Time

Note: Anti-establishment score rescaled 1-10 for German parties for all years in manifesto dataset.

Figure 2.9 displays over time anti-establishment sentiment for the CDU, SPD, Die Linke, and

34 AfD. Over time, the CDU/CSU and the SPD - Germany’s largest mainstream parties - have remained fairly static (with the exception of a CDU/CSU spike in the late 1990s) in their anti- establishment sentiment. More interestingly, an increase in AfD’s anti-establishment sentiment from their first manifesto year (2013) to their most recent manifesto in 2017. This could be attributed to ever increasing anti-immigrant sentiment, and a desire to position itself even further from the mainstream. Another popularly identified extreme party in Germany is Die Linke (The Left) (Cofféand Plassa 2010). The party is a 2007 creation of the merger between the Party of Democratic So- cialism (PDS) and the Electoral Alternative for Labor and Justice (WASG), has also experienced an increase in its anti-establishment sentiment over time. The party claims to be the representative of East German identity (Doerschler 2015). Its manifesto does indeed reflect a distinctly anti-establishment sentiment. Mostly, the party advocates an anti-capitalist agenda, whereby they promote equality of wages and the reduction of poverty. They are also nationalist, but not in the way the AfD is. The party promotes the preservation of German culture, but with the expansion of refugee rights and the “critical examination of German 19th and 20th century histories.”17 They also want a stop to privatization, and demand that the decisions about privatization be decided by popular referenda.

2.2.5 France

Figures 2.10 and 2.11 display French parties positions in 2017, and over time, respectively. What is perhaps the most telling thing about the current French parties’ positioning is that Macron’s centrist and globalist party, En Marche!, is the highest scoring French party in anti-establishment sentiment. Also, the Socialists expressed a high degree of anti-establishment sentiment in their 2017 manifesto, reﬂective of the increasingly polarized nature of French politics (Evans and Ivaldi 2017). The National Front, perhaps the most noted “anti-establishment” party in France, rates lower in this sentiment than the Republicans, the Socialists, and En Marche!. This is not to indicate that the National Front is somehow benign. Its xenophobic rhetoric is disturbing and divisive. It simply illustrates that while the focus has been on the party’s radical

17From page 62 of the 2017 manifesto, available from https://manifesto-project.wzb.eu/.

35 rhetoric, we’ve ignored the emergence of anti-establishment sentiment from the center. Further, this evidence that scholarship has conﬂated “radical right” with “anti-establishment.” The National Front is decidedly on the far right of the political spectrum. Their rhetoric is alarming. That fact does not necessarily indicate that they are anti-establishment. En Marche! also calls for the restoration of French culture and pride, though it is not at the expense of non-white Frenchmen, or immigrants. The party’s insistence on increased citizen involvement speaks to its desire for the use of elements of direct democracy.

Figure 2.10: French Parties’ Anti-Establishment Sentiment 2017

Note: Anti-establishment score rescaled 1-10 for French parties for 2017.

Over time trends are of interest too. The Socialist Party appears volatile in its anti-establishment sentiment. The UMP (Republicans) have slowly ampliﬁed their anti-establishment rhetoric over time. The National Front, though initially high in their anti-establishment sentiment, appears to have tempered its rhetoric, focusing more on an anti-immigration and anti-EU message. While that may push the party further to the right, it does not necessarily indicated an increasing anti- establishment sentiment. These trends reﬂect the general volatility which the French system is apparently experiencing. The recent presidential election year, 2017, saw the unprecedented con- vergence of a centrist challenge and party polarization (Evans and Ivaldi 2018). The Socialist party

36 in particular has been in a state of ﬂux, as candidate fractionalization drove the party in divergent directions (Evans and Ivaldi 2017).

Figure 2.11: French Parties’ Anti-Establishment Sentiment Over Time

Note: Anti-establishment score rescaled 1-10 for French parties for all years in manifesto dataset.

2.2.6 Netherlands

Figure 2.12: Dutch Parties’ Anti-Establishment Sentiment 2012

Note: Anti-establishment score rescaled 1-10 for Dutch parties for 2012.

37 The commonly identiﬁed anti-establishment party in the Netherlands is the Party for Freedom (PVV), led by the ﬂamboyant and xenophobic Geert Wilders. Though their most recent election was in 2017, the CMP data ends at 2012. In 2012, the PVV scores the highest of all the parties in its anti-establishment rhetoric. It should be noted, however, that compared to other Dutch right of center parties, it isn’t placed as far to the right as one may expect. In fact, it is no further to the right (in terms of its RILE score) than the current governing party, the People’s Party for Freedom and Democracy (VVD). Its distinction from the mainstream rests primarily in its degree of anti-establishment sentiment, which has been heretofore singularly characterized as “radically right.”

Figure 2.13: Dutch Parties’ Anti-Establishment Sentiment Over Time

Note: Anti-establishment score rescaled 1-10 for Dutch parties for all years in manifesto dataset.

What distinguishes the party is its anti-mainstream rhetoric, such as its call for binding referenda, the abolition of the Senate, and its desire to reduce the number of members in parliament.18 Similarly, the 50PLUS party (a party that speciﬁcally advocates for the elderly), though centrist, also asserts that the mainstream has abandoned the people (in this case, the elderly). They also

18From page 29 of its 2012 manifesto.

38 advocate for a few large departures from the status quo, such as eliminating the provincial level of governance.19 As is typically seen in the other countries, the mainstream parties in the Netherlands, the VVD and the Christian Democratic Appeal (CDA), have been consistently low in their anti-establishment sentiment over time. The PVV entered into Dutch politics in 2006, well above its mainstream counterparts in its anti-establishment rhetoric. While the party has toned down this sentiment, it remains well above the mainstream parties. Its current support of the VVD in parliament, may account for this moderation in rhetoric

2.2.7 Austria

The Austrian data is from 2013, and therefore may not be as reﬂective of current trends in the country. The Freedom Party (FPO),¨ has been termed “anti-establishment” in the media. While assuredly a right-wing party, is actually low in terms of its anti-establishment sentiment.

Figure 2.14: Austrian Parties’ Anti-Establishment Sentiment 2013

Note: Anti-establishment score rescaled 1-10 for Austrian parties for 2013.

19from page 8 of the party’s 2012 manifesto.

39 Figure 2.15: Austrian Parties’ Anti-Establishment Sentiment Over Time

Note: Anti-establishment score rescaled 1-10 for Austrian parties for all years in manifesto dataset.

The most anti-establishment party of the year was Team Stronach, which has since been dis- solved. The Austrian Green Party is distant from the mainstream parties in its anti-establishment sentiment. The other decidedly anti-establishment party, NEOS, is a centrist party that is pro- European, but advocates for direct democracy in policy making. From Figure 2.15, we can see that FPO¨ entered the manifesto data with an extreme anti- establishment sentiment, but has since tempered, appearing more in line with mainstream parties. In their place, the Greens have appeared to overtake the FPO¨ as the most anti-establishment party in Austria. While Green parties are often considered anti-establishment in their early years in a country (Abedi 2002), they have tended to moderate this sentiment with increasing institutionalization (Abedi 2002). The Austrian case presents an interesting look at party transformation overtime. The FPO¨ has it roots in German fascism (Evans 2005). At its entrance into Austrian politics, it did indeed have to distinguish itself from the mainstream. Its earlier manifestos, from the 1990s, indicate a deep-seated distrust of the mainstream parties who have damaged Austrian pride and performance.20 However, the most recent Green Party manifesto opens with,“Corruption, frater-

20From the 1999 manifesto.

40 nity, and a lack of transparency have damaged conﬁdence in Austrian politics.”21 It further touts its role in uncovering massive corruption scandals from the other parties, FPO¨ included. The party’s desire to position itself as distant from other mainstream parties enhances its anti-establishment rhetoric. As the graphs illustrate, parties express varying degrees of anti-establishment sentiment, and can reside anywhere on the ideological spectrum. These numerous examples underscore the notion that was is anti-establishment is not necessarily radical right (e.g. M5S, Podemos, Austrian Greens, En Marche!), and parties that are radically right are not necessarily anti-establishment (e.g. FPO).¨ Parties can and do shift over time in this sentiment as well. Some parties such as Lega Nord, National Front, and FPO,¨ have tempered their anti-establishment rhetoric over time. These country assessments provide a degree of face validity. Below, I attempt to provide additional validity by comparing my measure to existing categorizations of parties.

2.3 Validity of the Measure

To my knowledge, only one other measure of anti-establishment sentiment exists. In their 2017 work, Polk et al (2017) use the 2014 Chapel Hill Expert Surveys (CHES) to assess the level of anti-elite salience in European parties’ rhetoric. They then validate their measure using CMP quasi- statements on corruption. Their measure is incredibly useful, and good ﬁrst step in recognizing the phenomenon of anti-establishment parties. However, the measure is only for 2014 year, and thus does not track how parties shift in this rhetoric over time. Because my measure incorporates manifesto data, it can update with each new election. Further, my measure permits analysis of parties’ past anti-establishment sentiment. Also, while the CHES measure does account for anti- elite sentiment on the right and the left, it does implicitly discount the prevalence of the movement from the center. Further, it does not account for inclusive and exclusive nationalism as a component part of anti-establishment sentiment. Below, I illustrate the correlation of my measure with the CHES measure. I compare their scores for 2014 year with my scores for the years surrounding 2014. The ﬁgure below demonstrates the correlation of the two measures.22

21From page 7 of the 2013 manifesto, available from https://manifesto-project.wzb.eu/.

22I include all countries the CHES assesses, even those not used for demonstration (e.g. Greece and Hungary) earlier in this chapter.

41 Figure 2.16: Anti-Establishment Sentiment v. CHES Measure

Note: Anti-Establishment sentiment is on the y-axis, compared to the CHES measure on the x-axis. Correlation of the two measures is 0.48, and is statistically signiﬁcant at the 0.05 level. Dashed line represents a perfect 45°angle, and a perfect correlation.

The measures are positively correlated at 0.48, and signiﬁcant at the 0.05 level. While a moderately strong correlation, it is not a perfect match. The CHES measure does not necessarily take nationalism and its ability to be either exclusive or inclusive into account when examining the placement of the parties. Some noteworthy deviations are that they place the PVV higher in anti-elite salience than I do in anti-establishment sentiment.23 To further explore similarities and diﬀerences between the two measures, I run a regression similar to one Polk et al. (2017) assess. I use their measure of corruption salience as the depen-

23The CHES measure does not take into account nationalism. I run additional analyses excluding nationalism from my measure (available in the Appendix). There still exists variation between my measure and the CHES measure, indicating that the inclusion of nationalism does not account for the variation between the two measures.

42 dent variable, and use both my measure and the CHES anti-elite salience measure as independent variables. I control for ideology as measured by Polk et al. (2017).24 Table 2.5 displays the results.

Table 2.5: Predicting Corruption Salience. Anti-Estabishment Sentiment and CHES Anti-Elite Salience Measure

(1) Corruption Salience

Anti-Establishment Sentiment 0.360*** (0.0863) CHES Anti-Elite Salience 0.334*** (0.0494) Ideology (GAL TAN) -0.189*** (0.0393) Constant 2.534*** (0.396) Country Fixed Eﬀects Y Observations 100 R-squared 0.850

Note: Robust standard errors clustered by party in parentheses. ** p<0.01, ** p<0.05, * p<0.1.

My measure and the CHES measure have roughly the same positive and statistically significant influence on predicting the level of corruption salience for a political party. While this does signal that my measure is valid, both measures can roughly tell us the same thing about corruption salience in 2014. I maintain though, that because my measure relies on CMP data, it can be updated as parties shift their rhetoric from one election to the next. The CHES measure can only tell us about party rhetoric in and around 2014. This misses how parties evolve in their anti- establishment sentiment and it also misses newer parties (i.e. En Marche!). Figure 2.25 compares the fitted values for both measures. 24The Polk et al (2017) paper used GAL-TAN as a measure of ideology, as opposed to RILE scores. Both are widely used in the displine, and vary slightly. For example, GAL-TAN places M5S slightly further toward the center of the ideological spectrum than RILE does.

43 (a) Original Anti-Establishment Score

(b) CHES Anti-Elite Salience Measure

Figure 2.17: Predicting Corruption Salience:Anti-Establishment Sentiment v. CHES Anti-Elite Salience

Note:From the regression presented in Table 2.5.

44 In addition to the Polk et al. (2017) classification, there exists a qualitative list by Abedi (2002). I run a probit model to assess how well my measure predicts a positive categorization of a party as an anti-establishment party according to Abedi’s critera. Table 2.6 presents the results. Using Abedi (2009), I code each “anti-political establishment party” listed in his appendix as a “1,” and establishment parties listed as a “0.” Any party listed in the manifesto data that was not listed in Abedi’s appendix, was not used in the analysis. Because Abedi’s assessment ends prior to 2009, many parties of interest are not available, such as the Italian Five-Star Movement, the German AfD, and France’s En Marche! Several countries are also not presented in the analysis, such as Spain, Greece, Portugal, Bulgaria, Croatia, the Czech Republic, Estonia, Latvia, Lithuania, Poland, and Hungary, as they are missing from Abedi’s categorization. A bivariate regression reveals that a one unit increase in a party’s anti-establishment score, increases the probability of being categorized as an anti-establishment party (according to Abedi’s classification scheme) by roughly 26%. Additionally, a correlation test reveals that there is indeed a positive correlation between my measure and Abedi’s categorization with a correlation coefficient of 0.28, statistically significant at the 0.05 level.

Table 2.6: Comparing the Measure to Abedi’s Classiﬁcation

(2) Probability of Being Categorized as Anti-Political Establishment (Abedi 2009)

Anti-Establishment Score 0.192*** (0.049) Ideology 0.012 (0.12) Constant -1.069*** (0.642)

Observations 615

Note: Robust standard errors clustered by party in parentheses. ** p<0.01, ** p<0.05, * p<0.1.

As an additional test of the measure’s validity, I run an analysis using statements on the European Union as a dependent variable. A disdain for European integration is a common trend among European anti-establishment parties (Hooghe and Wilson 2002). Parties from the right and

45 the left coalesce around the issue of euroscepticism in the European Parliament, making for some strange bedfellows, such the Italian Five-Star Movement and the United Kingdom Independence Party-who disagree on issues such as immigration, and the role of the state in economics, but share a similar dislike for integration policies. Eurosceptic rhetoric is not incorporated into the measure itself, and therefore presents a somewhat ideal test. I expect that parties high in anti-establishment rhetoric will also be high in eurosceptic rhetoric.25 The Table 2.7 displays results from a regression, where positive rhetoric about the EU is the dependent variable. It is a continuous measure where a higher number indicates a larger portion of the manifesto is dedicated to positive statements about the EU and its policies. The main independent variable is the anti-establishment score. I also control for ideology, and use country and year ﬁxed eﬀects.

Table 2.7: Predicting Sentiment Toward the EU

(3) Positive EU Statements

Anti-Establishment Sentiment -0.214*** (0.0722) Ideology 0.0132 (0.0222)

Constant 1.803*** (0.301)

Country Fixed Eﬀects Y Year Fixed Eﬀects Y Observations 1,543 R-squared 0.147

Note: Robust standard errors in parentheses. ** p<0.01, ** p<0.05, * p<0.1. Only countries that are members of the European Union used in the analysis.

An increase in a party’s anti-establishment score leads to a statistically significant decrease in its use of positive statements about the European Union. Ideology appears to not have a statistically significant effect on whether a party expresses negative rhetoric about the EU. While largely seen as a domain of radical right-wing parties, euroscepticism is more likely to be within the purview

25I do maintain that being eurosceptic is neither a necessary nor a suﬃcient condition for a party being classiﬁed as anti-establishment. As noted earlier, many parties such as Spain’s Podemos and France’s En Marche! are actually pro-European. However, most anti-establishment parties express a degree of Euroscepticism.

46 of most European anti-establishment parties, rather than the domain of a particular ideological group.

2.4 Conclusion

While the rise of the anti-establishment in Europe has garnered a lot of attention, the discussion of the movement has lacked conceptual clarity and a systematic measurement scheme. Only one measure scheme presented by Polk et al. (2017) has attempted such an endeavor. This chapter has attempted to provide a potential solution to both issues. By conceiving anti-establishment sentiment as a separate dimension from ideology, scholars can avoid conflating the anti-establishment movement with the radical-right. Examining manifesto data for key features of the movement: distrust of political elites who have created a national crisis, remedying said crisis by using direct democracy to depart from the status quo, and a desire to put the nation first, permits us greater understanding of the anti-establishment movement’s scope. Using manifesto data to quantify the level of anti-establishment sentiment allows scholars the opportunity to assess trends in parties’ rhetoric over time, and the ability to recognize the degree to which all parties exhibit this sentiment. The conflation of the anti-establishment movement with the radical-right and the vague of terming the movement as “populist” has served to mislead scholars in assessing what this movement actually is: an anti-political elite movement unbound by ideology. The measurement proposed in this paper takes an important step in properly recognizing anti-establishment parties, and fully assessing their behavior, the sources of their support, and their transformation over time. By divorcing the anti-establishment from ideology, scholars can better understand this new class of party. There are some short comings to the proposed measure. Primarily, manifesto data does not change frequently. While changes to manifestos can be informative of the trajectory of a party, changes are slow, often taking several years. Party officials use blogs, speeches, and social media to communicate with constituents. The incorporation of these informal sources may provide additional insights in future iterations of this measure. Additionally, this measure cannot give scholars a clear demarcation between establishment and anti-establishment. While I argue that this is, by and large, a strength of this conceptualization, the line between the mainstream and the outsiders remains fuzzy, and will primarily depend on country context.

47 This measurement provides additional usefulness beyond classification. Researchers should be able to leverage this measure to answer a variety of questions. Several questions emerge from looking at the country-level data. For example, scholarship could examine why parties shift their rhetoric. Future work could explore how the entry of an anti-establishment party impacts the rhetoric of mainstream parties. Additionally, research could analyze how winning seats could affect the anti-establishment sentiment of a party. In other words, as parties become institutionalized, will they still express the same degree of anti-establishment sentiment? Other areas of research may also provide fruitful insights. For example, how does the presence of anti-establishment parties affect political institutions? Are governments less stable when these parties hold portfolio assignments? Further, how do these parties behave once they are in office? These parties spend a great deal of time outside of politics “making noise” (Williams 2006). It stands to reason that this behavior may extend to inside parliament as well, where they may be more active in parliamentary questions, bill introduction, and speeches. Finally, scholars could examine the influence of these parties’ presence in parliaments on mainstream parties. Electoral successes of these parties may influence the level of anti-establishment rhetoric among more mainstream parties. If anti-establishment tactics prove to be successful, mainstream parties may attempt to replicate this rhetoric. The phenomenon of anti-establishment parties is one worthy of careful discussion. Parties from the left, right, and center are increasingly viable and are positioned to alter politics dramatically. Scholars should be mindful of this movement’s distinct qualities apart from ideology.

48 CHAPTER 3

LETTING THE OUTSIDERS IN: TRUST AND ELECTORAL SUPPORT OF THE ANTI-ESTABLISHMENT

European anti-establishment parties have become increasingly popular, often attaining the second or third most seats in their domestic parliaments. Parties such as the Party for Freedom (Netherlands), the Five-Star Movement (Italy), En Marche! (France), and, most recently, Alterna- tive for Deutschland (Germany), demonstrate that anti-establishment parties from the left, right, and center are evolving from a marginal electoral position (Mudde 2007) to a viable and potentially influential role in domestic politics. In light of these parties’ recent electoral successes, it is important to understand why voters support the anti-establishment. As elaborated in Chapters 1 and 2, much of the discussion about anti-establishment parties revolves around the radical right, and as such, theories about their electoral support are constructed with the radical-right in mind. These theories emphasize the role of cultural and economic fears, but are insufficient to explain support for anti-establishment parties broadly. I argue that voters who possess low levels of institutional trust, or diffuse support, will be most likely to support these parties who assert that the political elite have mishandled public resources. In the American context, we observe that low levels of trust can lead to the support of third parties (Hetherington 1999, Peterson and Wrighton 1998); however, this relationship has been little explored in Europe. Given the anti- elite nature of these parties, and their exploitation of individuals’ low level of institutional trust, examining trust can illuminate support for anti-establishment parties from the ideological left, right, and center. Often, scholars pay attention to classic “demand-side” and “supply-side” explanations, which argue that radical right wing parties’ xenophobic and populist rhetoric resonates with voters who possess deep economic and cultural fears, and that these parties manipulate advantageous attributes of the electoral system (Golder 2016, Bustikova 2014, Betz 1994, Mudde 2010, Inglehart 1977, Tajfel and Turner 1979). Though these explanations seem sufficient to account for the rise of non-

49 traditional parties (e.g. Golder 2016), they are constructed primarily with the radical right in mind, potentially ignoring anti-establishment parties from elsewhere on the political spectrum. Given the increasing viability of anti-establishment parties across Europe, it is important to understand why voters support the anti-establishment parties across Europe. Leveraging the 2014 Eurobarometer data, I explore in this chapter the phenomenon of anti- establishment parties in several contexts. First, I examine trust’s relationship to the selection of a domestic anti-establishment party through a cross-national analysis. I then explore this relationship in predicting votes for anti-establishment parties for the European Parliament. In both contexts, high levels of institutional trust reduces the probability of voting for an anti-establishment party. Then, using country-level election surveys, I explore how trust relates to support for three electorally successful anti-establishment parties: the right-wing Party for Freedom (PVV) in the Netherlands, the left-wing Five-Star Movement (M5S) in Italy, and the centrist Macron and En Marche! in France. In each case, an anti-establishment party rose to either a significant position in the domestic parliament, or in the case of France, the presidency. I then explore an election in which an anti-establishment party did not attain parliamentary seats: the AfD in the 2013 German general election. Using these elections, I address what characteristics, beyond the traditional demand-side explanations of economic and cultural fears, prompt voters to choose an anti-establishment party. I argue that mainstream political parties, both in government and in opposition, are often viewed as ineffective and detached from the electorate. In most cases, an individual’s level of institutional trust made the difference in selecting between an established opposition party and an anti-establishment alternative. Low institutional trust prompted many citizens to seek not just any alternative to the incumbent government, but a distinctly anti-establishment one. This relationship holds even when controlling for the usual explanations of radical right-wing support, such as ideology, economic class, attitudes on immigration, employment sector, etc. I find that trust in institutions influenced voters to support anti-establishment parties. Specifi- cally, low trust is connected with support of such parties from the right and left, but not necessarily centrist parties. Further, low trust influences individuals to support anti-establishment parties over established opposition parties. These findings contribute to our understanding of the wider phenomenon of anti-establishment parties, extending beyond traditional explanations of radical right- wing support. Further, the findings highlight the importance of considering the anti-establishment

50 as a unique form of opposition in an increasingly anti-establishment friendly Europe, and underscore the importance of institutional trust in electoral behavior.

3.1 Vote Choice for the Anti-Establishment

Why would a voter select an outsider party that espouses an agenda largely divorced from the status quo? Vote choice is the subject of a large literature. The famous “Michigan Model” espouses that vote choice is based on a long term attachment to partisanship (Campbell and Converse 1960). Economic voting models theorize that voters base their choices on perceived economic gains (Lewis-Beck and Nadeau 2000). I argue that to best understand voters’ motivations toward anti-establishment parties, one must examine institutional trust. Institutional trust can be defined in different ways. Here, I conceive of institutional trust as deep-seated trust in domestic institutions (Citrin and Luks 2001). Trust in institutions, or diffuse support, reflects the confidence in the overarching political system (Norris 2005, Easton 1965). I argue that it is the decline in diffuse support of institutions that influences individuals to select anti-establishment politicians and parties. Scholarship has noted that in a consolidated democracy, this loss of institutional trust may prompt voters to seek outsider, or populist alternatives (Canovan 1981, Hetherington 1999, Hartleb 2015). Further, low institutional trust has been linked to vote choice of third parties in the United States and Canada as a way to express displeasure with the status quo (Hetherington 1999, Peterson and Wrighton 1998, Bélangerand Nadeau 2005). As in Canada, European voters regularly have the option to select from more than two parties. This makes the choice of supporting opposition parties more complicated, as some opposition parties are part of the establishment. For example, the UK Labour Party currently serves as the establishment opposition party, counter to the Conservatives. I contend that low diffuse support, or institutional trust, motivates voters to support anti-establishment parties in a multi-party system in the same way it motivates voters to select third parties in the context of a two-party system such as the United States. Low trust in the system drives voters toward an anti-establishment alternative, and away from establishment parties, even those in opposition. Influential scholarship has demonstrated that trust is influenced by poor incumbent performance (Citrin 1974, Citrin and Green 1986). There is also ample evidence that the relationship between political discontent and votes for the anti-establishment can run in either direction (Rooduijn, Van

51 Der Brug, and De Lange 2016), or that these parties’ support is based on policy preferences (Van der Brug 2003). Certainly, the occurrence and magnitude of events, such as a political or economic crisis, may lead voters to blame incumbents for the poor handling of state interests (Duch and Stevenson 2008), and consequently reduce trust and diffuse support for institutions (Citrin 1974, Citrin and Green 1986, Svolik 2013). While the story of incumbent missteps affecting citizen trust is compelling, numerous work rigorously demonstrates that institutional trust influences vote choice, particularly for North American third parties (Hetherington 1998, Hetherington 1999, Peterson and Wrighton 1998, Bélangerand Nadeau 2005). Further, many studies contend that the decline of trust gives rise to protest parties (Gamson 1968, Betz and Immerfall 1998), indicating that trust is a precursor to the selection of anti-establishment parties. In fact, the emergence of new parties may represent a pre-existing condition of political distrust amongst the electorate (Miller and Listhaug 1990). Along this point, many scholars are in agreement that “trust encompasses deep-seated feelings” (Citrin and Luks 2001, 11) about the political system (Norris 1999, Inglehart 1995, Inglehart and Norris 2017, Abramson, Ellis, and Inglehart 1997). Further, scholars note that institutional trust is both stable over time (Peterson and Wrighton 1998), and is more strongly linked to diffuse support of the overarching political system and institutions than it is to policy attitudes or incumbent assessment (Bélangerand Nadeau 2005, Peterson and Wrighton 1998, Hetherington 1998, Miller and Listhaug 1990, Miller 1974). This link and the empirical observation by many scholars that trust is on a continuing downward spiral, indicate that institutional trust, or diffuse support, being influenced by incumbent evaluations is unlikely (Peterson and Wrighton 1998). In this vein, I argue that institutional trust is theoretically a precursor to an individual’s vote choice. Low levels of trust and a lack of diffuse support for the system influence whether a citizen selects a traditional, establishment party or an anti-establishment or “outsider” party. The selection of an anti-establishment alternative is a reflection of declining systemic trust possessed by some citizens. While it is certainly possible that government performance can influence trust. I argue that it is unlikely that the selection of one party or candidate over another, i.e. vote choice, impacts long term, deep-seated feelings about government institutions, or diffuse support. Extant literature largely focuses on voting for the radical right, implicitly equating radical- right wing parties to the anti-establishment movement. Research espouses that voters’ economic

52 or cultural fears influence the choice of a radical right-wing party (Bustikova 2014, Inglehart and Norris 2017, Duch and Stevenson 2008, Fitzgerald et al. 2012, Golder 2003). Scholarship argues that those citizens who are economically displaced by modernization (Inglehart and Norris 2017), economic downturn (Golder 2016, Jackman and Volpert 1996), immigrant labor (Knigge 1998, Ivarsflaten 2008) or trade agreements (low skill, low mobility) (Betz 1994, Lucassen and Lubbers 2012, Lubbers, Gijsberts, and Scheepers 2002), or those in areas otherwise affected by economic recession are more supportive of radical right-wing parties and policies that spout protectionism. Individuals who may have lost jobs under establishment parties’ leadership may seek out the radical parties or anti-establishment parties offering to remedy economic woes (Lewis-Beck and Mitchell 1993, Golder 2003). In Europe, two trends are noteworthy: the support of radical and anti- establishment parties as a backlash toward European integration (Hooghe and Wilson 2002), and the attitudes toward immigrants stemming from the recent refugee crisis (Fitzgerald et al. 2012). While these demand side explanations are empirically supported, they are constructed almost wholly with the radical right in mind, conflating radical right parties with all anti-establishment groups. Parties, such as the Italian Five Star Movement (M5S), who reside on the ideological left, but may draw previously inactive supporters from elsewhere on the ideological spectrum (Natale 2014, Franzosi, Marone, and Salvati 2015), are often neglected. The conflation of the radical right with the anti-establishment has also led scholars to ignore other potential explanations for their popularity, including the decline in institutional trust, assuming that xenophobic attitudes are the most consequential for parties on the right such as the Dutch Party for Freedom (PVV) (Rydgren 2008). It is the decline of diffuse support for a system believed to be irretrievably broken that drives support for outsider, or anti-establishment parties. As noted with radical right parties, outsider parties are able to amplify concerns of the public and portray them as widespread crises (Hartleb 2015, Williams 2006, Rooduijn, Van Der Brug, and De Lange 2016, Van der Brug 2003). When trust is low, support for existing political structures, politicians, and parties wanes. If individuals believe their political system is rife with corruption, then the most logical course of action is to seek out remedies outside of existing political institutions. Anti-establishment parties thrive in a low-trust environment. These parties often allege severe corruption on the part of the political elite (Bustikova 2014, Betz 1994, Abedi and Lundberg 2009,

53 Albertazzi and McDonnell 2008, Inglehart and Norris 2017, McDonnell and Newell 2011, Taggart 2004). Informal sources such as blogs, and statements to the press reveal this sentiment. From the left, Five-Star founder Beppe Grillo asserts, “political amateurs are conquering the world because the ‘experts’ destroyed it,” indicating widespread corruption and conspiracies by the political elite to swindle or cheat the everyday citizen.1 From the right, we hear similar allegations. Frauke Petry, former leader Alternative for Deutsch- land (AfD), once said, “The fact that the AfD exists is a symptom of the failure of the traditional parties, who have lost the ability to solve problems.”2 PVV leader, Geert Wilders often asserts that the Dutch government incompetently handles large problems, specifically immigration (Van Kessel 2011). Even centrists, like En Marche! allege that traditional leadership is “irre- sponsible to the people.”3 This sentiment, that the political elite has failed the common citizen, is expressed by anti-establishment parties from all areas of the political spectrum. This sentiment reflects a vertical populist notion (as identified in Chapters 1 and 2) that there exists a fundamental conflict between the “pure people,” and the “corrupt elite”(Mudde 2004, 543). Thus, anti-establishment parties occupy a unique place in domestic political systems. They represent not only an opposition to the governing incumbent parties, but also any mainstream party (Abedi 2009). Presenting themselves as a unique brand of opposition, they tout an agenda that is counter the status quo, asserting that traditional politicians have somehow violated the public trust, and therefore, all traditional establishment parties, even those in opposition, cannot be considered adequate replacements for the incumbent (Hartleb 2015). While it has been recently noted that these parties can emerge from both the left and right (Abedi 2009, Abedi and Lundberg 2009, Hartleb 2015, Mosca 2014, Golder 2016), a unified theory and analysis of voters’ propensity to anti-establishment parties from across the ideological spectrum has not been fully realized. Recognizing that these parties represent a unique form of opposition, I argue that low institutional trust will influence voters to seek electoral alternatives outside of mainstream parties. While

1In an interview with Euronews November 14, 2016. http://www.euronews.com/2016/11/14/political-amateurs- are-conquering-the-world-beppe-grillo-tells-euronews.

2See “Mea Culpa from Merkel on Migrants After Election Drubbing.” http://www.euronews.com/2016/09/19/germany-mea-culpa-from-merkel-on-migrants-after-election-drubbing. September 19, 2016.

3From page 4 of the party’s 2017 manifesto, available from https://manifesto-project.wzb.eu/.

54 we may accept that low trust may prompt a voter to choose a third party (Hetherington 1999), we do not know if trust will inﬂuence a voter to choose the anti-establishment option even over the challenger and the incumbent. Those voters who are higher in institutional trust will seek out an establishment alternative-whether it is an incumbent or an opposition party. When institutional trust is low, even the opposition will not provide suﬃcient amount of change for the anti-establishment voter, and they will then seek outsiders to govern (Canovan 1981). Because of their unique position in the political system, and their insistence that the current political elite are corrupt, anti-establishment parties’ messages resonate best with citizens low in institutional trust. When frustrated with the system, the rational course is to seek remedies outside the current political environment. Therefore, those with low levels of institutional trust will seek out anti-establishment alternatives to the mainstream and traditional status quo, including establishment opposition parties.

Hypothesis: Voters low in institutional trust are more likely to support an anti-establishment party than any establishment party, including the incumbent and opposition parties.

3.1.1 The European Elections

The 2014 European Parliament elections saw an unprecedented number of eurosceptic parties elected to the body. In the aftermath of both a serious ﬁnancial crisis, and the refugee crisis, eurosceptic parties from the left and right enjoyed newfound electoral successes. For example, the United Kingdom Independence Party (UKIP) was the top party from the UK elected to the European Parliament. European elections are interesting context. They often are cited as “second order” elections (Hix and Marsh 2007, Reif and Schmitt 1980, Van der Eijk, Franklin et al. 1996), where European voters view these elections as less important than domestic elections to parliament. This often results in lower voter turnout for the European contest (Hix and Marsh 2007). Many parties that won seats in the European Parliament, did not even exist at the time of the previous election in 2009. For example, Italy’s Five Star Movement was founded in 2012, and Greece’s Golden Dawn was not politically viable in Greek domestic politics until 2012 (Ellinas 2013). The proliferation of these parties into the European Parliament alarmed many mainstream

55 European politicians, who feared such parties would seek to obstruct the integration progress of the body.

3.1.2 From the Right and Left and Center: Anti-Establishment Parties in the Netherlands, Italy, France, and Germany

I analyze four countries with noteworthy anti-establishment movements. In three of these cases (Netherlands 2010, Italy 2013, and France 2017), an anti-establishment party enjoyed considerable success. In the fourth case, Germany 2013, two anti-establishment parties, representing both the left and the right, either lost seats in parliament (Die Linke), or failed to attain seats (AfD).4 I argue that regardless of the relative success of these parties, low trust should still drive voters toward an anti-establishment alternative. The “success” cases: the Dutch PVV in 2010, the Italian M5S in 2013, and the 2017 first round success of centrist En Marche! and right wing National Front, share certain interesting parallels. In all cases, an anti-establishment party rose to incredible electoral success. In the Netherlands and Italy, anti-establishment parties became the third largest party or group in the domestic parliament.5 In the French case, a brand new party from the center captured the presidency in competition with a xenophobic right wing National Front, shutting out the nation’s two major parties for the first time since the 1950s. The Dutch PVV, the Italian M5S, and the French National Front are headed by flamboyant leaders who regularly use the media to draw attention to their rhetoric. In the French case, Emmanuel Macron was a political insider who, heretofore, had not been publicly elected to office. In the Dutch and Italian cases, the establishment opposition party closest ideologically to the anti-establishment party won the election (The Dutch People’s Party for Freedom and Democracy (VVD), and the Italian Democratic Party (PD)). In France, the final contest was between two “anti-establishment” candidates: Macron, a centrist, and the radically right National Front’s Marine Le Pen. Even the “non-successful” anti-establishment case, Germany in 2013, demonstrates two anti-establishment parties of interest. Specifically, an institutionalized anti-establishment party from the left, Die Linke lost seats, while the right wing AfD gained steam, yet still failed to attain seats in the Bundestag. This highlights how in each country

4In the 2013 election, Die Linke lost seats, but still served as the third largest party in parliament. AfD fell just shy of the 5% threshold required to obtain seats in the Bundestag.

5In the case of the Italian Five-Star Movement, the party was actually the largest single winning party, but refused to join any coalitions. Therefore, it is the third largest parliamentary group.

56 voters had a choice on two dimensions: ideology; and antipathy toward elites. All countries were facing integration questions, economic crisis, and immigration challenges.

3.1.3 The Netherlands 2010

The Netherlands is no stranger to xenophobic and anti-elite voting (Rydgren 2008, Akkerman 2005). So-called populist parties, such as List Pym Fortuyn (LPF), have enjoyed a certain amount of popularity, but have rarely been electorally successful (Van Kessel 2011). In the past decade, Dutch voters have been cited as being “increasingly detached” from the establishment parties (Van Kessel 2011, 69). The latest party to capture the detached Dutch voter is the Party for Freedom (PVV), led by Geert Wilders. Under Wilders’ leadership, the PVV has been able to ﬁll a void left by the collapse of the LPF (Vossen 2010), which catered to those voters who were anti-immigrant, anti-foreign, and perceived the mainstream to hold “anti-Dutch” values (Akkerman 2005). Dutch voters who feel that mainstream political establishment parties are incapable of handling the current problems facing the country, in this case immigration, are drawn to the PVV (Van Kessel 2011, Vossen 2010). These voters follow a pattern of desiring harsher immigration restrictions and tougher punishments for crime in general (Akkerman 2005, Vossen 2010). In this vein, Wilders has been a champion for these voters. Not only has he been sanctioned for anti-Islam rhetoric, notably making a blatantly anti-Islam movie in 2008, but he was also banned from the United Kingdom in 2009 due to his hostile and anti-Islam language (Vossen 2010). Citizens in the Netherlands most likely to support the PVV and other “outsider” parties have been found to possess high levels of protest attitudes, distinguishing them from mainstream voters (Schumacher and Rooduijn 2013). The Netherlands in 2010 provided fertile ground for the rise of the Wilders and the PVV. While the party won a handful of seats in 2006, they were largely inconsequential. However, as the refugee crisis deepened in severity, the PVV, with its largely xenophobic agenda, was able to gain ground. Successfully portraying itself as the party fully capable to remedy the perceived incompetencies of the mainstream (i.e. increases in immigration from outside the EU), the PVV propelled into third place, gaining ﬁfteen seats, and backing a minority government led by the People’s Party for Freedom and Democracy (VVD).

57 3.1.4 Italy 2013

Like the Netherlands, the Italian political environment in 2013 was primed for an anti-establishment success. Italy is no stranger to volatility and scandal. Notably, nationwide judicial investigations in the early 1990s ultimately led to multiple politicians’ resignations (Vannucci 2009a) and the rise of the ﬁrst “outsider” Silvio Berlusconi, and his new party Forza Italia (FI) in 1994. Berlusconi’s downfall in 2011 prompted the birth of a new anti-elite party. In 2009, Beppe Grillo, a stand-up comedian, and Gianroberto Casaleggio, a internet strategist, formed the Movimento Cinque Stelle, the “Five Star Movement” (M5S). The movement rejected traditional politics and attempted to defend the protection of public amenities and speak for the mass public, who, it argued, were abandoned by the corrupt elite (Diamanti 2014, Mosca 2014, Natale 2014). After great success in the 2012 regional and local elections, the new party entered parliament in 2013 with the third highest number of seats. Because the party refused to form coalitions with any of the establishment parties, it served as the largest opposition party in the Italian Chamber of Deputies. In 2018, the party attained the larges vote share of any single party, placing it in a governing coalition with the xenophobic Lega Nord. Italy’s long history with political scandal makes the rise of a party, such as the Five-Star Movement, unsurprising. The culmination of scandals, and economic and political crises served to erode public trust in institutions. The erosion of public trust led Italian voters to seek out an anti-establishment alternative from the status quo.

3.1.5 France 2017

The French 2017 Presidential election represented a departure from status quo politics. The result meant that the two major parties: the Socialist and the Republicans (UMP), were shut out of the presidency for the ﬁrst time since the 1950s. Also, for the ﬁrst time in modern French politics, three unique events converged: a strong centrist challenger, fragmentation within parties, and party polarization (Evans and Ivaldi 2018). These forces served to both embolden the National Front, a xenophobic mainstay of French politics since the 1970s, but also gave legitimacy to Emmanuel Macron’s new party, En Marche! (Evans and Ivaldi 2017). While this particular election was distinct from previous French elections, it still maintained some features of past contests. For example, centrists in France regularly depend on attractive candidates who lack the “usual badges

58 of oﬃce” (Lewis-Beck and Nadeau 2015, 27). This is certainly true for Emmanuel Macron, who depsite serving as Francois Hollande’s Economy Minister, successfully painted himself as a political newcomer. The other major force present in this election was the infamous Marine Le Pen and the National Front. Le Pen is credited with increasing her party’s viability by ﬁrst, rescinding her father’s membership;6 and, second, shifting the party’s rhetoric to a more populist discourse (Stockemer and Barisione 2017). Marine Le Pen termed herself a “crisis manager” and portrayed her restoration of French glory as “My Project” (Stockemer and Barisione 2017, 104). Her anti-immigrant sentiment was strengthened by the political context, which had been largely dominated by terrorism since the Bataclan Theater attack in 2015 (Faucher and Garcia 2016).

3.1.6 Germany 2013

Anti-immigrant sentiment was also strong in the context of the 2013 German Bundestag election. Perhaps more salient was the aftermath of the Eurozone crisis and the role Germany played in ensuring economic security across the continent. Angela Merkel’s view of an increasingly united Europe caused tension within her own party, resulting in the founding of the AfD by disgruntled CDU members (Jankowski, Schneider, and Tepe 2017). Indeed, previous studies of AfD voting behavior have cited AfD supporters as being eurosceptic, whereby they believe Germany is too involved in the EU, and fear that Germany will be the caretaker of the Union’s poorer states (Wagner, Lewandowsky, and Giebler 2015). Though radical right parties have been largely irrelevant to German politics (Decker 2008), AfD has recently changed this trend. Its success in 2017 was preceded by gains in the 2013 election, and later in the 2014 European Parliament election, by emphasizing a crisis message (Jankowski, Schneider, and Tepe 2017). While the party did not win seats in 2013, it did manage to win 4.7% of the party list vote, just shy of the 5% threshold required to be seated. Their success can be attributed to their positioning both to the right and above of the CDU/CSU in ideology and authoritarianism (Franzmann 2016). AfD was not the only anti-establishment contender in 2013. Die Linke, a left wing party that aims to represent East German interests (Coﬀ´eand Plassa 2010, Doerschler 2015) also competed. Though the party managed to garner sixty-four seats in the

6Jean Marie Le Pen is a noted Holocaust denier. Marine Le Pen removed her father from the party in order to salvage the party’s image. See Stockemer et al 2017 for further detail.

59 Bundestag, and was the third largest party, they actually lost twelve seats from 2009 to 2013. The party itself claims that German government forgets about the common people, and advocates for increases in citizen participation as a check on suspect politicians.

3.2 Data and Methods

Succinctly, I argue that low institutional trust will increase support for anti-establishment parties. Further, voters with high levels of institutional trust will seek out establishment opposition parties. I use the 2014 Eurobarometer, the 2010 Dutch Parliamentary Election Study, the 2013 Italian National Election Study (ITANES), the 2017 French National Election Study, and the 2013 German National Election Study to assess the influence of trust on vote choice and support for anti-establishment parties. As discussed in Chapters 1 and 2, anti-establishment groups express a fundamental distrust of political elites, asserting charges of corruption, calling for changes to the status quo via tools of direct democracy, and expressing a nation-first attitude, either asking to place native citizens first in the priorities of elected officials, or reasserting the prominence of the nation in the international community. In this regard, center parties, such as En Marche! and left-wing parties such as the Five Star Movement7 have much in common with some right wing parties such as the PVV and the United Kingdom Independence Party (UKIP). 8

3.2.1 Dependent Variables

The dependent variable in all contexts is vote choice. For the cross-national analysis, and for the individual country-level analyses, I use probit models to assess the probability of a citizen voting for an anti-establishment party. The dependent variables are dichotomous measures where “1” represents a vote for the anti-establishment party, and “0” indicates a vote for any other party. While this provides an initial glance at how likely a voter is to choose an anti-establishment alternative over any other party, the more serious question is when will a voter choose the anti-establishment

7There is some debate as to whether the Five Star Movement constitutes a true “left” party. Indeed, recent developments with the party indicated that it may indeed blur the lines between the left and the right. However, at the time of the 2013 election, the party was decidedly left-leaning. In fact, its RILE score was -49.03 (from the Comparative Manifesto Project), indicating it is a left-wing party. The Chapel Hill Expert Survey (CHES) also identiﬁes it as a center left party. See Chapters 1 and 2.

8In fact, the Five Star Movement often allies itself with UKIP in the European Parliament.

60 over an establishment alternative, such as the incumbent or the opposition. Therefore, I use multinomial logit models to test the likelihood of a voter selecting the anti-establishment party relative to the other establishment alternatives: the incumbent and the opposition.

The Netherlands. In the Netherlands, this is fairly straight forward. The incumbent government was run by a coalition headed by Christian Democratic Appeal (CDA)-and elections were called after its resignation. The Labour Party (PvdA) governed in coalition with the CDA. The main establishment challenger was the People’s Party for Freedom and Democracy (VVD), a centre- right party, and the main anti-establishment challenger was the PVV-situated on the far right. I code the dependent variable for the Dutch multinomial analysis as follows: “1” for the incumbent CDA, “2” for the PvdA, “3” for the VVD, and “4” for the PVV.

Italy. Italian parties, like those in many other multi-party systems, align themselves electorally with other parties (Bartolini, Chiaramonte, and D’Alimonte 2004, Golder 2006). Italy had three major establishment coalitions running in 2013. The ﬁrst was the incumbent Centre Coalition, led by Civic Choice (SC) and joined by Union of the Democratic Center (UdC), and the party for the Future and Freedom of Italy (FLI). Also, two opposition coalitions formed. The “Left Alliance” was led by the Democratic Party (PD) and joined by the Left Ecology and Freedom Party (SEL), the Democratic Centre (DC), the Crocetta List, and the South Tyrolean People’s Party (SVP). Finally, the “Right Coalition” was led by the People of Freedom (PdL) and joined by the Brothers of Italy, Lega Nord, La Destra (the Right), the Pensioners Party, the Great South (MPA), and Moderates in Revolution (MIR). Therefore, the dependent variable for the multinomial analysis is a four-value variable taking the value of “1” if the respondent voted for the incumbent coalition, “2” if she voted for the Left Alliance, “3” if she voted for the Right Coalition, and “4” if she voted for the M5S.9

France. In France, I examine the 2017 presidential election. France holds two rounds of elections, where the top two candidates in the ﬁrst round progress to the second round. As in the Netherlands and Italy, I also probit models to assess the probability of an individual voting for an

9Lega Nord, a xenophobic and radical right party, is also considered anti-establishment (Abedi 2009)(See also Chapter 2). In this election they ran in a coalition with establishment opposition parties. Ideally, I would be able to run an analysis predicting the votes for Lega Nord. However, in the survey used, only 22 respondents reported voting for Lega Nord. To assuage any concerns that voting for the Right Alliance may indicate an anti-establishment preference, I run the Italian multinomial analysis dropping the 22 Lega Nord voters, available in the Appendix.

61 anti-establishment party versus all other parties. In the ﬁrst round, I identify En Marche! and the National Front as the two anti-establishment parties. I run two probit models to assess the probability of voting for these parties. To assess the likelihood of selecting an anti-establishment party over other parties in the ﬁrst round, I run a multinomial logit where the candidate from the incumbent party, the Socialist Party, is the base category. The second round contest was between Emmanuel Macron and Marine Le Pen. I use a probit model, where “1” indicates a vote for Macron, and “0” indicates a vote for Le Pen.

Germany. Germany uses both party lists and constituency level districts to elect its parliament. Like France, I also identify two anti-establishment parties. AfD, and Die Linke. I run a total of four probit models, two for each party (party list and constituency level), to assess the probability of selecting the AfD or Die Linke over all other parties. I then run two multinomial logit models (party list and constituency level), using the incumbent party, CDU/CSU as the reference category.

3.2.2 Independent Variables

The main explanatory variable in the analysis is institutional trust. Because I am interested in diﬀuse support, I look to capture how much trust the citizen has in domestic political institutions (Norris 2005). In the Eurobarometer, I use “trust in national parliament” as a measure of institutional trust. I also use “trust in the European Union.” Each is coded in the trusting direction, using a four point scale, where “4” indicated the highest level of expressed trust. The country-level studies have considerable variation in questions available and appropriate to measure institutional trust. Below I discuss how I constructed the variable for each of the country studies.

Netherlands. For the Dutch survey, I construct a variable institutional trust using principal component analysis (PCA). Three variables are used to construct the measure: how much trust does the individual have in parties, how much the individual trusts the parliament, and how much the respondent trusts the government. Each variable positively loads onto the latent concept with an eigenvalue of 2.23, meaning that together the component explains 74% of the variation in the latent concept, which I identify as trust.10 The ﬁnal variable, is comprised of adding the variables

10Component loadings are as follows: trust in parties: 0.5990, trust in parliament: 0.5336, trust in government: 0.5970.

62 together, weighted by their respective loadings. It is coded in the trusting direction, and scaled to range from 1-10.

Italy. In Italy, variable construction is similar. Three questions are used to construct the variable: how much trust the respondent has in parliament, how much trust she has in parties, and how much she has in the president. Once again, each variable positively loads onto a single component, with an eigenvalue of 1.77, explaining approximately 59% of the variance in the latent concept.11 As with the Netherlands, the trust variable is constructed by adding together the weighted variables, and rescaling the variable to range between 1-10.

France. The trust variable in France is created using four questions: How much the respondent believes politicians care about common people, how trustworthy politicians are, whether or not the respondent believes politicians are the main problem in France, and if the respondent agrees that politicians only care about the rich. These are admittedly diﬀerent from what is used in the Dutch and Italian cases, however, these variables do positively load onto the same component, with an eigenvalue of 2.27, collectively explaining 57% of the variance in the latent concept.12 As with the previous two cases, I added together the weighted variables, and rescaled the variable to range between 1-10.

Germany. In the German survey, there are no overt trust questions. Instead, I opt to use satisfaction with democracy as an alternative measure for diﬀuse support. This is consistent with past research that also uses satisfaction with democracy as an indicator of how an individual evaluates the regime in practice (Linde and Ekman 2003, Armingeon and Guthmann 2014). The variable ranges 1-4, with 4 indicating the highest level of satisfaction with democracy. Previous studies have linked poor democratic satisfaction with voting for the AfD (see Jankowski et al 2017).

3.2.3 Control Variables

While the main focus of the analysis is diffuse support (institutional trust), I control for how specific support may influence vote choice. Therefore, I look to capture how positively the re-

11Component loadings are as follows: trust in parliament: 0.6292, trust in parties: 0.6050, trust in the president: 0.4879.

12Component loadings are as follows-politicians don’t care: 0.4935, politicians are trustworthy: 0.4527, politicians are the main problem: 0.5104, politicians care about the rich: 0.5395.

63 spondent evaluates the incumbent government. For the Eurobarometer analyses, respondents were asked about their approval of their national government. The variable is binary, where “1” indicates approval of the national government. In the Netherlands, I use government satisfaction to assess the level of specific support. Respondents were asked to rate their level of satisfaction with the government on a 5 point scale from “Very Dissatisfied” to ”Very Satisfied.” In the Italian case respondents are asked how positively they rate the incumbent Prime Minister’s (in this case, Mario Monti) economic policies. Respondents were asked to rate Monti’s policies on a scale from 0-10, where “0” is the most negative evaluation and “10” is the most positive. In France, respondents were asked about their level of satisfaction with the Hollande government, which ranges 1-4, where “4” indicates a high level of satisfaction. In Germany, respondents are asked to evaluate both the incumbent government and the opposition. In both cases, the variable ranges 1-11, where “11” indicates the most positive evaluation. These variables ask the respondent to reflect on the current government (i.e. the prime minister and his/her cabinet), I maintain that they capture the level of specific support the respondent has for the current government and its policies. I also control for a number of potential confounding variables. For all country-level analyses, I control for respondents’ gender (female) (Betz 1994, Lubbers, Gijsberts, and Scheepers 2002), education (possession of a college degree in Italy, and levels of education in the Netherlands, France, and Germany), employment status: “1” for employed, “0,” otherwise (Arzheimer and Carter 2006), marital status (“1” for married, “0,” otherwise). For the Netherlands, France, and Germany I also control for income. In the Netherlands and Italy, I control for age and economic class (whether the respondent is employed as an unskilled worker in Italy, or if one considers herself “working class” in Netherlands) (Golder 2016). In Italy, France, and Germany I control for frequency of religious attendance, and whether or not the respondent is Catholic (Lewis-Beck and Nadeau 2015). In France and Germany, I am also able to include personal views of the national economy and personal finances. In addition to demographic variables, I control for respondents’ ideology. A person may vote for a particular party simply because it most closely matches her political preferences. For all analyses, the variable ideology is a self-rating 10 point scale, where “10” is the most conservative value. I also control for a person’s level of political interest, as those who are more interested in politics are more likely to seek out information. In all cases, the anti-establishment garners a lot

64 of media coverage, whether it’s Geert Wilders’ explosive statements on Islam, Grillo’s antics (such as swimming the Strait of Messina for the 2012 regional elections), interviews with Marine Le Pen, or even the prominence of German politics in European integration. This may mean that those tuning-in to the coverage of the election were influenced to support an anti-establishment party, simply by the volume of attention devoted to these candidates. Finally, I control for a few policy attitude variables. Immigration is a salient topic in Europe and has been used to explain radical right-wing support (Ivarsflaten 2008, Golder 2016). It is especially imperative that I control for immigration attitudes given the provocative rhetoric from parties on the right. For example, Geert Wilders and the PVV regarding to Islam and immigrants. In all analyses higher values of the variable Immigration Opinion reflect a more positive attitude toward immigration/immigrants. In the French analyses, I use four questions about immigration: whether the respondent thinks immigrants are good for the economy, benefit French culture, negatively impact crime, or take away benefits from citizens. Finally, many “anti-establishment” parties in Europe are eurosceptic, and this is certainly the case with the PVV, the M5S, the National Front, and the AfD (Vossen 2010, Franzosi, Marone, and Salvati 2015). Though I maintain in earlier chapters that euroscepticism is neither a necessary, nor a sufficient condition to be considered anti- establishment. In country-level analyses, Support for the EU is a variable that measures how positively the respondent feels about the European Union (EU), with higher values corresponding to more positive views of the EU. In Germany, respondents were asked whether they blamed the EU for the economic crisis. I use this variable to tap into EU attitudes, where “1” indicates that an individual levies blame on the European Union for economic woes.

3.3 Results 3.3.1 Cross-National Results

Table 3.1 Presents the results from Model 1, assessing how trust relates to voting for an anti- establishment party for national parliament, as measured in the 2014 Eurobarometer survey.

65 Table 3.1: Probit Model: Voting for an Anti-Establishment Party for National Parliament

(1) Eﬀect on Probability of Voting for an AE Party: Domestic Level

Trust in National Parliament -0.105*** (0.0247) Trust in EU -0.109*** (0.0250) Ideology 0.0875*** (0.00655) Immigration -0.0341*** (0.00607) Attachment to Europe 0.00971 (0.0435) Approval of National Government -0.379*** (0.0423) Positive View of the National Economy -0.0423 (0.0277) Positive View of Personal Finances -0.0129 (0.0295) Interest in Politics 0.0488** (0.0210) Religious Attendance -0.0348*** (0.00966) Catholic -0.146*** (0.0551) Western Europe -0.00815 (0.220) Rural -0.114*** (0.0388) Unemployed -0.0230 (0.0349) Female -0.101*** (0.0356) Married -0.102*** (0.0355) Voted for the European Parliament -0.0208 (0.0418) Constant -1.538*** (0.205) Country Fixed Eﬀects Y Observations 11,862

Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Country Fixed Eﬀects. From the 2014 Eurobarometer.

Trust in the national parliament has a negative and statistically signiﬁcant inﬂuence on the probability of voting for an anti-establishment party on the domestic level. This is also true for an

66 individual’s trust in the European Union, and their approval of the national government. Opinion on immigration also influences support for an anti-establishment party. The more supportive a respondent is of increasing immigration, the less likely you are to vote for an anti-establishment party on the domestic level. Ideology is also statistically significant, the further to the right an individuals’ orientation is, the more likely she is to vote for an anti-establishment party on the domestic level. Another variable that has significant influence on anti-establishment voting is interest in politics, those who are more interested are more likely to vote for an anti-establishment party. However, economic explanations do not seem to matter. How the respondent views the national economy, or her personal finances does not appear to influence anti-establishment support. This is also true for employment status. Figure 3.1 displays the predicted probabilities for voting for an anti-establishment party on the domestic level across different levels of trust.

Figure 3.1: Probability of Voting for an AE Party: Domestic Level

Note: Predicted probabilities of voting for an anti-establishment party for national parliament with 95% conﬁdence intervals. All other variables held at their mean values.

67 At the lowest levels of expressed trust in the national parliament, a respondent has a roughly 14% probability of voting for an anti-establishment party. At the highest levels of trust, the respondent has a 9% probability of voting for an anti-establishment party. While trust does appear to influence an individual’s probability of voting for an anti-establishment party on the domestic level, how does it influence an individual’s choice in supporting an anti- establishment party over the incumbent? Table 3.3 presents the multinomial logit results, where I assess how trust influences the choice between different types of parties: the incumbent, establishment opposition, and the anti-establishment. The table compares three contests: establishment incumbent versus an establishment opposition party, anti-establishment party versus an establishment opposition party, and an anti-establishment party versus the incumbent. To fully capture a voter’s decision calculus, Table 3.2 presents Model 2, the multinomial logit. Columns 1 and 3 display the results using the domestic incumbent party as the base category. Col- umn 2 presents the results using the establishment opposition party as the base category. Column 1 should be interpreted as the likelihood of selecting the establishment opposition party relative to the incumbent. Column 2 should be interpreted as selecting an anti-establishment party relative to the establishment opposition, and Column 3 is the likelihood of selecting an anti-establishment party relative to the incumbent. Trust has a negative and statistically significant influence on choosing any opposition over the incumbent. Those who are more trusting of their national parliaments are less likely to choose both an establishment opposition party or an anti-establishment party over the incumbent party. This is not particularly surprising. However, the more interesting result is from Column 2. Trust makes an individual statistically significantly less likely to select the anti-establishment party over the establishment opposition. As trust goes up, an individual is 14% less likely to vote for the anti-establishment relative to the establishment opposition. Some of the control variables do present interesting results. First, those who are more approving of their national government are more likely to select the establishment opposition over the incumbent, but they are less likely to select to an anti-establishment party over the opposition. This may suggest that the fundamental difference between diffuse and specific support is playing out in vote choice for anti-establishment parties. Further, just as in the probit model, a positive view of personal finances has no influence on the choice between the anti-establishment party and either

68 the incumbent or the establishment opposition. Once again, interest in politics plays a positive and statistically signiﬁcant role in the selection of an anti-establishment party relative to both the incumbent and the establishment opposition.

Table 3.2: Model 2, Multinomial Results: Voting for an AE Party on the Domestic Level

(2) Establishment Opposition AE Party AE Party v. the Incumbent v. Opposition v. the Incumbent Trust in National Parliament -0.149*** -0.153*** -0.302*** (0.0297) (0.0510) (0.0542) Trust in EU -0.0849*** -0.148*** -0.233*** (0.0302) (0.0511) (0.0545) Ideology -0.0282*** 0.178*** 0.150*** (0.00818) (0.0146) (0.0155) Immigration 0.0374*** -0.0853*** -0.0479*** (0.00719) (0.0126) (0.0135) Attachment to Europe -0.00949 0.00521 -0.00428 (0.0559) (0.0900) (0.0972) Approval of National Government 0.258*** -0.931*** -0.672*** (0.0502) (0.0935) (0.0992) Positive View of National Economy 0.124*** -0.124*** -0.0899 (0.0313) (0.0584) (0.0618) Positive View of Personal Finances -0.0701* -0.0143 -0.0844 (0.0368) (0.0605) (0.0652) Interest in Politics -0.0118 0.136*** 0.124*** (0.0251) (0.0433) (0.0460) Religious Attendance -0.00749 -0.0546** -0.0621*** (0.0112) (0.0217) (0.0229) Catholic -0.255*** -0.0611 -0.317** (0.0633) (0.121) (0.128) Western Europe -0.295* 17.9 17.61 (0.160) (2,595) (2,595) Rural -0.0421 -0.112 -0.154* (0.0451) (0.0826) (0.0869) Unemployed -0.0827** 0.0340 -0.0486 (0.0421) (0.0738) (0.0784) Female -0.00727 -.214*** -0.221*** (0.0426) (0.0749) (0.0796) Married -0.0747* -0.148** -0.223*** (0.0431) (0.0743) (0.0792) Voted in the European Elections 0.0385 0.0200 0.0585 (0.0495) (0.0891) (0.0942) Constant 1.870*** -20.83 -18.96 (0.181) (2,595) (2,595) Country Fixed Eﬀects Y Y Y Observations 13,895 13,895 13,895

Note: Robust Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1. County Fixed Eﬀect. From the 2014 Eurobarometer. For columns 1 and 3, the base category is the incumbent party in each country, for column 2, the base category is the establishment opposition in the country.

69 Cross-nationally, it does appear that trust plays a role in determining vote choice for the anti- establishment. Broadly, those who possess higher levels of institutional trust are less likely to cast a vote for an anti-establishment party. Many anti-establishment parties have not enjoyed a great deal of electoral success on the domestic level, but they have made strides in the European Parliament, particularly in the 2014 European election. Table 3.3 displays the results from the probit model (Model 3) assessing voting for an anti-establishment party on the European level. As expected, trust has a negative and statistically significant influence on voting for an anti- establishment party for European Parliament. This is also true of how much the respondent trusts the EU itself. As is the case on the domestic level, views about the national economy or personal finances do appear to influence an individual’s vote for an anti-establishment party, nor does employment status. Being interested in politics, does make one more likely to vote for an anti- establishment party for the European Parliament, as it does on domestic vote choice. This could indicate that those who are interested in politics are exposing themselves to more information. Anti-establishment parties are fond of using the media to their benefit. This is true in Italy, where the M5S heavily relies on social media to relay their rhetoric. This may also indicate that those individuals who are eurosceptic, are keenly interested and aware of how sending an anti-establishment party to the European Parliament may obstruct the business of that institution. Further analysis is needed to tease out the relationship between interest in politics and voting for anti-establishment parties. Figure 3.2 displays the predicted probabilities of voting for an anti-establishment party for the European Parliament, based on trust in national parliaments and trust in the European Union. Many of the usual explanations for voting for the radical right-wing do hold for anti-establishment vote choice in the European Parliament. Unsurprisingly, attitudes toward immigration influence vote choice for anti-establishment parties. However, economic concerns do not seem to influence vote choice for the anti-establishment. This underscores the notion that preferences for voting for the right-wing are not one and the same with preferences for the anti-establishment.

70 Table 3.3: Probability of Voting for an Anti-Establishment Party: European Parliament

(3) Eﬀect of Probability of Voting for an AE Party: European Parliament

Trust in National Parliament -0.0862*** (0.0264) Trust in EU -0.234*** (0.0266) Ideology 0.0571*** (0.00678) Immigration -0.0699*** (0.00639) Attachment to Europe -0.0962** (0.0473) Approval of National Government -0.177*** (0.0438) Positive View of the National Economy 0.00282 (0.0289) Positive View of Personal Finances 0.0138 (0.0338) Interest in Politics 0.0540** (0.0226) Religious Attendance -0.0263*** (0.0101) Catholic -0.164*** (0.0613) Western Europe -1.077*** (0.124) Rural -0.0553 (0.0404) Unemployed 0.00448 (0.0368) Female -0.128*** (0.0375) Married -0.0452 (0.0375) Constant 1.026*** (0.162) Country Fixed Eﬀects Y Observations 9,574

Note: Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1. Country Fixed Eﬀects. From the 2014 Eurobarometer.

71 Figure 3.2: Probability of Voting for an AE Party: European Parliament

Note: Predicted probabilities of voting for an anti-establishment party for European Parliament, based on trust in national parliaments and trust in the EU, with 95% conﬁdence intervals. All other variables held at their mean values.

As trust in the national parliament increases, the probability of voting for an anti-establishment party for European Parliament decreases. This is a modest decrease, however. At the lowest level of trust, an individual has a roughly 15% probability of voting for the anti-establishment, and a 12% probability at the highest levels of trust. Trust in the European Union has a more dramatic impact on an individual’s probability of voting for an anti-establishment party. At the lowest level of trust, a voter has a 22% probability of voting for an anti-establishment party to the European Parliament, but at the highest level of trust, this drops to less than 10%.

3.3.2 Country-Level Results

The cross-national analysis from the Eurobarometer is a somewhat helpful staring point in assessing the relationship between trust and voting for anti-establishment parties. However, a

72 more helpful data source for assessing how trust relates to votes on the domestic level would be a domestic level survey. Below is an analysis of the Dutch 2010, Italian 2013, the French 2017 elections, and the 2013 German elections.

The Netherlands. Table 3.4 displays probit results, assessing the probability of voting for the PVV, a right-wing anti-establishment party. Figure 3.3 presents predicted probabilities.

Table 3.4: Main Results-Netherlands

(4) Probability of Voting for the Party for Freedom (PVV)

Institutional Trust -0.149*** (0.0376) Satisfaction with Incumbent Government -0.242*** (0.0653) Age -0.0185*** (0.00397) Female -0.203** (0.102) Married 0.104 (0.116) Education -0.257*** (0.0493) Employed -0.0761 (0.143) Immigration -0.232*** (0.0718) Support for the EU -0.301*** (0.0735) Interest -0.110 (0.101) Income -0.0168* (0.00957) Working Class 0.268** (0.115) Ideology 0.148*** (0.0234) Constant 2.749*** (0.454)

Observations 1,640

Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1, two-tailed

73 Figure 3.3: Predicted Probability of Voting for the PVV: Probit Model

Note: Predicted probabilities with 95% conﬁdence intervals. All other variables held at their mean values.

Institutional trust has a statistically significant and negative influence on voting for the anti- establishment. In the Netherlands, voters expressing the lowest levels of institutional trust had a 20% probability of voting for the PVV over all other parties. When moving from the lowest level of trust to the highest, this probability drops to nearly roughly 3%. This result holds even when controlling for the usual explanations of voting for the radical right, such as immigration opinion, opinion on the EU, and being a working class voter. Intriguingly, is that counter to the results from the 2014 Eurobarometer, interest in politics appears to play no role in 2010 Dutch voting for the PVV. It should be noted that most of the usual suspects in predicting right-wing voting, such as education, working class, immigration attitudes, are also significantly connected to voting for the PVV, a right-wing anti-establishment party. I will note though, that trust is still significant, even in the presence of these variables.

74 While the probit models are informative of an individual’s probability of voting for a particular party (PVV) versus all other parties, it does not reveal much about the actual decision calculus a voter makes in casting her ballot. When voting, an individual chooses between the incumbent and multiple challengers. The Multinomial model captures this decision calculus. Because every election can be viewed as a referendum on the incumbent, I use the incumbent party (the Dutch CDA) as the reference category. This captures the likelihood of voting for all opposition parties (establishment and anti-establishment) over the incumbent. Table 3.5 displays the results of the multinomial analysis. In the Dutch case, trust does not play a significant role in predicting votes for the establishment opposition parties over the incumbent. However, it does play a statistically significant role in influencing the selection of an anti-establishment party, in this case the PVV, relative to the incumbent. Exponentiating the coefficient reveals that with an increase in each level of trust, a voter is roughly 20% less likely to vote for the PVV relative to the incumbent, the CDA. This may indicate that diffuse support only plays a role when deciding between anti-establishment and mainstream parties (government or opposition), and that specific support (i.e. satisfaction with government) influences an individual’s choice between mainstream parties. While it may be easy to see how trust may play a role in vote choice between an anti- establishment party and the incumbent, how does it influence vote choice between an anti-establishment party and an establishment opposition party? Table 3.6 displays the results of a probit model where “1” represents a vote for PVV, and “0” represents a vote for the VVD, the right of center establishment opposition party in 2010. Figure 3.4 displays the predicted probabilities. Trust plays a negative and statistically significant role in supporting the PVV (anti-establishment) over the VVD (establishment opposition). At the lowest level of trust, an individual has 56% probability of voting for the PVV, relative to the VVD. This drops to 11% at the highest level of trust. Ideology plays no significant role, as both parties are situated right of center. Further, employment status and immigration opinion are also not statistically significant, indicating that political trust may matter more in the choice between an anti-establishment part and an establishment opposition party of similar ideological leaning. Ideology may sort an individual to the left or right, but trust may sort her to the mainstream or the anti-establishment.

75 Table 3.5: Model 5, Netherlands-Multinomial Results

(5) Voting for the Voting for the Voting for the PvdA v. Incumbent VVD v. Incumbent PVV v. Incumbent

Institutional Trust 0.0195 0.0679 -0.280*** (0.0993) (0.0795) (0.0988) Satisfaction with Government -0.985*** -0.973*** -1.251*** (0.167) (0.144) (0.182) Age -0.0237** -0.0162** -0.0492*** (0.0100) (0.00798) (0.0103) Female -0.996*** -0.548*** -1.008*** (0.253) (0.202) (0.263) Married -0.202 -0.302 -0.130 (0.251) (0.212) (0.286) Education -0.184 0.375*** -0.404*** (0.127) (0.113) (0.132) Employed -0.0769 0.306 -0.174 (0.343) (0.279) (0.371) Immigration 0.208 -0.268** -0.545*** (0.160) (0.135) (0.181) Support for the EU 0.150 0.00255 -0.491** (0.215) (0.170) (0.202) Interest 0.390 0.0578 -0.0767 (0.250) (0.204) (0.261) Income 0.0294 0.0201 -0.0225 (0.0225) (0.0190) (0.0241) Working Class 0.225 -0.324 0.463 (0.303) (0.275) (0.305) Ideology -0.887*** 0.164*** -0.0339 (0.0759) (0.0581) (0.0788) Constant 8.256*** 2.045* 12.34*** (1.353) (1.226) (1.436)

Observations 1,060 1,060 1,060

Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Multinomial results of voting for the three opposition parties over the incumbent, CDA (voting for the incumbent is the base category). Column 3 captures the relative probability of voting for the PVV (anti-establishment) over the incumbent.

Other control variables are of interest. Specifically, income and working-class status influence support for the PVV over the VVD. Further, low support for the EU influences voters to support the PVV over the VVD. This is unsurprising, given the PVV’s very public and notable Eurosceptic position. Finally, level of education plays a statistically significant role in predicting support for the PVV over the VVD. Those who are less educated are more likely to vote for the PVV over the VVD. This should also be unsurprising, as education relates to work sector.

76 Table 3.6: Probability of Voting for the PVV over the VVD

(6) Eﬀect on Probability of Selecting the PVV over the VVD

Institutional Trust -0.204*** (0.0541) Satisfaction with Incumbent Government -0.188** (0.0907) Age -0.0169*** (0.00584) Female -0.245* (0.142) Married 0.0694 (0.160) Education -0.444*** (0.0752) Employed -0.263 (0.187) Immigration -0.126 (0.0900) Support for the EU -0.344*** (0.0986) Interest -0.152 (0.138) Income -0.0218* (0.0123) Working Class 0.456*** (0.163) Ideology -0.0467 (0.0417) Constant 5.619*** (0.756)

Observations 545

Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1, two-tailed

77 Figure 3.4: Predicted Probability of Voting for the PVV over the VVD

Note: Predicted probabilities with 95% conﬁdence intervals. All other variables held at their mean values.

Italy. Table 3.7 displays the ﬁrst results from Italy. Model 7 assess the probability of voting for the M5S, over all other parties. Similar patterns for anti-establishment support are present in Italy. At the lowest level of trust, there is a nearly 30% probability that a respondent will vote for the M5S. With an increase of each level of trust, the probability of voting for the M5S decreases by approximately 10% points. At the highest level of trust, a respondent has a 3% probability of voting for the M5S. Figure 3.5 displays the predicted probabilities.

78 Table 3.7: Italian Probit Model: Predicting Support for the Five Star Movement

(7) Probability of Voting for the Five Star Movement

Institutional Trust -0.150*** (0.0316) Evaluation of Incumbent -0.0331 (0.0234) College 0.164 (0.111) Married 0.140 (0.118) Catholic 0.181 (0.174) Employed -0.117 (0.103) Unskilled Worker 0.437** (0.218) Ideology -0.0700*** (0.0199) Religious Services Attendance -0.0111 (0.0213) Immigration Opinion 0.0189 (0.0322) Interest 0.114* (0.0625) Support for the EU -0.0875 (0.0666) Female 0.0339 (0.100) Age -0.0174*** (0.00346) Constant 0.549* (0.320)

Observations 1,083

Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1, two tailed

In Italy, it is especially important to note that some classic explanations for supporting radical right-wing parties, such as demographics, employment status, level of education, and opinion of immigration did not have a statistically significant influence on the the probability of voting for the M5S over all other parties. This is initial evidence that other factors, such as institutional trust, may matter more for selecting anti-establishment parties, and this may particularly hold for those parties on the left. Ideology is significant, with right wing identification decreasing the probability of voting for the M5S. Here, interest in politics is positively associated with voting for the M5S,

79 Figure 3.5: Predicted Probability of Voting for the M5S: Probit Model

Note: Predicted probabilities with 95% confidence intervals. All other variables held at their mean values. consistent with Eurobarometer results. I posit that this is largely due to Beppe Grillo’s flamboyant antics, such as swimming the Strait of Messina, and holding public rallies. Table 3.8 presents the multinomial results from Italy (Model 8). As in the case of the Nether- lands, I used the incumbent party at the time (Civic Choice) as the base category. Institutional trust does not appear to influence a voter’s support of the establishment right- wing opposition (Right Coalition) relative to the incumbent, nor does it impact selecting the Left Alliance, another establishment opposition group, relative to the incumbent. However, as predicted, trust does decrease a respondent’s likelihood of choosing the Five-Star Movement relative to the incumbent. Specifically, an increase in the level of trust corresponds to 12% decrease in the probability of voting for the M5S relative to the incumbent. Institutional trust serves to decrease a voter’s likelihood of voting for the anti-establishment party. Unsurprisingly, evaluations of the incumbent are significant in explaining votes against the incumbent. I maintain that this is further

80 evidence that there is a difference in diffuse and specific support’s influence over vote choice for the anti-establishment. These results hold even when controlling for the usual demand-side explanations of radical right-wing support, such as immigration opinion, employment status and sector, and education.

Table 3.8: Italy-Multinomial Results

(8) Voting for the Voting for the Voting for the Left v. Incumbent Right v. Incumbent M5S v. Incumbent

Institutional Trust 0.130 -0.0883 -0.325*** (0.104) (0.111) (0.108) Evaluations of the Incumbent -0.411*** -0.527*** -0.580*** (0.0856) (0.0926) (0.0898) College -0.588 -0.665* -0.354 (0.361) (0.401) (0.371) Married -0.482 -0.458 -0.292 (0.393) (0.423) (0.410) Catholic -1.025 -0.961 -0.956 (0.896) (0.987) (0.909) Employed -0.197 -0.207 -0.443 (0.353) (0.397) (0.365) Unskilled Worker 0.626 0.657 1.296 (1.154) (1.168) (1.125) Ideology -0.858*** 0.520*** -0.368*** (0.0979) (0.102) (0.0936) Religious Attendance 0.00684 -0.0914 -0.0822 (0.0800) (0.0893) (0.0824) Immigration Opinion 0.133 -0.0809 0.0736 (0.109) (0.126) (0.115) Interest 0.266 -0.0858 0.362 (0.222) (0.238) (0.228) EU Opinion 0.0742 -0.107 -0.0793 (0.239) (0.247) (0.241) Female -0.288 -0.406 -0.226 (0.335) (0.370) (0.351) Age -0.000106 0.0104 -0.0317** (0.0122) (0.0128) (0.0126) Constant 7.151*** 2.052 9.126*** (1.482) (1.640) (1.515)

Observations 759 759 759

Note: Robust standard errors in parentheses. ** p<0.01, ** p<0.05, * p<0.1. Multinomial results of voting for the three opposition parties over the incumbent, Centre Coalition (voting for the incumbent is the base category). Column 3 captures the relative probability of voting for the M5S (anti-establishment) over the incumbent.

81 Table 3.9: Probability of Voting for the M5S over the Left Alliance

(9) Probability of Voting for M5S over Left Alliance

Trust -0.265*** (0.0465) Evaluation-Incumbent Government -0.114*** (0.0335) College 0.185 (0.153) Married 0.203 (0.152) Catholic -0.0133 (0.227) Employed -0.198 (0.142) Unskilled Worker 0.424 (0.322) Ideology 0.265*** (0.0379) Religious Services Attendance -0.0394 (0.0293) Immigration Opinion -0.0186 (0.0445) Interest in Politics 0.0195 (0.0860) Support for EU -0.0500 (0.0895) Female 0.0132 (0.138) Age -0.0204*** (0.00473) Constant 1.287*** (0.457)

Observations 522

Note: Robust standard errors in parentheses. ** p<0.01, ** p<0.05, * p<0.1, two tailed. Probit model assess the probability of voting for the M5S over the Left Alliance. Coeﬃcients are not directly interpretable.

82 Figure 3.6: Predicted Probability of Voting for the M5S over the Left Alliance

Note: Predicted probabilities with 95% conﬁdence intervals. All other variables held at their means.

Model 9, displayed in Table 3.9, assesses the probability of voting for the M5S relative to the Left Alliance, the establishment opposition closest to the M5S ideologically, where “1” represents a vote for the M5S, and “0” indicates a vote for the Left Alliance. Once again, institutional trust has a negative and statistically significant influence on voting for the anti-establishment party. Figure 3.6 displays the predicted probabilities. At the lowest level of trust, an individual has a 63% probability of selecting the M5S over the Left Alliance, whereas she only has a 6% probability of doing so at the highest levels of trust. Ideology is also statistically significant, where the further to the right an individual identifies herself, the more likely she is to vote for the M5S as opposed to the Left Alliance. This may be due to the more established nature of the member parties of the Left Alliance, and the relatively new nature of the M5S. This could be further explored, as the ideologically unconstrained nature of the M5S could spell interesting consequences for policy and voting in Italy.

83 France. The French analysis departs from the analyses of the Netherlands and Italy, due to the fact the French survey pertains to the presidential election. Table 3.10 displays results for two different probit models, assessing the probability of voting for Macron or Le Pen in the first round. Trust does not appear to play the same role in France as it did in the Netherlands and Italy. In the first round of the presidential election, institutional trust increased the probability of voting for Emmanuel Macron from 10% at the lowest level of trust, to 40% at the highest level of trust. Despite my argument that he represents a different type of anti-establishment candidate (see Chapter 2), trust appears to play an enhancing role in this context. Further, while being trusting of political institutions does decrease the probability of voting for Marine Le Pen (this is expected), it is not a statistically significant relationship. Other results are of interest. Ideology is positive and statistically significant in predicting votes only Le Pen, indicating that the the further to the right a person self-identifies, the more likely she was to vote for Le Pen. While unsurprising for Le Pen, Macron marketed himself as a centrist, and was a member of a Socialist government under Hollande. We do know that French parties in 2017 were extremely polarized (Evans and Ivaldi 2018). The results here may be a consequence of left wing voters moving toward the Socialist Party. Also potentially surprising is how immigration perceptions influenced votes in the first round. Believing that immigrants are good for the French economy negatively and statistically significantly reduced the probability of voting for Macron in the first round, as it does for Le Pen. This was the only immigration attitude that statistically significantly influenced votes for Macron in the first round. The only other immigration attitude that statistically significantly influenced votes for Le Pen was believing that immigrants are good for French culture, agreement with which negatively influence voting for Le Pen. The evidence that economic concerns about immigration influenced votes for Macron, but cultural concerns did not, is intriguing. It appears that both economic and cultural fears about immigration played a role in selecting Marine Le Pen and the National Front. This may suggest that there is indeed a difference within the anti-establishment movement between exclusive and inclusive nationalism. The former only being a component part of the right- wing anti-establishment. Finally, support for the EU was statistically significant in predicting votes for both candidates, with those who expressed support for the organization more likely to vote for

84 Macron, while those who are less supportive more likely to vote for Le Pen. This is not particularly surprising, given Le Pen’s intense eurosceptic rhetoric.

Table 3.10: Probability of Voting for Macron or Le Pen in the 1st Round

(10) (11) Macron Le Pen

Institutional Trust 0.115*** -0.0217 (0.0352) (0.0428) Female 0.127 0.0985 (0.117) (0.129) Education 0.0826 -0.0556 (0.0509) (0.0532) Income 0.0300 -0.0523 (0.0323) (0.0348) Positive View of the National Economy 0.0500 -0.153** (0.0633) (0.0769) Ideology 0.0424 0.160*** (0.0289) (0.0347) Personal Financial Strain -0.0810 0.195 (0.126) (0.137) Satisfaction with Hollande Government 0.337*** -0.0921 (0.0835) (0.104) Immigrants are good for the economy -0.109* -0.138** (0.0630) (0.0658) Immigrants are good for French culture 0.0823 -0.171** (0.0648) (0.0765) Immigrants do not negatively impact crime 0.00421 -0.0612 (0.0714) (0.0729) Immigrants do not take away beneﬁts 0.0350 -0.0574 (0.0611) (0.0768) Support for the EU 0.364*** -0.329*** (0.112) (0.0935) Interest in Politics 0.102 0.0752 (0.0692) (0.0705) Catholic 0.217* 0.176 (0.122) (0.142) Religious Services Attendance -0.0973 -0.0675 (0.0666) (0.0831) Married 0.303** -0.209 (0.127) (0.145) Constant -4.165*** 0.665* (0.466) (0.401)

Observations 820 820

Note: Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1.

85 (a) Macron (b) Le Pen

Figure 3.7: France 2017. Predicted Probabilities of Voting for Macron or Le Pen in the First Round

Note:Predicted probabilities of voting for Macron/Le Pen in the First Round of the Presidential Election with 95% conﬁdence intervals. Control variables held at their mean values.

To further explore the breakdown of votes in the first round, I run a multinomial logit to assess the calculation in selecting one candidate over another, just as I did with the Eurobarometer, Netherlands, and Italy. Table 3.11 displays the multinomial results, where the base category is voting for BenoˆıtHamon, the Socialist Party Candidate. Institutional trust decreased the probability of voting for any other candidate over Hamon. However, this was only statistically significant for those choosing to vote for Melanchon. Once again, believing that immigrants are good for the economy negatively influenced votes for Le Pen and Macron. The final contest was between Macron and Le Pen. Table 3.12 displays results for the probit model assessing votes in the second round in the French presidential election, where “1” indicates voting for the Macron, and “0” indicates a vote for Le Pen. Institutional trust is positively and statistically significantly connected to voting for Macron over Le Pen in the second round. Further, believing that immigrants are good for the economy positively and statistically significantly relates to voting for Macron, while the other immigration attitudes do not appear to have an influence. Unsurprisingly, conservative ideology is connected to votes for Le Pen.

86 Table 3.11: Multinomial Results: Votes in the First Round

(12) Melanchon Fillon Le Pen Macron v. Hamon v. Hamon v. Hamon v. Hamon

Institutional Trust -0.281** 0.160 -0.0605 0.122 (0.120) (0.137) (0.147) (0.116) Female -1.634*** -1.177** -0.932* -0.873** (0.410) (0.466) (0.477) (0.403) Education 0.0863 -0.100 -0.144 0.0384 (0.170) (0.197) (0.195) (0.172) Income 0.0892 0.174 -0.00943 0.118 (0.108) (0.125) (0.124) (0.108) Positive View of National Economy -0.231 -0.209 -0.253 -0.0703 (0.230) (0.255) (0.267) (0.226) Ideology -0.150 1.191*** 0.956*** 0.576*** (0.128) (0.163) (0.156) (0.134) Personal Financial Strain 0.554 -0.521 0.356 -0.0275 (0.418) (0.498) (0.489) (0.426) Satisfaction with Hollande Government 0.192 -0.283 -0.295 0.391 (0.300) (0.353) (0.357) (0.302) Immigrants are good for the economy -0.340 -0.559** -0.797*** -0.624*** (0.235) (0.265) (0.261) (0.241) Immigrants are good for French culture -0.0726 0.0511 -0.377 0.0316 (0.240) (0.270) (0.270) (0.241) Immigrants don’t negatively impact crime 0.0851 -0.279 -0.223 -0.153 (0.269) (0.289) (0.294) (0.265) Immigrants are not taking away beneﬁts 0.264 -0.208 -0.108 0.0420 (0.190) (0.230) (0.239) (0.190) Support for the EU -1.103*** -0.441 -1.498*** -0.371 (0.414) (0.495) (0.463) (0.455) Interest in Politics -0.316 0.136 0.0251 0.189 (0.251) (0.287) (0.290) (0.251) Catholic -0.175 0.823* 0.470 0.389 (0.391) (0.485) (0.483) (0.388) Religious Services Attendance 0.211 0.662** 0.140 0.174 (0.240) (0.262) (0.285) (0.237) Married 0.550 0.0129 -0.379 0.631 (0.402) (0.480) (0.481) (0.401) Constant 6.970*** -2.856 6.157*** -0.669 (1.683) (2.068) (1.936) (1.767)

Observations 629 629 629 629

Note: Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. The base category is voting for Hamon, the Socialist Candidate.

87 Table 3.12: Probability of Voting for Macron or Le Pen in the Second Round

(13) Eﬀect on Probability of Voting for Macron v. Le Pen

Institutional Trust 0.127** (0.0519) Female 0.166 (0.154) Education -0.0314 (0.0630) Income 0.0750* (0.0436) Positive View of the National Economy 0.0807 (0.0842) Ideology -0.125*** (0.0406) Personal Financial Strain -0.368** (0.156) Satisfaction with Hollande Government 0.251** (0.114) Immigrants are good for the economy 0.239*** (0.0753) Immigrants are good for French culture 0.102 (0.0872) Immigrants do not negatively impact crime 0.108 (0.0854) Immigrants do not take our beneﬁts 0.0905 (0.0790) Support for EU 0.627*** (0.120) Interest 0.0159 (0.0866) Catholic -0.327* (0.167) Religious Services Attendance 0.210** (0.0871) Married 0.415** (0.177) Constant -3.530*** (0.555)

Observations 569

Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1.

88 Figure 3.8: France 2017. Predicted Probabilities of Voting for Macron v. Le Pen in the Second Round

Note: Predicted probabilities of voting for Macron over Le Pen in the Second Round of the Presidential Election with 95% conﬁdence intervals. Control variables held at their mean values.

Figure 3.8 displays the predicted probabilities for trust. Going from the lowest to the highest level of political trust, results in a 30% increase in the probability of voting for Macron v. Le Pen. It should be noted, however, that even at the lowest level of trust, an individual has a 60% probability of voting for Macron over Le Pen. This speaks to the general trend in French politics, where the National Front may actually perform well in the ﬁrst round of presidential and legislative elections, but often under performs in the second round of voting due to the strategic behavior on the part of mainstream parties, and the generally unconstrained nature of the National Front’s voters (Mayer and Perrineau 1992).

Germany 2013. I have demonstrated how the relationship between trust and vote choice operates in the context of elections where anti-establishment parties won seats in parliament (the Netherlands and Italy), or attained executive oﬃce (France). Below, I present results from Germany in 2013, where AfD (on the right) and Die Linke (on the left) did not perform as well. The most recent German election (2017) saw an incredible rise in the popularity of the AfD, but the party

89 narrowly failed to attain seats in 2013, and Die Linke lost seats in 2013. How did trust influence vote choice in this context? Table 3.13 displays results for probit models, assessing the probability that a person would vote for the AfD both in party list voting, and in constituency level voting. Figures 3.9 and 3.10 display the predicted probabilities for voting for AfD and Die Linke, respectively. As satisfaction with democracy increases, the probability of voting for either the AfD or Die Linke decreases. This is the same effect in both party list and constituency seats. For AfD, the probability of voting for the party decreases from 12% to close to 0% in party list voting, and goes from 7% to close to 0% in constituency level voting. The influence of democratic satisfaction is roughly the same for voting for Die Linke. In party list voting, the probability of voting for Die Linke goes from roughly 16% to 8%, and in constituency level voting, it decreases from 16% to approximately 3%. Interestingly, only in party list voting for AfD did immigration opinion have a negative and statistically significant influence. Higher income also negatively and statistically significantly influenced votes for Die Linke. In all models, ideology is statistically significant, with right-wing ideology predicting votes for the AfD, and left-wing ideology influencing votes for Die Linke. As in the other analyses, I include a multinomial regression, where the base category is voting for the incumbent (CDU/CSU), comparing how satisfaction with democracy influences vote choice. Tables 3.13 and 3.14 display these results. In the multimomial model assessing party list voting, satisfaction with democracy negatively and statistically significantly influences selecting either anti-establishment party (AfD/Die Linke), relative to the incumbent. Satisfaction with democracy decreases the likelihood of voting for the AfD relative to the incumbent by roughly 75%, and it decreases the likelihood of voting for Die Linke by approximately 55%. Satisfaction with democracy also decreases the likelihood of voting for the FDP relative to the CDU/CSU. Interest in politics is positively and statistically significantly related to voting for both the AfD and Die Linke over the the incumbent. Results are similar in Table 3.15, the multinomial model assessing constituency level voting. Satisfaction with democracy is negative and statistically significant in influencing votes for the AfD and Die Linke, relative to the incumbent. This translates to an approximate 78% decrease in the likelihood of selecting Afd, and 62% decrease in the likelihood of selecting Die Linke. It also statistically significantly decreases the likelihood of voting for the Greens by approximately 40%.

90 Table 3.13: Democratic Satisfaction and Voting for Anti-Establishment Parties-Germany 2013

(14)

AfD-PR AfD-SMD Die Linke-PR Die Linke-SMD

Satisfaction with Democracy -0.396*** -0.470*** -0.230** -0.325*** (0.114) (0.128) (0.106) (0.106) Positive Evaluation of Government -0.0168 -0.118*** -0.0224 -0.0202 (0.0405) (0.0426) (0.0338) (0.0348) Immigration Opinion -0.243*** -0.0529 0.0515 0.0445 (0.0785) (0.0930) (0.0472) (0.0476) Blame EU for Economic Crisis 0.333 -0.598 0.143 0.270 (0.247) (0.578) (0.228) (0.233) Income 0.0704 0.0441 -0.102*** -0.0824** (0.0498) (0.0631) (0.0363) (0.0400) Married -0.157 -0.313 0.361** 0.245 (0.167) (0.196) (0.141) (0.154) Female -0.264 -0.151 0.0514 -0.148 (0.161) (0.203) (0.119) (0.128) Born in Germany -0.329 -0.261 0.116 0.504 (0.312) (0.318) (0.446) (0.554) Positive Evaluation of Opposition -0.00748 -0.0337 -0.0456 -0.0294 (0.0288) (0.0368) (0.0309) (0.0325) Negative Evaluation of Government 0.383** -0.0949 0.239* 0.205 (0.190) (0.240) (0.135) (0.149) Ideology 0.176*** 0.162*** -0.400*** -0.363*** (0.0344) (0.0465) (0.0521) (0.0552) Positive View of National Economy -0.00752 0.0579 0.0191 0.0117 (0.0833) (0.118) (0.0893) (0.0879) Positive View of Personal Finances -0.297*** -0.102 -0.00550 -0.0925 (0.101) (0.120) (0.101) (0.0989) Employed 0.0781 0.227 0.202 0.229 (0.178) (0.210) (0.135) (0.150) Rural 0.0724 0.0740 0.0984 0.0962 (0.154) (0.176) (0.147) (0.160) Catholic 0.133 0.196 -0.439** -0.529*** (0.154) (0.196) (0.186) (0.183) Interest 0.0836 0.0844 0.180*** 0.0334 (0.0729) (0.0934) (0.0654) (0.0641) Education 0.0804 -0.0891 -0.0510 -0.0424 (0.0601) (0.0864) (0.0598) (0.0635) Religious Attendance 0.0354 0.0647 -0.188*** -0.136** (0.0610) (0.0742) (0.0672) (0.0605) Constant -0.937 -0.773 1.594** 1.780** (0.625) (0.734) (0.719) (0.725)

Observations 1,161 1,161 1,161 1,161

Note: Robust standard errors clustered by electoral district in parentheses. *** p<0.01, ** p<0.05, * p<0.1.

91 (a) AfD-PR Seats (b) AfD-SMD

Figure 3.9: Germany 2013. Predicted Probability of Voting for AfD

Note:Predicted probabilities of voting for AfD, proportional and single member district seats with 95% conﬁdence intervals. Control variables held at their mean values.

(a) Die Linke-PR Seats (b) Die Linke-SMD

Figure 3.10: Germany 2013. Predicted Probability of Voting for Die Linke

Note:Predicted probabilities of voting for Die Linke, proportional and single member district seats with 95% conﬁdence intervals. Control variables held at their mean values.

92 Table 3.14: Multinomial Results. Eﬀects on Probability of Selecting Parties Over the CDU/CSU- Party List Voting

(15) SPD FDP AfD Die Linke Greens

Satisfaction with Democracy -0.215 -0.575* -1.368*** -0.797*** -0.381 (0.226) (0.333) (0.287) (0.296) (0.252) Positive Evaluation of Government -0.622*** 0.0449 -0.382*** -0.554*** -0.628*** (0.0667) (0.124) (0.110) (0.0925) (0.0813) Immigration Opinion 0.170 0.00388 -0.397** 0.277** 0.414*** (0.117) (0.196) (0.200) (0.130) (0.129) Blame EU for Economic Crisis -0.0158 -0.470 0.916* 0.223 -0.573 (0.473) (0.889) (0.510) (0.582) (0.660) Married -0.0333 -0.629 -0.642 0.525 -0.328 (0.243) (0.592) (0.419) (0.353) (0.316) Female 0.165 0.771* -0.764* 0.172 0.153 (0.230) (0.468) (0.397) (0.294) (0.262) Born in Germany 0.935 -0.288 -0.363 0.955 -0.0502 (0.659) (0.873) (0.658) (1.286) (0.913) Positive Evaluation of Opposition 0.621*** -0.0780 0.172** 0.252*** 0.467*** (0.0745) (0.0988) (0.0874) (0.0872) (0.0915) Negative Evaluation of Government 0.623** 0.597 1.182*** 1.026*** 0.891*** (0.280) (0.633) (0.400) (0.324) (0.339) Ideology -0.605*** 0.0508 0.0399 -1.224*** -0.581*** (0.0982) (0.121) (0.108) (0.158) (0.106) Positive View of National Economy -0.445** -0.291 -0.408* -0.406* -0.598** (0.198) (0.280) (0.247) (0.244) (0.255) Positive View of Personal Finances 0.0510 -0.0995 -0.620** -0.126 -0.137 (0.196) (0.364) (0.283) (0.287) (0.252) Employed 0.0847 -0.480 0.271 0.545* 0.228 (0.229) (0.612) (0.452) (0.326) (0.274) Rural -0.0937 -0.976* 0.166 0.316 0.240 (0.280) (0.555) (0.390) (0.358) (0.370) Catholic 0.104 0.532 0.114 -0.653* -0.00909 (0.269) (0.435) (0.419) (0.375) (0.358) Interest 0.353** 0.407 0.365* 0.592*** 0.0298 (0.138) (0.258) (0.202) (0.173) (0.169) Income 0.0108 -0.0369 0.121 -0.130 0.271** (0.0711) (0.182) (0.137) (0.103) (0.110) Education -0.158 0.171 0.150 -0.144 0.167 (0.100) (0.248) (0.155) (0.131) (0.144) Religious Attendance -0.143* -0.0964 0.0960 -0.465*** 0.0218 (0.0819) (0.166) (0.141) (0.162) (0.0938) Constant 3.456** -0.827 4.988*** 8.895*** 3.101 (1.415) (2.237) (1.718) (2.085) (1.957)

Observations 1,077 1,077 1,077 1,077 1,077

Note: Robust standard errors clustered by electoral district in parentheses. *** p<0.01, ** p<0.05, * p<0.1, two tailed. Voting for the CDU/CSU is the base category.

93 Table 3.15: Multinomial Results. Eﬀects on Probability of Selecting Parties Over the CDU/CSU- Constituency Level Voting

(16) SPD FDP AfD Die Linke Greens

Satisfaction with Democracy -0.315 0.292 -1.478*** -0.949*** -0.495* (0.232) (0.334) (0.392) (0.304) (0.292) Positive Evaluation of Government -0.753*** -0.0659 -0.660*** -0.649*** -0.683*** (0.0775) (0.209) (0.136) (0.0991) (0.101) Immigration Opinion 0.200** 0.133 0.0188 0.222* 0.211 (0.0961) (0.217) (0.248) (0.120) (0.135) Blame EU for Economic Crisis -0.483 -0.156 -0.343 0.0847 -1.149 (0.435) (0.914) (1.021) (0.611) (0.895) Married 0.0221 -2.302*** -0.824 0.553 -0.385 (0.239) (0.804) (0.547) (0.345) (0.327) Female 0.0779 -1.502* -0.390 -0.258 0.0599 (0.232) (0.858) (0.580) (0.327) (0.306) Born in Germany 1.009 13.94*** -0.0163 2.322 1.695* (0.642) (1.020) (0.612) (1.608) (1.017) Positive Evaluation of Opposition 0.768*** -0.0504 0.243** 0.444*** 0.600*** (0.0843) (0.149) (0.113) (0.100) (0.101) Negative Evaluation of Government 0.109 -0.271 0.110 0.620 0.539 (0.307) (1.099) (0.608) (0.389) (0.374) Ideology -0.599*** -0.0157 -0.0396 -1.181*** -0.592*** (0.0980) (0.133) (0.159) (0.170) (0.117) Positive View of National Economy -0.305* -0.0746 0.0155 -0.352 -0.459* (0.180) (0.563) (0.340) (0.243) (0.256) Positive View of Personal Finances 0.192 -0.505 -0.296 -0.114 -0.106 (0.193) (0.656) (0.325) (0.288) (0.290) Employed 0.360* -1.373* 1.173** 0.852*** 0.653** (0.218) (0.725) (0.556) (0.322) (0.308) Rural -0.223 0.0897 0.0790 0.0927 0.0937 (0.292) (0.760) (0.472) (0.385) (0.385) Catholic 0.0358 -14.16*** 0.155 -0.922** -0.0436 (0.268) (0.566) (0.496) (0.397) (0.385) Interest 0.264* -0.150 0.257 0.231 0.135 (0.153) (0.356) (0.273) (0.189) (0.213) Income -0.0478 0.187 0.0728 -0.210** 0.0876 (0.0687) (0.151) (0.187) (0.105) (0.0987) Education -0.222** 0.452 -0.300 -0.150 0.183 (0.0904) (0.334) (0.232) (0.142) (0.154) Religious Attendance -0.185** -0.0547 0.0652 -0.420*** -0.139 (0.0850) (0.206) (0.201) (0.147) (0.135) Constant 3.477** -15.43*** 4.436* 8.506*** 1.822 (1.576) (2.660) (2.465) (2.370) (2.061)

Observations 1,076 1,076 1,076 1,076 1,076

Note: Robust standard errors clustered by electoral district in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Voting for the CDU/CSU is the base category.

94 Table 3.16: Eﬀects on Probability of Voting for the AfD over Die Linke

(17) (18) Party List Constituency Level

Satisfaction with Democracy -0.359** -0.135 (0.182) (0.238) Positive Evaluation of Government 0.00162 -0.101 (0.0652) (0.0754) Immigration Opinion -0.436*** -0.179 (0.143) (0.140) Blame EU for Economic Crisis 0.0650 -1.193* (0.391) (0.690) Married -0.641** -0.718* (0.305) (0.411) Female -0.135 0.0439 (0.330) (0.379) Born in Germany -0.573 -0.0618 (0.558) (0.799) Positive Evaluation of Opposition -0.0750 -0.131 (0.0737) (0.0818) Negative Evaluation of Government Handling of Crisis 0.231 -0.0320 (0.344) (0.327) Ideology 0.535*** 0.429*** (0.0918) (0.0933) Positive View of National Economy 0.271 0.0241 (0.258) (0.262) Positive View of Personal Finances -0.523** -0.0606 (0.250) (0.233) Employed -0.370 0.194 (0.350) (0.428) Rural 0.0938 -0.0258 (0.252) (0.367) Catholic 0.579* 0.664** (0.322) (0.327) Interest 0.177 0.242 (0.183) (0.187) Income 0.306*** 0.169 (0.0901) (0.125) Education -0.109 -0.116 (0.107) (0.162) Religious Attendance 0.354*** 0.327*** (0.124) (0.117) Constant -2.234* -2.688* (1.191) (1.523)

Observations 198 152

Note: Robust standard errors clustered by electoral district in parentheses. *** p<0.01, ** p<0.05, * p<0.1.

Finally, I assess the probability of voting for AfD, relative to Die Linke. Table 3.16 displays

95 the results from two probit models, where the dependent variable indicates voting for AfD, relative to Die Linke, in party list voting and constituency voting, respectively. Figure 3.11 displays the predicted probabilities from the party list voting. In party list voting, satisfaction with democracy has a negative and statistically significant influence on voting for AfD relative to Die Linke, but this relationship is not significant in constituency level voting. Respondents expressing the lowest level of satisfaction with democracy have a roughly 35% probability of voting for AfD versus Die Linke. At the highest expressed levels of democratic satisfaction, this drops to 20%. Immigration opinion also has a statistically significant and negative influence on choosing the AfD over Die Linke on in party list voting. Only a few predictors are statistically significant in both party list and constituency level voting: ideology, married, Catholic, and religious service attendance. Unsurprisingly, increasingly right-wing ideology corresponds to an increased probability of voting for AfD. Those who are married as less likely to vote for AfD, relative to Die Linke. Both Catholicism and religious services attendance are positively associated with selecting AfD over Die Linke.

Figure 3.11: Predicted Probability of Voting for AfD over Die Linke-Party List

Note: Predicted probabilities with 95% conﬁdence intervals. All other variables held at their mean values.

96 3.4 Discussion

An anti-establishment party represents a unique form of opposition, one that situates itself as an electoral alternative to failed and seemingly corruption-plagued mainstream parties. What drives voters to support such a drastic alternative to the status quo? Are the classic explanations of radical right wing support appear insufficient to explain electoral preferences for anti-establishment parties from the left, right, and center. The current analysis suggests that they are not. The recent successes of parties such as the Five-Star Movement, AfD, En Marche!, and the PVV, prompt us to consider additional factors behind their support. It appears that trust plays a role in vote choice not only on the domestic level, but also on the European level. In the context of the European Parliament, the Netherlands, Italy, and Germany, those voters with low levels of institutional trust are more likely to support the anti-establishment parties than the incumbent or the establishment opposition. In many cases, it also appears that those with high levels of trust are just as likely to support the incumbent or an establishment opposition alternative. This indicates that there is a difference in the minds of establishment supporters between broad systemic performance and the performance of specific policies and politicians. Additional research is needed to explore whether these relationships are conditional. For now, results indicate that those who voted for the PVV, the Five-Star Movement, AfD, and Die Linke possessed low trust in the system, whereas those who were simply dissatisfied with policies of the incumbent selected an establishment alternative. This suggests that there is a difference in how diffuse support and specific support influence voters’ behavior regarding anti-establishment and mainstream parties. Further, in the cases of the Netherlands and Italy, trust influenced vote choice between an establishment opposition party and a similarly ideologically positioned anti-establishment party. This suggests that the two separate dimensions of ideology and anti-establishment sentiment exist for voters, just as they do for parties. My expectations did not hold for the French case. In some ways, this is unsurprising given the vastly different nature of this presidential election, from the other cases’ parliamentary elections. It may be that in the context of a majoritarian system, trust’s relationship to vote choice may take on a different nature. Combined with the unique attributes of the 2017 Presidential election, further exploration of the French cases is needed. Given these differences, low trust did play a role in the selection of Le Pen in the second round of voting. This may indicate that Macron’s image

97 may in fact be that of an insider. He did serve as Hollande’s economics minister, despite running with his own newly created party. Other interesting patterns emerge from the French analysis. Specifically, the role of immigration attitudes in this election presents interesting food for thought. In the first round, negative perceptions about immigrations’ impact on the economy influenced votes for Macron and Le Pen, while only cultural antipathy (immigration’s impact on culture, crime, etc.) related to votes for Le Pen. This might suggest an intriguing relationship between the type of immigration fears and the relative positioning of parties along both the ideological and the anti-establishment dimensions. In short, anti-immigration attitudes may not be completely and solely owned by the right. As argued in the previous two chapters, the anti-establishment encompasses more than just the radical right-wing. Therefore, explanations for their electoral support should also extend beyond the traditional predictors of radical right-wing support. As anti-establishment parties continue to increase their electoral viability, we must update our understanding of the sources of their support. The contexts of these particular elections are also illuminating. The 2014 European Parliamen- tary elections saw an incredible influx of these parties in the aftermath of economic crisis. The 2010 Dutch election saw the PVV filling a void left by the defunct List Pim Fortuyn. The party seems to exist at the intersection of radical-right and anti-establishment, indicating that scholars may need to take special care when theorizing about voter support for these outsider parties. In 2013, Italian politics was in disarray, permitting the entrance of a largely politically inexperienced, M5S. At the time, M5S was distinctly left, but drew supporters from across the ideological spectrum. Germany’s role in the eurozone crisis and the migration crisis has heightened nationalism, paving the way for AfD to gain ground. The 2017 French election was historical, in that the two major parties ceded control of the presidency for the first time since in nearly seven decades. Ideology is only part of the formula of their attraction to voters. They appeal to those voters who lack trust and are dissatisfied with the status quo. Those who lack trust come from all socio-economic backgrounds and levels of education. Ideology may sort voters toward parties on one dimension, but trust appears to sort voters along the anti-establishment dimension. The widespread appeal of these parties cannot be denied. Other parties, such as Podemos in Spain, the Danish People’s Party, UKIP, and the FPO´ in Austria are set to alter their countries’ domestic politics and relationship with the EU. The latest electoral victories of 2017 and 2018

98 demonstrate this reality acutely. AfD finally did succeed in attaining seats in the German Bun- destag. As the largest opposition party, they represent the first radical right-wing party since World War II to serve in parliament. Italian politics is dominated by a coalition between the M5S and Lega Nord, two anti-establishment parties that are attacking Italy’s membership in the eurozone. These parties flaunt their differences from the mainstream, while stressing the failures of traditional parties. While certain policy areas may divide them, such as immigration, low trust does appear to be a unifying explanation for their popularity. As research in this area progresses, it will become increasingly important to consider other aspects of the anti-establishment phenomenon. Understanding how these parties behave and how receptive voters are toward the rhetoric, activities, and ideological positioning of theses parties will be key in determining why voters would select seemingly unqualified and extreme candidates. Understanding the dynamics of these parties and their voters will require more than the traditional literature on the radical right. Trust in traditional European parties and politicians appears to be irretrievably broken in the minds of voters. They are seeking a solution outside of status-quo politics. Whether this solution comes from the left or the right may not matter. What may be more salient, is whether it comes from the outside the political status quo or from within. Low trust appears to translate into increased support for anti-establishment parties, from all areas of the ideological spectrum. How will mainstream European parties compete with charges of widespread corruption, ineptitude, and calls for populist solutions to the policy problems? The increasing viability of these parties stands to change status quo policy. Now that these parties are indeed electorally viable, and in a position to not only win seats, but run governments, what will define their policy agendas? In the next chapter, I explore how a party’s anti-establishment sentiment influences its behavior in parliaments.

99 CHAPTER 4

FROM MAKING NOISE TO MAKING POLICY: PARLIAMENTARY ACTIVITIES OF THE ANTI-ESTABLISHMENT

As the previous chapters illustrate, the anti-establishment movement encompasses deep-seated sentiments about the current political elite. These parties assert that rampant corruption on the part of the political elite has led the country into a crisis that only be solved by the people and a restoration of national pride. The previous chapter explored the sources of these parties’ electoral support, ﬁnding that institutional trust on the part of the electorate has potential to explain support for left and right wing anti-establishment parties. These parties have long served as noise makers in their respective political systems. Either excluded from political power, or relegated to a minor role within their systems, these parties have had little opportunity to play an active role in setting the domestic political agenda. How- ever, as discussed and demonstrated in chapter 3, anti-establishment parties have gained a certain degree of electoral viability, achieving seats in parliaments and in local government (e.g. the Netherlands, Italy, France, and Germany) and even working to support minority governments (e.g. the Netherlands). With this recent surge of anti-establishment parliamentary representation in domestic legislatures, attention should shift from how these parties derive support (Bustikova 2014, Golder 2003, Golder 2016), to how they behave once seated in parliaments. In this chapter, I assess anti-establishment members’ behavior across two dimensions: level of parliamentary activity and agenda area focus. Anti-establishment parties are well known for “making noise” (Williams 2006). Outside of parliament, this often has meant holding rallies, issuing press releases, or maintaining blogs (e.g. the Italian Five Star Movement). I surmise that, within parliaments, these parties’ behavior is often obstructionist, many times delaying legislation, and aimed toward garnering attention to their cause. Because they typically do not hold many seats or positions of power, this behavior may maximize their exposure, permitting them to cultivate a personal reputation among constituents (Sozzi 2016). This behavior is akin to that of opposition

100 parties, well known to be more active in parliament, i.e. asking more questions (Vliegenthart and Walgrave 2011). While parliamentary questions may not seem obstructionist, it is important to remember that within the context of a parliamentary system, questions remain one of the few tools possessed by opposition parties to act in an obstructionist manner. Questions provide an opportunity to challenge the governing party, put forth their agenda, and delay the business of parliament. Members of parliaments (MPs) and political parties may have a variety of factors that dictate their behavior: re-election concerns, party dynamics, institutional constraints, and even individual preferences (Mayhew 1974, Vliegenthart and Walgrave 2011, Jones and Baumgartner 2005, Budge and Farlie 1983, Rihoux and Rüdig 2006, Ström1997). While it has been noted that radical parties do sometimes temper their rhetoric when incorporated into governing coalitions (Albertazzi and McDonnell 2008), anti-establishment parties’ behavior within the context of domestic parliaments remains somewhat understudied. How do members of anti-establishment parties behave differently than their establishment counterparts, specifically members of establishment opposition parties? While we may expect that anti-establishment parties behave in a manner consistent with other opposition parties, i.e. increased levels of activity and challenging the government (Green- Pedersen 2007), it is important to understand that anti-establishment parties are a unique brand of opposition party. In fact, anti-establishment parties pride themselves as an alternative to all establishment parties, even those that do not occupy the majority of the legislature or hold government portfolios (Abedi 2009). For this reason, I argue that anti-establishment parties may use their new found electoral viability to further demonstrate their unique message and position in the political environment. Members from anti-establishment parties may also differ from their establishment counterparts in the types of issues on which they focus. There exist a plethora of issues on which politicians and parties may focus, however political actors have scarce time to devote to issues (Jones and Baumgartner 2005). This reality of legislating means that when MPs and parties choose one issue over another to attend to, they are making a meaningful choice about their agenda (Vliegenthart and Walgrave 2011, Green-Pederson and Mortensen 2009). Scholars have paid much attention to issue ownership of mainstream and niche parties (Wagner 2011, Ezrow 2008, Meguid 2005). Newer and smaller parties are recognized as more likely to adopt novel and emerging issues (Meguid 2005). This

101 strategy may change as parties become increasingly electorally viable (Meyer and Wagner 2013). While anti-establishment parties certainly represent emerging and newer parties in many systems, I argue that this does not necessarily dictate that their agenda will be a niche one. While potentially motivated to pursue issues they perceive as neglected by the establishment (Adams, Clark, Ezrow, and Glasgow 2006), I maintain that anti-establishment parties, due to their desire to draw attention to their movements, will attempt to cultivate issue ownership over core issues in an eﬀort to enhance their electoral credibility. I assess anti-establishment behavior across two dimensions: level of parliamentary activity and issue area of attention. I leverage European Parliamentary data from the eighth session to assess parties’ behavior. Anti-establishment parties have a tendency to perform better in the European elections, relative to domestic parliamentary elections. The 2014 European Parliamentary election in particular saw the seating of several new parties, such as the Italian Five Star Movement, Spanish Podemos, and the AfD, and gains for institutionalized parties such as the National Front, the Danish Peoples Party, and Lega Nord. I also use Italian Parliamentary data across ﬁve sessions (1996-2014) to observe how these parties behave in a domestic context. The Italian political environment is no stranger to party volatility and several anti-establishment parties, such as Forza Italia, Lega Nord, and the Five-Star Movement. I demonstrate that anti-establishment parties, and their members, occupy a unique position within parliaments. They are more active and adopt issues traditionally owned by mainstream parties. Not only do they challenge the ruling party and government, but they also present roadblocks to establishment opposition parties. Examining both the level of activity (number of questions), and the issue area of focus (context of parliamentary questions), permits me to move beyond where extant literature ends (explaining the popularity of such parties). Instead, I provide additional insight into the behavior and motivations of anti-establishment parties once they are in the business of making policy.

4.1 Parliamentary Activities, MPs, and Party Dynamics

A great deal of scholarship is devoted to agenda-setting activities of governments, with most focus centering on the choices government and politicians make in prioritizing one issue over another. Jones (2001) observes that governments prioritize issues that are more important to them, but scholarship has also acknowledged that some issues (defense, international aﬀairs, the economy,

102 government operations, and the rule of law) represent “core” issues prioritized over all others (Jennings 2011). Furthermore, because of limited time and resources, governments and politicians are bounded in their decision-making (Simon 1971). Two competing theories exist to explain what prompts parties to adopt particular agendas. Those who posit that parties pursue issues over which they have a comparative advantage claim that certain parties have “issue ownership” over particular policy areas (Budge and Farlie 1983, Petrocik 1996). This can be viewed as a “top-down” process, by which certain parties are perceived to be more competent in certain areas than other parties (Klüver and Sagarzazu 2016). Contrary to this view is the notion that parties attempt to “ride the wave” of public opinion, hoping to signal to voters that they care about salient issues. This is a “bottom-up” process, by which parties take their cues from the electorate directly (Ansolabehere and Iyengar 1994, Klüver and Sagarzazu 2016). In this vein, a variety of factors, such as re-election concerns, party dynamics, and personal preferences may dictate the activities of parties and members of parliaments (Müllerand Saalfeld 1997, Mayhew 1974, Vliegenthart and Walgrave 2011). In the context of European parliamentary systems, parties control how their members behave. Parties themselves pursue agenda items as a reflection of their base. Generally parties are risk-averse, maintaining their agenda focus as loyalty to their constituents, and to avoid alienating their base (Kitschelt 1994, Ström1990). In this respect, mainstream parties cultivate issue ownership over those issues that reflect their comparative advantage (Budge and Farlie 1983). These well established and mainstream parties, because of their electoral viability will keep their policy agendas stable (Budge, Ezrow, and McDonald 2010). Some parties pursue niche agendas, attempting to proclaim ownership over issues seemingly neglected by the establishment (Meguid 2005, Meyer and Wagner 2013, Wagner 2011, Adams et al. 2006, Ezrow 2008). However, electoral incentives may prompt these niche parties to pursue more mainstream agendas (Meyer and Wagner 2013). Also, as parties begin to participate in the government (i.e., when the anti-establishment actually wins seats), they may seek to switch from niche to mainstream agendas (Rihoux and Rüdig2006, Abedi and Lundberg 2009). This particular trend has been noted with Italian party Lega Nord (Albertazzi, McDonnell, and Newell 2011). A radical right-wing and anti-establishment party, Lega Nord was able to maintain a position in a coalition for a number of years. The consequence of this is that Lega Nord tempered

103 its rhetoric and adopted more mainstream stances on issues, enhancing their electoral viability (Albertazzi, McDonnell, and Newell 2011). Furthermore, other anti-establishment parties have the ability to move the agendas or otherwise affect mainstream and governing parties. As found in Denmark, parliamentary questions raised by opposition parties do have an impact on a prime minister’s speeches, and the ruling party’s agenda (Green-Pederson and Mortensen 2009). Parliamentary activities are an important aspect of how parties and their members in parliament convey issue priorities and signal that they are doing the work of representation (Sozzi 2016). Specifically, parliamentary questions and interpellations have become increasingly important for parties and MPs. Because legislation in a parliamentary system is dominated by the ruling party, and not the rest of parliament (Ström,Müller,and Bergman 2003), the function of MPs outside of the government has shifted from law-making to the critique of government action and policies (Green-Pedersen 2007). Generally, scholarship has found that opposition parties are less constrained than those in the government and tend to ask more questions (Vliegenthart and Walgrave 2011, Green-Pedersen 2007, Wiberg and Koura 1994). Opposition parties serve as an additional check on government action. Parliamentary activities of the opposition are designed to both challenge the government and vocalize their positions.

4.2 The Anti-Establishment Agenda and Legislative Activities

The extant literature on the rise of anti-establishment or outsider parties has largely focused on how these groups aﬀect their domestic political contexts by “making noise”(Williams 2006). Typically these fringe parties have not held important positions in their domestic governments, and in many areas, they are not even electorally viable. Instead, they have been noted as driving political discourse through activities, such as protest. As they have become increasingly electorally viable, it is important to consider how anti-establishment parties inﬂuence parliamentary agendas. I argue that anti-establishment, or “outsider,” parties represent a unique brand of opposition. Critiquing status quo policies and programs is a mainstay of anti-establishment political activity, as they desire to challenge mainstream establishment parties, both those in government and those in opposition (Abedi 2009, Schedler 1996, Bustikova 2014). In its need to cultivate issue ownership, anti-establishment actors in parliament must pursue agenda areas of particular saliency (i.e. core issues). Once in parliament, anti-establishment parties will challenge establishment elites on issues

104 classically attributed to mainstream parties. They will rarely pursue agendas for niche topics, as one of their goals is to usurp issue ownership away from the establishment. They view themselves as an alternative to the establishment in multiple policy areas (Abedi 2009), meaning that they must assert their authority on multiple mainstream topics. The content of their agenda will fundamentally reﬂect a desire to discuss salient topics that they believe establishment elites have mishandled. This may be especially true when competing with establishment parties that occupy the same space along the ideological spectrum. By adopting issues of greater saliency, they not only successfully draw attention to their cause, but they also increase their electoral viability by adopting issues important to voters (Meyer and Wagner 2013). As they become electorally viable and obtain seats in parliament, these anti-establishment parties will behave in a manner similar to, but more active than, mainstream opposition parties. It is well documented that opposition parties tend to increase their level of parliamentary activities (Vliegenthart and Walgrave 2011, Green-Pedersen 2007). This serves to increase their exposure and signal to constituents that they are productive parliamentarians. In the case of the anti- establishment, the goal is to attract as much attention as possible to the party’s agenda, both to signal competence to constituents, but also to further exacerbate diﬀerences with established political elite. Because these parties tend to hold fewer seats in the parliament, and are often excluded from governing coalitions, parliamentary activities such as interpellations and questions provide the opportunity to demonstrate their political prowess. The more active anti-establishment MPs are, the more they can credibly claim to challenge the status quo. For anti-establishment MPs, the vocalization of their agenda is just as important as the content of that agenda itself. Anti-establishment parties, from both sides of the ideological spectrum, are in natural competition with ideologically similar establishment parties. Anti-establishment politicians will directly challenge establishment parties with similar ideological leanings by attempting to steal away issue ownership. They then sell themselves as credible alternatives to the political establishment by actively bucking establishment authority and cultivating issue ownership traditionally attributed to classic establishment parties (Ansolabehere and Iyengar 1994, Budge and Farlie 1983). These parties use broad murmurings of dissatisfaction as a way to turn negative sentiment into cause for alarm, thereby folding mainstream issues into their agenda (Williams 2006). Therefore, anti- establishment politicians will not only be more active than their establishment counterparts, but

105 they will also attempt to steal away issue ownership by focusing on salient core issues typically pursued by the mainstream parties.

H1: Anti-Establishment politicians will be active parliamentarians, particularly they will be more active than members of the establishment opposition, asking more parliamentary questions.

H2: Anti-Establishment politicians will devote their questions to core issues (macro-economics, ﬁnance, law, international aﬀairs, defense, and government operations), at the same level or ex- ceeding establishment party politicians.

H3: Both level of activity and issue prioritization will be conditioned by a party’s ideology.

4.3 Data and Methods

To assess the inﬂuence of being a member of an anti-establishment party on parliamentary activity, I use European Parliamentary data from May 2014 -May 2017 (8th Session), and Italian parliamentary question data across ﬁve sessions from 1996 to 2014. Both datasets are coded according to the Comparative Agendas Project guidelines. The Comparative Agendas Project (CAP) is a cross national endeavor to collect and code agenda setting data according to a common codebook. Agenda items, such as parliamentary questions are coded according to content and categorized according to issue area.12 To measure a politician’s level of activity, I examine her number of parliamentary questions. Other types of activities could be used, such as bill introductions and sponsorships. However, questions often allow members to pursue areas of personal, constituent, or party concern with little interference from party leadership (Green-Pederson and Mortensen 2009, Green-Pedersen and Mortensen 2010, Vliegenthart and Walgrave 2011). Also, because the legislative agenda (i.e. bill

1For more information on countries’ speciﬁc coding guidelines, intercoder reliability, and to access the master codebook, please visit www.comparativeagendas.net.

2The Italian dataset is coded and provided by the Italian Agendas Project available from https://www.comparativeagendas.net/italy.

106 introductions) are dominated by the ruling party, parliamentary questions are representative of the agenda concerns for those outside the ruling party (Str¨om,M¨uller, and Bergman 2003). For the European Parliamentary data, I was able to account for all members’ level of parliamentary activity (i.e. the number of parliamentary questions asked) and demographic information for the time period between May 2014 and May 2017. There are over 30,000 parliamentary questions asked in this time period. I coded a random sample of 1,000 questions according to their major topic.3

Table 4.1: Data Availability by Case

Institution Data Years Available European Parliament Parliamentary Questions May 2014-May 2017 Italy Chamber of Deputies Parliamentary Questions 1996-2014

Note: Italian data available from the Italian Agendas Project, www.comparativeagendas.net/italy. European Union data available from the European Union oﬃcial website http://www.europarl.europa.eu/portal/en. Questions were coded according to Comparative Agendas Codebook guidelines.

4.3.1 Dependent Variables

I conduct analyses on the European Parliament and the Italian Chamber of Deputies separately. For both bodies, I analyze the same main dependent variables. First, for Hypothesis 1, I use a variable that indicates the number of questions asked by a politician. I use a negative binomial regression to assess the predicted number of questions asked by a politician according to her anti- establishment status, as developed in Chapter 2. Second, for Hypothesis 2 and 3, I conduct a series of analyses to determine the issue prioritization for members according to their parties’ anti-establishment status. First, I use a binary variable that indicates whether a politician introduced a parliamentary question pertaining to a core issue. Core issues are traditionally identified as economics, law, defense, international affairs, and government operations (Jennings 2011). These issues are traditionally seen as “core” issues that are prioritized over others (Jones 2001), and are salient national issues. The variable I created, core issues, values “1” a question pertains to macro-economics, finance (jointly considered “economics”), law and

3Undergraduate research assistants assisted in the coding of questions. Intercoder reliability was over 90%. Dis- crepancies were reconciled in a meeting between myself and the coders.

107 crime, defense, international aﬀairs, and government operations. I use probit models to assess the probability that a politician will introduce a question pertaining to a core issue. Second, I analyze the proportion of a members’ agenda devoted to core issues, according to her parties’ anti-establishment score. It is a continuous variable (proportion core), so I utilize an ordinary least squares model (OLS) to analyze the relationship between these variables. I assess the relationship between anti-establishment sentiment and agenda both among all members of a the legislative body and between anti-establishment members and establishment opposition members.

4.3.2 Explanatory and Control Variables

The main explanatory variable is a party’s anti-establishment score, as calculated in Chapter Two. Anti-Establishment Score ranges from 1-10, and is continuous. Higher values indicate a greater degree of anti-establishment sentiment present in a party’s manifesto. I expect that as parties’ anti-establishment scores increase, politicians will become more active in parliaments, and this will inﬂuence their attention to salient, core issues. To assess how anti-establishment politicians’ behavior diﬀers from that of establishment opposition members, I use a binary variable Anti- Establishment where “1” indicates a member of an anti-establishment party, and “0” indicates a member of an establishment opposition party.4 I also include several control variables. I control for a handful of individual level variables, party level, and country level variables. For the analyses of European Parliament, I control for an individual member’s gender (Cowley and Childs 2003), years in service, and age. In Italy, I control for the members’ gender. As core issues tend to be under the purview of governing parties (Jennings 2011), I control for whether the party asking the question is a member of a domestic governing coalition, using a binary variable, where “1” indicates the party part of the governing coalition. Because parliamentary questions are typically used by opposition parties, and not governing parties (Green-Pederson and Mortensen 2009), I control for the percentage of seats held by the party in domestic parliament. Because parties are perceived to have a comparative advantage over certain issues (Budge and Farlie 1983), I control for asking parties’ ideology. I use RILE scores from the Comparative Manifesto Project (CMP). RILE scores range from -100 (perfect Left) to 100 (perfect Right). I rescale this variable, Ideology, to range from 1-10, for ease

4Parties whose anti-establishment score was above the median (5) were coded as anti-establishment.

108 of interpretation. To test Hypothesis 3, I include an interaction term between anti-establishment and ideology.

4.4 Results 4.4.1 European Parliament

Anti-establishment parties have been more successful in gaining seats in the eighth Session of the European Parliament than in most domestic parliamentary elections. Parliamentary questions in the European parliament are often used by individual Members of European Parliament (MEPs) to express an interest of personal importance to them or their constituents. The constraints parties place on their members varies according to domestic level institutions (Hix 2004). A member’s behavior is both a function of her national party and her European party group (Hix 2004). Further, because of variations in ambition, members may choose to appease the party group or the national party (Meserve, Pemstein, and Bernhard 2009), though party group cohesion is typically high in the European parliament (Carrubba and Gabel 1999). What does this mean for the relatively new anti-establishment parties, who have largely joined eurosceptic parliamentary groups? Broadly, I expect these parties to be more active than their mainstream counterparts, especially as compared to those who serve in domestic establishment opposition parties. As opposition parties (either within or outside of parliament) in their home countries, they have an incentive to vocalize their agendas once elected. Further, because many of these parties are indeed part of eurosceptic groups in the EU Parliament, they will use parliamentary questions to obstruct the business of the EU. Table 4.3 displays the results from negative binomial regressions with and without an interaction between anti-establishment score and ideology. In the non-interactive model (Model 1), increases in anti-establishment score influence MEPs to ask more questions. Figure 4.1 displays the results from two negative binomial regressions, one without an interaction between anti-establishment score and ideology, and one including this interaction. Going from the lowest anti-establishment score to the highest yields a significant jump in the predicted number of questions, going from approximately sixty-six questions to roughly two-hundred ninety-six questions at the highest levels of anti-establishment sentiment. Ideology is not statistically significant in either model.

109 Table 4.2: European Parliament. Eﬀect of Anti-Establishment Score on the Number of Questions Asked

(1) (2) PQs PQs

Anti-Establishment 0.165*** 0.428** Score (0.0594) (0.202) Ideology -0.0616 0.0603 (0.0824) (0.135) AE Score x -0.0542 Ideology (0.0390) Female 0.0521 0.0594 (0.144) (0.145) Age -0.00784 -0.00698 (0.00613) (0.00622) Member of Domestic -0.296* -0.299* Governing Coalition (0.168) (0.169) Share of Seats -0.117 -0.130 in Domestic Parliament (0.580) (0.584) Years as MEP -0.0302** -0.0303** (0.0124) (0.0126) Constant 5.082*** 4.429*** (0.664) (0.874)

Observations 1,137 1,137

Note: Robust standard errors clustered by party in parentheses. *** p<0.01, ** p<0.05, * p<0.1.

(a) Predicted Number of Questions (b) Marginal Eﬀect of Ideology on Predicted Number of Questions

Figure 4.1: Predicted Number of Questions-European Parliament May 2014-May 2017

Note:Results of Negative Binomial Regressions with 95% conﬁdence intervals. Control variables held at their mean values.

110 The interaction between anti-establishment score and ideology is not statistically significant either, indicating that ideology has little influence on the level of MEPs’ level of activity. This is suggestive evidence that ideology and anti-establishment sentiment should not be conflated, and do indeed have disparate influences on legislative behavior. As Figure 4.1 illustrates, right-leaning ideology appears to have a moderating effect on anti- establishment score’s influence on number of questions asked. Despite the fact that these effects are not statistically significant, it does pose interesting food for thought. While the radical right-wing is often credited for being the loudest and most noteworthy movements in Europe currently, it appears as though ideology plays little to no role in determining the level of activity of a MEP.

Table 4.3: European Parliament. Number of Parliamentary Questions Anti-Establishment Parties versus Establishment Opposition

(3) (4) PQs PQs

Anti-Establishment 0.425** 1.712*** (0.168) (0.622) Ideology -0.0417 0.0656 (0.0776) (0.0835) Anti-Establishment x Ideology -0.253** (0.115) Female -0.0649 -0.0651 (0.130) (0.133) Age -0.0102** -0.00898* (0.00464) (0.00460) Member of Domestic Governing Coalition 0.274 0.229 (0.378) (0.370) Share of Seats in Domestic Parliament -0.400 -0.367 (0.602) (0.612) Years as MEP -0.0291** -0.0278** (0.0135) (0.0138) Constant 5.370*** 4.713*** (0.565) (0.644)

Observations 827 827

Note: Robust standard errors clustered by party in parentheses. *** p<0.01, ** p<0.05, * p<0.1.

It should be expected that any opposition party will be more active than governing parties (Green-Pederson and Mortensen 2009). I speciﬁcally argue that anti-establishment parties will be more active than establishment opposition parties, due to their desire to challenge all mainstream

111 parties. Table 4.3 displays results assessing the difference between anti-establishment parties and establishment opposition parties. Figure 4.2 displays the predicted number of questions. In both Models 3 and 4, anti-establishment MEPs ask more question than establishment opposition politicians. Once again, right-wing ideology has a moderating effect on this relationship. In the non-interactive model, going from an establishment opposition party to an anti-establishment party increases a MEP’s number of questions by forty questions. The interactive model reveals a moderating effect, with increasing right-wing ideology reducing an anti-establishment member’s level of activity by roughly twenty questions. To further demonstrate the differences in predicted questions, Figure 4.3 displays the predicted number of question for four groups: left-wing anti- establishment members, right-wing anti-establishment members, left-wing establishment opposition members, and right-wing establishment opposition members.

(a) Predicted Number of Questions (b) Marginal Eﬀect of Ideology on Predicted Number of Questions

Figure 4.2: Predicted Number of Questions Anti-Establishment Parties versus Establishment Op- position

Note: Results of Negative Binomial Regressions with 95% conﬁdence intervals. Control variables held at their mean values.

112 Figure 4.3: European Parliament. Predicted Number of Questions By Anti-Establishment Status and Ideology

Note: Predicted number of questions with 95% conﬁdence intervals. All other variables held at their mean values.

Among the four groups, left-wing anti-establishment politicians are the most active, predicted to ask approximately one-hundred twenty questions. There is no discernible diﬀerence between right-wing anti-establishment members and left-wing establishment opposition members, though both ask more questions than right-wing establishment opposition members. This begs the question as to why would right-wing ideology temper anti-establishment members’ behavior? Perhaps right- wing anti-establishment parties, such as the National Front, are more institutionalized, shaping their members’ behavior more like establishment opposition parties. Additional research is needed to examine the depth of this relationship. However, this is additional evidence that right-wing behavior and anti-establishment behavior should not be conﬂated. We should not assume certain behaviors based solely on ideological leaning. In addition to level of activity, I also assess the agenda areas of MEPs. Tables 4.4 and 4.5 present the results of probit models, assessing the probability of a MEP asking a question pertaining to a core issue.

113 Table 4.4: European Parliament. Eﬀect on Probability of Asking a Question Devoted to a Core Issue

(5) (6) Focus on Core Issues Focus on Core Issues

Anti-Establishment Score 0.0291 0.134 (0.0363) (0.134) Ideology -0.0130 0.0405 (0.0475) (0.0776) Anti-Establishment Score x Ideology -0.0209 (0.0268) Female 0.0675 0.0722 (0.0891) (0.0897) Age -0.00164 -0.00121 (0.00438) (0.00449) Member of Domestic Governing Coalition -0.0201 -0.0188 (0.122) (0.121) Share of Seats in Domestic Parliament -0.107 -0.125 (0.338) (0.337) Years as MEP 0.00922 0.00943 (0.00944) (0.00946) Constant -0.308 -0.608 (0.353) (0.517)

Observations 876 876

Note: Robust standard errors clustered by party in parentheses. *** p<0.01, ** p<0.05, * p<0.1

(a) Non-Interactive (b) Marginal Eﬀect of Ideology

Figure 4.4: European Parliament. Predicted Probability of Asking a Question Pertaining to a Core Issue Across Anti-Establishment Scores

Note: Predicted probabilities with 95% conﬁdence intervals. Control variables held at their mean values.

114 The results of Models 5 and 6 reveal that while anti-establishment score does have a positive influence on the probability of asking a question on a core issues, this is not a statistically significant relationship. However, none of the covariates have a statistically significant influence on MEPs’ attention to core issues. Figure 4.4 plots the predicted probabilities for a MEP to ask a question pertaining to a core issue across anti-establishment scores. The direction of the effects for both the base model and the interactive model echo the findings on activity, where right-wing ideology has a moderating influence on anti-establishment status. Table 4.5 displays the results comparing anti-establishment parties to establishment opposition parties. Figure 4.5 plots the marginal effect of ideology on the probability of asking a question about a core issue. Once again, the effects are in the expected direction, but are not statistically significant.

Table 4.5: European Parliament. Eﬀect on Probability of Asking a Question Devoted to a Core Issue Anti-Establishment Parties versus Establishment Opposition

(7) (8) Focus on Core Issues Focus on Core Issues

Anti-Establishment 0.0471 0.770 (0.111) (0.539) Ideology -0.0191 0.0470 (0.0522) (0.0656) Anti-Establishment x Ideology -0.142 (0.105) Female 0.0348 0.0411 (0.108) (0.109) Age -0.000690 0.000128 (0.00505) (0.00516) Member of Domestic Governing Coaltion 0.0643 0.0446 (0.216) (0.207) Share of Seats in Domestic Parliament 0.0189 0.000744 (0.437) (0.429) Years as MEP 0.00110 0.00246 (0.0137) (0.0139) Constant -0.227 -0.633 (0.343) (0.438)

Observations 669 669

Note: Robust standard errors clustered by party in parentheses. *** p<0.01, ** p<0.05, * p<0.1.

115 (a) Non-Interactive

(b) Marginal Eﬀect of Ideology

Figure 4.5: European Parliament. Predicted Probability of Asking a Question Pertaining to a Core Issue Anti-Establishment Parties versus Establishment Opposition

Note: Predicted probabilities with 95% conﬁdence intervals. Control variables held at their mean values.

Next, I examine what proportion of a MEP’s agenda is devoted to core issues. Table 4.6 displays the results using the proportion of a members’ agenda devoted to a core issue as the dependent variable.

116 Table 4.6: European Parliament. Proportion of Questions Devoted to Core Issues Across Anti- Establishment Scores

(9) (10) Focus on Core Issues Focus on Core Issues

Anti-Establishment Score 0.0110 0.0508 (0.0139) (0.0509) Ideology -0.00490 0.0153 (0.0180) (0.0293) Anti-Establishment Scores x Ideology -0.008 (0.0101) Female 0.0256 0.0272 (0.0338) (0.0340) Age -0.000620 -0.000459 (0.00166) (0.00171) Member of Domestic Governing Coalition -0.00752 -0.00702 (0.0456) (0.0455) Share of Seats in Domestic Parliament -0.0392 -0.0460 (0.127) (0.126) Years as MEP 0.00349 0.00357 (0.00361) (0.00361) Constant 0.379*** 0.266 (0.135) (0.197)

Observations 876 876 R-squared 0.006 0.008

Note: Robust standard errors in parentheses.*** p<0.01, ** p<0.05, * p<0.1

Once again, no variables are significant, though they are in the expected direction. This may indicated that the anti-establishment does not behave vastly differently from establishment opposition parties. They are not focusing on niche issues, and are just as likely to devote time to core issues as mainstream parties. While I did argue that anti-establishment politicians would attempt to steal away issue ownership from the establishment, the results here indicate that they are at the very least, keeping pace with the mainstream. Figure 4.6 plots the predicted proportions across anti-establishment scores, and the marginal effect of ideology on the proportion of questions pertaining to core issues.

117 (a) Non-Interactive

(b) Marginal Eﬀect of Ideology

Figure 4.6: European Parliament. Predicted Proportion of Agenda Pertaining to a Core Issue Across Anti-Establishment Scores

Note: Predicted probabilities with 95% conﬁdence intervals. Control variables held at their mean values.

As in the previous models, right-wing ideology does appear to have a moderating (albeit not statistically significant) influence on the dependent variable. Table 4.7 examines the difference between anti-establishment parties and establishment opposition parties. Figure 4.7 displays the results graphically.

118 Table 4.7: European Parliament. Proportion of Questions Devoted to Core Issues Anti- Establishment Parties versus Establishment Opposition

(11) (12) Focus on Core Issues Focus on Core Issues

Anti-Establishment 0.0180 0.292 (0.0422) (0.204) Ideology -0.00733 0.0178 (0.0198) (0.0250) Anti-Establishment x Ideology -0.0538 (0.0397) Female 0.0134 0.0153 (0.0412) (0.0416) Age -0.000266 3.95e-05 (0.00193) (0.00197) Membership in Domestic Governing Coalition 0.0249 0.0173 (0.0836) (0.0795) Share of Seats in Domestic Parliament 0.00751 0.000991 (0.166) (0.162) Years as MEP 0.000422 0.000939 (0.00522) (0.00527) Constant 0.410*** 0.256 (0.132) (0.167)

Observations 669 669 R-squared 0.003 0.009

Note: Robust standard errors clustered by party in parentheses.*** p<0.01, ** p<0.05, * p<0.1.

(a) Non-Interactive (b) Marginal Eﬀect of Ideology

Figure 4.7: European Parliament. Predicted Proportion of Agenda Pertaining to a Core Issue Across Anti-Establishment Parties versus Establishment Opposition Scores

Note: Predicted probabilities with 95% conﬁdence intervals. Control variables held at their mean values.

119 As was demonstrated by all the models above, the effects are in the expected direction, but lack statistical significance. Further research is need to tease out this relationship. Interestingly, right-wing ideology appears to have a moderating influence on behavior in the eighth session of the European parliament. Additional research on the institutionalization of parties may shed light on this relationship. Some of the right-wing parties in the European Parliament, such as the National Front, and the Danish People’s Party have existed longer than left-wing parties, such as Five-Star Movement, Podemos, and various Pirate Parties. Perhaps these parties behave more like establishment and mainstream parties as they become institutionalized. To further examine what MEPs are devoting their attention to, I display descriptive data for governing parties, establishment opposition parties and anti-establishment parties in Table 4.8. I disaggregate the core issues variable into its component parts.

Table 4.8: Core Issue Attention in the European Parliament

Governing Parties Establishment Opposition Anti-Establishment Macro-Economics 8.4% 13.8% 12.8% Law & Crime 19.3% 12.5% 20.1% Finance 17.7% 17.8% 18.1% International Relations 32% 36.8% 26% Defense 3.4% 3.3% 5.9% Government Operations 19.3% 15.8% 17.2%

Note: Averages for MEPs in each category.

It appears that anti-establishment parties devote more of their agendas to the core issue of law. Difference of means testing reveals that this is a statistically significantly higher proportion compared to either the governing parties, or the establishment opposition parties. They also devote more time to government operations than establishment opposition parties. They also statistically significantly devote less time to international relations than the establishment opposition parties (significant at p < 0.05). Figure 4.8 displays anti-establishment issue attention according to left and right-wing orientation.

120 Figure 4.8: European Parliament. Disaggregated Core Issue Attention of Left-Wing and Right- Wing Anti-Establishment Parties

Note: Proportion of an anti-Establishment MEP’s agenda devoted to core issues.

Table 4.9: Non-Core Issue Attention in the European Parliament

Governing Parties Establishment Opposition Anti-Establishment Civil Rights 9.6% 9.9% 6.3% Health 6.4% 9.9% 11.2% Agriculture 21.7% 18.2% 19.6% Labor 7.6% 8.3% 6.3% Education 0.6% 3.4% 3.1% Environment 10.2% 12% 13% Energy 6.4% 4.2% 2.7% Immigration 12.8% 12.5% 15.18% Transportation 3.2% 5.7% 6.7% Social Welfare 1.2% 0.5% 0.5% Housing 1.3% 1% 1.8% Technology 10.3% 5.2% 5.4% Trade 7.6% 6.3% 6.3% Public Lands 0% 1.6% 1.3% Culture 0.6% 1% 0.5%

121 When comparing left and right wing anti-establishment parties, left-wing anti-establishment parties appear to devote more of their agendas to international affairs, law, and defense than their right-wing counterparts. This is somewhat surprising given right-wing parties’ broad ownership of these types of issues (Budge, Ezrow, and McDonald 2010). Difference of means testing reveals that these are statistically significant differences at the p < 0.01 level. Examining non-core issues reveals other interesting patterns. First, all types of parties in the European Parliament appear to devote a substantial portion of their agendas to agriculture. As they do to immigration. This is not surprising, given the saliency of the recent refugee crisis. Anti- establishment parties devote more of their questions to immigration, but this is not a statistically significant difference. Anti-establishment parties also devote less attention to civil rights than do establishment opposition parties, and this is a statistically significant difference as well (p < 0.05).

4.4.2 Italian Parliament

Next, I examine these relationships within the context of a domestic parliament. The Italian political environment is no stranger to political upheaval and volatility. In 1994, the country elected a political outsider, Silvio Berlusconi, and his new party, Forza Italia, into parliament. The party was devoted to replacing the traditional, and corrupt Italian parties, many of whom had been implicated in corruption charges just a few years earlier (Vannucci 2009b). Over the course of the next ten years, Italian politics saw the entrance and exit of many parties. The most recent anti-establishment contender is the Five-Star Movement (M5S), that went from a largely online political movement in 2012, to the third largest political group in parliament in 2013, to co-leader of the governing coalition in 2018. Italian politics within the Chamber of Deputies is governed by numerous rules. Individual members may submit questions to the Premier and the governing party in the form of parliamentary questions. Like many countries in Western Europe, party discipline is controlled and members’ questions broadly reﬂect the agenda of the party. However, Italy also experiences party switching quite frequently, where members defect from more tightly disciplined parties (Heller and Mershon 2008). Individual members are making the decision to submit questions, and while these decisions are largely decided by partisan concerns, Italian MPs may experience less stringent partisan control over their legislative activities.

122 Table 4.10 presents the results from two negative binomial regressions, with and without the interactive term between anti-establishment score and ideology. Figure 4.9 presents the predicted number of questions graphically. In Italy, it appears as though a party’s anti-establishment score increases the number of questions asked, but this is not a statistically significant relationship. Ideology is statistically significant in Italy, and in the non-interactive model, right-leaning ideology increases the number of questions asked by a MP. This effect is substantively large as well, going from approximately one question at the most left-leaning ideological score, to roughly one-hundred forty-six questions at the most right-leaning ideological score.

Table 4.10: Italian Parliament. Inﬂuence of Anti-Establishment Score on Number of Parliament Questions Asked

(13) (14) PQs PQs

Anti-Establishment 0.0182 0.189 Score (0.0347) (0.122) Ideology 0.493*** 0.555*** (0.0925) (0.128) AE Score x Ideology -0.0311 (0.0224) Female -0.222 -0.197 (0.165) (0.168) Member of Governing 0.277*** 0.280*** Coalition (0.0894) (0.0884) Seat Share -3.335*** -3.437*** (0.602) (0.608) Constant -0.177 -0.483 (0.595) (0.723)

Observations 4,529 4,529

Note: Robust standard errors clustered by MP in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Session ﬁxed eﬀects included.

123 (a) AE Scores (b) Ideology

Figure 4.9: Italian Parliament. Predicted Number of Questions Across Anti-Establishment Scores and Ideology

Note: Predicted number of questions with 95% conﬁdence intervals. Graph A displays the predicted number of questions across anti-establishment scores for Model 13. Graph B displays the predicted number of questions across RILE scores for Model 13. All control variables held at their mean values.

Table 4.11: Italian Parliament. Anti-Establishment Members versus Establishment Opposition Members

(15) (16) PQs PQs

Anti-Establishment 0.394 -1.434 (0.272) (1.015) Ideology 0.0782 0.00445 (0.0615) (0.0839) Anti-Establishment x Ideology 0.311** (0.154) Female 0.0291 -0.00302 (0.159) (0.151) Member of Governing Coalition -0.419** -0.474** (0.199) (0.187) Share of Seats in Domestic Parliament -3.471*** -3.113*** (0.550) (0.549) Constant 1.951*** 2.317*** (0.400) (.510)

Observations 2,889 2,889

Note: Robust standard errors clustered by MP in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Session ﬁxed eﬀects included.

124 Table 4.11 assesses Italian MP’s level of activity, comparing anti-establishment members and establishment opposition members. In Model 13, the effect is in the expected direction, though it is not statistically significant. In Model 14, the interaction between anti-establishment status and ideology is statistically significant at the p < .10 level. Counter to the findings from the European Parliament, right-wing ideology has an enhancing influence on anti-establishment members’ level of activity. Figure 4.10 displays the results graphically.

(a) Non-Interactive

(b) Marginal Eﬀect of Ideology

Figure 4.10: Italian Parliament. Predicted Number of Questions Anti-Establishment Parties versus Establishment Opposition

Note: Predicted number of questions with 95% conﬁdence intervals. All control variables held at their mean values.

125 Figure 4.11: Italian Parliament. Predicted Number of Questions by Anti-Establishment Status and Ideology

Note: Predicted Number of Questions from Model 15, with 95% conﬁdence intervals. All other variables held at their mean values.

To further assess the diﬀerences between left-wing and right-wing anti-establishment and establishment opposition members, I display Figure 4.11. The results in Italy appear opposite of the results from the European Parliament. Here, right-wing anti-establishment MPs are the most active, whereas left-wing establishment opposition MPs are the least active, and no discernible diﬀerence exists between left-wing anti-establishment MPs and those members part of right-wing establishment opposition parties. Italian parties are volatile and there is a tendency in this time period for anti-establishment parties of both ideological leanings to be part of the the governing coalition. For example, Lega Nord, Forza Italia, and Rifondazione Comunista (the Communist Re- foundation Party) have all been part of governing coalitions and are considered anti-establishment for certain periods of time in this analysis. Next, I assess the attention to core issues in the Italian Parliament. Table 4.12 displays the results of models assessing the probability of asking a question pertaining to a core issue. Models 17 and 18 examine the relationship between anti-establishment score, and the interaction of anti- establishment score and ideology, respectively. Figure 4.12 displays the predicted probabilities.

126 Table 4.12: Italian Parliament.Eﬀect on Probability of Asking a Question Devoted to a Core Topic

(17) (18) Focus on a Core Issue Focus on a Core Issue

Anti-Establishment Score 0.0567*** 0.142** (0.0133) (0.0584) Ideology 0.0423 0.0733* (0.0269) (0.0374) Anti-Establishment Score x Ideology -0.0157 (0.0104) Female -0.0836 -0.0763 (0.0587) (0.0584) Member of the Governing Coalition 0.00904 0.0137 (0.0485) (0.0493) Share of Seats -0.0214 -0.0796 (0.170) (0.180) Constant -0.502*** -0.655*** (0.179) (0.224)

Observations 4,524 4,524

Note: Effect on probability of of asking a question on a core issue, across anti-establishment scores.Robust standard errors clustered by MP in parentheses.*** p<0.01, ** p<0.05, * p<0.1. Session fixed effects included.

(a) Non-Interactive (b) Marginal Eﬀect of Ideology

Figure 4.12: Italian Parliament. Probability of Asking Question Pertaining to a Core Issue Across Anti-Establishment Scores

Note: Predicted probabilities across anti-establishment scores with 95% conﬁdence intervals. All control variables held at their mean values.

127 Anti-establishment score is positive and statistically significant in both models. The interaction between anti-establishment score and ideology is not statistically significant, though it does reflect a moderating influence of right-wing ideology, as observed in the European Parliament. According to Model 17, the probability of an Italian MP asking a question at the lowest level of anti-establishment sentiment is 45%, and rises to 63% at the highest level anti-establishment sentiment. This is suggestive evidence that anti-establishment sentiment increases attention to core issues, as predicted in Hypothesis 2. Table 4.13 displays results comparing anti-establishment MPs to establishment opposition MPs.

Table 4.13: Italian Parliament. Issue Prioritization. Eﬀects on the Probability of Asking a Question Devoted to a Core Issue: Anti-Establishment versus Establishment Opposition

(19) (20) Focus on a Core Issue Focus on a Core Issue

Anti-Establishment 0.146* 1.287*** (0.0866) (0.435) Ideology 0.00879 0.0610* (0.0290) (0.0340) Anti-Establishment x Rile -0.194*** (0.0722) Female -0.0962 -0.0816 (0.0672) (0.0679) Member of Governing Coalition -0.281*** -0.250** (0.102) (0.102) Share of Seats 0.777*** 0.560** (0.253) (0.269) Constant -0.212 -0.484** (0.185) (0.206)

Observations 2,885 2,885

Note: Robust standard errors clustered by MP in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Session ﬁxed eﬀects included.

128 (a) Non-Interactive (b) Marginal Eﬀect of Ideology

Figure 4.13: Italian Parliament. Marginal Eﬀect of Ideology on the Probability of Asking a Question Pertaining to a Core Issue Anti-Establishment Parties versus Establishment Opposition

Note: Eﬀects on predicted probabilities with 95% conﬁdence intervals. All control variables held at their mean values.

Figure 4.14: Italian Parliament. Predicted Probability of Anti-Establishment Members Asking a Question on a Core Issue-Across Ideology

Note: Predicted probabilities with 95% conﬁdence intervals. All other variables held at their mean values.

129 Being a member of an anti-establishment party increases the probability of asking questions pertaining to a core issue by roughly 10%. Right-wing ideology has a moderating effect on this probability. Figure 4.14 displays the predicted probabilities across RILE scores. Extreme right-wing anti-establishment members only have a 28% probability of asking a question pertaining to a core issue, whereas centrist anti-establishment members have a 54% probability of asking on a core issue, and extreme left-wing members have a 73% probability of doing so. Table 4.14 analyzes issue attention using proportion of members’ agendas devoted to core issues as the dependent variable. Figure 4.15 displays the predicted proportion of members’ agendas. In both Models 21 and 22, increases in a member’s anti-establishment score correspond to a 2.1% increase in the proportion of a member’s agenda devoted to core issues. The interaction between anti-establishment score and ideology is not statistically significant, though the coefficient does indicate a moderating influence of right-wing ideology on anti-establishment members’ agendas.

Table 4.14: Italian Parliament. Issue Prioritization. Proportion of Questions Devoted to Core Issues Across Anti-Establishment Scores

(21) (22) Focus on Core Issues Focus on Core Issues

Anti-Establishment Score 0.0207*** 0.0494** (0.00479) (0.0229) Ideology 0.00993 0.0203 (0.0103) (0.0142) Anti-Establishment Score x Ideology -0.00528 (0.00401) Female -0.0324 -0.0300 (0.0216) (0.0216) Member of Governing Coalition -0.0174 -0.0158 (0.0153) (0.0155) Seat Share 0.0452 0.0257 (0.0604) (0.0633) Constant 0.289*** 0.237*** (0.0676) (0.0849)

Observations 4,528 4,528 R-squared 0.029 0.031

Note: Robust standard errors clustered by MP in parentheses.*** p<0.01, ** p<0.05, * p<0.1. Session ﬁxed eﬀects included.

130 (a) Non-Interactive

(b) Marginal Eﬀect of Ideology

Figure 4.15: Marginal Eﬀect of Ideology on the Proportion of Agenda Devoted to Core Issues Across Anti-Establishment Scores

Note: Conditional marginal eﬀect of ideology on the proportion of a MP’s agenda focused on core issues with 95% conﬁdence intervals.

The predicted proportion of a member’s agenda devoted to core issues ranges from just under 40% at the lowest level of anti-establishment sentiment to just under 60% at the highest level of anti-establishment sentiment. Table 4.15 displays results comparing anti-establishment members to establishment opposition members. Here, the results from the interactive model (Model 24) are statistically signiﬁcant. Right-wing ideology moderates the inﬂuence of anti-establishment membership on the proportion of a member’s agenda focused on core issues. Figure 4.16 displays the results graphically.

131 Table 4.15: Italian Parliament. Issue Prioritization. Proportion of Questions Devoted to Core Issues. Anti-Establishment Parties versus Establishment Opposition

(23) (24) Focus on Core Issues Focus on Core Issues

Anti-Establishment 0.0410 0.335** (0.0291) (0.157) Ideology -0.00492 0.00863 (0.0102) (0.0113) Anti-Establishment x Ideology -0.0499* (0.0263) Female -0.0387 -0.0349 (0.0258) (0.0262) Member of Governing Coalition -0.0998*** -0.0916*** (0.0352) (0.0355) Seat Share 0.275*** 0.219** (0.0924) (0.0956) Constant 0.415*** 0.344*** (0.0641) (0.0677)

Observations 2,889 2,889 R-squared 0.022 0.027

Note: Robust standard errors clustered by MP in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Session ﬁxed eﬀects included.

(a) Non-Interactive (b) Marginal Eﬀect of Ideology

Figure 4.16: Proportion of Agenda Devoted to Core Issues. Anti-Establishment Parties versus Establishment Opposition

Note: Conditional marginal eﬀect of ideology on the proportion of a MP’s agenda focused on core issues with 95% conﬁdence intervals.

132 Moving from the establishment opposition to an anti-establishment party increases the proportion of a member’s agenda devoted to core issues by approximately 10%. Once again, right-wing ideology tempers this eﬀect. Figure 4.17 displays the predicted proportions for anti-establishment members across ideology.

Figure 4.17: Proportion of Anti-Establishment Members’ Agendas Devoted to Core Issues - Across Ideology

Note: Predicted proportions with 95% conﬁdence intervals. All other variables held at their mean values.

Extreme right-wing anti-establishment members are predicted to devote roughly 26% of their agendas to core issues, while centrist anti-establishment members are predicted to devote 46% of their agendas to core issues, and extreme left-wing anti-establishment members are predicted to devote 63% of their agendas to such issues. Further research is needed to tease out this relationship between right-wing ideology, anti-establishment membership, and legislative behavior. For now, evidence suggests that there is a diﬀerence between being anti-establishment and ideology, and that both attributes may inﬂuence politicians’ behavior. Finally, I disaggregate the core issue variable into its component parts and compare issue attention across three groups: governing parties, establishment opposition parties, and anti-establishment

133 parties. Table 4.16 displays the disaggregated core issues, and Table 4.17 displays all other agenda areas.

Table 4.16: Core Issue Attention in the Italian Parliament

Governing Parties Establishment Opposition Anti-Establishment Macro-Economics 17.9% 14.6% 17.1% Law & Crime 29.3% 33.5% 27.6% Finance 14.3% 11.8% 13% International Relations 11.2% 8.1% 8.1% Defense 5.7% 5.4% 6.1% Government Operations 21.6% 26.5% 28.1%

Note: Average proportions across groups.

Anti-Establishment members of the Italian Parliament devote a higher proportion of their agenda to macro-economics than establishment opposition members, this a statistically signiﬁ- cant diﬀerence (p < 0.10). They also devote a proportion (p < 0.01) of their agenda to government operations than governing parties. Figure 4.18 displays the disaggregated core issues variable for anti-establishment members, divided by left-wing and right-wing party membership.

Figure 4.18: Italian Parliament. Disaggregated Core Issue Attention Left-Wing and Right-Wing Anti-Establishment Parties

Note: Proportion of an Anti-Establishment MP’s agenda devoted to core issues.

134 Left-wing anti-establishment members devote a statistically significantly higher proportion of their agendas to macro-economics than right-wing members (p < 0.01). Right-wing anti- establishment members devote a higher proportion of their agendas to defense and international affairs, significant at the p < 0.05 level, than left-wing anti-establishment members. In this respect, it does appear that anti-establishment parties do attempt to co-opt agenda areas owned by ideologically similar establishment parties.

Table 4.17: Non-Core Issue Attention in the Italian Parliament

Governing Parties Establishment Opposition Anti-Establishment Civil Rights 5.2% 5% 3.4% Health 16.8% 11% 16% Agriculture 7.6% 6% 7.8% Labor 11.3% 10.1% 13.2% Education 8.6% 10.4% 9.1% Environment 7.9% 7.3% 8.4% Energy 4.2% 4.7% 3% Immigration 5.8% 6.5% 6.1% Transportation 18.1% 20.1% 17.3% Social Welfare 3.2% 5.5% 3.5% Housing 2.7% 2.8% 2.4% Technology 3% 4.2% 2.3% Trade 1.3& 1% 1.4% Public Lands 1.7% 2.4% 2.1% Culture 2.8% 3.1% 3.7%

Note: Average proportions across groups.

When examining non-core issues, anti-establishment members devote a statistically signiﬁcantly higher proportion of their agendas to health and labor (p < .05) than establishment opposition members.

4.5 Discussion

How does membership in an anti-establishment party influence members’ behavior? It appears that results are mixed. In the context of the European Parliament, increases in parties’ anti- establishment scores increase the number of questions asked, though right-wing ideology has a moderating effect on this relationship. This may not be the case in the Italian Parliament, where anti-establishment sentiment does not appear to have a statistically significant influence on the level of a member’s activity. Further, there are disparate effects in the two bodies with regard to issue

135 attention. In the European Parliament, anti-establishment sentiment does not have a statistically significant influence on core issue attention. Whereas in the Italian parliament, anti-establishment status increases attention to core issues. Though I failed to find statistically significant support for the notion that anti-establishment sentiment increases core issue attention in the European Parliament, disaggregated core issues did yield some interesting patterns, such as left-wing anti-establishment MEPs’ attention to traditionally owned right-wing issues such as defense and international affairs. Further, interesting patterns emerge from the analysis of the Italian parliament, such as the increased level of activity of right- wing anti-establishment politicians as compared to left-wing anti-establishment members. However, right-wing anti-establishment Italian MPs are considerably less likely to devote their agendas to core issues, as compared to left-wing anti-establishment members. Additional research could explore the differences among anti-establishment parties according to their ideological orientation. Further research should also explore the differences between politicians’ behavior in European parliament and within the context of domestic parliaments. Perhaps these parties, given their general eurosceptic orientation, feel emboldened in the European parliament, whereas in a domestic context, they may be uncertain of their role. Given that many of these parties are new to parliaments, and their leaders new to politics, it may be that their domestic agendas are still unconstrained and unclear. This analysis should be the first step in assessing the behaviors of these unique parties. Further research in the parliamentary activities of anti-establishment parties should also incorporate more detailed individual level data on members of parliaments. A members’ gender, years of service, and political experiences have been shown to influence legislative behavior. While data availability precluded these things being considered fully in this analysis, future work incorporating these factors will provide additional insight into anti-establishment legislative behavior. Also, future research should also examine how these parties influence the agendas of mainstream parties. How does the entrance of an ideologically similar anti-establishment party impact a traditional party? The importance of understanding anti-establishment parliamentary activities cannot be over- stated. Scholarship has devoted considerable attention to how these parties cultivate support. While an important question, the focus of scholarship should shift from how these parties win votes, to how they approach governance once in office. Anti-establishment parties are winning in-

136 creasing vote shares in their respective domestic parliaments. Recently, Alternative for Deutschland (AfD) won the the third largest seat share in the German Bundestag. As the largest opposition party in parliament, they stand to inﬂuence policy dramatically. This is particularly important in the case of Italy, where an unholy union between a radical-right and xenophobic party (Lega Nord), and an ideologically unconstrained and politically inexperienced party (M5S) is set to govern Italy in the wake of the 2018 election. How this new “Italy First” coalition will impact domestic Italian politics is yet to be determined, but signs indicate that policies could veer drastically from the status quo. These parties are noted for behaving as obstructionist noise-makers within the context of domestic parliaments. They rail against the political status quo, claiming their superior ability to represent the people, and govern eﬀectively and corruption-free. It is an easy claim to make while outside of the political arena. Now that these parties are present in parliaments, what they do is incredibly important to the stability of their domestic governments.

137 CHAPTER 5

LEFT, RIGHT, AND EVERYTHING IN BETWEEN: FINAL THOUGHTS ON THE ANTI-ESTABLISHMENT MOVEMENT

The emergence and electoral success of anti-establishment parties has consumed the attention of scholars and pundits alike. Academic discussion has revolved around the success and agendas of these parties, yet defining the anti-establishment movement remains a convoluted and difficult task. This dissertation is meant to serve as a step to clarify the anti-establishment movement, providing a definition of an anti-establishment party, and proposing a classification scheme through which to better understand the movement, and its consequences. First, I presented a definition of the anti-establishment that focuses on their anti-elite sentiment, and have provided a systematic way of classifying these parties that is separate and distinct from ideology. I have also presented an additional explanation for the source of their electoral viability: low diffuse support, or institutional trust. Finally, I demonstrated some interesting patterns in the legislative behavior of these parties’ members. This dissertation provides an initial step in conceiving these parties as a distinct movement, and offers several pathways of future research. First, I provided a cohesive definition of the anti-establishment movement. Anti-establishment sentiment consists of a deep-seated mistrust of political elites, who are alleged to have hurled the country into a crisis. In this, I have attempted to clarify the populist nature of the anti- establishment movement. Instead of relying on vague references to the use movement’s use of populism, I instead identify this movement as “vertically populist” (Hartleb 2015). These parties draw a distinction between the people and the political elite. Anti-establishment politicians assert that the only way to remedy such a crisis is to use the tools of direct democracy to depart from a political status quo that is damaging the nation. Finally, those in the anti-establishment movement promote a nation-first attitude, where they seek to restore national glory, either through exclusive or inclusive nationalism.

138 Second, in Chapter Two, I demonstrated that these parties can and should be measured along a dimension that is separate and distinct from ideology. These parties appear everywhere along the traditional left-right spectrum. Anti-establishment sentiment should not be solely under the purview of the radical right, and, as such, it is a mistake to classify all such parties as radical right- wing. Further, I argued that anti-establishment parties should be measured along a continuum, demonstrating that a binary classification of the movement does not adequately capture the nuances of these parties’ messaging. The measure is a needed addition to the literature on anti-establishment parties. First, it helps clarify the distinction between anti-establishment parties and the radical right. Second, it provides a systematic way to organize and classify these parties. Finally, it permits the analysis of these parties over time, whereby scholars can assess parties’ development across elections. While the measure does indeed provide an important addition to the literature, it could benefit from additional data and more frequent updating. Because it is based on manifesto data, its updating is only as frequent as new elections. However, politicians constantly communicate to their constituents through press releases, speeches, etc. For example, the recent electoral success of the M5S in Italy is largely credited to the party’s nearly exclusive use of social media in its campaign.1 These parties are not only unconventional with regard to their messaging and outsider status, but they are also running unconventional campaigns. An updated measure that incorporates this approach to party rhetoric and messaging would also prove beneficial. Third, I demonstrated an additional explanation for these parties’ source of electoral support: low diffuse support, or institutional trust. These parties flourish in a political environment wrought with declining political trust. While attitudes toward certain policy areas, such as immigration and economics, tap into the deep-seated cultural and economic anxieties expressed by these parties’ supporters, these factors do not explain support for anti-establishment parties across the ideological spectrum. Depending on the other contextual factors, a variety of explanations, including declining diffuse support can help explain voting for anti-establishment parties. In the cases of the European Parliament, the Netherlands, Italy, and Germany, low diffuse support predicted vote choice for an anti-establishment party, whether that party existed on the left or right. This held when considering traditional explanations of radical right-wing support, such as economic motivations (i.e. unemployment), or cultural misgivings (i.e. immigration fears).

1http://www.euvisions.eu/italian-elections-twitter/

139 In fact, in the case of the left-wing Italian M5S, these traditional factors of right-wing support did not account for supporting the left-wing anti-establishment party. This is suggestive evidence that theories constructed with the radical right in mind are insufficient in accounting for anti- establishment support across the ideological spectrum. Further, in 2013 Germany, these factors also did not appear to dramatically influence left or right-wing anti-establishment support in party list voting. Trust had mixed results in the French case, with increased trust predicting votes for Macron, and decreased trust predicting votes for Le Pen in the second round of voting. While the evidence did not bear out in France, the unique nature of the 2017 Presidential election (Evans and Ivaldi 2018) may warrant additional analyses of how a centrist anti-establishment party achieves success. However, evidence suggests that low diffuse support can explain support for anti-establishment parties from the left and the right. The evidence from Italy and the Netherlands helps lend credible support to this theory. In both countries, individuals possessing low diffuse support and low specific support were more likely to vote for anti-establishment parties (the PVV and the M5S) than the incumbent or establishment opposition. However, if an individual was high in diffuse support but evaluated the government poorly (low specific support), there were more likely to select the establishment opposition. This is evidence that each type of evaluation (diffuse and specific support) plays a unique role in voters’ selection of anti-establishment parties from across the ideological spectrum. Finally, I demonstrated that anti-establishment sentiment has implications for parliamentary behavior. The level of parliamentary activity and the agendas of anti-establishment politicians has not been explicitly examined in the context of the current anti-establishment movement. I have shown that increases in anti-establishment sentiment can indeed impact the behavior of politicians. Results in Chapter Four were mixed, but interesting patterns of legislative behavior should be noted. In the context of the European Parliament, anti-establishment sentiment increased the number of parliamentary questions asked by a MEP, compared to parties who hold domestic governing positions and those who represent domestic establishment opposition parties. In Italy, this did not bear out, though the effects were in the expected direction. However, when assessing issue attention, it appears as though, at least in the Italian Parliament, anti-establishment status does indeed increase attention to core issues. Further research is needed to tease out various effects, including the influence of being a right or left-wing anti-establishment member.

140 For now, it appears as though the ideological leaning of an anti-establishment party has the potential to influence members’ behavior. In the European Parliament, right-wing ideology had a moderating effect on the number of parliamentary questions an anti-establishment member asked. In the Italian parliament, it appears as though it enhanced members’ core issue attention, particularly among issues such as defense and international affairs, traditionally owned by the right. In this context, anti-establishment parties do indeed attempt to co-opt mainstream issues as part of their agenda in an effort to steal away issue ownership and promote their own competence. This dissertation opens the possibility for further research on the anti-establishment movement and its consequences. Using the measure presented in Chapter Two, scholars can explore several questions. One use of the measure is to examine how parties shift in their rhetoric over time. While it was observed and acknowledged in Chapter Two that parties do indeed vary over time in their level of anti-establishment sentiment, further research could attempt to explain the reasons for these changes. It is possible that the longer these parties serve in parliament, the more likely they are to moderate their anti-establishment rhetoric. Feeling compelled to work with other parties, or attain ministerial positions, these parties may temper their anti-mainstream sentiment, as has been observed with certain right-wing parties (e.g. Lega Nord) (Albertazzi and McDonnell 2008). What other reasons exist for the shifts in a party’s anti-establishment sentiment? It could be that actually governing or being a member of a governing coalition softens anti-establishment sentiment, as it is hard to critique the political elite once you are part of it. This leads to a larger question of when and how parties shift from anti-establishment to mainstream status. I argued extensively in Chapter One, that a dichotomous approach to the categorization of parties ignores the nuances of parties’ orientation to politics. However, scholars and others are compelled to think of parties as occupying one particular part of the spectrum. As these parties gain electoral strength and begin to occupy government positions, rallying against the political elite may not come across as credible. In other words, how does a party claim to be anti-elite, anti-mainstream, i.e. anti- establishment, when they are part of the government, and therefore themselves part of the political elite? This could depend on several factors. For instance, there may be a fundamental difference between anti-establishment parties who arise from outside of politics, formed by non-politicians (M5S) from those who are made up of defectors from mainstream parties (Macron). In the first

141 case, we may observe radical changes from the status quo and unconstrained ideological positioning. Agendas stemming from these parties may be malleable. In the latter example, agendas may be more stable, and the transition from candidate to policy maker may be smoother than what previous non-politicians experience. This raises another area to consider in future research: the electoral benefit of the anti- establishment label. Does anti-establishment rhetoric increase a party’s chances of winning? This is most likely dependent on the political environment in which these parties campaign and run for office. One could imagine that in the context of a country that routinely struggles with corruption and scandal (e.g. Italy), anti-establishment rhetoric would have great appeal. Further institutional research could examine which political environments are most friendly to anti-establishment parties. While extant research often focuses on why individuals prefer these and other potentially extreme parties, future work could examine these parties’ successes in the context of a crisis, scandal, or other institutional factors. Supply-side studies of the radical right’s popularity could and should be updated to account for the popularity of anti-establishment parties from across the ideological spectrum. Further, how much does the entrance of an anti-establishment party in the political environment influence the orientation and agendas of mainstream and traditional parties? For example, the Danish People’s Party lends to support to the minority Danish government. They have been successful in pushing legislation that targets Jewish and Muslim minorities’ religious practices, such as dietary restrictions and circumcision. Additional areas of research could focus on the anti-establishment parties’ influence over existing institutions. Beyond examining legislative agendas, which I have done to some degree in Chapter Four, further work would elaborate on this topic. Research could explore bill introductions and legislative effectiveness. Further, additional research could explore how different demographic groups, such as women, behave within the context of an anti-establishment party. Additionally, how do these parties impact broader institutional stability? Does the inclusion of an anti-establishment party in parliament or a governing coalition hasten the government’s collapse? Further, research could explore how anti-establishment parties influence foreign policy and membership in international organizations, (e.g. UKIP’s influence in the Brexit vote). Finally, this line of research could easily extend beyond Europe. While the anti-establishment wave has been salient on the continent, we can observe the popularity of these parties and candi-

142 dates elsewhere. For example, in the 2016 United States Presidential contest, establishment and “anti-establishment” candidates competed in both the party primaries and in the general election. During the primary contests, anti-establishment or outsider candidates in both parties challenged status quo politicians. Donald Trump and Bernie Sanders stood in opposition to what they alleged was a tired, out of touch, and corrupt slate of mainstream candidates. In the U.S. general election, Donald Trump, a political outsider, campaigned against a distinctly establishment candidate, Hillary Clinton. Anti-establishment sentiment is also recently observed in Mexico. AndrésManuel López Obrador, representing the “Together We’ll Make History” party, has been credited by news outlets as “striking a populist chord” among the Mexican electorate.2 Obrador, a career politician, is challenging mainstream Mexican parties and politics, riding a wave of anti-establishment sentiment that is both left-wing in its ideological orientation, and anti-elite in its messaging. Because the Comparative Manifesto Data is cross-national, any party included in the database can have a calculated anti-establishment score. There is no reason that this line of research cannot extend beyond Europe. Scholarship can examine this important phenomenon and develop general theories that apply to anti-establishment parties cross-nationally. The implications of this line of research are numerous. Anti-establishment parties stand to dramatically alter the trajectories of their domestic governments. These parties represent a complex movement that ideological position cannot fully explain. Does the term anti-establishment even begin to capture the complexities of these parties’ orientation and behaviors? I think not. As illustrated in the previous chapters, these parties stem from a variety of backgrounds and espouse unconstrained agendas. While the common thread of anti-corruption, crisis, and nationalism begins to characterize the movement, the term “anti-establishment” itself may be misleading. These parties seek not only to buck the system, but additionally, to control it, replacing the current political elite. Many of these parties are shockingly and alarmingly unconstrained in their ideological orientation. The Italian M5S demonstrates this particular reality. While distinctly an economic left party when it first entered Italian politics, its agenda has seemingly veered to the right. This shift has been attributed to the party’s alliance with the xenophobic Lega Nord. However, the left-right

2See NPR-https://www.npr.org/2018/06/04/616917883/populist-message-appears-to-be-striking-a-cord-with- mexicans-during-presidential

143 orientation of the party should not be viewed as its most important aspect. Instead, scholarship should focus on its, and other parties,’ use of rhetoric that aims to drastically change the political status quo using tools of direct democracy. In the case of Italy, this could mean a further disconnec- tion from the European Union. While the Italian constitution requires an amendment to remove the country from the Euro, the continuing popularity of the M5S and Lega Nord could result in enough support to make this otherwise challenging reform a distinct possibility. In the context of Europe, anti-establishment parties’ rise in domestic governments could pro- foundly impact European Union stability. The recent Brexit vote demonstrates the reality of this instability. As the United Kingdom prepares to exit the Union, other countries are similarly influenced. Further, many of these parties are skeptical of increases in immigration. The AfD’s significant electoral gains and rise in popularity should be alarming to all, as the party espouses the white-washing of Europe, and a return to German nationalism. These are the policy consequences of these parties’ rise in prominence in domestic parliaments. However, the largest implication has to do with their source of support. As individuals become increasingly distrusting of political elites and parties, these anti-establishment groups will only become more popular. As people in democratic societies are increasingly disaffected, this can only spell increases in support for potentially dangerous and provocative political movements.

144 APPENDIX A

ANTI-ESTABLISHMENT SCORES

A.1 All European Countries

Table A.1: Full List of European Parties and Anti-Establishment Scores for the Most Current Manifesto Year-European Union Countries Plus Norway, Switzerland, and Iceland

Country Party Year Anti-Establishment Score United Kingdom Green (UK) 2015 6.417469 United Kingdom UKIP 2015 8.610013 United Kingdom SDLP (UK) 2015 7.07102 United Kingdom Sinn Fein 2015 1.192117 United Kingdom Plaid Cymru 2015 1.098695 United Kingdom Labour (UK) 2015 3.213324 United Kingdom UUP (UK) 2015 2.750751 United Kingdom DUP (UK) 2015 10 United Kingdom Liberal Democrats (UK) 2015 1.524127 United Kingdom Conservative (UK) 2015 3.592563 United Kingdom SNP (UK) 2015 .9999985 Italy RC 2013 10 Italy Autonomy Progress Aosta Valley 2013 1 Italy Lega Nord (ITA) 2013 4.25542 Italy PdL 2013 4.25542 Italy PD (ITA) 2013 2.7424 Italy SEL 2013 3.713893 Italy Civic Choice 2013 4.570474 Italy M5S (ITA) 2013 8.237235 Italy UdC 2013 5.960831 Italy SVP 2013 1.802234 Italy CD 2013 1.689636 Italy Brothers of Italy 2013 4.351317 Italy Labour and Freedom List 2013 3.455822 Spain EAJ (PNV)-Basque 2016 1.154277 Spain PP (ESP) 2016 2.132394 Spain ERC: Catalan 2016 3.087706 Spain En Marea (Galicia) 2016 2.362255 Spain United Podemos Coalition 2016 3.059725 Spain Podemos (Catalonia) 2016 1.661968 Spain CC-PNC 2016 1

145 Table A.1 Continued

Country Party Year Anti-Establishment Score Spain Ciudadanos 2016 3.014103 Spain Podemos 2016 2.756603 Spain PSOE (ESP) 2016 3.019901 Spain EHB(ESP) 2016 10 Spain Valencian Coalition (ESP) 2016 2.033001 Spain CDC(ESP) 2016 1.384678 Germany Greens(GER) 2017 1.927055 Germany CDU (GER) 2017 1.37521 Germany AfD (GER) 2017 10 Germany Die Linke 2017 2.535386 Germany SPD (GER) 2017 1.703756 Germany FDP (GER) 2017 1 France Greens(FRA) 2017 .9999996 France National Front (FRA) 2017 2.232702 France MoDem (FRA) 2017 9.026548 France UDI (FRA) 2017 2.677124 France PRG (FRA) 2017 1.917616 France En Marche! (FRA) 2017 10 France Socialists (FRA) 2017 8.788499 France LFI (FRA) 2017 4.718736 France UMP (FRA) 2017 2.677124 France Communist(FRA) 2017 1.111125 Netherlands CDA (Neth) 2012 2.183752 Netherlands SP 2012 3.103496 Netherlands 50PLUS 2012 6.462727 Netherlands D66 2012 2.074029 Netherlands Green Left 2012 1.916899 Netherlands VVD (Neth) 2012 1.911962 Netherlands PvdD 2012 3.060697 Netherlands CU 2012 1.10574 Netherlands SGP 2012 1.000001 Netherlands PvdA (Neth) 2012 2.326439 Netherlands PVV (Neth) 2012 9.999999 Austria Greens(AT) 2013 5.880282 Austria FPO (AT) 2013 1.519183 Austria NEOS (AT) 2013 5.481254 Austria OVP (AT) 2013 1.620654 Austria Team Stronach (AT) 2013 10 Austria SPO (AT) 2013 .9999997

146 Table A.1 Continued

Country Party Year Anti-Establishment Score Finland True Finns (FIN) 2011 10 Finland Left Wing (FIN) 2011 1.364602 Finland Swedish People’s Party (FIN) 2011 1.367743 Finland National Coalition (FIN) 2011 1.214679 Finland Social Democrats (FIN) 2011 .9999971 Finland Greens (FIN) 2011 2.442274 Finland Finnish Centre 2011 3.218756 Finland Christian Democrats (FIN) 2011 2.42602 Hungary Jobbik (HUN) 2014 3.604597 Hungary DK 2014 1.531755 Hungary Together 2014 2014 5.329805 Hungary FIDESZ 2014 1 Hungary MSzP (HUN) 2014 5.429105 Hungary LMP 2014 10 Sweden Center (SWE) 2014 1.000003 Sweden Greens (SWE) 2014 1.694384 Sweden SAP (SWE) 2014 1.553982 Sweden Left Party (SWE) 2014 1.000003 Sweden Liberals (SWE) 2014 1.63986 Sweden KD (SWE) 2014 2.122449 Sweden Moderate (SWE) 2014 1.269486 Sweden Swedish Dem (SWE) 2014 9.999995 Greece ANEL (GRE) 2015 2.576534 Greece SYRIZA (GRE) 2015 6.542151 Greece LAE (GRE) 2015 1 Greece PASOK (GRE) 2015 1.858502 Greece EK (GRE) 2015 9.999998 Greece Golden Dawn (GRE) 2015 10 Greece ANEL (GRE) 2015 6.34 Greece POTAMI (GRE) 2015 5.67 Greece Communist (GRE) 2015 4.931148 Greece Golden Dawn (GRE) 2015 5.226047 Greece POTAMI (GRE) 2015 5.686853 Greece PASOK (GRE) 2015 1.406616 Ireland Familiy of the Irish 2016 1.106196 Ireland Workers and Unemployment Action 2016 4.073951 Ireland Independent Alliance 2016 9.999999 Ireland We Ourselves 2016 1.846439 Ireland Green Party 2016 1.479204 Ireland People Before Proﬁt Alliance 2016 1.967765

147 Table A.1 Continued Country Party Year Anti-Establishment Score Ireland Social Democrats 2016 2.859584 Ireland Labour Party 2016 1.1717 Ireland Soldiers of Destiny 2016 1.034276 Ireland Anti-Austerity Alliance 2016 1 Denmark Danish Social-Liberal Party 2011 1.63746 Denmark Liberal Alliance 2011 1.000003 Denmark Liberals 2011 1.469652 Denmark Danish Peoples Party 2011 7.065014 Denmark Red-Green Unity List 2011 5.669547 Denmark Conservative Peoples Party 2011 10 Denmark Socialist Peoples Party 2011 1.66014 Denmark Social Democratic Party 2011 3.697573 Portugal Social Democratic Party 2011 1.859643 Portugal Ecologist Party The Greens’ 2011 .9999998 Portugal Social Democratic Center-Popular Party 2011 3.824241 Portugal Left Bloc 2011 10 Portugal Portuguese Communist Party 2011 4.279967 Portugal Socialist Party 2011 2.043855 Switzerland Geneva Citizens’ Movement 2011 2.53854 Switzerland Green Liberal Party 2011 1.952778 Switzerland Swiss Peoples Party 2011 9.999998 Switzerland FDP.The Liberals 2011 4.494187 Switzerland Social Democratic Party of Switzerland 2011 1.000001 Switzerland Christian Social Party 2011 1.000001 Switzerland Green Party of Switzerland 2011 2.216552 Switzerland Christian Democratic Peoples Party of Switzerland 2011 1.519267 Switzerland Protestant Peoples Party of Switzerland 2011 1.000001 Switzerland Conservative Democratic Party of Switzerland 2011 2.996733 Belgium New Flemish Alliance 2010 1 Belgium Christian Democratic and Flemish 2010 1.882539 Belgium Christian Social Party 2010 3.805836 Belgium Open Flemish Liberals and Democrats 2010 5.027481 Belgium List Dedecker 2010 10 Belgium Socialist Party Diﬀerent 2010 7.046293 Belgium Reform Movement 2010 4.224425 Belgium Ecologists 2010 9.002278 Belgium Francophone Socialist Party 2010 1.424673

148 Table A.1 Continued Country Party Year Anti-Establishment Score Belgium Green! 2010 5.761166 Belgium Flemish Interest 2010 1.53748 Norway Centre Party 2009 4.680617 Norway Socialist Left Party 2009 7.589408 Norway Progress Party 2009 2.989397 Norway Norwegian Labour Party 2009 4.854283 Norway Christian Peoples Party 2009 9.999997 Norway Conservative Party 2009 .9999933 Norway Liberal Party 2009 7.563675 Czech Republic Communist Party of Bohemia and Moravia 2013 9.796195 Czech Republic Czech Social Democratic Party 2013 3.42685 Czech Republic Green Party 2013 3.194897 Czech Republic Civic Democratic Party 2013 .9999998 Czech Republic ANO 2011 2013 9.027879 Czech Republic Tradition, Responsibility, Prosperity 09 2013 6.160468 Czech Republic Christian and Democratic Union - Czech People’s Party 2013 10 Czech Republic Tomio Okamura’s Dawn of Direct Democracy 2013 8.918381 Croatia Coalition for Prime Minister 2016 1.000002 Croatia Even Stronger Istria Coalition 2016 1.000002 Croatia Bridge of Independent Lists 2016 4.65819 Croatia People’s coalition 2016 2.761999 Croatia Croatian Democratic Union 2016 10 Poland Law and Justice 2011 8.56399 Poland Democratic Left Alliance 2011 9.999997 Poland Civic Platform 2011 1.074257 Poland German Minority 2011 1.000002 Poland Palikot’s Movement 2011 1.362828 Poland Polish Peasants’ Party 2011 1.935969 Slovenia Zoran Jankovi’s List - Positive Slovenia 2011 1.509115 Slovenia For Real 2011 3.686006 Slovenia Liberal Democracy of Slovenia 2011 .9999998 Slovenia Slovenian People’s Party 2011 4.310893 Slovenia Social Democratic Party 2011 2.977317 Slovenia Slovenian National Party 2011 10 Slovenia Democratic Party of Pensioners of Slovenia 2011 6.922722 Slovenia Slovenian Democratic Party 2011 3.768581 Slovenia Gregor Virant’s Civic List 2011 1.369738 Slovenia New Slovenian Christian Peoples Party 2011 2.309385 Estonia Pro Patria and Res Publica Union 2015 8.170476 Estonia Social Democratic Party 2015 1.856552 Estonia Estonian Center Party 2015 .9999985 Estonia Conservative People’s Party of Estonia 2015 9.999999 Estonia Free Party 2015 4.19396 Estonia Estonian Reform Party 2015 3.340838

149 Table A.1 Continued Country Party Year Anti-Establishment Score Lithuania Lithuanian Peasant and Green Union 2012 1.246377 Lithuania Labour Party 2012 1 Lithuania Liberal Movement 2012 2.73295 Lithuania Lithuanian Social Democratic Party 2012 1.217274 Lithuania Order and Justice 2012 10 Lithuania The Way of Courage 2012 9.452346 Lithuania Homeland Union - Lithuanian Christian Democrats 2012 1.152652 Lithuania Election Action of Lithuanias Poles 2012 6.438651 Bulgaria Citizens for European Development of Bulgaria 2017 8.439117 Bulgaria Movement for Rights and Freedoms 2017 1.000001 Bulgaria Will 2017 9.999999 Bulgaria BSP for Bulgaria 2017 7.463573 Bulgaria United Patriots 2017 4.733954 Cyprus Ecological and Environmental Movement 2011 4.049773 Cyprus European Party 2011 10 Cyprus Democratic Coalition 2011 2.083842 Cyprus Progressive Party of the Working People 2011 1 Cyprus Democratic Party 2011 5.90388 Cyprus United Democratic Union of Cyprus 2011 5.409941 Romania Social Liberal Union 2012 10 Romania People’s Party - Dan Dianconescu 2012 1.000001 Romania Hungarian Democratic Alliance of Romania 2012 3.328792 Romania Right Romania Alliance 2012 1.665703 Luxembourg Alternative Democatic Reform Party 2013 4.113176 Luxembourg The Greens 2013 3.156027 Luxembourg The Left 2013 3.566448 Luxembourg Christian Social Peoples Party 2013 10 Luxembourg Democratic Party 2013 2.275777 Luxembourg Socialist Workers Party of Luxembourg 2013 1 Malta Nationalist Party 1998 1 Malta Labour Party 1998 10 Iceland Pirate Party 2013 10 Iceland The Alliance - Social Democratic Party of Iceland 2013 1.538356 Iceland Bright Future 2013 6.412155 Iceland Left Green Movement 2013 1.012735 Iceland Progressive Party 2013 1.000001

150 Figure A.1: Alternative Anti-Establishment Measure-Cross National Placement

Figure A.2: Alternative Measure of Anti-Establishment Sentiment-Nationalism Excluded

151 Figure A.3: CHES v. Anti-Establishment Sentiment, Nationalism Excluded

Table A.2: Predicting Corruption Salience. Anti-Establishment Sentiment and CHES Anti-Elite Salience Measure - Nationalism Excluded

(1) Corruption Salience

Anti-Establishment Sentiment 0.323*** (0.066) Anti-Elite Salience 0.374*** (0.046) Ideology (GAL-TAN) -0.143*** (0.039) Constant 2.184*** (0.420) Country Fixed Eﬀects Y Observations 98 R-squared 0.845

Note: Nationalism excluded from the Anti-Establishment Sentiment measure. Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1.

152 Figure A.4: Predicting Corruption Salience

Note: From the regression in Table A.2. Nationalism excluded from the Anti-Establishment Sentiment measure.

A.2 Greek Parties

Figure A.5: Cross National Party Placement Including Greece

153 Figure A.6: Greek Parties’ Anti-Establishment Placement 2015

Figure A.7: Greek Parties Over Time

154 APPENDIX B

ADDITIONAL MODELS

B.1 Netherlands

Table B.1: Netherlands. Probability of Voting for the PVV-Disaggregated Trust Measure

(1) (2) (3) Vote for PVV Vote for PVV Vote for PVV

Trust Parliament -0.406*** (0.0945) Trust Parties -0.164* (0.0957) Trust Government -0.366*** (0.0905) Satisfaction with Government -0.257*** -0.295*** -0.247*** (0.0639) (0.0624) (0.0647) Age -0.0182*** -0.0180*** -0.0182*** (0.00399) (0.00388) (0.00395) Female -0.199* -0.196* -0.201** (0.102) (0.102) (0.102) Married 0.0927 0.116 0.105 (0.117) (0.115) (0.115) Education -0.256*** -0.260*** -0.256*** (0.0490) (0.0490) (0.0493) Employed -0.0603 -0.0840 -0.0646 (0.143) (0.140) (0.143) Immigration Opinion -0.228*** -0.252*** -0.225*** (0.0712) (0.0703) (0.0713) EU Opinion -0.301*** -0.336*** -0.297*** (0.0732) (0.0720) (0.0722) Interest in Politics -0.120 -0.113 -0.138 (0.102) (0.100) (0.100) Income -0.0164* -0.0164* -0.0165* (0.00955) (0.00943) (0.00963) Working Class 0.281** 0.281** 0.272** (0.115) (0.113) (0.114) Ideology 0.147*** 0.142*** 0.152*** (0.0235) (0.0228) (0.0232) Constant 2.975*** 2.612*** 2.835*** (0.471) (0.471) (0.458)

Observations 1,657 1,647 1,655 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

155 Table B.2: Netherlands. Probability of Voting for the PVV over the VVD-Disaggregated Trust Measure

(1) (2) (3) PVV v. VVD PVV v. VVD PVV v. VVD

Trust Parliament -0.406*** (0.0945) Trust Parties -0.164* (0.0957) Trust Government -0.366*** (0.0905) Satisfaction with Government -0.257*** -0.295*** -0.247*** (0.0639) (0.0624) (0.0647) Age -0.0182*** -0.0180*** -0.0182*** (0.00399) (0.00388) (0.00395) Female -0.199* -0.196* -0.201** (0.102) (0.102) (0.102) Married 0.0927 0.116 0.105 (0.117) (0.115) (0.115) Education -0.256*** -0.260*** -0.256*** (0.0490) (0.0490) (0.0493) Employed -0.0603 -0.0840 -0.0646 (0.143) (0.140) (0.143) Immigration -0.228*** -0.252*** -0.225*** (0.0712) (0.0703) (0.0713) EU Opinion -0.301*** -0.336*** -0.297*** (0.0732) (0.0720) (0.0722) Interest -0.120 -0.113 -0.138 (0.102) (0.100) (0.100) Income -0.0164* -0.0164* -0.0165* (0.00955) (0.00943) (0.00963) Working Class 0.281** 0.281** 0.272** (0.115) (0.113) (0.114) Ideology 0.147*** 0.142*** 0.152*** (0.0235) (0.0228) (0.0232) Constant 2.975*** 2.612*** 2.835*** (0.471) (0.471) (0.458)

Observations 1,657 1,647 1,655 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

156 B.2 Italy

Table B.3: Italy. Probability of Voting for M5S-Disaggregated Trust Measure

(1) (2) (3) Vote for M5S Vote for M5S Vote for M5S

Trust Parliament -0.196*** (0.0698) Trust Parties -0.382*** (0.0782) Trust President -0.189*** (0.0616) Evaluation of Monti -0.0502** -0.0492** -0.0418* (0.0228) (0.0227) (0.0242) College 0.145 0.131 0.120 (0.111) (0.111) (0.110) Married 0.118 0.107 0.144 (0.117) (0.116) (0.117) Catholic 0.0917 0.129 0.117 (0.173) (0.172) (0.173) Employed -0.126 -0.115 -0.146 (0.102) (0.102) (0.103) Unskilled Worker 0.418** 0.393* 0.457** (0.212) (0.217) (0.217) Ideology -0.0672*** -0.0629*** -0.0757*** (0.0196) (0.0197) (0.0198) Religious Service Attendance -0.00770 -0.00684 -0.00785 (0.0212) (0.0210) (0.0212) Immigration Opinion 0.0104 0.0222 -0.00621 (0.0317) (0.0320) (0.0313) Interest in Politics 0.119* 0.111* 0.111* (0.0620) (0.0624) (0.0618) EU Opinion -0.127** -0.118* -0.0711 (0.0642) (0.0648) (0.0662) Female 0.0348 0.0101 0.0424 (0.0988) (0.101) (0.0983) Age -0.0184*** -0.0183*** -0.0184*** (0.00339) (0.00339) (0.00344) Constant 0.566* 0.720** 0.647** (0.323) (0.327) (0.322)

Observations 1,100 1,103 1,103 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

157 Table B.4: Italy. Probability of Voting for M5S over the Left Alliance-Disaggregated Trust Measure

(1) (2) (3) M5S v. Left M5S v. Left M5S v. Left

Trust Parliament -0.365*** (0.0941) Trust Parties -0.570*** (0.104) Trust President -0.338*** (0.0817) Evaluation of Monti -0.133*** -0.136*** -0.113*** (0.0322) (0.0333) (0.0335) College 0.198 0.0978 0.0724 (0.152) (0.149) (0.150) Married 0.196 0.134 0.179 (0.149) (0.148) (0.154) Catholic -0.0717 -0.0782 -0.0916 (0.218) (0.219) (0.214) Employed -0.194 -0.170 -0.206 (0.137) (0.138) (0.138) Unskilled Worker 0.415 0.363 0.431 (0.316) (0.323) (0.325) Ideology 0.266*** 0.268*** 0.239*** (0.0373) (0.0374) (0.0354) Religious Services Attendance -0.0350 -0.0324 -0.0349 (0.0283) (0.0290) (0.0285) Immigration Opinion -0.0250 -0.0227 -0.0706* (0.0432) (0.0439) (0.0427) Interest in Politics 0.0165 0.0106 0.0171 (0.0844) (0.0860) (0.0825) EU Opinion -0.150* -0.105 -0.0251 (0.0867) (0.0889) (0.0900) Female 8.19e-05 -0.0308 0.0321 (0.134) (0.137) (0.134) Age -0.0222*** -0.0215*** -0.0223*** (0.00448) (0.00457) (0.00467) Constant 1.279*** 1.411*** 1.495*** (0.451) (0.451) (0.453)

Observations 531 531 531 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

158 Table B.5: Italy. Multinomial Model with Lega Nord Voters Excluded

Left Alliance Right Coalition M5S v. v. Incumbent v. Incumbent Incumbent Institutional Trust 0.135 -0.0841 -0.323*** (0.102) (0.121) (0.111) Evaluation of Monti Government -0.404*** -0.523*** -0.567*** (0.0926) (0.121) (0.0965) College -0.594* -0.586 -0.345 (0.351) (0.386) (0.359) Carried -0.431 -0.388 -0.255 (0.399) (0.453) (0.415) Catholic -1.028 -0.884 -0.982 (0.831) (1.054) (0.867) Employed -0.172 -0.295 -0.406 (0.368) (0.444) (0.382) Unskilled Worker 0.579 0.874 1.242 (1.166) (1.158) (1.192) Ideology -0.843*** 0.548*** -0.357*** (0.0947) (0.111) (0.0933) Religious Service Attendance 0.00272 -0.0877 -0.0857 (0.0753) (0.0952) (0.0781) Immigration Opinion 0.132 -0.0605 0.0767 (0.115) (0.133) (0.124) Interest in Politics 0.245 -0.116 0.328 (0.207) (0.244) (0.215) EU Opinion 0.0797 -0.112 -0.0653 (0.234) (0.270) (0.245) Female -0.287 -0.401 -0.227 (0.321) (0.382) (0.339) Age -0.000889 0.00907 -0.0322** (0.0130) (0.0132) (0.0133) Constant 7.072*** 1.604 9.057*** (1.542) (1.838) (1.573)

Observations 738 738 738 738 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

159 B.3 France

Table B.6: France. Probability of Voting for Macron in the First Round-Disaggregated Trust Measure

(1) (2) (3) (4) Vote for Macron Vote for Macron Vote for Macron Vote for Macron

Trust Politicians 0.124** (0.0550) Politicians care 0.136*** (0.0504) Politicians not the problem 0.0720 (0.0527) Politicians do not care about rich 0.141*** (0.0504) Female 0.0883 0.0777 0.0822 0.0799 (0.114) (0.114) (0.114) (0.114) Education 0.0969** 0.0750 0.102** 0.0724 (0.0490) (0.0492) (0.0494) (0.0500) Income 0.0318 0.0370 0.0332 0.0308 (0.0311) (0.0317) (0.0318) (0.0314) Positive view of the national economy 0.0625 0.0651 0.0742 0.0629 (0.0616) (0.0616) (0.0617) (0.0617) Ideology 0.0547** 0.0531* 0.0569** 0.0440 (0.0278) (0.0283) (0.0280) (0.0285) Positive view of personal ﬁnances -0.113 -0.121 -0.121 -0.0963 (0.124) (0.124) (0.125) (0.123) Satisfaction with Hollande 0.349*** 0.355*** 0.371*** 0.365*** (0.0804) (0.0810) (0.0816) (0.0809) Immigrants economy -0.0985 -0.0879 -0.108* -0.101 (0.0610) (0.0626) (0.0615) (0.0618) Immigrants French culture 0.104* 0.0841 0.0987 0.0938 (0.0624) (0.0631) (0.0630) (0.0630) Immigrants crime 0.00600 -0.00541 0.00513 -0.00824 (0.0687) (0.0693) (0.0699) (0.0694) Immigrants beneﬁts 0.0470 0.0487 0.0457 0.0376 (0.0582) (0.0596) (0.0595) (0.0601) Support for EU 0.403*** 0.409*** 0.397*** 0.376*** (0.111) (0.111) (0.113) (0.112) Interest in Politics 0.115* 0.114* 0.121* 0.116* (0.0681) (0.0672) (0.0677) (0.0676) Catholic 0.163 0.189 0.209* 0.155 (0.121) (0.120) (0.122) (0.120) Religious Services Attendance -0.104 -0.104 -0.0914 -0.102 (0.0646) (0.0653) (0.0654) (0.0636) Married 0.315** 0.345*** 0.317** 0.328*** (0.127) (0.127) (0.127) (0.126) Constant -4.297*** -4.120*** -4.230*** -3.942*** (0.460) (0.461) (0.470) (0.466)

Observations 836 833 825 834 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

160 Table B.7: France. Probability of Voting for Le Pen in the First Round. Disaggregated Trust Measure

(1) (2) (3) (4) Vote for Le Pen Vote for Le Pen Vote for Le Pen Vote for Le Pen

Trust Politicians -0.0695 (0.0671) Politicians care people -0.0534 (0.0617) Politicians not the problem 0.0711 (0.0655) Politicians do not care the rich -0.0635 (0.0646) Female 0.0980 0.107 0.0975 0.0930 (0.129) (0.129) (0.129) (0.128) Education -0.0604 -0.0512 -0.0619 -0.0479 (0.0522) (0.0529) (0.0523) (0.0531) Income -0.0529 -0.0543 -0.0596* -0.0519 (0.0344) (0.0344) (0.0345) (0.0344) Positive view national economy -0.162** -0.163** -0.173** -0.161** (0.0750) (0.0745) (0.0760) (0.0746) Ideology 0.160*** 0.160*** 0.155*** 0.162*** (0.0343) (0.0345) (0.0344) (0.0346) Positive view personal ﬁnances 0.208 0.214 0.212 0.200 (0.136) (0.135) (0.136) (0.136) Satisfaction with Hollande -0.0800 -0.0892 -0.114 -0.0882 (0.105) (0.104) (0.106) (0.105) Immigrants the economy -0.139** -0.146** -0.142** -0.148** (0.0647) (0.0649) (0.0653) (0.0656) Immigrants French Culture -0.169** -0.161** -0.188** -0.166** (0.0736) (0.0737) (0.0752) (0.0756) Immigrants crime -0.0694 -0.0660 -0.0549 -0.0576 (0.0722) (0.0716) (0.0726) (0.0719) Immigrants beneﬁts -0.0524 -0.0495 -0.0650 -0.0457 (0.0761) (0.0759) (0.0766) (0.0754) Support for the EU -0.331*** -0.335*** -0.354*** -0.324*** (0.0919) (0.0910) (0.0931) (0.0928) Interest in Politics 0.0818 0.0785 0.0636 0.0839 (0.0709) (0.0706) (0.0703) (0.0707) Catholic 0.199 0.199 0.167 0.195 (0.141) (0.140) (0.142) (0.141) Religious Services Attendance -0.0715 -0.0755 -0.0728 -0.0730 (0.0831) (0.0827) (0.0832) (0.0830) Married -0.223 -0.228 -0.217 -0.227 (0.145) (0.145) (0.145) (0.145) Constant 0.748* 0.648 0.760* 0.599 (0.399) (0.398) (0.405) (0.406)

Observations 836 833 825 834 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

161 Table B.8: France. Probability of Voting for Macron over Le Pen in the Second Round. Disaggre- gated Trust Measure

(1) (2) (3) (4) Macron Macron Macron Macron v. Le Pen v. Le Pen v. Le Pen v. Le Pen

Trust Politicians 0.158** (0.0792) Politician care about people 0.230*** (0.0775) Politicians are not the biggest problem -0.0156 (0.0832) Politicians do not care only about the rich 0.0961 (0.0699) Female 0.184 0.165 0.152 0.187 (0.149) (0.152) (0.149) (0.149) Education -0.0124 -0.0483 -0.0159 -0.0335 (0.0610) (0.0626) (0.0618) (0.0617) Income 0.0833** 0.0936** 0.0843** 0.0844** (0.0423) (0.0427) (0.0427) (0.0420) Positive view of the national economy 0.112 0.104 0.116 0.115 (0.0817) (0.0820) (0.0823) (0.0810) Ideology -0.114*** -0.122*** -0.105*** -0.118*** (0.0395) (0.0394) (0.0399) (0.0399) Positive view of personal ﬁnances -0.369** -0.359** -0.402*** -0.359** (0.154) (0.153) (0.155) (0.154) Satisfaction with Hollande Government 0.269** 0.291** 0.300** 0.298*** (0.113) (0.113) (0.117) (0.112) Immigrants are good for the economy 0.237*** 0.255*** 0.254*** 0.246*** (0.0725) (0.0723) (0.0759) (0.0731) Immigrants are good for French Culture 0.116 0.0755 0.122 0.110 (0.0814) (0.0816) (0.0858) (0.0854) Immigrants do not negatively impact crime 0.109 0.106 0.0878 0.0908 (0.0826) (0.0830) (0.0838) (0.0830) Immigrants do not steal beneﬁts 0.0984 0.0925 0.129* 0.0905 (0.0760) (0.0777) (0.0775) (0.0761) Support for the EU 0.647*** 0.651*** 0.681*** 0.648*** (0.116) (0.118) (0.119) (0.118) Interest in Politics 0.0229 0.00890 0.0474 0.0172 (0.0852) (0.0871) (0.0844) (0.0846) Catholic -0.328** -0.341** -0.301* -0.314* (0.162) (0.162) (0.165) (0.162) Religious Services Attendance 0.155* 0.190** 0.214** 0.173** (0.0875) (0.0898) (0.0861) (0.0877) Married 0.416** 0.438** 0.421** 0.405** (0.173) (0.173) (0.177) (0.173) Constant -3.649*** -3.437*** -3.643*** -3.360*** (0.559) (0.562) (0.573) (0.553)

Observations 581 578 572 581 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

162 APPENDIX C

ISSUE ATTENTION OF NON-CORE ISSUES

C.1 European Parliament

Figure C.1: European Parliament. Attention to Non-Core Issues Among Anti-Establishment Par- ties By Ideology

C.2 Italy

Figure C.2: Italian Parliament. Attention to Non-Core Issues Among Anti-Establishment Parties By Ideology

163 BIBLIOGRAPHY

Aalberg, Toril, and Claes H de Vreese. 2016. “Introduction: Comprehending populist political communication.” In Populist political communication in Europe. Routledge , 13–21.

Abedi, Amir. 2009. Anti-Political Establishment Parties: A Comparative Analysis. London and New York: Routledge.

Abedi, Amir, and Thomas Carl Lundberg. 2009. “Doomed to Failure? UKIP and the Organisational Challenges Facing Right-Wing Populist Anti-Political Establishment Parties.” Parliamentary Aﬀairs 62(1): 72–87.

Abedi, Arim. 2002. “Challenges to Established Earties: The Eﬀects of Party System Features on the Electoral Fortunes of Anti-Political-Establishment Parties.” European Journal of Political Research 41(4): 551–583.

Abramson, Paul R, Susan Ellis, and Ronald Inglehart. 1997. “Research in context: Measuring value change.” Political Behavior 19(1): 41–59.

Adams, James, Michael Clark, Lawrence Ezrow, and Garrett Glasgow. 2006. “Are Niche Parties Fundamentally Diﬀerent from Mainstream Parties? The Causes and the Electoral Conse- quences of Western European Parties’ Policy Shifts, 1976-1998.” American Journal of Political Science 50(3): 513–529.

Agerberg, Mattias. 2017. “Failed expectations: Quality of government and support for populist parties in Europe.” European Journal of Political Research .

Akkerman, Tjitske. 2005. “Anti-immigration parties and the defence of liberal values: The excep- tional case of the List Pim Fortuyn.” Journal of Political Ideologies 10(3): 337–354.

Albertazzi, Daniele, and Duncan McDonnell. 2008. Twenty-First Century Populism:The Spectre of Western European Democracy. New York: Palgrave McMillan.

Albertazzi, Daniele, Duncan McDonnell, and James L. Newell. 2011. “Di Lotta e Di Governo: The Lega Nord and Rifondazione Comunista in Oﬃce.” Party Politics 17(4): 471–487.

Ansolabehere, Stephen, and Shanto Iyengar. 1994. “Riding the Wave and Claiming Ownership Over Issues.” Public Opinion Quarterly 58: 335–357.

Armingeon, Klaus, and Kai Guthmann. 2014. “Democracy in crisis? The declining support for national democracy in European countries, 2007–2011.” European Journal of Political Research 53(3): 423–442.

164 Arzheimer, Kai. 2015. “The afd: Finally a successful right-wing populist eurosceptic party for germany?” West European Politics 38(3): 535–556.

Arzheimer, Kai, and Elisabeth Carter. 2006. “Political Opportunity Structures and Right-Wing Extremist Party Success.” European Journal of Political Research 45(3): 419–443.

Bale, Tim, Stijn Van Kessel, and Paul Taggart. 2011. “Thrown around with abandon? Popular understandings of populism as conveyed by the print media: A UK case study.” Acta Politica 46(2): 111–131.

Bartolini, Stefano, Alessandro Chiaramonte, and Roberto D’Alimonte. 2004. “The Italian Party System Between Parties and Coalitions.” West European Politics 27(1): 1–19.

B´elanger, Eric,´ and Richard Nadeau. 2005. “Political trust and the vote in multiparty elections: The Canadian case.” European journal of political research 44(1): 121–146.

Betz, Hans-Georg, and Stefan Immerfall. 1998. The new politics of the right: Neo-populist parties and movements in established democracies. Macmillan.

Betz, Hans George. 1994. Radical Right Wing Populism. London: McMillan Press.

Biorcio, Roberto. 2014. “The Reasons for the Success and Transformations of the 5 Star Movement.” Contemporary Italian Politics 6(1): 37–53.

Budge, I., and D. Farlie. 1983. Explaining and Predicting Elections. London: Allen & Unwin.

Budge, Ian, Lawrence Ezrow, and Michael D. McDonald. 2010. “Ideology, Party Factionalism and Party Change: An Integrated Dynamic Theory.” British Journal of Political Science 40: 781– 804.

Bustikova, L. 2014. “Revenge of the Radical Right.” Comparative Political Studies 47: 1738–1765.

Campbell, Angus, and Philip E. Converse. 1960. “Converse/Warren E. Miller/Donald E. Stokes, The American Voter.” New York: Wiley.

Canovan, Margaret. 1981. Populism. Houghton Miﬄin Harcourt P.

Carrubba, Cliﬀord, and Matthew Gabel. 1999. “Roll-Call Votes and Party Discipline in the Euro- pean Parliament: Reconsidering MEP Voting Behavior.”.

Citrin, Jack. 1974. “Comment: The political relevance of trust in government.” American Political Science Review 68(3): 973–988.

Citrin, Jack, and Donald Philip Green. 1986. “Presidential leadership and the resurgence of trust in government.” British journal of political science 16(4): 431–453.

165 Citrin, Jack, and Samantha Luks. 2001. “Political trust revisited: D´ej`avu all over again?” What is it about government that Americans dislike , 9–27.

Citrin, Jack, Donald P Green, Christopher Muste, and Cara Wong. 1997. “Public opinion toward immigration reform: The role of economic motivations.” The Journal of Politics 59(3): 858– 881.

Coﬀ´e,Hilde, and Rebecca Plassa. 2010. “Party policy position of Die Linke: A continuation of the PDS?” Party Politics 16(6): 721–735.

Cowley, Philip, and Sarah Childs. 2003. “Too Spineless to Rebel? New Labour’s Women MPs.” British Journal of Political Science 33: 345–365.

Decker, Frank. 2008. “Germany: Right-wing populist failures and left-wing successes.” In Twenty- First Century Populism. Springer , 119–134.

Diamanti, Ilvo. 2014. “The 5 Star Movement: a Political Laboratory.” Contemporary Italian Politics 6(1): 4–15.

Doerschler, Peter. 2015. “Die Linke: Still an Eastern Cultural Icon?” German Politics 24(3): 377– 401.

Duch, Raymond, and Randolph Stevenson. 2008. The Economic Vote: How Political and Economic Institutions Condition Election Results. Cambridge: Cambridge University Press.

Dunteman, George H. 1989. Principal components analysis. Number 69 Sage.

Dutch Parliamentary Election Study. 2010. http://www.persistent-identiﬁer.nl/?identiﬁer= urn:nbn:nl:ui:13-9x4l-vy.

Easton, David. 1965. A Systems Analysis of Political Life. New York: Wiley.

Ellinas, Antonis A. 2013. “The rise of Golden Dawn: The new face of the far right in Greece.” South European Society and Politics 18(4): 543–565.

Erlingsson, Gissur O,´ and Mikael Persson. 2011. “The swedish pirate party and the 2009 european parliament election: Protest or issue voting?” Politics 31(3): 121–128.

Evans, Jocelyn AJ. 2005. “The dynamics of social change in radical right-wing populist party support.” Comparative European Politics 3(1): 76–101.

Evans, Jocelyn, and Gilles Ivaldi. 2017. “An atypical honeymoonelection? Contextual and strategic opportunities in the 2017 French legislative elections.” French Politics 15(3): 322–339.

166 Evans, Jocelyn, and Gilles Ivaldi. 2018. “The 2017 Presidential Election: Continuity and Change.” In The 2017 French Presidential Elections. Springer , 17–44.

Ezrow, Lawrence. 2008. “Research Note: On the Inverse Relationship Between Votes and Proximity for Niche Parties.” European Journal of Political Research 47: 206–220.

Faucher, Florence, and N´uriaGarcia. 2016. “France.” European Journal of Political Research Political Data Yearbook 55(1): 99–105.

Fitzgerald, Jennifer et al. 2012. “Anxious Publics: Worries About Crime and Immigration‘.” Comparative Political Studies 45: 477–506.

Franzmann, Simon T. 2016. “Calling the Ghost of Populism: The AfD’s Strategic and Tactical Agendas until the EP Election 2014.” German Politics 25(4): 457–479.

Franzosi, Paolo, Francesco Marone, and Eugenio Salvati. 2015. “Populism and Euroscepticism in the Italian Five Star Movement.” The International Spectator 50(2): 109–124.

Gamson, William A. 1968. “Stable unrepresentation in American society.” American Behavioral Scientist 12(2): 15–21.

Golder, Matt. 2003. “Explaining variation in the success of extreme right parties in Western Europe.” Comparative Political Studies 36(4): 432–466.

Golder, Matt. 2016. “Far right parties in Europe.” Annual Review of Political Science 19: 477–497.

Golder, Sona Nadenichek. 2006. “Pre-Electoral Coalition Formation in Parliamentary Democra- cies.” British Journal of Political Science 36(2): 193–212.

Goodwin, Matthew, and Caitlin Milazzo. 2017. “Taking back control? Investigating the role of immigration in the 2016 vote for Brexit.” The British Journal of Politics and International Relations 19(3): 450–464.

Green-Pedersen, Christoﬀer. 2007. “The Growing Importance of Issue Competition: The Changing Nature of Party Competition in Western Europe.” Political Studies 55: 608–628.

Green-Pedersen, Christoﬀer, and Peter B Mortensen. 2010. “Who sets the agenda and who responds to it in the Danish parliament? A new model of issue competition and agenda-setting.” European Journal of Political Research 49(2): 257–281.

Green-Pederson, Christoﬀer, and Peter Mortensen. 2009. “Who Sets the Agenda and Who Responds to it in the Danish Parliament? A New Model of Issue Competition and Agenda Setting.” European Journal of Political Research 49: 257–281.

167 Groppi, Tania. 2012. “The Impact of the Financial Crisis on the Italian Written Constitution.” HeinOnline.

Hameleers, Michael, L Bos, and CH de Vreese. 2016. “They did it.” The Eﬀects of Emotionalized Blame Attribution in Populist Communication. Communication Research, prepublished online .

Hartleb, Florian. 2013. “Anti-elitist cyber parties?” Journal of Public Aﬀairs 13(4): 355–369.

Hartleb, Florian. 2015. “Here to Stay: Anti-Establishment Parties in Europe.” European View 14(1): 39–49.

Heller, William B, and Carol Mershon. 2008. “Dealing in discipline: party switching and legislative voting in the Italian Chamber of Deputies, 1988–2000.” American Journal of Political Science 52(4): 910–925.

Hetherington, Marc J. 1998. “The political relevance of political trust.” American political science review 92(4): 791–808.

Hetherington, Marc J. 1999. “The eﬀect of political trust on the presidential vote, 1968–96.” American Political Science Review 93(2): 311–326.

Hix, Simon. 2004. “Electoral institutions and legislative behavior: Explaining voting defection in the European Parliament.” World politics 56(2): 194–223.

Hix, Simon, Abdul Noury, and Gerard Roland. 2006. “Dimensions of politics in the European Parliament.” American Journal of Political Science 50(2): 494–520.

Hix, Simon, and Michael Marsh. 2007. “Punishment or protest? Understanding European parliament elections.” Journal of Politics 69(2): 495–510.

Hooghe, Liesbet, Gary Marks, and Carole J. Wilson. 2002. “Does Left/Right Structure Party Positions on European Integration?” Comparative Political Studies 35: 965–989.

Ignazi, Piero. 2003. Extreme right parties in Western Europe. Oxford University Press on Demand.

Indridason, Indridi H. 2014. “The collapse: Economic considerations in vote choice in Iceland.” Journal of Elections, Public Opinion & Parties 24(2): 134–159.

Inglehart, Ronald. 1977. The Silent Revolution. Princeton: Princeton University Press.

Inglehart, Ronald. 1995. “Public support for environmental protection: Objective problems and subjective values in 43 societies.” PS: Political Science & Politics 28(1): 57–72.

168 Inglehart, Ronald, and Pippa Norris. 2017. “Trump and the Populist Authoritarian Parties: The Silent Revolution in Reverse.” Perspectives on Politics 15(2): 443–454.

Italian National Election Studies 2013 Post-Election Study. 2013. www.itanes.org.

Ivarsﬂaten, Elisabeth. 2008. “What Unites Right-Wing Populists in Western Europe? Re-examining Grievance Mobilization Models in Seven Successful Cases.” Comparative Political Studies 41(1): 3–23.

Jackman, Robert W, and Karin Volpert. 1996. “Conditions Favouring Parties of the Extreme Right in Western Europe.” British Journal of Political Science 26(4): 501–521.

Jankowski, Michael, Sebastian Schneider, and Markus Tepe. 2017. “Ideological alternative? Analyz- ing Alternative f¨urDeutschland candidates ideal points via black box scaling.” Party Politics 23(6): 704–716.

Jennings, Will, Shaun Bevan Arco Timmermans Gerard Breeman Sylvaln Brouard Laura Chaqués- Bonafont Christoffer Green-Pedersen Peter John Peter B. Mortensen and Anna M. Palau. 2011. “Effects of the Core Functions of Government on the Diversity of Executive Agendas.” Comparative Political Studies 44: 1001–1030.

Jones, Bryan. 2001. Politics and the Architecture of Choice. Chicago: Chicago University Press.

Jones, Bryan, and Frank Baumgartner. 2005. The Politics of Attention: How Government Priori- tizes Attention. Chicago: University of Chicago Press.

Kitschelt, Herbert. 1994. The Transformation of European Social Democracy. Cambridge: Univer- sity of Cambridge Press.

Kl¨uver, Heinke, and Inaki Sagarzazu. 2016. “Setting the Agenda or Responding to Voters? Political Parties, Voters, and Issue Attention.” West European Politics 39: 380–398.

Knigge, P. 1998. “The Ecological Correlates of Right-Wing Extremism in Western Europe.” Euro- pean Journal of Politics 34: 249–279.

Lewis-Beck, Michael S., and Glenn E. Mitchell. 1993. “French Electoral Theory: The National Front Test.” Electoral Studies 12.

Lewis-Beck, Michael S, and Glenn E Mitchell II. 1993. “French electoral theory: the National Front test.” Electoral Studies 12(2): 112–127.

Lewis-Beck, Michael S, and Richard Nadeau. 2000. “French electoral institutions and the economic vote.” Electoral Studies 19(2): 171–182.

169 Lewis-Beck, Michael S, and Richard Nadeau. 2015. “Explaining French elections: The presidential pivot.” French Politics 13(1): 25–62.

Linde, Jonas, and Joakim Ekman. 2003. “Satisfaction with democracy: A note on a frequently used indicator in comparative politics.” European journal of political research 42(3): 391–408.

Lipset, Seymour M, and Stein Rokkan. 1967. “Cleavage structures, party systems, and voter alignments: an introduction.”.

Lubbers, Marcel, M´erove Gijsberts, and Peer Scheepers. 2002. “Extreme Right-Wing Voting in Western Europe.” European Journal of Political Research 41(3): 345–378.

Lucassen, Geertje, and Marcel Lubbers. 2012. “Who Fears What? Explaining Far-Right-Wing Preference in Europe by Distinguishing Perceived Cultural and Economic Ethnic Threats.” Comparative Political Studies 45(5): 547–574.

Mair, Peter, and Cas Mudde. 1998. “The party family and its study.” Annual Review of Political Science 1(1): 211–229.

Mayer, Nonna, and Pascal Perrineau. 1992. “Why do they vote for Le Pen?” European Journal of Political Research 22(1): 123–141.

Mayhew, David. 1974. Congress: The Electoral Connection. New Haven: Yale University Press.

Mazzoleni, Gianpietro, Julianne Stewart, and Bruce Horsﬁeld. 2003. The media and neo-populism: A contemporary comparative analysis. Greenwood Publishing Group.

McDonnell, Duncan, and James L. Newell. 2011. “Outsider Parties in Government in Western Europe.” Party Politics 17(4): 443–452.

McGann, H Kitschelt AJ, and Herbert Kitschelt. 1995. “The Radical Right in Western Europe: A Comparative Analysis.” Ann Arbor: University of Michigan Press .

Meguid, B.M. 2005. “Competition Between Unequals: The Role of Mainstream Party Strategy in Niche Party Success.” American Political Science Review 99: 347–359.

Meserve, Stephen A, Daniel Pemstein, and William T Bernhard. 2009. “Political ambition and legislative behavior in the European Parliament.” The Journal of Politics 71(3): 1015–1032.

Meyer, Thomas M., and Markus Wagner. 2013. “Mainstream or Niche? Vote-Seeking Incentives and the Programmatic Strategies of Political Parties.” Comparative Political Studies 46(10): 1246– 1272.

Miller, Arthur H. 1974. “Political issues and trust in government: 1964–1970.” American political science review 68(3): 951–972.

170 Miller, Arthur H, and Ola Listhaug. 1990. “Political parties and conﬁdence in government: A comparison of Norway, Sweden and the United States.” British Journal of Political Science 20(3): 357–386.

Mosca, Lorenzo. 2014. “The Five Star Movement: Exception or Vanguard in Europe?” The International Spectator 49: 36–52.

Mudde, Cas. 2004. “The Populist Zeitgeist.” Government and Opposition 39(4): 541–563.

Mudde, Cas. 2007. Populist Radical Right Parties in Europe. Vol. 22 Cambridge University Press Cambridge.

Mudde, Cas. 2010. “The Populist Radical Right: A Pathological Normalcy.” West European Politics 33: 1167–1186.

M¨uller,Jan-Werner. 2017. What is populism? Penguin UK.

M¨uller,W., and T. Saalfeld. 1997. Members of Parliament in Western Europe. London: Routeledge.

Natale, Paolo. 2014. “The Birth, Early History and Explosive Growth of the Five Star Movement.” Contemporary Italian Politics 6(1): 16–36.

Norris, Pippa. 1999. Critical citizens: Global support for democratic government. OUP Oxford.

Norris, Pippa. 2005. Radical Right: Voters and Parties in the Election Market. Cambridge: Cam- bridge University Press.

Orriols, Lluis, and Guillermo Cordero. 2016. “The breakdown of the Spanish two-party system: the upsurge of Podemos and Ciudadanos in the 2015 general election.” South European Society and Politics 21(4): 469–492.

Pasquino, Gianfranco. 2008. “Populism and democracy.” In Twenty-First Century Populism. Springer , 15–29.

Peterson, Geoﬀ, and J Mark Wrighton. 1998. “Expressions of distrust: Third-party voting and cynicism in government.” Political Behavior 20(1): 17–34.

Petrocik, J. 1996. “Issue Ownership in Presidential Elections.” American Journal of Political Science 38: 100–123.

Polk, Jonathan, Jan Rovny, Ryan Bakker, Erica Edwards, Liesbet Hooghe, Seth Jolly, Jelle Koedam, Filip Kostelka, Gary Marks, Gijs Schumacher et al. 2017. “Explaining the salience of anti-elitism and reducing political corruption for political parties in Europe with the 2014 Chapel Hill Expert Survey data.” Research & Politics 4(1): 2053168016686915.

171 Ramiro, Luis, and Raul Gomez. 2017. “Radical-left populism during the great recession: Podemos and its competition with the established radical left.” Political Studies 65(1 suppl): 108–126.

Reif, Karlheinz, and Hermann Schmitt. 1980. “Nine second-order national elections–a conceptual framework for the analysis of European Election results.” European journal of political research 8(1): 3–44.

Rihoux, B., and W. R¨udig.2006. “Analyzing Greens in Power: Setting the Agenda.” European Journal of Political Research 45: S1–S33.

Rooduijn, Matthijs, Wouter Van Der Brug, and Sarah L De Lange. 2016. “Expressing or fuelling discontent? The relationship between populist voting and political discontent.” Electoral Stud- ies 43: 32–40.

Rydgren, Jens. 2008. “Immigration sceptics, xenophobes or racists? Radical right-wing voting in six West European countries.” European Journal of Political Research 47(6): 737–765.

Sartori, Giovanni. 1976. “Party and party systems.” A Framework for Analysis. Cambridge, Cam- bridge UP .

Schedler, Andreas. 1996. “Anti-Political-Establishment Parties.” Party Politics 2: 291–312.

Schumacher, Gijs, and Matthijs Rooduijn. 2013. “Sympathy for the ‘devil’? Voting for populists in the 2006 and 2010 Dutch general elections.” Electoral Studies 32(1): 124–133.

Simon, Herbert A. 1971. “Designing Organization for an Information-Rich World.” In Computers, Communication, and the Public Interest, ed. Martin Greenberger. Baltimore: Johns Hopkins University Press , 37–72.

Sola, Jorge, and C´esarRendueles. 2018. “Podemos, the upheaval of Spanish politics and the challenge of populism.” Journal of Contemporary European Studies 26(1): 99–116.

Sozzi, Fabio. 2016. “Asking territories: the constituency orientation of Italian and French members of the European Parliament.” Italian Political Science Review/Rivista Italiana di Scienza Politica 46(2): 199–217.

Stockemer, Daniel, and Mauro Barisione. 2017. “The newdiscourse of the Front National under Marine Le Pen: A slight change with a big impact.” European Journal of Communication 32(2): 100–115.

Str¨om,K. 1990. “A Behavioral Theory of Competitive Political Parties.” American Political Science Review 34: 565–598.

172 Str¨om,Kaare. 1997. “Rules, Reasons and Routines: Legislative Roles in Parliamentary Democra- cies.” In Members of Parliament in Western Europe, eds. Wolfgang C. M¨uller,and Thomas Saalfeld. London: Frank Cass and Company Limited , 155–174.

Ström,Karre, Wolfgang C. Müller,and TorbjörnBergman. 2003. Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press.

Svolik, Milan W. 2013. “Learning to love democracy: electoral accountability and the success of democracy.” American Journal of Political Science 57(3): 685–702.

Taggart, Paul. 2004. Populism. Buckingham: Open University Press.

Tajfel, Henri, and John C Turner. 1979. “An Integrative Theory of Intergroup Conﬂict.” The Social Psychology of Intergroup Relations 33(47): 74.

Van der Brug, Wouter. 2003. “How the LPF fuelled discontent: Empirical tests of explanations of LPF support.” Acta Politica 38(1): 89–106.

Van der Eijk, Cees, Mark N Franklin et al. 1996. “Choosing Europe? The European electorate and national politics in the face of union.”.

Van Kessel, Stijn. 2011. “Explaining the electoral performance of populist parties: The Netherlands as a case study.” Perspectives on European Politics and Society 12(1): 68–88.

Vannucci, Alberto. 2009a. “The Controversial Legacy of ‘Mani Pulite’: A Critical Analysis of Italian Corruption and Anti-Corruption Policies.” Bulletin of Italian Politics 1(2): 233–264.

Vannucci, Alberto. 2009b. “The controversial legacy of Mani Pulite: A critical analysis of Italian corruption and anti-corruption policies.” Bulletin of Italian politics 1(2): 233–64.

Vliegenthart, Rens, and Stefaan Walgrave. 2011. “Content Matters: The Dynamics of Parliamen- tary Questioning in Belgium and Denmark.” Comparative Political Studies 44: 1031–1059.

Vossen, Koen. 2010. “Populism in the Netherlands after Fortuyn: Rita Verdonk and Geert Wilders Compared.” Perspectives on European Politics and Society 11(1): 22–38.

Wagner, Aiko, Marcel Lewandowsky, and Heiko Giebler. 2015. “Alles neu macht der Mai?” In Die Europawahl 2014. Springer , 137–148.

Wagner, M. 2011. “Deﬁning and Measuring Niche Parties.” Party Politics 18: 845–864.

Wiberg, Matti, and Antii Koura. 1994. “The Logic of Parliamentary Questioning.” In Parliamen- tary Control in the Nordic Countries, ed. Matti Wiberg. Tampere: Finnish Political Science Association , 179–222.

173 Williams, Michelle Hale. 2006. The Impact of Radical Right-Wing Parties in West European Democ- racies. New York: Palgrave MacMillan.

174 BIOGRAPHICAL SKETCH

Teresa Cornacchione earned her Bachelors degree in political science from the University of West Florida in 2002, and her Masters degree in 2005. She spent nine years as a public high school teacher before pursuing her doctorate at Florida State University. Her research interests revolve around anti-establishment and radical parties, comparative political behavior, and gender and politics.

175