Voting for a change
The democratic lure of
populist radical right parties
in voting behavior
Manuscript committee:
Prof.dr. M.J.A.M. Verkuyten (chairman, Utrecht University)
Prof.dr. A. Akkerman (University of Groningen / Radboud University Nijmegen / VU
University Amsterdam)
Prof.dr. M. Bovens (Utrecht University)
Prof.dr. W. van der Brug (University of Amsterdam)
Prof.dr. P. Scheepers (Radboud University Nijmegen)
The research presented in this book has been funded by the NWO Graduate Training Program
Grant (2008/2009) awarded to the research school Interuniversity Center for Social Science
Theory and Methodology (ICS).
Cover design
Cover image
Printed by
ISBN
Anke Struijs
Dreamstime LLC
Ridderprint BV
978-90-393-6271-6
© 2015 Tim Immerzeel
All rights reserved. No part of this publication may be reproduced or transmitted in any
form of by any means, electronic or mechanical, including photocopy, recording, or any
information storage or retrieval system, without permission in writing the author. The
copyright of the articles that have already been published, has been transferred to the
respective journals.
Voting for a change
The democratic lure of populist radical
right parties in voting behavior
Stemmen voor de verandering
De democratische lokroep van populistisch radicaal
rechtse partijen in stemgedrag
(met een samenvatting in het Nederlands)
PROEFSCHRIFT
ter verkrijging van de graad van doctor aan de Universiteit Utrecht
op gezag van de rector magnificus, prof.dr. G.J. van der Zwaan,
ingevolge het besluit van het college voor promoties in het openbaar te
verdedigen op vrijdag 6 februari 2015 des middags te 12.45 uur
door
Timotheus Immerzeel
geboren op 30 juni 1986
te Vlaardingen
Promotor: Prof.dr.ir. A.G. van der Lippe
Co-promotoren: Dr. E. Jaspers
Dr. M. Lubbers
“Het blijft een onomstotelijke wet van de
geschiedenis dat zij juist tijdgenoten verbiedt de
grote bewegingen die hun tijd bepalen al in hun
vroegste stadium te herkennen.”
Stefan Zweig (De wereld van gisteren, p.346)
Voor Erik
Contents
Chapter 1. Introduction
9
1.1 The democratic lure of European populist radical right parties
11
1.2 Defining the populist radical right
16
1.3 Three waves of populist radical right research
17
1.4 Research problem and contributions of the dissertation
25
1.5 Research questions
27
1.6 Data
32
Chapter 2. Competing with populist radical right parties
37
2.1 Introduction
39
2.2 Theory and hypotheses
41
2.3 Data and methods
44
2.4 Results
47
2.5 Conclusion
53
Part 1. Populist radical right parties and non-voters
Chapter 3. Populist radical right parties mobilizing ‘the people’?
59
3.1 Introduction
61
3.2 Theory and hypotheses
63
3.3 Data and methods
67
3.4 Results
72
3.5 Conclusion
79
Chapter 4. A Dutch ‘populist radical right revolt’?
83
4.1 Introduction
85
4.2 Theory and hypotheses
87
4.3 Data and methods
91
4.4 Results
98
4.5 Conclusion
104
Part 2. Populist radical right parties and cleavage voters
Chapter 5. Religion and populist radical right voting
109
5.1 Introduction
111
5.2 Theory and hypotheses
113
5.3 Data and methods
115
5.4 Results
118
5.5 Conclusion
122
Chapter 6. Gender and populist radical right voting
125
6.1 Introduction
127
6.2 Theory and hypotheses
129
6.3 Data and methods
132
6.4 Results
138
6.5 Conclusion
144
Chapter 7. Conclusion
147
7.1 Introduction
149
7.2 Main findings
149
7.3 Main conclusions
157
7.4 Contributions
160
7.5 Limitations and suggestions for further research
162
7.6 Concluding remarks
164
Appendices
167
Dutch summary (Nederlandse samenvatting)
191
References
203
Acknowledgements (Dankwoord)
217
Curriculum vitae
223
ICS Dissertation series
227
Chapter 1
Introduction
9
Introduction
1.1 The democratic lure of European populist radical right parties
It was only four days after his assassination that thousands of people went to see Pim
Fortuyn’s funeral procession in his Dutch hometown of Rotterdam on May 10, 2002. The
right-wing populist Fortuyn was shot by Green activist Volkert van der Graaf and his death
had a deep impact on Dutch society and its citizens. The Dutch public TV broadcaster NOS
reported that more than 1.5 million people—10% of the Dutch population—followed the
procession live on television. Only 5 days after the funeral, Fortuyn’s political party Lijst
Pim Fortuyn [List Pim Fortuyn, LPF] won 26 seats in the Dutch parliamentary elections,
representing 17% of the Dutch electorate.
This remarkable political event has often been labeled a turning point in Dutch politics
and has received significant academic attention (e.g., Couwenberg, 2004; De Lange, 2012a;
Koopmans & Muis, 2009; Lucardie & Voerman, 2012; Van der Brug, 2003; Van Holsteijn &
Irwin, 2003). Not only was Fortuyn’s murder the first Dutch political assassination since the
17th century, Pim Fortuyn was also the first Dutch populist radical right (PRR) leader who
was able to attract a large number of people as the political ‘new kid on the block’ (Krouwel
& Lucardie, 2008). Accordingly, the event raised questions about why the PRR was able to
break through so fast in the Dutch political arena.
Similar emergences of PRRs have been observed in other Eastern and Western European
countries in recent decades. In 1999, many European citizens and the European Union were
shocked by the sudden popularity of Jörg Haider’s Freiheitliche Partei Österreichs [Freedom
Party of Austria, FPÖ], which won somewhat more than 21% of the vote. For many, the
formation of an FPÖ government with the mainstream Christian-democratic party Österreichische Volks Partei [Austrian People’s Party, ÖVP] was the straw that broke the camel’s
back. To paraphrase Stefan Zweig, who looked back upon the rise of right-wing radicalism
in the early decades of the 20th century in his autobiographic Welt von Gestern: “it felt as if
the ‘Golden Age of Security’ had ended” (Berezin, 2011; Zweig, [1943] 2011). The fear of the
FPÖ culminated in an Austrian boycott by some European Union member states.
More recent experiences in Finland and the Netherlands have shown that PRRs are
still able to electorally surprise contemporary democracies. The Finnish Perussuomalaiset
[(True) Finns Party, PS] experienced their electoral breakthrough in 2011, when it won
almost 20% of the vote and became the third party in the parliamentary elections, just
after the social democrats. In the Netherlands, the Dutch Partij van de Vrijheid [Party of
Freedom, PVV] made a similar breakthrough in 2010 when it won more than 15% of the
vote, resulting in a strong negotiation position and becoming a (half-)member of a Dutch
coalition. Although Wilders’ entrance into a supporting-government was not followed by
a European boycott, people feared his popularity and labeled him the embodiment of 21stcentury fascism (e.g., Riemen, 2010). Some groups called for legal action against Wilders,
stating that his ideology and rhetoric encouraged hatred and disruption. Despite the fact
11
Chapter 1
that Wilders was acquitted of all charges, his rhetoric was deemed to be on the edge of
legality and judges warned him to be careful.
In other Western European countries, comparable parties have joined the government
(e.g., Akkerman, 2012; De Lange, 2012b). See Table 1.1 and Figure 1.1 for an overview of
the popularity of the PRR’s popularity in Europe in 2014 and the various PRRs that have
taken office during the last two decades. Due to the PRR’s persistent popularity and the
more responsible roles that these parties have taken in European democracies, it is safe
to expect that these PRRs are here to stay (Zaslove, 2008). Consequently, research into
this successful party family has grown immensely over the last three decades and could be
labeled a new research tradition (see for overviews Backes & Moreau, 2012; Mudde, 2007;
Rydgren, 2007; Van der Brug & Fennema, 2007).
The presence and popularity of the PRR have caused pundits and researchers to question
and investigate its political consequences. On the one hand, some argue that the PRR may
threaten democracy because its political style and ideological program fuels polarization
and distrust of politics and politicians (e.g., Oosterwaal & Torenvlied, 2010). Consequently,
people may turn away from politics, resulting in lower electoral turnout (cf. Ansolabehere
& Iyengar, 1995). Decreasing electoral turnout is generally interpreted as bad for democracy because it calls into question the representativeness and legitimacy of the legislature
(Mair, 2006a; Mair, 2006b; Van Reybrouck, 2013). On the other hand, scholars have suggested potential positive effects of PRR success on electoral turnout. It is argued that the
PRR’s populist, nativist program fosters electoral competition and gives citizens the ability
to become passionately involved in politics, thus leading to higher electoral turnout (e.g.,
Mouffe, 2005). Although the heavily debated question of whether PRRs constitute a ‘threat
or corrective’ to democracy is often drawn into the normative, the question of whether
there is a relationship between the popularity of the PRR and voting behavior is a theoretical question that can be empirically tested. However, there has been scant research on the
question of the political behavioral consequences of the PRR. Do PRRs, indeed, play a role
in encouraging or discouraging citizens to go to the polling booth, and why? Moreover,
will those people, who are known to vote for the same parties repeatedly, reconsider their
habitual vote?
The aim of this study is to add to the emerging research tradition on the PRR by providing insight into the effect of the PRR’s popularity on citizen’s voting behavior. In recent decades, the PRR has become the proponent of a nativist ideology that greatly contributes to
a cultural conflict between adherents of mono- and multicultural societies that has recently
changed contemporary European political arenas (Berezin, 2011; Kriesi et al., 2008; Van
der Brug & Van Spanje, 2009). Together with a traditional economic conflict, this cultural
conflict structures the most important ‘axes around which policy positions have crystal-
12
Introduction
lized’ at both the citizen and the party levels, thus changing the party system (Pellikaan, De
Lange, & Van der Meer, 2007: p. 295).
In addition to the PRR’s achievement in contributing to party system change, the PRR
in Europe is seen as the most successful source of a populist zeitgeist (Mudde, 2004). PRR
leaders have spread an ideology that places the homogenous, ‘pure’ people in an antagonistic relationship with the ‘corrupt, untrustworthy’ elite and the cultural ‘others’ (Abts &
Rummens, 2007; Canovan, 1999; Mudde, 2007; Taggart, 2000). PRR leaders have consistently argued that the political elite has not listened to the widespread grievances of the
electorate. Moreover, the PRR has opposed the consensual style of politics (Mouffe, 2005),
has spoken the same language as the ‘pure and good’ people (Van Leeuwen, 2009) and has
blamed the elite for failing to properly address the issues that threaten the common man:
cultural diversity, globalization, immigration, integration and Europeanization (Berezin,
2011; Betz, 1994). Despite the fact that leftist and centrist populist parties are also emerging
in some Eastern and Western European countries (March, 2011; March, 2007; Učeň, 2007),
populism has disproportionately attached itself to the European radical right. The populist
left party family has been a marginal movement in most European democracies and unlike
the PRR, has not made much of a contribution to this new cultural conflict. Accordingly, I
study the effect of the success of the European populist radical right on voting behavior, i.e.,
the decision of whether to vote (turnout) or to reconsider one’s usual vote (cleavage voting).
So far, only anecdotal evidence tends to underline the hypothesized relationship between
the popularity of the PRR and voting behavior. For instance, the Dutch Fortuyn revolution
(Couwenberg, 2004) has been said to strongly influence polarization and political turmoil
(Van Holsteijn & Irwin, 2003; Vossen, 2010), and electoral turnout increased from 73%
in 1998, before his popularity, to 79% in 2002, after his impressive increase in popularity
(Koopmans & Muis, 2009). Although this corroborates the hypothesized positive relationship between the success of the PRR and electoral turnout, a more comprehensive study
of the influence of the PRR in both Eastern and Western European countries on voting
behavior is lacking. This study is a first step toward filling this theoretical and empirical
lacuna.
By providing theoretical and empirical scrutiny of the debate of the impact of the PRR
on voting behavior, I speak to Robert Jansen’s claim that ‘sociologists cannot longer afford
to sidestep the analytical challenges […] of the [academic] problem of populism’ (Jansen,
2011: p.75). Despite the fact that much is known about the PRR’s electoral sociology (Arzheimer, 2012), little is known about the impact of these parties on the voting behavior of
various social groups. Inspired by Jansen’s definition of populism as a type of political practice that uses populist rhetoric to attract people to the political (Jansen, 2011), I investigate
the extent to which the PRR is able to attract disengaged people to politics and whether it is
able to stimulate people to reconsider their voting behavior.
13
14
The Future of Austria
BZÖ
Date
latest elections
Table 1.1 List of PRRs in 35 European countries in the period 2000-2014, their popularity in 2014 and government experience
Name (in English)
Party (Abbr.) Name
% of popular
vote in
latest elections
LN
Northern League
9,1%
4,1%
0,8%
TT - Tvarka ir teisingumas
Tvarka ir teisingumas
Order and Justice
Order and Justice
ADR Alternativ Demokratesch
Alternativ Demokratesch
Reformpartei
Alternative Democratic Reform Party
Reformpartei Order and
Alternative
Tvarka ir teisingumas
Justice Democratic Reform Party
7,3%
PVV Partij voor de
Partij
voor de Vrijheid
Party for Freedom
Vrijheid
Party for Freedom
LPF Lijst Pim Fortuyn
Lijst Pim Fortuyn
List Pim Fortuyn
List Pim Fortuyn
ToN Trots op Nederland
Trots op Nederland
of the Netherlands
Proud of theProud
Netherlands
Lega Nord
‫האיחוד‬
HaIhud HaLeumi
National
Union/
Jewish Home
HaIhud
HaLeumi
National
Union/ Union/
Jewish
Home
‫הלאומיהאיחוד הלאומי‬
HaIhud
HaLeumi
National
Jewish
Home
Italy FN
FN
La Destra–Fiamma
Tricolore
The Right–Tricolour
Flame
La
Destra–Fiamma
Tricolore
The
Right–Tricolour
Flame
FN
La Destra–Fiamma Tricolore
The Right–Tricolour Flame
LN Lega Nord Lega Nord
Northern League
LN
Northern League
LatviaLatvia
Latvia
Lithuania Lithuania
TT
Luxembourg
Luxembourg
ADR
Lithuania
TT
NetherlandsNetherlands
PVV
LPF
ToN
Italy
Italy
Austria
Belgium
Country
2000-2005
Governingperiod
-
1
1
12,3%
17-04-2011
- 2001-2011*2001-2011
12,3%
15-09-201115-09-2011
0,4%
06-03-2011
0,4%
10-06-2012a06-03-2011 19,1%
17-04-2011
19,1%
22-09-2013b17-04-2011 a
a
13,4%
10-06-2012
13,4%
10-06-2012
b
b
22-09-20130,1%
b
0,1%
22-09-201322-09-2013
b
b
17-06-2012
1,5%
1,5%
22-09-201322-09-2013
1,6%
17-06-2012
06-04-2014
2013-present
1,6%
17-06-2012
16,7%
06-04-2014
2013-presen
16,7%
27-04-201306-04-2014 -2013-present
27-04-2013
27-04-2013
25-02-2011
25-02-201125-02-2011
22-01-2013
2001-2003
23,3%
23,3%
22-01-201322-01-2013
2001-2003 2001-200
2006-present
2006-presen
2006-present
9,1%
22-01-2013
2009-presen
22-01-2013
2009-present
9,1%
22-01-2013
2009-present
0,8%
25-02-2013
0,8%
25-02-2013
25-02-2013
4,1%
4,1%
25-02-201325-02-2013
2001-2006 2001-200
25-02-2013
2001-2006
2008-2011 2008-201
- 2008-2011
17-09-2011
17-09-2011
7,3%
17-09-2011
7,3%
28-10-201228-10-2012
6,6%
20-10-2013
6,6%
28-10-201220-10-2013 10,1%
10,1%
12-09-201212-09-2012
2010-2012*2010-2012
0,0%
200
0,0%
12-09-201212-09-2012 2002
0,0%
0,0%
12-09-201212-09-2012
-
06-03-2011
% of popular
% of popular
29-09-2013
vote in 2005-2006
Date
Governing
vote in
Date
Governinglatest
elections
latest elections
perio
7,8%
13-06-2010
latest elections
latest elections
period
20,5%
29-09-2013
2000-200
20,5%
0,5%
13-06-201029-09-2013 - 2000-2005
3,5%
3,5%
29-09-201329-09-2013
2005-2006 2005-200
7,3%
12-05-2013
7,8%
7,8%
13-06-201013-06-2010
3,0%
04-12-2011
0,5%
0,5%
13-06-201013-06-2010
7,3%
22-05-2011
7,3%
12-05-201312-05-2013
3,0%
3,0%
04-12-201104-12-2011
26-10-2013
22-05-2011
22-05-2011
12,3%
15-09-2011
2001-2011*
26-10-2013
26-10-2013
3,5%
Country
Party
(Abbr.)
Name
Name (in English)
VB
Vlaams
Belang (Flanders)
FlemishName
Interest
Party
(Abbr.)
Name
(in English)
AustriaFPÖ
FPÖ FreiheitlicheFreiheitliche
Partei Österreichs
Freedom Party of Austria
Partei Österreichs
Freedom
FN
Front National (Wallonia)
National
Front Party of Austria
BZÖ Bundnis Zukunft
Bundnis
Zukunft Österreichs
Future of Austria
BZÖ
Österreichs
The Future The
of Austria
Bulgaria
ATAK
Săyuz
Ataka Belang (Flanders)
National Union Attack
VBNatsionalen
Vlaams
Flemish Interest
Belgium Belgium
VB
Vlaams Belang
(Flanders)
Flemish Interest
Croatia
HSP
StrankaFront
PravaNational (Wallonia)
Croation Party of Rights
FN Hrvatska
National Front
FN
Front National
(Wallonia)
National Front
ATAK
National
Union Attack
Cyprus
- NatsionalenNatsionalen
Bulgaria Bulgaria
ATAK
Sǎyuz AtakaSǎyuz Ataka
National Union
Attack
Croatia
HSP
Hrvatska
Stranka
Prava
Croation
Party of Rights
Croatia
HSP
Croation Party of Rights
Czech
Republic
- Hrvatska Stranka Prava
CyprusCyprus
Denmark
DFp
Dansk Folkeparti
Danish People’s Party
CzechCzech
Estonia
EIP
Eesti Iseseisvuspartei
Estonian Independence Party
0,4%
Republic
Republic
DFpPerussuomalaiset
Dansk Folkeparti
Danish
People's Party
Finland
PS
(True) Finns
Party
19,1%
Denmark Denmark
DFp
Dansk Folkeparti
Danish
People's
Party
EstoniaEIP
EIP Eesti Iseseisvuspartei
Eesti Iseseisvuspartei
Estonian Independence Party
Estonia
Estonian
France
FN
Front National
National
Front Independence Party
13,4%
Finland
PS
Perussuomalaiset
Finland
PS
(True) Finns(True)
Party Finns Party
Germany
REP
Die Perussuomalaiset
Republikaner
The Republicans
0,1%
FranceFN
FN
Front National
National Front
France
Front National
National Front
NPD
Nationaldemokratische
Partei
Deutschlands
National
Democratic
Party
of
Germany
1,5%
REP Die Republikaner
Die Republikaner
The Republicans
Germany Germany
REP
The Republicans
Greece
LAOS
Laikós
Orthódoxos
Synagermós
The People’s
Orthodox
Rally Democratic
1,6%
NPD
Nationaldemokratische
Partei Deutschlands
National
Party of Germany
NPD
Nationaldemokratische
Partei Deutschlands
National
Democratic
Party of Germany
GreeceLAOS
LAOS
Laikós Orthódoxos
People'sRally
Orthodox Rally
Hungary
Jobbik
Jobbik
Jobbik The People'sThe
16,7%
Greece
Laikós Orthódoxos
SynagermósSynagermós
Orthodox
Jobbik Jobbik
Jobbik
Jobbik
Hungary Hungary
Jobbik
Jobbik
Iceland
IcelandIceland
Ireland
IrelandIreland
Israel
Beiteinu Yisrael Beiteinu
Israel Our Home Israel Our Home
23,3%
Israel ‫ל‬
‫ ל‬Yisrael
Israel
Yisrael Beiteinu
Israel Our Home
Bundnis Zukunft Österreichs
Table 1.1 List ofFPÖ
PRRs in 35Freiheitliche
EuropeanPartei
countries
in the period 2000-2014,
experience
Austria
Österreichs
Freedomtheir
Partypopularity
of Austria in 2014 and government
20,5%
29-09-2013
Country
One Table:
Table 1.1
One Table:
Table 1.1
Chapter
Chapter
1 1
Table 1.1 List of PRRs in 35 European countries in the period 2000-2014, their popularity in 2014 and government experience
Chapter 1
Slovenska nacionalna stranka
Lipa
SNS
Lipa
Slovenia
Srpska radikalna stranka
(DE) Schweizerische Volkspartei
(FR) Union Démocratique du Centre
(IT) Unione Democratica di Centro
(IT) Lega dei Ticinesi
(DE) Schweizer Demokraten
(FR) Démocrates Suisses
(IT) Democratici Svizzeri
Milliyetçi Hareket Partisi
SVP
LdT
Dem
Switzerland
British National Party
United Kingdom Independence Party
BNP
UKIP
United Kingdom
0,0%
UK Independence Party
British National Party
All-Ukrainian Union ‘Freedom’
3,1%
1,9%
10,4%
13,0%
0,0%
Swiss Democrats
Nationalist Movement Party
0,8%
26,6%
Swiss People’s Party
Ticino League
5,7%
-
0,0%
1,8%
4,6%
4,6%
11,7%
1,5%
0,3%
0,0%
16,3%
0,0%
Sweden Democrats
-
The Lime Tree Party (Linden)
Slovenian National Party
Slovak National Party
Serbian Radical Party
Liberal Democratic Party Russia
Greater Romania Party
National Renovator Party
League of Polish Families
Progress Party
6,6%
10,1%
20-10-2013
06-05-2010
06-05-2010
28-10-2012
12-06-2011
23-10-2011
23-10-2011
23-10-2011
19-09-2010
04-12-2011
04-12-2011
10-03-2012
06-05-2012
04-12-2011
09-12-2012
05-06-2011
09-11-2011
09-09-2013
12-09-2012
12-09-2012
12-09-2012
-
-
-
-
-
-
-
2000-present
-
-
-
-
2006-2010
-
-
-
-
2006-2007
2001-2005*
-
2002
2010-2012*
Sources: Akkerman, 2012; De Lange, 2012b; Norwegian Social Science Data Services (NSD), 2014.
Notes: *Supporting member of a minority cabinet.
a
Based on the vote in the first round; b Based on the first constituency ballot.
Some PRRs did not receive any votes at the latest elections. However, they are included in this table as they were more popular before and, thus, have been included in some
of the following chapters.
Vseukrayinske obyednannia ‘Svoboda’
MHP
Svoboda
Turkey
Ukraine
Sverigedemokraterna
SD
Sweden
-
-
Spain
Slovenská národná strana
SRS
SNS
Partidul România Mare
Liberal’no-Demokraticheskaya Partiya Rossii
PRM
LDPR
Romania
Russian Federation
Serbia
Liga Polskich Rodzin
Partido Nacional Renovador
LPR
PNR
Poland
Portugal
Slovakia
Fremskrittspartiet
FrP
Norway
List Pim Fortuyn
Proud of the Netherlands
Lijst Pim Fortuyn
Trots op Nederland
LPF
ToN
Alternative Democratic Reform Party
Party for Freedom
Alternativ Demokratesch Reformpartei
Partij voor de Vrijheid
ADR
PVV
Luxembourg
Netherlands
Introduction
15
Chapter 1
Figure 1.1 The popularity of Populist Radical Right parties in Europe 2014
Source: Norwegian Social Science Data Services (NSD), 2014.
1.2 Defining the populist radical right
The stark increase in scientific publications on the PRR has led to much debate about what
constitutes the radical right. Parallels with fascist movements in the 1930s abound; scholars, politicians, pundits and the public use the terms far-right, extreme-right, right-wing
extremist, and radical right interchangeably; extreme-right activism and radical-right voting are seen as two sides of the same coin; and a wide array of core characteristics has been
given to these parties, ranging from merely anti-immigrant (e.g., Dahlström & Sundell,
2012; Van der Brug, Fennema, & Tillie, 2005) to nativist, authoritarian, populist or even
more (Mudde, 2007).
Despite the difficulty in defining the core characteristics of this new party family,
scholars often refer to the same parties (Rydgren, 2007). In this section, I introduce three
characteristics that are generally acknowledged as defining features of the PRR family
(Mudde, 2007; Norris, 2005; Rydgren, 2007): nativism (anti-immigrant and nationalism),
authoritarianism and populism. Nativism refers to the ideology that natives should be the
sole inhabitants of a nation-state and ‘nonnative elements (ideas, persons) are fundamen16
Introduction
tally threatening to the homogeneous nation-state’ (Mudde, 2007: p.19). Nativism is a
combination of a preference for the native and xenophobia (a fear/hatred of everything that
is non-native). In contrast to pure notions of nationalism or racism, it does not encompass
a hierarchy of ethnicities with one ethnicity being superior to others. Rather, it embraces
the idea that people from different cultures and ethnicities cannot live together: societies
should be mono-ethnic or mono-cultural (Fennema, 2005; Mudde, 2007; Rydgren, 2007).
Consequently, immigrants with a different socio-cultural background are not necessarily
regarded as inferior but are seen as incompatible with the mainstream majority culture, and
should therefore leave the country to live in a country with a compatible culture.
In addition to nativist ideology, the PRR is also known for its authoritarian appeal and
strong emphasis on law and order. Mudde (2007: p.23) notes that an important part of
the PRR’s ideology is the ‘belief in a strictly ordered society in which infringements of
authority are to be punished severely’. Following Altemeyer’s famous F-scale, authoritarianism can be viewed as individuals’ social-psychological disposition to be submissive to and
respectful of authorities and to be aggressive against those who deviate from the authority’s
norms and values (Altemeyer, 2005). Furthermore, scholars have mentioned populism as
the third characteristic of the PRR-family. Populism is defined as a thin ideology that states
that society is divided into two homogenous groups: ‘pure people’ with a common will that
is embodied by the PRR and the untrustworthy ‘corrupt elite’ formed by the established
political parties (Jagers, 2006). Based on these defining features, and in line with other
studies addressing the categorization of the PRR, I made a selection of political parties
that belong to the PRR-family (Ennser, 2012; Mudde, 2007; Norris, 2005; Rovny, 2013). In
Table 1.1, I present an overview of these PRRs, their popularity in 2014 and experience with
governance in the various Eastern and Western European countries.
1.3 Three waves of populist radical right research
In a recent review of the PRR, Backes and Moreau (2012: p.9) note that the ample scientific
literature about the PRR has become almost unmanageable for scholars. Three decades of
research on the PRR has formed the basis of a research tradition that has made great steps
toward unveiling the social, political and cultural basis of these parties, thus rendering the
PRR a ‘terra cognita’ (cf. Backes & Moreau, 2012). This research tradition into the PRRs can
be largely divided into three waves of research with different research questions and theoretical avenues. These three waves did not necessarily happen in succession chronologically,
but they can be seen as signs or routes into the abundant literature on the PRR.
17
Chapter 1
1.3.1 First wave: Defining the PRR and the demand-side explanations
The first wave in PRR research has primarily focused on the definition and the demand-side
explanations for these parties. Whereas in earlier works scholars drew parallels with fascist
movements and parties (e.g., Eatwell, 2000; Ignazi, 1992; Ignazi, 2003), in later definitions
scholars have taken a more nuanced view, especially compared to some pundits, and have
distanced themselves from comparisons to the fascist movements and parties of the 1930s
(e.g., Mudde, 2007). An increasing number of scholars have demarcated the PRR family as
a distinct party family that may have some similarities to the earlier fascist parties but that
cannot be regarded as neo-fascist.
In addition to this problem of definition, the first wave of research encompasses
demand-side explanations for why the PRR has emerged and become popular (Rydgren,
2007). The essential theories of this first wave depart from the notion that certain macrostructural processes changed the attitudes, values, and interests of the European electorates,
which in turn sparked voter demand for the PRR. Three macro-structural processes have
been labeled as pivotal to understanding the emergence and success of these parties: globalization/modernization (e.g., Berezin, 2011; Betz, 1994; Ignazi, 1992), immigration (e.g.,
Fennema, 2005; Olzak, 1992) and widespread political discontent (e.g., Lubbers, Gijsberts,
& Scheepers, 2002; Norris, 2005).
Two theories that explicitly refer to the processes of modernization and globalization
are the anomie or social disintegration theory (Arendt, [1951] 1973; Kornhauser, 1959) and
the losers of modernization theory (Betz, 1994; Kriesi et al., 2008). The former, Durkheimian theory is directly linked to traditional mass society theories that modernization and
urbanization cause people to increasingly disintegrate from their communities, families
and societies. Due to these crumbling social structures, isolated individuals attempt to cope
with their experienced loss of identity by replacing it with other (metaphysical) identities,
such as the ethnic-national (Fennema & Tillie, 1998). Isolated individuals are available to
nativist movements and parties such as the PRR, which may provide them with a ‘quasicommunity’ or alternative feeling of belonging and identity (Arendt, [1951] 1973; Fennema
& Tillie, 1998). Thus, scholars studying the PRR’s electorate hypothesize that those voters
with fewer social contacts, who are less integrated and who are not members of any social,
religious or political organization will be more likely to vote for PRRs (Rydgren, 2009;
Rydgren, 2011; Zhirkov, 2014). The studies by Rydgren (2009, 2011) and Zhirkov (2014),
however, show that in both Eastern and Western European countries, this theory does not
hold water: they do not find differences in the propensity to vote for a PRR between those
with either a small or a large stock of social capital.
The ‘losers of modernization’ theory also departs from the notion that modernization
and globalization are the major drivers of the PRR’s success (Betz, 1994; Ignazi, 1992).
According to the ‘losers of modernization’ theory, globalization leads to the decay of
18
Introduction
traditional social structures, with people becoming increasingly individualized. Globalization gives way to a ‘flux of contextualized identities’ and pressures traditional institutions,
norms and values (Betz, 1994: p.29). Moreover, globalization fundamentally threatens
national economies as multi-national industries and firms transfer their primary manual
activities to cheaper (non-Western) countries. This means that industrial and blue-collar
workers are more likely to lose their jobs or to be forced into lower-paying jobs in the new
global economy (Givens, 2004). The theory postulates that these processes lead to a greater
cultural and economic threat that has divided society into two groups: the ‘winners’, who
are able to cope with these developments and applaud the new configuration; and the ‘losers’, who feel insecure and threatened by such developments (Betz, 1994; Kriesi et al., 2008).
Consequently, the losers are expected to demand PRRs that propose anti-globalization
(and anti-European Union), nationalist and culturally conservative measures. This leads to
generally supported hypotheses that less educated, unemployed, self-employed, blue-collar
workers and men are more likely to vote for the PRR (e.g., Fontana, Sidler, & Hardmeier,
2006; Givens, 2004; Lubbers et al., 2002).
These modernization theories aside, ethnic competition theory focuses on immigration
as an explanation for the emergence and popularity of the PRR (Coenders, Lubbers, &
Scheepers, 2006; Lubbers et al., 2002; Olzak, 1992; Rydgren, 2007). The ethnic competition theory asserts that with the higher influx of (non-Western) immigrants into European
societies, immigrants are increasingly perceived by natives as competitors in the (labor,
housing or marriage) market. Being ‘in competition’ with immigrants over scarce resources
will trigger exclusionary reactions because of feelings of being threatened (Fennema, 2005).
These reactions may result in a stronger demand for the PRR, which favors nativist policies to reduce competition from immigrants for scarce resources by rejecting their equal
access to resources (Olzak, 1992). Based on this theory, various scholars have hypothesized
that in regions with more immigrants and higher levels of unemployment, the PRR will be
more popular (e.g., Coffé, Heyndels, & Vermeir, 2007; Knigge, 1998; Lubbers et al., 2002).
These studies consistently find that in regions (i.e., countries, provinces, neighborhoods)
where more (non-Western) immigrants live, people are indeed more likely to vote for the
PRR than in other regions. Coffé and others (2007) find that it is primarily the presence
of Muslim immigrants in Belgian regions that is a positive predictor for PRR-voting. The
results related to unemployment rates are mixed: most studies do not find a positive effect
of the regional unemployment rate on the likelihood of voting for the PRR (e.g., Lubbers et
al., 2002). Golder (2003) concludes in his cross-national comparative study of 165 elections
in 19 countries that unemployment only matters when the level of immigration is high.
Moreover, scholars have tested expectations about groups that are the most likely to be
in ethnic competition over scarce resources, i.e., less educated, unskilled workers and unemployed men with strong nativist attitudes who aim for the same jobs and social benefits
19
Chapter 1
as the immigrants (e.g., Givens, 2004; Ivarsflaten, 2008; Lubbers & Scheepers, 2000). In line
with ethnic competition theory, Ivarsflaten (2008) finds that anti-immigrant attitudes are
the most robust and most important predictor of PRR-voting in all countries under study.
With respect to structural factors, various scholars find support for the ethnic competition
thesis that the social base of the PRR’s electorate is formed by men who are less educated
workers and are over-represented in the unskilled blue-collar sector (e.g., Arzheimer, 2012;
Lubbers et al., 2002).
This first wave theorized that widespread political discontent and dissatisfaction with
democracy also fueled the emergence and popularity of the PRR, regardless of the ideological incentives (Hooghe, Marien, & Pauwels, 2011; Lubbers et al., 2002; Van der Brug, 2003).
The reason for the electorate’s political dissatisfaction often relates to the abovementioned
macro-societal processes, which have not been adequately addressed by the established,
mainstream parties. As a result, citizens feel abandoned by the political elite and have developed negative attitudes toward the mainstream, established political parties (cf. Betz, 1994).
Moreover, stealth democracy theory postulates that not politics and democracy per se but
also democratic procedures have increasingly fueled political dissatisfaction (Hibbing &
Theiss-Morse, 2002). According to stealth democracy theory, citizens feel increasingly unable to control their country’s democratic functioning and perceive a mismatch between
how political decisions ideally should be made versus the practical reality. Ordinary citizens
believe that experts are too decisive and that the opportunity to control the decision-making
process is too small (ibid.). This mismatch makes ordinary citizens more negative toward
government processes and could lead to demands for the PRR that may function as watchdog for the vox populi and a promise to ‘protect’ the people from the ‘corrupt’ mainstream
parties (Mudde, 2004). Therefore, various scholars expect that politically dissatisfied people
who feel abandoned by their political elites are more likely to vote for the PRR (e.g., Lubbers
& Scheepers, 2000; Lubbers et al., 2002; Schumacher & Rooduijn, 2013). Although Van der
Brug (2003) does not find political protest as important predictor of voting for Fortuyn,
various studies provide evidence that political dissatisfaction is indeed an important predictor of PRR-voting (e.g., Hooghe et al., 2011; Schumacher & Rooduijn, 2013).
Despite the fact that individual demand-side explanations have found evidence for most
of their hypotheses, scholars have pointed to problematic issues. Why did the PRR become
more popular in some countries than in other countries? Why is there no credible popular
PRR in the two largest European countries, Germany and the United Kingdom, where globalization, modernization and political protest are also flourishing? Why did some of these
PRRs become popular in the 1980s, whereas PRRs in other countries did not become popular until the 2000s? To illustrate, Koopmans and Muis (2009) ask why Fortuyn was able to
break through in Dutch politics in only six months in 2001, whereas his Dutch predecessor
Janmaat had not been able to become popular six years earlier. Macro-structural conditions
20
Introduction
(i.e., immigration levels, globalization and political protest) did not change much over that
period, so why in 2001 but not in 1995?
1.3.2 Second wave: Supply-side explanations
Researchers increasingly acknowledge that looking only at individual demand-side factors did not provide a full understanding of the emergence and popularity of the PRR.
Individual factors alone could not explain why the PRR was successful in some countries
but not in others during certain periods. This was the impetus for second-wave studies
to incorporate supply-side elements to explain cross-national and over-time variations in
the PRR’s success (e.g., Art, 2006; Arzheimer & Carter, 2006; Arzheimer, 2009; Kitschelt &
McGann, 1995; Koopmans & Muis, 2009). These supply-side elements generally refer to
arguments about the PRR’s party program, a country’s political opportunity structure or
political space and the internal party organization of the PRR itself (e.g., Art, 2006; Art,
2011; Berezin, 2011; Rydgren, 2007). I note that most of the presented arguments complement the first-wave explanations. However, I stress that these second-wave explanations
generally take the individual explanations as constant and focus on party politics and the
political opportunities of the PRRs related to other parties’ positions and the political system in which these parties operate.
Supply-side explanations that address the political opportunity structure primarily
employ spatial analyses of the political arena to investigate the conditions under which
new political parties have the potential to emerge (e.g., Dahlström & Sundell, 2012; Tarrow,
1998). For instance, it is easier for new parties to emerge if voters are generally volatile in
their voting behavior than if all voters are captured in their own social or religious groups
that have their own parties for which they consistently vote (e.g., Aarts & Thomassen, 2008;
Lipset & Rokkan, 1967; Mair, 2008). Because the arguments of this approach apply to the
emergence of all possible new parties, PRR-specific arguments have to be postulated to
understand why it is that exactly the PRR has benefited from certain opportunity structures
(Rydgren, 2007).
Arzheimer and Carter (2006) postulate that there is a particular political opportunity
structure in a particular country at a particular point in time that is formed by (quasi-)
permanent institutional settings, such as the electoral system and the level of federalism (cf.
Swank & Betz, 2003; Van der Brug et al., 2005); and by somewhat more volatile party system
variables, such as the ideological position of the mainstream right party on immigration
and the degree of ideological convergence among established parties (cf. Kitschelt & McGann, 1995). Others have added the increasing importance of the socio-cultural cleavage
dimension (e.g., Kriesi et al., 2008; Pellikaan et al., 2007) and the more prevalent political
role played by the media (e.g., Koopmans & Muis, 2009). The hypotheses derived by these
21
Chapter 1
scholars generally follow the same logic: the more openness in the political opportunity
structure, the more likely that the PRR will emerge and become popular.
To give an example, Arzheimer and Carter (2006) hypothesize that the more the
established parties converge ideologically, the more opportunity for the PRR to position
itself as a real alternative to the established parties, and thus the more popular the PRR
becomes. However, they find no support for this hypothesis. Neither do they find that the
ideological position of the major right-wing party is associated with success of the PRR,
which suggests that these political opportunity variables do not play a role in explaining the
variety of PRR success stories. In contrast, they do find that a grand coalition government is
beneficial for the PRR’s success. Moreover, Koopmans and Muis (2009) find strong support
for the idea that there should be space for the PRR to voice its ideology in the media. When
Fortuyn was more visible in the media and the leaders of other parties expressed sympathy
for his analysis of societal problems and solutions, Fortuyn became more popular with the
electorate. These results underline the notion that the more legitimate and unique the PRR’s
message, the more successful the PRR. Thus, Koopmans and Muis (2009) conclude that the
opportunity structure is pivotal to understanding why Pim Fortuyn was able to attract very
much support in a relatively limited period.
Next to the political opportunity structure, the internal party characteristics of the
PRR have been mentioned as supply-side explanations (most notably: Art (2011)). How
these parties are organized, whether their leaders are charismatic and their level of internal
consensus have all been argued to partly explain why some PRRs are successful and others
are not. The basic tenet of the argument is that once parties are more attractive (with charismatic leaders), better organized and have achieved a trustworthy and stable image, they
become more popular among the electorate. Thus, scholars began to investigate the role
of party characteristics and the charisma of PRR leaders in the PRR’s success (Art, 2011;
Arzheimer & Carter, 2006; Lubbers & Scheepers, 2000; Lubbers, 2001a; Lubbers et al., 2002;
Van der Brug et al., 2005). Based on interviews with activists belonging to PRRs, Art (2011)
concludes that ‘organization matters in particular’: all of his interviewees mentioned professional party organization as ‘the ultimate source of their parties’ successes and failures’ (Art,
2011). These qualitative results are underlined by the study of Lubbers and others (2002),
which shows that variation in the quality of a PRR organization and the level of charisma of
a party leader across different PRRs are important in explaining why some PRRs are more
popular than others in some countries during some time periods.
1.3.3 Third wave: Political and societal consequences of PRR popularity
After three decades of research on the PRR, and the experience of the PRR’s growing
popularity in European democracies, a new, third wave of questions has emerged that
focuses on the consequences of the PRR’s success (e.g., Mudde & Rovira Kaltwasser, 2012).
22
Introduction
As illustrated by the examples in the introduction, much of the public and many political
pundits have problematized the PRR’s impact on democracy. Some PRR leaders have been
summoned and in some cases —including those of the Dutch Centrumdemocraat [Centre
Democrats, CD] Hans Janmaat and the former leader of the French Front National [National Front, FN] Jean-Marie Le Pen— have been found guilty of extremist, anti-democratic
rhetoric that sows hatred and societal division. In Germany, the Nationaldemokratische
Partei Deutschlands [National-democratic Party of Germany, NPD] and the Deutsche Volksunion [German People’s Party, DVU] are under constant surveillance by the government.
In general, the arguments why the PRR threatens democracy are as follows. The PRRideology is nativist, xenophobic, politically untrustworthy and irresponsible (e.g., Kitschelt
& McGann, 1995; Rosanvallon, 2008). Moreover, the PRR ‘reject[s] individual and social
egalitarianism’ and attributes to the people a ‘superior common sense’ (Betz, 1994: p.4).
From this, the PRR derives a notion of the people having a common will, the ‘volonté
generale’, and being homogeneous. This is at odds with pluralism and the protection of
minorities, which have become the constitutional foundations of many European liberal
democracies (Abts & Rummens, 2007). When PRRs become popular, therefore, pundits
and scholars argue that the foundations of liberal democracy are under pressure, even
more so when PRR leaders talk about constitutional changes. When Fortuyn (in 2002)
and Wilders (in 2006) wanted to change the constitution by adapting or erasing the first
non-discrimination article, a polarized debate unfolded about the threat of these parties for
Dutch democracy (Novum/ANP, 2006).
However, there have also been arguments in favor of the PRR, which state that the
PRR should be seen as a corrective for democracy. It is argued that mainstream political
parties in European democracies have become too consensual, rational and pragmatic to
overcome political and social conflicts, thus preventing voters from becoming passionately
involved in politics (Mouffe, 2005). Various scholars claim that contemporary politics and
democracies have become largely depoliticized (Mair, 2006b; Mouffe, 2005). In line with
decreasing numbers of political partisans and members (Van Biezen, Mair, & Poguntke,
2012), voter turnout (e.g., Franklin, 2004) and levels of political interest (Bovens & Wille,
2010), Mair (2006b) even concludes that European democracies have become ‘democracies
without demos’. Scholars postulate that this growing apolitical portion of the electorate may
feel attracted by the PRR, which addresses the problems of the ‘common man’ (referring to
immigration, crime, and globalization), uses clear language (Van Leeuwen, 2009), emphasize popular sovereignty and seeks to restore the ‘will of the people’ (e.g., Canovan, 1999;
Rydgren, 2007). Thus, it is suggested that the PRR may trigger political interest and that its
mere existence and popularity can thus be interpreted as a corrective for democracy.
The potential mobilizing role of the PRR has been one of the hypothesized ‘corrective’ consequences belonging to the third wave of the PRR research tradition (Mudde &
23
Chapter 1
Rovira Kaltwasser, 2012); so far, however, it has received only scant attention. Most of the
attention in the third wave has been paid to ‘threatening’ consequences related to policy
outcomes (e.g., Akkerman, 2012; De Lange, 2012b; Zaslove, 2004), the strategies of other
political parties (e.g., Bale, Green-Pedersen, Krouwel, Luther, & Sitter, 2010; Van Kersbergen & Krouwel, 2008; Van Spanje, 2010) and the social environment in which PRRs emerge
(Mudde, 2012). For example, Akkerman and De Lange (2012) have studied the impact
of the PRR’s electoral success and of being a member of a government on stricter policy
outcomes related to integration and immigration (e.g., Akkerman, 2012; De Lange, 2012b).
Akkerman found that it is not the PRR’s participation in government per se that results in
tougher policy outcomes related to immigration and integration. Rather, she shows that
the presence of a mainstream right wing in office is associated with tougher measures on
immigration and integration (Akkerman, 2012). Still, this could imply that the PRR can
influence policy making indirectly via its impact on other parties’ positions. Following that
line of reasoning, other scholars have investigated whether the PRR’s success influences
other parties’ policy positions (e.g., Bale et al., 2010; Van Spanje, 2010). Remarkably, Van
Spanje (2010) has concluded, based on various expert surveys performed in Western European countries, that in general, all other political parties have become more restrictive
with respect to immigration and integration due to the PRR’s success. Therefore, there is
evidence that the PRRs has an indirect influence on policy outcomes.
Final, some examine more socially ‘threatening’ consequences of the PRR’s success, such
as social polarization, anti-immigrant violence and the fueling of anti-immigrant attitudes
(e.g., Braun, 2011; Dunn & Singh, 2011; Oosterwaal & Torenvlied, 2010). Only a few studies
have addressed the effect of the PRR’s success on political behavior. With respect to political
party membership, Quintelier (2008) finds that over the years, the Belgium PRR Vlaams
Belang [Flemish Interest, VB] has recruited more than twenty-five thousand members. This
suggests that the PRR’s success can positively affect political membership levels. Likewise,
De Lange and Akkerman (2012) show that since 1997, political trust and satisfaction
with democracy in Belgium have tended to increase with the rise of the VB. In addition,
Fallend (2012) concludes that the Austrian FPÖ fueled polarization and addressed issues
neglected by other parties, such as immigration and integration. Accordingly, over the period 1996-2001, the party seemed to give voice to an apolitical part of the electorate, which
increasingly felt that politicians listened to them. Regarding voting behavior, De Lange and
Akkerman (2012) find that electoral turnout numbers in Belgium have decreased since
1997, whereas the VB has become more and more popular, which contradicts the idea that
the PRR attracts apolitical, disengaged people. In contrast, it seems that with the success of
the VB, more people have disengaged from politics.
24
Introduction
1.4. Research problem and contributions of the dissertation
In this dissertation, I will be ‘standing on the shoulders’ of other PRR scholars by combining the three waves of PRR research to understand how the PRR’s popularity relates to
voting behavior. Much of the third wave’s research has paid attention to the consequences
of the PRR in the field of policy outcomes and political party stances related to immigration
and integration (Akkerman, 2012; Van Spanje, 2010; Zaslove, 2004). This is not strange: the
third wave of research bore out the observation that with its increasing popularity, the PRR
began to assume responsible positions in the government (e.g., De Lange, 2012b). Given
the PRR’s nativist ideology, scholars paid attention to the extent to which the PRR was
successful in implementing policies derived from its nativist ideology and thus investigated
whether governments that included members of the PRR introduced tougher policies on
immigration and integration.
However, the PRR’s success might also have consequences for the electorate’s political
behavior. With the PRR’s populist appeal, it declares itself to be the embodiment of the
people and thus claims to ‘know what the people want’. By doing so, the PRR can trigger
politically disengaged people to become actively or passionately involved in politics, leading to higher levels of electoral turnout (e.g., Jansen, 2011; Mouffe, 2005). With researchers
consistently demonstrating that fewer people turn out at elections (Blais & Rubenson, 2013;
Enos & Fowler, 2012; Franklin, 2004; Mair, 2006b) and actively participate in politics (e.g.,
Van Biezen et al., 2012), and keeping in mind that popular political involvement in elections
is highly valued by many political philosophers, politicians, and political scientists (e.g.,
Dahl, 1998; De Tocqueville, [1835-1840] 2000; Mair, 2006a; Putnam, 2000; Rosanvallon,
2008; Thomassen, Van Ham, & Andeweg, 2014; Van Reybrouck, 2013), it is remarkable that
there has been no systematic study into the potential ability of the PRR to attract former
non-voters to the polling booth.
Moreover, if the PRR contributes to replacing traditional social cleavages and conflicts
with a novel cultural conflict (Pellikaan et al., 2007), it poses the theoretical question of
whether it also affects the voting behavior of individuals whose votes are based upon those
traditional conflicts. Do we observe that these ‘cleavage voters’ re-evaluate their traditional
votes and base their voting decisions on this new cultural line of conflict? In other words,
does the PRR influence the propensity of cleavage voters to reconsider their normal or
‘habitual’ voting behaviors (cf. Aldrich, Montgomery, & Wood, 2011; Gerber, Green, &
Shachar, 2003) and is it therefore ‘cleavage-breaking’?
This dissertation will address both the hypothesized success of the PRR in attracting
non-voters to the polling booth and in attracting those voters who are said to be locked up
in their traditional social cleavage. Earlier research has not shown where the PRR’s voters
‘come from’: did they belong to an apolitical part of the electorate triggered by the PRR to
return to voting, as suggested by third wave PRR researchers? Are the PRR’s voters volatile
25
Chapter 1
voters from other parties (cf. Van der Meer, Lubbe, Van Elsas, Elff, & Van der Brug, 2012)?
Conversely, are traditional ‘cleavage-voters’ also lured by the PRR (cf. Rydgren, 2012)? The
overarching research question guiding this dissertation therefore reads as follows:
To what extent does the popularity of the European PRR influence voting behavior, i.e., electoral turnout and cleavage voting?
In this dissertation, I contribute to the PRR research tradition in the following ways. First, I
combine theoretical insights that belong to the various waves of the PRR research tradition.
The focus on the consequences of the PRR’s success for voting behavior is an extension
of the third wave of PRR research that investigates political and social consequences but
simultaneously builds upon theoretical insights from the first and second waves. Demandside theories, such as ethnic competition theory, developed in the first wave, and supply-side
theories addressing (for example) internal party characteristics, developed in the second
wave, can potentially inform our understanding of whose voting behavior is primarily
affected by the rise and popularity of the PRR. An integration of the three waves of PRR
research will help us improve our knowledge to why and the extent to which the PRR is able
to influence the political behavior of eligible citizens.
Second, I provide more comprehensive, empirical scrutiny of the scientific and public
debate about the PRR’s ability to include nonvoting parts of society that have disengaged
from the political realm. Notably, Mudde and Rovira Kaltwasser (2012) have recently
provided a more systematic look at the consequences of populism in Europe and the
Americas for, amongst others, political behavior. The various contributions in their book
are important initial steps toward a better understanding of the PRR’s influence on political
behavior, but due to the exploratory nature of these studies, they provide only weak evidence for the notion that the PRR may have an impact on voting behavior (e.g., De Lange
& Akkerman, 2012; Fallend, 2012). These scholars somewhat loosely show that various
time-varying macro-indicators, such as PRR success and voter turnout, are significantly
related. Therefore, these studies run the risk of the ecological fallacy. On the one hand, it is
possible that with the increasing popularity of the PRR, more nonvoting citizens (re)turn
to the polling booth. On the other hand, it could very well be that the popularity of PRR
affects mainstream voters who feel appalled by the PRRs’ nativist, anti-elitist ideology and
thus are more likely to vote. The PRR’s success then triggers people to speak out and send
a signal to the other parties to keep ‘the rascals out’ (cf. Van Spanje, 2011). Accordingly, a
more comprehensive study is necessary to formulate more far-reaching conclusions about
the PRR’s influence on voting behavior and in particular, on electoral turnout.
Moreover, it is often suggested that the PRR can flourish because of the increasing
number of the population of voters who have detached themselves from traditional social
26
Introduction
and religious cleavages (cf. Inglehart & Norris, 2000; Lipset & Rokkan, 1967) and have
become floating voters (Koopmans & Muis, 2009; Mair, 2008; Van Holsteijn & Irwin, 2003).
However, second- and third-wave research suggests that with the introduction of typical
PRR issues and policy proposals, a new sociocultural cleavage incorporating the issue of
immigration has developed (Pellikaan et al., 2007; Rydgren, 2007). As a result, the PRR has
the potential to form new, ‘catch-all’ parties that attract people from different social strata
and therefore fundamentally change the party system (Oosterwaal & Torenvlied, 2010;
Pellikaan et al., 2007). Thus, if the PRR’s issues play a more prominent role in the political
opportunity structure (e.g., Koopmans & Muis, 2009), the PRR may break through more
traditional social and structural cleavages and have the potential to attract those parts of
the electorate that traditionally vote for other parties (cf. Aarts & Thomassen, 2008). Earlier
first- and second-wave studies have demonstrated that, for example, Christian voters and
women are underrepresented in the PRR’s electorate, which may contradict that hypothesis.
Conversely, studies do also suggest that lower-class workers are lured by the PRR into saying farewell to their traditional left-wing parties (e.g., Oesch, 2008; Rydgren, 2012). Thus,
the PRR may also affect the voting behavior of religious people and women, i.e., cleavage
voters, who are least expected to change their voting behavior. Thus far, third-wave PRR
research has not paid much attention to that possibility.
Third, I take a cross-national and over-time comparative perspective. The earlier case
studies addressing the relationship between the PRR’s success and political behavior neither
show that it is specifically the PRR that drives the change in electoral turnout nor that there
is a general PRR effect on electoral turnout in both Eastern and Western European countries.
It could very well be that along with the changes in PRR popularity, various other changes
in country-specific conditions have had an impact on electoral turnout. Thus, based on the
earlier studies it is difficult to formulate coherent, conclusive claims about the influence of
the PRR’s popularity on electoral behavior. Therefore, I formulate cross-national, over-time
comparisons of the influence of the PRR in both Eastern and Western European countries.
Below, I describe how these contributions are reflected in the various empirical chapters
and research sub-questions.
1.5 Research questions
The overarching research question has been divided into two parts. These two parts reflect
the potential influence of the PRR on (1) non-voters, and (2) cleavage voters. The first part
of this dissertation specifically addresses the influence of the PRRs on the non-voters. It
aims at answering the question of whether the PRR has the ability to attract non-voters
to the polling booth. This part will consist of two chapters: (1) a cross-national, over-time
comparison of the effect of the PRR on individual propensity to turn out for elections in
27
Chapter 1
both Eastern and Western European countries; and (2) a longitudinal study in one country,
the Netherlands, which addresses the dynamic question of whether non-voters are most
likely to turn to the PRR once they decide to vote.
The second part of the dissertation addresses the potential cleavage-breaking impact
of the PRR. This part will specifically focus on Western European countries where certain
cleavages have most strongly dominated 20th- and 21st-century (party) politics. The Communist experience and legacy has made it impossible for some cleavages to emerge in East
and Central European countries (Judt, 2010). Moreover, because it is argued that PRRs
may become ‘catch-all’ parties, we focus on those countries with a successful PRR, where
cleavage-breaking patterns should have become most visible. Specifically, I focus on the
traditional religious cleavage (Knutsen, 2004; Lipset & Rokkan, 1967) and the gender cleavage in voting behavior (Inglehart & Norris, 2000; Knutsen, 2001).
Before I turn to these two parts, in my first empirical chapter, Chapter 2, I make a
pan-European comparison of the ideological differences between the PRR and other party
families. This overview will inform the results of the following chapters and is necessary to
establish the ‘uniqueness’ and exclusivity of the PRR’s ideology. This is important because
second- and third-wave research suggests that the PRR can only be successful in influencing people’s voting behavior if it takes a unique position in the political opportunity
structure (Kitschelt & McGann, 1995; Koopmans & Muis, 2009; Mudde & Rovira Kaltwasser, 2012), and because earlier research shows that other parties have incorporated some of
the PRR’s stances on immigration and integration (Van Spanje, 2010). Thus, I first provide
an overview of cross-party differences on ‘typical’ PRR issues—immigration, nationalism,
authoritarianism, populism and anti-establishment images—and investigate the extent of
the PRR’s ‘uniqueness’.
In Chapter 2, I also introduce a variety of possible explanations for the distance between
the PRR and other parties with respect to their different characteristics. Consistent with
second- and third-wave PRR research, the differing party characteristics of the PRR and
other parties as potential explanations will be introduced, i.e., electoral success and being
part of a government. Taking insights from spatial competition theories (e.g., Dahlström &
Sundell, 2012; Meguid, 2005), it is argued that these different characteristics of the PRR and
other parties are likely to influence the distance between the PRR and those other parties.
The first research question thus reads as follows:
1. To what extent are PRR characteristics and those of other parties associated with the distance
on various ideological, policy issues and style dimensions between the PRR and other parties?
28
Introduction
1.5.1 The PRR and its influence on non-voters
After having studied the exclusiveness of the PRR’s programs, I turn to the first part of the
dissertation that addresses the influence of the PRR on non-voters. In this part, I formulate
research questions 2 and 3, which strongly coincide with the notion that the PRR can
‘increase participation by the inclusion of marginalized groups in society (…)’ (Mudde &
Rovira Kaltwasser, 2012: p.20). This can occur via the popular encouragement to vote, or
the introduction of novel, unique party programs that address the problems of the ‘common
man’, such as immigration and globalization. Thus, the PRR may have something that could
increase democratic citizenship and political awareness among people who are otherwise
excluded from the political process.
In the first empirical chapter of this part, Chapter 3, I explicitly concentrate on the influence of the PRR’s success on electoral turnout. Because the literature addressing the political
behavioral consequences of the PRR’s success is primarily focused on one country and is of
an exploratory and descriptive nature (De Lange & Akkerman, 2012; Fallend, 2012), I adopt
a cross-national pan-European perspective and investigate the influence of successful PRRs
on levels of voter turnout in both Eastern and Western European countries. Drawing upon
different theories, such as electoral competition (Franklin, 2004), I hypothesize that successful PRRs may foster voter turnout because they are passionate mobilizers that fulfill
a watchdog function and reintroduce electoral competition. Additionally, I argue in line
with the ‘silent majority’ literature that if the PRR is able to attract non-voters to the polling
booth, it will be especially successful in doing so among those who favor PRR policies and
are known to be over-represented in the PRR-electorate: the lower-educated, the blue-collar
workers, the unemployed, the politically dissatisfied, and those with negative attitudes toward immigrants. In addition, I develop a competing hypothesis that successful PRRs may
actually inhibit voter turnout because they introduce a more negative, hardened tone to
politics that further triggers distrust toward politics, politicians and democracy. Consistent
with the turnout literature (Blais & Rubenson, 2013; Franklin, 2004), I expect that this will
especially discourage those who are new to democracy: youngsters and Eastern Europeans.
Accordingly, the second research question reads as follows:
2. What is the effect of the PRR on levels of voter turnout in Europe? Who is encouraged to vote
or discouraged from voting by the popularity of the PRR?
In the second chapter of this part, Chapter 4, I provide an alternative test of the PRR’s potential mobilizing role by investigating non-voters’ propensity to vote for the PRR instead of
for other parties. This dynamic question asks for longitudinal analyses of over-time changes
in individual voting behavior. If it is the case that the PRR is more successful in encouraging
the electorate that feels underrepresented and has disengaged from politics, then the PRR
29
Chapter 1
should be shown draw disproportionately more support than other parties from people
who previously did not vote. Thus, PRRs may have the special potential to attract nonvoters to vote for them in future elections.
Typically, this question belongs to the volatility literature, which defines volatility as
people changing their party choice over time, thus resulting in changing percentages of
seats won by political parties over time (e.g., Gallagher, Laver, & Mair, 2006; Mair, 2008;
Pedersen, 1979; Van der Meer et al., 2012). However, volatility not only arises out of people
who decide to vote differently in different years but also is caused by people who previously
decided not to vote but later decide to vote for a party and vice versa. I aim to integrate the
electoral volatility literature and the third-wave PRR literature about the consequences of
the PRR’s success to investigate how non-voters change their voting behavior. More specifically, I investigate whether former non-voters decide to vote and whether they have the
tendency to vote for the PRR instead of for another party. I will use longitudinal Dutch data
because the Netherlands is famous for its unpredictable political system and high levels of
electoral volatility (Mair, 2008). Although this chapter does not include a specific test for
the influence of the PRR on electoral turnout, it does provide insights in whether and why
non-voters are more likely to turn to PRRs, rather than to other parties. This leads to the
following research question:
3. To what extent are (former) Dutch non-voters more likely to vote for a PRR instead of for
another party and how can that change be explained?
1.5.2 The PRR and its influence on cleavage voters
The second part of the thesis addresses the PRR’s potential to attract cleavage voters who
are consistently found to vote for the same party (e.g., Knutsen, 2001; Knutsen, 2004;
Van der Brug, Hobolt, & De Vreese, 2009). Scholars have argued that the weakening of
traditional class and religion cleavages (Andeweg & Irwin, 2005; Andeweg, 1982) together
with processes of globalization and migration, have fueled the emergence of new attitudinal cleavages (Aarts & Thomassen, 2008). Consequently, if typical PRR issues and policy
proposals have contributed to the development of a new cleavage, then those parties have
the potential to become new ‘catch-all’ parties that attract people from different social strata
and therefore fundamentally change the party system (Oosterwaal & Torenvlied, 2010; Pellikaan et al., 2007). Hence, if the PRR’s issues play a more prominent role in the political and
discursive opportunity structure (e.g., Koopmans & Muis, 2009), then I expect them also
to break through more traditional social and structural cleavages and to have the potential
to mobilize those parts of the electorate that traditionally vote for other parties (cf. Aarts &
Thomassen, 2008).
30
Introduction
First-wave research on the PRR has consistently found a clear social base of PRRsupporters (e.g., Lubbers et al., 2002). Self-employed people, blue-collar workers and
people with lower to middle levels of education who have more negative attitudes toward
immigrants are overrepresented in the PRR’s electorate (Lubbers, 2001a). Furthermore,
research indicates that non-religious instead of religious people (Arzheimer & Carter, 2009;
Billiet, 1995), and primarily men instead of women (Givens, 2004) vote for these parties.
Scholars have often presented the cleavage argument as a rationale for why these groups are
underrepresented in the PRR’s electorate: i.e., these groups are locked in their own social
cleavage and thus vote accordingly. Religious people vote for religious parties (Arzheimer
& Carter, 2009) and women vote for women’s parties, i.e., left parties (Betz, 1994; Edlund
& Pande, 2002; Inglehart & Norris, 2000; Inglehart & Norris, 2003). In addition, because
PRRs are ‘Männerparteien’ (Mudde, 2007: p.90) and generally secular (Camus, 2007), they
are not attractive to women and religious people.
In this second part, I argue that in countries with a popular PRR, both women and
religious voters are challenged to reconsider their habitual voting behavior. With respect to
the religious cleavage, I argue that in countries with a popular PRR, the gradually declining
group of Christians who have always voted for Christian parties are posed with a serious dilemma: to vote for a Christian party or the PRR? With the increasing number of
immigrants, other religions —in particular Islam— have become more visible in Western
European countries. Moreover, popular PRRs position themselves as the ‘defenders’ of the
traditional Judeo-Christian values and norms. I expect that in particular, orthodox Christians may feel religiously threatened by the increasing prevalence of other religions, which
could intensify their anti-immigrant attitudes (Abu Raiya, Pargament, Mahoney, & Trevino,
2008). This in turn may cause nativist PRRs that claim to defend Judeo-Christian values and
norms to be preferred over Christian parties that do not have or do not emphasize a nativist
stance. The fourth research question thus reads as follows:
4. To what extent are PRRs able to attract religious voters who are not generally expected to
vote for these parties? If so, why?
With respect to the gender cleavage, scholars have also investigated why women are
less likely than men to vote for PRRs (Fontana et al., 2006; Gidengil, Hennigar, Blais, &
Nevitte, 2005; Givens, 2004; Rippeyoung, 2007). Surprisingly, neither gender differences
in structural individual characteristics (such as position in the labor market) nor gender
differences in socio-political attitudes (such as authoritarianism, anti-immigrant attitudes
and political dissatisfaction) that belong to typical first-wave explanations can fully account
for the gender gap in PRR-voting.
31
Chapter 1
It is, however, important to note that the few studies to investigate the gender gap in
PRR-voting have a limited comparative scope, focusing on a restricted number of European
countries. Consequently, little is known about the presence and magnitude of a gender gap
in PRR-voting in a wider range of Western European countries and the differentiation of
the gender gap among those countries. Due to the limited number of countries examined
by previous studies, there has been a focus on individual-level explanations, and little attention has been paid to cross-national variations in the size of the gender gap. However,
descriptive results of earlier studies show evidence for such variations over time within and
between countries (e.g., Mudde, 2007). Inspired by the supply-side arguments of the second
wave of PRR research, I investigate the extent to which the populist image of the PRR, as
political outsiders and as users of populist rhetoric, may explain cross-national variations
in the PRR gender gap. Taking insights from Roth (1989), I introduce an alternative mechanism that argues that given women’s lesser political interest, women are more hesitant than
men to vote for political outsiders. Next, I argue that the polarizing, populist style of the
PRR’s discourse is typically masculine and therefore keeps women from voting for a PRR
(cf. Mayer, 2002). However, as a party loses its outsider image and becomes more popular,
I anticipate that women will be more likely to vote for these parties, which decreases the
gender gap. If this occurs, I would find evidence for the potential of the PRR to attract
women’s votes. The fifth research question thus reads as follows:
5. To what extent is there a gender gap in Western European PRR-voting, and how can this
gap be explained? To what extent can the image of a populist outsider explain cross-national
variation in the gender gap in PRR-voting?
1.6 Data
To answer the research questions and to test the hypotheses developed in the various
chapters I will use different data sets. To obtain an idea about the cross-national political
opportunity structures and to compare some characteristics of the PRRs in both Eastern
and Western European countries for the first research question, I use my own administered
Expert Survey data set: the Expert Judgment Survey for European Political Parties in 2010
(Immerzeel, Lubbers, & Coffé, 2011). To answer the second research question, I use a combination of the six waves of the European Social Survey (ESS, 2002-2012), the European
Election Database (Norwegian Social Science Data Services (NSD), 2014) and the Comparative Political Dataset-III 1990-2010 (Armingeon et al., 2012). For the third research
question, addressing the dynamic question of whether non-voters have the tendency to
vote for PRRs instead of other parties, I use the LISS (Longitudinal Internet Studies for the
Social Sciences) panel data administered by CentERdata (Tilburg University, 2007-2014).
32
Introduction
Lastly, for the research questions regarding the ability of PRRs to attract religious people
and women, I use the European Values Study (EVS, 2010), combined with the European
Election Database (Norwegian Social Science Data Services (NSD), 2014) and my own
Expert Judgment Survey for European Political Parties in 2010 (Immerzeel et al., 2011).
Below, I present a short description of the various data sources.
1.6.1 Expert Judgment Survey of European Political Parties in 2010
The Expert Judgment Survey of European Political Parties (Immerzeel et al., 2011) is a
data set consisting of scores on various characteristics of political parties given by political
scientists and other experts of political parties in 38 European countries, including both
Eastern and Western European countries, Turkey and Israel. The overall response rate of
the experts was 37.7% (see Appendix 1.A for the country-specific response rates and the
reliability analyses). Generally, the data set is the follow-up of the Expert Judgment Survey
for West European Political Parties 2000 which aim was ‘to discover opportunity structures for political parties with respect to the immigration theme as well as organizational
aspects of political parties’ (Lubbers, 2001b: p.2). The dataset has a similar aim and generally extends the Expert Judgment Survey 2000 by adding other questions that deal with
typical PRR-party characteristics: immigration restrictiveness, populism, authoritarianism,
nationalism, and (anti-)establishment image.
The data have been administered with an online survey consisting of 10 questions in
the period October-December 2010. The questionnaire consists of three parts: first, the
experts were asked to rate the political parties in the country they were most familiar with
on the often used general left-right-wing scale and the economic left-right-wing scale (also
see the Chapel Hill Expert Survey (Hooghe et al., 2010)). Second, experts were asked to
rate political parties on an array of typical PRR-characteristics (see above): immigration
restrictiveness, populism, authoritarianism, nationalism, and anti-establishment image.
Lastly, experts were asked to rate political parties on their quality of organization, level of
consensus within the party and the charisma of its party leader (see Appendix 1.A for an
overview of the exact wording of the questions and the answer categories). Each expert was
asked to position all the parties running in the last elections within their country that won
at least 0.2% of the votes on each of the dimensions on a scale ranging from 0 to 10, with
10 referring to the PRR-position (thus: very restrictive concerning immigration, strongly
supporting tough measures to fight crime, etc.).1
1.6.2 The European Social Survey
The European Social Survey (ESS, European Social Survey Team 2014: 2002-2003, 20042005, 2006-2007, 2008-2009; 2010-2011 and 2012-2013) is a high-quality, standardized,
1
For more information about the Expert Survey, see Appendix 1.A and 1.B.
33
Chapter 1
cross-sectional, national representative survey, conducted in the years 2002, 2004, 2006,
2008, 2010 and 2012, which is part of the European Science Foundation’s initiative to obtain
a cross-cultural comparative research design. In total, 36 European countries participated
at least in one of the six ESS waves. Respondents were selected by means of strict probability
samples of the resident populations aged 15 years and older. Response rates of 70% apply
to most countries. The ESS-data contain information about voting behavior, a variety of
attitudes towards politics, immigrants and politics and an extensive list of demographic and
structural characteristics.2
1.6.3 Longitudinal Internet Studies for the Social sciences
The LISS (Longitudinal Internet Studies for the Social sciences) panel data is a representative sample of Dutch individuals who participate in monthly Internet surveys and
is administered by CentERdata (Tilburg University, 2007-2014). The panel is based on a
true probability sample of households drawn from the population register. Households that
could not otherwise participate are provided with a computer and Internet connection. A
longitudinal survey is fielded in the panel every year, covering a large variety of domains
including work, education, income, housing, time use, political views, values and personality.3
1.6.4 The European Values Study
The 2008 Wave of the European Values Study (EVS, 2010) is a standardized, cross-sectional,
national representative survey in which 47 European countries participated. In every country, respondents were selected by means of stratified random samples of the resident populations aged 18 years and older. The target sample size of 1,500 respondents per country was
generally met, but the response rates differed considerably among countries, ranging from
roughly 35% to 85%. The data contain information on party choice, various structural and
demographic background characteristics, different dimensions of religiosity and opinions
and attitudes towards a broad range of issues.4
1.6.5 The Comparative Political Data Set III 1990-2010
The Comparative Political Data Set-III 1990-2010 (Armingeon et al., 2012) consists of
a variety of political and institutional data for a group of 35 OECD and/or EU-member
countries for the period 1990-2010. Amongst others, it contains information about the
electoral results of all political parties participating in the legislative national elections held
between 1990 and 2010 in all European countries. Data are mainly based upon two earlier
data sets created at the University of Berne, Institute of Political Science and funded by the
2
3
4
34
For more information about the ESS: http://www.europeansocialsurvey.org/
For more information about the LISS: http://www.lissdata.nl/
For more information about the EVS: http://www.europeanvaluesstudy.eu/
Introduction
Swiss National Science Foundation: The Comparative Political Data Set I (CPDS I) and The
Comparative Political Data Set II (CPDS II).5
1.6.6 European Election and Referendum Database
The electoral results for 2012 and later were gathered using the European Election and Referendum Database (Norwegian Social Science Data Services (NSD), 2014). The construction of this European Election and Referendum Database belonged to the European Sixth
Framework Research Programme, called “Citizens and Governance in a Knowledge-based
Society” which aim was to identify the determinants of civic participation in a variety of
elections. The database includes all outcomes of parliamentary, presidential, European, and
European Union related elections and referenda in 35 European countries since 1990. The
data are collected from original sources, prepared and made available by the Norwegian
Social Science Data Services (NSD).6
5
6
For more information about the Comparative Political Dataset III: http://www.ipw.unibe.ch/content/team/
klaus_armingeon/comparative_political_data_sets/index_eng.html
For more information about the European Elections and Referenda Database: http://www.nsd.uib.no/european_election_database/
35
Chapter 2
Competing with populist
radical right parties
This chapter was written together with Marcel Lubbers and Hilde Coffé. A slightly different
version of this chapter has been accepted for publication in Party Politics.
Abstract
The popularity of European populist radical right parties has led to investigations into the
distances between PRRs and other parties’ stances regarding immigration. This chapter
adds to this literature by investigating the distance between PRRs and the other parties on
a wider variety of typical PRR policy and style issues. Based on our own Expert Survey,
we consider ideological (immigration, nationalism, law and order) and style dimensions
(anti-establishment and populism). Furthermore, we examine to what extent characteristics of other parties and PRRs (ideological position, electoral success, being in office) are
associated with these distances. The results show that right-wing (neo-)conservative and
Christian-democratic parties are closest, while green and socialist parties are furthest to
PRRs regarding ideological dimensions. The opposite is found for the style dimensions.
Additionally, we show that the other parties’ characteristics, rather than those of PRRs, are
associated with ideological and style distances between PRRs and other parties.
37
Competing with populist radical right parties
2.1 Introduction
Over the last few decades, various European countries have witnessed increasing electoral
popularity of PRRs. Some PRRs have become part of government, such as the Austrian
Freedom Party (between 2000 and 2005), or supported minority governments, like the
Danish People’s Party (between 2001 and 2011). Whereas this increased success has led to a
substantial number of studies explaining PRRs’ electoral growth (see for overviews Mudde,
2007; Rydgren, 2007; Van der Brug & Fennema, 2007), scholars have only recently turned
to studying the effect of PRRs’ electoral success on policy outcomes (e.g., Akkerman, 2012),
social outcomes, such as anti-immigrant violence (e.g., Braun, 2011; Mudde, 2007), and
policy positions of other parties (e.g., Bale et al., 2010; Van Spanje, 2010).
Since the immigration issue is generally considered to be the major PRRs’ selling point
(Ivarsflaten, 2008; Mudde, 2007), most of this research on the effect of PRRs’ success on
policy outcomes and policy positions of other parties has focused on PRRs’ influence on
immigration-related issues. For example, Akkerman (2012) studied the effect of having a
PRR in office or supporting a minority government on the implementation of new immigration and integration policies in nine Western countries. A broader, cross-national European
study by Van Spanje (2010) examined the effect of PRR success on all other parties’ policy
stances regarding immigration and integration.
The almost exclusive focus on the issue of immigration in previous studies is understandable in the light of the often heard assertion that ethnic exclusionism is the sole raison
d’être of PRRs (e.g., Fennema, 2005), but it omits the growing consensus that PRRs are not
solely single-issue parties. They are also characterized by other ideological issues such as
nationalism and law and order, and more style-related elements like populism and antiestablishment image (Mudde, 2007). Recently, this brought researchers to study the extent
to which mainstream parties have reacted with populist rhetoric as a response to success
of PRRs (Rooduijn, De Lange, & Van der Brug, 2012). In this chapter, we aim to add to the
scant literature on the relation between PRRs’ successes to other parties’ policy positions
considered as characteristic of PRRs. In particular, we will look at the distance on different
measures between PRRs and other parties. The measures we will consider are three policy
issues (immigration, nationalism and law and order) and two style-related characteristics
(anti-establishment image and populism). The nationalist ideology of PRRs refers to protecting the homogeneity of the nation and consequently, the neutralization of any possible
threats to the homogeneous nation. PRRs’ stance related to law and order is exemplified by
a demand for harder punishments and zero tolerance. Furthermore, PRRs are characterized
by their populist anti-establishment strategy: PRR party leaders claim to represent the vox
populi, and to be the real representatives of ordinary citizens. They also construct an image
of themselves as being “in opposition to the political class” and established parties, which
are criticized for having lost touch with ordinary people and focusing on obsolete issues
39
Chapter 2
(Rydgren, 2007: p.245-46). Jagers (2006) formulates that populism can be defined both as
political style or discourse and as political ideology. We do question to what extent the
directly policy-related, ideological issues of nationalism, migration and law and order can
be distinguished from the style dimension, and how the other party families in Europe
differ from the PRRs.
In addition to looking at a variety of policy and style dimensions, this chapter contributes
to the existing research in two more ways. First, whereas the Eastern European perspective
is often lacking in research on PRRs (Bustikova, 2009; Mudde, 2007), this chapter has a
pan-European perspective. Such a pan-European approach is needed as it enables us to
explore potential differences in the association of PRR-characteristics and other parties’
policy stances in older Western European and newer Eastern European democracies (e.g.,
Mudde & Rovira Kaltwasser, 2012). Second, we look at both characteristics of PRRs and of
other parties and how they relate to the distance between PRRs and other parties on the
different policy and style dimensions. PRRs’ characteristics comprise their electoral success
(and change therein) and whether they are part of (or supporting) a (minority) government. Similar characteristics are introduced for the other parties, next to the ideology of the
other parties. Taking insights from spatial competition theories (e.g., Dahlström & Sundell,
2012; Downs, 1957; Meguid, 2005), we argue that these different characteristics of PRRs
and those of other parties are likely to be associated with the distance between the PRRs
and the other parties.
In sum, the two main research questions motivating this chapter are: (1) How close
do PRRs and other parties stand on different policy and style dimensions typical of the
PRRs?, and (2) To what extent are PRRs characteristics and those of other parties associated
with the distance on various ideological, policy issues and style dimensions between PRRs
and other parties? To answer our research questions, we use our own online administered
Expert Judgment Survey of European Political Parties 2010 (Immerzeel et al., 2011) that
enables us to study 25 Eastern and Western European countries with at least one PRR. Since
the expert survey introduced various questions which were not included in earlier expert
surveys, and was the first to include all political parties in such high number of Eastern
and Western European countries, we cannot merge it with earlier administered surveys.
This is unfortunate as it would have allowed us to take a longitudinal approach and draw
conclusions about causal effects, whereas relying on only one survey limits us to drawing
conclusions about associations (rather than causal effects). Yet, the survey used in this
chapter has the major advantage of allowing us to look at a wide variety of policy issues and
style components and among a high number of Eastern and Western European countries.
40
Competing with populist radical right parties
2.2 Theories and hypotheses
The ongoing electoral popularity of PRRs since the mid-1980s has led to a substantial
number of studies explaining its success. More recently, some scholars have observed that
PRRs influence policy making and other parties’ policy positions. Some argued that with
the growth of PRRs, other parties can no longer ignore the issues PRRs have put on the
political agenda and must incorporate them into their own program. Indeed, parties that
take extreme positions on issues, similarly to PRRs, are particularly known to be successful in putting it on the political agenda (Wagner, 2012) and as such to challenge other
(mainstream) parties to also focus and position on these issues (Green-Pedersen & Van
Kersbergen, 2002).
Standard spatial competition theories (e.g., Dahlström & Sundell, 2012; Downs, 1957;
Meguid, 2005) emphasize the strategic importance of competing parties’ policy positions.
Meguid (2005: p.348) describes two possible strategies for parties competing for votes:
“movement towards (policy convergence) and movement away from (policy divergence)
a specific competitor in a given policy space.” Policy divergence refers to increasing policy
distance with political competitors and encourages voter flight to the competing party. By
contrast, policy convergence entails occupying a position close(r) to that of the competitor.
The rationale behind policy convergence is that a party draws voters away from a threatening competing party by occupying a position close to that party. With a PRR as a competing
party, the convergence strategy would incite other parties to search a position close to the
PRR, for example by taking a tougher stance on immigration or law and order. Even though
some have suggested that this strategy is not always successful in gaining voters back from
PRRs (Dahlström & Sundell, 2012), previous research had indicated that PRRs do influence
other parties’ policy positions (e.g., Van Spanje, 2010). We propose that the extent to which
other parties will incorporate policies from PRRs, and thus decrease the distance with the
PRRs’ positions, depends on different characteristics of both PRRs and other parties.
2.2.1 Characteristics of other parties
Starting with other parties’ characteristics, the spatial competition theory logically predicts
that right-wing and conservative parties are more strongly threatened by PRRs’ success
than left-wing parties (Van Spanje, 2010). Right-wing and conservative parties’ original
ideological position is closer to that of PRRs and they are thus more likely to lose voters
to the PRRs. As a result, they will be more inclined to move towards the PRRs, hoping to
draw voters away from PRRs (Van Kersbergen & Krouwel, 2008). Moreover, given the tradition of defending the nation and its culture among right-wing parties (Bale, 2008: p.463)
and their ’ownership‘ of issues such as cultural identity and national pride (Van Spanje,
2010: p.567), it is easier for right-wing (conservative and Christian democratic) parties to
adopt a hard line on immigration compared with mainstream left-wing (social democratic)
41
Chapter 2
parties (Van Kersbergen & Krouwel, 2008). This is also in line with the findings of Akkerman (2012) who has shown that center-right parties in particular implement tougher
immigration and integration policies when a country’s PRR is successful. Hence, our first
hypothesis is straightforward: H1. The distance between PRRs and right-wing (conservative
and Christian-democratic) parties on PRRs central policy issues (immigration, nationalism,
law and order) is smaller compared with the distance between PRRs and left-wing parties (i.e.
social democratic parties, socialist and green parties).
However, research has consistently found that working class voters, who typically vote
for mainstream left-wing, i.e. social democratic and socialist parties, are over-represented
among the PRR-electorate (Coffé, 2008; Rydgren, 2012). Consequently, social democratic
and socialist parties seem to compete with PRRs for the same voters, suggesting that the
left-wing parties also have incentives to move towards PRRs in order to attract voters (back)
to their own parties. Research by Bale et al. (2010) confirms that left-wing parties in different Western European countries have responded to the electoral challenge of the PRRs
by taking a tougher stance on issues related to immigration and integration. Hence, the
expectation that right-wing parties will be more likely to move towards PRRs, compared
with mainstream left-wing parties, may not be so straightforward as expected based on the
spatial competition theory. This is in line with Meguid’s modified spatial theory (2005),
which contends the relevance of both proximal and non-proximal competitors’ positions
in shaping parties’ success. The research from Halikiopoulou et al. (2012) also shows that
nationalism is a common denominator of both the PRRs and the radical left. Hence, we
formulate an alternative hypothesis: H1a. The distances between PRRs and right-wing (conservative and Christian democratic) parties are comparable to the distances between PRRs
and left-wing (social democratic and socialist) parties on PRRs central policy issues (immigration, nationalism, law and order).
The style dimension
PRRs’ anti-establishment, populist style is often similar to the contemporary green and
socialist parties’ political style. Emerging from the new social movements of the 1960s,
green parties organized themselves into more ‘traditional’ political parties during the
1990s (Burchell, 2001) and have gained electoral success in the last decades. Despite their
tendency to downplay their radical political style due to Realos (Realists) surpassing Fundis
(Fundamentalists) in many of the green movements (Burchell, 2001; March & Mudde,
2005), green parties are not considered as mainstream parties but often seen as ‘New Politics’ niche parties that favor more elite-challenging (Poguntke, 1993) and popular forms
of democracy (Burchell, 2001; Donovan & Karp, 2006). The emphasis of green parties on
more direct forms of democracy where citizens are expected to contribute more actively
to decision-making processes via the use of e.g., referenda and citizen initiatives, relates to
42
Competing with populist radical right parties
PRRs’ stance that the voice of the ordinary people should be heard (e.g., Canovan, 1999;
Donovan and Karp, 2006). This had led some scholars to suggest that the green parties’
level of anti-institutionalism shows similarities with the contemporary populist, antiestablishment style of PRR (Thijssen, 2012).
Similarly, socialist parties take a niche position, in particular in Western Europe, and
present themselves as a true alternative to established parties (March, 2011). Over the last
decades, traditional socialist parties have mainly developed into social-populist parties,
standing up for ‘the people’ instead of only, as traditionally, for the proletariat (March &
Mudde, 2005). For example, the Dutch Socialist Party (SP) has developed from an originally Maoist party defending the proletariat into a proponent of the people (Lucardie &
Voerman, 2012). They use fierce anti-elite slogans such as ‘Stem Tegen!’ (Vote Against) and
‘Jij.Ik.Wij.’ (You. Me. We.) (March & Mudde, 2005), which implicitly distinguish ‘the moral
people’ from ‘the corrupt elite’ (Mudde, 2004). Hence, whereas both green and socialist
parties are expected to take a clearly different position than the PRRs on policy issues,
we anticipate that they stand closer to PRRs on the populist, anti-establishment dimensions compared with the mainstream parties. Following modified spatial theory (Meguid,
2005), we may thus expect the green and socialist parties to more prominently utilize their
populist and anti-establishment rhetoric and style to compete with PRRs for voters. Our
second hypothesis is thus: H2. The distance between PRRs and greens and socialist parties on
the populist, anti-establishment style dimensions is smaller than the distance between PRRs
and mainstream parties.
Size of other parties
In addition to the ideological background of other parties, we expect their popularity to
be associated with their distance with PRRs on both the policy and style dimensions. In
particular, to the extent that smaller parties are niche parties which do not focus primarily
on gaining votes but on their unique policy stances, we expect party popularity to be negatively related to their distance with the PRRs. Smaller parties are more likely to emphasize
their unique stance and thus less likely to adapt their positions in order to gain votes from
a competitor. Hence, translating this idea to the PRR as competitor, our third hypothesis
reads: H3. The ideological and style distance between PRRs and other parties is larger, the
smaller the electoral support for the other party.
A final characteristic which we anticipate to be relevant when comparing the distance
between other parties and PRRs is whether or not the other party is in government. As Van
Spanje (2010: p.568) pointed out, being in office is associated with different constraints
which make it more difficult for government parties to change their policy positions while in
government. Moreover, it is also riskier for parties to make bold statements on policy issues
closer to PRR policy positions, which are generally considered controversial. Furthermore,
43
Chapter 2
a governmental party is a central part of the political establishment, and thus very unlikely
to hold an anti-establishment or outspoken populist discourse. Thus, our fourth hypothesis
is: H4. The ideological and style distance between PRRs and other parties is smaller when the
other party is not in office.
2.2.2 PRR-ideology and style characteristics
We do not only expect characteristics of other parties to be associated with their distance
with PRRs, we also anticipate PRRs’ characteristics to be relevant. Specifically, we expect
policy convergence, and thus a smaller distance between PRRs and other parties when
PRRs are electorally successful. Facing stronger electoral competition from PRRs, other
parties have an incentive to move towards the PRRs’ position and to adopt closer policy and
style positions, and encourage voters to substitute the PRR to their own party (Van Spanje,
2010). Hence, our fifth hypothesis reads: H5. The more successful the PRR, the smaller the
distance between PRRs and other parties on all ideological and style dimensions.
Finally, once a PRR is in (or supporting a minority) government, we expect the distance
between PRRs and other parties to be smaller. Being in office contains some constraints
and makes it harder for PRRs, as for other parties, to take tough and controversial stances.
Furthermore, it is more difficult for PRRs in government to hold an outspoken antiestablishment discourse and to comment constantly on coalition parties or parties which
they support. Moreover, being part of government legitimizes to a certain extent the ideas
of PRRs, motivating other parties to feel less restraint in copying PRR positions. Thus, our
sixth and final hypothesis is: H6. The distance between PRRs and other parties on all ideological and style dimensions is smaller when the PRR is in office or supports the government.
2.3 Data and methods
To answer our research questions, we rely on our own online Expert Judgment Survey for
European Political Parties 2010 (Immerzeel et al., 2011). Although expert surveys have been
criticized to produce data which are biased by (theoretical) expectations regarding parties’
positions, Benoit and Laver (2006) have extensively compared the reliability of expert surveys compared with other sources, and confirmed their comparability and usefulness (see
Dalton, 2009). Moreover, the inter-coder reliability measures for this specific expert survey
show very high Cronbach’s alphas (generally above .90) for all dimensions (see Appendix
1.A). This indicates that the experts in every country generally agreed on the positions
of the different parties regarding the ideological and style dimensions. The expert survey
focused specifically on the political opportunity structures for PRRs and policy stances
related to core PRR issues.
44
Competing with populist radical right parties
The survey was conducted successfully in 33 countries: in all 28 countries of the European Union except Luxembourg, plus Iceland, Israel, Norway, Serbia, Switzerland and
Turkey. In total 2,310 political scientists and political sociologists were e-mailed and invited
to participate in the survey. Scholars’ names were selected from the homepage of political
science departments of universities across countries included in the survey. With 870 experts completing the questionnaire in all the countries, the response rate was 37.7 percent,
which is comparable to Huber and Inglehart’s (1995) expert survey (41 percent). As this
study addresses the distances between PRRs and other parties in countries with a PRR, we
only included those 25 countries where one (or more) PRRs received at least 0.2 percent
of the votes in the last national election before the survey. For Belgium, we investigated
Flanders and Wallonia separately as these regions have different party systems.7
2.3.1 Operationalization
Dependent Variables
The survey included questions related to parties’ positions on the core ideological topics
and style dimension of the radical right: immigration, nationalism, law and order. The
style dimension includes populism and anti-establishment image. Each expert was asked to
position all the parties running in the last elections within their country that won at least
0.2 percent of the votes on each of the dimensions on a scale ranging from 0 to 10, with
10 referring to the populist radical right-wing position (e.g., very restrictive concerning
immigration, and strongly supporting tough measures to fight crime). The exact wording
of the questions is presented in Appendix 1.A.
A factor analysis revealed that the policy stances on immigration, nationalism, law
and order relate to one dimension, and that anti-establishment forms a second one, with
populism loading on both (see Appendix 2.A Table 2.A1). This latter finding underlines
Jagers’ (2006) notion that populism can both be seen as a political style and as an ideology.
To measure our dependent variables we calculated the differences between the positions
of the PRR on a dimension and the position of the other parties on that same dimension in
the respective country.8 Hence, the dependent variable refers to the distance between the
PRRs and another party on a particular dimension. A higher score refers to a larger distance
between a party and the PRR in the respective country. We used the PRR classification of
Mudde (2007), which is ‘the most well-founded attempt’ to classify PRRs lately, widely used
by other scholars (Ennser, 2012: p.156). As Mudde’s study does not cover the most recent
PRRs in all countries under study, we used a variety of other studies that explicitly refer to
Mudde’s definition when defining a party as PRR (Arter, 2010; Ennser, 2012; Ford & Good7
8
There was no successful PRR in Cyprus, Czech Republic, Iceland, Ireland, Latvia, Lithuania, Malta and
Spain in the latest election before the interview was taken. Hence, these countries were excluded from our
analyses.
If more than one PRR exists in a country, we summed and divided the results by the number of PRRs.
45
Chapter 2
win, 2014; Hirsch-Hoefler, Canetti, & Pedahzur, 2010; Mudde, 2012; Werts, Scheepers, &
Lubbers, 2013).9 Table 2.A2 provides an overview of the PRRs included in our analyses.
Independent Variables
We look at two broad groups of explanations for the distance between PRRs and other
parties’ positions: (1) characteristics of other parties and (2) characteristics of PRRs.
Other parties’ characteristics
The other parties, in comparison with the PRRs, have been grouped into eight different
party families: neo-conservatives (neoliberal-populists and ethno-regionalists), conservatives, Christian democrats, liberals, social democrats, greens, socialists/communists and
others (see Ennser, 2012). For this classification, we used Ennser’s (2012) classification of
the Western European party families, but decided to separate the Christian democrats and
the conservative parties, and add a different category of (radical left) socialist/communist
parties (cf. March & Mudde, 2005). Moreover, we distinguished a group of neo-conservative
parties that have become popular lately. The parties belonging to this neo-conservative
group are only present in six countries and are often mentioned by Mudde (2007: p.46-52)
and others as ethno-regionalist/nationalist or neo-liberal populist parties (e.g., Hooghe
et al., 2011). For non-EU member countries, we consulted the websites of the respective
political parties and followed Klingemann et al. (2006) to categorize these parties. In the
analyses, the social democrats are the reference category. Table 2.A3 in Appendix 2.A provides an overview of all the parties and their classification included in our study.
Next to party family, we included a dummy measuring whether the party was a governmental party in 2010 (year of data collection). Moreover, we added the party’s popularity
measured by the percentage of received votes in the last national elections before the survey
was taken, and additionally controlled for the percentage point change in the party’s popularity in those elections compared with the previous elections.
PRR characteristics
The PRR popularity in a country is measured by the percentage of votes the PRR scored in the
national elections held in the country closest before 2010. If more than one PRR competed
in the elections, we summed the results. We also included the percentage point change in
the PRRs’ popularity between these elections and the previous elections. Furthermore, we
introduced a measure that indicates whether the PRR is in office or supporting a minority
government. Table 2.1 offers descriptive statistics of all variables included in our analyses.
9
46
Despite not being mentioned in any scientific literature, we did include the Slovenian LIPA-party as PRR as
it was formed by former SNS-politicians just before the 2008 parliamentary elections.
Competing with populist radical right parties
Table 2.1 Descriptive statistics of the 205 parties included in the analyses
Min
Max
Mean
S.D.
Distance immigration
-0.56
8.94
4.06
2.07
Distance nationalism
0.10
8.62
4.22
2.08
-2.95
7.62
2.83
2.08
Distance law and order
Distance populism
-1.25
7.43
3.03
1.84
Distance anti-establishment
-6.90
7.57
2.72
2.93
0
1
36.5%
Eastern EU
Party family
Neo-conservative
0
1
6.1%
Conservative
0
1
10.5%
Christian democrat
0
1
21.0%
Liberal
0
1
17.7%
Social democrat
0
1
17.1%
Green
0
1
8.8%
Socialist/communist
0
1
11.6%
Other
0
1
7.2%
Party’s popularity (centered)
0.20
52.73
0.0
12.42
Party’s change in popularity
-23.90
39.70
0.16
6.63
0
1
32.6%
0.20
30.00
10.20
9.14
-9.80
14.50
0.83
5.90
0
1
11.5%
6.33
9.96
8.85
Party in government
PRR characteristics (country)
PRR’s popularity
PRR’s change in popularity
PRR in government
PRR’s position: immigration
0.88
PRR’s position: nationalism
8.16
9.89
9.36
0.43
PRR’s position: law and order
4.20
9.64
8.20
1.29
PRR’s position: populism
6.33
9.52
8.50
0.73
PRR’s position: anti-establishment
1.20
8.98
6.35
1.82
Source: Expert Judgment Survey of European Political Parties 2010 (Immerzeel et al., 2011).
2.4 Results
2.4.1 Descriptive Analyses
To answer our first research question, Table 2.2 and Figure 2.1 present the average policy
positions of the different party families.
Our results indicate that the PRRs score highest on the five domains considered in our
study, as one would expect given that these are the central foci and characteristics of the
47
Chapter 2
Table 2.2 Average party family position on core PRR-issues (standard deviations between brackets)
Party family:
Populist radical right
(N=34)
Neo-conservative
(N=11)
Conservative
(N=19)
Christian democrat
(N=38)
Immigration
Nationalism
Law and
Order
Populism
Antiestablishment
Leftright
Economic
left-right
8.97
9.38
8.34
8.61
6.64
8.43
5.96
(0.85)
(0.44)
(1.22)
(0.78)
(1.76)
(0.96)
(1.46)
7.31
8.33
6.81
8.11
6.24
7.13
6.05
(1.56)
(1.05)
(1.69)
(0.94)
(1.36)
(1.76)
(1.78)
5.91
6.42
6.52
6.31
3.10
6.54
6.25
(1.53)
(1.51)
(1.24)
(1.63)
(1.87)
(1.27)
(1.43)
5.86
6.10
6.25
5.53
2.89
6.51
5.83
(1.38)
(1.44)
(1.06)
(1.88)
(1.36)
(0.92)
(1.35)
Liberal
4.47
4.08
5.56
4.94
2.91
5.73
6.52
(N=32)
(1.64)
(1.46)
(1.25)
(1.23)
(1.53)
(1.38)
(1.35)
4.19
4.43
5.16
5.27
2.55
3.90
4.00
(N=31)
(1.30)
(1.33)
(0.87)
(1.28)
(0.99)
(0.94)
(1.07)
Green
3.09
3.04
3.67
3.98
4.49
3.58
3.77
(N=16)
(1.23)
(1.61)
(1.09)
(1.05)
(0.93)
(0.90)
(1.06)
Social democrat
Socialist/communist
(N=21)
2.83
4.16
3.52
5.58
6.28
2.13
1.84
(1.56)
(2.06)
(1.47)
(1.47)
(1.94)
(1.54)
(1.23)
Other
5.04
6.22
5.49
6.09
4.84
5.70
5.25
(N=13)
(1.99)
(1.66)
(1.96)
(1.02)
(2.02)
(2.22)
(1.77)
Total
(N=215)
5.43
5.81
5.87
6.02
4.20
5.66
5.16
(2.37)
(2.43)
(1.90)
(1.92)
(2.23)
(2.25)
(1.95)
Source: Expert Judgment Survey of European Political Parties 2010 (Immerzeel et al., 2011).
Note: All variables range between 0 and 10 (see Appendix 1.A for exact wording of the items).
Figure 2.1 Average policy positions by party family in Europe
Source: Expert Judgement Survey of European Political Parties 2010 (Immerzeel et al., 2011).
48
Competing with populist radical right parties
PRRs. The PRR family scores high on anti-immigration (8.97) and nationalism (9.38). It
also has an outspoken strict position related to law and order (8.34) and is characterized by
a populist (8.61) and anti-establishment discourse (6.64).
Comparing the average positions of PRRs with positions of the other party families in
Europe, the gaps on the immigration and nationalism policy domains are generally the
largest. The descriptive data also show that the scores of the neo-conservative group are
closest to the PRRs on all the ideological PRR-issues and populism. Their score on the antiestablishment dimension (6.24) is also relatively close to the score of the PRRs, although
the score of the socialist/communist party family is slightly higher on that dimension (6.28
vs. 6.24). It lends credence to call these neoconservative parties ‘borderline cases’ as they
are closest to PRRs on almost every dimension, but slightly less radical (cf. Mudde, 2007).
After the neo-conservative parties, the conservatives have the smallest gap with the
PRRs. This party family has an average score of 5.91 on the immigration policy, 6.42 on
the nationalism policy, and 6.31 on populism. The Christian democrats follow on the immigration and nationalism policies, trailed by the ‘other’ parties, the liberal parties and the
social democrats. The socialists have the lowest scores on immigration, whereas the greens
have the lowest score on nationalism. The greens and socialists also have the lowest score
on law and order. Christian democrats and conservatives have a smaller distance to the
PRRs than social democrats with respect to law and order. In addition, the liberals are on
average somewhat stricter concerning immigration and law and order than the mainstream
left-wing parties.
Regarding populism, the party families do not differ much, which could be a sign of a
‘populist Zeitgeist’ (Mudde, 2004). Apart from the green and liberal parties, all other party
families score 5.27 (the social democrats) or higher. Concerning the anti-establishment
style dimension, the PRRs score 6.64 and are closely followed by the socialists (6.28) and
the neo-conservatives (6.24). The conservatives, Christian democrats, liberals and social
democrats all score below 3.11; the greens 4.49 and other parties 4.84.
2.4.2 Multivariate Analyses
To answer our second research question and explore to what extent PRR characteristics and
those of other parties are associated with the distance between PRR and other parties, we
performed a multilevel analysis with maximum likelihood estimation for each dimension
considered in our study. Multilevel analyses are used, because our data have a two-level data
structure (Hox, 2002; Snijders & Bosker, 1999): namely the characteristics of other parties (party-level) are nested in countries with varying PRR-characteristics (country-level).
Table 2.3 presents the results of our analyses.
Starting with the association between the other parties’ characteristics and the distance
between PRRs and the other parties, Table 2.3 indicates that the other parties’ ideological
49
Chapter 2
Table 2.3 Results of the multilevel analyses of distances between PRRs and other parties’ positions
Nationalisma
Immigration
Law and order
Populism
Anti-establishment
b
S.E.
b
S.E.
b
S.E.
b
S.E.
b
S.E.
-3.11 ***
.56
Party characteristics
Party family
Neo-conservative
-3.55 ***
.52
-3.81 ***
.67
-2.29 ***
.47
-3.70 ***
.46
Conservative
-2.00 ***
.40
-2.10 ***
.37
-1.57 ***
.36
-1.39 ***
.35
-.61
.43
Christian democrat
-1.84 ***
.33
-1.69 ***
.38
-1.35 ***
.30
-.70 *
.30
-.24
.36
.20
.43
-.89 *
.33
-.42
.32
-.35
.39
Liberal
-.58
Social democrat (ref.)
Ref.
Greens
1.02 *
.45
1.40 **
.44
.89 *
.41
.03
.40
-1.50 **
.49
Socialist/communist
1.07 *
.42
.23
.48
1.11 **
.38
-1.34 ***
.38
-3.41 ***
.46
-1.40 **
Other
.36
Ref.
Ref.
Ref.
Ref.
-1.25 ***
.49
-2.00 **
.55
-.94 *
.44
.44
-1.59 **
.53
Party’s popularity (centered)
-.02 *
.01
-.01
.01
-.04 **
.01
-.05 ***
.01
.02
.01
Party’s change in popularity
.02
.02
.00
.01
.01
.01
-.00
.01
-.04 *
.02
-.12
.27
.07
.31
-.23
.24
.12
.24
PRR’s popularity
-.03
.02
-
-
.00
.03
.02
.02
-.03
PRR’s change in popularity
-.03
.04
-
-
.02
.04
-.06
.04
-.05
.07
.24
.59
-
-
.13
.68
.64
.62
.91
1.18
Eastern Europe
-.63 ~
.38
-.14
.28
-1.44 **
.43
-1.16 **
.39
-1.78 *
.73
Intercept
5.55 ***
.41
5.01 ***
.33
4.09 ***
Party level-variance
Party in government
.82 **
.29
PRR characteristics (country)
PRR in government
.43
4.03 ***
.40
4.53 ***
.05
.67
1.70 ***
.19
-
-
1.37 ***
.16
1.33 ***
.15
1.94 ***
.22
Country level-variance
.48 ***
.22
-
-
.76 ***
.28
.60 ***
.23
2.60 ***
.81
R2-party level
.53
2
-
.50
.40
.52
.35
R -country level
.24
-
.54
.48
R2-total
.49
.44
.52
.43
.43
Ni (Parties)
181 181 181 181
181
26 26 26 26
26
Nj (Countries)
Source: Expert Judgment Survey of European Political Parties 2010 (Immerzeel et al., 2011).
Notes: Sign. ~p< .10, *p<.05, **p<.01, ***p<.001 (two-tailed, maximum likelihood estimation)
Ni=181; Nj=26 (24 countries and the two Belgian regions Wallonia and Flanders).
a
As an empty model showed that there was no significant evidence for country-level variance in nationalism, we
show the results of a multivariate OLS regression excluding the country-level variables and using cluster-corrected
standard errors.
background is associated with the distance with the PRRs and thus mirrors the descriptive findings presented above. For all PRR-issues, except for anti-establishment, the neopopulist ethno-regionalist group is closest to the PRR-position. In line with Hypothesis
50
Competing with populist radical right parties
1, conservative and Christian democratic parties are substantially more likely to have a
position close to PRRs’ position on immigration, nationalism and law and order compared
with the social democrats. This refutes the alternative Hypothesis 1A. Liberal parties stand
appreciably closer to PRRs on law and order compared with social democrats, but not on
immigration and nationalism. Both the greens and socialists show a significantly larger
distance with PRRs than the social democrats on both immigration and law and order. Only
the greens show a significantly larger distance with PRRs on nationalism than the social
democrats and PRRs.
The picture is different when we look at the style dimensions: populism and antiestablishment. As anticipated in Hypothesis 2, green and socialist parties are generally
positioned closer to PRRs on the anti-establishment dimension than the social democrats
and the other mainstream parties (conservatives, Christian democrat, and liberal parties)10,
ceteris paribus. All party families except the greens and liberals stand closer to PRRs compared with the social democratic parties when considering the populist style of discourse,
which is not in line with Hypothesis 2. However, we do find in line with Hypothesis 2 that
the socialist parties are closer to PRRs on the populist dimension than the social democrats,
Christian democrats, and liberals. Hence, we only find partial evidence for Hypothesis 2.
Parties’ popularity, measured by their electoral success, is generally negatively associated with the distance with PRRs, supporting Hypothesis 3: the larger the party, the smaller
the distance with PRRs’ positions. This is the case for all issues, except for the positions on
nationalism and anti-establishment. For the latter, the findings suggest that a party’s increasing electoral success is associated with a smaller distance with the anti-establishment
position of the PRRs. Our final hypothesis related to the characteristics of the other parties
(Hypothesis 4) suggested that parties in office will have a larger distance with PRRs. Our
results indicate that this is only the case for the anti-establishment dimension. Hence, Hypothesis 4 can only be confirmed for that particular style dimension and not for populism.
Turning to the association between PRRs’ characteristics and the distance between
PRRs and other parties, the analyses are easy to summarize. Hypothesis 5 and 6 suggested
that the distance between the other parties and the PRRs would be smaller when PRRs are
more successful or when they are in government. The parameters do not reach conventional
levels of significance for any of the dimensions and our Hypotheses 5 and 6 can thus not be
corroborated.
The analyses in Table 2.3 indicate that the effect of Eastern Europe is significant and
negative for all style dimensions and policy issues, except for nationalism. This indicates
that the parties in Eastern Europe generally stand closer to PRRs than in Western Europe
(see Table 2.A4 in Appendix 2A). Therefore, we also explored possible interactions between
10 We changed the reference categories to check whether the results were significantly different as well for the
other party families.
51
Chapter 2
Table 2.4 Results of the multilevel analyses of distances on immigration, law and order, and populism
between PRRs and other parties’ positions and interaction effects for Eastern and Western European
countries
Immigration
b
Law and order
S.E.
b
Populism
S.E.
b
S.E.
Party family
Neo-conservative
-3.97 ***
.72
-3.26 ***
.61
-3.99 ***
.60
Conservative
-2.44 ***
.67
-2.09 ***
.48
-1.18 **
.48
Christian democrat
-2.12 ***
.53
-1.34 ***
.38
-.56 *
.38
Liberal
-1.29 **
.42
-1.02 *
.40
-.65 *
.40
Social democrat (ref.)
Greens
Socialist/communist
Other
Ref.
Ref.
Ref.
.78
.49
.93 *
.45
1.10 *
.49
1.39 **
.44
-1.41 ***
.44
-1.55 *
.71
-1.23 ~
.64
-1.83 **
.63
-.03 *
.01
-.03 ~
.01
-.03 *
.01
.03
.03
.02
.02
-.01
.02
Party in government
-.05
.33
-.19
.30
-.41
.30
PRR’s popularity
-.03
.03
.01
.03
.03
.03
PRR’s change in popularity
-.08 ~
.04
-.10 *
.05
.43
.56
.47
.62
.70
.68
-1.52 *
.72
-1.71 *
.72
-1.40 ~
.77
Party’s popularity
Party’s change in popularity
PRR in government
Eastern Europe
-.03
-.10 ~
.44
.05
Interaction-effects
Eastern Europe*…
Neo-conservative
1.27
.98
2.19 *
.89
.67
.88
Conservative
1.09
.77
.96
.70
-.38
.69
Christian democrat
.43
.67
.21
.60
-.34
.60
Liberal
1.80 *
.69
.47
.63
.76
.62
Social democrat (ref.)
Ref.
Greens
Socialist/communist
Ref.
Ref.
-.06
1.49
-.50
1.36
-.32
1.35
-2.07 *
1.00
-1.92 *
.90
.41
.89
Other
.75
.96
.51
.87
.87
.86
Party’s popularity
.03
.02
-.01
.02
Party’s change in popularity
-.04
.03
-.02
.03
.01
.03
Party in government
.56
.54
.19
.49
1.20 *
.49
PRR’s popularity
.00
.04
-.01
.05
-.03
.05
PRR’s change in popularity
.12 ~
.06
.22 **
.07
.10
.08
Intercept
5.79 ***
.45
4.14 ***
.45
Party level-variance
1.49 ***
1.22 ***
-.04 ~
4.16 ***
.02
.48
1.18 ***
Country level-variance
.34 ***
Ni (Parties)
181 181 181 26 26 26 Nj (Countries)
.52 ***
.68 ***
Source: Expert Judgment Survey of European Political Parties 2010 (Immerzeel et al., 2011).
Notes: Sign. ~p< .10, *p<.05, **p<.01, ***p<.001 (two-tailed, maximum likelihood estimation)
Ni=181; Nj=26 (24 countries and the two Belgian regions Wallonia and Flanders)
No significant interaction effects (p<.05) were found for the nationalism and anti-establishment dimension and were
thus excluded from this table.
52
Competing with populist radical right parties
the other parties’ and PRRs’ characteristics, and Eastern versus Western Europe in Table
2.4. The aim of this analysis was to explore whether some characteristics of the other parties
or the PRRs were related differently to the distance between PRRs and other parties in
Eastern and Western Europe.
With regard to other party characteristics, we found a limited number of significant
interaction effects. Of the party families, Western European liberals are somewhat more
restrictive regarding immigration than their social democratic counterparts, whereas in
Eastern Europe the liberals and social democrats are similarly restrictive regarding immigration. Furthermore, Eastern European socialists do not significantly differ from the social
democrats in their position on law and order, whereas they are somewhat more restrictive
towards immigration when compared with their social democratic counterparts. Moreover,
Western European neo-conservatives are somewhat tougher on law and order than their
social democratic counterparts, whereas in Eastern Europe the neo-conservatives and
social democrats are more similar regarding their position on law and order. Additionally,
we find some evidence that being in office as an Eastern European party is significantly
associated with a higher distance with PRRs on populism.
Lastly, with regard to PRR characteristics, we find that an increase in PRR’s popularity
is associated with a slightly closer position of all other parties to the PRRs on immigration,
law and order, and populism in Western Europe. This corroborates Hypothesis 5 and is
consistent with Van Spanje’s (2010) Western European study which concluded that as PRRs
become more popular, the other parties become stricter regarding immigration. For Eastern Europe, the significant interaction effects show that such effects have not been found,
and we thus find only partial evidence for Hypothesis 5.
2.5 Conclusion
The aims of this chapter were to investigate the ‘unique’ position PRRs take in the political
opportunity structure and to add to the literature on the possible consequences of PRRs’
success in Europe. Even though PRRs may not have a direct influence on policy making
because they have been seldom part of a ruling government, they may indirectly influence
policy making by influencing other (including governmental) parties’ stances (Williams,
2006). Whereas earlier research generally investigated the consequences of PRRs’ success
on the immigration stances of other parties (e.g., Akkerman, 2012; Van Spanje, 2010), we
suggested that a more comprehensive overview of the relationship between PRRs and other
parties’ stances required a more detailed investigation into a wider variety of issue and
style dimensions, typically associated with PRRs. In particular, we looked at (I) ideological
dimensions: immigration, nationalism, and law and order; and (II) style dimensions: populism and anti-establishment image. Based on our own Expert Judgment Survey for Euro53
Chapter 2
pean Political Parties 2010 in both Eastern and Western European countries, in this chapter
we investigated how close PRRs and other parties are positioned on these ideological and
style dimensions, and examined to what extent other parties’ characteristics (ideological
background, popularity, and being in office) and PRR characteristics (popularity and being
in office) are associated with these distances.
Regarding the other party characteristics, we found evidence that on the core ideological issues of PRRs (immigration, nationalism and law and order) socialist parties show the
largest distance, followed by the greens and social democrats. This is in line with Meguid’s
spatial competition model (2005), stating that those parties whose original ideology is
closer to the PRR will take similar stances, as they are competing for the same voters. Although it is often suggested that ideologically, the parties can be represented by a horseshoe
shape (see Halikiopoulou, Nanou, & Vasilopoulou, 2012), such that the socialists are closer
to PRRs than the center and right-wing parties, we found no evidence of that regarding the
issues of immigration, nationalism and law and order. With regard to nationalism this is
surprising, because Halikiopoulou et al. (2012) showed that radical left and radical right
parties have comparable anti-EU stances. This may be due to our measure of nationalism,
which is rather broad and might not refer to the Eurosceptic stances that these parties are
also known for.
Yet, we did find a horseshoe shape related to the anti-establishment position of both
party families. Our analyses revealed that the positions of the socialist (and green) parties
on the anti-establishment dimension are much closer to the PRRs’ position than the positions of the center parties. This speaks to the importance of non-proximal competitors in
the modified spatial theory (Meguid, 2005), which suggests that parties on the other end
of the political spectrum will incorporate comparable political styles to compete for voters.
Radical ‘niche’ parties on both ends of the continuum may learn from each other’s strategies
to obtain as many votes as possible.
Our results also demonstrated that successful parties have on average a closer position to PRRs (except for the nationalism and anti-establishment dimension). This may
indicate that the year 2010 (when the survey was held) was prototypical for the European,
‘Verrechtse’ (Bale, 2003), populist Zeitgeist (Mudde, 2004) where parties that adopt more
right-wing populist policies are more popular among the electorate. This is however in
contrast to the findings from Rooduijn et al. (2012) who performed content analyses of the
parties’ programs in fewer countries, and came to the conclusion that there is no relation
between popularity of PRRs and populist rhetoric in mainstream parties’ programs. An
area of research worth pursuing is to incorporate more time-points and countries in future
research, and rely on different assessments of populist rhetoric from parties, to advance
either conclusion more convincingly.
54
Competing with populist radical right parties
Regarding PRRs’ characteristics, we did not find evidence that when PRRs are in office,
the other parties would be likely to be closer to PRR positions. Given the few countries with
PRRs in office, this obviously needs more attention in future studies in order to draw firm
conclusions. However, we did find that an increase in PRR electoral share is associated with
a position closer to the PRR in the Western European countries on immigration and law
and order. This is consistent with Van Spanje’s (2010) study on Western European countries
and the position change between 2000 and 2004. Yet, for the Eastern European countries
and the other dimensions (nationalism, populism and anti-establishment) such effects have
not been found. Due to lack of longitudinal Expert Survey data on the dimensions under
study, we did not take a dynamic approach. Future research based on longitudinal data
could successfully use such an approach. This would help to solve part of the unavoidable
endogeneity problem inherent to our analysis: is it because of electoral success that other
parties become closer to PRRs, or is it the opposite: other parties become more successful
once they take a position that more closely resembles the PRRs’ stance?
To conclude, we have shown that the presence of PRRs is generally only relatively
weakly related to the ideological and style dimensions of other parties. Instead, the other
parties’ own ideology and popularity seems to play a major role in the position they take
on typical PRR issues. Moreover, our pan-European approach has provided new empirical
insights in the differences in party systems across Eastern and Western European countries.
We argue that future research should theorize and further investigate about why Western
European PRRs seem to be more radical and unique in their typical stances than their
Eastern European counterparts.
55
Part I
Populist radical right parties
and non-voters
57
Chapter 3
Populist radical right parties mobilizing
‘the people’?
This chapter was written together with Mark Pickup. A slightly different version of this
chapter is currently under review at an international journal. An earlier draft was presented
at the SFU-Roundtable Sessions at Simon Fraser University, Burnaby/Vancouver (Canada,
November 2012).
Abstract
The increasing popularity of European PRRs has often been argued as either a corrective
or threat to democracy. In this chapter, we provide empirical scrutiny to these opposing
claims and investigate the impact of successful PRRs on levels of voter turnout. We argue
that successful PRRs may either foster voter turnout, because they are passionate mobilizers that fulfill a watchdog function and re-introduce electoral competition; or inhibit
voter turnout, because they introduce a more negative, hardened tone to politics that
further triggers distrust towards politics, politicians and democracy. To investigate these
possibilities, we examine whether various societal groups are affected by the popularity of PRRs. Using the six available waves of the European Social Survey for the period
2002-2012, we apply hierarchical fixed effect models to test the (de)mobilization potential
of PRRs in 19 Western and 14 Eastern European countries. Our results suggest that the
higher the popularity of PRRs the lower the voting propensity of all Eastern European
citizens and particularly young people, whereas in Western Europe the more successful
the PRR is, the higher the propensity to go to the polls among citizens who are already
politically interested.
59
Mobilizing ‘the people’
3.1 Introduction
European democracies have witnessed an increase in the popularity of PRRs over the last
three decades (Mudde, 2007). In Chapter 1, we showed that parties which combine nativist,
authoritarian and populist stances, such as the Schweizerische Volkspartei [Swiss People’s
Party, SVP] and the Hungarian Jobbik [Jobbik, the Movement for a Better Hungary],
have received more than 15% of the popular vote lately. The electoral successes of PRRs
have often been labeled as landmark political events, or even revolutions (e.g., the Dutch
‘Fortuyn-revolt’ (Couwenberg, 2004; Koopmans & Muis, 2009)), and have fuelled much
political turmoil. These observations appear to be at odds with the claim that contemporary
politics and democracies have become largely depoliticized (Mair, 2006b; Mouffe, 2005).
Based on decreasing numbers of political party partisans and members (Van Biezen et al.,
2012), Mair (2006) even concluded that our European democracies have become ‘democracies without demos’. Hence, it is puzzling that these PRRs receive so much popular support
and attention.
Various scholars claim, however, that PRRs have become popular exactly because of the
widespread lack of political interest (Betz, 1994; Mair, 2006b; Mouffe, 2005; Rydgren, 2007).
Some argue that the mainstream political parties in European democracies have become too
consensual, rational and pragmatic to overcome political and social conflicts, preventing
voters from becoming involved in politics (Mouffe, 2005). PRRs that address the problems
of the common man (referring to migration, crime, and globalization), use clear language
that everyone can understand (Van Leeuwen, 2009), emphasize popular sovereignty and
want to restore the ‘will of the people’ (Donovan & Karp, 2006; Mudde, 2007), could attract
an apolitical ‘silent majority’, thus leading to higher levels of turnout (Mudde & Rovira
Kaltwasser, 2012; Rovira Kaltwasser, 2012). In contrast, it could be argued that PRRs engage
in a negative, uncivil rhetoric and use a harsh polarizing political style (Van Dooremalen,
2011). Their uncivil exchanges of political views could cause negative reactions towards
politicians, the government and the whole electoral systems. This, in turn, may discourage
people to turnout at elections (Mutz & Reeves, 2005).
What these theoretical notions suggest is that the electoral success of PRRs either
inspires or discourages people to speak up and become active in politics. Hence, PRRs
may play a mobilizing or demobilizing role when it comes to political engagement, and
thus, their mere existence and popularity could be interpreted as a corrective or threat for
democracy (Mudde & Rovira Kaltwasser, 2012). Yet, the degree to which PRRs are able to
affect such popular political engagement remains hitherto largely unknown. We argue that
if PRR success affects political engagement at all, it should be mainly observed in voting
behavior, which is the least demanding form of political engagement (e.g., De Rooij, 2009).
Therefore, we pose the following research questions: what is the effect of the PRR on levels
61
Chapter 3
of voter turnout in Europe? Who is encouraged to vote or discouraged from voting by the
popularity of the PRR?
Although several studies have addressed the impact of PRR success on people’s antiimmigrant (Dunn & Singh, 2011; Semyonov, Raijman, & Gorodzeisky, 2006; Wilkes, Guppy,
& Farris, 2007) and authoritarian attitudes (Andersen & Evans, 2003), little is known about
the effect of PRR success on political behavior, and voter turnout in particular. The few studies addressing this relationship have, moreover, remained on an exploratory level and show
mixed evidence with regard to type of behavior. Quintelier (2008) found that the Vlaams
Belang [Flemish Interest, VB] has recruited more than twenty-five thousand member over
the years, which suggests that successful PRRs can positively affect political membership
levels. In contrast, De Lange and Akkerman (2012) showed that Belgian voter turnout has
been decreasing since 1997 whereas the VB has become more and more popular in that
period, which contradicts the mobilization thesis.
With only few studies addressing the relationship between PRR-success and voter
turnout, the question of whose political behavior is affected by PRR-success remains hitherto unanswered. In spite of the almost unmanageable literature investigating the PRRs’
electorate (e.g., Arzheimer, 2012; Backes & Moreau, 2012; Lubbers et al., 2002), we do not
know whether successful PRRs trigger formerly non-voters to go the polling booth. Neither
do we know whether there are certain social groups that may actually be appalled by the
popularity of PRR and therefore decide to stop going to the polls, or feel urged to speak
out against the ‘rascals’ (cf. Andersen & Evans, 2003; Van Spanje, 2011). This is surprising,
because there are good theoretical reasons to expect that increasing PRR-popularity has
differential, catalyzing or inhibiting, effects on voter turnout of various social groups (e.g.,
Ansolabehere & Iyengar, 1995; Kahn & Kenney, 1999).
This chapter tries to fill these theoretical and empirical lacunae by conducting comparative analyses into the effect of PRR success on voter turnout in 19 Western and 14 Eastern
European countries covering the period 2002-2012. As such, this chapter greatly extends
the scope of earlier studies by including more countries over a longer period of time and by
investigating differential impacts of PRR-success on various social groups in these societies.
We make a distinction between Eastern and Western European countries, because Chapter
2 showed that there are remarkable differences between the Eastern and Western European
PRRs in terms of ideology and style (cf. Bustikova, 2009). Moreover, the contexts in which
the PRRs have emerged and flourished differ substantively across the younger democracies
of the Eastern European post-communist countries and the Western European countries
that have a longer tradition of democracy and electoral competition (Rydgren, 2011). The
comparative, cross-national nature of this chapter responds to Mudde’s (2012) claim that
more cross-national research is necessary to understand the impact of PRRs on people’s
political behavior; and enables us to provide empirical scrutiny of the debate on the con62
Mobilizing ‘the people’
sequences of populism on democracy (see Mudde & Rovira Kaltwasser, 2012). To test our
hypotheses we use all six waves of the European Social Survey (ESS, 2002-2012) and apply
hierarchical fixed effect logistic modeling which enables us to identify the effect of successful PRRs on levels of voter turnout in Europe.
3.2 Theory and hypotheses
In this section, we first establish how various prominent theories explain low levels of voter
turnout. Then, we introduce explanations for how successful PRRs may foster or deteriorate
voter turnout. Based on these explanations, we derive hypotheses about the effect of PRR
success on voter turnout in general and of various groups in society in particular.
It is generally found that a large group of people in society have disengaged from politics, resulting in lower levels of voter turnout. Various arguments have been postulated to
explain this trend in turnout decline. First, the ‘generation school’ (cf. Blais & Rubenson,
2013) addresses the changing attitudes towards democracy, politics and civic duty across
generations. According to various scholars (e.g., Campbell, 2006; Putnam, 2000), these
changes have been caused by the decrease in significance and the involvement in traditional civic associations such as the church and the labor unions. As ‘schools of democracy’
(De Tocqueville, [1835-1840] 2000), these associations used to foster social trust across
different social strata and (in)directly helped to the development of norms that promote
political and civic engagement, including the civic norm to go to the polls in the event of
elections. However, due to modernization processes (such as the ‘technological transformation of leisure’, mobility and increasing female labor force participation (Putnam, 1995)),
traditional civic life declined and the development of widespread norms that turning out at
elections is appreciated and good for one’s country was hindered. Consequently, younger
cohorts generally have a lower sense of civic duty, and are less interested in politics, which
explains why fewer people turnout at elections now than 30 years ago (Blais, 2006; Blais &
Rubenson, 2013; Dalton, 2008; Inglehart & Norris, 2003).
Second, the electoral competition or ‘context school’ (cf. Blais & Rubenson, 2013) argues that turnout has been declining due to characteristics of the elections rather than the
structural or attitudinal changes of the electorate (Franklin, 2004). Over the last decades,
people have felt that elections serve less of a function to democracy (e.g., Rosenstone &
Hansen, 1992), because elections have become less and less competitive (Franklin, 2004).
Likewise, Mouffe (2005) argued that politics has become a politics of consensus in which
traditional political parties rationally and pragmatically try to overcome political and
social conflicts. The decrease in polarization and electoral competition, accompanied with
converging ideological stances among the mainstream parties, e.g., about immigration and
multiculturalism (Arzheimer & Carter, 2006; Koopmans & Muis, 2009), have made it dif63
Chapter 3
ficult for ordinary citizens to distinguish the different parties. Consequently, mainstream
political parties have disabled citizens from becoming involved in politics, leading to lower
levels of turnout (Mouffe, 2005).
Thirdly, it is argued that more people have stopped going to the polls, because they
are dissatisfied with contemporary politics. This political dissatisfaction could stem from
the perceived mismanagement of mainstream parties dealing with problems accompanying
globalization and modernization that have deeply transformed European countries (Betz,
1994). Moreover, it is argued that citizens feel increasingly unable to control the democratic functioning of a country and that they perceive a mismatch between how political
decisions ideally should be made vis-á-vis the practical reality (Hibbing & Theiss-Morse,
2002). Ordinary citizens think that experts are too decisive and the opportunity to control
the process of decision-making is too small. This mismatch makes ordinary citizens more
negative towards government processes (Hibbing & Theiss-Morse, 2002), resulting in lower
levels of voter turnout.
3.2.1 PRR as mobilizers
Although widespread turnout decline is found in most European democracies, some scholars have argued that PRRs may be able to turn the tide of rising political quiescence. Here,
we present various arguments that refer to how PRRs may be able to foster general voter
turnout and then develop hypotheses on who is particularly likely to be mobilized.
With their new position in the political and discursive opportunity structure (Koopmans & Muis, 2009), PRRs evoke electoral competition and polarization (Mouffe, 2005). In
line with the ‘context school’, it is argued and empirically shown that in times of heightened
electoral competition, parties will try harder to distinguish themselves from the other parties with the use of political advertisements and by actively campaigning on the streets. This
helps voters to be better able to recall candidates and to be more familiar with campaign
issues, and thus leads to higher levels of voter turnout (e.g., Anderson & Beramendi, 2012;
Gimpel, Kaufmann, & Pearson-Merkowitz, 2007; Pacheco, 2008). With the politicization of
novel issues regarding immigration and integration, some argued that PRRs have contributed to the emergence of a new line of cultural conflict that also other parties had to address
(e.g., Pellikaan et al., 2007; Van Spanje, 2010). Polarization and fierce disagreement between
the mono- and the multiculturalists may give citizens the idea that there is something at
stake and that it is important to voice one’s opinion in the event of elections (Franklin, 2004).
Altogether, we expect that once PRRs become more popular and receive more credibility
(e.g., Koopmans & Muis, 2009), electoral competition increases, and thus we hypothesize:
H1a. The more popular a PRR in a country, the higher the level of voter turnout.
64
Mobilizing ‘the people’
Although electoral competition theory leads us to expect that increasing PRR-popularity is associated with a universal higher turnout, we suppose that some social groups are
particularly triggered to go the polls by the rise of PRR-popularity.
The most obvious group that may be extra excited to turnout is the group of people who
sympathize with the ideological stances of the PRR. As first-wave PRR-research showed
(see Chapter 1), this mainly concerns the ‘losers of modernization’, who face disproportionally more ethnic competition with immigrants: the lower-educated, the unemployed,
and the blue collar workers with negative attitudes towards immigrants (e.g., Betz, 1994;
Ivarsflaten, 2008; Lubbers et al., 2002; Olzak, 1992). This group is also the group with fewer
cognitive and material resources who, in the absence of a popular PRR, are less likely to
go to the polls (e.g., Bovens & Wille, 2010; De Rooij, 2009; Verba, Schlozman, & Brady,
1995). We argue that especially they may feel that there is finally a party that listens to their
grievances (regarding immigration and ethnic competition (Fennema, 2005)), and speaks
the language they can understand (Van Leeuwen, 2009). Therefore, we expect them to be
more likely to go to the polls in elections with a popular PRR: H1b. The more popular a PRR
in a country, the higher the levels of voter turnout among the lower educated, the unemployed,
the blue-collar workers, and/or those with negative attitudes towards immigrants.
Another group of people that we may expect to go the polling booth, which they will
most likely not do if a PRR is less successful, is the group of politically dissatisfied people
who have lost their interest in politics. Inspired by first-wave PRR studies (see Chapter 1;
Hooghe et al., 2011; Lubbers & Scheepers, 2000), we argue that these groups do not necessarily share the ideological convictions of the PRRs, but want to ‘send a signal’ to the other
parties that they have done something wrong (cf. Franklin, 2004: p.43). As we argued above,
they may be dissatisfied with politics because it has developed into a rational engagement
of consensus-seeking that leaves no room for passionate involvement (e.g., Mouffe, 2005).
Or they may feel abandoned by their political elites who have made politics and democratic
decision-making into an opaque affair where the opportunity to control is too limited (e.g.,
Hibbing & Theiss-Morse, 2002). Since PRRs stick up for the vox populi and promise to
‘protect’ the people from the ‘corrupt’ mainstream parties, they may be interpreted as the
ideal popular watchdog (Mudde, 2004). Therefore, they could motivate politically dissatisfied people who have lost their interest in politics and feel abandoned by their political elites
to go to the polls again, and we may thus expect: H1c. The more popular a PRR in a country,
the higher the level of voter turnout among the politically dissatisfied people and/or those with
a lack of interest in politics.
The last group of people that we anticipate to be mobilized by the PRR-success is those
who strongly oppose its radical ideology. PRRs have addressed issues about immigration
and integration that were labeled politically incorrect and appalling by the other parties,
often resulting in a cordon sanitaire (Coffé, 2005; Mayer, 2002). Such an exclusionary reac65
Chapter 3
tion of the other, mainstream parties may mirror the part of the electorate who finds PRRs
and their ideological stances repellent: those with positive attitudes towards immigrants,
the higher educated and the politically satisfied. They argue that PRRs are actually a ‘threat’
to democracy and despise its growing popularity (see Mudde & Rovira Kaltwasser, 2012:
p.16). We argue that those people with strong anti-PRR sentiments are even more inclined
to go to the polls, if PRRs become more popular, to keep the PRRs away from political
power (cf. Lubbers & Gijsberts, 2004). Hence, we hypothesize: H1d. The more popular a
PRR in a country, the higher the level of voter turnout among the politically satisfied and/or
interested, the higher educated and those with positive attitudes towards immigrants.
3.2.2 PRRs as demobilizers
Having hypothesized four types of positive effects of successful PRRs on voter turnout, we
now introduce opposing hypotheses for the potential negative effect of successful PRRs on
voter turnout. We draw on the theory of negative campaigning and its potential effects on
turnout and then develop hypotheses on who is particularly likely to be discouraged to go
the polls.
The theory of negative campaigning has been a vehemently debated theory since its introduction in the ‘90s (Ansolabehere & Iyengar, 1995). It claims that negative campaigning
results in lower levels of political engagement. The ‘mudslinging’ and negative way of putting aside political opponents perpetuates, in the media, an image of politicians as uncivil,
corrupt, incompetent, and untrustworthy. Consequently, voters are given the impression
that these political candidates are incapable of fulfilling a responsible, representative job
which fuels dissatisfaction with the political system and, in turn, results in a lower incentive
to vote for any candidate (Kahn & Kenney, 1999).
The evidence for the general hypothesis is mixed (e.g., Jackson, Mondak, & Huckfeldt,
2009), but scholars have argued and shown that it is the uncivil and negative tone of the
campaigning strategies in particular that negatively affects voter turnout (Brooks & Geer,
2007; Kahn & Kenney, 1999). According to these scholars, political actors engaging in media interactions that violate social norms regarding every-day face-to-face discourse will
reaffirm that politicians cannot be trusted. Their uncivil exchanges of political views cause
negative reactions towards politicians, the government and the whole electoral systems,
which discourage people to vote (Mutz & Reeves, 2005). We argue that PRRs, with their
harsh anti-immigrant, anti-establishment and populist style of doing politics, engage in
a negative, uncivil rhetoric. Scholars have documented how the harsh and negative tone
of these PRRs caused higher levels of debate polarization (e.g., Van Dooremalen, 2011).
Based on this observation, and following the negative campaigning theory, we argue that as
PRRs introduce a more negative, hardened tone to the debate, and become more popular,
more people are offended and therefore more distrustful towards politics, politicians and
66
Mobilizing ‘the people’
democracy, which leads to the hypothesis: H2a. The more popular a PRR in a country, the
lower the level of voter turnout.
Whereas Hypothesis 1d stated that those citizens who oppose PRRs are more likely to
go to the polls with growing PRR-popularity, negative campaigning theory predicts that
anti-PRR citizens would refrain from going to the polls as they are appalled by the harsh
anti-immigrant, uncivil rhetoric of the PRRs. Hence, we hypothesize: H2b. The more popular a PRR in a country, the lower the level of voter turnout among the politically satisfied and/
or interested, the higher educated and/or those with positive attitudes towards immigrants.
Moreover, in line with the ‘generation school’ it has been argued and convincingly
shown that the negative effect of the uncivil tone and rhetoric in the political arena affects
especially those people that are relatively new or marginally involved in politics (Kahn
& Kenney, 1999). The argument goes that people whose voting behavior has become a
habit (the ‘habitual’ voters (Plutzer, 2002)), have developed a kind of immunity for the
type of campaigning. Yet, for those who are not really interested in and have relatively
little experience with politics, it is argued that the negative ‘mudslinging’ keeps them away
from politics as it ‘resonates with their typical stereotypes about politics and politicians,
leading them to stay home on election day’ (Kahn & Kenney, 1999: p.885). Therefore, we
assume that East European citizens that have less experience with electoral competition
and negative campaigning than their Western European counterparts, young people who
are new in politics, and people who are not interested in politics at all are most affected by
PRRs’ negative campaigning. Therefore, we hypothesize: H2c. The more popular a PRR in a
country, the lower the levels of voter turnout among the Eastern Europeans, youngsters, and/
or politically disinterested.
3.3 Data and Methods
To test our hypotheses, we use the six available waves from the European Social Survey
(ESS, European Social Survey Team 2012: 2002-2003, 2004-2005, 2006-2007, 2008-2009;
2010-2011 and 2012-2013). In total, 36 European countries participated in at least one
of the six ESS waves.11 The ESS is a high-quality, standardized, cross-sectional, national
representative survey, conducted in the years 2002, 2004, 2006, 2008, 2010 and 2012, which
is part of the European Science Foundation’s initiative to obtain a cross-cultural comparative research design. Respondents were selected by means of strict probability samples of
the resident populations aged 15 years and older. Response rates of 70% apply to most
countries.12 We excluded those people who were not eligible to vote or did not provide
information about their voting behavior and moreover, we performed listwise deletion for
11 We decided to only focus on European countries and therefore excluded Israel from our analysis. Moreover,
we excluded Ukraine from our analysis due to too many missing values for some of our key variables.
12 For more information about the ESS: http://www.europeansocialsurvey.org
67
Chapter 3
the individual-level data, which results in 230,504 individuals (N1) who are nested in 33
countries (N2), 6 time-periods (N3) and 146 country*time periods (N4).
In addition, we use the Comparative Political Data Set III 1990-2010 (Armingeon et
al., 2012). This data set contains information about the electoral results of all political parties participating in the legislative elections held between 1990 and 2010 in all European
countries, Australia, Canada, New-Zealand, and USA. The electoral results for the 2012
wave were gathered using the European Election Database (Norwegian Social Science Data
Services (NSD), 2014).13
3.3.1 Operationalization
Dependent variable
Voting is based on the following item: ‘Some people don’t vote nowadays for one reason or
another. Did you vote in the last national election in [month/year]?’ Respondents could answer: ‘yes’, ‘no’, and ‘non-eligible’.14 We excluded the non-eligible people from our analysis.
Independent variables – Country-wave level
PRR success is the percentage of votes received by the populist radical right party in a
country in the latest election before the respective wave of the European Social Survey. The
choice of PRRs was based on the classification by Mudde (2007).15
Independent variables – Individual-level
Educational level is the ISCED-score for the highest level of achieved education provided
by the ESS and was divided into five categories: 1 ‘basic/primary education (ISCED 0/1)’, 2
‘lower secondary education (ISCED 2)’, 3 ‘higher secondary education (ISCED 3)’, 4 ‘postsecondary, non-tertiary education (ISCED 4)’, and 5 ‘tertiary education (ISCED 5-6)’. These
are entered into the model as a series of dummies. We took the first category (basic/primary
education) as reference category. To test our interaction effect of lower education and PRRsuccess, we created a dummy with value 1 for those with at maximum ‘lower secondary
education’ and value 0 for all those who have achieved higher levels of education.
Employment status is based on the item that asks respondents for the best description of
their current main activity for the last 7 days. We recoded this item into a dummy variable
13 For more information about the EED: http://www.nsd.uib.no/european_election_database
14 The electoral turnout numbers in the ESS are somewhat higher than the official national turnout numbers,
but the correlation between them is high: Pearson R=.82. In some cases, respondents from different ESSwaves are referred to the same last national election. In additional analyses (not shown here but available
upon request), we found that adding a dummy-variable measuring whether the last election was more than
2 years before the interview was held did not substantively change our conclusions. Furthermore, note that
this chapter takes actual voting behavior as dependent variable, whereas in the subsequent chapters we use
voting intention as dependent variable.
15 See Appendix 3.A, Table 3.A1 for the list of PRRs in the countries under study.
68
Mobilizing ‘the people’
with value ‘1’ for ‘unemployed, actively searching for a job’ and value ‘0’ for ‘being employed
or otherwise inactively unemployed’.
Blue-collar worker is based on the item that asks for the kind of occupation respondents
currently have. We used the item that asks for the sector people work in. As mainly the
production workers (instead of the managers) are threatened by ethnic competition, we
used the ISCO occupational scale (Ganzeboom & Treiman, 1996) to only select those with
the lower occupational status within the blue-collar sector. If a respondent has missing
information on the sector item, we used Oesch’ (2008) ISCO-transformation that partly
accounts for the sector in which people work. If a respondent has missing information on
the ISCO occupational scale, we used the sector item instead. We recoded this variable
into a dummy variable with value ‘1’ for ‘blue collar workers’, and ‘0’ for ‘other sector/never
had a paid job’. Respondents who are currently unemployed, but reported to be blue-collar
worker in their last job also received ‘1’ on the dummy variable.
Anti-immigrant attitudes measures the respondent’s attitudes towards immigrants (Schneider, 2008). It is based on the following three items that were available in all 6 waves of the
ESS: ‘Would you say it is generally bad or good for [country’s] economy that people come
to live here from other countries?’ with an 11 point answer scale 0 ‘bad for the economy’
- 10 ‘good for the economy’; ‘Would you say that [country’s] cultural life is undermined or
enriched by people coming to live here from other countries’ with an 11 point answer scale
0 ‘cultural life is undermined’ – 10 ‘cultural life is enriched’; and ‘Is [country] made a worse
or better place to live by people coming here from other countries?’ with an 11 point answer
scale 0 ‘worse place to live’ – 10 ‘better place to live’. We took the average score of these
three items (Cronbach’s alpha=.85); rescaled the variable such that a higher score indicated
a more negative attitude towards immigrants; treated the scale as an interval-variable; and
performed country-wave specific centering before entering it in our hierarchical fixed effect
model.
Political dissatisfaction measures the level of dissatisfaction with democracy and the
government in one’s country and is based on the following two items: ‘Now thinking about
the [national] government, how satisfied are you with the way it’s doing its job?’, and ‘On
the whole, how satisfied are you with the way democracy works nowadays in [country]?’
For both items, respondents could answer on a 10-point scale running from ‘0 extremely
dissatisfied’ to ’10 extremely satisfied’. We calculated the average score of these two highlycorrelated items (Pearson R=.62) ; took the value of one item if the other item was missing;
and rescaled the variable such that a higher score on this scale referred to a higher level of
political dissatisfaction. We treated the variable as interval variable and performed countrywave specific centering before entering it in our hierarchical fixed effect model.
Political interest is based on the item ‘How interested would you say you are in politics?’
Respondents could answer on 1-4 scale with 1 ‘not at all interested’ to 4 ‘very interested’.
69
Chapter 3
We treated the variable as interval variable and performed country-wave specific centering
before entering it in our hierarchical fixed effect model.
Young measures the age of the respondent and is recoded into a dummy variable with
score ‘1’ for those who are younger than 30 years old and score ‘0’ for those who are older
than 29 years old.
Control variables
We also control for variables that are associated with voter turnout and political participation in general (e.g., De Rooij, 2009):
Female is a dummy variable with score ‘1’ for females and score ‘0’ for males.
Marital status measures the current legal marital status of a respondent and was recoded
into three categories: ‘having a partner’ (married, or cohabiting), ‘ever had a partner’ (divorced, separated, and widowed), and ‘single/never had a partner’. These are included as
binary variables with the partner-group as the reference category.
Religious attendance is based on the item: ‘Apart from special occasions, such as weddings and funerals, about how often do you attend religious services nowadays?’ Respondents could answer on a 7-point scale running from 1 ‘every day’ to 7 ‘never’. Due to the
non-normal distribution of this variable we recoded the variable into three binary variables,
‘never’, ‘sometimes (i.e. once a year-less than once a month)’ and ‘often (at least once a
month)’ and used the never-group as reference category.
Urbanization identifies the area of a respondent’s residence and has five categories: in
a farm/at the countryside, country village, town/small city, suburb/outskirts or big city.
Again, these are included as a series of binary variables with farm/countryside is used as
the reference category.16
3.3.2 Methods
To test our hypotheses, we use multilevel or hierarchical methods that account for the hierarchical (cross-classified) nature of our data: respondents are nested in a specific country
(the country-level) and in a specific wave (the wave-level), and because we are interested
in changes over time (in PRR-popularity) within countries, respondent are also nested in
countries in a specific wave (country*time) (Hox, 2002; Rabe-Hesketh & Skrondal, 2008;
Snijders & Bosker, 1999).
For Hypothesis 1a and 2a, we look at the general effect of PRR-success on the individual’s
propensity to vote. To test these hypotheses, we use a hierarchical logistic regression with
fixed effects for countries and years and a random intercept for country*wave units. Let
Pvotei,j,t be the probability that individual i, from country j, at time t voted in the previous
election.
16
70
See Appendix 3.A, Table 3.A2 for the descriptive statistics of all variables.
For Hypothesis
1a and1a2a,
theatgeneral
effecteffect
of PRR-success
on theon the
For Hypothesis
andwe2a,look
we at
look
the general
of PRR-success
individual’s
propensity
to vote.
To test
hypotheses,
we use
hierarchical
logistic
individual’s
propensity
to vote.
Tothese
test these
hypotheses,
wea use
a hierarchical
logistic
regression
with with
fixed fixed
effectseffects
for countries
and years
and aand
random
intercept
for ‘the
regression
for countries
and years
a random
intercept
for people’
Mobilizing
be𝑖𝑖,𝑗𝑗,𝑡𝑡
thebeprobability
that individual
𝑖𝑖, from
country
𝑗𝑗, at 𝑗𝑗, at
country*wave
units. units.
Let 𝑃𝑃𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣
the probability
that individual
𝑖𝑖, from
country
country*wave
Let 𝑖𝑖,𝑗𝑗,𝑡𝑡
𝑃𝑃𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣
time 𝑡𝑡time
voted
in the in
previous
election.
𝑡𝑡 voted
the previous
election.
𝑃𝑃𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑖𝑖𝑖𝑖𝑖𝑖
𝑃𝑃𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑖𝑖𝑖𝑖𝑖𝑖
� = 𝛽𝛽�00=+𝛽𝛽00
�+ 𝛽𝛽�
𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙
�
𝛽𝛽𝑘𝑘 𝑥𝑥𝑘𝑘,𝑖𝑖𝑖𝑖𝑖𝑖
�
𝑘𝑘 𝑥𝑥𝑘𝑘,𝑖𝑖𝑖𝑖𝑖𝑖
1 − 𝑃𝑃𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣
1 −𝑖𝑖𝑖𝑖𝑖𝑖
𝑃𝑃𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑖𝑖𝑖𝑖𝑖𝑖
𝑘𝑘=2:𝐾𝐾
𝑘𝑘=2:𝐾𝐾
𝛽𝛽00 =𝛽𝛽𝛽𝛽001 𝑃𝑃𝑃𝑃𝑃𝑃
+ 𝜁𝜁𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗 +
+ 𝜁𝜁𝜂𝜂𝑗𝑗𝑗𝑗𝑗𝑗 + 𝜏𝜏𝜂𝜂𝑡𝑡𝑗𝑗++𝜀𝜀𝜏𝜏𝑗𝑗𝑗𝑗𝑡𝑡 + 𝜀𝜀𝑗𝑗𝑗𝑗
= 𝛽𝛽𝑗𝑗𝑗𝑗
1 𝑃𝑃𝑃𝑃𝑃𝑃
2
2
𝜁𝜁𝑗𝑗𝑗𝑗 ~𝑁𝑁�𝜇𝜇
𝜁𝜁𝑗𝑗𝑗𝑗 ~𝑁𝑁�𝜇𝜇
𝜁𝜁 , 𝜎𝜎𝜁𝜁 �𝜁𝜁 , 𝜎𝜎𝜁𝜁 �
Where:
Where:
𝜋𝜋 2 𝜋𝜋 2
𝜀𝜀𝑖𝑖𝑖𝑖𝑖𝑖 ~𝑁𝑁
�0,
�
𝜀𝜀𝑖𝑖𝑖𝑖𝑖𝑖
~𝑁𝑁 �0,
�
3
3
𝑖𝑖 ≡ 𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖
𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖
𝑖𝑖 ≡ 𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖
{Austria,
Belgium,
…, Switzerland,
Turkey}
𝑗𝑗 ≡ 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖:
{Austria,
𝑗𝑗 ≡ 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖:
Belgium,
…, Switzerland,
Turkey}
𝜁𝜁𝑗𝑗𝑗𝑗 are country*wave specific intercepts
{2002,
𝑡𝑡 ≡ 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡
2004,2004,
2006,2006,
2008,2008,
2010,2010,
2012}2012}
{2002,
𝑡𝑡 ≡ 𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖:
𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖:
specific
intercepts
(fixed
effects)
𝜂𝜂𝑗𝑗 are18𝜁𝜁𝑗𝑗𝑗𝑗country
areAppendix
specific
intercepts
18 country*wave
See
3.A, Table
3.A2
the for
descriptive
statistics
of all variables.
See Appendix
3.A,
Tablefor
3.A2
the descriptive
statistics
of all variables.
specific
intercepts
(fixed
effects)
specific
intercepts
(fixed
effects)
𝜏𝜏𝑡𝑡 are𝜂𝜂wave
𝑗𝑗 are country
84
wave specific intercepts (fixed effects)
𝜏𝜏𝑡𝑡 and𝜏𝜏𝑡𝑡𝜂𝜂are
𝑗𝑗 are included to control for unobserved wave and country level heterogeneity.
84
𝜏𝜏𝑡𝑡 and 𝜂𝜂𝑗𝑗 are included to control for unobserved wave and country level heterogeneity.
As the dependent variable of interest, electoral turnout, is a dichotomous variable
As the dependent variable of interest, electoral turnout, is a dichotomous variable
(withAsvalue
0 ‘not vote’
and value
1 ‘vote’),
we will use hierarchical
logistic regression
thevalue
dependent
variable
of interest,
electoral
is a dichotomous
variable (with
(with
0 ‘not vote’
and value
1 ‘vote’),
we will turnout,
use hierarchical
logistic regression
value
0
‘not
vote’
and
value
1
‘vote’),
we
will
use
hierarchical
logistic
regression
model (e.g.,
model
(e.g.,
Rabe-Hesketh
&
Skrondal,
2008).
This
model
links
the
log
odds
model (e.g., Rabe-Hesketh & Skrondal, 2008). This model links the log odds of the of the
Rabe-Hesketh
& Skrondal,
2008). Thisofmodel
links and
the log
odds of thevariables.
individual voting
individual
voting
with
a linear
individual
country*time
individual
voting
with
a linearfunction
function of individual
and country*time
variables. The The
propensity with a linear function of individual and country*time variables. The effect of
) isthe
(𝛽𝛽1(𝛽𝛽)1is
of PRR
success
theaverage
average effect
odds
of the
individual
effectfor
forthe
theloglog
odds
of average
the average
individual
effecteffect
of PRR
success
PRR success (β1) is the
average effect for the log odds of the average individual turning out
turning
outvote
to vote
(controllingfor
forother
other variables).
variables).
turning
out (controlling
to
(controlling
to vote
for other variables).
include fixed wave specific intercepts (𝜏𝜏𝑡𝑡 ) to eliminate systematic, acrossWe We
include
fixed wave
( 𝑡𝑡)) to
systematic, across-time
to eliminate
eliminate systematic,
acrossWe
include
fixed
wave specific
specific intercepts
intercepts (𝜏𝜏
time
differences
(wave-level
heterogeneity).
By
doing
so,
we
control
for
possible
heterogeneity).
By doing
so, weso,
control
for possible
unmeasured
timedifferences
differences(wave-level
(wave-level
heterogeneity).
By doing
we control
for possible
unmeasured that
Z-variables
thatboth
influence
both individual’s
probability
of turnout
and success.
PRR
Z-variables
influence
individual’s
probability
of turnout
and PRR
In fact,
unmeasured Z-variables that influence both individual’s probability of turnout and PRR
success.
In
fact,
it
eliminates
any
time
varying
pattern
that
is
common
to
all
countries
that
it eliminates any time varying pattern that is common to all countries that could lead to
success.
Inlead
fact,toresult.
ita eliminates
anyTo
time
varying
pattern
thateconomic
is common
to
all countries
that
a could
spurious
To illustrate,
European-wide
economic
and and
political
crises
spurious
result.
illustrate,
European-wide
political
crisesor terrorist
events
(such
as the (such
Madrid-Atocha
stationEuropean-wide
bombings)
thateconomic
influence
all
European
countries
could
to a spurious
result.
political
orlead
terrorist
events
as To
the illustrate,
Madrid-Atocha
station bombings)
that and
influence
all crises
are
controlled
for
by
including
these
fixed
wave
specific
intercepts.
Europeanevents
countries
are controlled
for by including these
fixed bombings)
wave specificthat
intercepts.
or terrorist
(such
as the Madrid-Atocha
station
influence all
To eliminate
the
possible
effects
of
time-invariant,
country
specific
factors (country
To eliminate
the possible
of time-invariant,
specific intercepts.
factors
European countries
are controlled
foreffects
by including
these fixedcountry
wave specific
level
heterogeneity)
on
both
the
probability
of
an
individual’s
vote
and
the
PRR success, we
(country level heterogeneity) on both the probability of an individual's vote and the PRR
To
eliminate
the
possible
effects
of
time-invariant,
country
specific
factors
include fixed country specific intercepts (ηj). The result is that β1 is estimated on
the basis
success, we include fixed country specific intercepts �𝜂𝜂𝑗𝑗 �. The result is that 𝛽𝛽1 is
(country
level heterogeneity)
on that
bothdiffer
the probability
of an individual's
vote
PRR
of changes
within countries
from one country
to the next.
Forand
ourthe
hypotheses,
estimated on the basis of changes within countries that differ from one country to the
we areweestimating
effect
of changes
in the
level of �𝜂𝜂
PRR�.success
on theisprobability
The result
that 𝛽𝛽1 isof an
success,
include the
fixed
country
specific
intercepts
𝑗𝑗
next.
For
our
hypotheses,
we
are
estimating
the
effect
of
changes
in
the
level
of
PRR
individual turning out to vote.
estimated
ononthe
of changes
within turning
countries
success
the basis
probability
of an individual
out tothat
vote.differ from one country to the
next. For our
we and
are estimating
the effect
of changes
the level
of PRR
In hypotheses,
Hypotheses 1b-d
Hypotheses 2b-c,
we predict
that theineffect
of PRR
success
differs
across individuals
depending
on some
success
on the
probability
of an individual
turning
out individual
to vote. characteristics such as
level of education and employment status. To allow the effect of PRR success to differ
In Hypotheses 1b-d and Hypotheses 2b-c, we predict that the effect of PRR
71
Chapter 3
In Hypotheses 1b-d and Hypotheses 2b-c, we predict that the effect of PRR success
differs across individuals depending on some individual characteristics such as level of education and employment status. To allow the effect of PRR success to differ across individuals
depending on some individual characteristic x2, we estimate the following model:
𝑃𝑃𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑖𝑖𝑖𝑖𝑖𝑖
𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 �
� = 𝛽𝛽00 + � 𝛽𝛽𝑘𝑘 𝑥𝑥𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖
1 − 𝑃𝑃𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑖𝑖𝑖𝑖𝑖𝑖
𝑘𝑘=2:𝐾𝐾
𝛽𝛽00 = 𝛽𝛽1 𝑃𝑃𝑃𝑃𝑃𝑃𝑗𝑗𝑗𝑗 + 𝛽𝛽𝐾𝐾+1 𝑃𝑃𝑅𝑅𝑅𝑅𝑗𝑗𝑗𝑗 × 𝑥𝑥2,𝑖𝑖𝑖𝑖𝑖𝑖 + 𝜁𝜁𝑗𝑗𝑗𝑗 + 𝜂𝜂𝑗𝑗 + 𝜏𝜏𝑡𝑡 + 𝜀𝜀𝑗𝑗𝑗𝑗
𝜁𝜁𝑗𝑗𝑗𝑗 ~𝑁𝑁�𝜇𝜇𝜁𝜁 , 𝜎𝜎𝜁𝜁2 �
𝜀𝜀𝑖𝑖𝑖𝑖𝑖𝑖 ~𝑁𝑁 �0,
𝜋𝜋 2
�
3
because we hypothesize
in Hypothesis
that may
there be
maydifferential
be differential
Lastly, becauseLastly,
we hypothesize
in Hypothesis
2c that 2c
there
effects of
effects
of
PRR-success
for
Eastern
and
Western
European
citizens,
we
decided
to
provide
PRR-success for Eastern and Western European citizens, we decided to provide separate
analyses
forWestern
Eastern and
Western European
countries.
Thisalso
wasinformed
also informed
analysesseparate
for Eastern
and
European
countries.
This was
by by
our findour
findings
of
Chapter
2
that
show
significant
differences
in
the
position
PRRs
take
ings of Chapter 2 that show significant differences in the position PRRs take in theinpolitical
the political
opportunity
structure
between
Eastern andEuropean
Western European
countries.
opportunity
structure
between
Eastern
and Western
countries.
3.4 Results
3.4 Results
In Table 3.1, we present the results of the hierarchical fixed effect logistic regression
In Tablemodel
3.1, we
resultstheofmain
the hierarchical
effect
logisticbehavior.
regression
in present
which wethe
included
effect of PRRfixed
success
on voting
We model
in whichdistinguish
we included
the main
PRR
success on
voting
behavior.
distinguish
bebetween
three effect
models:ofthe
all-country
model
(including
bothWe
Eastern
and
tween three
models:
the
all-country
model
(including
both
Eastern
and
Western
European
Western European countries), and the separate models for Eastern and Western European
countries),
and the separate models for Eastern and Western European countries.
countries.
The resultsThe
in the
all-countries
models
(Model
1) of1)Table
3.13.1
show
is no asresults
in the all-countries
models
(Model
of Table
showthat
thatthere
there is
sociation
between
PRR
success
and
voter
turnout.
The
hierarchical
fixed
effect
model
no association between PRR success and voter turnout. The hierarchical fixed effect does
not show
thatdoes
the more
or that
lessthe
popular
PRRpopular
becomes,
lower the
level the
of voter
model
not show
more ora less
a PRRthe
becomes,
the lower
level ofturnout,
which contradicts
both
Hypothesis
1a
and
2a
that
either
a
general
positive
or
negative
voter turnout, which contradicts both Hypothesis 1a and 2a that either a general positive effect
on turnout
is to be expected. Yet, Models 2 and 3 show that there are differences between
or negative effect on turnout is to be expected. Yet, Models 2 and 3 show that there are
Eastern differences
and Western
European countries. We find no significant association between PRR
between Eastern and Western European countries. We find no significant
success and voting behavior in the Western European countries, whereas a negative associaassociation between PRR success and voting behavior in the Western European countries,
tion is found in the Eastern European countries. The latter finding corroborates Hypothesis
whereas a negative association is found in the Eastern European countries. The latter
2c, stating that Eastern Europeans who have less experience with electoral competition and
finding corroborates Hypothesis 2c, stating that Eastern Europeans who have less
negative campaigns will be discouraged by PRR-success to go to the polls.
experience with electoral competition and negative campaigns will be discouraged by
The negative main effect in the Eastern European model of Table 3.1 implies that with an
PRR-success to go to the polls.
increase of one standard deviation in PRR success in a certain country in a certain year, the
odds of voting are (1-e(-.016*7.478=)) 11% lower. To clarify, we show in Figure 3.1 the predicted
probabilities of voting for different levels of PRR success, holding all other variables86 at their
72
Mobilizing ‘the people’
Table 3.1 Main effect of the PRR-success on voter turnout a
Model 1 All
DV=Voting
b Model 2 West
s.e.
b s.e.
Model 3 East
b s.e.
Country*Wave level
PRR success
-.000
.006
.009
.007
-.016 ~
.009
Individual level
Education
Primary (ref.)
Lower Secondary
-.173 **
.021
-.149 **
.024
-.170 **
.039
Upper Secondary
.016
.020
.037
.024
.035
.038
Post-secondary/Non-tertiary
.195 **
.033
.203 **
.040
.236 **
.060
University (Tertiary)
.334 **
.022
.311 **
.027
.413 **
.043
Unemployed (actively)
-.425 **
.025
-.530 **
.037
-.289 **
.039
Blue collar-worker
-.148 **
.015
-.107 **
.021
-.180 **
.021
.012 **
.003
.017 **
.004
-.091 **
.003
-.075 **
.004
-.109 **
.004
.658 **
.007
.687 **
.009
.624 **
.011
Young
-.650 **
.016
-.702 **
.021
-.583 **
.026
Female
.072 **
.012
.037 *
.015
.126 **
.018
Anti-immigrant attitudes
Political dissatisfaction
Political interest
.005
.004
Marital status
Partner (ref.)
-
Ever partner
-.305 **
.015
-.341 **
-
.017
-.253 ***
-
.022
Never partner
-.390 **
.016
-.409 **
.021
-.358 ***
.026
Urbanization
Farm/countryside (ref.)
-
-
-
-
.014 ~
-
Country village
-.014
.028
-.045
.035
Town/small city
-.170 **
.027
-.171 **
.035
-.078
.059
.059
Suburbs/outskirts
-.161 **
.030
-.176 **
.037
-.039
.067
Big city
-.325 **
.029
-.342 **
.037
-.218 **
.060
Religious attendance
Never (ref.)
-
Sometimes
.269 **
.013
.308 **
.017
.204 **
.021
Often
.537 **
.017
.581 **
.021
.463 **
.026
2.034 **
.063
1.861 **
.163
1.445 **
.175
Constant
-
-
-
-
σ2 Individual-level
π2/3
π2/3
π2/3
σ2 Intercept-Country-Wave
.039 **
.038 **
.031 **
379.48 **
242.81 **
2
LR-Chi
N1 (Individuals)
N2 (Countries)
N3 (Waves)
N4 (Country*Waves)
738.14 **
230,504
154,699
33
19
6 146
6 94
-
75,805
14
6 52
Notes: **p<.01, *p<.05, ~<.10 (two-tailed).
Country and wave dummies not shown here.
a
The results remain robust when PRR success is operationalized as ‘change in PRR success in % of votes as compared
to the penultimate election’.
73
Chapter 3
Figure 3.1 Predicted voting-probability at different levels of PRR success in Eastern European countries
(at 0=.76, at 18=.71)
means in the East European model. In the most extreme cases of PRR success in Eastern
Europe (0% vs. 17.32% ), we find that the probability of voting drops from .780 if the PRR
received 0% of the votes to .679 if the PRR-received 17.32% of the votes, ceteris paribus.
In Hypotheses H1b-d and H2b-c, we expected potential differential effects on mobilization
of specific groups in society. For the test of these hypotheses, we turn to our analyses that
include interaction effects of PRR success with the different individual characteristics. In
Table 3.2, we present the results of the interaction effects for Western European countries,
and in Table 3.3 for the Eastern European countries.
Table 3.2 and 3.3 suggest that there is mixed evidence for Hypotheses 1b-d and 2b-c,
and that variation exists between Eastern and Western European countries in terms of
which groups are mobilized or demobilized. We start discussing the results of the Western
European countries in Table 3.2.
In Hypothesis 1b, we predicted that those who are more threatened by the processes of
globalization and immigration–i.e. the lower educated, the unemployed, blue collar workers and those with negative attitudes towards immigrants–are more likely to turnout once a
PRR is more popular. If we take a look at the interaction effects in Table 3.2, we do not find
any evidence for this hypothesis. The Western European lower educated, unemployed, blue
collar workers and those with more negative attitudes towards immigrants are not more
likely to turnout in elections once the PRR is more popular. On the contrary, the lower
educated, blue-collar workers and those with strong anti-immigrant attitudes are less likely
to turnout once a PRR is more successful. In addition, the interaction effect of PRR-success
74
PRR-succes
+
.003
.008
.035
+
π2/3
-.002 ~
-.002
+
+
+
.035 **
+
.001
.001
.012
s.e.
π2/3
.001
.009
+
+
-.078 **
+
+
+
+
b Model 5
.001
.007
.012
s.e.
154,699
19
6 94
N1 (Individuals)
N2 (Countries)
N3 (Waves)
N4 (Country*waves)
94
6 19
154,699
507.87 **
.069 **
.040 **
94
6 19
154,699
398.49 **
.114 **
.039 **
94
6 19
154,699
384.71 **
.028 **
.039 **
94
6 19
154,699
660.73 **
.006 **
.039 **
94
6 19
154,699
592.09 **
.006 **
.038 **
Notes: **p<.01, *p<.05, ~ p<.10 (two-tailed).
Intercept, country and wave dummies not shown here, a Controlled for gender, marital status, urbanization, religion (not shown here).
+ = Included as control variable in analysis.
379.10 **
LR-Chi2
σ2 Slope-Country-Wave
94
6 19
154,699
688.73 **
.042 **
.039 **
π2/3
.012 **
.012
+
.600 **
+
+
+
+
+
b Model 6
.003
.008
.031
s.e.
.008
-.695 **
+
+
+
+
+
+
b Model 7
94
6 19
154,699
687.15 **
.162 **
.036 **
.038 **
π2/3
-.008 *
.010
+
+
+
+
-.060 ~
+
b Model 4
σ2 Intercept-Country-Wave
.007
.007
.066
+
s.e.
π2/3
π2/3
-.006
.010
+
+
+
+
+
-.468 **
+
b Model 3
.002
.004
.008
.043
s.e.
π2/3
π2/3
-.014 **
.008
+
+
+
+
+
+
-.163 **
b Model 2
σ2 Individual-level
.003
.024
.010
.004
.004
.023
.037
.027
s.e.
Young*PRR
Interest*PRR
Dissatisfied*PRR
Anti*PRR
Blue-collar*PRR
Unemployed*PRR
Lower education*PRR
Cross-level interactions
-.001
.696 **
-.725 **
Political interest
Young
Country*Wave level
.012 **
-.076 **
Anti-immigrant attitudes
Political dissatisfaction
-.527 **
-.124 **
Unemployed, active
Blue collar
-.231 **
b Lower education
Model 1
b s.e.
Model 0
Table 3.2 Interaction effects of PRR-success and individual-level variables on electoral turnout, Western European countries a
.005
.007
.062
s.e.
Mobilizing ‘the people’
75
76
+
.006
.009
.045
π2/3
.002
.001
.009
.010
s.e.
Model 5
-.016 ~
+
+
+
-.009
+
+
+
b +
π2/3
.001
-.015 ~
+
.001
.009
.011
s.e.
Model 6
-.113 **
+
+
+
+
b 6 52
N3 (Waves)
N4 (Country*waves)
52
6 14
75,805
364.79 **
.092 **
.036 **
52
6 14
75,805
239.70 **
.042 **
.030 **
52
6 14
75,805
245.43 **
.023 **
.030 **
52
6 14
75,805
270.24 **
.001 **
.030 **
52
6 14
75,805
273.57 **
.002 **
.030 **
Notes: **p<.01, *p<.05, ~ p<.10 (two-tailed).
Intercept, country and wave dummies not shown here, a Controlled for gender, marital status, urbanization, religion (not shown here).
+ = Included as control variable in analysis.
14
75,805
N1 (Individuals)
N2 (Countries)
249.62 **
LR-Chi2
σ2 Slope-Country-wave
52
6 14
75,805
421.91 **
.034 **
.031 **
π2/3
-.010 ~
-.016 ~
+
.005
.009
.042
s.e.
Model 7
.694 **
+
+
+
+
+
b -.012
52
6 14
75,805
345.88 **
.089 **
.029 **
.009
.009
.072
s.e.
Model 8
-.390 **
+
+
+
+
+
+
b .031 **
π2/3
.023 **
-.020 *
+
+
+
+
-.308 **
s.e.
Model 4
σ2 Intercept-Country-wave
.009
.009
.076
+
b π2/3
π2/3
.008
-.016 ~
+
+
+
+
+
-.349 **
+
s.e.
Model 3
-.029 **
.009
.008
.071
b π2/3
π2/3
.045 **
-.026 **
+
+
+
+
+
+
-.521 **
s.e.
Model 2
σ2 Individual-level
.009
.026
.011
.004
.004
.021
.039
.022
b Young*PRR
Interest*PRR
Dissatisfied*PRR
Anti*PRR
Blue-collar*PRR
Unemployed*PRR
Lower education *PRR
Interactions
PRR success
-.016 ~
.636 **
-.596 **
Political interest
Young
Country*wave level
.002
-.110 **
Anti-immigrant attitudes
Political dissatisfaction
-.307 **
-.234 **
Unemployed, active
Blue collar
-.261 **
s.e.
Model 1
Lower education
b Table 3.3 Interaction effects of PRR-success and individual-level variables on electoral turnout, Eastern European countries a
Chapter 3
Mobilizing ‘the people’
and being unemployed does not reach significance. Altogether, these findings lead us to
conclude that Hypothesis 1b cannot be confirmed for Western Europe.
Hypothesis 1c stated that those who are more politically dissatisfied and who lack
interest in politics are generally more likely to vote with a more popular PRR, whereas in
Hypothesis 1d we expected that those who are more politically satisfied and who are very
interested in politics are more likely to turnout if a PRR becomes more popular. Concerning
political (dis)satisfaction, we do not find a significant interaction effect, which implies that
both politically dissatisfied and satisfied are not more or less likely to turnout once a PRR
is more popular. For political dissatisfaction, neither Hypothesis 1c nor Hypothesis 1d is
supported by our results.
Regarding political interest, we do find a significant interaction effect in Model 6 of
Table 3.2: those who are more interested in politics are more likely to turnout once a PRR
become more popular, which is in line with Hypothesis 1d. In that same Model 6, the bcoefficient belonging to the main-effect of PRR-success refer to the effect of PRR-success on
the likelihood to turnout of those who are not politically interested. Since that b-coefficient
is not significant (b=.012, s.e.=.008, p>.05), we conclude that Western European citizens
who lack political interest are generally not affected by the popularity of a PRR. Together
with the finding that the interaction effects of PRR-success and political dissatisfaction does
not reach significance, we thus conclude that H1c is not corroborated for Western Europe.
In Hypothesis 1d, we also predicted that people who are repelled by the PRR-ideology–i.e.
higher-educated citizens and/or citizens with more positive attitudes towards immigrants–
are more inclined to go to the polls once a PRR becomes more popular. The non-significant
effect of the main-effects of PRR-success in Models 1 and 4 of Table 3.2 that refer to the
higher educated (b=.008, s.e.=.008, p>.05) and those with more positive attitudes towards
immigrants (b=-.002, s.e.=.001, p>.05), respectively, indicate that mobilization effects of
anti-PRR citizens do not seem to occur in Western European countries.
Regarding H2b and H2c, we hypothesized that the more popular a PRR in a country is, the lower the likelihood to turnout at elections among the politically satisfied, the
politically interested and disinterested, the higher educated, those with positive attitudes
towards immigrants, the Eastern Europeans and young people. The results can be easily
summarized: only Eastern European citizens seem to be more hesitant to go the polls with
a more popular PRR as we showed in Table 3.1. Regarding any of the other hypothesized
social groups in Western Europe, the results in Table 3.2 support neither Hypothesis 2b nor
2c. Surprisingly, Table 3.2 shows that those who are found to be overrepresented among
the PRR-electorate (also see Chapter 1)–the lower-educated, blue-collar workers, and those
who have more negative attitudes towards immigrants (the most important group of the
PRRs’ electorate (Ivarsflaten, 2008))–are actually less likely to vote if the PRR is successful
and thus, in general, seem demobilized by PRRs.
77
Chapter 3
In Table 3.3 we present the results of the Eastern European countries. We can be short
in describing the tests of Hypotheses 1b-1d that expected positive effects of turnout with
growing PRR-success among various social groups. All these hypotheses are not supported
by our findings. Table 3.3 does show that the interaction effects of PRR success with bluecollar workers and lower education attitudes are both significantly positive. However, we
cannot conclude from these effects that blue-collar workers and people with lower level
of education are more mobilized with PRR success. As the conditional marginal effects in
Figures 3.2 and 3.3 show: people with lower education and those working in the blue-collar
sector are not mobilized by successful PRRs, whereas people with a higher education and
those working in another sector are demobilized by successful PRRs. Hence, the negative
effect for higher-educated citizens corroborates H2b, which states that higher educated
groups in society are demobilized by PRR success. H2b is also supported by findings for the
other social groups: the politically interested (Model 6), the politically satisfied (Model 5),
and those with more positive attitudes towards immigrants (Model 4) are all less likely to
turnout once a PRR is more popular.
Furthermore, Table 3.3 shows that the interaction effect of PRR success with being
young and the effect of PRR-success on those with a lack of political interest (the main
effect of PRR-success in Model 6 (b=-.016, s.e.=.009, p<.10)) are significantly negatively
associated with the propensity to vote. Hence, in line with H2c, youngsters and people who
lack any interest in politics tend to increasingly avoid the polling booth with popular PRRs.
Altogether, in Eastern Europe our results are generally more in line with the demobilizing hypothesis H2b and H2c. We find especially strong negative effects for the young on
their propensity to turnout with growing PRR success. This underlines H2c stating that
Figure 3.2 Conditional marginal effect of PRR success for high and low educated citizens in East European countries (with 95% C.I.)
78
Mobilizing ‘the people’
Figure 3.3 Conditional marginal effect of PRR success for blue-collar workers and other-sector workers
in East European countries (with 95% C.I.)
young people are new in the political arena and feel opposed to the negative way of doing
politics. Surprisingly, we do not find any evidence of mobilization or demobilization of the
politically dissatisfied, the unemployed people or the people with negative attitudes about
immigrants in Eastern Europe.
3.5 Conclusion
In this chapter, we have added empirical scrutiny of the scientific and societal debate about
the consequences of PRR success in contemporary European societies. In particular, we
have investigated the potential impact of PRR success on the level of voter turnout in both
Eastern and Western European countries. Based on the premise that PRRs endeavor to
fulfill their radical democrat promise to restore the voice of the politically disengaged
people, we hypothesized that the PRRs might be a corrective to democracy by mobilizing
this group into politics. In contrast, we identified the potential consequences of the negative
‘anti’-ideology and rhetoric. We hypothesized that rather than attracting the politically disengaged into politics, these PRRs may indirectly further the distance between the politically
disengaged and the polls, whereas they may mobilize those people who feel repelled by the
PRR-ideology and want to keep PRRs away from power.
Interestingly, although earlier research showed that much of the typical PRR electorate
is formed by lower-educated, unemployed people, blue-collar workers and those with negative attitudes towards immigrants (see Chapter 1, and e.g., Lubbers (2001a)), these groups
do not seem to belong to the politically disengaged that did not vote and was excited to go
79
Chapter 3
to the polls when PRRs became more successful. Rather, these results suggest that PRRs
receive most of their votes from the segment of the electorate that was already voting for
another party, and changed its vote when there was a PRR (cf. Van der Meer et al., 2012).
Moreover, we find that the lower educated, the unemployed, the blue-collar workers, and
those with negative attitudes towards immigrants and politics are generally demobilized or
not affected in their voting propensity in both Eastern and Western European countries.
This could be explained by the notion that these typical PRR-supporters want a responsive
argument that is in line with their wishes, but without themselves having to go to the polls
(cf. Hibbing & Theiss-Morse, 2001; Hibbing & Theiss-Morse, 2002; Mudde, 2004). If a PRR,
their favorite popular watchdog which shares their ideological wishes, becomes popular
(which people notice in polls before the election takes place), they may anticipate that a sufficient number of people are voting for it, which makes the free-rider option of not turning
out the preferred option.
Generally, our results reveal substantive East-West differences regarding the effect of
PRR success on voter turnout. In Western Europe, we find that the probability of voting increases amongst the politically interested once a PRR is more successful. This latter finding
could indicate that PRRs trigger political interest and become successful, because more politically interested vote for them. On the other hand, the politically interested people could
be more inclined to go to the polls to keep the ‘rascals out’ (cf. Van Spanje, 2011). In the
next chapter, we use Dutch individual panel-data to investigate whether those who acquire
more political interest are also more likely to vote for a PRR or whether they opt for another
party. Furthermore, we find a decreased voting propensity amongst the lower educated and
those with negative attitudes towards immigrants when a PRR becomes more successful.
This is an interesting finding, because it implies that growing PRR-success widens the gap in
turnout between higher and lower educated citizens. This corroborates Bovens and Wille’s
(2010) claim that Western European democracies have turned into ‘diploma-democracies’,
and additionally suggests that PRRs’ mission to more properly represent the ‘voice of the
people’ does not translate into a more egalitarian involvement in elections of the people
themselves.
In Eastern Europe, we find that generally most people are demobilized if PRRs are more
successful, although some groups more than others. Young political newcomers are most
affected by the success of PRRs: they are increasingly less likely to vote if a PRR is more
successful. Furthermore, our results demonstrate that typical PRR-supporters are least affected by the rise of popularity. To illustrate, the lower educated and blue-collar workers are
generally not affected, whereas the higher-educated and other-sector workers are generally
demobilized by more popular PRRs. These negative effects of PRR success on voter turnout
in Eastern Europe are an important indication that PRRs using a strong anti-rhetoric in
countries with a shorter history of democracy may pose a ‘threat to democracy’. We urge
80
Mobilizing ‘the people’
future researchers to investigate the conditions that foster or inhibit such negative effects
and provide explanations for why such negative effects may exist.
Altogether, there is a little impact of PRR success on voting propensity in Western
Europe (also see Mudde (2012)), whereas in Eastern Europe PRRs seem to form a ‘threat to
democracy’ when it comes to electoral turnout. We have hypothesized that these differences
exist because Eastern European citizens may be relatively new to politics and therefore are
more strongly influenced by the negative, anti-rhetoric of Eastern European PRRs. However, this does not explain why Western European youngsters, who are also new to politics,
are not influenced by popular Western European PRRs. An explanation might derive from
the results in Chapter 2 that show that Eastern European PRRs are, compared with Western
European PRRs, ideologically closer to their competitors. This could indicate that Western
European PRRs are seen as niche parties that should not be taken seriously and thus do
not really affect people’s voting behavior. In Eastern Europe, however, where this radical
rhetoric is more mainstream and it is more difficult for established parties to function
as buffers against the PRRs (e.g., Coffé, 2008), the increase of PRR popularity may more
strongly nurture the idea that politics in general is uncivil, thus resulting in lower levels
of turnout (Mutz & Reeves, 2005). More comprehensive research is needed to substantiate
these possible explanations.
Based on the findings of this chapter, we postulate that we should be cautious in concluding that PRRs influence on the (non-)voting behavior of citizens is negligible. Not only did
we demonstrate that some social groups are affected more by PRR-popularity than others,
we also acknowledge that the PRR’s mere presence and popularity could have prevented the
turnout rates to be even lower. Furthermore, the entrance and popularity of PRRs into the
political arena did perhaps not attract many non-voters, but did influence others who were
already voting: otherwise these parties would not have become so popular. Hence, PRRs do
influence people’s decision to change their voting behavior (cf. Rydgren, 2012), but it seems
that they are more successful in doing so among those who have voted before than among
those who have not gone to the polls (yet). In Chapter 5 and 6, we elaborate on that latter
notion by investigating whether these PRRs are also able to attract those, who are said to
be locked up in their own cleavage, the cleavage-voters, and always vote for the same party
(e.g., Aarts & Thomassen, 2008; Knutsen, 2004).
In addition, we urge researchers to investigate more broadly the potential effect of
PRRs on other forms of political behavior. As Rosanvallon (2008) notes that populism is
the purest form of political protest and rejectionist behavior; we argue that a thorough
examination of the mobilization potential of PRRs through protests and demonstrations,
both in the ‘real world’ and on the internet, is needed.
81
Chapter 4
A Dutch ‘populist radical right revolt’?
This chapter was written together with Eva Jaspers and Marcel Lubbers. A slightly different
version of this chapter is currently under review at an international journal.
Abstract
PRRs may have negative and positive consequences for contemporary liberal democracies.
In this chapter, we investigate the potential positive effect of PRRs to politically mobilize
formerly non-voters, thus increasing electoral turnout. It examines whether non-voters
are more likely to turn to PRRs and tests explanations for why they are more likely to turn
to a PRR instead of another party. Using the six waves of the Dutch Longitudinal Internet
Studies for the Social Sciences (LISS) covering the period 2007-2013, it was found that if
non-voters decide to vote in a next election, they are indeed more likely to turn to Dutch
PRRs than to other parties. Our results suggest that the shift from non-voting to a vote for
a PRR is mainly related to changes in the PRR’s party leader evaluation among formerly
non-voters. Remarkably, changes in typical ideological PRR-attitudes, i.e. anti-immigrant,
anti-EU attitudes and political dissatisfaction, are not associated with higher probabilities
of voting for a PRR rather than another party.
83
A Dutch ‘populist radical right revolt’
4.1 Introduction
Four days after his assassination, thousands of people turned out for Pim Fortuyn’s funeral
procession in Rotterdam on May 10, 2002. The Dutch public TV broadcaster NOS reported
that more than 1.5 million people (10% of the Dutch population) followed the procession
live on television. Five days after his funeral, his political party received 26 seats in the
Dutch parliamentary elections, representing 17% of the Dutch electorate.
This remarkable political event has often been labeled a turning point in Dutch politics
and has received significant academic attention (e.g., Koopmans & Muis, 2009; Lucardie &
Voerman, 2012; Van der Brug, 2003). Not only had his assassination been the first Dutch
political assassination since the 17th century, Fortuyn was the first Dutch populist radical
right (PRR) leader who appealed to the broader population, raising questions about why
the PRR was able to break through. Moreover, Mair (2008) showed that the Dutch 2002
parliamentary election, in which 30.7% of seats changed, was among the most volatile
elections in post-war European history. Given that this indicator of voting instability underestimates the level of individual voting instability, and in spite of empirical evidence
that most people’s voting behavior has remained stable (e.g., Aldrich et al., 2011), 30.7% is
an exceptionally high level that together with high volatility levels in later elections (Van
Holsteijn, 2011) underlines the general trend towards ‘floating behavior’ in Dutch elections
(Mair, 2008). Additionally, the elections of 2002 are renowned for breaking the long term
trend of decreasing election turnout. In 2002, voter turnout increased to 79%, up from
73.2% in 1998, which suggests that previous non-voters decided to vote in the 2002 election.
These figures seem paradoxical in the context of scholars debating contemporary
depoliticizing Western democracies (Mair, 2006b; Mouffe, 2005) that are faced with continuously decreasing voter turnout (e.g., Blais & Rubenson, 2013; Franklin, 2004), political
party membership (e.g., Van Biezen et al., 2012), and other forms of political engagement.
In contrast, the ‘Fortuyn-revolt’ (Couwenberg, 2004) suggests that with the entrance and
increasing popularity of PRR-parties, people may be inspired to reconsider their ‘habitual’
non-voting behavior and turn out at elections. PRRs may be able to excite this type of
behavior because they claim to address the problems of the ‘common man’ (Van Leeuwen,
2009), emphasize popular sovereignty and aim to restore the ‘will of the people’ (e.g.,
Donovan & Karp, 2006; Mudde, 2007). Thus, PRRs may function as a curative for depoliticizing democracies by turning the tide of rising political quiescence into more political
engagement. Therefore, they may not only have negative but also positive consequences for
contemporary liberal democracies (see Mudde & Rovira Kaltwasser, 2012).
In this chapter, we aim to further substantiate the debate over the relationship between
PRRs’-success and political behavior. Specifically, we focus on the segment of the volatile
electorate that did not vote in one election but decided to vote for a party in a subsequent
election. We study whether these former non-voters are more likely to vote for a PRR
85
Chapter 4
than for another party upon voting in a subsequent election. Thus, we pose the following
research question: to what extent are (former) Dutch non-voters more likely to vote for a
PRR instead of another party, and how can that change be explained?
To answer our research question, we use the Politics and Values-Module of the Dutch
6-wave panel survey, the Longitudinal Internet Survey for the Social Sciences, which are
administered by CentERdata (Tilburg University, 2007-2014). This dataset enables us to
follow a large representative group of Dutch citizens who reported on their voting intentions, political attitudes and socio-economic backgrounds over the period 2007-2013. We
focus on those respondents who indicated that they would not go to the polls if there were
elections in one wave and who reported their voting intentions in the succeeding waves.
Thus, all formulated hypotheses concern these individuals. As such, we are able to detect
whether and why respondents change their intention from not voting at all to voting for a
PRR or another party.
The LISS-panel data do not allow us to test the changes in voting behavior during
Fortuyn’s heyday, as his party was a short-lived party and was unable to fulfill Fortuyn’s
political dreams (e.g., Akkerman, 2012; Lucardie & Voerman, 2012). Nevertheless, scholars
have often claimed that Fortuyn’s legacy opened up the political opportunity structure in
the Netherlands for other PRRs to become popular (Pellikaan et al., 2007; Vossen, 2010).
The 2006 entrance of Wilders’ PRR Partij voor de Vrijheid [Party for Freedom, PVV] in
Dutch parliament, its enduring popularity, and the popularity of Verdonk’s PRR Trots op
Nederland [Proud of the Netherlands, ToN] in 2008 and 2009 bear witness to this claim.
Whereas ToN never managed to obtain a seat in Dutch parliament, the electoral success of
the PVV increased from 5.9% in 2006 to 15.4% in 2010 (Van Holsteijn, 2011). After supporting a minority government, the PVV lost some of its appeal and received 10.1% of the
votes in the most recent elections in 2012. The PVV has grown into a stable political actor
receiving votes from a substantial portion of the Dutch population in various elections.
Hence, if PRRs have the ability to attract non-voters to the polling booth, then we should
also observe these effects for ToN and PVV in the period under study (2007-2013).
Generally, we improve upon the literature in the following ways: first, we employ an
individual panel study to follow the voting intentions of the same respondents over several
years in a Western European country. Second, we derive hypotheses regarding stability
(e.g., Aldrich et al., 2011; Denny & Doyle, 2009; Fowler, 2006) and instability in voting
behavior (Mair, 2008) that can be tested in the open and unpredictable Dutch political
system, which faces high levels of electoral volatility (Mair, 2008; Van der Meer et al., 2012).
Third, we provide empirical scrutiny of the debate over the mobilization success of PRRs
and more closely show whether and why people turn their non-vote into a PRR-vote rather
than another party vote.
86
A Dutch ‘populist radical right revolt’
4.2 Theory and hypotheses
The flourishing literature about electoral volatility demonstrates that a growing portion of
the population is changing its votes between elections (e.g., Gallagher et al., 2006; Mair,
2008; Pedersen, 1979; Van der Meer et al., 2012). At the same time, these studies show
that most eligible people continue to vote for the same parties in subsequent elections.
This behavioral continuity, known as habitual voting, has recently gained more prominence
in the turnout literature and addresses why non-voters tend to be habitual non-voters
(Aldrich et al., 2011; Denny & Doyle, 2009; Fowler, 2006; Franklin, 2004; Gerber et al.,
2003; Green & Shachar, 2000). In this section, we derive hypotheses about the tendency of
people to remain non-voters. Then, we make use of electoral competition and mobilization
theories that explain why some people might change their non-votes into PRR-votes, and
may theoretically support the notion that electorates are becoming more and more ‘adrift’
(Andeweg, 1982).
4.2.1 Stability in non-voting behavior
Various scholars have theorized why people tend to be stable in their (non-)voting behavior. These scholars introduced the concept of habitual voting, which states that voting is a
gradually acquired habit and that one’s voting behavior in one election is highly similar to
one’s voting behavior in subsequent elections (e.g., Plutzer, 2002).
Voters are more likely to vote and non-voters are more likely to non-vote in future
elections (Green & Shachar, 2000: p.569). It is argued that voting stability is the result of a
cost-benefit consideration: once people have invested in turning out to their first election
– learning to cast a vote and distinguishing among the political parties (Plutzer, 2002) - the
marginal costs to go to the polls at every subsequent election become smaller. Thus, the cost
barriers of voting shrink over time and the likelihood of turning out grow. For non-voters,
the cost-barriers remain high at every future election, and their likelihood of turning out at
future elections remains low.
Moreover, voting is interpreted as a self-reinforcing process: through voting, people feel
that they actively contribute to politics, thus strengthening their efficacy, which increases
their likelihood of turning out in later elections. In contrast, non-voting people feel less
politically efficacious and are less likely to turnout in following elections (Denny & Doyle,
2009). Additionally, it is argued that if individuals do not vote in one election, they will
tend to keep their future voting behavior in line with their earlier voting behavior ‘as being
inconsistent generates psychological discomfort’ (Denny and Doyle 2009: p.19). Lastly,
some scholars have argued that if people do not vote, they do not enter the political arena,
and therefore are not targeted by political campaigns and advertisements. This hinders
the development of political interest and leaves their low probability of turning out to an
election unchanged (Huckfeldt & Sprague, 1992). Based on these arguments, we derive the
87
Chapter 4
following hypothesis: H1. Individuals who do not vote in one wave are more likely not to vote
than to vote in a subsequent wave.
The abovementioned arguments also imply that once people become more politically
efficacious and interested in politics–whether through political campaigns or other mechanisms affecting political efficacy–they will re-evaluate their cost-benefit considerations to
turn out, which affects their propensity to turn out in future elections. Thus, we hypothesize:
H2. Individuals who (do not vote in one wave and) acquire greater political efficacy or interest
over time are more likely to vote than not to vote in a subsequent wave.
4.2.2 From non-voting to a vote for PRRs
Although recent literature has found evidence for habitual voting in countries with majoritarian systems, such as the United States (Green & Shachar, 2000) and Great Britain
(Denny & Doyle, 2009), scholars have also observed high levels of volatility in countries
with proportional multi-party systems, such as the Netherlands (Mair, 2008). These observations are based on macro-indicators that calculate volatility as the aggregated cumulative
electoral seat gains of all winning parties in a certain election as a percentage of all seats to
be gained (Gallagher et al., 2006). Scholars have demonstrated that people are increasingly
likely to change their habitual voting behavior, leading to increasing volatility percentages
of 15-20% on average in Dutch parliamentary elections in the 1990s and 2000s (Mair, 2008).
Hence, while most people seemed to have voted for the same party, a large minority also
changed its voting behavior.17
Volatility results from people voting for different parties in subsequent elections,
and from people who change their initial vote into a non-vote or vice versa. The change
from non-voting into voting has been of particular concern in theories of mobilization
(Mudde & Rovira Kaltwasser, 2012). Because non-voting has frequently been argued as
being detrimental to the democratic functioning of society, scholars have quite extensively
studied the reasons for non-voting (e.g., Blais, 2006). They showed that some people do not
vote because they lack some crucial cognitive resources, e.g., political efficacy, or material
resources, e.g., income (Verba et al., 1995); others do not vote because they are dissatisfied
with their contemporary politics (Hooghe et al., 2011).
The entrance and popularity of new parties, such as PRRs, have been argued to heighten
electoral competition and polarization, which is beneficial for election turnout (e.g., Franklin, 2004; Mouffe, 2005). Namely, heightened electoral competition triggers all parties to
campaign more intensely to communicate their unique profile (Franklin, 2004). This helps
17 Because we are dealing with macro-indicators of vote change and not individual measures of vote change,
15-20% is an underestimation of true individual electoral volatility: if parties receive the same percentage
of votes in two subsequent elections from two completely different groups of voters, then the individual
measure of volatility is much larger than the macro-indicator of volatility.
88
A Dutch ‘populist radical right revolt’
people to be better able to recall candidates and be more familiar with campaign issues,
which may trigger the non-voters to vote (Pacheco, 2008).
Although electoral competition theory states that heightened competition is related to
higher turnout, it does not theorize about how PRRs specifically may influence the turnout
of the disengaged non-voters. We argue that the combination of a nativist and populist
ideology (Mudde, 2007) is especially attractive to politically disengaged non-voters. The
PRRs’ emphasis on bringing the ‘voice of the people’ back into the political realm (Canovan,
1999), and therefore using the clear language that the common man in the street can understand may be especially attractive for non-voters. Thus, we expect: H3. Individuals who
(do not vote in one wave and) decide to vote in a subsequent wave are more likely to vote for
a PRR than for any other party.
We focus on two important explanations of why PRRs specifically have the potential to
attract non-voting people to the polling booth: individual changes in typical PRR-attitudes,
and changes in evaluations of the PRR-leader. Scholars have argued that the weakening
of the traditional class and religion cleavages together with processes of globalization and
migration have fuelled the emergence of new attitudinal cleavages (Aarts & Thomassen,
2008). Following ethnic competition (Olzak, 1992; Scheepers, Gijsberts, & Coenders, 2002)
and the ‘losers of modernization’ theory (Betz, 1994), it is argued that issues of globalization and immigration in particular have divided society in two groups: the anti-immigrant
traditionalists who favor national identity and reject the European Union, and the multicultural progressive global citizens who favor the European Union (cf. Kriesi et al., 2008).
Ethnic competition theory states that there is greater (perceived) competition with
immigrants over scarce resources such as labor and welfare state benefits, leading to the
development of anti-immigrant attitudes among natives who are in ethnic competition (Rydgren, 2007; Scheepers et al., 2002) Additionally, the ‘losers of modernization’ thesis posits
that a large group of people with fewer cognitive and material resources feel threatened
by globalization and immigration because they are unemployed or work in sectors where
the risk of becoming unemployed is relatively high (Betz, 1994). Due to the inability to
accept and adapt to socio-cultural transformations, people feel insecure and anxious. It is
argued that mainstream political parties have constantly ignored these sentiments or have
not taken them seriously, which has fuelled dissatisfaction with the mainstream political
parties (e.g., Rydgren, 2007) and thus caused people to turn away from politics and become
non-voters.
We argue that PRRs may attract these non-voters who develop more negative attitudes
towards immigrants, the EU or the mainstream political parties. First, PRRs propose antiimmigrant, anti-EU policies that follow from their nativist ideology that all non-native elements (immigrants, the European Union, and mainstream parties with pro-EU preferences)
are fundamentally threatening to the nation-state (Mudde, 2007). Second, by referring to a
89
Chapter 4
homogeneous people with a unique national history and identity, PRRs may help to create a
collective form of identification that functions as a ‘new anchor of passionate mobilization’
for non-voters (Mouffe, 2005: p.55). Third, PRRs may give the non-voters the idea that
there is ‘finally’ a party that listens to their grievances, which could trigger them to ‘send a
message’ to mainstream parties that have presumably ignored their problems (cf. Franklin,
2004: p.43). Based on these considerations we expect the following: H4. Individuals who (do
not vote in one wave and) develop more negative attitudes towards immigrants or the EU or
become more politically dissatisfied are more likely to vote for a PRR rather than for any other
party in the subsequent wave.
In addition to the nativist stance, PRRs have been characterized by their populism
(Mudde, 2007). Schumacher and Rooduijn (2013) have argued and shown that two aspects
of populism have been crucial to understanding votes cast for populist parties: anti-elitism
and the evaluation of the party leaders. The anti-elitist notion posits that society is divided
into the ‘naturally good and pure’ people, and the ‘essentially bad, untrustworthy’ elite, two
groups that are fundamentally opposed to each other (Mudde, 2007). Populist parties are
opposed to the mainstream political parties because these parties ignore the interests of the
people, serve their own interests exclusively, and have increasingly turned politics into an
uncontrollable and rational engagement (e.g., Mouffe, 2005). As we hypothesized in H4, it
may be expected that formerly non-voting people will vote for these parties because they
share with PRRs a growing dissatisfaction with the mainstream political parties.
The second aspect to which Schumacher and Rooduijn (2012) refer is the importance of
the populist leader. PRR-leaders use clear language and position themselves as the embodiment of the people to function as controllers or watchdogs over the elite. By doing so,
they make politics more personal and give the impression that they know exactly what the
people want (e.g., Weyland, 2001). Naturally, to be successful thus implies that the ‘people’
see the leader as the embodiment of the people and, at the very least, like the PRR-leader.
Van Holsteijn and Andeweg (2010) have shown that personalization and party-leader
evaluation are especially important for populist parties and it can thus be expected that
once non-voting people become more positive towards a populist leader, they will also be
more likely to vote for a populist party. Because the Socialistische Partij [Socialist Party, SP]
has also been characterized as a populist party (e.g., Lucardie & Voerman, 2012), we expect
that, next to changes in the PRR-leader sympathy, SP-leader sympathy also influences the
voting behavior of non-voters. Hence, we hypothesize: H5. Individuals who (do not vote in
one wave and) develop more positive attitudes towards the PRR-leader or the SP-leader are
more likely to vote for the PRR or the SP than any other party in the subsequent wave.
90
A Dutch ‘populist radical right revolt’
4.3 Data and methods
To answer our research questions and test our hypotheses, we use the Politics and ValuesModule of the Dutch 6-wave panel survey Longitudinal Internet Studies for the Social Sciences (LISS), which is administered by CentERdata (Tilburg University, the Netherlands).
This dataset follows a large, representative group of Dutch respondents and includes a wide
array of topics including voting intentions, political attitudes and socio-economic backgrounds for the period 2007-2013 (i.e., 6 waves).18 The panel consists of 8,204 individuals
coming from 5,000 households and is based on a true probability sample of households
drawn from the population register by Statistics Netherlands (CBS). The original sample
comprised 10,150 addresses, of which 8,204 respondents consented to participate in the
panel (80.8%).19 The Politics and Values-module is part of the core module of the LISSpanel that panel-members must fill out online annually. Households that could not otherwise participate are provided with a computer. To improve representativeness, CentERdata
added two refreshment samples to the original core panel in 2009 and in 2012. The 2009
refreshment was a stratified sample aimed at improving the sample representativeness
regarding household type, age and ethnicity. The 2012 refreshment was randomly drawn
from the population register by Statistics Netherlands (Tilburg University, 2007-2014).
The response rates of the various waves of the panel were high: 6,811 (83%) of the 8,204
approached respondents who were willing to participate in the panel filled out the questionnaire in Wave 1 in December 2007-January 2008; 72.8% of the 8,289 respondents in Wave
2 in December 2008-January 2009; 68% of the 9,398 respondents in Wave 3 in December
2009-January 2010; 73.6% of the 7,328 respondents in Wave 4 in December 2010-January
2011; 80.5% of the 7,372 respondents in Wave 5 in December 2011-January 2012; and lastly,
85.7% of the 6,692 respondents in Wave 6 in December 2012-January 2013. In total, 10,232
respondents (N2) have participated in at least one wave of the LISS-panel. Together, these
respondents filled out 35,695 questionnaires over the various years, meaning that respondents participated in an average of 3.49 waves. We excluded 595 person*year observations
(1.7%) from respondents who were not eligible to vote and were under 18 years old on the
day the interview was taken. Additionally, we excluded 11 person*year observations from
respondents who reported inconsistent information about their gender or age.20
Because we are mainly interested in people who filled out the questionnaire in two
subsequent waves and said that they would not go to the polls in the first of the succeeding
waves, we only included those respondents who mentioned at least once that they would not
go to the polls if there were general elections on the day of the interview and whose voting
intention one wave later are known. Based on that decision, 9,145 respondents (89.4%) and
18 More information about the LISS-panel can be found at: https://www.lissdata.nl/lissdata/About_the_Panel
19 For more information about the recruitment of the first panel, see Scherpenzeel (2009).
20 Nine respondents reported some deviant value for gender and age as compared to their overall pattern. The
deviant observations were excluded from the analysis.
91
Chapter 4
31,944 person*year observations (91.0%) were excluded from the analyses. After exclusion,
we were left with 1,087 respondents (10.6%) who indicated that they would not go to the
polls at least once in 3,148 person*year observations. Lastly, we excluded those observations for which the respondent did not know whether or what to vote for. We performed
person-specific mean imputation (see further) and included those respondents that had
information on all relevant dependent and independent variables in at least two subsequent
waves. The final sample of our analyses comprised 2,861 person*year observations (N1)
nested in 1,077 respondents (N2) for an average of 2.66 observations per respondent. We
found that the non-voters are somewhat underrepresented in the LISS-panel data (see
Tables 4.1a and 4.1b), which is typical for surveys: abstainers are less likely than voters
to participate in surveys (Heath, McLean, Taylor, & Curtice, 1999), and respondents may
report that they vote, whereas they would not go to the polls in the event of real elections
(for incidental reasons). Despite the under-representation of this group of non-voters, we
have no reason to expect that the theorized explanations for this group of non-voters work
differently for non-voters in general.
Table 4.1a Percentage of non-voters and votes for the different parties in the Dutch parliamentary
[Tweede Kamer] elections of 2006, 2010, 2012 (in %)a
2006
Not vote
PVV
ToN
2010
2012
N
%
N
%
N
%
2,409,505
19.65%
3,054,199
24.44%
3,227,587
25.43%
579,490
4.72%
1,454,493
11.64%
950,263
7.49%
0
0.00%
52,937
0.42%
0
0.00%
SP
1,630,803
13.30%
924,696
7.40%
909,853
7.17%
CDA
2,608,573
21.27%
1,281,886
10.26%
801,620
6.32%
PvdA
2,085,077
17.00%
1,848,805
14.79%
2,340,750
18.45%
VVD
1,443,312
11.77%
1,929,575
15.44%
2,504,948
19.74%
GL
453,054
3.69%
628,096
5.03%
219,896
1.73%
D66
193,232
1.58%
654,167
5.23%
757,091
5.97%
CU
390,969
3.19%
305,094
2.44%
294,586
2.32%
SGP
153,266
1.25%
163,581
1.31%
196,780
1.55%
PvdD
179,988
1.47%
122,317
0.98%
182,162
1.44%
50+
0
0.00%
0
0.00%
177,631
1.40%
Other
120,919
0.99%
50,354
0.40%
88,655
0.70%
Blank
16,315
0.13%
26,976
0.22%
37,988
0.30%
12,264,503
100.00%
12,497,176
100.00%
12,689,810
100.00%
Source: Kiesraad (2014).
Note: aInvalid votes belong to the blank category.
92
A Dutch ‘populist radical right revolt’
Table 4.1b Percentage of non-voters and votes for the different parties in the LISS-data (in %)
2008
2009
2010
2011
2012
2013
N
%
N
%
N
%
N
%
N
%
N
%
Not vote
505
9.85%
410
9.68%
554
10.92%
309
7.13%
477
10.39%
462
9.63%
PVV
268
5.23%
332
7.84%
681
13.43%
549
12.68%
476
10.37%
460
9.59%
ToN
724
14.12%
143
3.38%
59
1.16%
0
0,00%
0
0.00%
0
0.00%
SP
628
12.25%
454
10.72%
426
8.40%
443
10.23%
705
15.36%
558
11.63%
CDA
885
17.26%
727
17.16%
692
13.65%
455
10.51%
365
7.95%
439
9.15%
PvdA
609
11.88%
636
15.01%
506
9.98%
554
12.79%
493
10.74%
809
16.86%
VVD
417
8.13%
427
10.08%
533
10.51%
916
21.15%
932
20.30%
555
11.57%
GL
339
6.61%
251
5.93%
346
6.82%
341
7.87%
230
5.01%
92
1.92%
D66
162
3.16%
413
9.75%
703
13.86%
339
7.83%
439
9.56%
521
10.86%
CU
240
4.68%
150
3.54%
191
3.77%
144
3.32%
139
3.03%
166
3.46%
SGP
102
1.99%
91
2.15%
73
1.44%
67
1.55%
91
1.98%
86
1.79%
PvdD
131
2.55%
88
2.08%
128
2.52%
65
1.50%
84
1.83%
85
1.77%
0
0.00%
0
0.00%
0
0.00%
0
0.00%
0
0.00%
290
6.05%
Other
50+
13
0.25%
4
0.09%
9
0.18%
27
0.62%
22
0.48%
40
0.83%
Blank
105
2.05%
110
2.60%
170
3.35%
122
2.82%
138
3.01%
234
4.88%
5,128 100.00% 4,236 100.00% 5,071 100.00% 4,331 100.00% 4,591 100.00% 4,797 100.00%
Source: LISS Panel (Tilburg University, 2007-2014).
Note: Numbers are based on all respondents without missing information about their voting intention.
4.3.1 Operationalization
In this section we describe the operationalization of the dependent, independent and
control variables.
Dependent variable
Voting measures the type of voting option respondents would choose if there were general
elections on the day the interview was taken. Respondents could indicate which party they
would vote for, or whether they would vote blank, would not vote, were non-eligible, did
not know or did not want to say what they would do. The respondents who did not know
or want to say what they would do or were non-eligible were excluded from the sample. To
test Hypothesis 2, we constructed a dummy variable with value 0 for ‘non-vote’ and 1 for a
‘vote for a party or a blank vote’. To test Hypotheses 4 and 5, we constructed the following
variable containing five categories: ‘non-vote,’ ‘vote PRR (PVV/ToN),’ ‘vote SP,’ ‘vote other
party,’ and ‘blank.’
93
Chapter 4
Independent variables
Political efficacy is based on the following six items: ‘1. Parliamentarians do not care about
the opinions of people like me,’ ‘2. Political parties are only interested in my vote and not in
my opinion,’ ‘3. People like me have no influence at all on government policy,’ ‘4. I am well
capable of playing an active role in politics,’ ‘5. I have a clear picture of the most important
political issues in our country,’ and ‘6. Politics sometimes seems so complicated that people
like me can hardly understand what is going on.’ Respondents could answer ‘This is not
true’ (0) or ‘This is true’ (1). We recoded the answer categories of items 1, 2, 3, and 6 such
that the value 0 reflects no political efficacy and value 1 reflects political efficacy. We then
counted all the 1 values and made a sum scale running from 0, ‘no political efficacy’ to
6, ‘much political efficacy.’ Because this variable is highly right-skewed, we constructed a
dummy-variable with value 0 for people who score lower than 2 on that sum scale and value
1 for those who score 2 or higher.
Political interest is measured by the item: ‘Are you very interested in political topics,
fairly interested or not interested?’ Respondents could answer on a 3-point scale, which was
recoded into a dummy variable with value 0 being ‘not interested,’ and value 1 being ‘fairly/
very interested.’
Political dissatisfaction is a mean scale (Cronbach’s α =.94) constructed on the basis of
the items that asked for the respondents’ level of (dis)satisfaction with the following six
institutions: government, parliament, political parties, politicians, the European parliament
and democracy. Respondents could answer all items on a 10-point scale running from 0,
‘very dissatisfied’ to 10, ‘very satisfied.’ The items were recoded such that a higher score for
the variable indicated greater political dissatisfaction. Those respondents with a missing
value in a certain wave were assigned the average value of the values they reported in the
other waves. Respondents without any value on political dissatisfaction in one of the waves
received the average of the other respondents belonging to the specific wave.21 We treat the
variable as an interval-variable.
Anti-immigrant attitudes is a mean scale (Cronbach’s α =.81) constructed on the basis
of the following six items: ‘Where would you place yourself on a scale of 1 to 5, where 1
means that immigrants can retain their own culture and 5 means that they should adapt
entirely?’, ‘It is good if society consists of people from different cultures,’ ‘It should be made
easier to obtain asylum in the Netherlands,’ ‘Legally residing foreigners should be entitled
to the same social security as Dutch citizens,’ ‘There are too many people of foreign origin
or descent in the Netherlands,’ and ‘It does not help a neighborhood if many people of
foreign origin or descent move in.’ Respondent could answer all items on a scale running
from 1, ‘fully disagree’ to 5, ‘fully agree.’ The items were recoded such that a higher score
21 To check for robustness, we performed additional analyses with alternative specifications of the missing
values (assigning the neutral position on a scale, or listwise deletion). The results of these analyses lead to
the same conclusions.
94
A Dutch ‘populist radical right revolt’
on the variable indicated a negative attitude towards immigrants. We performed the same
procedure for missing values as reported for political dissatisfaction and treat the variable
as an interval-variable.
Anti-EU attitudes is based on the following item: ‘Some people and parties say that
European unification should go further. Others think that European unification has already
gone too far. Where would you place yourself on a scale from 1-5 where 1 means that
European unification should go further and 5 means that European unification has already
gone too far?’ We performed the same procedure for missing values as reported for political
dissatisfaction and treat the variable as an interval-variable.
PRR-leader and SP-leader sympathy are based on the following items: ‘What do you
think of (…)’ the PVV-leader ‘(…) Geert Wilders?’ and the SP-leader: ‘(…) Agnes Kant/
Emile Roemer?’ Respondents could answer on a scale from 0-10 with 0 indicating ‘very
unsympathetic’ to 10 ‘very sympathetic.’ We performed the same procedure for missing
values as reported for political dissatisfaction and treat the variables as interval-variables.
Control variables
We also included some control variables that are known to be indicators of voter turnout
(Verba et al., 1995): Female is a dummy-variable with value 1 for females and 0 for males.
Age is a time-invariant variable that measures the age of the respondent when the respondent was interviewed for the first time.
Partner is a dummy-variable that measures whether a respondent lives together with
a partner in the same household. Respondents with a partner received value 1 and those
without a partner received value 0.
Child is a dummy variable that measures whether a respondent has ‘0’ children or ‘1
child or more.’
Education is measured as the highest level of obtained education for which the respondent received a diploma. We recoded the variable using the following five dummy variables:
‘primary education,’ ‘secondary (VMBO/HAVO/VWO),’ ‘intermediate vocational (MBO),’
‘tertiary education (HBO/WO)’ and ‘other.’
Employment status measures the primary activity of a respondent and was recoded into
the following dummy-variables: ‘being employed (incl. freelance, family business),’ ‘being
unemployed,’ ‘student’ and ‘retired.’
Religious attendance is based on the item: ‘Aside from special occasions such as weddings
and funerals, how often do you attend religious gatherings nowadays?’ We recoded this
variable as a dummy variable with value 1 for ‘regular churchgoers: who attend religious
gatherings once a month or more’ and value 0 for ‘non- or irregular churchgoers: those who
attend religious gatherings fewer than once a month.’ We performed the same procedure for
missing values as reported for political dissatisfaction.
95
Chapter 4
Table 4.2 Descriptive statistics of the dependent, independent, and control variablesa
Mean
S.D.
Range
Not vote
0.785
-
0/1
SP
0.026
-
0/1
PVV/ToN
0.046
-
0/1
Other
0.099
-
0/1
Blank
0.044
-
0/1
Political efficacy
0.386
-
0/1
Political interest (1=yes)
0.528
-
0/1
Political dissatisfaction
5.715
1.953
0-10
Anti-immigrant attitudes
3.373
0.666
1-5
Anti-EU attitudes
3.682
1.063
1-5
Sympathy PVV-leader
3.300
2.633
0-10
Sympathy SP-leader
5.002
1.970
0-10
Dependent variable
Vote which party
Independent variables
Control variables
Female (1= yes)
0.553
-
0/1
44.899
16.054
18-88
Partner (1=yes)
0.744
-
0/1
Children (1 =yes)
0.480
-
0/1
Primary
0.082
-
0/1
Secondary
0.318
-
0/1
MBO
0.363
-
0/1
HBO/WO educated
0.178
-
0/1
Other
0.059
-
0/1
Employed
0.543
-
0/1
Unemployed
0.241
-
0/1
Student
0.076
-
0/1
Retired
0.140
-
0/1
Religious attendance (1=regular churchgoer)
0.099
-
0/1
N1 (Person*year-observations)
2861
N2 (Respondents)
1077
N3 (Households) b
874
Age at 1st wave participated
Education
Employment
Source: LISS Panel (Tilburg University, 2007-2014).
Notes: aThese descriptive statistics are based on the pooled panel data and are not controlled for clustering at the
respondent’s level.
b
As the number of respondents is close to the number of households, we ignore the nesting of the respondents in the
households.
96
A Dutch ‘populist radical right revolt’
In Table 4.2 we present the descriptive statistics of the dependent, independent and
control variables in the pooled dataset (i.e. all waves together).
4.3.2 Methods
The LISS-panel survey enables us to study repeated measures of the voting intentions of
Dutch respondents over several years. In this study, we have formulated five hypotheses that
ask for different types of statistical techniques. Hypotheses 1 and 3 address the probability
of non-voters in one election to non-vote or vote for a PRR in the subsequent election. To
test these hypotheses, we make use of transition analyses that show the observed probability
of being in the next-wave voting category compared to the current voting category.
In addition, the LISS-data enable us to analyze variation across respondents and variation within respondents over time (Cameron & Trivedi, 2010). Because the same respondents have been observed repeatedly, we cannot assume that the observations in the panel
data are completely independent. Therefore, panel data need richer estimation methods and
models, such as fixed effect and random effect models that control for the dependency of
observations and correlations of errors over time and within individuals (ibid.: p.235). Since
this chapter is mainly interested in changes within individuals that may be associated with
binary (voting vs. non-voting) and multiple nominal (non-voting, PRR-voting, SP-voting,
other-party and blank voting) outcome variables, we employ logistic and multinomial
logistic fixed effect regressions (e.g., Allison, 2009). We choose fixed effect models because
these only use within-individual variation and allow the unobserved variables to have any
possible association with the observed variables. In contrast, random effect models assume
that the unobserved variables are uncorrelated with the variables of interest, an assumption
that is often violated resulting in biased estimates (Allison, 2009: p.3).
Because Hypothesis 2 addresses a binary outcome variable, voting vs. non-voting, we use
fixed effect logistic regression with conditional maximum likelihood (Allison, 2009: p.32).
To obtain these models, we restructure our dataset into ‘long’ format: for every year that a
respondent is observed there is a unique record, and all records from the same respondent
share the same ID-number. The fixed effect logistic model allows the data to be unbalanced,
i.e., persons can be missing in one or more waves. Because the fixed effect logistic regression only uses within-individual variation over time, respondents who indicated that they
would not go to the polls in the subsequent waves are excluded from the analysis. Moreover,
time-invariant variables such as gender are excluded from the analyses.
For hypotheses 4 and 5, our dependent variable of interest is a nominal variable with
five categories: non-voting, PRR-voting, SP-voting, other-voting and blank. It is therefore
necessary to apply a multinomial logistic regression (e.g., Long & Freese, 2006). We use
Allison’s (2009) hybrid population-averaged fixed effect multinomial logistic regression
97
Chapter 4
model.22 For this method, we have created variables that indicate the respondent’s overall
mean and the wave-specific deviation from that mean for all time-varying predictors. Timeinvariant variables such as gender have no wave-specific deviation and are therefore only
included as the mean variables in the model. Then, we estimate an ordinary multinomial
logistic model including the mean and deviation variables and use cluster-corrected robust
standard errors to control for dependency across observations from the same respondent.
4.4 Results
In this section, we begin by describing the transition probabilities of the voting intention
variable in Table 4.3 to test Hypotheses 1 and 3. Then, we turn to the results of the (multinomial) logistic fixed effect regressions in Tables 4.4 and 4.5 to test Hypotheses 2, 4 and 5.
In Table 4.3 we show the transition probabilities for everyone who reported their voting
intentions in at least two consecutive waves (this is not based on the non-voter specific
sample described above). In Hypothesis 1, we predicted that non-voters are generally more
likely to non-vote rather than to vote in a next election. This prediction is confirmed by
Table 4.3: 60.70% of the non-voters will also intend not to vote in the next wave. This indicates that not intending to vote in the current wave most likely leads to being a non-voter,
rather than a voter, in the next wave (see the statistical test in Table 4.A2 of Appendix 4.A).
Moreover, this table shows that the majority is stable in its voting behavior.
These results largely underline the theoretical notion that non-voting is habitual and
that people who did not intend to vote once are more likely to non-vote in the future. At the
same time, we find that there is evidence for volatility: people do break their habitual nonvoting patterns. We hypothesized that this is mainly due to a change in political interest or
political efficacy (H2) and now focus on the non-voter specific sample. To test Hypothesis
2, we turn to Table 4.4, which reports the results for the fixed effect logistic regressions of
voter turnout probability.23
Consistent with Hypothesis 2, we find in Table 4.4 that a change in an individual’s political interest is strongly associated with a heightened probability of intended turnout. Based
on the coefficients in Table 4.4, we find that a change from no political interest into political
interest is associated with a (e.418*100%≈) 52% increase in the odds of turnout intention in
the next wave. We do not find such a significant effect of the change in political efficacy.
This means that non-voters who become more politically efficacious are not more likely to
vote in the subsequent wave, which partly contradicts Hypothesis 2.
22 So far, there is no statistical package that enables researchers to model a pure fixed effect multinomial
logistic regression model (Allison 2009). Pforr (2013) has developed such an implementation for Stata, but
he is still awaiting the decision to allow him to implement it (personal communication, December 2013).
23 As a robustness check, in Table 4.A2 of Appendix 4.A we show the results of an ordinary pooled logistic
regression with robust standard errors for clustering at the respondent’s level.
98
A Dutch ‘populist radical right revolt’
Table 4.3 Transition Analysis for people that have information on the voting behavior variable in at least
two consecutive waves of the LISS Panel
PARTY t1:
VVD
84
62
19
40
5
37
124
1,547
60.70
8.40
4.72
2.20
5.43
4.01
1.23
2.59
0.32
2.39
8.02
100
PVV/ToN
132
1,506
101
71
53
303
17
49
12
75
65
2,384
%
5.54
63.17
4.24
2.98
2.22
12.71
0.71
2.06
0.50
3.15
2.73
100
SP
61
64
1,317
26
192
37
83
106
8
75
29
1,998
%
3.05
3.20
65.92
1.30
9.61
1.85
4.15
5.31
0.40
3.75
1.45
100
51
47
43
1,700
58
207
17
105
35
33
26
2,322
2.20
2.02
1.85
73.21
2.50
8.91
0.73
4.52
1.51
1.42
1.12
100
2,159
VVD
%
GL
%
D66
%
CU/SGP
%
Other
%
Blank
Other Blank
Total
PvdA
%
CU
34
PvdA
D66
CDA
%
GL
73
CDA
SP
PARTY t2:
130
%
PVV/ToN
939
Not vote
Not vote
87
41
149
27
1,512
37
78
159
4
29
36
4.03
1.90
6.90
1,25
70.03
1.71
3.61
7.36
0.19
1.34
1.67
100
79
135
28
97
37
1,733
10
187
11
75
51
2,443
3.23
5.53
1.15
3.97
1.51
70.94
0.41
7.65
0.45
3.07
2.09
100
27
11
111
13
140
11
583
129
8
27
15
1,075
2.51
1.02
10.33
1.21
13.02
1.02
54.23
12.00
0.74
2.51
1.40
100
26
20
103
37
156
137
92
983
3
29
18
1,604
1.62
1.25
6.42
2.31
9.73
8.54
5.74
61.28
0.19
1.81
1.12
100
5
8
13
43
7
5
11
15
545
18
3
673
0.74
1.19
1.93
6.39
1.04
0.74
1.63
2.23
80.98
2.67
0.45
100
761
28
43
28
20
26
12
24
16
18
524
22
3.68
5.65
3.68
2.63
3.42
1.58
3.15
2.10
2.37
68.86
2.89
100
99
41
25
15
24
41
11
13
2
19
94
384
%
25.78
10.68
6.51
3.91
6.25
10.68
2.86
3.39
0.52
4.95
24.48
100
Total
1,534
2,046
1,991
2,083
2,289
2,585
945
1,802
651
941
483
17,350
8.84
11.79
11.48
12.01
13.19
14.90
5.45
10.39
3.75
5.42
2.78
100
%
Source: LISS Panel (Tilburg University, 2007-2014).
Notes: The total number reflects the total number of transitions: the sum of the total number person*year observations per respondent minus 1.
a
See Table 4.A1 in Appendix 4.A for the statistical test that non-voters are more likely to vote for PVV/ToN than for
any other party.
In Hypothesis 3, we predicted that for the part of the electorate that breaks its habitual
non-voting pattern, a PRR is the most attractive party for which to vote. Table 4.3 shows
that PVV/ToN receives relatively the greatest number of votes from former non-voters:
8.4% of the votes in a subsequent wave come from former non-voters. The populist left SP
follows with 4.7%. Table 4.A1 in Appendix 4.A confirms that former non-voters are indeed
significantly more likely to turn to PRRs than to any other party, and we thus find evidence
99
Chapter 4
Table 4.4. Results of the fixed effect binomial logistic model of voting behavior (1=vote, 0= non-vote,
robust standard errors) a
Independent variables:
b
S.E.
Political efficacy (1=high)
0.243
0.164
Political interest (1=yes)
0.418 **
0.182
Control variables b
Partner (1=yes)
-0.878
0.523
Children (1=yes)
0.693
0.516
-0.291
0.995
0.400
1.031
HBO/WO educated
-0.851
1.185
Other
-0.099
0.840
Unemployed
-0.275
0.431
Student
-0.374
0.615
Retired
-0.247
0.640
0.173
0.401
t3
0.928 **
0.196
t4
2.559 **
0.220
t5
1.274 **
0.240
2.686 **
0.236
Education (primary=ref.)
Secondary
MBO
Employment status (employed=ref.)
Religious attendance (1=regular churchgoer)
Wave (t2=ref.) c
t6
N1
1,648d
N2
564
Log-likelihood
-443 Source: LISS Panel (Tilburg University, 2007-2014).
Notes: *.01<p<.05, **p<.01.
a
See Table 4.A2 in Appendix 4.A for a pooled logistic regression model that ignores the panel structure of the data.
b
The control variables age at first wave and gender were omitted from the analyses because of no-within variance.
c
Due to our selected respondents, the first wave of a respondent always refers to a situation where the respondent did
not vote. This implies that the dependent variable is perfectly predicted by including the first wave dummy. As such,
we exclude the first-wave and use the second wave-dummy as reference category for the wave-variable.
d
Those respondents who did not change in the dependent variable at all (always non-vote) are excluded in a fixed
effect model (i.e. 1,213 person*year observations were excluded from the analyses).
for Hypothesis 3 that if non-voters go to the polls, they are more likely to vote for a PRR
than for another party.
Lastly, we discuss the tests of Hypotheses 4 and 5. We turn to Table 4.5 that shows
the results of the fixed effect multinomial logistic regression of voting behavior. In Models
1 and 2, we provide the results of the comparisons between the non-voters, PRR-voters,
100
0.236
0.042
Anti-immigrant attitudes
Anti-EU attitudes
Sympathy PVV-leader
Anti-EU attitudes
0.037
**
0.185
-0.266
Political dissatisfaction
Anti-immigrant attitudes
*
**
-0.332
-0.765
Political efficacy
Political interest
0.380
-0.467
-0.679
Employed
Religious attendance
Means:
1.079
0.793
Education (high)
0.770
0.087
0.131
0.044
0.173
0.181
0.565
0.581
-0.002
-0.850
Partner
Child
-0.572
Age c
0.394
0.939
0.265
0.098
0.027
-2.368
-0.313
-0.667
0.705
-1.405
0.108
*
0.417
*
**
**
0.164
0.218
0.074
0.288
0.315
1.045
0.621
0.945
0.833
0.887
0.066
0.191
0.462
0.124
0.325
-0.572
0.229
-0.434
0.451
-1.206
-0.101
-0.538
-2.215
0.337
0.271
-0.264
0.489
0.113
-0.644
0.681
**
*
**
0.191
0.270
0.085
0.386
0.407
1.081
0.737
1.210
1.263
1.081
0.086
0.211
0.484
0.132
0.444
0.448
S.E.
0.050
0.203
0.184
-0.599
0.037
b
0.032
0.038
-0.310
0.180
-0.746
-0.345
-0.613
-0.415
0.744
-0.904
0.064
-0.563
-0.108
Control variables
0.048
0.359
b
*
**
**
**
*
*
Not-vote
vs. other
Sympathy SP-leader
**
S.E.
SP
vs. other
0.074
0.121
0.511
0.077
-0.001
0.100
b
PRR
vs. other
Model 1
Sympathy PVV-leader
0.113
0.293
0.079
0.274
*
0.184
*
-0.616
Political interest
Political dissatisfaction
0.241
S.E.
b
Not-vote
vs. other
0.032
Political efficacy
Deviations:
0.032
0.088
0.138
0.051
0.172
0.181
0.574
0.378
1.110
0.582
0.792
0.047
0.084
0.063
0.113
0.297
0.079
0.276
0.242
S.E.
0.608
0.401
0.127
0.176
0.478
-0.564
-2.045
-0.571
-0.416
1.059
-0.357
0.197
-0.181
0.673
0.047
0.062
0.202
0.083
0.226
b
**
*
*
**
PRR
vs. other
Model 2
S.E.
0.061
0.177
0.269
0.094
0.328
0.360
1.529
0.653
1.391
0.942
1.011
0.081
0.112
0.104
0.214
0.444
0.128
0.424
0.399
Table 4.5 Results of the fixed effect multinomial logistic regression of party choice of former non-voters (robust standard errors) ab
-0.150
0.255
-0.170
0.351
-0.461
0.613
-1.331
-0.107
-0.259
-1.652
-0.139
0.258
0.420
0.254
-0.288
0.299
0.182
-0.795
0.695
b
*
**
**
*
*
SP
vs. other
0.068
0.194
0.286
0.095
0.375
0.421
1.168
0.907
1.424
1.195
1.320
0.091
0.164
0.127
0.220
0.520
0.144
0.462
0.434
S.E.
A Dutch ‘populist radical right revolt’
101
102
0.150
0.219
.609
0.048
-0.210
2.860
0.113
2,861
1,077
Employed
Religious attendance
Constant
Pseudo R2
N1
N2
1,077
2,861
0.113
-7.474
-1.318
0.084
-0.780
0.034
0.357
0.001
-0.120
b
**
*
*
PRR
vs. other
Model 1
1.099
0.585
0.267
0.395
0.268
0.302
0.008
0.232
S.E.
1,077
2,861
0.113
-1.025
-0.998
-0.076
-0.103
0.143
-0.980
-0.006
-0.058
b
**
SP
vs. other
1.154
0.617
0.303
0.385
0.307
0.306
0.009
0.287
S.E.
1,077
2,861
0.172
3.147
-0.216
0.036
0.131
0.241
-0.289
-0.004
-0.188
-0.037
B
*
Not-vote
vs. other
S.E.
0.705
0.221
0.151
0.178
0.158
0.161
0.005
0.131
0.045
1,077
2,861
0.172
-6.120
-1.212
0.241
-0.567
0.209
0.184
0.007
0.008
-0.302
b
**
*
**
PRR
vs. other
Model 2
S.E.
1.381
0.546
0.277
0.417
0.289
0.333
0.008
0.247
0.080
1,077
2,861
0.172
-5.545
-0.997
0.020
-0.074
0.135
-0.905
-0.007
-0.027
0.525
b
**
**
**
SP
vs. other
S.E.
1.468
0.627
0.306
0.374
0.302
0.315
0.010
0.293
0.099
Source: LISS Panel (Tilburg University, 2007-2014).
Notes: *.01<p<.05, **p<.01
The results for the other comparisons (e.g., ‘blank vs. not-vote’) are not shown here for space requirements.
a
Additional analyses show that the results are similar when listwise-deletion was performed and when the values of the participants who answered ‘don’t know’ on the various
scales (e.g., anti-immigrant attitudes) were imputed with the scales’ midpoint.
b
In a random effect multinomial logistic model, it is assumed that the mean and the deviation coefficients are identical. This hybrid model relaxes that assumption as we are
not interested in the between-variation here.
c
Since the deviation in age and time-waves are perfectly correlated, we do not need to include wave-dummies in this analysis.
0.157
0.178
0.248
0.125
Child
Education (high)
**
0.159
-0.265
Partner
0.005
-0.188
-0.004
0.131
S.E.
Female
b
Not-vote
vs. other
Age
Control variables:
Sympathy SP-leader
Table 4.5 Results of the fixed effect multinomial logistic regression of party choice of former non-voters (robust standard errors) ab (continued)
Chapter 4
A Dutch ‘populist radical right revolt’
SP-voters and the other party-category. For Hypothesis 4, we focus on Model 1, whereas in
Model 2, we include the PRR-leader and SP-leader sympathy variables to test Hypothesis
5. The ‘mean’-variables in Table 4.5 basically refer to the between-person variation, in
which we are not presently interested. The coefficients of interest belong to the ‘deviation’variables. These deviation-coefficients are comparable to fixed effect coefficients, such as
those presented in Table 4.4, because they are estimated using only within-person variation
(Allison 2009: p.40).
In Hypothesis 4, in line with earlier studies on the determinants of PRR-voting, we
expected that it is the change in typical PRR-attitudes that may explain why people decide
to vote for a PRR rather than for another party. Therefore, we hypothesized that once nonvoters develop more negative attitudes towards immigrants and/or the European Union
and/or grow more politically dissatisfied, the probability to vote for a PRR rather than
another party increases. In the comparison between PRRs and other parties in Model 1
in Table 4.5, we find that this hypothesis cannot be confirmed with regard to any of the
typical PRR-attitudes. As non-voting individuals develop more negative attitudes towards
immigrants, the EU or politics in general, they are not more likely to vote for a PRR rather
than another party. The deviation coefficients for neither anti-immigrant attitudes nor antiEU attitudes nor political dissatisfaction reach significance, which contradicts Hypothesis
4. The significant ‘mean’-coefficients for the typical PRR-attitudes accompanying the ‘PRR
vs. other’-comparison, however, indicate that non-voters who turn to PRRs are on average
more negative towards immigrants, the EU and politics than other non-voters. This suggests that non-voters can only be mobilized by PRR-leaders if they share the PRR-ideology.
Lastly, we turn to the test of Hypothesis 5 which stated that having greater sympathy
for a populist leader explains why people are more likely to vote for a populist party in a
subsequent wave. In Model 2 in Table 4.5, where we compare the PRR to the other parties,
we find that the deviation coefficients of the party leader variables corroborate the findings:
a more positive evaluation of the PVV-leader Wilders induces people to vote for a PRR
rather than another party. A one unit increase in the PRR- or SP-leader sympathy is associated with, respectively, a significant (e.673*100%≈) 96% or (e.418*100%≈) 52% higher relative
risk ratio of voting for a PRR or SP than voting for another party. We thus find evidence
that developing more positive feelings towards the leader of the PVV or the SP over time
is significantly associated with an increased propensity to vote in the subsequent wave for
the PVV or the SP. These results are in line with Schumacher and Rooduijn (2013) who also
find that sympathy for populist leaders plays an important role in the decision to vote for
a populist party.
103
Chapter 4
4.5 Conclusion
In this chapter, we departed from the scholarly debate about the positive and negative consequences of PRRs. Based on the idea that PRRs have the potential to mobilize politically
disengaged voters and could thus be seen as a corrective for depoliticizing democracies (see
Mudde & Rovira Kaltwasser, 2012), we argued that successful PRRs may trigger non-voters
to reconsider their ‘habitual’ non-voting behavior. We theorized that, in general, non-voters
are most likely to be habitual in their voting behavior but that with increasing political
efficacy and interest, their propensity to vote also increases. Next, we argued that PRRs are
particularly attractive alternatives for them because these parties have recently positioned
themselves as the embodiment of the people, use clear language and listen to the grievances
of the common man on the street. Therefore, we hypothesized that voting for a PRR rather
than another party can be explained by the change in typical PRR-attitudes and sympathy
for the party leader among the (habitual) non-voters. By analyzing the LISS-panel (Tilburg
University, 2007-2014), we were able to test these hypotheses using binary and multinomial
logistic fixed effect regressions for the period 2007-2013.
Our results show that despite increasing levels of electoral volatility in the Netherlands,
people are generally most likely to be stable in their voting behavior. These results speak
to the growing literature on habitual voting (Plutzer, 2002): (non-)voters are on average
most likely to (non-)vote for the same party in subsequent elections. Yet, this study also
demonstrated that a minority of the non-voters do ‘break through’ their habitual voting
pattern. In particular, we found that a change in one’s interest in politics can explain why
people who do not normally vote become more likely to go to the polls in a subsequent election. This is not a novel finding (e.g., Verba et al., 1995), but it is striking that only a change
in political interest, rather than a change in structural individual characteristics–such as
employment status–can explain why non-voters decide to go to the polls in later elections.
It could be that these structural changes take place later in life and that the habitual (non-)
voting pattern has already set in. Admittedly, the question of how political interest is triggered remains.
Our expectation was that PRRs are especially able to trigger this political interest and
that non-voters therefore tend to vote for a PRR once they decide to vote. This hypothesis is
confirmed by our results. We find that if non-voters decide to vote in a subsequent election,
they are generally more likely to vote for a PRR than for any other party. This result is in line
with the notion that PRRs may be able to ‘mobilize excluded sections of society’ (Mudde
& Rovira Kaltwasser, 2012: p.21). Surprisingly, the results of our fixed effect multinomial
logistic analyses show that this specific volatility in changing from non-voting to a PRRvote cannot be explained by a change in the typical PRR-attitudes. Non-voters turning to
PRRs seem to have some ideological overlap with PRRs, but a change in sympathy for PRRleaders is particularly strongly associated with a PRR-vote. This is an important finding as
104
A Dutch ‘populist radical right revolt’
it suggests that changes in ideological considerations among non-voters do not explain why
non-voters turn to a PRR, and thus, that other factors should explain why PRRs become
successful. In line with Schumacher and Rooduijn (2013), we find that the evaluation of
party leaders is decisive in understanding why PRRs are successful and able to attract politically disengaged voters.
Still, we note that there are some shortcomings of this chapter that prevent us from
conclusively claiming that PRRs are able to mobilize people politically. A major shortcoming of this chapter is that we use voting intention rather than real voting behavior. Despite
the fact that these two constructs are strongly associated, it leaves us with the question of
whether people actually behave differently in subsequent elections and shift from being
non-voters to PRR-voters. Voting intention is the best proxy for volatility in voting behavior
among non-voters, however, as we do not have a panel study that includes sufficient nonvoters who changed their votes in two or more consecutive parliamentary elections. Future
research should utilize longitudinal data that follow a larger panel of people over a longer
period of time.
In addition, this chapter does not provide evidence of whether people are also appalled
by the anti-immigrant, anti-elite discourse and are therefore more inclined to vote for
another party or stop voting. Moreover, we are unable to explain why some people change
earlier than others: why is it that some people turned to a PRR in 2008 and others in 2009?
In Chapter 5 and 6, we suggest that religious people and women may be more hesitant
to vote for a (new) PRR than non-religious people and men, but that once PRR grow in
popularity and become more established, these differences may disappear. This is especially
interesting in periods when PRRs become part of governments and thus become part of the
establishment that they usually see as untrustworthy. Hence, we need further comparative
research to more comprehensively test these hypotheses and discover whether there are differences between various contexts and across different social groups (in line with Chapter
3) in the extent to which, when and why individuals are motivated or demotivated to go to
the polls.
Nevertheless, this is the first study that uses longitudinal panel data to investigate the
mobilization success of PRRs. Based on this chapter we tentatively conclude that a Dutch
PRR-leader who is appreciated by the electorate may well have the ability to trigger some
non-voters to take an active part in democracy. This could be regarded as a curative for
democratic functioning in a Western-European society.
105
Part II
Populist radical right parties
and cleavage voters
107
Chapter 5
Religion and populist radical right voting
This chapter was written together with Eva Jaspers and Marcel Lubbers. A slightly
different version of this chapter is published as: Immerzeel, T., E. Jaspers & M. Lubbers
(2013). Religion as Catalyst or Restraint of Radical Right Voting? West European Politics
36(5): 946-968, DOI: 10.1080/01402382.2013.797235. Earlier drafts were presented at
the Annual Meeting of the International Society of Political Psychology (ISPP, Chicago,
United States, July 2012) and at the ‘Dag van de Sociologie’ (Utrecht, the Netherlands,
May 2012).
Abstract
It is often found that religious people are underrepresented among the populist radical
right electorate, despite PRRs’ claim of being defenders of the Judeo-Christian society.
In this chapter, we investigate this paradoxical finding and examine to what extent two
dimensions of religion–practice and belief–play a role in voting for a PRR across seven
Western European countries. Using the European Values Study from 2008, we find that
the religiously active people are indeed less likely to vote for a PRR, because they tend
to vote for a Christian party. However, we challenge the common wisdom that religion
restrains from PRR-voting alone and show that orthodox believers in three countries
(Belgium, Norway and Switzerland) have more negative attitudes towards immigrants
and therefore are more likely than their mainstream counterparts to vote for a PRR.
109
Religion and populist radical right voting
5.1 Introduction
It is often shown that religiosity is negatively related to the likelihood of PRR-voting: people
who more often attend religious services are less likely to vote for PRRs (Billiet, 1995; Lubbers & Scheepers, 2000; Lubbers et al., 2002). This empirical regularity is explained by the
idea that religious people are more integrated in religious communities that have the norm
to vote for a Christian party, which makes them more likely to vote for traditional Christian
parties rather than PRRs that promote intolerant, xenophobic ideologies (Billiet, 1995). Additionally, PRRs have not been regarded as credible alternatives to religious voters because
Christian religiosity is a non-issue in PRRs party programs (Arzheimer & Carter, 2009).
Recently, however, the Western European anti-immigrant PRRs which are mainly
secular movements (Camus, 2007) emphasize the religious divide by presenting themselves as the safeguards of the ‘Judeo-Christian societies’ that defend the people from the
Islamic threat (e.g., Wilders, 2012). Camus (2007) showed that some of the contemporary
European PRRs have their ‘roots in the alliance between religious fundamentalism and
counter-revolutionary ideas’ which contrasts Arzheimer and Carter’s idea. Moreover, Israeli scholars find that the motives of Israeli people to vote for a religious fundamentalist
party largely coincide with the motives to vote for a PRR (Hirsch-Hoefler et al., 2010),
though we note that they dropped religiosity as an explanatory factor. The overview of the
recent positions of PRRs on religion and earlier empirical evidence point at a paradoxical
situation, which could indicate that–in Allport’s famous words–there is something about
religion that ‘makes and unmakes for’ (Allport, 1966) PRR-voting.
In this chapter, we focus on this paradoxical relationship and investigate what role
religion, as classic societal cleavage (Lipset & Rokkan, 1967), has for the voting behavior
of native Christian inhabitants of contemporary Western European secularizing societies
(Achterberg et al., 2009; Knutsen, 2004; Van der Brug et al., 2009). Previous studies on
religiosity as main cleavage in voting behavior show that Christian people mainly vote for
Christian parties. However, as suggested above, there are reasons to expect that PRRs are
credible alternatives for Christian people.
One reason addresses the most important incentive generally found to vote for a PRR:
anti-immigrant sentiments or attitudes (Ivarsflaten, 2008). The increasing number of immigrants has fueled the prevalence of other religions and in particular Islam, in Western
European countries, with religious leaders propagating fundamental different truth claims
than their Christian counterparts. We anticipate that especially orthodox Christians, that
is Protestants and Catholics with orthodox beliefs, may feel religiously threatened by the
increasing prevalence of other religions brought to the countries by immigrants (Abu Raiya
et al., 2008), which could intensify their anti-immigrant attitudes. This in turn may make
anti-immigrant PRRs preferred over Christian parties that do not have or do not emphasize
an anti-immigrant stance. Thus, we counter the norm-compliance argument that Christian
111
Chapter 5
people tend to vote for Christian parties, and anticipate that there is more in religion that
may drive the vote for a PRR instead of a Christian party. Our central research question
is therefore: To what extent are PRRs able to attract religious voters who are not generally
expected to vote for these parties? If so, why? By answering this question, we provide more
comprehensive insight in how and to what extent the religious cleavage affects people’s
voting behavior in secularizing countries.
Our main contribution lies in the Weberian distinction between religious practice and
the content of one’s religious beliefs that both may strongly affect one’s (voting) behavior
(Driskell, Embry, & Lyon, 2008; Weber, [1930] 2003). This theoretically and empirically
broader conceptualization of religiosity is inspired by prejudice scholars who show that
these dimensions of religiosity are differentially related to prejudice (e.g., Reitsma, 2007;
Rowatt & Franklin, 2004; Scheepers et al., 2002; Scheepers, Gijsberts, & Hello, 2002; Van
der Slik & Konig, 2006) and, moreover, builds upon earlier studies relating religion to PRRvoting that only introduce church attendance and denomination as indicators of religiosity
(e.g., Billiet, 1995). Notably, Arzheimer and Carter (2009) introduce a broader conceptualization - encompassing attendance, belonging, praying, and subjective religiosity - but use
a composite measure that inhibits investigation of the potential different relations between
the religious dimensions and PRR-voting. We hypothesize and test whether religious practice, i.e. church attendance, and religious beliefs, i.e. orthodoxy, are differentially associated
with the likelihood of PRR-voting.
Furthermore, we investigate the religious practice evidence that higher integration
into churches relates to a lower likelihood of PRR-voting. We argue that integration in
Christian communities may be accompanied by anti-immigrant attitudes, but that this does
not translate into a PRR-vote as long as integration is strong. Once people disintegrate
from churches, the norms of voting for a Christian party become less pressing and these
anti-immigrant attitudes (Mudde, 2007) become more important in the decision to vote for
a PRR. Thus, we expect and test whether one’s anti-immigrant attitudes become more vital
the less people attend a religious service, making the vote for a PRR more likely.
To test our hypotheses, we use the 2008 round of the European Values Study (EVS,
2010) that encompasses much information about the two dimensions of people’s religiosity,
party preference, anti-immigrant attitudes, and various socio-demographic characteristics.
We focus on native, Christian believers and their voting behavior in seven Western European countries. We choose these seven Western European countries where the traditional
religious cleavage has strongly dominated 20th and also 21st century party politics in these
Western European countries, whereas the communist legacy has made it almost impossible
for this cleavage to emerge in Eastern Europe (e.g., Judt, 2010; Knutsen, 2004; Lipset &
Rokkan, 1967; Van der Brug et al., 2009). Moreover, in line with various scholars, we argued
that PRRs may have greatly contributed to a new line of cultural conflict that, next to the
112
Religion and populist radical right voting
traditional economic conflict, largely structures party competition and voter’s party affiliations (e.g., Aarts & Thomassen, 2008; Kriesi et al., 2008; Pellikaan et al., 2007). Therefore,
PRRs have the potential to become ‘catch-all’ parties that contribute to the crumbling of
traditional cleavages like the religious cleavage. We reckon that if we want to observe and
explain such cleavage-breaking patterns of PRRs, we should focus on countries that are
known for having had a strong Christian democratic tradition and have a popular and
credible PRR nowadays.
In the following section, we present our theoretical arguments and hypotheses regarding the different effects of the religious dimensions on the likelihood of PRR-voting. Next,
we present a description of the data and the applied methods. Then we present our results
and conclude and discuss our most important results and implications for further research
in the last section.
5.2 Theory and hypotheses
To improve our understanding of the relationship between religiosity and PRR-voting,
we have argued that a distinction should be made between two dimensions of religiosity:
religious practice and religious beliefs. In this section, we present theoretical arguments
about how these two dimensions are differentially related to PRR-voting. We emphasize
that all presented arguments concern the native Western European Christian believers and
their voting behavior and that we refer to that group when we mention ‘the religious people’.
5.2.1 Religious practices
Religious practices usually refer to church attendance and membership. In contrast with
religious beliefs, religious practices have been often related to the likelihood of radical right
voting (e.g., Billiet, 1995; Lubbers & Scheepers, 2000). Particularly the social disintegration theory (Arendt, [1951] 1973) has inspired first-wave PRR-scholars (see Chapter 1) to
include measures of religious practices. Namely, according to Arendt ([1951] 1973: p.317),
it was only natural that ‘the masses of a highly atomized society (…) have tended toward an
especially violent nationalism’. Individuals living in divided, socially disintegrated societies
are particular likely to support ethno-nationalist and populist politics (Rydgren, 2007),
because PRRs have a nationalist program that offers the socially disintegrated a sense of
belonging (Fontana et al., 2006).
Based on this theory, scholars argued that being an active member of a religious community is an expression of social integration, and would thus result in a lower likelihood to
PRR-voting. Moreover, Lubbers and Scheepers (2000) argue that these religious communities spread the norms of voting for a Christian party, rather than another party, and that it
thus can be expected that the more integrated one is in church, the more likely one is to vote
113
Chapter 5
for a Christian party rather than a PRR. Hence, this leads to the following hypothesis: H1.
The more religiously integrated one is (more often attends religious services), the more likely
one is to vote for a Christian party instead of a PRR.
The abovementioned hypothesis assumes that highly integrated religious people–independent of their other attitudes–vote for a Christian party. As Arzheimer and Carter (2009)
put it: religion has an ‘encapsulating’ effect on party choice. It is not necessarily because
they differ from PRR-voters in their attitudes that they do not vote for a PRR–they follow
the norm to vote for a Christian party, even when the PRR is proximal closer. Obviously,
once religious people become less integrated they will be less likely to adhere to the norm
to vote for a Christian party–the religious ‘encapsulating’ effect decreases–and other attitudes become more important for their party choice. To test the encapsulating effect of
religiosity stricter than before, we propose that strong anti-immigrant attitudes will not
result in a PRR-vote only among highly integrated religious members. Hence, we expect an
interaction effect of religious practices with anti-immigrant attitudes: H2. Anti-immigrant
attitudes become more important for the decision to vote for a PRR instead of a Christian
party, the less religiously integrated one is.
5.2.2 Religious beliefs
Next to religious practices, religion entails a content dimension that deals with the theological teachings and beliefs–which has mainly been neglected in PRR-research. In psychological literature much attention has been paid to this dimension, referring to religious
fundamentalism as the most extreme position towards theological teachings (e.g., Abu
Raiya et al., 2008; Altemeyer & Hunsberger, 1992; Altemeyer, 2005; Blogowska & Saroglou,
2011). The psychological literature provides the religious coping theory (Pargament, 1997)
that describes how religious beliefs may be associated with anti-immigrant attitudes and
could help to understand how religious beliefs may relate to PRR-voting.
Religious coping theory states that people are ‘proactive, goal-directed human beings
who constantly search for meaning and significance in their lives’ (Abu Raiya et al., 2008).
Once people relate this quest to a sacred god or another divine object, it inspires people to
preserve and protect the sacred from various threats and violations by developing distinct
coping mechanisms. One such threat is the prevalence of another religion which truth is
based on another god or divine object and thus undermines the fundamental truth of one’s
own sacred god or object. It is assumed that people who do not relate their significance to
a divine god will not feel such religious threat: it is the divine connection which is fundamental to various religions that drives religious coping.
This theory has been tested in multicultural Western European societies where increasing numbers of Muslims are interpreted as threats to Christians. Various scholars investigated whether Christians may be more prejudiced towards Muslims as a result of coping
114
Religion and populist radical right voting
with the potential threat of Muslims (e.g., Abu Raiya et al., 2008; Rowatt & Franklin, 2004;
Van der Slik & Konig, 2006). In line with their results, we argue that one’s religious coping
method is associated with the content of one’s religious beliefs and anticipate that people
with an orthodox belief system–based on ‘teachings that clearly contain the fundamental
truth about humanity and deity’ (Altemeyer, 2005)–may develop more negative attitudes
towards (Muslim) immigrants24 than people who have a more mainstream, tolerant belief
emphasizing Christian charity. This is because their own truth-claim is (passively and actively) undermined by groups not accepting the traditional Christian teachings (Scheepers
et al., 2002). In line with Abu Raiya et al. (2008) who show that more orthodox people are
more likely to feel religiously threatened by the presence of Muslims, we expect orthodox
people to have stronger anti-immigrant attitudes than more mainstream, tolerant believers.
A consequence of these stronger anti-immigrant attitudes is the vote for a political
party taking an anti-immigrant position as well (Mudde, 2007). In some countries religious
fundamentalist (RF) parties exist, e.g., the re-reformed SGP in the Netherlands (HirschHoefler et al., 2010), that are strong articulators of this threat. In most Western European
countries such RF parties do, however, not exist and in that case PRRs may be evaluated by
orthodox believers as the parties that best ‘safeguard the Judeo-Christian societies’. Thus,
we hypothesize that: H3. Orthodox believers have more negative attitudes towards immigrants than mainstream believers, and therefore are more likely to vote for a PRR instead of
a Christian party – H3a. Except for the Netherlands where a Religious Fundamentalist party
exists.
5.3 Data and methods
To test our hypotheses we use the 2008 round of the European Values Study (EVS, 2010),
which is a standardized, cross-sectional survey that is a collection of representative samples
of 1500 people per country in 47 European countries.25 In all countries, respondents were
selected by means of stratified random samples of the resident populations aged 18 years
and older. Since we are interested in how PRRs may affect the role of the religious cleavage
in secularizing Western European countries, we focus on the following seven countries:
Austria, Belgium (Flanders), Finland, Italy, the Netherlands, Norway, and Switzerland.
These countries have a rather similar party composition regarding presence and popularity of PRRs (and sufficient PRR voters for our analyses), and generally have at least one
Christian democratic party that has received a fairly stable number of votes over the last two
24 In the countries under study Muslims are considered as most important non-Western immigrant group, with
their increasing presence (instead of other groups of immigrants) being an important macro-indicator of
PRR-voting (Coffé et al., 2007). The proposed anti-immigrant measure can therefore be seen as a proxy of
anti-Muslim sentiments.
25 For more information about the EVS: http://www.europeanvaluesstudy.eu/
115
Chapter 5
decades. We acknowledge that next to Christian parties, also the presence and popularity
of conservative parties may influence the individual vote of Christian people. The membership in the European People’s Party of both conservative and Christian parties illustrates
that conservative and Christian parties tend to have similar ideologies. Moreover, we find
in Chapter 2 that conservative and Christian democratic parties take similar positions on
various typical PRR-issues; and earlier studies show that Christian people have a tendency
to vote for conservative parties in countries with an unpopular or no Christian party (e.g.,
Van der Brug et al., 2009). In addition, the Christian parties generally fill the ‘conservative
gap’ in the political spectrum in the absence of an explicitly conservative party (e.g., the
Netherlands). For these reasons and because of statistical power problems that arise with
small numbers of Christian voters in the EVS (2010), we decided to allocate the conservative and Christian-voters into the same category in four countries (Belgium, Switzerland,
Finland and Norway).
Furthermore, since we are interested in how Christian respondents vote, emphasizing
the difference between orthodox and mainstream believers, we excluded respondents who
were non-Christian: those who adhered to another religious denomination (e.g., the Muslims) and the non-believers. This resulted in a total N of 1,943 respondents with a Christian
belief system in seven countries.
5.3.1 Operationalization
Dependent variable
Party choice uses the item ‘What party would you vote for, if there were general elections
tomorrow?’ which is asked when people assent to the question ‘Would you go voting if there
were general elections tomorrow?’ People that would not go voting or were not eligible to
vote are excluded from the analysis. We distinguish the PRR, Christian, and conservative
parties based on the following three sources: the membership of the political parties in the
European Parliament parties, the congruency in policy position on various topics according
to the Expert Judgment Survey of European Political Parties 2010 (Immerzeel et al., 2011)
and the PRR-categorization of Mudde (2007).26 All other party families (social democratic,
liberal, green, communist/socialist), other parties, and blank ballots were taken together in
the ‘other party’-dummy.
Independent variables
Religious practice is operationalized by religious attendance. Religious attendance is measured by the item: ‘Apart from weddings, funerals and christenings, about how often do you
attend religious services these days?’ Respondents could answer from 1 ‘never’ to 7 ‘more
than once a week’. Due to the non-normal distribution of this variable in various countries,
26
116
See Appendix 5.A, Table 5.A1 for an overview of the political parties in the different countries.
Religion and populist radical right voting
we decided to recode this variable in three categories ‘attends less than once a year’, ‘attends
once a year – less than once a month’, and ‘attends once a month or more’.
Religious belief, or orthodox belief, is constructed by a religious particularism item and
five doctrinal belief items. Religious particularism is measured by the question: ‘Which
of the following statements best describes your view? 1. There is only one true religion; 2.
There is only one true religion, but other religions do contain some basic truths as well; 3.
There is not one true religion, but all great world religions contain some basic truths; and
4. None of the great religions have any truths to offer.’ The doctrinal belief items refer to the
extent respondents say that they believe in the following five religious concepts: God, sin,
hell, heaven, and life after death. Respondents could answer for every concept with 1 ‘yes’
and 0 ‘no’. We took the mean score on these five items.
Based on these items we constructed the orthodox belief dummy variable with value
1 for both those who believe that ‘there is only one true religion (particularism value 1)’
and; those who believe that ‘there is only one true religion, but other religions do contain
some basic truths as well’ (particularism value 2) and score .60 or higher at the doctrinal
beliefs sum-item. The value 0 of the orthodox belief dummy reflects the mainstream (nonorthodox) believers. These are the respondents who have value 2 or 3 of the particularism
scale and believe in at least one of the five religious concepts. All other respondents who
do not belong to either two groups are labeled the non-believers and are excluded from the
analysis.
Anti-immigrant attitudes (Scheepers et al., 2002; Schneider, 2008) is a latent variable
that measures the attitudes towards immigrants and immigration. Respondents indicated
their position about the following bi-polar statements on a scale from 1 to 10: ‘A country’s
cultural life is undermined by immigrants versus is not undermined’; ‘Immigrants make
crime problems worse versus do not make crime problems worse’; ‘Immigrants are a strain
on a country’s welfare system versus are no strain to the welfare system’; ‘In the future the
proportion of immigrants will become a threat to society versus will not become a threat’. A
fifth item asked for the opinion on the number of immigrant in the country: ‘Today [in one’s
country], there are too many immigrants’. For this latter item respondents could answer on
a 1-5 scale, with 1 ‘disagree strongly’ to 5 ‘agree strongly’. Due to the different scaling of the
answer categories, we standardized the items before constructing the latent variable.27
Controls
We control for a number of individual-level socio-demographic characteristics which have
been linked to populist radical right voting. We include gender in our analysis and coded
men as 0 and 1 for female respondents (Givens, 2004). Age is the age of the respondent
when interviewed in 2008. Occupation type is measured by the item that asks for the kind
27
See Appendix 5.A, Table 5.A2 for the cross-national measurement invariance of this latent construct.
117
Chapter 5
of occupation respondents currently have. The reported occupations were summarized on
the ISCO occupational scale (Ganzeboom & Treiman, 1996) and were then recoded in line
with Oesch’s occupational class measure. Oesch’ measure also accounts for the sector in
which people work (Oesch, 2006; Oesch, 2008). Since the ‘manual (production) workers’
are found to be more likely to vote for PRRs (e.g., Lubbers et al., 2002) we added this category as dummy variable into the analysis. Respondents who are currently unemployed, but
reported to be manual production worker in their last job also received ‘1’ on the respective
dummy variable. Education is operationalized by the number of completed years in education, which is based on the international ISCED-measure that asks for the highest level of
completed education. These categories were then recoded into the corresponding number
of years of education completed. Political interest is an interval variable based on the item
‘How interested would you say you are in politics?’ Respondents could answer on a 1-4 scale
with 1 being ‘not interested’ and 4 being ‘very interested in politics’.
5.3.2 Method
To test our hypotheses we apply logistic regression in a SEM-framework for the seven
countries comparing the PRR voters with the Christian voters. For the countries with both
a conservative and a Christian party (Belgium, Switzerland, Finland and Norway), we
include both the conservative and Christian voters (the most attractive parties for religious
people in these countries (Van der Brug et al., 2009)) and compare these to the PRR-voters.
The presented model is inspired by Arzheimer and Carter (2009) and generally produces
good model fit in all seven countries (see Appendix 5.A, Table 5.A4). We add the theoretically relevant religious belief dimension as independent variable and model the covariance
between the religious belief dimension and the religious practice dimension (see Figure
5.1). Since our hypotheses mainly deal with the comparison between Christian parties and
PRRs (and not with the other voters and non-voters), we only present the results of the
comparison of these two party families.
5.4 Results
We first present in Table 5.1a and Table 5.1b the results of the tests of hypotheses 1 and
2. These hypotheses deal with the direct effects of the religious practice dimension and
the interaction effect of anti-immigrant attitudes and religious practice on voting behavior
without having modeled any mediation paths. Then, we present the results of the mediation
model in Table 5.2 and discuss the test of hypothesis 3.
Table 5.1a shows in line with earlier research that religious attendance is negatively associated with voting for a PRR instead of a Christian party in four of the seven countries,
which corroborates Hypothesis 1. Those who more often attend religious services are more
118
Religion and populist radical right voting
Figure 5.1 The (summarized) SEM-model as estimated in MPlus 6.12
Notes:
1. + : Expected positive relationship, - : Expected negative relationship
2. Arrows for the error terms are not shown here.
3.All dependent variables (religious attendance, orthodox vs. mainstream belief, anti-immigrant attitudes and
PRR-voting) were controlled for by the same variables (gender, age, education, occupation type). Additionally,
anti-immigrant attitudes and PRR-voting were controlled for by political interest. For clarity, the arrows of these
control variables were not shown here.
4. For general model fit measures in the countries under study, see Appendix 5.A, Table 5.A4.
likely to vote for a Christian party instead of a PRR than those who sometimes and never
attend a religious service in these countries. Interestingly, in Switzerland, Finland and
Italy, we do not find evidence for this hypothesized effect and we therefore conclude that
Hypothesis 1 is only partly confirmed by our results.
Further, with hypothesis 2, we expected that the ‘encapsulating effect’ of religious practice decreases once people disintegrate from their religious communities and that other
attitudes, i.e. anti-immigrant attitudes, become more decisive in the choice to vote for a
PRR. The results presented in Table 5.1b, where we modeled this interaction effect, show
that this hypothesis is not corroborated for any country. We do not find any evidence that
anti-immigrant attitudes become more important the more people disintegrate from their
churches and we thus find no evidence for Hypothesis 2.
Mediation hypothesis
In this subsection we discuss the results of Hypothesis 3 and Hypothesis 3a that addressed
the indirect effect of religious beliefs on PRR vs. Christian voting. With H3 we hypothesized
that orthodox believers have more negative attitudes towards immigrants than mainstream
119
Chapter 5
Table 5.1a Direct effect religious practice on PRR-voting vs. Christian votinga
AT
BE
CH
FI
IT
NL
NO
Attendance (sometimes=ref.):
Never
.956*
.652
.619
.279
.324
.643*
.183
(.463)
(.486)
(.385)
(.417)
(.634)
(.379)
(.359)
-1.108*
-2.190*
-.069
-.040
.050
-1.690*
-1.078*
(.371)
(1.146)
(.389)
(.557)
(.312)
(.369)
(.389)
3652.48
3216.16
3587.83
2807.20
4738.52
5549.62
3876.61
Often
AIC
BIC
3755.85
3316.85
3689.60
2905.09
4848.78
5664.86
3979.53
Sample-Size Adjusted BIC
3663.91
3224.93
3597.67
2813.19
4756.79
5572.84
3887.59
261
238
247
216
331
393
257
N
Notes: *p<.05 (one-tailed).
Unstandardized regression coefficients (Maximum Likelihood Robust-estimator).
a
Controlled for gender, age, education, and occupation type.
Table 5.1b Direct effect religious practice and interaction effect anti-immigrant attitudes and religious
practice on PRR-voting vs. Christian votinga
AT
BE
CH
FI
IT
NL
NO
Attendance (sometimes=ref.):
Never
1.056*
.926
.739*
.299
.573
.883*
.216
(.567)
(.750)
(.377)
(.453)
(.627)
(.489)
(.351)
-1.001*
-1.692
-.041
.147
.279
-1.500*
-1.561*
(.661)
Often
(.369)
(1.192)
(.424)
(.562)
(.375)
(.517)
Anti-immigrant attitudes
1.312*
1.583*
1.441*
1.035*
1.097*
1.278*
.574*
(.302)
(.808)
(.311)
(.317)
(.379)
(.485)
(.227)
.622
-.505
-.809
-.065
-.650
-.614
-.019
Anti-immigrant*Never
(.798)
(.901)
(.493)
(.471)
(.692)
(.577)
(.376)
Anti-immigrant*Often
-.466
-.758
-.074
-.590
-.569
-.415
1.647
(.449)
(.773)
(.518)
(.585)
(.423)
(.648)
(1.102)
AIC
3653.64
3219.62
3588.93
2810.15
4740.58
5551.87
3875.54
BIC
3764.14
3327.26
3697.72
2914.78
4858.45
5675.06
3985.56
Sample-Size Adjusted BIC
3665.86
3229.00
3599.45
2816.55
4760.12
5576.70
3887.28
261
238
247
216
331
393
257
N
Notes: *p<.05 (one-tailed).
Unstandardized regression coefficients (Maximum Likelihood Robust-estimator).
a
Controlled for gender, age, education, and occupation type.
120
Religion and populist radical right voting
believers and therefore, are more likely to vote for a PRR instead of a Christian party. Only
for the Netherlands we did not expect to find such a relationship, because of the presence
of a religious fundamentalist party (H3a). Table 5.2 shows the decomposition of the total
effect of religious orthodoxy into the direct and indirect effect. Concerning the direct effect
of religious orthodoxy, Table 5.2 shows that only in Norway a positive direct effect of orthodoxy remains when mediation via anti-immigrant attitudes is modeled. Yet, in Austria the
orthodox believers are less likely than the mainstream believers to vote for a PRR instead of
a Christian party, which partly contradicts H3.
Regarding the indirect effect of religious orthodoxy, we do find partly support for our
third hypothesis. In Belgium, Switzerland and Norway, we find that the indirect path of orthodoxy via anti-immigrant attitudes (the ‘via anti-immigrant’-row at the lower end of the
table) is significantly positively associated with the likelihood of voting for a PRR instead
of a Christian or conservative party. This implies that the orthodox believers in these three
countries have more negative attitudes towards immigrants than their mainstream counterparts, which enhances their likelihood to vote for a PRR instead of a Christian party,
and thus supports H3. Additionally, in Switzerland the total positive effect of orthodoxy on
PRR-voting is fully driven by the indirect effect of orthodoxy. Yet, it should be noted that
the indirect effect of orthodoxy is somewhat smaller than the direct effect in Belgium and
Norway.
Table 5.2 Decomposition effect religious practice and belief on PRR vs. Christian Party
Often-attend on PRR vs. Christian party
Total indirect
AT
BEa
CHa
FI
IT
NL
NO
-.196*
-.074*
-.291*
-.058
-.068*
-.112*
-.097*
(Via anti-immigrant attitudes)
(.044)
(.037)
(.104)
(.130)
(.035)
(.041)
(.058)
Direct
-.274*
-.596*
.035
-.155
.015
-.530*
-.730*
(.103)
(.171)
(.163)
(.362)
(.127)
(.162)
(.292)
Total
-.470*
-.670*
-.256*
-.213
-.052
-.642*
-.827*
(.091)
(.161)
(.131)
(.379)
(.127)
(.163)
(.302)
AT
BEa
CHa
FI
IT
NL
NO
Orthodox beliefs on PRR vs. Christian party
.033
.100*
.362*
.052
.026
.073
.103*
(Via anti-immigrant attitudes)
(.045)
(.044)
(.123)
(.128)
(.029)
(.049)
(.057)
Direct
-.261*
.259
-.132
.149
-.008
-.030
.505*
(.094)
(.203)
(.183)
(.369)
(.127)
(.205)
(.297)
Total
-.228*
.359*
.229*
.201
.018
.043
.608*
(.100)
(.193)
(.142)
(.378)
(.130)
(.216)
(.304)
Total indirect
Notes: *p<.05 (one-tailed).
Unstandardized regression coefficients (Weighted Least Square Mean Variance-adjusted estimator).
a
The results are robust if we compare the PRR to the Christian parties alone.
121
Chapter 5
In line with Hypothesis 3a, we find that in the Netherlands there is no (in)direct relationship between religious belief and voting for a PRR instead of a Christian party, which
is due to the finding that Dutch orthodox believers are not more anti-immigrant than
mainstream believers. In the other countries a negative indirect effect of orthodoxy is not
found, because the orthodox believers are not more anti-immigrant than their mainstream
counterparts (see Appendix 5.A, Table 5.A4), or orthodox believers are just more likely to
vote for a Christian party rather than a PRR, which is the case in Austria.
In addition, Table 5.2 demonstrates that next to the direct effect of the religious practice
dimension, we find that the indirect effect of the religious practice dimension also plays a
role: those who more often go to church are generally less supportive of the anti-immigrant
attitudes and therefore less likely to vote for a PRR instead of a Christian party. These
findings coincide with claims and results of (Putnam, 2000) and others, who argue that
religious activism generally fosters tolerance towards others and social trust (e.g., Newton,
2001). Interestingly, this also holds for Switzerland and Italy where the religious practice
dimension did not directly affect the likelihood of voting for a PRR instead of a Christian
party. Only in Finland, none of the dimensions of religion play a role in the choice to vote
for the (True) Finns Party versus the Christian and conservative alternative.
5.5 Conclusion
In this chapter we presented a test of the paradoxical notion that there may be something
in Christian religion that makes and something that unmakes for PRR-voting among native, Western European Christian believers. To clarify this notion, we departed from the
Weberian assumption that religiosity involves more than one’s religious practices, but also
entails a content, belief dimension that may be relevant in explaining one’s voting behavior
(cf. Driskell et al., 2008; Weber, [1930] 2003). We theorized that this belief dimension helps
to solve the puzzle of what it is in religion that makes for PRR-voting, whereas the religious
practice dimension illuminates what aspect of religion unmakes for PRR-voting. Regarding
the religious practice dimension we proposed the (i) norm-compliance mechanism, stating
that those who more strongly adhere to the norm of voting for a Christian party are more
likely to vote for a Christian party instead of a PRR. Regarding the religious belief dimension, we postulated the (ii) threat mechanism, stating that those who feel more religiously
threatened by Muslim immigrants are more likely to vote for a PRR instead of a Christian
party. By analyzing wave 4 of the European Values Study (2010), we were able to test these
underlying mechanisms using mediation-analysis in a SEM-framework. Inspired by social
cleavage theory (Knutsen, 2001; Knutsen, 2004; Lipset & Rokkan, 1967) we studied seven
countries with popular PRRs and a stable Christian or conservative party that has traditionally attracted native Christian people to vote for.
122
Religion and populist radical right voting
When looking at the religious practice dimension, this chapter indeed shows in line with
earlier research (Billiet, 1995; Lubbers & Scheepers, 2000) that the religiously active people
are more likely to vote for a Christian party instead of a PRR in the majority of countries
under study. Nevertheless, this finding cannot really tell us whether the norm-compliance
mechanism, stating that the religiously active are better integrated in a religious community
and more strongly adhere to the norm of voting for a Christian party, drives the relationship. For that reason we have tested it more strictly and assumed that if this mechanism
is at work, strongly integrated Christian people will, independent of their anti-immigrant
attitudes, be more likely to vote for a Christian party instead of a PRR. Whereas for less
integrated Christian people we expected that their anti-immigrant attitudes become more
decisive in their voting choice and the likelihood of voting for a PRR instead of a Christian
party will increase accordingly. This hypothesis has however not been confirmed for any of
the countries.
This may have several reasons that challenge the often postulated norm-compliance
mechanism. At first, our results suggest a confirmation of Putnam (2000) and Newton’s
(2001) mechanism that religious activeness fosters tolerance towards others, which reduces
the development of negative attitudes towards immigrants, and consequently lowers the
probability of voting for a PRR. Although, we note that this mechanism cannot fully account for the negative effect of the religious practice dimension. Furthermore, we speculate
that the similarity in anti-immigrant stances of Christian (and conservative) parties as
compared to their PRR-counterparts, as we showed in Chapter 2, could make the role of
anti-immigrant attitudes in voting for a Christian party similarly important in voting for
a PRR. All in all, the significance of Arendt’s social disintegration theory in understanding
the vote for a PRR may have been overestimated (also see Chapter 1 and Rydgren, 2009;
Rydgren, 2011; Zhirkov, 2014), and we therefore urge further research to more closely
investigate the possible other mechanisms that underlie the relationship between religious
practice and PRR vs. Christian-voting.
When it comes to the religious belief dimension, the findings provide some evidence
that one’s religious conviction affects one’s voting behavior as well. In Belgium, Switzerland
and Norway, our findings show that orthodox believers are generally more anti-immigrant
than mainstream, tolerant believers, which in turn is associated with a higher probability
of voting for a PRR instead of a Christian party. These results partly counter the religious
cleavage idea that religious people tend to vote for a religious party (Van der Brug et al.,
2009). We demonstrate that among the Christian religious people there may be much more
variation in belief systems that differentially influences one’s voting choice. PRRs which are
generally secular parties and have a strict stance on immigrant issues can be credible alternatives for religious people with a distinct, orthodox belief system in these three countries.
123
Chapter 5
Nonetheless, in line with our expectations, this does not hold for the Dutch orthodox
believers. We anticipated that the unique presence of a religious orthodox party, the rereformed party SGP, which advocates a similar anti-immigrant stance as the Dutch PRR,
and combines it with other more traditional, religious fundamentalist stances, will lower
the likelihood of orthodox believers to take a PRR as credible alternative. Moreover, our
results show that the Dutch orthodox believers are not more anti-immigrant than the
mainstream believers.
Yet, the question lingers: why do we not find such differential effects of the religion
dimensions for the other three countries? For the Finnish and Italian cases, we conjecture
that cooperation between the PRR and the Christian party in these countries during the
EVS-field work period may be of influence. In Finland, the (True) Finns Party and the
smaller Christian Democrats were allies in the 2009 European Parliamentary elections, and
in Italy the (conservative) Christian-democratic Pópolo della Libertá [Free People, PdL]
and the PRR Lega Nord [Northern League, LN] were in the same government in 2010. Such
cooperation may explain why orthodox Christian voters feel equally represented by the
two parties and thus do not favor the Christian party over the PRR. Regarding the general
direct negative effect of religious belief in Austria, one could argue that a longer experience
of cooperation between the PRR and the Christian party (in Austria the Freiheitliche Partei
Österreichs [Freedom Party Austria, FPÖ] and the Christian Österreichische Volks Partei
[Austrian People’s Party, ÖVP] formed a coalition government from 1999-2006) inhibits
Christian orthodox voters to vote for the PRR: the experience of long-lasting cooperation
with PRRs that do not favor specific Christian values may drive orthodox voters to vote
for Christian parties again. These are, however, tentative explanations that deserve more
attention in future research.
All in all, this chapter has provided evidence for the double role religion can play in the
decision to vote for either a Christian party or a PRR. On the one hand, religious involvement in general leads to lower levels of PRR-support, but on the other hand, we show that
the orthodox believers in some countries are more likely than mainstream, tolerant believers to vote for a PRR. We showed with sophisticated structural equation models that the differential relationships of the two dimensions of religion with for instance ethnic prejudice
(Scheepers et al., 2002), volunteering (Reitsma, 2007) and political participation (Driskell
et al., 2008) can for certain countries also be generalized to voting behavior. Therefore,
we emphasize that to further our knowledge about how religion drives political behavior,
future research should carefully investigate and theorize about how its different dimensions
may relate to other forms of political behavior.
124
Chapter 6
Gender and populist radical right voting
This chapter was written together with Hilde Coffé and Tanja van der Lippe. A slightly
different version of this chapter will be published as: Immerzeel, T., H. Coffé & T. van
der Lippe, Explaining the Gender Gap in Radical Right Voting: A Cross-National
Investigation in 12 Western European Countries (forthcoming). Comparative European
Politics, DOI: 10.1057/cep.2013.20. Earlier drafts were presented at the 6th European
Consortium of Political Research (ECPR, Reykjavik, Iceland, August 2011) and at the
‘Dag van de Sociologie’ (Ghent, Belgium, May 2011).
Abstract
It is common wisdom in PRR-research that men are overrepresented among the PRR
electorate. In this chapter, we explore whether a PRR gender gap exists across 12 Western
European countries and examine how this gap may be explained. Using the European
Values Study (2010), we find a PRR gender gap that remains substantial after controlling for socioeconomic and political characteristics. However, our results indicate strong
cross-national variation in the size of the gap. Explanations for these differences are
explored by looking at the outsider image and the populist discourse style of the PRRs,
which are hypothesized to keep women from voting for the PRRs. Our results do not
support this expectation: differences in party characteristics do not account for crossnational differences in the gender gap. Implications of these findings and suggestions for
further research are discussed.
125
Gender and populist radical right voting
6.1 Introduction
PRRs have become popular in many Western European countries over the last three decades. In different Western European countries PRRs have been in office or supported a
minority government. Together with their electoral rise, PRRs and their electorate have received substantial scholarly attention (see Rydgren, 2007). Empirical investigations into the
PRRs’ electorate led scholars to conclude that gender is important in explaining voting for
PRRs (Givens, 2004; Mayer, 2002; Mudde, 2007; Norris, 2005), with mainly men voting for
PRRs. Obviously, a gender gap in party preference is not new. Ever since the introduction
of women’s right to vote, gender differences in voting behavior have received scholarly and
social interest because they have major consequences for the electoral outcome of parties
and the composition of the party system (e.g., Inglehart & Norris, 2000).
Despite the fact that gender differences in voting behavior and the rise of PRRs have
been amply studied, research describing and explaining the gender gap in PRR-voting is
relatively scarce (Fontana et al., 2006; Gidengil et al., 2005; Givens, 2004; Rippeyoung, 2007).
Moreover, while focusing on individual level structural and attitudinal characteristics, most
of these studies have not been able to find an explanation for the gender gap (notable exception is Gidengil et al. (2005) who show that political dissatisfaction and social conservatism
explain the gender gap in Canadian Alliance voting). This is surprising because we know
that various other gender gaps in party choice have been explained by gender differences in
structural characteristics and attitudes. For example, in Britain and Australia, gender differences in occupational attainment explained why women tend to vote more conservatively
than men; whereas in the United States, gender differences in political and cultural attitudes
make women more likely to vote for the Democrats than the Republicans (Inglehart &
Norris, 2000; Studlar, McAllister, & Hayes, 1998).
Furthermore, the few studies investigating the gender gap in PRR-voting have a limited
comparative scope, focusing on Canada or a restricted number of European countries.
Consequently, little is known about the presence and magnitude of a gender gap in PRRvoting in a wide range of Western European countries and the differentiation of the gender
gap between the countries.
In this chapter, we aim to improve the current knowledge about the gender gap in
PRR-voting in Western Europe by looking at a larger number of countries than previous
empirical research and by exploring explanations for cross-national differences in the PRR
gender gap. As we are interested in the ‘cleavage-breaking’ impact of PRRs, we focus on
Western European countries where realigning gender differences in voting behavior have
been best observed: Inglehart and Norris (2000) showed that the process of gender realignment has not developed in post-communist Eastern European countries. First, we describe
the gender gap in 12 Western European countries and test various explanations for the
gender gap in these countries. To compare PRRs cross-nationally, we draw upon Mudde’s
127
Chapter 6
(2007) definition about the European PRRs that describes three shared characteristics:
nativism (anti-immigrant and nationalism), authoritarianism and populism. Based on this
definition, we assume that people vote for the party that is closest in issue position to their
own opinions and structural position, and we examine to what extent mainstream socioeconomic characteristics and political attitudes (nativism, authoritarianism and populism)
may explain the PRR-gender gap. Hence, the first question we pose is to what extent is there
a gender gap in Western European populist radical right voting, and, if there is a gender
gap, to what extent can this gap be explained by gender differences in socioeconomic characteristics and PRR-attitudes?
Additionally, we explore how cross-national differences in the gender gap may be explained. Due to the limited number of countries in previous studies, there has been a focus
on individual level explanations, and little attention has been paid to cross-national variation in the gender gap size. Yet, descriptive results of earlier studies show evidence for such
variation over time within and between countries (e.g., Mudde, 2007). Inspired by Kitschelt
and McGann (1995), we account for party characteristics as potential explanations. In particular, we investigate to what extent the populist image of PRRs, as political outsiders and
users of a populist rhetoric, may explain variation in the PRR gender gap. Taking insights
from Roth (1989), we claim that, given women’s lesser political interest, women are more
hesitant to vote for political outsiders than men. Next, in line with Mayer (2002) we argue
that the polarizing, populist style of discourse of PRRs is typically masculine and therefore
keeps women from voting for a PRR. However, as a party loses its outsider image and employs a less populist style of discourse, we anticipate that women will become less restrained
to vote for these parties, which will decrease the gender gap. In sum, our second research
question asks: to what extent can the image of a populist outsider explain cross-national
variation in the gender gap in PRR-voting?
To test our hypotheses, we use the cross-national European Values Study (EVS, 2010)
that contains information on party choice, background characteristics, and political attitudes. We use our Expert Judgment Survey of European Political Parties 2010 (Immerzeel
et al., 2011) to obtain information about the political outsider image and the populist style
of discourse of the PRRs. The combination of these datasets enables us to explore and
explain the gender gap in PRR-voting in 12 Western European countries simultaneously
and test to what extent cross-national differences can be explained by PRRs’ characteristics.
Before turning to our analyses, the next section describes relevant theories and introduces
our hypotheses.
128
Gender and populist radical right voting
6.2 Theory and hypotheses
To answer our first research question, we present an overview of the theoretical arguments
and hypotheses on the individual-level explanations of the gender gap. The hypotheses will
be discussed according to the three characteristics of Mudde’s (2007) definition of PRRs:
nativism, authoritarianism and populism. After having described our hypotheses related to
individual-level explanations, we turn to our second research question and present hypotheses regarding party characteristics that could explain potential cross-national variation in
the gender gap in PRR-voting.
6.2.1 Individual-level explanations for the gender gap
Nativism
In his definition, Mudde (2007) emphasizes the strong sense of nativism that characterizes PRRs: natives should be the sole inhabitants of a nation-state and ‘nonnative elements
(ideas, persons) are fundamentally threatening to the homogeneous nation-state’ (Mudde,
2007: p.19). Nativism is a combination of a preference for the native and xenophobia (a
fear/hatred of everything that is non-native), and various arguments have been formulated
about the reasons why men are more likely to be attracted by PRRs’ nativist stance than
women.
First, it is argued that modernization and globalization of Western societies are fundamentally threatening national economies, particularly blue-collar sectors with high
numbers of manual workers, because multi-national industries are transferring their
manual jobs to cheaper (non-Western) countries. This means that industrial manual workers are more likely to lose their jobs or to be forced into lower-paying jobs in the new
global economy (Givens, 2004). Because men are overrepresented in these occupations,
men experience an insecure economic situation, making them the ‘losers of modernization’
(Betz, 1994). They develop insecure and resentful sentiments that lead to positive attitudes
toward anti-immigrant policies (Studlar et al., 1998).
Second, the ethnic competition theory (Fennema, 2005; Olzak, 1992; Rydgren, 2007;
Scheepers et al., 2002) asserts that immigrants are seen as competitors in the labor market.
Being ‘in competition’ with immigrants over scarce resources triggers exclusionary reactions because workers experience feelings of threat (Fennema, 2005). Such reactions may
result in voting for PRRs, which want to reduce competition from immigrants over scarce
resources by rejecting equal access to resources for immigrants (Olzak, 1992). Because
immigrants mainly compete with people from lower strata and because men are overrepresented in these strata (Givens, 2004), it is argued that men are more likely to have
anti-immigrant attitudes than women. Moreover, women more often work at home or do
not search for a job (Rink, Phalet, & Swyngedouw, 2009). Therefore, they do not feel the
competition over jobs and are less likely to have strong anti-immigrant attitudes. Based
129
Chapter 6
on both theories, we expect that structural characteristics may explain the gender gap in
PRR-voting: H1a. Men are more likely to be manual workers and are therefore more likely to
vote for PRRs than women.
Additionally, we expect that anti-immigrant attitudes may explain the gender gap: H1b.
Men have more negative attitudes towards immigrants than women and are therefore more
likely to vote for PRRs than women.
Authoritarianism
PRRs are also known for their authoritarian appeal and strong emphasis on law and order.
Mudde (2007: p.23) notes that an important part of the PRRs’ ideology is the “belief in a
strictly ordered society in which infringements of authority are to be punished severely”.
Various theories describe why men are expected to be attracted by this authoritarian stance.
Researchers have argued that with modernization, gender egalitarian attitudes have become more mainstream in many Western societies (Betz, 1994), which affected men and
women differently (Inglehart & Norris, 2000; Inglehart & Norris, 2003; Togeby, 1994). In
particular, these modernization processes caused the erosion of traditional, authoritarian
values that made men fear a loss of status (Ignazi, 1992; Ignazi, 2003). ‘The shift in cultural
values (…) has undercut the traditional dominance of the White male’ (Gidengil et al.,
2005: p.1176), which fuelled a ‘silent counter-revolution’ where men nostalgically reacted
to postmodern values by favoring parties that emphasized traditional views about the order
of society (Ignazi, 2003).
Second, socio-psychological theories focus on gender differences in authoritarian attitudes. Gilligan (1982) maintains that men emphasize the strictness of the law and individual
responsibility by solving practical morality issues. Women, by contrast, treat these issues
in terms of conflicting responsibilities that should be solved by the collective in a world
where people care for each other (Gilligan, 1982). Research on attitudes towards military
use and defense spending in the United States confirm these arguments (Kaufmann &
Petrocik, 1999); women are more strongly opposed to the use of force and more supportive
of compassionate policies (Frankovic, 1982; Gilens, 1988). Similarly, Gidengil et al. (2005)
show that men have stricter attitudes towards law and order issues, which helps to close the
gender gap in voting for the Canadian Alliance party.
Based on these theories, we expect that authoritarian attitudes may explain the gender
gap in PRR-voting: H2. Men have more authoritarian attitudes than women and are therefore
more likely to vote for PRRs than women.
Populism
According to Mudde (2007), the third characteristic that PRRs share is populism. Populism
is defined as an ideology that states that the society is divided into two groups: the ‘pure
130
Gender and populist radical right voting
people’ that have a common will which is embodied by the PRRs; and the distrustful ‘corrupt elite’ that are formed by the other established political parties. Because PRRs reference
established parties as corrupt, they are often seen as anti-establishment parties that are subject to a cordon sanitaire (Bustikova, 2009; Kitschelt, 2007; Mayer, 2002; Mudde, 2007). So
far, the anti-establishment image of PRRs has only been implicitly related to the gender gap
in PRR-voting. However, we argue that it may be relevant in explaining the gender gap in
PRR-voting. Due to socialization and life experiences, women are less politically interested
and active than men (Burns, 2007; Coffé & Bolzendahl, 2010; Dalton, 2008; Lovenduski,
1986; Mudde, 2007; Verba, Burns, & Schlozman, 1997). Similar to Mudde (2007), we argue
that less politically interested and active people are likely to become aware of new political
developments later in time than politically interested and active citizens. This makes citizens who are less politically interested and active less likely to vote for new ‘extreme’ parties,
and more likely to follow the widespread norm to vote for an established party. Given that
women are generally less politically interested and active, they are expected to be more
likely to vote for established parties rather than political outsiders with an ‘extremist’ image such as PRRs (Kitschelt & McGann, 1995; Mayer, 2002; Mudde, 2007; Norris, 2005).28
Hence, we hypothesize: H3. Women are less politically interested and active than men and are
therefore less likely to vote for PRRs than men.
6.2.2 Party characteristic explanations for cross-national variation in gender gap
There is ample agreement among scholars that nativism, authoritarianism and populism
are defining features of PRRs. However, there is variation in the strength of these party
characteristics among various PRRs (e.g., Arter, 2010). This is a crucial assumption to
answer our second research question addressing cross-national variation in the gender gap.
We maintain that PRRs are generally anti-establishment outsiders with a populist discourse
style (Mudde, 2007), but argue that the extent to which a PRR has a political outsider image and uses a populist discourse style varies across countries, which influences men and
women’s likelihood to vote for a PRR. Consequently, we assume that these characteristics
may explain cross-national variation in the gender gap in PRR- voting.
Regarding the political outsider image, Roth (1989) maintained and showed in the case
of the German Green-party that women are more hesitant to new political developments,
whereas men are aware earlier of new political developments and therefore more likely
to vote for a relatively new political party. Logically, it follows that when a PRR loses its
28 One could argue that as ‘modernization losers’, men are more dissatisfied with the established parties and
are therefore more likely to vote for an anti-establishment party (cf. Lubbers & Scheepers, 2000; Lubbers
et al., 2002; Van der Brug, 2003). However, amongst others Coffé and Bolzendahl (2010) have shown
that women are less satisfied with political institutions than men. Hence, political dissatisfaction cannot
explain the gender gap in PRR-voting. We performed additional analyses including a measure of political
dissatisfaction. The results of these analyses were similar to those presented here. Hence, we excluded this
variable from our analyses.
131
Chapter 6
‘newcomer’ outsider image and obtains a stable position in the political spectrum, women
will be as likely as men to vote for a PRR. Hence, we expect the following: H4a. The weaker
the image of a PRR as a political outsider, the smaller the gender gap in PRR-voting will be.
Accordingly, we expect that in countries where PRRs are more electorally successful, the
general perceptions about voting for the PRR change. As more people vote for the party,
it becomes more ‘normal’ to do so. Moreover, once a PRR becomes larger, questions arise
about the extent to which a PRR is actually radical right. In the Netherlands, for example, a
research project that concluded that PVV was radical right (Moors, Balogh, Van Donselaar,
& De Graaff, 2009) led to a public debate in which many people doubted the radical content
of the party. Furthermore, PRRs have played a prominent role in the functioning of a minority government and have cooperated as a tolerating partner with other political parties
in some countries. In doing so, PRRs show that they become part of the establishment,
which may reduce their outsider image. In summary, we argue that the political outsider
image weakens in countries where the PRR has become electorally successful: H4b. The
more electorally successful a PRR, the less it will be seen as a political outsider and the smaller
the gender gap in PRR-voting will be.
Various scholars argued that PRRs employ a populist, simplistic ‘us versus them’ discourse in which the antagonistic groups (the elite, the immigrants) are seen as unreliable
deviant free-riders (Papadopoulous, 2002). This polarizing tone belongs to the PRRs’
populist ideology, but can also be seen as a separate ingredient of the message. Naidoo
and Lord (2008) have convincingly shown that the utterance of a message is important in
getting a public committed to a speaker’s ideological content. We argue that the populist,
i.e. polarized and simplifying, style of discourse is typically masculine (Tannen, 1990) and
strongly opposes the more compromising feminine discourse styles (Huddy & Terkildsen,
1993). Following Mayer (2002) who argued that mainly men are attracted by masculine
verbal violence, we hypothesize that: H4c. The less populist the discourse style of a PRR is, the
smaller the gender gap in PRR-voting will be.
6.3 Data and Methods
To test our hypotheses, we used the 2008 round of the European Values Study (EVS, 2010).
In total, 47 European countries participated in this standardized, cross-sectional, national
representative survey. In every country, respondents were selected by means of stratified
random samples of the resident populations aged 18 years and older. The target sample size
of 1,500 respondents per country was generally met, but the response rates differed considerably among countries, ranging from roughly 35% to 85%.29 Because not all countries have
a PRR, and because we are interested in Western European countries, we focused on the
29 For more information about the EVS: http://www.europeanvaluesstudy.eu/
132
Gender and populist radical right voting
12 Western European countries that have at least 20 PRR-voters in the sample. The original
number of respondents in the 12 counties was 17,777. We excluded 2,250 respondents who
initially stated that they would not go to the polls (948 males and 1,302 females). Then, we
excluded 3,751 voters who were missing items about which party they would vote for (1,635
males and 2,116 females) and performed listwise deletion, resulting in 10,643 respondents.
Moreover, to assess the party characteristics, we draw on the Expert Judgment Survey of
European Political Parties 2010 (Immerzeel et al., 2011), which we described in Chapter 1.
6.3.1 Operationalization
Dependent variable
PRR-voting was measured by the question, ‘Would you vote if there were general elections
tomorrow?’30 If assenting, ‘What party would you vote for if there were general elections
tomorrow?’ Based on the classification of Mudde (2007), we coded the PRR votes as ‘1’ and
other parties, invalid or blank votes as ‘0’. In line with earlier PRR-research (e.g., Lubbers
et al., 2002), we performed analyses that compared PRR-voters with other party voters and
excluded the non-voters. In addition, we excluded the respondents who did not specify a
party they would vote for but who would vote.
Independent variables
Individual level
Our main explanatory variable, gender, is measured as a dichotomous variable with men
coded as 0 and women as 1.
The main structural explanatory characteristics introduced in our models are employment status and occupation. Employment status measures the respondents’ current main
daily activity and is represented by three dummies: employed, self-employed, inactive/
unemployed.
Occupation measures the kind of occupation respondents currently have. The reported
occupations were summarized on the ISCO occupational scale (Ganzeboom & Treiman,
1996) and were then rearranged according to Oesch’ occupational class measure, which
accounts for the sector in which people work (Oesch, 2006; Oesch, 2008). The scale
distinguishes the following categories: ‘manual workers’, ‘service workers’, ‘office clerks’,
‘socio-cultural professionals’, ‘technical professionals’, and ‘managers and administrators.’
Oesch’ scale distinguished between self-employed and employee. Because the two selfemployed categories are very small in our sample, we decided to place all self-employed in
one separate dummy (see above). For those respondents who are currently unemployed, the
30 ‘Populist Radical Right voting’ draws on the expected vote rather than the actual vote. In the 2008 wave of
the European Social Survey (ESS, 2008) respondents did indicate their actual vote. The number of PRRvoters is highly similar in both datasets.
133
Chapter 6
occupational type of the last job is reported. For those respondents who never had a paid
job, we added a separate dummy ‘never had a paid job’.31
In addition to socioeconomic, structural characteristics, we enter a variety of political
attitudes and characteristics and explore to what extent these can account for the PRRgender gap. Anti-immigrant attitudes (Scheepers et al., 2002; Schneider, 2008) is an interval
mean scale (Cronbach’s alpha=.87) that measures the attitudes towards immigrants and
immigration. Respondents gave their opinion about the following five statements: ‘Immigrants take jobs away from natives’; ‘A country’s cultural life is undermined by immigrants’;
Immigrants make crime problems worse’; ‘Immigrants are a strain on a country’s welfare
system’; and ‘In the future, the proportion of immigrants will become a threat to society’.
For all items, respondents could answer on a 1-10 scale, with 1 ‘disagree strongly’ to 10
‘agree strongly’.32
Authoritarian attitudes were measured by two items in the survey. First, we used the
item that asks what the respondent thinks is the most important issue for the country in the
next 10 years (Abramson & Inglehart, 1995). We made a dummy with 1 for those who chose
‘maintaining order in the nation’ and 0 for those who chose one of the other three aims:
‘giving the people more say in important political decisions’; ‘fighting rising prices’; and
‘protecting freedom of speech.’ Second, we used the item that asked respondents whether it
is either good or bad to have a strong leader who governs the country and does not have to
bother with parliament and elections. Respondents could answer on a scale ranging from 1
‘very bad’ to 4 ‘very good’. Due to the non-normal distribution of the variable, we recoded
these categories into a dummy variable with 0 ‘bad’ and 1 ‘good’.
Political interest is an interval variable based on the item ‘How interested would you say
you are in politics?’ Respondents could answer on a 1-4 scale with 1 being ‘not interested’
and 4 being ‘very interested in politics’.
Political action is an interval mean scale (Cronbach’s alpha=.74) based on 5 items that ask
for whether a person has ever participated in a certain political action, and if not, whether
they would do so. The political actions were signing a petition, joining a boycott, joining an
unofficial strike, attending a lawful demonstration, and occupying buildings or factories.
Respondents could answer with 1 ‘have done’, 2 ‘would do that’, and 3 ‘would never do’ for
all of these 5 items. We recoded the scale such that it runs from 1 ‘low intention (would
never do any action)’ to 3 ‘high intention (have done all actions)’.
Finally, we control for a number of socio-demographic characteristics that have been
linked to PRR-voting and that may influence the gender gap. Age is the age of the respon31 We are aware that Oesch’ occupational scale does not perfectly represent a sector scale. However, because
the argument is mainly about industries where people perform manual work, we see the operationalization
of Oesch as most appropriate, given the information available in the EVS survey.
32 Nativism also addresses a preference for the native. Due to the lack of good measures for nationalistic
attitudes, we performed additional analyses with strict attitudes towards citizenship. Including this scale did
not, however, change our main conclusions, and we therefore excluded this variable from our analyses.
134
Gender and populist radical right voting
dent when interviewed in 2008. Education is operationalized by the international ISCEDmeasure that asks for the highest level of completed education, which we recoded into
three dummies: ‘primary/lower secondary’, ‘higher secondary’ and ‘tertiary’ education. For
marital status, we created the following dummies: ‘having a partner’, ‘never had a partner’,
and ‘had a partner’ (divorced, separated, and widowed). The degree of urbanization is
measured by four dummies that measure how many inhabitants live in the respondent’s
place of residence: ‘<5,000’; ‘5,001-20,000’; ‘20,001-100,000’; and >100,000 inhabitants.
Lastly, we control for religiosity by using two measures: religious denomination and religious
attendance.33 Religious denomination is operationalized by three dummies: non-religious,
Christian and other religion. Religious attendance is operationalized by the item ‘Apart
from weddings, funerals and Christenings, about how often do you attend religious services
these days?’ Respondents could answer on a scale from 1 ‘never’ to 7 ‘weekly’. Due to the
U-shaped distribution of this variable, we decided to recode this variable into three dummies: ‘never’, ‘sometimes’ (including the categories ‘less than once a year’-‘only on specific
holidays’), and ‘often’ (‘once a month’ or more).
Party characteristics
Three party level characteristics are considered: the political outsider image, the electoral
success and the political discourse style of PRRs. The political outsider image of the PRR is
a measure based on the Expert Judgment Survey for European Political Parties 2010 (Immerzeel et al., 2011). The experts were asked: ‘Consider the following political parties in your
country, and place them on the scale that concerns the extent to which the party is considered
as anti-establishment (0=not anti-establishment, 10=very anti-establishment)’. We calculated
the mean score of all experts on this issue for all political parties in their country. Then,
we used the mean score of the PRR as the absolute measure of political outsider image. In
countries with multiple PRRs, we took the average of their scores on this item. To deal with
the notion that the outsider image also depends on the outsider scores of the other parties in
a country, we constructed a relative measure of political outsider image. We took the average
difference between the outsider score of the PRR and the outsider score of all other parties that received at least 5% of the votes in the latest elections and which are considered as
establishment parties, i.e. the mean score on the outsider scale was lower than 5.
Electoral success of the PRR is operationalized in two ways. The absolute success is based
on the percentage of PRR-voters in the latest elections before 2008 (when the survey data
were collected). Data were retrieved from official electoral institutes in the countries under
study. In countries with multiple PRRs, we took the sum of their percentages. The relative
success is measured by how successful the PRR is as compared to the most electorally suc33 Religion has been introduced as explanatory factor for gender differences in voting behavior (e.g. Studlar
et al., 1998). However, Rippeyoung (2007) showed that religiosity cannot explain the gender gap in PRRvoting, which made us decide to include it as a control variable.
135
Chapter 6
cessful party in a country. We took the ratio of the two percentages. To illustrate, if the most
successful party in a country received 40% of the votes and the PRR received 15% of the
votes, then this variable scored: 15/40=0.375. The inclusion of such relative measurement is
important because in countries with a (quasi) two-party system, 15% is a small percentage,
whereas in a multi-party system, a party attaining 15% of the votes may be considered as a
relatively large political player.
Populist discourse style is measured by an item of the Expert Survey (Immerzeel et al.,
2011): ‘Consider the following political parties in your country, and evaluate their style of
discourse (0 = non-populist, 10 = very populist)’. We calculated the mean score of all experts
on this issue for all political parties in their country. Then, we used the mean score of the
PRR as the absolute measure of populist discourse style. In countries with multiple PRRs,
we took the average of their scores on this item. To deal with the notion that the populist
discourse style may also be present among other parties in different countries and PRRs are
thus not necessarily unique in their populism, we constructed a relative measure of populist
discourse style. We took the average difference between the populist score of the PRR and
the populist score of all other parties that received at least 5% of the votes in the latest
elections. Table 6.1 presents an overview with the descriptive statistics of our dependent,
independent and control variables.
6.3.3 Analytical strategy
To test the individual-level hypotheses, we utilized logistic regression models with robust
standard errors that correct for clustering at the country level. Because our first goal is to
explain the gender gap in PRR-voting, we are interested in the gender effect on the probability of PRR-voting. Therefore, in the first model, we only present the general gender effect
on PRR-voting with control variables to investigate whether there is indeed a significant
negative effect of being female. In the second model, we add our hypothesized structural
characteristics to investigate to what extent the gender gap is due to gender differences in
socioeconomic characteristics. Our political attitudes are added in a third model. If the
inclusion of these variables leads to a decrease of the gender effect on PRR-voting between
the models, we are able to (partly) explain the gender gap in PRR-voting.
To assess whether there is cross-national variation in the size of the gender gap, we also
present separate analyses for each country. Ideally, one would perform multilevel modeling
where a random slope of the gender effect on PRR-voting is added and slope variance is
explained by cross-level interactions of the party variables with gender (Hox, 2002; RabeHesketh & Skrondal, 2008). However, the number of countries included in this chapter
(N2=12) is too limited to perform a robust multilevel analysis (Hox, 2002). Therefore, we
apply an exploratory approach and correlate the country-specific gender coefficients with
the party variables retrieved from the Expert Survey.
136
Gender and populist radical right voting
Table 6.1 Descriptive statistics dependent, independent and control variablesa
Mean
women
Mean
S.D.
PRR-voting
.09
-
0/1
.07
Female
.50
-
0/1
-
Employment status
Range
Mean
men
Sig.b
.11 ***
-
Employed
.50
-
0/1
.48
.52 ***
Self-employed
.07
-
0/1
.04
.09 ***
Inactive/unemployed
.43
-
0/1
..47
.39 ***
Occupation
Never had a paid job
.07
-
0/1
.10
.04 ***
Production worker
.19
-
0/1
.09
.29 ***
Service worker
.17
-
0/1
.24
.10 ***
Office clerks
.12
-
0/1
.15
.07 ***
Socio-cultural professionals
.16
-
0/1
.22
.10 ***
Technical professionals
.08
-
0/1
.03
.12 ***
Managers and administrators
.22
-
0/1
.17
.27 ***
5.86
2.16
1-10
5.78
5.93 ***
Anti-immigrant attitudes
Strict att. towards law and order
Having strong leader
.24
-
0/1
.23
.24
Maintaining order important
.29
-
0/1
.29
.30
Political interest
2.69
.90
1-4
2.56
Political action
1.77
.48
1-3
1.74
49.11
17.45
17-99
48.97
Age
Education (ISCED)
2.83 ***
1.80 ***
49.33
Primary or lower
.25
-
0/1
.27
.23 ***
Secondary, non-tertiary
.43
-
0/1
.41
.45 ***
Tertiary
.32
-
0/1
.32
.32
Marital Status
Partner
.57
-
0/1
.54
.59 ***
Ever had partner
.18
-
0/1
.23
.13 ***
Never had partner
.26
-
0/1
.23
.28 ***
Urbanization
<5,000 inhabitants
.25
-
0/1
.24
.25 *
5-20,000 inhabitants
.26
-
0/1
.26
.26
20-100,000 inhabitants
.28
-
0/1
.27
.28
>100,000 inhabitants
.22
-
0/1
.23
.21 *
Religious denomination
None
.29
-
0/1
.26
.31 ***
Christian
.69
-
0/1
.71
.66 ***
Other
.02
-
0/1
.02
.02
137
Chapter 6
Table 6.1 Descriptive statistics dependent, independent and control variables(continued)a
Mean
S.D.
Mean
women
Range
Mean
men
Sig.b
Religious attendance
Never
.38
-
0/1
.35
.41 ***
Sometimes
.40
-
0/1
.40
.41
Often
.22
-
0/1
N1
10,643
.26
5309
.18 ***
5334
Political outsider image
Absolute image score
7.01
1.57
3.8-9.0
Relative image score
4.39
1.53
2.2-6.4
Electoral success of PRR
Absolute success
10.96
8.32
2.2-19.4
Relative success
0.42
0.41
.06-1.51
Populist style of discourse
Absolute populist measure
9.04
0.35
8.5-9.5
Relative populist measure
3.90
.93
2.5-5.2
N2
12
Sources: European Values Study (2010); Expert Judgment Survey of European Political Parties 2010 (Immerzeel et
al., 2011).
Notes: ***p<.001 **p<.01 *p<.05 (two-tailed tests).
a
These numbers reflect the average mean differences after cluster correction for countries.
b
These tests are based on bivariate regression analyses with robust standard errors corrected for country-clusters and
indicate to what extent gender differences are significant.
6.4. Results
6.4.1 Descriptive analyses
Before turning to our multivariate analyses, Table 6.2 presents an overview of the gender gap
in PRR-voting in all countries under study. We find that, in general, in comparison to women,
men are more likely to vote for PRRs compared with any other party. These descriptive results
also show considerable variation in the gender gap. The gender gap is largest in Norway,
where the electoral successful Progress Party receives almost twice as many male than female
votes. Additionally, in Austria, Switzerland, and the Netherlands, we find large gender gaps
with men being much more likely to vote for PRRs than women. One exception is France,
where we find the same percentage of men and women voting for PRRs.34
34 Mayer (2002) does find a significant gender gap in PRR-voting in France. However, in the ESS (2008), we find
nearly the same percentage of men (2.7%) and women (2.3%) having voted for PRRs. In earlier waves of the
ESS we do find gender gaps in France. Hence, the year 2008 seems exceptional for France in that there is no
significant gender gap in PRR-voting.
138
Gender and populist radical right voting
Table 6.2 Gender gap in PRR-voting in 12 Western European countries
Definition: % of (wo)men voting for PRR (excluding non-voters)
Country
EVS 2008
Party
Austria
FPÖ, BZÖ
N
N (total
PRR-voter)
% Men
% Women
% Gap
752
131
22.3
13.3
9.0
Belgium
FN, VB
1,292
67
6.8
3.7
3.1
Denmark
Danske FP
1,076
97
10.4
7.5
2.9
Finland
True Finns
577
68
13.2
10.2
3.0
France
FN, MNR
1,147
26
2.3
2.3
0.0
Germany
Rep., NPD/NVU
1,032
20
2.6
1.3
1.3
Greece
LAOS
861
30
5.4
1.8
3.6
Italy
Lega Nord, F. Tricolore
667
67
11.4
8.5
3.1
Luxembourg
ADR (Comité d’Action)
Netherlands
PVV, ToN
Norway
Progress Party
Switzerland
SVP, Swiss Democrats
Total
712
22
3.2
3.0
0.2
1,123
116
13.0
8.0
5.0
860
187
27.9
14.6
13.3
544
125
25.9
19.5
6.4
10,643
956
11.1
6.8
4.3
Source: European Values Study (2010).
6.4.2 Multivariate analyses
We now turn to our multivariate analyses. Table 6.3 presents the results of the logistic
regression model with robust standard errors and country-clusters to test hypotheses 1-3.
These analyses include all countries simultaneously.
In line with our expectations and previous research on PRR-voting, our findings (Model
1) reveal a negative effect of being female on PRR-voting: women have 39% (1-e-.497) lower
odds of voting for a PRR than men.35 Model 1 further shows in line with earlier research
(e.g., Lubbers et al., 2002) that being younger, having lower education, and attending religious meetings less often are strong indicators of PRR-voting. In Model 2, we add our
hypothesized structural explanatory variables. We find that the negative effect remains but
decreases compared to Model 1. This indicates that our structural characteristics do partly
mediate the gender effect on PRR-voting. Men are more likely to be self-employed and
are overrepresented in the manual worker category–see Table 6.1–which seems to partly
explain the gender gap in PRR-voting. Thus, we find some evidence for Hypothesis 1a.
Our analysis also reveals that socio-cultural professionals, who are mainly women, are
substantially less likely to vote for a PRR compared to manual workers.
Model 3 indicates that the gender gap remains, even when political attitudes are controlled for. When adding anti-immigrant attitudes (H1b), authoritarian attitudes (H2),
35 To obtain the standardised effect, we use the formula e(b-coefficient*S.D.) for a positive b-coefficient and the
formula 1-e (b-coefficient*S.D.) for a negative b-coefficient.
139
Chapter 6
Table 6.3 Logistic regression analyses for PRR-voting in 12 Western European countries (robust standard errors, dummies for countries not shown)
Model 1
Female
Model 2
Model 3
b s.e.
b. s.e.
b s.e.
-.497 ***
.066
-.423 ***
.079
-.395 ***
.094
Employment status
Employed (ref.)
-
Self-employed
.421 *
.169
.391 *
.195
Inactive/Unemployed
.162 **
.062
.099 *
.056
-.801 ***
.140
-.634 ***
.188
Occupation
Never had a paid job
Manual (production) workers (ref.)
-
-
Service workers
-.004
.160
.068
.187
Office clerks
-.175 *
.093
.065
.132
Socio-cultural professionals
-.821 ***
.165
Technical professional
-.234
.176
.013
.170
Managers and administrators
-.248
.172
-.026
.175
Anti-immigrant attitudes
-.440 **
.157
.505 ***
.038
Strict att. towards law and order
Having strong leader
.238 *
.143
Maintaining order important
.020
.076
Political interest
-.039
.069
Political action
-.154
.177
.004
-.013 ***
.003
Control variables
Age
-.007
.004
-.009 *
Education
Primary or lower (ref.)
-
Secondary, non-tertiary
-.259 **
.098
-.200 *
.094
-.033
.104
Tertiary
-1.183 ***
.136
-.880 ***
.153
-.416 ***
.115
Marital Status
Partner (ref.)
Ever had partner
-.005
.104
-.004
.105
-.048
.100
Never had partner
-.115
.106
-.089
.080
-.043
.078
.036
.134
.061
.131
.045
.127
Urbanization
<5,000 inhabitants (ref.)
5,001-20,000 inhabitants
-
-
-
20,001-100,000 inhabitants
-.118
.148
-.079
.144
-.045
.159
>100,000 inhabitants
-.343 *
.162
-.293 *
.157
-.108
.156
.099
.092
.095
.088
-.047
.089
-.350
.211
-.447 *
.198
-.191
.283
Religious denomination
None (ref.)
Christian
Other
-
-
-
Religious attendance
Never
Sometimes
Often
.381 ***
.083
-
.363 ***
.081
-
.328 *
.106
-
-.350 *
.162
-.335 *
.159
-.288 *
.151
Constant
-.650 *
.258
-.556 **
.196
-4.063 ***
.597
Pseudo R2
.13
.14
.24
10,643
10, 643
10, 643
N1
Source: European Values Study (2010); Note: ***p<.001 **p<.01 *p<.05 (one-tailed tests).
140
Gender and populist radical right voting
political interest and activeness (H3), we see that the gender effect only slightly decreases:
the odds of PRR-voting are 33% (1-e-.395) lower for women. Hence, even though men are in
general more anti-immigrant, more politically interested and more active than women (see
Table 6.1), these characteristics cannot explain the gap. Consequently, we must conclude
that we find little support for our hypotheses on political attitudes.
Although attitudes only slightly decrease the gender gap, Model 3 shows that attitudes
do strongly affect PRR-voting. Given Mudde’s (2007) definition of the PRRs as being nativist, it is not surprising that anti-immigrant attitude is the best predictor of PRR-voting (cf.
Ivarsflaten, 2008): a one standard deviation increase in the anti-immigrant scale is related
to 198% (e.505*2.16) higher odds of voting for a PRR. In addition, authoritarian attitudes play
a role in explaining PRR-voting; stating that it is good thing to have a strong leader who
governs the country and does not have to bother with parliament and elections is related to
27% (e.238) higher odds of voting for a PRR. In contrast, we do not find that rating maintaining order as the most important aim of the country’s government, being politically active or
having political interest are positively related with PRR-voting.
6.4.3 Country-specific analyses
To explore whether there is variation in the size of the gender gap, we now move on to
the country-specific analyses. Table 6.4 presents three models: bivariate, structural and
attitudinal models. Three important patterns emerge in Table 6.4. First, for those countries
with a gender gap, we find that structural characteristics generally explain at least part of
the gender gap in PRR-voting, which corroborates Hypothesis 1a. In particular, we find that
the gender gap ceases to be significant after inclusion of the structural items in Denmark
and Switzerland.36 In both countries, this is mainly due to the overrepresentation of men
in the manual occupations and the overrepresentation of women in the socio-cultural
professions (see Güveli, Need, & De Graaf, 2007). Moreover, in Denmark, women are overrepresented in the highest level of education. This level of education is negatively associated
with PRR-voting. In Belgium, Greece, the Netherlands and Norway, we find that the gender
gap declines after inclusion of the structural items, which partly corroborates H1a; men are
more likely to be employed as manual workers, which partly explains why they are more
likely to vote for a PRR. The only exception is Austria where we do not find evidence for
H1a. Despite the overrepresentation of men in the manual occupations and in the selfemployed group, the gender gap remains the same after inclusion of these characteristics.
Second, we find limited evidence for our hypotheses on political attitudes (H1b: anti-immigrant, H2: authoritarianism, and H3: political interest and activism) in most countries.
Only in Norway does the gender gap decrease by including the attitudinal items. The main
reason is that Norwegian men are much more anti-immigrant and authoritarian than Nor36
The country-specific descriptive results are not shown here but are available upon request.
141
Chapter 6
wegian women, which aids in explaining the gender gap. By contrast, in Austria, Belgium
and Greece, the gender gap increases after inclusion of the political attitudes. This can be
explained by the higher levels of political interest and activism among men. In contrast to
what we had expected, these higher levels of political interest and activism are associated
with a smaller likelihood of PRR-voting in these three countries.
Third, and most interestingly, the bivariate models in Table 6.4 show that in only 7 of the
12 countries a significant gender gap exists. We find no significant gender differences in the
probability of PRR-voting in Finland, France, Italy, or Luxembourg. In the case of France,
Germany, and Luxembourg, we are cautious in interpreting the results due to the small
number of PRR-voters (N=20-30) in the sample. The pattern remains after controlling for
all individual-level variables (see Models 2 and 3).
Altogether, these results show a mixed pattern of the gender gap in PRR-voting and
indicate more cross-national differences in the extent of the gender gap than has been presumed in previous research. In some countries, there is no gender gap; in other countries,
the gender gap decreases after inclusion of structural characteristics; in some countries,
the gender gap is persistent and does not change after inclusion of explanatory variables;
and finally, in some countries, the gender gap increases. This suggests that gender differences in PRR-voting may be mainly due to differences in party characteristics rather than
Table 6.4 Logistic regression results for women’s PRR-voting (robust standard errors in parentheses) in
12 Western European countries, separate models
Country
2. Structurala
1. Bivariate
3. Attitudesb
N
Austria
-.62 ***
(.20)
-.63 ***
(.22)
(.26)
752
Belgium
-.65 **
(.26)
-.58 *
(.31)
-1.00 **
(.35)
1,292
Denmark
-.37 *
(.22)
.19
(.29)
.22
(.31)
1,076
Finland
-.30
(.26)
-.47
(.33)
-.35
(.36)
577
France
-.00
(.40)
-.24
(.45)
-.31
(.61)
1,147
-.68
(.47)
-.51
(.49)
.52
(.78)
1,032
(.42)
-.91 *
(.43)
-.95 *
(.44)
861
667
Germany
Greece
-1.15 **
-.82 ***
Italy
-.32
(.26)
-.07
(.30)
-.22
(.35)
Luxembourg
-.08
(.44)
-.24
(.58)
-.04
(.55)
712
Netherlands
-.55 **
(.20)
-.46 *
(.23)
-.45 *
(.25)
1,123
Norway
-.82 ***
(.18)
-.57 **
(.20)
-.34
(.21)
860
Switzerland
-.37 *
(.21)
-.36
(.24)
-.38
(.27)
544
Source: European Values Study (2010).
Notes: ***p<.001, **p<.01, *p<.05 (one-tailed tests).
a
Model controlling for age, education, employment status, occupation, marital status, religious attendance, religious
denomination, and urban residence.
b
Structural model plus anti-immigrant and authoritarian attitudes, political interest, and political activeness.
142
Gender and populist radical right voting
individual-level explanations. Hence, considerable variation in the size of the gender gap
requires contextual explanation.
6.4.4 Explaining cross-national variation in the gender gap
In this section, we explore to what extent the cross-national variation may be explained by
the party’s image as a political outsider and its populist discourse (both measured by expert
scores) and by the party’s electoral success. With Hypotheses 4a, we expected that the less a
PRR is perceived as an anti-establishment party, the smaller the gender gap in PRR-voting.
Table 6.5, however, shows no significant effect of the absolute and relative image score on
either the bivariate gender coefficient or the controlled gender coefficient
With Hypothesis 4b we anticipated that the more electoral successful the PRR was in earlier
elections, the smaller the gender gap in PRR-voting would be. Because the effect of female is
generally negative, we expected to find a significant positive r-coefficient. Table 6.5, however,
shows that there is no significant relationship the effect of being female in the country-specific
logistic regression analyses and the PRR’s absolute and relative electoral success. Lastly, we
hypothesized that the less populist the style of discourse a PRR employs, the smaller the gender
gap in PRR-voting (H4c). Table 6.5 does not display any significant effect and we thus cannot
corroborate this hypothesis. In sum, we do not find evidence for Hypotheses 4a, and 4b and 4c.
6.4.5 Summary of tests of hypotheses
Summarizing, concerning the individual-level hypotheses, we find evidence for H1a that
gender differences in employment status, occupational type and education partly explain
the gender gap in PRR-voting in most countries. Gender differences in anti-immigrant,
authoritarian and political attitudes do generally not aid to explain the gender gap in PRRvoting, thus contradicting Hypotheses 1b, 2 and 3. Regarding the party-level indicators, our
Table 6.5 Bivariate correlations b-coefficients (derived from Table 6.4: Model 1. Bivariate; and Model 3.
Attitudes) and party characteristic variables
Gender gap in PRR-voting
Party characteristics
Bivariate model
Multivariate (attitudinal model)
Absolute image score
.486
.324
Relative image score
.406
.379
Absolute success
-.022
-.213
Relative success
.014
-.225
Absolute populist score
.144
-.375
Relative populist score
.397
-.043
Sources: European Values Study (2010); Expert Judgment Survey of European Political Parties 2010 (Immerzeel et
al., 2011).
Note: *p< .05 (one tailed tests).
143
Chapter 6
results suggest that variation in anti-establishment image, populist rhetoric and electoral
success of the PRR cannot aid to explain the cross-national variation in the gender gap in
PRR-voting, and thus contradict Hypotheses 4a, 4b and 4c.
6.5 Conclusion
This chapter presented tests of various explanations for the gender gap in PRR-voting in
12 Western European countries. First, we examined to what extent there is a gender gap in
these countries and to what extent a wide range of individual-level factors, including both
structural and political attitudinal characteristics, can explain the gender gap in PRR-voting.
These individual-level explanations were based on the definition of PRRs by Mudde (2007),
who says that PRRs are generally nativist, authoritarian and populist. We hypothesized
that people vote for PRRs because their own attitudes, opinions and structural positions
coincide with the parties’ opinions and characteristics. Second, we developed and tested
new hypotheses to explain cross-national differences in the gender gap. In particular, we
explored to what extent cross-national variation can be explained by party characteristics,
in particular outsider image and populist discourse style.
Our analyses, based on the European Values Study (2010), reveal that men are on average significantly more likely to vote for PRRs, which confirms previous studies. However,
our analyses provide more detail, indicating that although gender differences in employment status, occupational types and education explain some part of the gender gap in
PRR-voting, the gap remains substantial once these characteristics and different attitudes
are controlled for. Furthermore, our separate country analyses indicate that considerable
cross-national variation exists in the size of the gap. Indeed, whereas we do find an overall
gender gap, there is no such gender gap in some countries. Hence, a first important finding
of this chapter challenges the common wisdom of a PRR gender gap and demonstrates a
more diverse picture. Additionally, in countries with a gender gap, explanations for the
gender gap differ. Whereas in some countries the gender gap can be (partly) explained by
structural characteristics, it remains substantial in other countries. In general, gender differences in political attitudes have limited explanatory power. In some cases, the gender gap
even increases once these attitudes are introduced. This indicates that if men and women
were to have similar attitudes (e.g., towards the presence of immigrants), the gender gap
would be larger.
To gain further understanding of the cross-national differences in the PRR gender gap,
we explored to what extent party characteristics, i.e., PRRs’ outsider image and populist
discourse style, may explain this cross-national variation. Taking insights from Roth (1989)
and Mudde (2007) who argue that these party characteristics are important in understanding the gender gap, we hypothesized that once PRRs are political insiders and use a less
144
Gender and populist radical right voting
populist discourse style; women will be more attracted to vote for these parties, resulting
in a smaller gender gap. However, our findings do not support our hypotheses. Yet, we
believe that the role of party characteristics deserves further attention in understanding
cross-national variation in the gender gap. Because the EVS had few European countries
with considerable numbers of PRR-voters, we could only perform an explorative analysis in
12 countries, which makes significant effects difficult to find. We should thus be careful in
concluding that party characteristics cannot help in explaining variation in the gender gap
in PRR-voting. Ideally, future research applies a random slope multilevel model incorporating more countries or time periods to replicate this examination and disentangle whether
these characteristics affect the gender gap in PRR-voting.
Moreover, we think that future research could investigate more carefully the crossnational differences in individual explanations of the gender gap. In some countries, the
gender differences in our explanatory characteristics are more substantial, which could
explain why, in these countries, a decrease of the gender gap is found. Further work can
investigate why these cross-national differences in explanatory characteristics may exist,
while also considering that other factors may be relevant in understanding male and female
PRR-voting behavior in different countries. For example, Coffé (2005) suggests that PRRs’
differing strategies, e.g., softening political stances concerning immigration, may be relevant when explaining the vote for a more established PRR compared to a new one.
For now, we believe this chapter added to the research on gender and PRR-voting by
exploring this gap across a wide range of Western European countries and including a
cross-national comparison. By doing so, we revealed substantial cross-national differences
in the gender gap in PRR-voting, which have been largely ignored in previous research.
Despite the inability of the party characteristics to explain cross-national differences in the
gender gap, we see this chapter as a starting point for developing new theoretical perspectives on party characteristics on how to explain the gender gap in PRR-voting behavior.
145
Chapter 7
Conclusion
147
Conclusion
7.1 Introduction
Ever since the entrance and growing popularity of populist radical right parties in Eastern
and Western European democracies in the 1980s, 1990s and 2000s, the scientific and public
debate about the why, the who, the where and the when of this contemporary phenomenon
has been increasingly pressing. In some countries, there were temporal peaks of PRR success, but in most European countries, these parties began to develop into serious political
contenders. Sometimes, their success resulted in entry into either a government or a supporting partnership in a minority government (e.g., De Lange, 2012b). It is hardly debatable
that the nativist, populist, and authoritarian PRR is here to stay (Zaslove, 2008).
The growing popularity and continuing emergence of new PRRs in the 1990s and 2000s
have triggered extensive political and scholarly debate (cf. Backes & Moreau, 2012). A focal
point in this debate concerns the consequences of the PRR for contemporary democracies
(Mudde & Rovira Kaltwasser, 2012; Rosanvallon, 2008). Some have claimed that the PRR
constitutes a serious threat to democracy because its leaders employ a radical notion of
democracy that emphasizes a homogeneous voice —the ‘voice of the people’— and excludes
minorities (Mudde, 2007). Others have noted that these PRRs actually correct democratic
deficiencies by speaking to a large group of citizens disillusioned with mainstream politicians
(Mudde & Rovira Kaltwasser, 2012). Citizens feel that there is someone who has ‘listened
to their grievances’ (Ivarsflaten, 2008) and who has enabled them to become passionately,
rather than rationally, involved in politics (cf. Mouffe, 2005).
This dissertation contributes to the ongoing public and scholarly debate about the consequences of PRR popularity for the voting behavior of European citizens. As Robert Jansen
(2012) has claimed, sociologists cannot longer sidestep the academic challenge of populism
and populist mobilization. Inspired by his notion of populism as a political practice that
uses populist rhetoric to include marginalized parts of society in the political realm, I
investigated how the PRR affects the voting behavior of citizens in Eastern and Western
European countries. The central aim of this study was to gain insight into the extent to
which popular PRRs influence (1) the decision of non-voters to vote (again), and (2) the
decision of cleavage voters to reconsider their habitual votes. In this final chapter, I will
discuss the main findings and evaluate them in light of the overarching research question.
Based on that evaluation I both reveal some limitations of this study and provide suggestions for future research. I conclude by summarizing the core insights of this dissertation.
7.2 Main findings
The overarching research question that guided this dissertation was: To what extent does the
popularity of the European PRR influence voting behavior, i.e., electoral turnout and cleavage
voting? This question is addressed in two parts based on the distinction of whether the PRR
149
Chapter 7
is able to attract (1) non-voters to the polling booth and (2) cleavage voters to reconsider
their party choice. In the first part (Chapters 3 and 4), I addressed the influence of the presence and popularity of the PRR on non-voters in Eastern and Western European countries.
The second part of this dissertation aimed to investigate the cleavage-breaking influence of
PRRs in Western European countries. This part focused on the question of whether people
who are known for their voting stability are also affected by the popularity and presence
of PRRs. Specifically, I asked whether religious people (Chapter 5) and women (Chapter 6)
reconsider their normal vote once PRRs become more successful. We focused on Western
European countries in that part, because religious and gender cleavages have been most
prominently visible and most strongly dominated 20th- and 21st-century (party) politics
(e.g., Inglehart & Norris, 2000; Knutsen, 2004; Van der Brug et al., 2009). The Communist
experience and legacy have made it impossible for some cleavages to emerge in Eastern
and Central European countries (Judt, 2010). Moreover, because it is argued that PRRs
may become ‘catch-all’ parties, we focused on those countries with successful PRRs where
cleavage-breaking patterns should have become most visible.
To help answer the abovementioned research questions, Chapter 2 presented a detailed
investigation into the uniqueness of the PRR’s party program and political style by comparing mainstream, established parties and PRRs on typical PRR-issues in 25 Eastern and
Western European countries in 2010. In particular, I looked at (i) ideological dimensions:
immigration, nationalism, and law and order; and (ii) style dimensions: populism and
anti-establishment image. Based on my own Expert Judgment Survey of European Political
Parties 2010 (Immerzeel et al., 2011), I investigated how close PRRs and other parties are
positioned on these ideological and style dimensions, and examined the extent to which
other parties’ characteristics (ideological background, electoral success, and being in office)
and PRR characteristics (electoral success and being in office) are associated with these
distances.
The results from Chapter 2 show that despite the findings of earlier research that all
other parties may have moved into the direction of the PRR’s immigration stance (Van
Spanje, 2010), there remain considerable differences in the ideological stances, populism
and anti-establishment between PRRs and other mainstream parties in both Eastern and
Western Europe. The results showed that the distances between PRRs and other parties
were generally larger in Western Europe than in Eastern Europe on all dimensions, with
the exception of nationalism. However, despite some small differences between Eastern
and Western European parties in the distances between the PRRs and some party families,
the PRR seems to take an exceptional position in the party system with its combination of
populism, nativism (anti-immigrant and nationalism) and authoritarianism (as suggested
by Mudde (2007)).
150
Conclusion
Although it is suggested that ideologically, the parties can be represented by a horseshoe
shape (e.g., Halikiopoulou et al., 2012) such that the socialists are closer to PRRs than the
center and right-wing parties, no evidence is found of that regarding the issues of immigration, nationalism, law and order and populism. In fact, neo-conservative, conservative and
Christian-democratic parties are closest to the PRR position on these typical PRR-issues.
Socialist parties show the largest distance from the PRRs, followed by the Greens and social
democrats. It is only with respect to the anti-establishment dimension that I found a horseshoe-shape: the positions of the socialist (and Green) parties on the anti-establishment
dimension are much closer to the PRR’s position than those of the center parties to the
PRR’s position. In addition, I found that the other parties’ position on typical PRR issues
primarily depend upon their own popularity: on average, more successful parties have a
closer position to that of the PRR.
Based on this investigation, I concluded that the PRR takes a ‘unique’ position in the
political opportunity structure and therefore may have a unique potential to affect people’s
voting behavior. Below, I will present the main findings related to the two parts of the
dissertation and then evaluate its overarching research question.
7.2.1 Part 1: The PRR and its influence on non-voters
In this part, I aimed to investigate the extent to which the PRR is able to attract non-voters
to the polling booth and accordingly, contribute to a higher turnout at elections. In Chapter
3, I studied the relationship between the PRR’s electoral popularity and the individual propensity to turn out at a national legislative election both in Eastern and Western European
countries. The research questions were: What is the PRR’s effect on voter turnout levels in
Europe? Who is encouraged or discouraged to vote by the PRR’s popularity? In this chapter, I
departed from the electoral competition theory (Franklin, 2004) by arguing that the PRR
strengthens the competition for votes by entering the political arena. By doing so, it may be
expected that the PRR has a generally positive impact on electoral turnout. However, the
PRR is also known for its radical extreme anti-form of politics (both in style and ideology,
as demonstrated in Chapter 2), which may negatively impact citizens’ trust in politics and
democracy. Thus, I also formulated the counter-hypothesis that popular PRRs, with their
radically anti-form of politics, may discourage people to turn out for legislative national
elections. Using the six waves of the European Social Survey (ESS, 2002-2012), I tested
these opposing hypotheses in both Eastern and Western European countries over the period 2002-2012.
The results did not show a positive influence of PRR’s success on electoral turnout.
In contrast, only Eastern European countries show an indication that the popularity of
the PRR actually has a negative effect on the individual propensity to turn out to vote. In
Western European countries, no effect is found. Despite some anecdotal evidence that some
151
Chapter 7
PRRs may be able to passionately mobilize citizens into politics (Mouffe, 2005), there does
not seem to be a general, universal PRR effect that increases the level of electoral turnout
in either Western or Eastern Europe. Hence, to speak of PRRs as ‘correctives of democracy’
is—with respect to electoral turnout—a misunderstanding.
However, one should be careful about concluding that the PRR has no impact at all on
electoral turnout. Based on first-wave PRR research (Betz, 1994; Rydgren, 2007) and more
general theories about turnout (Blais & Rubenson, 2013), I argued in Chapter 3 that there
might be specific social groups whose turnout may be either encouraged or discouraged by
the success of PRRs. The differential impact of PRRs on different segments of the electorate
may eventually lead to a general null-impact of PRR success. Using multiple cross-sections
of Eastern and Western European citizens in the period 2002-2012, I examined whether
the PRR affected the voting propensity of specific groups. On the basis of first-wave PRR
research, I hypothesized that the typical PRR-electorate (e.g., Arzheimer, 2012; Lubbers et
al., 2002), i.e., the lower-educated, blue-collar workers, and unemployed people, who are
politically dissatisfied and have negative attitudes toward immigrants, is more likely to vote
once the PRR has become more popular. Moreover, and in line with the turnout literature, I
argued that in particular, young and politically disengaged people may be prompted by the
popularity of the PRR to go to the polls.
Surprisingly, the results in Chapter 3 indicated that the typical PRR electorate is not
prompted to go to the polls by the PRR’s success. The blue-collar workers, the lower educated, those with anti-immigrant attitudes and those who are politically dissatisfied had a
voting propensity that either was generally discouraged or not affected in both Eastern and
Western European countries. Although these groups are overrepresented among the PRR’s
electorates (e.g., Lubbers et al., 2002), I argued that these groups do not seem to belong
to the segment of the electorate that usually stays home and is triggered to go the polls
when the PRR becomes successful. In contrast, I assumed that the PRR receives most of its
support from people who were already voting (before there was a popular PRR), and shifted
their party preferences once the PRR became more successful. In addition, I speculated that
this pattern could be explained by stealth democracy theory, which states that typical PRR
supporters want a responsive government without going to the polls themselves (cf. Hibbing & Theiss-Morse, 2002; Mudde, 2004). They may view the PRR, which presents itself
as the embodiment of the people, as the perfect candidate to control the decision-making
processes of the ‘corrupt’ mainstream parties and to deliver a responsive government. Once
those voters anticipate that the PRR has become more popular and that other people are
voting for the PRR (e.g., via opinion polls taking place before Election Day), they may feel
less of an urge to go to the polls, which explains why I did not find a positive mobilization
effect.
152
Conclusion
In Chapter 3, I found some substantive East-West differences regarding the effect of
PRR popularity on individual propensity to turn out at elections. In Eastern Europe, I found
that most people are actually discouraged from going to the polls. In particular, Eastern
European youngsters were negatively affected by the success of PRRs: they are much less
likely to vote once a PRR has become more popular. In Western Europe, such negative
effects of the popularity of the PRR on youngsters’ individual propensity to turn out have
not been found. Therefore, Western European youngsters do not seem to be affected by
the popularity of the PRR. I argued that these different results might be explained by the
findings of Chapter 2, which show that in Western Europe, PRRs hold a unique position
in the political opportunity structure compared to their Eastern European counterparts.
The latter are ideologically closer to the other established parties, which may indicate that
radical rhetoric is more evenly spread across the political spectrum (cf. Učeň, 2007). I
argued that Western European youngsters may see the PRR’s uncivil, aggressive style as
exceptional to the entire political system and therefore, that the PRR does not affect their
general trust in politics. Thus, the increasing popularity of the PRR might not translate
into lower turnout levels among Western European youngsters (Mutz & Reeves, 2005). In
Eastern Europe, however, where the radical, extreme form of politics is less concentrated
within one single party, youngsters might feel that the increasing popularity of the PRR and
its radical, uncivil rhetoric is a deficit of the entire political system, thus resulting in lower
turnout rates at elections that include popular PRRs.
In Chapter 3, I showed that different parts of the electorate are encouraged to vote,
whereas others are discouraged from voting. This partly explains why I found small or
no universal effects of PRR popularity on individual propensity to turn out. However,
the results from Chapter 3 did not show whether people who are encouraged to go to the
polls are also those who vote for the PRR. This is because I was unable to follow the same
individuals over time and study the party choice of those who are encouraged to vote.
As I hypothesized in that chapter, it may very well be that nonvoting citizens turn to the
polling booth because they are repelled by the PRR’s ideology and political style (Andersen
& Evans, 2003) and want to keep the ‘rascals out’ (cf. Van Spanje, 2011). This was partially
supported by the finding that in Western European countries with a more popular PRR, the
politically interested are more likely to turn out.
To be able to investigate where these non-voters go, in Chapter 4 I followed a specific
group of Dutch non-voters over 7 years and investigated whether the PRR is indeed most
attractive to non-voters once they (re)turn to voting. The research question guiding Chapter
4 was as follows: To what extent are (former) Dutch non-voters more likely to vote for a PRR
instead of for another party and how can that change be explained?
Using the Longitudinal Internet Studies for the Social Sciences (LISS, Tilburg University,
2007-2014), the voting intentions of a specific group of eligible, nonvoting Dutch citizens
153
Chapter 7
were followed over the period 2007-2013. I found that the majority of this nonvoting group
is most likely to remain nonvoting over time, providing evidence for the growing literature
on habitual voting (e.g., Plutzer, 2002). However, Chapter 4 also demonstrated that a large
minority of non-voters is unstable in its voting behavior and that people tend to ‘break
through’ their general voting patterns. I found that a change in one’s interest in politics
could explain why habitual non-voters are more likely to vote in a future election. Whereas
earlier studies also indicated that one’s interest in politics is important in understanding
voter turnout (cf. Verba et al., 1995), this chapter showed that primarily changes in political
interest, rather than changes in structural individual characteristics—such as employment
status, having a partner or a child—can explain why non-voters decide to vote in later
elections. I argued that this could be explained by the fact that these structural changes take
place later in life, when one’s habitual nonvoting behavior has already been set in motion.
However, these non-significant findings of changes in structural characteristics can also be
due to too little variability in the structural variables, whereas political interest is an attitude
that potentially more strongly varies over time and correlates, for example, with the amount
of attention paid by the media to politics (cf. Koopmans & Muis, 2009).
In Chapter 4, I furthermore tested the expectation that especially the PRR has a special
ability to trigger political interest and that non-voters turn to the PRR once they have
decided to vote. This hypothesis was supported. I found that if non-voters decide to vote
in a future election, they are generally more likely to vote for a PRR than for any other
party. Surprisingly, the results of the analyses showed that this specific volatility (i.e., from
nonvoting to a PRR vote) could not be explained by a change in typical PRR attitudes.
Although non-voters who shift to the PRR have more favorable PRR attitudes (regarding
immigration, politics and the EU) than habitual non-voters and those who shift to other
parties, the findings suggested that it is particularly the change in sympathy for PRR leaders
that is strongly associated with an increased likelihood to vote for a PRR instead of another
party. This finding coincides with the second-wave PRR research (e.g., Art, 2011; Lubbers
et al., 2002) and is an important finding because it suggests that changes in ideological
considerations among non-voters do not explain why nonvoting people turn to a PRR.
Rather, it seems that non-voters who are already favorable toward PRR-ideology can only
be mobilized if they also like the leader. Consistent with findings from Schumacher and
Rooduijn (2013), I found that party leader evaluation is decisive in understanding why the
PRR is able to attract politically disengaged voters.
7.2.2 Part 2: The PRR and its influence on cleavage voters
Whereas the first part of this dissertation focused on the potential mobilization of nonvoting citizens into the political arena by the PRR, in the second part of this dissertation, I
focused on the potential of the PRR to influence those who are generally labeled as cleav154
Conclusion
age voters (cf. Knutsen, 2004; Lipset & Rokkan, 1967) and are known for their stability of
their party choice (Aarts & Thomassen, 2008), to reconsider their habitual votes. More
specifically, in line with second- and third-wave PRR research, I argued that if the PRR
is able to fundamentally change the party system and the topics around which political
parties compete for votes (Pellikaan et al., 2007), then voters that are typically locked up in
their own cleavage would also be more likely to reconsider their vote (cf. Lipset & Rokkan,
1967). In particular, I looked at whether religious people and women in Western Europe are
affected by PRR success and whether they are likely to change their normal voting patterns.
First-wave PRR research has demonstrated that religious groups and women are generally underrepresented in the PRR electorate, but I argued that the presence and current
popularity of the PRR prompted them to reconsider their voting behavior and to vote for
the PRR.
In Chapter 5, I looked at whether the presence of a popular PRR may cause religious
voters to be more likely to vote for another party than for the Christian party. The research
questions were as follows: To what extent are PRRs able to attract religious voters who are
not generally expected to vote for these PRRs? If the PRR does have such an ability, why? I
departed from the notion that the PRR, despite being primarily secular (Camus, 2007),
has something to offer to Christian believers. The Western European anti-immigrant PRR
has recently presented itself as the safeguard of the ‘Judeo-Christian societies’ that defend
people from the Islamic threat (e.g., Wilders, 2012). Moreover, Israeli scholars have found
that Israeli people’s motives to vote for a religious fundamentalist party largely coincide with
the motives to vote for a PRR (Hirsch-Hoefler et al., 2010). Thus, there might be something
in religion that makes for PRR-voting.
Chapter 5 built on the Weberian assumption that religiosity entails two dimensions: a
practice and a belief dimension (Driskell et al., 2008; Weber, [1930] 2003). I expected that
both may be relevant in explaining a vote for a PRR or another party and that specifically
the belief dimension helps to solve the puzzle of why religious people would be tempted to
vote for a PRR. I postulated the (i) threat mechanism, which states that those with a more
orthodox fundamentalist belief system feel more religiously threatened by Muslim immigrants (e.g., Abu Raiya et al., 2008) and therefore are more likely than more mainstream,
tolerant believers to vote for a PRR instead of a Christian party. With respect to the religious
practice dimension, I proposed the (ii) norm-compliance mechanism, which states that
those who attend religious services more often will more strongly adhere to the norm of
voting for a Christian party and therefore are more likely to vote for a Christian party
instead of a PRR, notwithstanding their agreement with the PRR’s ideology (cf. Arendt,
[1951] 1973; Arzheimer & Carter, 2009).
By analyzing wave 4 of the European Values Study (EVS, 2010), I was able to test these
underlying mechanisms. With respect to the religious belief dimension, the findings provide
155
Chapter 7
some evidence that one’s religious conviction affects one’s vote for either a PRR or a Christian
party. In Belgium, Switzerland and Norway, the results show that orthodox believers have
generally more negative attitudes toward immigrants than mainstream believers, which in
turn is associated with a higher probability of voting for a PRR instead of a Christian party.
Thus, it seems that in these countries, orthodox believers feel more religiously threatened
by the presence of immigrants of a religion that has fundamentally different truth claims.
These results partly countered the religious cleavage idea that religious people tend to vote
for a religious party (Knutsen, 2004; Van der Brug et al., 2009) and therefore, there is some
evidence that PRRs can ‘break through’ traditional voting patterns. A PRR with a strict
stance on immigrant issues can be a credible alternative for religious people with a distinct,
orthodox belief system in these three countries.
In addition, the results in Chapter 5 suggested that religiously active people are more
likely to vote for a Christian party instead of a PRR in the majority of countries under study
(Billiet, 1995; Lubbers & Scheepers, 2000). Nevertheless, this finding could not determine
whether the norm-compliance mechanism, which states that the religiously active are better integrated into a religious community and more strongly adhere to the norm of voting
for a Christian party, drove that relationship. Accordingly, I assumed that if this mechanism
is at work, strongly integrated Christian people will, independent of their anti-immigrant
attitudes, be more likely to vote for a Christian party instead of the PRR. Conversely, for
less-integrated Christian people, I expected that anti-immigrant attitudes to be more
decisive in their voting choice and that the likelihood of voting for a PRR instead of a
Christian party would increase accordingly. However, this hypothesis was not supported.
I argued that this could be partly explained by the notion that even the slightest involvement in religious activities fosters tolerance toward others (Newton, 2001; Putnam, 2000).
Consequently, this reduces the development of negative attitudes towards immigrants, and
thus lowers the probability of voting for the PRR. Moreover, I argued that the significance of
the first wave social disintegration theory in understanding the vote for a PRR might have
been overestimated (cf. Rydgren, 2009; Rydgren, 2011).
Finally, in Chapter 6 I looked at the influence of the PRR on the gender cleavage in
party choice (cf. Inglehart & Norris, 2000; Inglehart & Norris, 2003). It is common wisdom
that there is a gender gap in PRR-voting with primarily men voting for PRRs. In this last
chapter, I explored whether such a PRR-gender gap exists across 12 Western European
countries and examined how this gap may be explained. The research questions were as follows: To what extent is there a gender gap in Western European PRR-voting, and how can this
gap be explained? To what extent can the image of a populist outsider explain cross-national
variation in the gender gap in PRR-voting? Based on analyses of the European Values Study
(2010), I indeed found a gender gap in PRR-voting that remains substantial after controlling
for socioeconomic and political characteristics. Only gender differences in socioeconomic
156
Conclusion
position, i.e., the overrepresentation of men in the blue-collar sectors, could partly explain
the gender gap in PRR-voting in some countries.
However, the results also indicated strong cross-national variations in the size of the
gender gap. I argued that this cross-national variation might be explained by the PRR’s
potential to break through women’s normal voting pattern. It is often theorized that women
shy away from voting for a PRR, because of their opposition to typically masculine populist
rhetoric (Huddy & Terkildsen, 1993; Mayer, 2002) and their lower inclination to vote for
political newcomers (e.g., Roth, 1989). Simultaneously, it was argued that once the PRR has
become popular and has transformed into a more stable, political actor, it loses its outsider
image and people become acquainted with its populist style of discourse. Along with that
argument, I had expected that women’s former hesitance to vote for a new party such as the
PRR would diminish once the PRR had become more popular and transformed itself into
a more mainstream political contender. In turn, this could have led to smaller differences
among men and women in their tendency to vote for the PRR. However, the results in
Chapter 6 did not support this expectation: differences in party characteristics did not account for cross-national differences in the gender gap. When the PRR is more popular and
populist or belongs to the establishment, the gender gap in voting behavior does not seem
to be affected. Thus, I did not find evidence that women will be more likely to reconsider
their voting behavior and vote for the PRR. However, I argued that one should be careful
in concluding that party characteristics cannot help to explain variations in the gender gap
in PRR-voting. Because I could only perform an explorative analysis in 12 countries, it is
difficult to find significant cross-country variation. Moreover, I could not account for other
potentially explanatory variables, such as the degree of masculine rhetoric that PRR-leaders
employ, in the various countries under study.
7.3 Main conclusions
The primary aim of this dissertation was to provide insight into the effect of the presence
and popularity of the PRR on citizens’ voting behavior, i.e., electoral turnout and cleavage
voting. Based on the five empirical chapters of this dissertation, I conclude that the PRR
generally has either no influence or a negative influence on electoral turnout. Despite the
unique ‘profile’ of the PRR as a populist, nativist contender and some anecdotal evidence
that the PRR’s electoral breakthroughs accompany rising turnout, these parties do not have
the general positive impact on electoral turnout that scholars had hypothesized. In Western
European countries, the PRR generally has no influence on electoral turnout, whereas in
Eastern European countries it has a negative effect.
Although popular PRRs do not have a generally positive influence on electoral turnout, Western European PRRs do encourage some social groups to turn out for national
157
Chapter 7
elections. These groups are, however, not the people who agree with the PRRs’ ideology,
but seem to be those groups that are actually repelled by them (cf. Andersen & Evans,
2003). Western Europeans who are more politically interested are more likely to turn out
with growing PRR popularity. They may be more inclined to keep the ‘rascals out’ (cf. Van
Spanje, 2011). In Eastern Europe, such positive effects on electoral turnout for those who
oppose PRR ideology were not found at all. In contrast, anti-PRR Eastern Europeans were
actually strongly discouraged from voting with increasing PRR popularity. More highly
educated, non-blue collar workers and politically interested citizens were more inclined to
stay home at national election time when the PRR became more popular.
We argued above that this might be explained by the different positions taken by PRRs
in Western versus Eastern European countries. However, these East-West differences may
also be caused by the fact that Eastern Europeans live in younger democracies and have less
experience with electoral competition than their Western European counterparts (cf. Blais,
2006; Lijphart, 1999; Mudde & Rovira Kaltwasser, 2012). Anti-PRR Western Europeans are
often socialized in a democratic regime where elections are accompanied by contentious
political campaigns. They may think of elections as a legitimate method of ‘protest’ against
the PRR and thus turn out to vote (cf. Hirschman, 1970; Thomassen et al., 2014), whereas
anti-PRR Eastern Europeans may interpret the PRR’s success as a confirmation of their image of democracy, politics and politicians as being untrustworthy and corrupt. This results
in lower levels of electoral turnout (Brooks & Geer, 2007; Kahn & Kenney, 1999).
Likewise, democracy’s different age in these two regions may also explain why I found
that the voting propensity of Western European youngsters is not affected by popular PRRs,
whereas Eastern European youngsters are strongly discouraged from voting when the PRR
is popular. Living as a Western European youngster in a country where electoral turnouts
are generally higher, where voting is to a greater extent interpreted as a civic duty, and
where voting is seen as the generally accepted means to influence politics (rather than
demonstrating, boycotting et cetera) might condition the PRR’s potentially negative impact
on electoral turnout in Western Europe (cf. Campbell, 2006). In addition, these findings
underline the claim of some scholars arguing that the PRR poses a threat to democracy: if
young people do not vote, their interests may be underrepresented in the legislature, which
may have long-term effects on future turnout. Some studies show that if youngsters do not
vote at the first opportunity, they will be more likely to become habitual non-voters (e.g.,
Plutzer, 2002). In the short term, this leads to lower levels of turnout among youngsters.
In the long term, this younger generation could pass down their habitual nonvoting to the
next generation, which eventually leads to even lower levels of turnout across the electorate
(cf. Blais & Rubenson, 2013). Hence, I conclude that for Eastern Europe, one should be
concerned about the popularity of the PRR because it seems to discourage young people
from turning out at elections.
158
Conclusion
With respect to the influence on non-voters, I conclude that the PRR is not the passionate mobilizer that earlier studies have suggested (e.g., Mouffe, 2005). There is some evidence
that non-voters who decide to vote in a future election turn to the PRR instead of to other
parties, but these changes are not as widespread as expected. Non-voters generally tend to
remain non-voters. A small group of non-voters decide to become voters primarily because
of the likability of the PRR’s political leader, but it seems that this only occurs if such voters’
ideological stances are in line with the PRR. If charismatic leadership is the primary reason
for non-voters to become mobilized, then it is not strange that no universal trend is found
toward greater electoral turnout. As Weber (2009) and Pels (2011) note, charismatic leadership is a social construct that waxes and wanes with the attention of the public and the
media. It may well be that the PRR’s mobilization of non-voters is actually temporal. This
may imply the existence of a group of indifferent voters who generally do not vote, but who
are more likely to vote if they like a leader. Moreover, the idiosyncrasies of PRR leadership,
which undoubtedly vary across political arenas and time, may explain the lack of consistent
country-level patterns in the PRR’s electoral mobilization reported in Chapter 3.
With respect to the effect on cleavage voters, however, I conclude that popular PRRs
do pose an ideological dilemma for religiously orthodox people. The findings showed that
the religiously orthodox are more likely than their mainstream counterparts to vote for the
PRR. This is primarily for ideological reasons: religiously orthodox people may feel that
the fundamental truth claims of Christian orthodoxy are more strongly threatened by the
presence of immigrants. Accordingly, the PRR has the potential to win the votes of religious
citizens who have traditionally supported Christian democratic parties. Over the long term,
this may imply that the PRR could contribute to the hollowing of Christian democracy
that, despite processes of secularization, has remained relatively stable in most Western
European countries (e.g., Kalyvas & Van Kersbergen, 2010; Van der Brug et al., 2009).
This study did not provide evidence that women are reconsidering their party choice
with the PRR’s growing popularity. Despite the fact that there is cross-national variation in
the differences between men and women in their tendency to vote for the PRR, that variation cannot be explained by the variation in these parties’ popularity or political style. That
the PRR has become an alternative for orthodox Christians but not for women may lie in
the notion that the traditional religious cleavage is becoming less relevant, other attitudinal
cleavages are becoming more important, and people are becoming more inclined to look for
alternatives that better represent their wishes and (religious) interests (Aarts & Thomassen,
2008). In the case of gender, women are actually in the process of realigning their party alliances and becoming even more inclined to vote for a left-wing ‘women’s’ party to exclude
the typical male parties from legislatures (Ignazi, 1992; Inglehart & Norris, 2000).
Altogether, I conclude that the PRR’s influence on non-voters and on cleavage voters is
small. Although earlier research has found that the PRR seems to affect the party system
159
Chapter 7
(e.g., Pellikaan et al., 2007), other political parties’ stances regarding immigration (Van
Spanje, 2010), and attitudes toward immigrants (Semyonov et al., 2006; Wilkes et al., 2007),
the results of this dissertation show that the PRR is not as influential for the voting decisions
of non-voters and cleavage voters. The claim that the PRR strongly motivates non-voters
and cleavage voters to reconsider their voting behavior should be nuanced based on the
findings presented in this dissertation. The PRRs may have somewhat more of an attraction for non-voters and religious orthodox people than do other parties, but there is no
indication that electoral turnouts are generally rising and the traditional social cleavages
are crumbling with the PRR’s increasing popularity (cf. Kalyvas & Van Kersbergen, 2010).
Nonetheless, because these parties have become popular and have received support from
many citizens, these results imply that the PRR does affect the voting behavior of those
segments of the electorate that were already voting. Moreover, it could be that the PRR
attracted a substantial portion of the previous voters who might have abstained if the PRR
was not there. Thus, although these parties did not seem to fuel much political support
among non-voters and cleavage voters, their mere presence and popularity might have
prevented the turnout rates from dipping even lower.
7.4 Contributions
With this dissertation, I have provided pan-European, cross-country and over-time empirical evidence of the debate of whether PRRs are able to influence voting behavior, i.e.,
electoral turnout and cleavage voting. As such, it tested the often-heard claim that PRRs
are ‘passionate’ drivers of popular engagement with politics and thus are a curative for
democracy. The various empirical chapters have contributed in different ways to the debate
about the consequences of PRR popularity and to the emerging PRR research tradition.
First, the focus on political behavioral consequences of popular PRRs is a novel and
important contribution to the third-wave PRR literature related to the ‘corrective’ and
‘threatening’ consequences of PRRs on contemporary liberal democracies. Whereas most
studies in that third wave addressed potentially policy-related and social consequences,
(e.g., Akkerman, 2012; De Lange & Akkerman, 2012; Mudde, 2012; Zaslove, 2004), this dissertation more explicitly studied the political behavioral consequences of PRR-popularity.
Accordingly, it strengthened the third wave by theorizing and testing how the PRR may
indirectly influence the policy and polity process via the individual decisions of voting citizens. In addition, earlier case studies addressing the relationship between the PRR’s success
and political behavior neither show whether it is specifically the PRR that drives the change
in electoral turnout nor whether there is a general PRR effect on electoral turnout in both
Eastern and Western European countries. The cross-national studies in this dissertation
enabled me to formulate conclusions that are more coherent and to make claims that are
160
Conclusion
more generalizable about the influence of the PRR’s popularity on electoral behavior. This
study’s cross-national and over-time comparison focusing on voting behavior allows it to
nuance the view of some scholars who see PRRs as the passionate mobilizers of renewed
political interest (Mouffe, 2005). The various chapters of this dissertation showed that
the PRR has no universal ‘corrective’ effect on electoral turnout and cleavage voting. For
Eastern Europe, the study showed even more strongly that the PRR actually has a generally
detrimental effect on electoral turnout.
Second, with this study I have combined theoretical insights belonging to the various
waves of the PRR research tradition. I have used first- and second-wave theories, such as
ethnic competition theory and losers of modernization theory, to understand why people
reconsider their voting behavior. The chapters provide novel empirical evidence for why
people shift their non-vote or cleavage-vote into a vote for a PRR, thus providing new
insights into existing theoretical stands. For example, changes in PRR-leadership considerations seem to be crucial in non-voters’ decisions to vote for a PRR. This lends credence
to second-wave PRR-theories related to party characteristics, such as leader charisma (e.g.,
Lubbers et al., 2002) and challenges first-wave demand-side theories, such as ethnic competition theory, which hypothesized that people vote for the PRR from ideological incentives
alone (e.g., Ivarsflaten, 2008).
In addition, I contributed to first- and second-wave PRR studies by theorizing and testing new explanations for the often-found religion and gender cleavages in PRR-voting. The
introduction of the religious orthodox dimension in Chapter 5 of this book informs firstwave studies that have consistently found that the PRR is not an alternative for religious
people (e.g., Arzheimer & Carter, 2009). I found that there might be reasons for religious
people to turn to PRRs, which shows that the relationship between religion and PRR-voting
is not as straightforward as previous studies suggested.
Similarly, Chapter 6’s cross-national investigation into the gender gap showed that
the ‘gender gap in PRR-voting’ is in fact a misconception. First-wave PRR studies consistently found that women were underrepresented in the PRR electorate (e.g., Mudde, 2007;
Rippeyoung, 2007). However, a further inspection of gender gaps showed that there are
more cross-national differences in the gender gap than suggested by previous studies and
thus asked for cross-national explanations and conditions for why these gender gaps differ
in size. Despite the fact that PRR’s popularity and its populist style of discourse could not
explain why the gender gap is smaller in some countries than in other countries, this study
contributed by theorizing and testing how these differences can be explained. Accordingly,
an integration of the three waves of PRR research helped to improve our knowledge of why
people vote for the PRR and the role played by the PRR itself in those voting decisions.
Third, this dissertation theorized about why the PRR could both positively and negatively affect electoral turnout and focused on the important question of whether some social
161
Chapter 7
groups are more or less likely to turn out with the popularity of the PRR. Accordingly, it
does not only contribute to the PRR research tradition but also added to the literature
about electoral turnout, electoral volatility and habitual voting. The specification of different social groups enabled me to show that despite the general null-impact of PRR success
on electoral turnout, there might very well be differential impacts on different groups of
people. Earlier research focusing on electoral turnout and the role of electoral competition
has not always incorporated such a strategy or has focused only on one social group (e.g.,
Blais & Rubenson, 2013).
7.5 Limitations and suggestions for future research
The research in this dissertation has a number of limitations that could be used to pose new
questions for future research. In this section, I focus on the more general limitations of the
research project and formulate suggestions for future research that go beyond the more
specific suggestions made in the various chapters.
It is important to note that this study focused on voting intentions and voting behavior
in national legislative elections and take for granted that these elections are all similar in
terms of how the PRR can influence voting behavior. However, anecdotal evidence suggests
that the first entrance of a PRR is particularly crucial in stirring up popular political engagement (e.g., the Fortuyn-revolt (Couwenberg, 2004)). Moreover, the study by Koopmans
and Muis (2009) shows that how these PRRs become popular and whether they are able
to mobilize political support largely depends upon their visibility and novelty and the
credibility provided by other party leaders in the media. Thus, not all legislative elections
are similar with respect to how the PRR may influence voting behavior. Therefore, future
research should distinguish between the effect of the PRR’s first entrance and the effect of
its later popularity on electoral turnout. It could very well be that ‘being new’ and ‘being
different’ are the best drivers of electoral competition and thus affect non-voters’ decisions
to vote more strongly than if the PRR was becoming a more stable political actors unable to
surprise and mobilize the people.
Furthermore, the focus on national legislative elections ignores how PRR breakthroughs
in ‘second-order elections’, such European Parliament or the municipal election, influence
the voting behavior of citizens in first- and second-order elections. The recent successes
of the United Kingdom Independence Party [UKIP] and the French Front National [National Front, FN] in European parliamentary and municipal elections show that without
the national first-past-the-post system, PRRs have the potential to attract citizens and attain
seats, whereas they are unable to do so at the national level. The PRR and other populist
parties seem to fare better in proportional, second-order European elections. Thus, there
could be institutional barriers that affect how the PRR may influence individual voting
162
Conclusion
instability. Where do PRR voters come from? Future research could pay more attention to
how elections with a different nature and different institutions can affect individuals’ voting
instability. What are the reasons for people to vote for a PRR in European elections and not
in national elections? What do these people vote for in an election that hinders the success
of the PRR?
With respect to the voting instability of voters, this study applied both cross-national
and over-time comparisons. However, the over-time comparison was only performed in
the Netherlands. That makes it difficult to formulate Europe-wide conclusions of whether,
why and when non-voters turn to the PRR instead of another party. The results in Chapter 3
show that there are East-West differences in the degree to which citizens’ turnout propensity
is affected by the success of the PRR. Given these findings, along with discussions about
the legitimacy of European democracies, it is valuable for future research to investigate
why Eastern European youngsters shy away from elections once the PRR becomes more
popular. Is it the image of politics and the lack of civic duty, as we suggested, or are other
mechanisms at work that keep them away from the polling booth? To understand whether
civic duty and the enforcement of social norms regarding turnout and cleavage-voting condition the impact of PRR success on non-voting and cleavage voting, I believe it is necessary
to account for the influence of the contexts in which these norms are formed and enforced.
That means that we should also study the role of partners, families, neighbors, communities
and countries (cf. Campbell, 2006) to understand why we find that individuals differ in
changing their voting behavior when faced with increasingly popular PRRs.
Moreover, the longitudinal study does not yet inform us about the encouraging impact
of the PRR on those voters who are repelled by the PRR’s ideology. This study provides a first
indication of whether people also are appalled by the rise of PRR, but I cannot tell whether,
for that reason, they are also more inclined to vote for a different party. Is it the case, as suggested in Chapter 3, that anti-PRR citizens voice their opinion by voting for a party that is
strongly opposes the PRR, or do they—in Hirschmann’s (1970) words—exit from politics?
In addition, do anti-PRR citizens react similarly in countries with a long versus short tradition of electoral democracy? To answer these questions, future research is urged to combine
different longitudinal panel data sets from different countries that follow citizens over a
longer period to more deeply investigate how changes in the popularity of different PRRs in
different countries influence voting behavior. This helps to explain potential cross-national
differences in non-voters’ tendency to become PRR voters; however, it also can help in
explaining why people turn to other parties or stop voting with the increasing popularity of
the PRR. Moreover, it provides us with better and more fine-grained insights about whether
voters become more inclined to vote for the PRR over time.
In addition to the nature of the elections and the extension of the data, I believe that
it is also important to study the effect of PRR popularity on the voting behavior of an163
Chapter 7
other typical social group: immigrants. As the scapegoats of the PRR’s nativist ideology,
one may expect them to be affected politically. Of course, it is known that immigrants are
legally excluded from voting from certain elections in certain countries. However, secondgeneration immigrants are allowed to vote and can vote in different types of elections. What
is the impact of PRRs on their voting behavior? Will they be triggered to vote or will they
be discouraged from voting once a PRR has become more successful? Are there any crossnational differences in the extent to which immigrants experience and behave politically in
the face of the popularity of the PRR?
Finally, this dissertation primarily focused on the consequences of PRR success on
voting behavior: both electoral turnout (nonvoting) and cleavage voting. The focus on
voting behavior, and electoral turnout specifically, is chosen because it is often assumed
as an indicator for the quality of democracy (e.g., Blais & Rubenson, 2013; Mair, 2006a;
Mair, 2006b; Van Reybrouck, 2013). Recently, however, it has been argued that although
electoral turnout numbers are declining, this does not necessarily imply that democracy is
under threat (e.g., Thomassen et al., 2014). It could very well be that citizens still believe
that there are enough parties that represent their interests and that they still believe that
(representative) democracy is the most appropriate and legitimate political system to govern a country (ibid.). Additionally, it has been argued that more unconventional forms of
political behavior, such as demonstrations and boycotts, become more standard as more
citizens participate in them (e.g., Dalton, 2008; Dodson, 2011; Rosanvallon, 2008). Although it remains questionable whether these other forms of political behavior have a real
impact on the political decision-making process and whether politicians really take notice
of these other forms of ‘between-election’ political behavior (see for an overview: Esaiasson
& Narud (2013)), a different manner of popular engagement with politics may replace the
conventional political action repertoire. Thus, it follows that the quality of democracy is
reflected not only by participation in elections but also by popular involvement in other
forms of political participation. Therefore, future research investigating the consequences
of PRR popularity on the quality of democracy should also include dependent variables
like attitudes toward the legitimacy and functioning of democracy and other, betweenelections, forms of political participation.
7.6 Concluding remarks
This book started by describing the Dutch experience of the sudden emergence and popularity of the PRR leader Pim Fortuyn. His ‘revolution’ has provided this dissertation’s first
building block, which aimed at understanding how and why the PRR is able to influence
non-voters and cleavage voters. Together, the five empirical chapters have shown that although Fortuyn seemed able to fuel popular engagement with politics, this is not a shared
164
Conclusion
phenomenon typical of the European PRR family. Departing from the notion that popular
PRRs bring more electoral competition and polarization by introducing a different political
style and party program, this book showed that in general, nonvoters and cleavage voters do
not reconsider their party choice. Nonvoters generally remain nonvoters, religious people
remain Christian-Democrat voters, and women remain non-PRR voters. These findings
highlight the importance of providing empirical scrutiny to the often-repeated claims in
the media that the PRR forms either a ‘threat’ or a ‘corrective’ to democracy. Interpreting
the European PRR and its politicians as radical democrats that know how to lure nonvoters
and cleavage voters to change their normal voting behavior is difficult to reconcile with the
main results of this dissertation.
165
Appendices
167
Appendices
APPENDIX 1.A
The Original Questionnaire of the Expert Judgment Survey of European
Political Parties 2010
1.
2.
3.
4.
Are you female or male? (1=female, 2=male)
What is your age?
From which country do you know best its political parties?
Would you please place the following political parties on the overall left-right dimension? (0 = very left, 10 = very right)
5. Would you please place the following political parties on the economic left-right dimension? (0 = very left, 10 = very right)
6. Consider the same parties, and place them on the following scale: restrictiveness concerning immigration (0 = Not restrictive at all, 10 = Very restrictive concerning immigration).
7. Consider the same parties, and evaluate their style of discourse (0 = Non-populist, 10 =
Very populist).
8. Consider the same parties, and place them on the scale that concerns the programmes
of the parties towards the law and order issue (0 = weakly support tough measures to
fight crime, 10 = strongly support tough measures to fight crime)
9. Consider the same parties, and place them on the scale that concerns the degree of
cosmopolitism or nationalism in the party programmes (0 = strongly advocates cosmopolitism, 10 = strongly advocates nationalism)
10.Consider the same parties, and place them on the scale that concerns the extent to
which the party is considered as anti-establishment (0 = not anti-establishment, 10 =
very anti-establishment)
11.Please give a mark from 1 to 10 for each party concerning the following aspects:
a.Organization of political party on national level (0=very badly, 10=very well organized)
b. Internal party consensus (0=very little, 10=very high consensus)
c. Charisma of party leader (0=very little charisma, 10=very much charisma)
Expert Selection and Response Rate
The Expert Judgment Survey was organized in 38 countries: all countries of the European
Union except Luxembourg, plus Croatia, Iceland, Israel, Norway, Serbia, Switzerland and
Turkey. Originally, the survey was also designed for experts of political parties in Albania,
Bosnia, Macedonia, Russia, and Ukraine. However and despite sending a printed version
169
Appendices
of the questionnaire to the experts in Bosnia, Russia and Ukraine, fewer than five experts
(i.e. the target response) of these countries cooperated, which made me decide to drop
these countries. In total 2,310 experts were e-mailed and invited to participate in the survey. Scholars’ names were selected from the homepage of political science departments of
universities across countries included in the survey. But also own contacts and names from
the homepage of the workgroup Democracy and Extremism were used. With 870 experts
completing the questionnaire in the selected countries (and, thus, rating 293 parties), the
overall response rate was 37.7% 37, which is not as high as the Expert Judgment Survey 2000
(with a 51.7% response rate), but still may be considered as reasonable when it is compared
for example with Huber’s and Inglehart’s (1995) expert survey: 41%. The fact that there are
more Eastern European countries included in the survey of 2010 may be the reason that
this response rate is significantly lower. Namely, the response is much lower for the Eastern
European countries than for the Western European countries.
Experts received three reminders with the question to complete the questionnaire. Some
experts have not completed the questionnaire of their own country, but another (nearby)
country. So there are two columns in Table 1.A1. The first column is the response rate of
the experts per country they belong to. The second column beholds the total number of
questionnaires completed per country. Because of the foregoing reason, the first and second
column differs. Some countries with high response are Belgium (54%), Netherlands (60%),
Norway (58%), and Portugal (70%). Ukraine had also a high response rate (50%), but only
two experts are approached, so the absolute response is low (1). The countries with a low
response are Czech Republic, Greece, Israel (all 20%), Macedonia (0%), Poland (14%), Russia (7%). Especially the response rate in Israel and Russia is disappointing. Israel: because
there are 106 experts approached and only 21 of them completed the questionnaire. Russia:
because only 1 expert of the 15 completed the questionnaire. The fact that Macedonia has
a response rate of 0% is not that surprising, because only 1 expert is approached and he or
she did not complete the questionnaire. As a final note about the response, the parties from
Northern Ireland are included in the questionnaire of the United Kingdom.38
Reliability Analyses
To test the reliability of the Expert Survey Data, I use the inter-coder reliability-method (see
Lubbers, 2001b). As Table 1.A2 shows, I generally find in every country high to extremely
high inter-coder reliability regarding most of the characteristics. Thus, I can be sure that
the experts in almost all countries have a similar idea about how the different parties in a
certain country should be rated regarding the party characteristics under study.
37 In addition, I deleted those experts (N=141) who rated the parties on fewer than 3 party characteristics.
Hence, the total number of experts in the dataset is 729.
38 The Expert Judgment Survey of European Political Parties 2010 is archived and freely available at Data
Archiving and Networked Services (DANS): https://easy.dans.knaw.nl/ui/datasets/id/easy-dataset:55725
170
Appendices
Table 1.A1 Response rate and absolute number of responses by country
Country
Responses/total
%
Albania
3/11
27%
2
Austria
23/56
41%
22
Belgium
38/70
54%
35
Bosnia-Herzegovina
4/12
33%
4
Bulgaria
8/35
23%
7
Croatia
13/35
37%
13
Cyprus
# of questionnaires
9/32
28%
8
Czech Republic
14/71
20%
14
Denmark
48/99
48%
47
Estonia
15/43
35%
16
Finland
53/131
40%
50
France
35/81
43%
37
Germany
32/98
33%
36
Greece
18/89
20%
20
Hungary
21/76
28%
19
8/15
53%
7
Iceland
Ireland
41/95
43%
40
Israel
21/106
20%
22
Italy
30/77
39%
36
Latvia
15/33
45%
15
Lithuania
37/86
43%
35
0/1
0%
0
Macedonia
Malta
5/8
63%
5
Netherlands
56/94
60%
42
Norway
46/80
58%
43
Poland
12/84
14%
14
Portugal
7/10
70%
6
Romania
24/76
32%
25
1/15
7%
3
Russia
Serbia
6/29
21%
8
Slovakia
12/67
18%
12
Slovenia
35/98
36%
34
Spain
27/70
39%
32
Sweden
53/114
46%
54
Switzerland
57/122
47%
54
Turkey
5/21
24%
6
Ukraine
1/2
50%
1
37/68
68%
38
United Kingdom
171
172
5
BE
BG
HR
CY
CZ
DK
EE
FI
FR
DE
GR
HU
IS
IE
IL
IT
LV
LI
MA
Croatia
Cyprus
Czech R.
Denmark
Estonia
Finland
France
Germany
Greece
Hungary
Iceland
Ireland
Israel
Italy
Latvia
Lithuania
Malta
38
NO
PL
PT
RO
Norway
Poland
Portugal
Romania
23
6
13
41
Netherlands NL
27
12
28
18
33
6
17
17
29
29
45
13
43
14
8
13
5
34
6
6
6
7
12
3
11
8
10
17
7
6
5
6
8
13
8
7
9
6
7
9
9
12
0,899
0,994
0,961
0,996
0,995
0,980
0,954
0,959
0,977
0,984
0,991
0,971
0,988
0,995
0,995
0,997
0,994
0,936
0,995
0,992
0,928
0,983
0,871
0,997
0,946
0,984
0,943
0,995
0,983
0,790
0,953
0,916
0,928
0,953
0,981
0,954
0,971
0,980
0,973
0,988
0,991
0,978
0,996
0,977
0,971
0,939
0,977
0,984
0,995
0,994
0,957
0,936
0,935
0,930
0,944
0,971
0,941
0,978
0,980
0,989
0,992
0,990
0,935
0,995
0,945
0,907
0,983
0,846
0,899
0,996
0,991
0,994
Imm. strictness
0,691
0,981
Bulgaria
0,995
Belgium
5
AT
Austria
21
Abb. Experts Parties Left-right Ec. Left-right
Country
Table 1.A2 Intercoder-Reliability Analyses
0,968
0,779
0,935
0,987
0,994
0,920
0,964
0,717
0,961
0,880
*
0,792
0,969
0,991
0,992
0,930
0,981
0,941
0,978
0,990
0,991
0,929
0,963
0,901
0,671
0,924
0,967
0,955
0,962
0,873
0,989
0,964
0,989
0,976
0,992
0,949
0,990
0,977
0,871
0,957
0,712
0,993
0,992
0,799
0,888
0,934
0,871
0,666
0,886
0,981
0,982
0,991
0,985
0,990
0,870
0,995
0,906
0,733
0,911
*
0,994
0,993
0,956
0,874
0,965
0,957
0,984
0,752
0,977
0,969
0,728
0,977
0,753
0,988
0,984
0,925
0,860
0,922
0,980
0,987
0,965
0,956
0,799
0,860
0,918
0,974
0,961
0,922
0,974
0,988
0,968
0,987
0,753
0,974
0,978
0,950
0,894
0,965
0,987
0,954
0,973
0,767
0,891
0,933
0,945
0,968
0,982
0,951
0,982
0,949
0,976
0,959
0,970
0,942
0,813
0,602
0,979
0,923
0,643
0,975
0,951
0,908
0,752
0,986
0,963
0,733
0,937
0,560
0,957
0,931
0,961
0,905
0,835
0,912
0,748
0,955
0,857
0,955
0,853
0,989
0,932
0,921
0,954
0,981
0,970
0,980
0,923
0,808
0,892
0,643
0,978
0,951
0,840
0,947
0,974
*
0,917
0,684
0,863
0,843
0,766
0,850
0,963
0,942
0,880
0,914
0,905
0,907
0,963
0,903
0,652
0,712
0,644
0,945
0,871
Populism Authoritarianism Nationalism Anti-establishment Organization Consensus Charisma
Appendices
40
41
6
SI
ES
SE
GB
Too few cases
Alpha below .7
Alpha increased after removal of some coders
Spain
Sweden
Switzerland CH
TR
Slovenia
Turkey
UK
Average
*
.592
.592
729
29
29
29
10
293
13
14
14
8
13
9
7
0,965
0,983
0,904
0,996
0,993
0,978
0,972
0,904
0,944
0,963
0,833
0,992
0,991
0,930
0,870
0,921
*
SK
0,797
Slovak Rep.
12
SB
Serbia
7
Abb. Experts Parties Left-right Ec. Left-right
Country
*
0,915
0,905
0,881
0,210
0,981
0,974
0,976
0,892
0,952
0,867
*
0,907
0,962
0,744
0,989
0,987
0,950
0,912
0,746
0,742
0,950
0,974
*
0,992
0,983
0,951
0,989
0,889
0,855
0,894
0,923
0,665
0,986
0,988
0,969
0,896
0,793
0,425
0,928
0,960
*
0,956
0,938
0,863
0,972
0,924
0,873
0,841
0,717
*
0,794
0,906
0,776
0,866
0,781
*
0,884
0,932
0,271
0,934
0,957
0,907
0,969
0,871
0,822
Populism Authoritarianism Nationalism Anti-establishment Organization Consensus Charisma
0,994
0,991
0,976
0,989
0,635
0,378
Imm. strictness
Table 1.A2 Intercoder-Reliability Analyses (continued)
Appendices
173
Appendices
Table 2.A1 Oblimin rotated factor solution of ideology and style dimension (N=215 parties)
Factor 1 (Ideology) loadings Factor 2 (Style) loadings
Communality
Immigration
.93
.01
.87
Nationalism
.84
.19
.86
Law & Order
.94
-.16
.80
Anti-establishment
-.04
.73
.51
Populism
.51
.53
.75
Source: Expert Judgment Survey of European Political Parties 2010 (Immerzeel et al., 2011).
Table 2.A2 List of 34 PRRs in the 24 European countries and 2 Belgian regions Flanders and Wallonia
Country
Party
(Abbrev.)
Name
Name (in English)
Austria
FPÖ
Freiheitliche Partei Österreichs
Freedom Party of Austria
BZÖ
Bundnis Zukunft österreichs
The future of Austria
Belgium (Flanders)
VB
Vlaams Belang
Flemish Interest
Belgium (Wallonia) FN
Front National
National Front
Bulgaria
ATAK
Natsionalen Săyuz Ataka
National Union Attack
Croatia
HSP
Hrvatska stranka prava
Croatian Party of Rights
Denmark
DFp
Dansk Folkeparti
Danish People’s Party
France
FN
Die Republikaner
The Republicans
National Democratic Party of Germany
MIEP
MIÉP-Jobbik a Harmadik Út
pártszövetség
MIÉP-Jobbik Third Way Alliance of
Parties
LAOS
Laikós Orthódoxos Synagermós
The People’s Orthodox Rally
HaIhud HaLeumi
La Destra–Fiamma Tricolore
Tvarka ir teisingumas
Lega Nord
Alternativ Demokratesch Reform
Front
National
Dansk
Folkeparti
Die
Republikaner
Eesti
Iseseisvuspartei
Nationaldemokratische
Partei Deu
Perussuomalaiset
Laikós
Front Orthódoxos
National Synagermós
Jobbik
Die Republikaner
- Nationaldemokratische Partei D
- Laikós Orthódoxos Synagermós
Yisrael
JobbikBeiteinu
HaIhud HaLeumi
La Destra–Fiamma Tricolore
Yisrael Beiteinu
Lega Nord
FN
DFp
REP
EIP
NPD
PS
LAOS
FN
Jobbik
REP
- NPD
- LAOS
‫ ל‬Jobbik
‫האיחוד הלאומי‬
FN
‫ל‬
LN
Greece
Israel
National Front
Nationaldemokratische Partei
Deutschlands
NPD
‫האיחוד הלאומי‬
FN
TT
LN
ADR
Hungary
Front National
Ireland
Greece
Israel
Hungary
Iceland
Ireland
Italy
Israel
REP
Estonian Independence Party
True Finns
Latvia
Italy
Lithuania
Luxembourg
Partij
- voor de Vrijheid
Lijst Pim Fortuyn
Tvarka
ir
teisingumas
Trots
op Nederland
Alternativ
Demokratesch Refor
Partij voor de Vrijheid
Lijst Pim Fortuyn
Trots op Nederland
Germany
Eesti Iseseisvuspartei
Perussuomalaiset
Netherlands
Latvia
Lithuania
Luxembourg
Netherlands
EIP
PS
PVV
LPF
TT
ToN
ADR
PVV
LPF
ToN
Estonia
Finland
Yisrael Beiteinu
Israel Our Home
HaIhud HaLeumi
National Union
Italy
LN
Lega Nord
Northern League
Netherlands
PVV
Partij voor de Vrijheid
Party for Freedom
TON
Trots op Nederland
Proud of the Netherlands
Czech
Bulgaria
Republic
Croatia
Denmark
Cyprus
Estonia
Czech
Finland
Republic
France
Denmark
Germany
Estonia
Finland
Greece
France
Hungary
Germany
Iceland
Norway
FrP
Fremskrittspartiet
Progress Party
Poland
LPR
Liga Polskich Rodzin
League of Polish Families
Portugal
PNR
Partido Nacional Renovador
National Renovator Party
Slovenia
174
Partidul România Mare
Greater Romania Party
Srpska radikalna stranka
Serbian Radical Party
SNS
Slovenská národná strana
Slovak National Party
SNS
Slovenska nacionalna stranka
Slovenian National Party
Belgium
Country
Austria
Bulgaria
Croatia
Belgium
Cyprus
Bundnis Zukunft Österreichs
Vlaams Belang (Flanders)
Name
Front
National Partei
(Wallonia)
Freiheitliche
Österreichs
Natsionalen
Sǎyuz Österreichs
Ataka
Bundnis Zukunft
Hrvatska
Prava
Vlaams Stranka
Belang (Flanders)
- Front National (Wallonia)
- Natsionalen Sǎyuz Ataka
Hrvatska Stranka Prava
Dansk Folkeparti
Eesti Iseseisvuspartei
Perussuomalaiset
Slovakia
PRM
SRS
BZÖ
VB
Party (Abbr.)
FN
FPÖ
ATAK
BZÖ
HSP
VB
- FN
- ATAK
HSP
DFp
EIP
PS
Romania
Serbia
Appendices
Table 2.A2 List of 34 PRRs in the 24 European countries and 2 Belgian regions Flanders and Wallonia
(continued)
Country
Party
(Abbrev.)
Lipa
Lipa
The Lime Tree Party (The Linden)
Sweden
SD
Sverigedemokraterna
Sweden Democrats
Switzerland
SVP
(Ge)Schweizerische Volkspartei
Swiss People’s Party
Name
Name (in English)
(Fr) Union Démocratique du Centre
(Ita) Unione Democratica di Centro
LdT
(Ita) Lega dei Ticinesi
Ticino League
Dem
(Ge) Schweizer Demokraten
Swiss Democrats
(Fr) Démocrates Suisses
(Ita) Democratici Svizzeri
Turkey
MHP
Milliyetçi Hareket Partisi
Nationalist Movement Party
United Kingdom
BNP
British National Party
British National Party
UKIP
United Kingdom Independence Party
United Kingdom Independence Party
Source: Expert Judgment Survey of European Political Parties 2010 (Immerzeel et al., 2011).
175
176
Atak
HSP
DFP
EIP
PS
FN
REP, NPD
LAOS
MIEPJOBBIK
YBeiteinu,
NU
LN
PVV, TON
FrP
LPR
PNR
5 Denmark
6 Estonia
7 Finland
8 France
9 Germany
10 Greece
11 Hungary
12 Israel
13 Italy
14 Netherlands
15 Norway
16 Poland
17 Portugal
FN
Belgium
(Wallonia)
4 Croatia
VB
2 Belgium
(Flanders)
3 Bulgaria
FPÖ, BZÖ
CPNT, MPF
HDSSB
N-VA, LDD
Populist
NeoRadical Right conservative
1 Austria
Party families:
Country:
PIS, SRP
SP, HØYRE
PDL
Likud
MDF
UMP
KOK
RL
KFP
RZS
Conservative
MN
V
D66, VVD
IdV
Kadima, AY
FDP
SFP, KESK
KESK, REF
V, RV, NyA
HNS, HSLS,
IDS
DPS-HÖH
MR
Open-VLD
Liberal
CDS-PP, PPD-PSD
PO, PSL
KFP
CDA, CU, SGP
MA, SVP, UdC
Shas, UTJ
FIDESZ-KDNP
NDSO
CDU-CSU
UDF
KD
RESP
KRD
HDZ, HSS
GERB, CK
cdH
CD&V
ÖVP
Christian
Democrat
B90Grünen
Les Verts
VIHR
EER
SFP
Ecolo
Groen!
Die
Grünen
Green
PS
LID
DNA
PVDA
PD, PSI
GL, PVDD
Avoda, Meretz Meimad,
HY
MSZP
PASOK
SPD
PS
SDP
SDE
SD
SDP
KB
PS
SP.A
SPÖ
Social
Democrat
Table 2.A3 List of all parties in the 24 European countries and 2 Belgian regions (Flanders and Wallonia)
BE, CDU
SV
SP
SA
Hadash
KKE, SYRIZA
Die Linke
LCR, LO, PC,
PRG
VAS
Ø
Communist
Socialist
Balad, Gil, JH, RT
HSU
Other
Appendices
SRS
SNS
SNS, LIPA
SD
SVP, SD, LdT
MHP
BNP, UKIP
19 Serbia
20 Slovakia
21 Slovenia
22 Sweden
23 Switzerland
24 Turkey
25 United Kingdom
11
SNP, EDP
GENÇ
SSN
L’S-HZDS
PNGCD
19
Tories
DP
M
SDS
SMKMostHID
PSS
UDMR
Conservative
32
LibDem
FDP, LPS
C, FL
LDS, ZARES
SaS
LDP, RP
PNL
Liberal
38
AKP
CSP, CVP, EDU,
EVP
KD
SLS-SMS
SDKÚ-DS, KDH
DSS-NS
PD-L
Christian
Democrat
Source: Expert Judgment Survey of European Political Parties 2010 (Immerzeel et al., 2011).
# Parties (N=215) 34
PRM
NeoRadical Right conservative
Populist
18 Romania
Country:
Party families:
31
Labour, SDLP
CHP, HYP
SPS
SAP
SD
SMER-SD
ZES, Romani
P
PSD+PC
Social
Democrat
16
Plaid
Cymru
GPS, GLP
MP
Green
Table 2.A3 List of all parties in the 24 European countries and 2 Belgian regions (Flanders and Wallonia) (continued)
21
Sinn Féin
IÇÇI, TKP
PDA
V
SPS-PUPS-US
Communist
Socialist
13
DUnionistP
SAADET, BTP,
ATP
DESUS
MK, LzS, NSS
Other
Appendices
177
Appendices
Table 2.A4 Average distances between PRRs and other parties’ positions in Eastern and Western European countries
Party family
Immigration
Nationalism
Authoritarianism
Populism
Anti-establishment
West East Sig. West East Sig. West East Sig. West East Sig. West
Sig.
2.01 1.18
.99 1.44
1.53 1.05
1.14
Conservative
2.46 2.87
2.76 2.91
1.45 1.50
2.68 1.66
4.82
1.68 **
Christian democrat
3.24 2.69
3.30 3.25
2.68
3.91 1.76 **
4.58
2.52 *
Liberal
4.10 4.75
5.04 5.44
3.01 1.49 **
3.82 3.19
4.45
2.75 *
Social democrat
4.90 4.05
5.15 4.53
3.67 2.15 *
3.88 2.36 **
4.56
2.63 *
Greens
6.33 4.45
6.49 5.83
5.24 3.05
4.81 1.33 *
2.80
2.17
Socialist/communist
6.92 2.55 ***
5.60 3.39 ~
5.73 3.04 *
3.24 2.44
Other
3.05 2.85 2.78 3.13 2.17 2.38 1.71 2.31 .67 ***
.24 *
East
Neo-conservative
Source: Expert Judgment Survey of European Political Parties 2010 (Immerzeel et al., 2011).
Note: Sign. ~p< .10, *p<.05, **p<.01, ***p<.001 (two-tailed).
178
1.54 -1.12 *
.61 -1.65
-.13
-.65 Appendices
Table 3.A1 PRRs in the Comparative Political Data Studies (period 2002-2012)
Country
PRR1
PRR2
PRR3
AT
FPÖ
BZÖ
-
BE
FN
VB
-
BG
Ataka
-
-
CH
SVP
Ticino League
SD
CY
ELAM
-
-
CZ
-
-
-
DE
Republikaner
NPD
-
DK
DFp
-
-
EE
EIP
-
-
ES
-
-
-
FI
PS
-
-
FR
FN
-
-
UK
UKIP
-
-
GR
LAOS
-
-
GB
UKIP
-
-
HR
HSP
HDSSB
-
HU
MIÉP
Jobbik
-
IE
-
-
-
IS
-
-
-
IT
LN
-
-
LT
TT
-
-
LUX
ADR
Lux for the Lux
-
LV
-
-
-
NL
LPF
PVV
-
NO
FrP
-
-
PL
LPR
-
-
PT
PNR
-
-
RO
PRM
PURM
-
RU
LDPR
-
-
SE
SD
-
-
SI
SNS
-
-
SK
SNS
PSNS
-
TR
MHP
-
-
XK
SNS
-
-
Sources: Mudde, 2007; Norris, 2005; Norwegian Social Science Data Services (NSD), 2014.
179
Appendices
Table 3.A2 Descriptive statistics all variables
Variable
Mean
S.D.
Range
Dependent variable
Voting (1=yes)
.78
0/1
Independent variables
Education
Primary (ref.)
.14
0/1
Lower Secondary
.18
0/1
Upper Secondary
.38
0/1
Post-secondary/Non-tertiary
.05
0/1
University (Tertiary)
.26
0/1
Unemployed (actively) (1=yes)
.05
0/1
Blue collar-worker (1=yes)
.17
Anti-immigrant attitudes
.03
2.02
-6.96-6.57
Political dissatisfaction
.03
2.00
-7.67-6.99
Political interest
.05
.86
-1.90-2.12
Young (1=yes)
.16
0/1
.54
0/1
0/1
Control variables
Female (1=yes)
Marital status
Partner (ref.)
.59
0/1
Ever partner
.19
0/1
Never partner
.22
0/1
Urbanization
Farm/countryside (ref.)
.06
0/1
Country village
.31
0/1
Town/small city
.30
0/1
Suburbs/outskirts
.12
0/1
Big city
.21
0/1
Religious attendance
Never (ref.)
.32
0/1
Sometimes
.41
0/1
Often
.27
0/1
Country*Wave level
PRR success
N1 (Individuals)
N2 (Countries)
N3 (Waves)
N4 (Country*Waves)
Note: Unweighted results of pooled data.
180
6.84
230,504
33
6
146
7.48
0-28.90
-1.252
(0.098)
16,696
6,022
0.062
N1 (Observations)
N2 (Respondents)
Pseudo R2
**
**
**
**
(0.068)
0.444
(0.087)
-0.398
(0.087)
-0.556
(0.076)
-0.496
(0.060)
-0.476
-
(0.203)
-1.382
**
**
**
**
**
**
(0.071)
0.223
(0.082)
0.364
(0.085)
-0.030
(0.076)
-0.190
(0.067)
-0.587
-
(0.158)
-0.562
**
**
*
**
**
PvdA
vs. PVV/ToN
(0.079)
-0.205
(0.093)
0.490
(0.091)
0.986
(0.084)
0.732
(0.075)
-0.086
-
(0.163)
-1.000
**
**
**
**
**
VVD
vs. PVV/ToN
(0.077)
-0.125
(0.091)
0.402
(0.089)
0.587
(0.085)
-0.048
(0.075)
-0.380
-
(0.157)
-0.535
**
**
**
**
SP
vs. PVV/ToN
(0.092)
-0.719
(0.140)
-0.808
(0.111)
0.076
(0.098)
0.277
(0.085)
-0.034
-
(0.280)
-1.219
**
**
**
**
GL
vs. PVV/ToN
(0.079)
-0.212
(0.094)
0.339
(0.096)
0.143
(0.090)
-0.161
(0.072)
0.224
-
(0.197)
-1.151
**
**
**
**
(0.090)
-0.607
(0.098)
0.183
(0.099)
0.012
(0.089)
-0.211
(0.068)
-0.280
-
(0.334)
-1.843
**
*
**
**
D66
CU/SGP
vs. PVV/ToN vs. PVV/ToN
(0.136)
-1.742
(0.146)
1.611
(0.154)
0.344
(0.156)
-0.196
(0.127)
-0.008
-
(0.212)
-0.234
**
**
*
Other vs.
PVV/ToN
(0.144)
-2.050
(0.168)
0.875
(0.173)
0.708
(0.172)
0.066
(0.170)
0.016
-
(0.146)
1.627
**
**
**
**
Blank
vs. PVV/ToN
Notes: *.01<p<.05, **p<.01.
These results are based upon the pooled data of couples of waves. These data comprise all respondents that provided information about their voting intention in two subsequent waves. Respondents that have answered their voting in for example three subsequent waves (e.g. waves 1, 2, and 3) pop up twice in this dataset (i.e. in wave-couple 1 &
2, and in wave couple 2 & 3). The robust standard errors control for the nesting of the different wave-couples in respondents.
a
The results show that non-voters in the first wave are generally most likely to non-vote in the next wave. However, if they vote, they are most likely to turn to the PVV/ToN:
all b-coefficients for the comparison between a certain party and the PVV/ToN are significantly negative, indicating that the vote for PVV/ToN is more likely than the vote
for the corresponding party.
(0.123)
0.186
(0.117)
0.544
(0.123)
-0.548
(0.118)
Constant
5&6
4&5
3&4
0.019
2&3
1 & 2 (ref.)
-
(0.117)
in 1st wave
Waves
3.265
Non-vote (vs. vote)
Not-vote
vs. PVV/ToN
CDA
vs. PVV/ToN
Voting Behavior in 2nd Wave
Table 4.A1 Results of the test that non-voters are most likely to turn to PVV/ToN rather than another party (robust standard errors in parentheses) a
Appendices
181
Appendices
Table 4.A2 Results of the pooled logistic regression model of voting behavior (1=vote, 0=non-vote)
Independent variables:
b
Robust S.E.
Political efficacy (1=high)
0.279 **
0.102
Political interest (1=yes)
0.443 **
0.101
Age at 1st wave participated a
0.005
0.004
Female (1=yes) a
0.119
0.091
Partner (1=yes)
0.178
0.107
Children (1=yes)
-0.127
0.103
-0.007
0.180
0.005
0.183
-0.355
0.207
0.162
0.221
Control variables
Education (primary=ref.)
Secondary
MBO
HBO/WO educated
Other
Employment status (employed=ref.)
Unemployed
-0.090
0.114
Student
-0.044
0.200
Retired
-0.080
0.174
-0.228
0.164
Religious attendance (1=regular churchgoer)
Wave (t2=ref.)
t3
0.365 *
t4
1.646 **
0.136
t5
0.115
0.168
t6
1.640 **
0.139
-2.670 **
0.301
Constant
N1
N2
Log-likelihood
0.158
2,861
1,077
-1,344 Source: LISS Panel (Tilburg University, 2007-2014).
Notes: *.01<p<.05, **p<.01.
a
In a pooled logistic regression model, time-constant variables such as age of participant at the first wave (s)he participated and gender can be normally included.
182
Norway
NO
FrP
PVV, ToN
Source: European Values Study (2010).
Netherlands
NL
Lega Nord, Fiamma Tric.
PS (True Finns)
Finland
Italy
FI
IT
SVP, SDem, Ticino, FDU
Switzerland
CH FPÖ, BZÖ
Austria
Belgium-Flanders Vlaams Belang
AT
Populist Radical Right
Parties
Party family
Country
BE
Conservative
12,36% KoK
10,15% 10,65% 22,11% Hoyre
126
191
22,18% Radicale
124
87
8,82% N-VA
79
18,85% -
Parties
61
%
Voters
154
#
-
171
-
-
190
86
- ÖVP
Parties
Christian
19,79% KRF
- CDA, CU
- UdC, PdL
26,99% SKP
15,38% CVP, CSP, EVP
12,28% CD&V
%
Voters
85
#
Table 5.A1 The political parties in the 7 countries under study (based on the total # of voters per country)
44
353
282
16
89
187
Parties
Orthodox
5,09% -
29,84% SGP
36,25% -
2,27% -
15,92% -
27,02% -
23,99% -
%
Voters
196
#
-
-
-
-
-
-
-
-
-
-
-
-
1,78%
%
Voters
21
#
864
1183
778
704
559
692
817
Votes
Total
Appendices
183
Appendices
Table 5.A2 Factor loadings of the anti-immigrant variables on the latent construct
Anti-immigrant attitudes BY:
AT
BE
CH
FI
IT
NL
NO
Anti-immigrant 1
.773*
.719*
.683*
.777*
.593*
.637*
.671*
(Cultural life undermined)
(.044)
(.052)
(.056)
(.055)
(.050)
(.047)
(.060)
Anti-immigrant 2
.766*
.733*
.703*
.777*
.724*
.680*
.583*
(Crime problems worse)
(.056)
(.055)
(.062)
(.063)
(.061)
(.053)
(.067)
Anti-immigrant 3
.677*
.767*
.807*
.824*
.690*
.746*
.620*
(Strain welfare system)
(.061)
(.058)
(.058)
(.052)
(.057)
(.046)
(.071)
Anti-immigrant 4
.869*
.890*
.796*
.877*
.871*
.889*
.781*
(Future threat to society)
(.048)
(.054)
(.053)
(.043)
(.047)
(.036)
(.061)
Anti-immigrant 5
.759*
.754*
.665*
.781*
.671*
.648*
.700*
(Too many immigrants)
(.059)
(.058)
(.057)
(.055)
(.069)
(.047)
(.055)
261
238
247
216
331
393
257
N
Notes: *p<.05 (one-tailed).
Unstandardized regression coefficients (MLR-estimator).
1. Cultural life undermined:
‘A country’s cultural life is undermined by immigrants’
2. Crime problems worse:
‘Immigrants make crime problems worse’
3. Strain welfare system:
‘Immigrants are a strain on a country’s welfare system’
4. Future threat to society:
‘In the future the proportion of immigrants will become a threat to society’
5. Too many immigrants:
‘Today [in one’s country], there are too many immigrants’
Source: European Values Study (2010).
184
Appendices
Table 5.A3 Logistic regression of voting behavior
AT
BE
CH
-.274*
-.596*
.035
(.103)
(.171)
(.163)
FI
IT
NL
NO
-.155
.015
-.530*
-.730*
(.362)
(.127)
(.162)
(.292)
a
Attendance (not often=ref.) :
Often
Orthodoxy (tolerant belief=ref.)
Orthodox belief
-.261*
.259
-.132
.149
.075
-.030
.505*
(.094)
(.203)
(.183)
(.369)
(.083)
(.205)
(.297)
Female
-.238
-.312
-.138
.051
.432*
-.038
-.504*
(.166)
(.276)
(.201)
(.246)
(.138)
(.173)
(.205)
-.219
.073
.218
.637*
-.077
-.008
-.361
Occupation type (else=ref.)
Manual-production worker
(.213)
(.319)
(.220)
(.336)
(.227)
(.225)
(.399)
Education (in years)
-.018
.010
-.031
-.064*
-.037*
-.011
-.034*
(.023)
(.024)
(.019)
(.038)
(.016)
(.015)
(.019)
Age
-.015*
-.004
-.004
-.015*
-.017*
-.006
.000
(.005)
(.009)
(.006)
(.008)
(.006)
(.004)
(.006)
Anti-immigrant attitudes
.459*
.343*
.492*
.449*
.343*
.330*
.285*
(.075)
(.123)
(.104)
(.083)
(.086)
(.072)
(.117)
.089
-.276*
-.182*
-.029
-.148
.029
-.077
(.101)
(.155)
(.105)
(.118)
(.098)
(.096)
(.124)
Political interest
RMSEA
.045
.028
.030
.039
.027
.039
.033
CFI
.959
.974
.974
.962
.972
.952
.973
TLI
.929
.956
.957
.935
.954
.921
.953
N
261
238
247
216
331
393
257
Notes: *p<.05 (one-tailed).
Unstandardized regression coefficients (Weighted Least Square Mean Variance-adjusted estimator).
a
For parsimony reasons we dropped the never-attend-dummy from our path analysis in AT, FI, and NO.
185
Appendices
Table 5.A4 Multivariate regression of anti-immigrant attitudes as mediator variable
AT
BE
CH
FI
IT
NL
-.427*
-.205*
(.094)
(.116)
NO
-.591*
-.128
-.197*
-.339*
-.340
(.207)
(.296)
(.089)
(.146)
(.274)
Attendance (not often=ref.):
Often
Orthodoxy (tolerant belief=ref.)
Orthodox belief
Female
.072
.291*
.735*
.117
.075
.222
.360
(.095)
(.129)
(.229)
(.290)
(.083)
(.162)
(.265)
.119
.092
.538*
-.269
.432*
.195
-.292
(.164)
(.161)
(.239)
(.185)
(.138)
(.149)
(.180)
.280
.120
.384
.259
.377*
.158
.100
Occupation type (else=ref.)
Manual-production worker
(.230)
(.211)
(.255)
(.288)
(.175)
(.190)
(.302)
Education (in years)
-.078*
.001
-.042*
-.025
-.026*
-.012
-.045*
(.022)
(.015)
(.022)
(.030)
(.011)
(.011)
(.017)
Age
.013*
.011*
.003
-.008
-.005
.005
.004
(.004)
(.005)
(.006)
(.006)
(.004)
(.004)
(.005)
Political interesta
-.384*
-.140*
.238*
-
.175*
-.159*
-.149
(.111)
(.095)
(.078)
(.118)
-
(.074)
(.075)
RMSEA
.045
.028
.030
.039
.027
.039
.033
CFI
.959
.974
.974
.962
.972
.952
.973
TLI
.929
.956
.957
.935
.954
.921
.953
N
261
238
247
216
331
393
257
Notes: *p<.05 (one-tailed).
Unstandardized regression coefficients (Weighted Least Square Mean Variance-adjusted estimator).
a
For parsimony reasons we dropped the political interest-variable as predictor-variable in FI.
186
Appendices
Table 5.A5 Correlations of dependent variables in the various countries
AT
Variables:
1
2
3
4
5
6
7
1. PRR-vote
x
x
x
x
x
x
x
2. Anti1
.379
x
x
x
x
x
x
3. Anti2
.457
.574
x
x
x
x
x
4. Anti3
.465
.372
.542
x
x
x
x
5. Anti4
.464
.666
.609
.501
x
x
x
6. Anti5
.459
.525
.508
.552
.608
x
x
7. Orthodoxy
-.333
-.026
.038
.000
-.054
-.040
x
8. Often-church
-.521
-.165
-.321
-.283
-.282
-.326
.224
BE
Variables:
1
2
3
4
5
6
7
8
1. PRR-vote
x
x
x
x
x
x
x
x
2. Anti1
.412
x
x
x
x
x
x
x
3. Anti2
.388
.533
x
x
x
x
x
x
4. Anti3
.269
.525
.594
x
x
x
x
x
5. Anti4
.510
.655
.644
.656
x
x
x
x
6. Anti5
.371
.498
.586
.562
.655
x
x
x
7. Orthodoxy
.111
.153
.171
.092
.085
.274
x
x
8. Never-church
.255
.063
.051
.018
.038
.061
-.155
x
9. Often-church
-.562
-.017
-.098
-.099
-.147
-.101
.299
-.698
CH
Variables:
1
2
3
4
5
6
7
8
1. PRR-vote
x
x
x
x
x
x
x
x
2. Anti1
.331
x
x
x
x
x
x
x
3. Anti2
.399
.438
x
x
x
x
x
x
4. Anti3
.456
.545
.615
x
x
x
x
x
5. Anti4
.387
.539
.508
.599
x
x
x
x
6. Anti5
.397
.445
.425
.433
.556
x
x
x
7. Orthodoxy
.091
.257
.324
.152
.338
.318
x
x
8. Never-church
.154
-.027
.056
.083
.009
.050
-.164
x
9. Often-church
-.161
-.101
-.158
-.178
-.136
-.258
.415
-.711
Variables:
1
2
3
4
5
6
7
1. PRR-vote
x
x
x
x
x
x
x
2. Anti1
.372
x
x
x
x
x
x
3. Anti2
.337
.533
x
x
x
x
x
4. Anti3
.403
.603
.735
x
x
x
x
5. Anti4
.337
.696
.670
.717
x
x
x
6. Anti5
.394
.637
.569
.605
.702
x
x
7. Orthodoxy
.029
.197
-.035
-.142
.055
-.016
x
8. Often-church
-.051
.060
-.043
-.146
.076
-.097
.807
FI
187
Appendices
IT
Variables:
1
2
3
4
5
6
7
8
1. PRR-vote
x
x
x
x
x
x
x
x
2. Anti1
.220
x
x
x
x
x
x
x
3. Anti2
.116
.375
x
x
x
x
x
x
4. Anti3
.298
.401
.459
x
x
x
x
x
5. Anti4
.362
.474
.645
.572
x
x
x
x
6. Anti5
.198
.401
.431
.485
.537
x
x
x
7. Orthodoxy
-.009
.146
-.085
.033
.007
-.040
x
x
8. Never-church
.036
-.031
-.027
.040
.007
-.062
-.238
x
9. Often-church
-.046
-.003
-.175
-.076
-.201
-.100
.338
-.445
NL
Variables:
1
2
3
4
5
6
7
8
1. PRR-vote
x
x
x
x
x
x
x
x
2. Anti1
.310
x
x
x
x
x
x
x
3. Anti2
.275
.393
x
x
x
x
x
x
4. Anti3
.324
.429
.612
x
x
x
x
x
5. Anti4
.379
.578
.585
.671
x
x
x
x
6. Anti5
.424
.461
.425
.429
.578
x
x
x
7. Orthodoxy
-.460
.081
-.017
-.099
-.009
-.061
x
x
8. Never-church
.328
.021
.015
.066
.070
.100
-.335
x
9. Often-church
-.616
-.106
-.125
-.163
-.163
-.157
.604
-.682
Variables:
1
2
3
4
5
6
7
1. PRR-vote
x
x
x
x
x
x
x
2. Anti1
.274
x
x
x
x
x
x
3. Anti2
.171
.312
x
x
x
x
x
4. Anti3
.126
.340
.536
x
x
x
x
5. Anti4
.318
.523
.390
.461
x
x
x
6. Anti5
.325
.469
.345
.392
.521
x
x
7. Orthodoxy
-.028
.050
.073
.100
.094
.002
x
8. Often-church
-.359
.091
-.142
-.014
-.035
-.105
.769
NO
Source: European Values Study (2010).
188
Appendices
Table 6.A1 Overview reliability scale and country data
PRR
Abs.
image
score
Rel.
image
score
Abs.
Succ.
Rel.
Succ.
Abs.
Popul.
Style
Rel.
Popul.
Style
N
Anti-immigrant
(α)
Austria
752
.92
FPÖ, BZÖ
6.19
2.89
15.15
.43
8.75
3.09
Belgium
1292
.88
Vlaams Belang
8.67
5.66
13.96
.75
9.35
4.87
Denmark
1076
.83
Danske FP
5.34
2.78
13.90
.53
8.50
3.23
Finland
577
.90
True Finns
6.82
4.27
4.05
.18
8.88
3.98
France
1147
.90
Front National
8.98
6.39
11.52
.15
9.52
4.79
Germany
1032
.87
Die Republikaner
8.32
6.23
2.20
.06
8.77
3.34
Greece
861
.85
LAOS
3.81
2.22
3.80
.09
9.35
2.48
Italy
667
.89
Lega Nord
7.29
4.10
5.19
.17
9.22
2.91
Luxembourg
712
.83
ADR
-
-
9.95
.28
-
-
Netherlands
1123
.87
PVV
8.47
6.30
5.90
.22
9.28
4.60
Norway
860
.83
Progress Party
6.91
3.53
22.10
.68
8.62
4.44
Switzerland
544
.87
SVP
6.34
3.92
29.40
1.51
9.23
10643
.87
Country
Total
5.18
Sources: European Values Study (2010); Expert Judgment Survey of European Political Parties 2010 (Immerzeel et
al. 2011).
189
Dutch summary
(Nederlandse samenvatting)
191
Nederlandse samenvatting
Stemmen voor de verandering
De democratische verlokking van populistisch radicaal rechts in stemgedrag
Het was slechts vijf dagen na zijn dood dat duizenden mensen in Rotterdam de straat
opgingen om de laatste eer te bewijzen aan Pim Fortuyn. Met de massale opkomst van
aanhangers en het grote publiek dat de processie van Fortuyn door Rotterdam rechtstreeks
volgde via televisie, werd eens te meer duidelijk hoeveel zijn in- en uittrede in de Nederlandse politiek maatschappelijk teweeg bracht. De verkiezingsoverwinning op 16 mei 2002
voor Lijst Pim Fortuyn was gigantisch en ging de boeken in als de meest volatiele verkiezing
in de moderne, naoorlogse Nederlandse geschiedenis (Mair, 2006). Fortuyn’s erfgenamen
wonnen met 26 zetels 17% van de 150 zetels en de verkiezingsopkomst schoot omhoog
van 72,6 naar 79 procent. Er werden in 2002 850.000 stemmen meer uitgebracht dan bij de
verkiezingen in 1998.
De opmerkelijke opkomst en populariteit van de rechts-populist Fortuyn is vaak bestempeld als een keerpunt in de Nederlandse politieke geschiedenis. De ‘Fortuyn-revolte’
was een aanleiding tot veel onderzoek naar de verklaring voor de opkomst en toenemende
populariteit van deze rechts-populist. In andere Oost- en West-Europese landen hebben
zich de afgelopen 30 jaar vergelijkbare doorbraken van populistisch radicaal rechts (PRR)
voorgedaan. In de jaren ’80 en ‘90 van de vorige eeuw werden bijvoorbeeld het Franse
Front National en het Belgische Vlaams Blok zeer populair. Maar ook recentelijk kon men
de United Kingdom Independence Party en de Finse Ware Finnen aan populariteit zien
winnen. De Europese PRR-familie is een partij familie waar rekening mee gehouden dient
te worden (Mudde, 2012; Zaslove, 2004).
Dat de doorbraak van deze nieuwe partij familie veel politieke en maatschappelijke onrust teweegbrengt, heeft te maken met de radicale visie van PRR op de samenleving. Deze
radicale visie steunt grotendeels op drie ideologische kenmerken: nativisme, autoritarisme
en populisme (Mudde, 2007). Het nativisme wordt gekenmerkt door een sterke voorkeur
voor het eigen volk, de natie, en een sterke afkeer van al datgene wat niet behoort tot het
eigen volk. Andere ‘niet-natieve’ elementen, zoals immigranten met andere ideeën en religie, die zich onder het eigen volk begeven vormen een fundamentele bedreiging voor het
eigen volk. Om die bedreiging tegen te gaan, zouden volgens nativisten deze niet-natieve
elementen moet vertrekken naar het deel van de wereld waar ze wel ‘thuis horen’.
Naast deze nativistische ideologie delen PRR-partijen een sterke, autoritaire nadruk op
de naleving van de wet. PRR-leiders vinden dat het samenleven rustig en ordelijk moet
verlopen en stellen daarom vaak strengere maatregelen voor om criminaliteit te bestraffen.
Als laatste worden PRR-partijen gezien als populistisch. Hierbij wordt gerefereerd aan Jagers (2006) die populisme in beginsel ziet als een ‘dunne’ ideologie die de samenleving in
twee groepen onderverdeelt: de onbetrouwbare ‘corrupte elite’ en het betrouwbare, gewone
193
Nederlandse samenvatting
volk dat begiftigd is met een pure en eerlijke volkswil. PRR identificeert zich graag met het
pure volk en ziet de andere gevestigde partijen als de corrupte elite die de stem van het volk
niet goed vertegenwoordigt.
Met de opkomst en stabiele populariteit van partijen die deze drie ideologische kenmerken aan elkaar verbinden, is er veel maatschappelijke en wetenschappelijk discussie
over de vraag of populistisch radicaal rechts een bedreiging vormt voor de Westerse liberale
democratieën en rechtsstaten. Staat de xenofobe, negatieve houding jegens immigrant, elite
en Europese Unie niet op gespannen voet met diversiteit, pluralisme en de legitimiteit van
de liberale democratie? Kan de sterkere nadruk op de wil van het, vooral homogene, nationale volk niet leiden tot marginalisering en onderdrukking van minderheden?
Er zijn echter ook mensen die stellen dat de opkomst en populariteit van PRR nodig is
om de kwaliteit van de democratie te waarborgen. Deze partijen kunnen als ‘hoeders van de
democratie’ worden gezien. Door hun nadruk op de ‘stem van het volk’, hun toegankelijke
taalgebruik en het benoemen van pijnlijke onderwerpen zou het ze lukken mensen weer te
betrekken bij politiek. Hun radicale visie doorbreekt taboes over problemen die samenhangen met de hedendaagse, multiculturele samenleving, waardoor burgers zich weer serieus
genomen voelen. Dit zou kunnen leiden tot meer betrokkenheid bij politiek en democratie
en daarmee een versterking van de legitimiteit van de democratie (vgl. Mouffe, 2005).
Deze laatste verwachting vormt het uitgangspunt van deze dissertatie. Dit boek geeft een
theoretische en empirische onderbouwing van de publieke en wetenschappelijke discussie
over de gevolgen van PRR op het stemgedrag van Europese burgers. Geïnspireerd door Jansen’s (2012) notie van populisme als een politieke praktijk die populistische retoriek bezigt
om gemarginaliseerde sociale groepen bij de politiek te betrekken, wordt in deze dissertatie
onderzocht in hoeverre PRR het stemgedrag van burgers in Oost- en West-Europa beïnvloedt. Meer specifiek wordt in dit boek onderzocht in hoeverre PRR de niet-stemmers naar
de stembus weet te lokken. Daarnaast wordt bekeken in hoeverre PRR in staat is mensen
naar zich toe te trekken die er om bekend staan dat ze vaak op dezelfde partij stemmen, de
‘scheidslijn’-stemmers.
Met het beantwoorden van deze twee vragen, wordt er meer empirische basis geboden
aan het maatschappelijke en wetenschappelijk debat over de gevolgen van PRR op huidige
Europese liberale democratieën. Het geeft meer duidelijkheid over de vraag of PRR-partijen
in staat zijn om politiek weer aantrekkelijker te maken en mensen erbij te betrekken.
Bovendien geeft het inzicht in de mate waarin traditionele scheidslijnen in de maatschappij
verbroken worden door deze nieuwe partij familie.
Om de twee deelvragen te beantwoorden is het proefschrift opgedeeld in vijf empirische
hoofdstukken. Hierna zullen de verschillende hoofdstukken uit dit proefschrift kort samengevat worden. De aanleiding, de hoofdvraag, en de belangrijkste resultaten worden
per hoofdstuk bespreken. Vervolgens zullen de conclusies worden samengevat en wordt
194
Nederlandse samenvatting
antwoord gegeven op de hoofdvraag in hoeverre en waarom PRR het stemgedrag van
Europese burgers beïnvloedt.
Populistisch radicaal rechts en de concurrenten
Om de onderzoeksvragen over de invloed van PRR op het stemgedrag van niet-stemmers
en ‘scheidslijn’-stemmers beter te kunnen beantwoorden, werd in het eerste empirische
hoofdstuk, Hoofdstuk 2, onderzocht hoe uniek het partijprogramma en de politieke stijl
van de PRR-partijen is. Op basis van de ‘Expert Judgment Survey of European Political
Parties 2010’ (Immerzeel et al., 2011), gehouden onder partij-experts uit meer dan 30 Europese landen, werd onderzocht hoe PRR zich verhoudt tot andere partijen op ideologische
(immigratie, nationalisme en autoritarisme) en stijl-gerelateerde onderwerpen (populisme
en anti-establishment retoriek).
De resultaten van dat hoofdstuk lieten inderdaad zien dat PRR een unieke ideologische positie inneemt in de Europese politieke arena’s. Op zowel de ideologische als de
stijl-gerelateerde kenmerken zijn er, gemiddeld genomen, behoorlijke verschillen in de
posities die partijen innemen ten opzichte van PRR. Neoconservatieve, conservatieve en
Christendemocratische partijen staan wat betreft ideologische kenmerken en populisme
het dichtst bij PRR-partijen, terwijl socialistische en Groene partijen het verst verwijderd
zijn van PRR op deze standpunten. Alleen wat betreft anti-establishment retoriek leken
de Groene en socialistische partijen meer op PRR dan de andere centrumpartijen. Verder
werd duidelijk dat Oost-Europese partijen over het algemeen dichter bij PRR stonden dan
de West-Europese partijen op hun standpunten jegens immigratie, striktheid in naleving
van de wet, populisme en anti-establishment retoriek. Echter, ook in Oost-Europa bleven
behoorlijke verschillen bestaan tussen PRR en de andere partijen en er werd daarom geconcludeerd dat PRR een uitzonderlijke positie inneemt in de Europese partijsystemen. Op
basis van deze resultaten was het aannemelijk dat PRR een unieke invloed heeft op het
stemgedrag van burgers (vgl. Mudde & Rovira Kaltwasser, 2012).
Populistisch radicaal rechts en niet-stemmers
Na de vaststelling dat PRR een unieke positie inneemt in de Europese politieke arena’s, werd
in het eerste deel van dit boek bekeken in hoeverre PRR in staat is om niet-stemmers naar
de stembus te lokken en zo een bijdrage levert aan hogere opkomstcijfers bij verkiezingen.
In Hoofdstuk 3 van dit proefschrift werd de relatie tussen het succes van PRR in een land
en de individuele bereidheid om te stemmen onderzocht. De onderzoeksvragen van dat
hoofdstuk waren: Wat is het effect van de populariteit van PRR op verkiezingsopkomst in
195
Nederlandse samenvatting
Europa? Wie wordt er aangemoedigd en wie ontmoedigd om te gaan stemmen door het
succes van PRR?
Vertrekkend vanuit de theorie van electorale competitie (Franklin, 2004) werd beargumenteerd dat de opkomst en populariteit van PRR leidt tot een sterkere competitie
om stemmen en daarmee de bereidheid van burgers om te gaan stemmen verhoogt. Kort
samengevat: het succes van PRR zou een positief effect kunnen hebben op de opkomstcijfers bij verkiezingen. Echter, het tweede hoofdstuk liet zien dat PRR bekendstaat om
een unieke, radicale vorm van politiek bedrijven (zowel ideologische als stijl-gerelateerd).
Verschillende wetenschappers betogen dat deze meer extreme, radicale vorm van politiek
burgers juist kan afkeren van politiek, omdat men het politieke bedrijf en de politici gaat
wantrouwen. Met andere woorden, de opkomst en populariteit van PRR zou een negatief
effect kunnen hebben op de opkomst bij verkiezingen. Om deze tegengestelde hypotheses
te toetsen voor zowel Oost- als West-Europese landen, werd in dit boek gebruik gemaakt
van de zes rondes van de European Social Survey (ESS, 2002-2012). Met behulp van
zogenoemde hiërarchische of multi-niveau modellen (bv. Hox, 2002) werd de relatie tussen
PRR-succes en individuele geneigdheid tot stemmen getoetst.
De resultaten uit Hoofdstuk 3 bevestigden de verwachting van de electorale competitie
niet. Burgers die in een land woonde waar PRR succesvoller was, waren niet meer geneigd
om naar de stembus te gaan dan burgers die in een land woonde met een minder populaire
PRR. In tegenstelling, in Oost-Europa werden aanwijzingen gevonden voor een negatieve
invloed van PRR: in Oost-Europese landen waar PRR succesvoller waren, leken burgers
minder geneigd om te gaan stemmen. In West-Europese landen werd zo’n negatief effect
niet gevonden en leken PRR-partijen geen invloed te hebben op de individuele beslissing
om te gaan stemmen. Ondanks anekdotisch bewijs voor een positief effect van PRR succes
op opkomstcijfers, werd in dit hoofdstuk geen universeel positief effect gevonden en bleek
de opkomst van PRR eerder geen effect te hebben op de individuele geneigdheid om te
stemmen.
Echter, verdere analyses op het individuele niveau lieten zien dat deze laatste observatie
niet te stellig moet worden geponeerd. Niet alleen werd er een licht negatief effect gevonden
in Oost-Europese landen, de resultaten lieten ook zien dat PRR wel degelijk het stemgedrag
van sommige specifieke groepen wist te beïnvloeden. Zo werd in West-Europese landen
gevonden dat mensen die meer geïnteresseerd zijn in politiek, meer geneigd zijn om te gaan
stemmen, terwijl laagopgeleiden en fabrieks- en handarbeiders juist minder geneigd zijn
om te gaan stemmen als PRR populairder is. Deze laatste bevinding was noemenswaardig,
aangezien eerder onderzoek heeft laten zien dat laagopgeleiden en fabrieks- en handarbeiders juist behoren tot het typische PRR-electoraat (bv. Lubbers et al., 2002). Het suggereerde dat West-Europese PRR-partijen succesvol werden door mensen die al stemden
voordat er een PRR was en hun stem verwisselden zodra PRR opkwam.
196
Nederlandse samenvatting
Een andere belangrijke bevinding uit dat hoofdstuk was dat het stemgedrag van OostEuropese jongeren sterk beïnvloed werd door de opkomst en succes van PRR. Hoewel in
Oost-Europa de hele bevolking minder geneigd is om naar de stembus te gaan als PRR
populair is, zijn het vooral de jongeren die thuisblijven. Dit was opmerkelijk, aangezien
West-Europese jongeren niet beïnvloed leken te worden door succesvolle PRR-partijen. Er
werd betoogd dat dit wellicht te maken heeft met de meer unieke positie die PRR-partijen
in West-Europa innemen, zoals Hoofdstuk 2 liet zien. West-Europese jongeren zouden de
radicale en minder beschaafde ideologie en stijl van PRR meer als uitzondering dan als
regel zien en daarom niet het vertrouwen in de gehele politiek en democratie verliezen.
Daarentegen zouden Oost-Europese jongeren de toenemende populariteit van PRR juist
als bevestiging zien van hun vermoeden dat het gehele politiek systeem niet deugt en zich
daarom afkeren van electorale politiek. Deze speculatie verdient echter meer aandacht in
vervolgonderzoek.
Hoewel Hoofdstuk 3 liet zien hoe verschillende delen van het electoraat op verschillende
wijze beïnvloed werden door de opkomst en populariteit van PRR, werd nog niet duidelijk
of diegene die meer geneigd waren om naar de stembus te gaan ook vaker zouden stemmen
op PRR. Het zou kunnen dat juist de mensen die zich tegen PRR afzetten meer geneigd
zijn om te gaan stemmen, omdat ze niet willen dat PRR te populair wordt (vgl. Andersen
& Evans, 2003). Om na te gaan of niet-stemmers daadwerkelijk meer geneigd zijn om PRR
te stemmen, werd in Hoofdstuk 4 het stemgedrag van een specifieke groep Nederlandse
niet-stemmers gevolgd voor een periode van 6 jaar (2007-2013). De onderzoeksvragen
van dat hoofdstuk waren: In hoeverre zijn Nederlandse (voormalige) niet-stemmers meer
geneigd te stemmen op een PRR-partij dan op een andere partij? Hoe valt zo’n eventuele
verandering in stemgedrag te verklaren?
Om deze onderzoeksvragen te beantwoorden werd er gebruik gemaakt van de ‘Longitudinal Internet Studies for the Social Sciences’ (LISS, Universiteit van Tilburg, 2007-2014).
Deze panelstudie maakte het mogelijk om de stemintentie van een groep Nederlandse
niet-stemmers te volgen over tijd. Uit de analyse van deze gegevens werden vier belangrijke
resultaten gevonden. Allereerst, ondanks het feit dat Nederland bekendstaat als een land
met veel zwevende kiezers en een zeer open, onvoorspelbaar politiek systeem (Mair, 2006b),
bleek dat de meerderheid van de niet-stemmers en stemmers stabiel is in haar stemgedrag.
Niet-stemmers bleven over het algemeen niet-stemmers en mensen die op een bepaalde
partij stemmen, zijn geneigd dat te blijven doen in latere verkiezingen. Ten tweede, als nietstemmers besloten te gaan stemmen, leek dat vooral te maken hebben met een verandering
in politieke interesse. Meer structurele veranderingen als het krijgen van een kind en het
verliezen van een baan leken niet samen te hangen met een verandering in stemgedrag. Ten
derde, als niet-stemmers besloten te stemmen bleken ze inderdaad vaker te kiezen voor
PRR dan voor elke andere partij afzonderlijk, wat in overeenstemming is met het argument
197
Nederlandse samenvatting
dat PRR ‘de hoeder is van de democratie’. Als laatste liet Hoofdstuk 4 zien dat het besluit als
niet-stemmer om later PRR te stemmen voornamelijk samenhangt met een verandering in
sympathie voor de leider van de PRR-partij. Veranderingen in ideologische overtuigingen
die passen bij PRR (zoals negatieve houdingen jegens immigranten, Europese Unie en
politiek in het algemeen) bleken niet samen te hangen met de keuze voor de PRR-partij.
Echter, niet-stemmers die overstapten naar PRR waren wel negatiever over immigranten,
de EU en de politiek in het algemeen, dan de overige niet-stemmers.
Populistisch radicaal rechts en ‘scheidslijn’-stemmers
In het tweede deel van deze dissertatie werd onderzocht in hoeverre belangrijke politieke
scheidslijnen, die traditioneel een belangrijke invloed hebben op stemgedrag (religie en
gender), beïnvloed werden door de opkomst van PRR. Wetenschappers lieten in eerdere
studies zien dat PRR in belangrijke mate bijdraagt aan de opkomst van een nieuwe
scheidslijn rond waarden van globalisering en immigratie. Deze immigratie/globalisering
scheidslijn verdeelt de samenleving in een groep van burgers die globalisering en immigratie toejuichen en burgers die zich juist in sterke mate afkeren van globalisering en immigratie (Betz, 1994; Kriesi et al., 2008; Pellikaan et al., 2007). Een belangrijke aanname in
deze literatuur is dat deze nieuwe scheidslijn de traditionele, sociaal-culturele scheidslijnen
vervangt, waarbij PRR een belangrijke vertegenwoordiger wordt van de anti-immigratie en
anti-globaliseringswaarden. Op basis van die literatuur werd de verwachting uitgesproken
dat PRR, als belangrijke representant van deze nieuwe scheidslijn, mensen die eerder opgesloten zaten in de traditionele scheidslijnen weet te verlokken op hun te stemmen (bv.
Aarts & Thomassen, 2008; Inglehart & Norris, 2000; Lipset & Rokkan, 1967). Om deze
verwachting te staven, werden in deel 2 van het boek twee belangrijke scheidslijnen onder
de loep genomen: de religieuze en de gender scheidslijn. Daarbij werd er ingezoomd op
West-Europese landen waar deze traditionele scheidslijnen het diepst geworteld waren in
de moderne geschiedenis en waar dit soort processen dus het duidelijkst zichtbaar gemaakt
konden worden (bv. Inglehart & Norris, 2000; Judt, 2010).
In Hoofdstuk 5 van dit boek werd onderzocht in welke mate de aanwezigheid van een
populaire PRR-partij de keuze van Christenen om voor een Christelijke partij te stemmen
deed wankelen. Hoewel PRR over het algemeen een seculiere beweging is, heeft ze zich
de afgelopen tijd steeds meer ontpopt als de verdediger van de westerse samenlevingen
die met hun ‘Joods-Christelijke’ tradities beschermd dienen te worden voor ‘Islamitische
dreiging’ (zie bijvoorbeeld het partijprogramma van de Partij voor de Vrijheid: Wilders,
2012). De verwachting werd uitgesproken dat PRR hiermee iets te ‘bieden’ heeft voor
met name orthodox-Christenen. Orthodox-Christenen die hun religie als de enige ware
religie beschouwen, zouden zich namelijk bedreigd kunnen voelen in hun fundamentele
198
Nederlandse samenvatting
levensovertuiging door het toenemende aantal Moslims in hun land. Zodoende zou het
voor hun logischer zijn om te stemmen op PRR dan voor minder orthodoxe Christenen.
Om deze verwachting te toetsen, werd de vierde ronde van de European Values Study
(EVS, 2010) gebruikt. Deze dataset maakte het mogelijk om in zeven verschillende landen
(België, Denemarken, Finland, Nederland, Noorwegen, Oostenrijk en Zwitserland) de
relatie te toetsen tussen Christelijke orthodoxie en het stemmen op een PRR-partij in plaats
van een Christelijke partij. Uit deze analyses bleek dat Christenen het meest geneigd zijn om
te stemmen op een Christelijke partij wat het belang van de traditionele religieuze scheidslijn grotendeels onderschrijft (bv. Knutsen, 2004;Van der Brug et al., 2009). Echter, in drie
landen (België, Noorwegen en Zwitserland) werd bewijs gevonden voor de verwachting dat
orthodox-Christenen meer negatief staan tegenover immigranten dan andere Christenen
en dat er dus een grotere kans is dat zij op een PRR-partij stemmen in plaats van op een
Christelijke partij. Het hoofdstuk suggereerde dat in landen waar een orthodox-Christelijke
partij actief is, zoals de SGP in Nederland, orthodox-Christenen niet vaker PRR stemmen
dan meer doorsnee gelovigen. Deze resultaten lieten dus gedeeltelijk zien dat PRR in staat
is om traditionele scheidslijnen te doorbreken: een PRR-partij met een negatieve houding
jegens immigranten kan een redelijk alternatief zijn voor orthodox-Christenen die niet de
mogelijkheid hebben om op een orthodox-Christelijke partij te stemmen.
In het laatste empirische hoofdstuk van dit boek, Hoofdstuk 6, werd de invloed van PRR
succes op de genderscheidslijn nader bekeken. Veel eerder onderzoek liet zien dat er een
genderverschil is in PRR-stemmen, waarbij vooral mannen zich aangetrokken voelen tot
deze partijen (bv. Givens, 2004; Mayer, 2002). In dit laatste hoofdstuk werd deze zogenaamde
‘gender-gap’ geanalyseerd aan de hand van de volgende onderzoeksvragen: In hoeverre is er
een gender-gap in West-Europees PRR-stemmen en hoe kan dit gender-verschil verklaard
worden? In hoeverre kan variatie in succes en populistisch anti-establishment imago van
PRR-partijen cross-nationale verschillen in de gender-gap verklaren?
Gebaseerd op analyses van de vierde ronde van de European Values Study (EVS, 2010)
werd er inderdaad een gender gap in PRR-stemmen gevonden. Gender-verschillen in typische PRR-attituden, zoals anti-immigranten en autoritaire houdingen konden deze gender
gap niet verklaren. Alleen genderverschillen in socio-economische positie bleken deels bij
te dragen aan de verklaring voor man-vrouw verschillen in PRR-stemmen.
Bovendien lieten de resultaten tamelijk veel variatie zien in de grootte van de gender
gap tussen de verschillende landen. Er werd beargumenteerd dat deze verschillen verklaard
konden worden door de variatie in succes en de mate van het populistisch imago van de
verschillende West-Europese PRR-partijen. Daarbij werd aangenomen dat wanneer PRR
populairder wordt en minder het imago heeft als politieke buitenstaander, vrouwen meer
geneigd zijn op deze partijen te stemmen (bv. Mudde, 2007). Gezien het feit dat vrouwen
minder politiek geïnteresseerd zijn (bv. Burns, 2007) en daarom minder snel op een nieuwe,
199
Nederlandse samenvatting
politieke buitenstaander zullen stemmen (bv. Roth, 1989), werd er voorspeld dat in landen
waar PRR populairder is en een minder populistisch imago heeft de gender gap kleiner
werd. Een verkleining zou dan betekenen dat de genderscheidslijn minder belangrijk wordt
met een toenemende populariteit van PRR.
De bevindingen uit Hoofdstuk 6 onderschreven deze verwachting echter niet: variatie
in de populariteit en populistisch imago van een PRR in een bepaald land hingen niet samen met de grootte van de gender gap in een land. Oftewel, deze bevinding liet zien dat
PRR geen rol lijkt te spelen in de overwegingen van vrouwen in hun stemgedrag en PRR dus
niet bijdraagt aan het doorbreken van de veel gevonden genderscheidslijn in stemgedrag.
Concluderend
Het belangrijkste doel van dit proefschrift was inzicht te verschaffen in het effect van de
opkomst en populariteit van PRR op het stemgedrag van Europese burgers. Daarbij werd
in het eerste deel aandacht besteed aan de invloed van PRR op niet-stemmers en dus (in)
direct op de opkomstcijfers bij verkiezingen. In het tweede deel werd nader onderzocht of
en hoeveel PRR het stemgedrag van burgers beïnvloedt die in hoge mate geneigd zijn te
stemmen op dezelfde partij (de zogenaamde ‘scheidslijn’-stemmers). Gebaseerd op de vijf
empirische hoofdstukken uit dit boek kan geconcludeerd worden dat PRR, ondanks haar
unieke profiel van populistische, nativistische en autoritaire uitdager van het establishment,
slechts beperkt in staat is om niet-stemmers en typische ‘scheidslijn’-stemmers te verlokken
hun gewoonlijke stemgedrag te heroverwegen.
Uit het eerste deel bleek dat toenemend succes van populaire PRR-partijen niet samengaat met een sterkere individuele neiging om naar de stembus te gaan. In West-Europa
bleek een succesvolle PRR-partij geen effect te hebben op de individuele bereidheid tot
stemmen. Alleen mensen die in politiek geïnteresseerd waren, gingen vaker naar de stembus wanneer PRR populairder was. In Oost-Europa bleek er eerder een negatieve invloed
te zijn van PRR op de opkomst bij verkiezingen. Voornamelijk jongeren, hogeropgeleiden
en politiek geïnteresseerden leken zich in sterkere mate af te keren van het electorale proces
wanneer PRR populairder was. Deze bevindingen staan in sterk contrast met de notie van
PRR als de ‘gepassioneerde katalysator’ van politieke betrokkenheid (vgl. Mouffe, 2005)
die bij kan dragen aan stijgende opkomstcijfers bij verkiezingen. Ze laten eerder zien dat
PRR een bedreiging kan vormen voor Oost-Europese landen. Als jongeren niet stemmen,
zouden hun belangen minder goed vertegenwoordigd worden in de nationale parlementen
en is het bovendien aannemelijk dat zij ook later in hun leven minder vaak zullen stemmen
(bv. Plutzer, 2002). Dat zou dus langetermijneffecten kunnen hebben op de opkomstcijfers
bij verkiezingen en de kwaliteit en legitimiteit van democratie ondermijnen.
200
Nederlandse samenvatting
Toch is voorzichtigheid geboden met te concluderen dat PRR alleen maar bedreigend is
voor Europese democratieën. Deze dissertatie liet ook zien dat de Nederlands PRR-partijen
wel in staat leken te zijn om meer voormalig niet-stemmers aan zich te binden dan welke
andere afzonderlijke partij ook. Hoewel deze groep niet omvangrijk was, bleken de veranderingen van een niet-stem naar een stem op PRR vooral samen te gaan met een verandering
in de sympathie voor de leider. Degene die overstapten bleken ideologisch wel meer overeenkomsten te hebben met PRR dan de andere niet-stemmers. Echter, als charismatisch
leiderschap de primaire reden is van niet-stemmers om te gaan stemmen op een PRR, zou
het kunnen dat dit positieve mobilisatie-effect van PRR op niet-stemmers tijdelijk is. Het
charisma van een leider is een sociaal construct dat door media- en publieke aandacht
aan constante verandering onderhevig is (Weber, 2009; Pels, 2011). Dit zou betekenen dat
de groep niet-stemmers die overstapten naar PRR wellicht ook sneller weer thuis zouden
blijven als ze de leider niet meer sympathiek genoeg vinden. Meer onderzoek is nodig om
die verwachting te staven.
Uit het tweede deel van deze dissertatie bleek dat de aanwezigheid van een succesvolle
PRR-partij in een land vrouwen niet aanspoort hun stemgedrag te heroverwegen. Er werd
geen relatie gevonden tussen de populariteit van PRR en de mate van genderverschillen in
PRR-stemmen. Daarentegen bleek dat orthodox-Christenen zich wel voor een ideologisch
dilemma gesteld zien staan. Orthodox-Christenen leken hun fundamentele waarheidsclaims
meer onder druk te voelen staan wanneer zij geconfronteerd werden met immigranten die
een andere religie aanhangen. Deze religieuze dreiging wordt door enkele PRR-partijen
benoemd en kan daarmee een reden zijn voor orthodox religieuzen om te stemmen op
PRR in plaats van een Christelijke partij. Op de lange termijn zou dit kunnen betekenen
dat PRR bijdraagt aan de verbrokkeling van de traditionele Christelijke scheidslijn die,
ondanks processen van secularisering, zo sterk is verankerd in veel West-Europese landen
(vgl. Kalyvas & Van Kersbergen, 2010; Van der Brug et al., 2009).
In dit boek was de ‘revolutionaire’ opkomst van rechts-populist Fortuyn en de daarmee gepaard gaande maatschappelijke en politieke onrust de aanleiding voor de vraag
of en hoe PRR het stemgedrag beïnvloedt van niet-stemmers en ‘scheidslijn’-stemmers.
Hoewel Fortuyn in staat leek te zijn om politiek vervreemde burgers (weer) te betrekken bij
politiek, hebben de vijf empirische hoofdstukken tezamen laten zien dat er geen universele
aantrekkingskracht uitgaat van Europese PRR-partijen op het politieke gedrag van burgers.
Wanneer PRR populairder wordt, blijven de meeste niet-stemmers niet naar de stembus
komen, blijven Christenen op Christelijke partijen stemmen, en blijven vrouwen minder
vaak stemmen op PRR. Deze bevindingen onderstrepen het belang van wetenschappelijk,
empirisch onderzoek om veelgehoorde claims over PRR bij te kunnen stellen. Dit proefschrift liet zien dat de notie van PRR als ‘hoeder van de democratie’ lichtelijk overdreven
blijkt te zijn.
201
References
203
References
Aarts, K., & Thomassen, J. (2008). Dutch voters and the changing party space 1989-2006. Acta Politica, 42(2/3),
203-234.
Abramson, P. R., & Inglehart, R. (1995). Value change in global perspective. Ann Arbor: University of Michigan
Press.
Abts, K., & Rummens, S. (2007). Populism versus democracy. Political Studies, 55(2), 405-424.
Abu Raiya, H., Pargament, K. I., Mahoney, A., & Trevino, K. (2008). When Muslims are perceived as a religious
threat: Examining the connection between desecration, religious coping, and anti-muslim attitudes. Basic
and Applied Social Psychology, 30(4), 311-325.
Achterberg, P., Houtman, D., Aupers, S., De Koster, W., Mascini, P., & Van der Waal, J. (2009). A Christian cancellation of the secularist truce? Waning Christian religiosity and waxing religious deprivatization in the
West. Journal for the Scientific Study of Religion, 48(4), 687-701.
Akkerman, T. (2012). Comparing radical right parties in government: Immigration and integration policies in
nine countries (1996-2010). West European Politics, 35(3), 511-529.
Aldrich, J. H., Montgomery, J. M., & Wood, W. (2011). Turnout as a habit. Political Behavior, 33(4), 535-563.
Allison, P. D. (2009). Fixed effects regression models. Los Angeles: Sage.
Allport, G. W. (1966). The religious context of prejudice. Journal for the Scientific Study of Religion, 5(3), 447-457.
Altemeyer, B. (2005). Fundamentalism and authoritarianism. In R. F. Paloutzian, & C. L. Park (Eds.), Handbook of
the psychology of religion and spirituality (pp. 378-393). New York: Guilford Press.
Altemeyer, B., & Hunsberger, B. (1992). Authoritarianism, religious fundamentalism, quest, and prejudice. International Journal for the Psychology of Religion, 2(2), 113-133.
Andersen, R., & Evans, J. A. J. (2003). Social-political context and authoritarian attitudes: Evidence from seven
European countries. CREST Working Paper (No. 104), Glasgow, UK.
Anderson, C. J., & Beramendi, P. (2012). Left parties, poor voters, and electoral participation in advanced industrial societies. Comparative Political Studies, 45(6), 714-746.
Andeweg, R. B., & Irwin, G. A. (2005). Governance and politics of the Netherlands. Houndmills: Palgrave MacMillan.
Andeweg, R. B. (1982). Dutch voters adrift: On explanations of electoral change (1963-1977). Leiden: Leiden
University.
Ansolabehere, S. D., & Iyengar, S. (1995). Going negative. How political advertisements shrink and polarize the
electorate. New York: The Free Press.
Arendt, H. ([1951] 1973). The origins of totalitarianism. Orlando: Harcourt Brace & Company.
Armingeon, K., Careja, R., Weisstanner, D., Engler, S., Potolidis, P., & Gerber, M. (2012). Comparative Political
Data Set III 1990-2010. Institute of Political Science, University of Bern, Switzerland.
Art, D. (2006). The European radical right in comparative-historical perspective. Paper Presented at the American
Political Science Association, August 31-September 3, Philadelphia, USA.
Art, D. (2011). Inside the radical right. The development of anti-immigrant parties in Europe. Cambridge: Cambridge University Press.
Arter, D. (2010). The breakthrough of another West European populist radical right party? The case of the True
Finns. Government and Opposition, 45(4), 484-504.
Arzheimer, K. (2009). Contextual factors and the extreme right vote in Western Europe, 1980-2002. American
Journal of Political Science, 53(2), 259-275.
Arzheimer, K. (2012). Electoral sociology - who votes for the extreme right and why - and when? In U. Backes,
& P. Moreau (Eds.), The extreme right in Europe. Current trends and perspectives (pp. 35-50). Göttingen:
Vandenhoeck & Ruprecht.
205
References
Arzheimer, K., & Carter, E. (2006). Political opportunity structures and right-wing extremist party success. European Journal of Political Research, 45(3), 419-443.
Arzheimer, K., & Carter, E. (2009). Christian religiosity and voting for West European radical right parties. West
European Politics, 32(5), 985-1011.
Backes, U., & Moreau, P. (Eds.) (2012). The extreme right in Europe. Current trends and perspectives. Göttingen:
Vandenhoeck & Ruprecht.
Bale, T. (2003). Cinderella and her ugly sisters: The mainstream and extreme right in Europe’s bipolarising party
systems. West European Politics, 26(3), 67-90.
Bale, T. (2008). Turning round the telescope. Centre-right parties and immigration and integration policy in
Europe. Journal of European Public Policy, 15(3), 315-330.
Bale, T., Green-Pedersen, C., Krouwel, A., Luther, K. R., & Sitter, N. (2010). If you can’t beat them, join them?
Explaining social democratic responses to the challenge from the populist radical right in Western Europe.
Political Studies, 58(3), 410-426.
Benoit, K., & Laver, M. (2006). Party policy in modern democracies. Routledge Research in Comparative Politics,
19, 1-289.
Berezin, M. (2011). Illiberal politics in neoliberal times. Culture, security and populism in the new Europe. Cambridge: Cambridge University Press.
Betz, H. (1994). Radical right-wing populism in Western Europe. New York: St Martin’s.
Billiet, J. B. (1995). Church involvement, ethnocentrism, and voting for a radical right-wing party: Diverging
behavioral outcomes of equal attitudinal dispositions. Sociology of Religion, 56(3), 303-326.
Blais, A. (2006). What affects voter turnout? Annual Review of Political Science, 9, 111-125.
Blais, A., & Rubenson, D. (2013). The source of turnout decline: New values or new contexts? Comparative Political
Studies, 46(1), 95-117.
Blogowska, J., & Saroglou, V. (2011). Religious fundamentalism and limited prosociality as a function of the target.
Journal for the Scientific Study of Religion, 50(1), 44-60.
Bovens, M., & Wille, A. (2010). Diplomademocratie. Over de spanning tussen meritocratie en democratie. Amsterdam: Bert Bakker.
Braun, R. (2011). The diffusion of racist violence in the Netherlands: Discourse and distance. Journal of Peace
Research, 48(6), 753-766.
Brooks, D. J., & Geer, J. G. (2007). Beyond negativity: The effects of incivility on the electorate. American Journal
of Political Science, 51(1), 1-16.
Burchell, J. (2001). Evolving or conforming? Assessing organisational reform within European green parties. West
European Politics, 24(3), 113-134.
Burns, N. (2007). Gender in the aggregate, gender in the individual, gender and political action. Politics & Gender,
3(1), 104-124.
Bustikova, L. (2009). The extreme right in Eastern Europe: EU accession and the quality of governance. Journal of
Contemporary European Studies, 17(2), 223-239.
Cameron, A. C., & Trivedi, P. K. (2010). Microeconometrics using Stata. College Station, Texas: Stata Press.
Campbell, D. E. (2006). Why we vote. How schools and communities shape our civic life. Princeton, NJ: Princeton
University Press.
Camus, J. (2007). The European extreme right and religious extremism. Středoevropské Politické Studie [Central
European Political Studies], 9(4), 263-279.
Canovan, M. (1999). Trust the people! Populism and the two faces of democracy. Political Studies, 47(1), 2-16.
206
References
Coenders, M., Lubbers, M., & Scheepers, P. (2006). ‘Het tolerante land’ in historisch en in landenvergelijkend
perspectief. In F. Van Tubergen, & I. Maas (Eds.), Allochtonen in Nederland in internationaal perspectief
(pp. 89-110). Amsterdam: Amsterdam University Press.
Coffé, H. (2005). Extreem rechts in Vlaanderen en Wallonië: Het verschil. Roeselare: Roularta Books.
Coffé, H. (2008). Social democratic parties as buffers against the extreme right: The case of Belgium. Contemporary
Politics, 14(2), 179-195.
Coffé, H., & Bolzendahl, C. (2010). Same game, different rules? Gender differences in political participation. Sex
Roles, 62(5-6), 318-333.
Coffé, H., Heyndels, B., & Vermeir, J. (2007). Fertile grounds for extreme-right parties: Explaining the Vlaams
Blok’s electoral success. Electoral Studies, 26(1), 142-155.
Couwenberg, S. W. (2004). Opstand der burgers. De Fortuyn-revolte en het demasqué van de oude politiek. Budel:
Civis Mundi jaarboek 2004.
Dahl, R. A. (1998). On democracy. London: Yale University Press.
Dahlström, C., & Sundell, A. (2012). A losing gamble. How mainstream parties facilitate anti-immigrant party
success. Electoral Studies, 31(2), 353-363.
Dalton, R. J. (2008). Citizenship norms and the expansion of political participation. Political Studies, 64(1), 76-98.
Dalton, R. J. (2009). Economics, environmentalism, and party alignments: A note on partisan change in advanced
industrial democracies. European Journal of Political Research, 48(2), 161-175.
De Lange, S. L. (2012a). Politiek en samenleving in het post-Fortuyn tijdperk. Beleid en Maatschappij, 39(2),
114-128.
De Lange, S. L. (2012b). Radical right-wing populist parties in offices - A cross-national comparison. In U. Backes,
& P. Moreau (Eds.), The extreme right in Europe. Current trends and perspectives (pp. 171-194). Göttingen:
Vandenhoeck & Ruprecht.
De Lange, S. L., & Akkerman, T. (2012). Populist parties in Belgium: A case of hegemonic liberal democracy? In
C. Mudde, & C. Rovira Kaltwasser (Eds.), Populism in Europe and the Americas. Threat or corrective for
democracy? (pp. 27-45). New York: Cambridge University Press.
De Rooij, E. (2009). Specialisation of political participation in Europe: A comparative analysis. Doctoral Thesis,
University of Oxford, Oxford, United Kingdom.
De Tocqueville, A. ([1835-1840] 2000). Democracy in America. Chicago: University of Chicago Press.
Denny, K., & Doyle, O. (2009). Does voting history matter? Analysing persistence in turnout. American Journal of
Political Science, 53(1), 17-35.
Dodson, K. (2011). The movement society in comparative perspective. Mobilization: An International Journal,
16(4), 475-494.
Donovan, T., & Karp, J. A. (2006). Popular support for direct democracy. Party Politics, 12(5), 671-688.
Downs, A. (1957). An economic theory of political action in a democracy. The Journal of Political Economy, 2,
135-150.
Driskell, R., Embry, E., & Lyon, L. (2008). Faith and politics: The influence of religious beliefs on political participation. Social Science Quarterly, 89(2), 294-314.
Dunn, K. P., & Singh, S. P. (2011). The surprising non-impact of radical right-wing populist representation on
public tolerance of minorities. Journal of Elections, Public Opinion & Parties, 21(3), 313-331.
Eatwell, R. (2000). The rebirth of the ‘extreme right’ in Western Europe? Parliamentary Affairs, 53(3), 410-414.
Edlund, L., & Pande, R. (2002). Why have women become left-wing? The political gender gap and the decline in
marriage. The Quarterly Journal of Economics, 117(3), 917-961.
Ennser, L. (2012). The homogeneity of West European party families: The radical right in comparative perspective.
Party Politics, 18(2), 151-171.
207
References
Enos, R. D., & Fowler, A. (2012). Can electoral competition mobilize under-represented citizens? Evidence from a
field experiment in the aftermath of a tied election. Working Paper, 1-35.
Esaiasson, P., & Narud, H. M. (Eds.) (2013). Between-election democracy. Colechester, UK: ECPR Press.
ESS Round 1: European Social Survey Round 1 Data (2002). Data file edition 6.2. Norwegian Social Science Data
Services, Norway – Data archive and distributor of ESS data. Retrieved, September 1/2012, from www.
europeansocialsurvey.org
ESS Round 2: European Social Survey Round 2 Data (2004). Data file edition 3.2. Norwegian Social Science Data
Services, Norway – Data archive and distributor of ESS data. Retrieved, September 1/2012, from www.
europeansocialsurvey.org
ESS Round 3: European Social Survey Round 3 Data (2006). Data file edition 3.3. Norwegian Social Science Data
Services, Norway – Data archive and distributor of ESS data. Retrieved, September 1/2012, from www.
europeansocialsurvey.org
ESS Round 4: European Social Survey Round 4 Data (2008). Data file edition 4.0. Norwegian Social Science Data
Services, Norway – Data archive and distributor of ESS data. Retrieved, September 1/2012, from www.
europeansocialsurvey.org
ESS Round 5: European Social Survey Round 5 Data (2010). Data file edition 2.0. Norwegian Social Science Data
Services, Norway – Data archive and distributor of ESS data. Retrieved, September 1/2012, from www.
europeansocialsurvey.org
ESS Round 6: European Social Survey Round 6 Data (2012). Data file edition 2.0. Norwegian Social Science
Data Services, Norway – Data archive and distributor of ESS data. Retrieved, January 1/2014, from www.
europeansocialsurvey.org
EVS Wave 4: European Values Study Wave 4 Data (2010). Integrated dataset, ZA4800 data file version 2.0.0.
Cologne, Germany: GESIS Data Archive.
Fallend, F. (2012). Populism in government: The case of Austria. In C. Mudde, & C. Rovira Kaltwasser (Eds.),
Populism in Europe and the Americas. Threat or corrective for democracy? (pp. 113-135). New York: Cambridge University Press.
Fennema, M. (2005). Populist parties of the right. In J. Rydgren (Ed.), Movements of exclusion: Radical right-wing
populism in the Western world (pp. 1-24). Hauppage, NY: Nova Science.
Fennema, M., & Tillie, J. (1998). Social isolation. Theoretical concepts and empirical measurement. In M. Fennema,
H. Schijf & C. van der Eijk (Eds.), In search of structure: Essays in social science methodology. Amsterdam,
the Netherlands: Het Spinhuis.
Fontana, M., Sidler, A., & Hardmeier, S. (2006). The ‘new right’ vote: An analysis of the gender gap in the vote
choice for SVP. Swiss Political Science Review, 12(4), 243-271.
Ford, R., & Goodwin, M. J. (2014). Revolt on the right. Explaining support for the radical right in Britain. Oxon:
Routledge.
Fowler, J. (2006). Habitual voting and behavioral turnout. Journal of Politics, 68(2), 335-344.
Franklin, M. N. (2004). Voter turnout and the dynamics of electoral competition in established democracies since
1945. New York: Cambridge University Press.
Frankovic, K. A. (1982). Sex and politics. New alignments, old issues. PS, 15(3), 439-448.
Gallagher, M., Laver, M., & Mair, P. (2006). Representative government in modern Europe. New York: McGraw Hill.
Ganzeboom, H. B. G., & Treiman, D. J. (1996). Internationally comparable measures of occupational status for the
1988 International Standard Classification of Occupations. Social Science Research, 25, 201-239.
Gerber, A. S., Green, D. P., & Shachar, R. (2003). Voting may be habit forming: Evidence from a randomized field
experiment. American Journal of Political Science, 47(3), 540-550.
208
References
Gidengil, E., Hennigar, M., Blais, A., & Nevitte, N. (2005). Explaining the gender gap in support for the new right:
The case of Canada. Comparative Political Studies, 38(10), 1171-195.
Gilens, M. (1988). Gender and support for Reagan: A comprehensive model of presidential approval. American
Journal of Political Science, 32(1), 19-49.
Gilligan, C. (1982). In a different voice. Cambridge, MA: Harvard University Press.
Gimpel, J. G., Kaufmann, K. M., & Pearson-Merkowitz, S. (2007). Battleground states versus blackout states: The
behavioral implications of modern presidential campaigns. The Journal of Politics, 69(3), 786-797.
Givens, T. E. (2004). The radical right gender gap. Comparative Political Studies, 37(1), 30-54.
Golder, M. (2003). Explaining variation in the success of extreme right parties in Western Europe. Comparative
Political Studies, 36(4), 163-178.
Green, D. P., & Shachar, R. (2000). Habit formation and political behaviour: Evidence of consuetude in voter
turnout. British Journal of Political Science, 30(4), 561-573.
Green-Pedersen, C., & Van Kersbergen, K. (2002). The politics of the ‘Third Way’. Party Politics, 8(5), 507-524.
Güveli, A., Need, A., & De Graaf, N. D. (2007). Socio-political, cultural, and economic preferences and behaviour
of the social and cultural specialists and the technocrats. Social class or education? Social Indicators
Research, 81(3), 597-631.
Halikiopoulou, D., Nanou, K., & Vasilopoulou, S. (2012). The paradox of nationalism: The common denominator
of radical right and radical left euroscepticism. European Journal of Political Research, 51(4), 504-539.
Heath, A., McLean, I., Taylor, B., & Curtice, J. (1999). Between first and second order: A comparison of voting behaviour in European and local elections in Britain. European Journal of Political Research, 35(3), 389-414.
Hibbing, J. R., & Theiss-Morse, E. (2001). Process preference and American politics: What the people want government to be. American Political Science Review, 95(1), 145-154.
Hibbing, J. R., & Theiss-Morse, E. (2002). Stealth democracy. Americans’ beliefs about how government should work.
Cambridge: Cambridge University Press.
Hirsch-Hoefler, S., Canetti, D., & Pedahzur, A. (2010). Two of a kind? Voting motivations for populist radical right
and religious fundamentalist parties. Electoral Studies, 29(4), 678-690.
Hirschman, A. (1970). Exit, voice and loyalty. Cambridge, MA: Harvard University Press.
Hooghe, L., Bakker, R., Brigevich, A., de Vries, C., Edwards, E., Marks, G., & Steenbergen, M. (2010). Reliability
and validity of measuring party positions: The Chapel Hill Expert Surveys of 2002 and 2006. European
Journal of Political Research, 49(5), 684-703.
Hooghe, M., Marien, S., & Pauwels, T. (2011). Where do distrusting voters turn if there is no viable exit or voice
option? The impact of political trust on electoral behaviour in the Belgian regional elections of June 2009.
Government and Opposition, 46(2), 245-273.
Hox, J. (2002). Multilevel analysis: Techniques and applications. London: Lawrence Elbaum Ass.
Huber, J., & Inglehart, R. (1995). Expert interpretations of party space and party locations in 42 societies. Party
Politics, 1(1), 73-111.
Huckfeldt, R., & Sprague, J. (1992). Political parties and electoral mobilization: Political structure, social structure,
and the party canvass. American Political Science Review, 86(1), 70-86.
Huddy, L., & Terkildsen, N. (1993). Gender stereotypes and the perception of male and female candidates. American Journal of Political Science, 37(1), 119-147.
Ignazi, P. (1992). The silent counter-revolution: Hypotheses on the emergence of extreme right-wing parties in
Europe. European Journal of Political Research, 22(1), 3-34.
Ignazi, P. (2003). Extreme right parties in Western Europe. Oxford: Oxford University Press.
209
References
Immerzeel, T., Lubbers, M., & Coffé, H. (2011). Expert Judgment Survey ofEuropean Political Parties 2010. Utrecht:
Utrecht University, NWO, Department of Sociology, archived at DANS: https://easy.dans.knaw.nl/ui/
datasets/id/easy-dataset:55725.
Inglehart, R., & Norris, P. (2000). The developmental theory of the gender gap: Women and men’s voting behaviour
in global perspective. International Political Science Review, 21(4), 441-462.
Inglehart, R., & Norris, P. (2003). Rising tide: Gender equality and cultural change around the world. Cambridge:
Cambridge University Press.
Ivarsflaten, E. (2008). What unites the populist radical right in Western Europe? Reexamining grievance mobilization models in seven successful cases. Comparative Political Studies, 41(1), 3-23.
Jackson, R. A., Mondak, J. J., & Huckfeldt, R. (2009). Examining the possible corrosive impact of negative advertising on citizens’ attitudes. Political Research Quarterly, 62(1), 55-69.
Jagers, J. (2006). Stem van het volk! Populisme als concept getest bij Vlaamse politieke partijen. Doctoral Thesis,
University of Antwerp, Antwerp, Belgium.
Jansen, R. S. (2011). Populist mobilization: A new theoretical approach to populism. Sociological Theory, 29(2),
75-96.
Judt, T. (2010). Postwar: A history of Europe since 1945. London: Vintage Books.
Kahn, K. F., & Kenney, P. J. (1999). Do negative campaigns mobilize or suppress turnout? Clarifying the relationship between negativity and participation. American Political Science Review, 93(4), 877-889.
Kalyvas, S. N., & Van Kersbergen, K. (2010). Christian democracy. Annual Review of Political Science, 13, 183-209.
Kaufmann, K. M., & Petrocik, J. R. (1999). The changing politics of American men: Understanding the sources of
the gender gap. American Journal of Political Science, 43(3), 864-887.
Kiesraad (2014). Databank Verkiezingsuitslagen [Databank Election results]. Retrieved January 1/2014 from http://
www.verkiezingsuitslagen.nl/.
Kitschelt, H. (2007). Growth and persistence of the radical right in postindustrial democracies: Advances and
challenges in comparative research. West European Politics, 30(5), 1176-1206.
Kitschelt, H., & McGann, A. J. (1995). The radical right in Western Europe: A comparative analysis. Ann Arbor:
University of Michigan Press.
Klingemann, H., Volkens, A., Bara, J., Budge, I., & McDonald, M. D. (2006). Mapping policy preferences II. Estimates for parties, electors, and governments in Central and Eastern Europe, European Union and OECD
1990-2003. Oxford: Oxford University Press.
Knigge, P. (1998). The ecological correlates of right-wing extremism in Western Europe. European Journal of Political Research, 34(2), 249-279.
Knutsen, O. (2001). Social class, sector employment, and gender as party cleavages in the Scandinavian countries:
A comparative longitudinal study. Scandinavian Political Studies, 24(4), 311-350.
Knutsen, O. (2004). Religious denomination and party choice in Western Europe: A comparative longitudinal
study from eight countries, 1970-97. International Political Science Review, 25(1), 97-128.
Koopmans, R., & Muis, J. (2009). The rise of right-wing populist Pim Fortuyn in the Netherlands: A discursive
opportunity approach. European Journal of Political Research, 48(5), 642-664.
Kornhauser, W. (1959). The politics of mass society. Glencoe, IL: Free Press.
Kriesi, H., Grande, E., Lachat, R., Dolezal, M., Bornschier, S., & Frey, T. (2008). West European politics in the age of
globalization. Cambridge: Cambridge University Press.
Krouwel, A., & Lucardie, P. (2008). Waiting in the wings: New parties in the Netherlands. Acta Politica, 43(2),
278-307.
Lijphart, A. (1999). Patterns of democracy: Government forms and performance in 36 countries. New Haven: Yale
University Press.
210
References
Lipset, S. M., & Rokkan, S. (1967). Cleavage structures, party systems and voter alignments: An introduction. In
S. M. Lipset, & S. Rokkan (Eds.), Party systems and voter alignments: Cross-national perspectives (pp. 1-64).
New York: Free Press.
Long, J. S., & Freese, J. (2006). Regression models for categorical and dependent variables using Stata. College Station, Texas: Stata Press.
Lovenduski, J. (1986). Women and European politics: Contemporary feminism and public policy. Amherst: University of Massachusetts Press.
Lubbers, M. (2001a). Exclusionistic electorates: Extreme right-wing voting in Western Europe. Amsterdam: Thela
Thesis.
Lubbers, M. (2001b). Expert Judgment Survey of Western-European Political Parties 2000. Nijmegen, The Netherlands: NWO, Department of Sociology, University of Nijmegen.
Lubbers, M., & Gijsberts, M. (2004). Buurtdynamiek in verkiezingsuitslagen. Sociale Wetenschappen, 47(1), 41-60.
Lubbers, M., Gijsberts, M., & Scheepers, P. (2002). Extreme right-wing voting in Western Europe. European
Journal of Political Research, 41(3), 345-378.
Lubbers, M., & Scheepers, P. (2000). Individual and contextual characteristics of the German extreme right vote in
the 1990s. European Journal of Political Research, 38(1), 63-94.
Lucardie, P., & Voerman, G. (2012). Populisten in de polder. Amsterdam: Boom Uitgevers.
Mair, P. (2006a). Polity-scepticism, party failings, and the challenge to European democracy. Wassenaar, the Netherlands: NIAS.
Mair, P. (2006b). Ruling the void? The hollowing of Western democracy. New Left Review, 42, 25-51.
Mair, P. (2008). Electoral volatility and the Dutch party system: A comparative perspective. Acta Politica, 43(1),
235-253.
March, L. (2011). Radical left parties in Europe. Oxon: Routledge.
March, L. (2007). From vanguard of the proletariat to Vox populi: Left-populism as a ‘shadow’ of contemporary
socialism. SAIS Review, 17(1), 63-77.
March, L., & Mudde, C. (2005). What’s left of the radical left? The European radical left after 1989: Decline and
mutation. Comparative European Politics, 3(1), 23-49.
Mayer, N. (2002). Ces Français qui votent Le Pen. Saint-Amand-Montrond: Flammarion.
Meguid, B. M. (2005). Competition between unequals: The role of mainstream party strategy in niche party success. American Political Science Review, 99(3), 347-359.
Moors, H., Balogh, L., Van Donselaar, J., & De Graaff, B. (2009). Polarisatie en radicalisering in Nederland. Tilburg:
IVA Beleidsonderzoek en advies.
Mouffe, C. (2005). The ‘end of politics’ and the challenge of right-wing populism. In F. Panizza (Ed.), Populism and
the mirror of democracy (pp. 50-71). London: Verso.
Mudde, C. (2004). The populist zeitgeist. Government and Opposition, 39(4), 542-563.
Mudde, C. (2007). Populist radical right parties. Cambridge: Cambridge University Press.
Mudde, C. (2012). The 2012 Stein Rokkan lecture: Three decades of populist radical right parties in Western Europe: So what? European Journal of Political Research, Presented at the ECPR Joint Sessions of Workshops
in Antwerp, Belgium, April 11.
Mudde, C., & Rovira Kaltwasser, C. (Eds.) (2012). Populism in Europe and the Americas. Threat or corrective for
democracy? New York: Cambridge University Press.
Mutz, D. C., & Reeves, B. (2005). The new videomalaise: Effects of televised incivility on political trust. American
Political Science Review, 99(1), 1-15.
Naidoo, L. J., & Lord, R. G. (2008). Speech imagery and perceptions of charisma: The mediating role of positive
affect. The Leadership Quarterly, 19(3), 283-296.
211
References
Newton, K. (2001). Trust, social capital, civil society, and democracy. International Political Science Review, 22(2),
201-214.
Norris, P. (2005). Radical right: Voters and parties in the electoral market. New York: Cambridge University Press.
Norwegian Social Science Data Services (NSD) (2014). European Election and Referendum Database. Retrieved
March 1/2014 from: http://www.nsd.uib.no/european_election_database/
Novum/ANP (2006). Wilders: Artikel 1 van de grondwet schrappen [Wilders: Abolish Article 1 of the constitution]. Trouw, March 21.
Oesch, D. (2006). Coming to grips with a changing class structure. An analysis of employment stratification in
Britain, Germany, Sweden and Switzerland. International Sociology, 21(2), 263-288.
Oesch, D. (2008). Explaining workers’ support for right-wing populist parties in Western Europe: Evidence from
Austria, Belgium, France, Norway, and Switzerland. International Political Science Review, 29(3), 349-373.
Olzak, S. (1992). Dynamics of ethnic competition and conflict. Stanford: Stanford University Press.
Oosterwaal, A., & Torenvlied, R. (2010). Politics divided from society? Three explanations for trends in societal
and political polarisation in the Netherlands. West European Politics, 33(2), 258-279.
Pacheco, J. S. (2008). Political socialization in context: The effect of political competition on youth voter turnout.
Political Behavior, 30(4), 415-436.
Papadopoulous, Y. (2002). National populism in Western European: An ambivalent phenomenon. Lausanne, Switzerland: Institut d’Études Politiques et Internationales.
Pargament, K. I. (1997). The psychology of religion and coping: Theory, research, practice. New York: Guilford.
Pedersen, M. N. (1979). The dynamic of European party systems: Changing patterns of electoral volatility. European Journal of Political Research, 7(1), 1-26.
Pellikaan, H., De Lange, S. L., & Van der Meer, T. (2007). Fortuyn’s legacy: Party system change in the Netherlands.
Comparative European Politics, 5(3), 282-302.
Pels, D. (2011). Het volk bestaat niet. Leiderschap en populisme in de mediademocratie. Amsterdam: De Bezige Bij.
Plutzer, E. (2002). Becoming a habitual voter: Inertia, resources, and growth in young adulthood. American Political Science Review, 96(1), 41-56.
Poguntke, T. (1993). Goodbye to movement politics? Organisational adaptation of the German green party.
Environmental Politics, 2(3), 379-404.
Putnam, R. (1995). Bowling alone: America’s declining social capital. Journal of Democracy, 6(1), 65-78.
Putnam, R. (2000). Bowling alone: The collapse and revival of the American community. New York: Simon &
Schuster.
Quintelier, E. (2008). Changing determinants of party memberships in Flanders. Paper presented at the Politicologenetmaal May 29-30, Nijmegen, the Netherlands.
Rabe-Hesketh, S., & Skrondal, A. (2008). Multilevel and longitudinal modeling using Stata. Texas: Stata Press.
Reitsma, J. (2007). Religiosity and solidarity. Dimensions and relationships disentangled and tested. Dissertation,
Radboud University Nijmegen, Nijmegen, the Netherlands.
Riemen, R. (2010). De eeuwige terugkeer van het fascisme [The eternal recurrence of fascism]. Amsterdam: AtlasContact.
Rink, N., Phalet, K., & Swyngedouw, M. (2009). The effects of immigrant population size, unemployment and
individual characteristics on voting for the Vlaams Blok in Flanders 1991-1999. European Sociological
Review, 25(4), 411-424.
Rippeyoung, P. L. F. (2007). When women are right: The influence of gender, work and values on European farright party support. International Feminist Journal of Politics, 9(3), 379-397.
Rooduijn, M., De Lange, S. L., & Van der Brug, W. (2012). A populist Zeitgeist? Programmatic contagion by
populist parties in Western Europe. Party Politics, Advance Access, 1-20.
212
References
Rosanvallon, P. (2008). Counter-democracy. Politics in an age of distrust. New York: Cambridge University Press.
Rosenstone, S. J., & Hansen, J. M. (1992). Mobilization, participation and democracy in America. New York:
MacMillan.
Roth, D. (1989). Sind die Republikaner die fünfte Partei? Aus Politik Und Zeitgeschichte, 41-42, 10-20.
Rovira Kaltwasser, C. (2012). The ambivalence of populism: Threat and corrective for democracy. Democratization, 19(2), 184-208.
Rovny, J. (2013). Where do radical right parties stand? Position blurring in multidimensional competition. European Political Science Review, 5(1), 1-26.
Rowatt, W. C., & Franklin, L. M. (2004). Christian orthodoxy, religious fundamentalism, and right-wing authoritarianism as predictors of implicit racial prejudice. International Journal for the Psychology of Religion,
14(2), 125-138.
Rydgren, J. (2007). The sociology of the radical right. Annual Review of Sociology, 33, 241-262.
Rydgren, J. (2009). Social isolation? Social capital and radical right-wing voting in Western Europe. Journal of Civil
Society, 5(2), 129-150.
Rydgren, J. (2011). A legacy of ‘uncivicness’? Social capital and radical right wing populist voting in Eastern
Europe. Acta Politica, 46(2), 132-157.
Rydgren, J. (2012). Class politics and the radical right. Oxon: Routledge.
Scheepers, P., Gijsberts, M., & Coenders, M. (2002). Ethnic exclusionism in European countries. European Sociological Review, 18(1), 17-34.
Scheepers, P., Gijsberts, M., & Hello, E. (2002). Religiosity and prejudice against ethnic minorities in Europe:
Cross-national tests on a controversial relationship. Review of Religious Research, 43(3), 242-265.
Scherpenzeel, A. (2009). Start of the LISS panel: Sample and recruitment of a probability-based internet panel.
Tilburg: CentERdata, the Netherlands.
Schneider, S. L. (2008). Anti-immigrant attitudes in Europe: Outgroup size and perceived ethnic threat. European
Sociological Review, 24(1), 53-67.
Schumacher, G., & Rooduijn, M. (2013). Sympathy for the ‘devil’? Voting for populists in the 2006 and 2010 Dutch
general elections. Electoral Studies, 32(1), 124-133.
Semyonov, M., Raijman, R., & Gorodzeisky, A. (2006). The rise of anti-foreigner sentiment in European societies,
1988-2000. American Sociological Review, 71(3), 426-449.
Snijders, T. A. B., & Bosker, R. J. (1999). Multilevel analysis: An introduction to basic and advanced multilevel
modeling. London: SAGE.
Studlar, D. T., McAllister, I., & Hayes, B. C. (1998). Explaining the gender gap in voting: A cross-national analysis.
Social Science Quarterly, 79(4), 779-798.
Swank, D., & Betz, H. (2003). Globalization, the welfare state and right-wing populism in Western Europe. SocioEconomic Review, 1(2), 215-245.
Taggart, P. (2000). Populism. Buckingham: Open University Press.
Tannen, D. (1990). You just don’t understand. women and men in conversation. New York: Ballantine.
Tarrow, S. (1998). Power in movement. Social movements and contentious politics. Cambridge, UK: Cambridge
University Press.
Thijssen, N. (2012). De jaren zestig herinnerd. Over gedeelde idealen uit een linkse periode. Amsterdam: Amsterdam
University Press.
Thomassen, J., Van Ham, C., & Andeweg, R. B. (2014). De wankele democratie. Heeft de democratie haar beste tijd
gehad? Amsterdam: Prometheus-Bert Bakker.
Tilburg University (2007-2014). Longitudinal Internet Studies for the Social sciences (LISS), Administered by
CentERdata, Tilburg, the Netherlands.
213
References
Togeby, L. (1994). Political implications of increasing numbers of women in the labor force. Comparative Political
Studies, 27(2), 211-240.
Učeň, P. (2007). Parties, populism, and anti-establishment politics in East Central Europe. SAIS Review, 17(1),
49-62.
Van Biezen, I., Mair, P., & Poguntke, T. (2012). Going, going, … gone? The decline of party membership in contemporary Europe. European Journal of Political Research, 51(1), 24-56.
Van der Brug, W. (2003). How the LPF fuelled discontent: Empirical tests of explanations of LPF support. Acta
Politica, 38(1), 89-106.
Van der Brug, W., & Fennema, M. (2007). What causes people to vote for a radical-right party? A review of recent
work. International Journal of Public Opinion, 19(4), 474-487.
Van der Brug, W., Fennema, M., & Tillie, J. (2005). Why some anti-immigrant parties fail and others succeed.
Comparative Political Studies, 38(5), 537-573.
Van der Brug, W., Hobolt, S. B., & De Vreese, C. H. (2009). Religion and party choice in Europe. West European
Politics, 32(6), 1266-1283.
Van der Brug, W., & Van Spanje, J. (2009). Immigration, Europe and the ‘new’ cultural dimension. European
Journal of Political Research, 48(3), 309-334.
Van der Meer, T., Lubbe, R., Van Elsas, E., Elff, M., & Van der Brug, W. (2012). Bounded volatility in the Dutch
electoral battlefield: A panel study on the structure of changing vote intentions in the Netherlands during
2006-2010. Acta Politica, 47(4), 333-355.
Van der Slik, F. W. P., & Konig, R. P. (2006). Orthodox, humanitarian, and science-inspired belief in relation to
prejudice against Jews, Muslims, and ethnic minorities: The content of one’s belief does matter. International Journal for the Psychology of Religion, 16(2), 113-126.
Van Dooremalen, T. (2011). The (temporary) end of the consensual Dutch political culture? The case of the immigration and integration issue. Unpublished Master Thesis, University of Amsterdam.
Van Holsteijn, J. J. M. (2011). The Dutch parliamentary election of 2010. West European Politics, 34(2), 412-419.
Van Holsteijn, J. J. M., & Andeweg, R. B. (2010). Demoted leaders and exiled candidates. Disentangling party and
person in the voter’s mind. Electoral Studies, 29(4), 628-635.
Van Holsteijn, J. J. M., & Irwin, G. A. (2003). Never a dull moment: Pim Fortuyn and the Dutch parliamentary
election of 2002. West European Politics, 26(2), 41-66.
Van Kersbergen, K., & Krouwel, A. (2008). A double-edged sword! The Dutch centre-right and the ‘foreigners
issue’. Journal of European Public Policy, 15(3), 398-414.
Van Leeuwen, M. (2009). ‘Clear’ vs. ‘woolly’ language use in political speeches: The case of the controversial
Dutch politician Geert Wilders. Paper Presented at the Annual Conference of the Poetics and Linguistics
Association, July 28-August 1, Middelburg, the Netherlands. Retrieved September 1/2011 from: http://
www.pala.ac.uk/resources/proceedings/2009/vanleeuwen2009.pdf.
Van Reybrouck, D. (2013). Tegen verkiezingen [Against elections]. Amsterdam: De Bezige Bij.
Van Spanje, J. (2010). Contagious parties: Anti-immigrant parties and their impact on other parties’ immigration
stances in contemporary Western Europe. Party Politics, 16(5), 563-586.
Van Spanje, J. (2011). Keeping the rascals in: Anti-political-establishment parties and their cost governing in
established democracies. European Journal of Political Research, 50(5), 609-635.
Verba, S., Burns, N., & Schlozman, K. L. (1997). Knowing and caring about politics: Gender and political engagement. The Journal of Politics, 59(4), 1051-1072.
Verba, S., Schlozman, K. L., & Brady, H. E. (1995). Voice and equality: Civic voluntarism in American politics.
Cambridge, MA: Harvard University Press.
214
References
Vossen, K. (2010). Populism in the Netherlands after Fortuyn: Rita Verdonk and Geert Wilders compared. Perspectives on European Politics and Society, 11(1), 22-38.
Wagner, M. (2012). When do parties emphasize extreme positions? How strategic incentives for policy differentiation influence issue importance. European Journal of Political Research, 51(1), 64-88.
Weber, M. ([1930] 2003). The Protestant ethic and the spirit of capitalism. Dover: Dover Publications Inc.
Weber, M. (2009). The sociology of charismatic authority. In H. H. Gerth, & C. Wright Mills (Eds.), From Max
Weber. Essays in sociology (pp. 245-252). New York: Routledge.
Werts, H., Scheepers, P., & Lubbers, M. (2013). Euro-scepticism and radical right-wing voting in Europe,
2002–2008: Social cleavages, socio-political attitudes and contextual characteristics determining voting
for the radical right. European Union Politics, 14(2), 183-205.
Weyland, K. (2001). Clarifying a contested concept: Populism in the study of Latin American politics. Comparative
Politics, 34(1), 1-22.
Wilders, G. (2012). Hún Brussel, óns Nederland. Verkiezingsprogramma 2012 PVV [Their Brussels, our Netherlands.
Party manifesto 2012 Party of Freedom]. Retrieved March 1/2014 from: http://www.pvv.nl/index.php/visie/
verkiezingsprogramma-2012.html .
Wilkes, R., Guppy, N., & Farris, L. (2007). Right-wing parties and anti-foreigner sentiment in Europe. American
Sociological Review, 72(5), 831-840.
Williams, M. H. (2006). The impact of radical right-wing parties in West European democracies. New York: Palgrave
Macmillan.
Zaslove, A. (2004). Closing the door? The ideology and impact of radical right populism on immigration policy in
Austria and Italy. Journal of Political Ideologies, 9(1), 99-118.
Zaslove, A. (2008). Here to stay? Populism as a new party type. European Review, 16(3), 319-336.
Zhirkov, K. (2014). Nativist but not alienated: A comparative perspective on the radical right vote in Western
Europe. Party Politics, 20(2), 286-296.
Zweig, S. ([1943] 2011). Die Welt von gestern [De wereld van gisteren]. Utrecht: Arbeiderspers.
215
Acknowledgements
(Dankwoord)
217
Acknowledgements
‘Democratie is als klei: ze plooit zich naar de tijd. De concrete vormen die ze aanneemt zijn
altijd geboetseerd door historische omstandigheden’.
David van Reybrouck (2013) – Tegen verkiezingen, De Bezige Bij, Amsterdam: p. 139.
Het is goed om nu, na 4 jaar onderzoek, een punt te zetten achter een langdurig proces
van nadenken, opschrijven, herschrijven, bijschaven, integreren, en analyseren. Er waren
dagen dat ik me verwant voelde met de onderzoeker uit het boek ‘Het onderzoek’ van Philip
Claudel: ‘Nadenken is als het laten draaien van een lege wasmachine. Het helpt om je te
vergewissen dat het nog functioneert, maar aan het eind van de dag is de vuile was nog
steeds vuil’. Wat kon een dag zonder letter op papier of met enkel mislukte analyses voelen
als een zinloze draaiing om de hete brij. Maar wat kon zo’n dag ook nuttig zijn voor de
volgende morgen.
Schrijven en onderzoeken kosten tijd, zoals borduren draad kost (Alexis Jenni, 2013, ‘In
tijden van Oorlog’). De concrete vorm die het proefschrift heeft aangenomen is zoals Van
Reybrouck’s democratie ‘geboetseerd’ door historische gebeurtenissen. Hadden migratie en
globalisering niet zo’n stempel gedrukt op onze hedendaagse Westerse samenlevingen, dan
had de roep om meer ‘eigen Volk eerst’ wellicht niet zo hard geklonken. Had Fortuyn met
zijn ‘At your service’ niet in 2001 het Nederlandse presidentschap tot missie uitgeroepen,
dan had mijn interesse voor politiek en samenleving zich misschien niet vertaald in een
studie sociologie. Had Kopspijkers niet al die politici prachtig gepersifleerd, dan had dit
proefschrift wel eens niet tot stand kunnen komen.
Opgeleid in een methodologisch individualistische school is het echter te makkelijk om
te schuilen achter zuiver de historische context. De ‘homo sociologicus’ handelt -al dan niet
rationeel- binnen de gegeven structurele, sociale en culturele omstandigheden, én is door
te handelen in staat om invloed uit te oefenen op die omstandigheden. Voilà, daar is het
‘Coleman-bootje’ uit het eerste college Sociale Problemen dat de rode lijn van mijn studie
en academische carrière zou vormen.
Graag wil ik hier diegenen bedanken die mij de instrumenten gaven om te kunnen
‘boetseren’ om dit proefschrift de uiteindelijke vorm te geven. Allereerst zijn dat natuurlijk
mijn begeleiders: Eva, Marcel en Tanja. Eva, dank je wel voor de fijne samenwerking. Jij
relativeerde, legde de vinger op de zere plek, kwam vaak vrolijk de kamer binnen om te
vragen hoe het ging en wist de juiste kritische vragen te stellen als dat nodig was. Het laatste
jaar is een moeilijke fase in je leven aangebroken. Ondanks dat ben je nauw betrokken
gebleven bij mijn voortgang. Ik ben je daar dankbaar voor en wil je veel sterkte wensen de
komende tijd. Marcel, jij hebt me doen inzien dat wetenschap meer is dan zomaar wat hypotheses afleiden: het gaat erom dat je meer wilt weten over hoe de werkelijkheid in elkaar
steekt. Dan moet je goed en kritisch nadenken over de mechanismen die ten grondslag
219
Acknowledgements
liggen aan je hypotheses. Ik wil je bedanken voor je inhoudelijke commentaren, de prettige
en effectieve samenwerking en de intensievere begeleiding in het laatste jaar. Tanja, naast
het feit dat jij me liet kennis maken met de Coleman-boot, wil ik je vooral bedanken voor
je fantastische begeleiding. Je las alles grondig, leverde commentaar dat meer dan eens
kopzorgen veroorzaakte, maar gaf bovenal veel vertrouwen. Ik kreeg alle ruimte om me
te ontwikkelen (‘Tim, is die Institutie-commissie niets voor jou?!’) en mijn proefschrift de
kant op te sturen die ik graag wilde. Heel veel dank daarvoor!
Beste Hilde, na ons eerste artikel vertrok jij naar Nieuw-Zeeland waardoor je niet meer
betrokken was bij de verdere begeleiding van mijn promotie. Dat vond ik jammer, maar
heeft uiteindelijk een verdere samenwerking niet in de weg gestaan. Wat heb ik veel geleerd
van jouw nauwkeurige formuleringen bij reacties aan reviewers en de zoveelste versie van
ons gender-paper. Met veel plezier denk ik ook terug aan onze IJsland-trip, dank daarvoor!
Dear Mark, in the fall of 2012 I had the fortunate opportunity to visit you, Eline and
David at the Simon Fraser University in Vancouver. We worked together on Chapter 3 of
this dissertation in coffee bars with some of the most terrible coffees ever (coffee snob,
right?!). I really liked the easy-going cooperation and learnt a great deal about pseudopanel approaches, Hausman-tests and the like. Although our paper has not been published
yet, it was a fine format for a course on longitudinal multilevel modelling that we taught
together in Slovenia. Thanks, it’s been a great pleasure to work with you!
Dear Yolanda and Ron, living together with you made the Vancouver-experience even
more ‘wonderful’! You gave veganism a more human touch, showed that biking without
helmets can in fact be a matter of judicial debate and taught me that beaches should also be
visited during winter time. I look forward to seeing you again.
Naast mijn begeleiders waren er natuurlijk ook andere collega’s en mede-aio’s die het
schrijven van een proefschrift makkelijker maakten. Esther, paranimf en favoriete (ex-) collega, wat is het heerlijk om uit eten te gaan, bioscopen plat te lopen en eindeloos vaak de cola
light’s en zero’s uit de schappen van de Spar te plukken. Die dissertatie werd er beter door.
Dank voor je vriendschap, aanmoedigingen en steun! Wike, Zolí, Sanne, Mariska: op onze
kamer was het soms stil, maar werd er toch voornamelijk gediscussieerd, hardop nagedacht
en gelachen. Het is mooi dat we nu als vrienden elkaar blijven volgen. Natuurlijk wil ik
ook Bärbel, Pim, Ingrid, Marjet, Mariëlle en Ellen bedanken voor borrels, alumni-avonden,
biologische dadels en gewoon zomaar een praatje op het secretariaat. Mirjam, bedankt voor
het stellen van de juiste vragen. Sabrina, Sara, Antonie, Sarah, Nienke, Vincenz, Lieselot,
Anne, Jesper, Susanne, MaSS-seminar leden, en andere staf: bedankt voor de mooie tijd in
Utrecht!
Ook buiten de academie hebben zoveel mensen direct en indirect meegeholpen en
meegedacht. Ik denk aan de lieve vrienden van de Silo-Cantorij, de Rede & Religie-club en
natuurlijk de VRICK-ers. Machteld, Christien, Jef: mijn leven wordt mooier, voller en rijker
220
Acknowledgements
door jullie! Dat is ‘bijzonder’. Leonard, paranimf en sociologen-vriend van het allereerste
uur: onze gedeelde passie voor religie, politiek, samenleving en het verzamelen en (zoveel
mogelijk) lezen van veel boeken heeft mij veranderd en geïnspireerd. Als een Stefan Zweig
geniet ik van je anekdotes en sociologische analyses, jouw altijd rake observaties, humor en
veel betere muzieksmaak. Bedankt daarvoor!
Berend en Hermien, Jonien, Gert en Davy, Edmon en Nadie: het is fijn om een Drentse famile te hebben en weer opgeladen aan de slag te gaan na een weekend samen in
Wapserveen. Dat Drents zal ik ooit echt leren. Verder natuurlijk pa en ma: jullie vertrouwen, liefde en aansporingen om mezelf uit te dagen en te ontwikkelen zijn de basis geweest
voor dit boek. Er zijn niet voldoende woorden om jullie te bedanken en ik wil daarom
alleen maar zeggen dat ik van jullie houd. Lieve broers, Lennart, Peter en Rick, een gesprek
wordt pas een discussie als mensen het lekker hartgrondig met elkaar oneens zijn. Onze
discussies, uiteenlopend van voetbal tot politiek, vormden en scherpten mij. Ze maakten
dat ik als de ‘linkste’ van de vier radicaal rechts heb bestudeerd. Ik vind het mooi dat ik deze
mijlpaal met jullie alle drie mag vieren: het is ook een beetje jullie proefschrift. Lilian en
Nienke, dat jullie ondanks deze vaak heftige discussies niet besloten de Immerzelen vaarwel
te zeggen, vind ik geniaal. Ik ben nog dankbaarder voor het feit dat we door jullie wat meer
nuchtere feminiene tegenspraak krijgen. Want hoewel gender een continuüm is, zijn de
broer-discussies tamelijk masculien van aard.
Als laatste wil ik mijn allerliefste schat bedanken. Nan, wat hebben we al veel beleefd
samen. Jij zag me als ik het even niet meer zag zitten, jij zei me dat ik niet altijd ’s avonds
door moest werken en gaf en geeft me bovenal de liefde en vrijheid om tot hier en verder
te komen. Ik hou van je! Tot slot wil ik Erik noemen. Op de dag voor mijn sollicitatie bij
ICS belde hij mij op en sprak voor mij de historische woorden: ‘Tim, je moet dit echt doen.’
Erik, je bent er niet meer bij, maar dit boek draag ik op aan jou.
221
Curriculum Vitae
223
Curriculum vitae
Tim Immerzeel was born in Vlaardingen, the Netherlands, on June 30 1986. After his
bachelor’s degree in Sociology (2008), he obtained his Research Master’s degree in the
international prestige master programme Sociology and Social Research (cum laude, 2010)
at Utrecht University. In September 2010, Tim started working as a PhD candidate at the
Interuniversity Center for Social Science Theory and Methodology (ICS) at the same university investigating the consequences of populist radical right success on voting behavior.
During his PhD, he supervised master- and bachelorstudents in Sociology in writing their
theses, taught various courses on social capital and networks, statistics, and social policy,
and was tutor of second- and third-year bachelorstudents. As member of the Education Advisory Committee and President of the Alumni Committee, he was also involved in policy
advice on various terrains. From September to December 2012, Tim was a visiting scholar
at the Simon Fraser University, in Vancouver, Canada, where he was hosted by Dr. Mark
Pickup. Together with Mark Pikcup he taught a course on longitudinal modelling with
the European Social Survey for an international audience in April 2013. Since September
2014, Tim has been working as a postdoctoral researcher at the Department of Sociology
of the VU University Amsterdam. He works on an international research project that aims
to answer the question ‘how people try to influence politics and why’. In this project, he is
responsible for a meta-analytic study reporting on the relationship between engagement in
party politics and movement politics.
225
ICS dissertation series
227
ICS Dissertation Series
The ICS series presents dissertations of the Interuniversity Center for Social Science Theory
and Methodology. Each of these studies aims at integrating explicit theory formation with
state of the art empirical research or at the development of advanced methods for empirical
research. The ICS was founded in 1986 as a cooperative effort of the universities of Groningen and Utrecht. Since 1992, the ICS expanded to the University of Nijmegen. Most of the
projects are financed by the participating universities or by the Netherlands Organization
for Scientific Research (NWO). The international composition of the ICS graduate students
is mirrored in the increasing international orientation of the projects and thus of the ICS
series itself.
1)
2)
3)
4)
5)
6)
7)
8)
9)
10)
11)
12)
13)
Cornelis van Liere (1990). Lastige leerlingen. Een empirisch onderzoek naar sociale
oorzaken van probleemgedrag op basisscholen. Amsterdam: Thesis Publishers.
Marco H.D. van Leeuwen (1990). Bijstand in Amsterdam, ca. 1800-1850. Armenzorg
als beheersings- en overlevingsstrategie. ICS-dissertation, Utrecht.
Ineke Maas (1990). Deelname aan podiumkunsten via de podia, de media en actieve
beoefening. Substitutie of leereffecten? Amsterdam: Thesis Publishers.
Marjolein I. Broese van Groenou (1991). Gescheiden netwerken. De relaties met vrienden en verwanten na echtscheiding. Amsterdam: Thesis Publishers.
Jan M.M. van den Bos (1991). Dutch EC policy making. A model guided approach to
coordination and negotiation. Amsterdam: Thesis Publishers.
Karin Sanders (1991). Vrouwelijke pioniers. Vrouwen en mannen met een ‘mannelijke’
hogere beroepsopleiding aan het begin van hun loopbaan. Amsterdam: Thesis Publishers.
Sjerp de Vries (1991). Egoism, altruism, and social justice. Theory and experiments on
cooperation in social dilemmas. Amsterdam: Thesis Publishers.
Ronald S. Batenburg (1991). Automatisering in bedrijf. Amsterdam: Thesis Publishers.
Rudi Wielers (1991). Selectie en allocatie op de arbeidsmarkt. Een uitwerking voor de
informele en geïnstitutionaliseerde kinderopvang. Amsterdam: Thesis Publishers.
Gert P. Westert (1991). Verschillen in ziekenhuisgebruik. ICS-dissertation, Groningen.
Hanneke Hermsen (1992). Votes and policy preferences. Equilibria in party systems.
Amsterdam: Thesis Publishers.
Cora J.M. Maas (1992). Probleemleerlingen in het basisonderwijs. Amsterdam: Thesis
Publishers.
Ed A.W. Boxman (1992). Contacten en carrière. Een empirisch theoretisch onderzoek
naar de relatie tussen sociale netwerken en arbeidsmarktposities. Amsterdam: Thesis
Publishers.
229
ICS Dissertation Series
14) Conny G.J. Taes (1992). Kijken naar banen. Een onderzoek naar de inschatting van arbeidsmarktkansen bij schoolverlaters uit het middelbaar beroepsonderwijs. Amsterdam:
Thesis Publishers.
15) Peter van Roozendaal (1992). Cabinets in multi party democracies. The effect of dominant and central parties on cabinet composition and durability. Amsterdam: Thesis
Publishers.
16) Marcel van Dam (1992). Regio zonder regie. Verschillen in en effectiviteit van gemeentelijk arbeidsmarktbeleid. Amsterdam: Thesis Publishers.
17) Tanja van der Lippe (1993). Arbeidsverdeling tussen mannen en vrouwen. Amsterdam:
Thesis Publishers.
18) Marc A. Jacobs (1993). Software: Kopen of kopiëren? Een sociaal wetenschappelijk
onderzoek onder PC gebruikers. Amsterdam: Thesis Publishers.
19) Peter van der Meer (1993). Verdringing op de Nederlandse arbeidsmarkt. Sector- en
sekseverschillen. Amsterdam: Thesis Publishers.
20) Gerbert Kraaykamp (1993). Over lezen gesproken. Een studie naar sociale differentiatie
in leesgedrag. Amsterdam: Thesis Publishers.
21) Evelien Zeggelink (1993). Strangers into friends. The evolution of friendship networks
using an individual oriented modeling approach. Amsterdam: Thesis Publishers.
22) Jaco Berveling (1994). Het stempel op de besluitvorming. Macht, invloed en besluitvorming op twee Amsterdamse beleidsterreinen. Amsterdam: Thesis Publishers.
23) Wim Bernasco (1994). Coupled careers. The effects of spouse’s resources on success at
work. Amsterdam: Thesis Publishers.
24) Liset van Dijk (1994). Choices in child care. The distribution of child care among mothers, fathers and non parental care providers. Amsterdam: Thesis Publishers.
25) Jos de Haan (1994). Research groups in Dutch sociology. Amsterdam: Thesis Publishers.
26) Kwasi Boahene (1995). Innovation adoption as a socio economic process. The case of the
Ghanaian cocoa industry. Amsterdam: Thesis Publishers.
27) Paul E.M. Ligthart (1995). Solidarity in economic transactions. An experimental study
of framing effects in bargaining and contracting. Amsterdam: Thesis Publishers.
28) Roger Th. A.J. Leenders (1995). Structure and influence. Statistical models for the dynamics of actor attributes, network structure, and their interdependence. Amsterdam:
Thesis Publishers.
29) Beate Völker (1995). Should auld acquaintance be forgot…? Institutions of communism,
the transition to capitalism and personal networks: the case of East Germany. Amsterdam: Thesis Publishers.
230
ICS Dissertation Series
30) Anna M. Cancrinus Matthijsse (1995). Tussen hulpverlening en ondernemerschap.
Beroepsuitoefening en taakopvattingen van openbare apothekers in een aantal West
Europese landen. Amsterdam: Thesis Publishers.
31) Nardi Steverink (1996). Zo lang mogelijk zelfstandig. Naar een verklaring van verschillen in oriëntatie ten aanzien van opname in een verzorgingstehuis onder fysiek kwetsbare
ouderen. Amsterdam: Thesis Publishers.
32) Ellen Lindeman (1996). Participatie in vrijwilligerswerk. Amsterdam: Thesis Publishers.
33) Chris Snijders (1996). Trust and commitments. Amsterdam: Thesis Publishers.
34) Koos Postma (1996). Changing prejudice in Hungary. A study on the collapse of state
socialism and its impact on prejudice against gypsies and Jews. Amsterdam: Thesis
Publishers.
35) Jooske T. van Busschbach (1996). Uit het oog, uit het hart? Stabiliteit en verandering in
persoonlijke relaties. Amsterdam: Thesis Publishers.
36) René Torenvlied (1996). Besluiten in uitvoering. Theorieën over beleidsuitvoering
modelmatig getoetst op sociale vernieuwing in drie gemeenten. Amsterdam: Thesis
Publishers.
37) Andreas Flache (1996). The double edge of networks. An analysis of the effect of informal networks on cooperation in social dilemmas. Amsterdam: Thesis Publishers.
38) Kees van Veen (1997). Inside an internal labor market: Formal rules, flexibility and
career lines in a Dutch manufacturing company. Amsterdam: Thesis Publishers.
39) Lucienne van Eijk (1997). Activity and well being in the elderly. Amsterdam: Thesis
Publishers.
40) Róbert Gál (1997). Unreliability. Contract discipline and contract governance under
economic transition. Amsterdam: Thesis Publishers.
41) Anne Geerte van de Goor (1997). Effects of regulation on disability duration. ICSdissertation, Utrecht.
42) Boris Blumberg (1997). Das Management von Technologiekooperationen. Partnersuche
und Verhandlungen mit dem Partner aus empirisch theoretischer Perspektive. ICSdissertation, Utrecht.
43) Marijke von Bergh (1997). Loopbanen van oudere werknemers. Amsterdam: Thesis
Publishers.
44) Anna Petra Nieboer (1997). Life events and well being: A prospective study on changes
in well being of elderly people due to a serious illness event or death of the spouse. Amsterdam: Thesis Publishers.
45) Jacques Niehof (1997). Resources and social reproduction: The effects of cultural and
material resources on educational and occupational careers in industrial nations at the
end of the twentieth century. ICS-dissertation, Nijmegen.
231
ICS Dissertation Series
46) Ariana Need (1997). The kindred vote. Individual and family effects of social class and
religion on electoral change in the Netherlands, 1956-1994. ICS-dissertation, Nijmegen.
47) Jim Allen (1997). Sector composition and the effect of education on wages: an international Comparison. Amsterdam: Thesis Publishers.
48) Jack B.F. Hutten (1998). Workload and provision of care in general practice. An empirical study of the relation between workload of Dutch general practitioners and the
content and quality of their Care. ICS-dissertation, Utrecht.
49) Per B. Kropp (1998). Berufserfolg im Transformationsprozeß. Eine theoretisch empirische Studie über die Gewinner und Verlierer der Wende in Ostdeutschland. ICSdissertation, Utrecht.
50) Maarten H.J. Wolbers (1998). Diploma inflatie en verdringing op de arbeidsmarkt.
Een studie naar ontwikkelingen in de opbrengsten van diploma’s in Nederland. ICSdissertation, Nijmegen.
51) Wilma Smeenk (1998). Opportunity and marriage. The impact of individual resources
and marriage market structure on first marriage timing and partner choice in the Netherlands. ICS-dissertation, Nijmegen.
52) Marinus Spreen (1999). Sampling personal network structures: Statistical inference in
ego-graphs. ICS-dissertation, Groningen.
53) Vincent Buskens (1999). Social networks and trust. ICS-dissertation, Utrecht.
54) Susanne Rijken (1999). Educational expansion and status attainment. A cross-national
and over-time comparison. ICS-dissertation, Utrecht.
55) Mérove Gijsberts (1999). The legitimation of inequality in state-socialist and market
societies, 1987-1996. ICS-dissertation, Utrecht.
56) Gerhard G. Van de Bunt (1999). Friends by choice. An actor-oriented statistical network
model for friendship networks through time. ICS-dissertation, Groningen.
57) Robert Thomson (1999). The party mandate: Election pledges and government actions
in the Netherlands, 1986-1998. Amsterdam: Thela Thesis.
58) Corine Baarda (1999). Politieke besluiten en boeren beslissingen. Het draagvlak van het
mestbeleid tot 2000. ICS-dissertation, Groningen.
59) Rafael Wittek (1999). Interdependence and informal control in organizations. ICSdissertation, Groningen.
60) Diane Payne (1999). Policy making in the European Union: An analysis of the impact of
the reform of the structural funds in Ireland. ICS-dissertation, Groningen.
61) René Veenstra (1999). Leerlingen-klassen-scholen. Prestaties en vorderingen van leerlingen in het voortgezet onderwijs. Amsterdam, Thela Thesis.
62) Marjolein Achterkamp (1999). Influence strategies in collective decision making. A
comparison of two models. ICS-dissertation, Groningen.
232
ICS Dissertation Series
63) Peter Mühlau (2000). The governance of the employment relation. A relational signaling
perspective. ICS-dissertation, Groningen.
64) Agnes Akkerman (2000). Verdeelde vakbeweging en stakingen. Concurrentie om leden.
ICS-dissertation, Groningen.
65) Sandra van Thiel (2000). Quangocratization: Trends, causes and consequences. ICSdissertation, Utrecht.
66) Rudi Turksema (2000). Supply of day care. ICS-dissertation, Utrecht.
67) Sylvia E. Korupp (2000). Mothers and the process of social stratification. ICS-dissertation, Utrecht.
68) Bernard A. Nijstad (2000). How the group affects the mind: Effects of communication in
idea generating groups. ICS-dissertation, Utrecht.
69) Inge F. de Wolf (2000). Opleidingsspecialisatie en arbeidsmarktsucces van sociale
wetenschappers. ICS-dissertation, Utrecht.
70) Jan Kratzer (2001). Communication and performance: An empirical study in innovation teams. ICS-dissertation, Groningen.
71) Madelon Kroneman (2001). Healthcare systems and hospital bed use. ICS/NIVELdissertation, Utrecht.
72) Herman van de Werfhorst (2001). Field of study and social inequality. Four types of
educational resources in the process of stratification in the Netherlands. ICS-dissertation,
Nijmegen.
73) Tamás Bartus (2001). Social capital and earnings inequalities. The role of informal job
search in Hungary. ICS-dissertation Groningen.
74) Hester Moerbeek (2001). Friends and foes in the occupational career. The influence of
sweet and sour social capital on the labour market. ICS-dissertation, Nijmegen.
75) Marcel van Assen (2001). Essays on actor perspectives in exchange networks and social
dilemmas. ICS-dissertation, Groningen.
76) Inge Sieben (2001). Sibling similarities and social stratification. The impact of family
background across countries and cohorts. ICS-dissertation, Nijmegen.
77) Alinda van Bruggen (2001). Individual production of social well being. An exploratory
study. ICS-dissertation, Groningen.
78) Marcel Coenders (2001). Nationalistic attitudes and ethnic exclusionism in a comparative perspective: An empirical study of attitudes toward the country and ethnic immigrants in 22 countries. ICS-dissertation, Nijmegen.
79) Marcel Lubbers (2001). Exclusionistic electorates. Extreme right wing voting in Western
Europe. ICS-dissertation, Nijmegen.
80) Uwe Matzat (2001). Social networks and cooperation in electronic communities. A
theoretical-empirical analysis of academic communication and internet discussion
groups. ICS-dissertation, Groningen.
233
ICS Dissertation Series
81) Jacques P.G. Janssen (2002). Do opposites attract divorce? Dimensions of mixed marriage and the risk of divorce in the Netherlands. ICS-dissertation, Nijmegen.
82) Miranda Jansen (2002). Waardenoriëntaties en partnerrelaties. Een panelstudie naar
wederzijdse invloeden. ICS-dissertation, Utrecht.
83) Anne Rigt Poortman (2002). Socioeconomic causes and consequences of divorce. ICSdissertation, Utrecht.
84) Alexander Gattig (2002). Intertemporal decision making. ICS-dissertation, Groningen.
85) Gerrit Rooks (2002). Contract en conflict: Strategisch Management van Inkooptransacties. ICS-dissertation, Utrecht.
86) Károly Takács (2002). Social networks and intergroup conflict. ICS-dissertation, Groningen.
87) Thomas Gautschi (2002). Trust and exchange, effects of temporal embeddedness and
network embeddedness on providing and dividing a surplus. ICS-dissertation, Utrecht.
88) Hilde Bras (2002). Zeeuwse meiden. Dienen in de levensloop van vrouwen, ca. 18501950. Aksant Academic Publishers, Amsterdam.
89) Merijn Rengers (2002). Economic lives of artists. Studies into careers and the labour
market in the cultural sector. ICS-dissertation, Utrecht.
90) Annelies Kassenberg (2002). Wat scholieren bindt. Sociale gemeenschap in scholen.
ICS-dissertation, Groningen.
91) Marc Verboord (2003). Moet de meester dalen of de leerling klimmen? De invloed
van literatuuronderwijs en ouders op het lezen van boeken tussen 1975 en 2000. ICSdissertation, Utrecht.
92) Marcel van Egmond (2003). Rain falls on all of us (but some manage to get more wet
than others): Political context and electoral participation. ICS-dissertation, Nijmegen.
93) Justine Horgan (2003). High performance human resource management in Ireland and
the Netherlands: Adoption and effectiveness. ICS-dissertation, Groningen.
94) Corine Hoeben (2003). LETS’ be a community. Community in Local Exchange Trading
Systems. ICS-dissertation, Groningen.
95) Christian Steglich (2003). The framing of decision situations. Automatic goal selection
and rational goal pursuit. ICS-dissertation, Groningen.
96) Johan van Wilsem (2003). Crime and context. The impact of individual, neighborhood,
city and country characteristics on victimization. ICS-dissertation, Nijmegen.
97) Christiaan Monden (2003). Education, inequality and health. The impact of partners
and life course. ICS-dissertation, Nijmegen.
98) Evelyn Hello (2003). Educational attainment and ethnic attitudes. How to explain their
relationship. ICS-dissertation, Nijmegen.
234
ICS Dissertation Series
99) Marnix Croes en Peter Tammes (2004). Gif laten wij niet voortbestaan. Een onderzoek
naar de overlevingskansen van joden in de Nederlandse gemeenten, 1940-1945. Aksant
Academic Publishers, Amsterdam.
100) Ineke Nagel (2004). Cultuurdeelname in de levensloop. ICS-dissertation, Utrecht.
101) Marieke van der Wal (2004). Competencies to participate in life. Measurement and the
impact of school. ICS-dissertation, Groningen.
102) Vivian Meertens (2004). Depressive symptoms in the general population: a multifactorial social approach. ICS-dissertation, Nijmegen.
103) Hanneke Schuurmans (2004). Promoting well-being in frail elderly people. Theory and
intervention. ICS-dissertation, Groningen.
104) Javier Arregui (2004). Negotiation in legislative decision-making in the European
Union. ICS-dissertation, Groningen.
105) Tamar Fischer (2004). Parental divorce, conflict and resources. The effects on children’s
behaviour problems, socioeconomic attainment, and transitions in the demographic
career. ICS-dissertation, Nijmegen.
106) René Bekkers (2004). Giving and volunteering in the Netherlands: Sociological and
psychological perspectives. ICS-dissertation, Utrecht.
107) Renée van der Hulst (2004). Gender differences in workplace authority: An empirical
study on social networks. ICS-dissertation, Groningen.
108) Rita Smaniotto (2004). ‘You scratch my back and I scratch yours’ versus ‘Love Thy neighbour’. Two proximate mechanisms of reciprocal altruism. ICS-dissertation, Groningen.
109) Maurice Gesthuizen (2004). The life-course of the low-educated in the Netherlands:
Social and economic risks. ICS-dissertation, Nijmegen.
110) Carlijne Philips (2005). Vakantiegemeenschappen. Kwalitatief en kwantitatief onderzoek naar gelegenheid en refreshergemeenschap tijdens de vakantie. ICS-dissertation,
Groningen.
111) Esther de Ruijter (2005). Household outsourcing. ICS-dissertation, Utrecht.
112) Frank van Tubergen (2005). The integration of immigrants in cross-national perspective: Origin, destination, and community effects. ICS-dissertation, Utrecht.
113) Ferry Koster (2005). For the time being. Accounting for inconclusive findings concerning
the effects of temporary employment relationships on solidary behavior of employees.
ICS-dissertation, Groningen.
114) Carolien Klein Haarhuis (2005). Promoting anti-corruption reforms. Evaluating the
implementation of a World Bank anti-corruption program in seven African countries
(1999-2001). ICS-dissertation, Utrecht.
115) Martin van der Gaag (2005). Measurement of individual social capital. ICS-dissertation, Groningen.
235
ICS Dissertation Series
116) Johan Hansen (2005). Shaping careers of men and women in organizational contexts.
ICS-dissertation, Utrecht.
117) Davide Barrera (2005). Trust in embedded settings. ICS-dissertation, Utrecht.
118) Mattijs Lambooij (2005). Promoting cooperation. Studies into the effects of long-term
and short-term rewards on cooperation of employees. ICS-dissertation, Utrecht.
119) Lotte Vermeij (2006). What’s cooking? Cultural boundaries among Dutch teenagers of
different ethnic origins in the context of school. ICS-dissertation, Utrecht.
120) Mathilde Strating (2006). Facing the challenge of rheumatoid arthritis. A 13-year
prospective study among patients and a cross-sectional study among their partners. ICSdissertation, Groningen.
121) Jannes de Vries (2006). Measurement error in family background variables: The bias
in the intergenerational transmission of status, cultural consumption, party preference,
and religiosity. ICS-dissertation, Nijmegen.
122) Stefan Thau (2006). Workplace deviance: Four studies on employee motives and selfregulation. ICS-dissertation, Groningen.
123) Mirjam Plantinga (2006). Employee motivation and employee performance in child
care. The effects of the introduction of market forces on employees in the Dutch childcare sector. ICS-dissertation, Groningen.
124) Helga de Valk (2006). Pathways into adulthood. A comparative study on family life
transitions among migrant and Dutch Youth. ICS-dissertation, Utrecht.
125) Henrike Elzen (2006). Self-management for chronically ill older people. ICS-Dissertation, Groningen.
126) Ayse Güveli (2007). New social classes within the service class in the Netherlands and
Britain. Adjusting the EGP class schema for the technocrats and the social and cultural
specialists. ICS-dissertation, Nijmegen.
127) Willem-Jan Verhoeven (2007). Income attainment in post-communist societies. ICSdissertation, Utrecht.
128) Marieke Voorpostel (2007). Sibling support: The exchange of help among brothers and
sisters in the Netherlands. ICS-dissertation, Utrecht.
129) Jacob Dijkstra (2007). The effects of externalities on partner choice and payoffs in exchange networks. ICS-dissertation, Groningen.
130) Patricia van Echtelt (2007). Time-greedy employment relationships: Four studies on the
time claims of post-Fordist work. ICS-dissertation, Groningen.
131) Sonja Vogt (2007). Heterogeneity in social dilemmas: The case of social support. ICSdissertation, Utrecht.
132) Michael Schweinberger (2007). Statistical methods for studying the evolution of networks and behavior. ICS-dissertation, Groningen.
236
ICS Dissertation Series
133) István Back (2007). Commitment and evolution: Connecting emotion and reason in
long-term relationships. ICS-dissertation, Groningen.
134) Ruben van Gaalen (2007). Solidarity and ambivalence in parent-child relationships.
ICS-dissertation, Utrecht.
135) Jan Reitsma (2007). Religiosity and solidarity – Dimensions and relationships disentangled and tested. ICS-dissertation, Nijmegen.
136) Jan Kornelis Dijkstra (2007). Status and affection among (pre)adolescents and their
relation with antisocial and prosocial behavior. ICS-dissertation, Groningen.
137) Wouter van Gils (2007). Full-time working couples in the Netherlands. Causes and
consequences. ICS-dissertation, Nijmegen.
138) Djamila Schans (2007). Ethnic diversity in intergenerational solidarity. ICS-dissertation, Utrecht.
139) Ruud van der Meulen (2007). Brug over woelig water: Lidmaatschap van sportverenigingen, vriendschappen, kennissenkringen en veralgemeend vertrouwen. ICS-dissertation,
Nijmegen.
140) Andrea Knecht (2008). Friendship selection and friends’ influence. Dynamics of networks and actor attributes in early adolescence. ICS-dissertation, Utrecht.
141) Ingrid Doorten (2008). The division of unpaid work in the household: A stubborn pattern? ICS-dissertation, Utrecht.
142) Stijn Ruiter (2008). Association in context and association as context: Causes and
consequences of voluntary association involvement. ICS-dissertation, Nijmegen.
143) Janneke Joly (2008). People on our minds: When humanized contexts activate social
norms. ICS-dissertation, Groningen.
144) Margreet Frieling (2008). ‘Joint production’ als motor voor actief burgerschap in de
buurt. ICS-dissertation, Groningen.
145) Ellen Verbakel (2008). The partner as resource or restriction? Labour market careers
of husbands and wives and the consequences for inequality between couples. ICSdissertation, Nijmegen.
146) Gijs van Houten (2008). Beleidsuitvoering in gelaagde stelsels. De doorwerking van
aanbevelingen van de Stichting van de Arbeid in het CAO-overleg. ICS-dissertation,
Utrecht.
147) Eva Jaspers (2008). Intolerance over time. Macro and micro level questions on attitudes
towards euthanasia, homosexuality and ethnic minorities. ICS-dissertation, Nijmegen.
148) Gijs Weijters (2008). Youth delinquency in Dutch cities and schools: A multilevel approach. ICS-dissertation, Nijmegen.
149) Jessica Pass (2009). The self in social rejection. ICS-dissertation, Groningen.
150) Gerald Mollenhorst (2009). Networks in contexts. How meeting opportunities affect
personal relationships. ICS-dissertation, Utrecht.
237
ICS Dissertation Series
151) Tom van der Meer (2009). States of freely associating citizens: Comparative studies
into the impact of state institutions on social, civic and political participation. ICSdissertation, Nijmegen.
152) Manuela Vieth (2009). Commitments and reciprocity in trust situations. Experimental
studies on obligation, indignation, and self-consistency. ICS-dissertation, Utrecht.
153) Rense Corten (2009). Co-evolution of social networks and behavior in social dilemmas:
Theoretical and empirical perspectives. ICS-dissertation, Utrecht.
154) Arieke J. Rijken (2009). Happy families, high fertility? Childbearing choices in the
context of family and partner relationships. ICS-dissertation, Utrecht.
155) Jochem Tolsma (2009). Ethnic hostility among ethnic majority and minority groups in
the Netherlands. An investigation into the impact of social mobility experiences, the local living environment and educational attainment on ethnic hostility. ICS-dissertation,
Nijmegen.
156) Freek Bucx (2009). Linked lives: Young adults’ life course and relations with parents.
ICS-dissertation, Utrecht.
157) Philip Wotschack (2009). Household governance and time allocation. Four studies on
the combination of work and care. ICS-dissertation, Groningen.
158) Nienke Moor (2009). Explaining worldwide religious diversity. The relationship between
subsistence technologies and ideas about the unknown in pre-industrial and (post-)
industrial societies. ICS-dissertation, Nijmegen.
159) Lieke ten Brummelhuis (2009). Family matters at work. Depleting and enriching effects
of employees’ family lives on work outcomes. ICS-dissertation, Utrecht.
160) Renske Keizer (2010). Remaining childless. Causes and consequences from a life Course
Perspective. ICS-dissertation, Utrecht.
161) Miranda Sentse (2010). Bridging contexts: The interplay between family, child, and peers
in explaining problem behavior in Eearly adolescence. ICS-dissertation, Groningen.
162) Nicole Tieben (2010). Transitions, tracks and transformations. Social inequality in
transitions into, through and out of secondary education in the Netherlands for cohorts
born between 1914 and 1985. ICS-dissertation, Nijmegen.
163) Birgit Pauksztat (2010). Speaking up in organizations: Four studies on employee voice.
ICS-dissertation, Groningen.
164) Richard Zijdeman (2010). Status attainment in the Netherlands, 1811-1941. Spatial
and temporal variation before and during industrialization. ICS-dissertation, Utrecht.
165) Rianne Kloosterman (2010). Social background and children’s educational careers. The
primary and secondary effects of social background over transitions and over time in the
Netherlands. ICS-dissertation, Nijmegen.
238
ICS Dissertation Series
166) Olav Aarts (2010). Religious diversity and religious involvement. A study of religious
markets in Western societies at the end of the twentieth century. ICS-dissertation,
Nijmegen.
167) Stephanie Wiesmann (2010). 24/7 Negotiation in couples transition to parenthood.
ICS-dissertation, Utrecht.
168) Borja Martinovic (2010). Interethnic contacts: A dynamic analysis of interaction between immigrants and natives in Western countries. ICS-dissertation, Utrecht.
169) Anne Roeters (2010). Family life under pressure? Parents’ paid work and the quantity
and quality of parent-child and family time. ICS-dissertation, Utrecht.
170) Jelle Sijtsema (2010). Adolescent aggressive behavior: Status and stimulation goals in
relation to the peer context. ICS-dissertation, Groningen.
171) Kees Keizer (2010). The spreading of disorder. ICS-dissertation, Groningen.
172) Michael Mäs (2010). The diversity puzzle. Explaining clustering and polarization of
opinions. ICS-dissertation, Groningen.
173) Marie-Louise Damen (2010). Cultuurdeelname en CKV. Studies naar effecten van kunsteducatie op de cultuurdeelname van leerlingen tijdens en na het voortgezet onderwijs.
ICS-dissertation, Utrecht.
174) Marieke van de Rakt (2011). Two generations of crime: The intergenerational transmission of convictions over the life course. ICS-dissertation, Nijmegen.
175) Willem Huijnk (2011). Family life and ethnic attitudes. The role of the family for attitudes towards intermarriage and acculturation among minority and majority groups.
ICS-dissertation, Utrecht.
176) Tim Huijts (2011). Social ties and health in Europe. Individual associations, crossnational variations, and contextual explanations. ICS-dissertation, Nijmegen.
177) Wouter Steenbeek (2011). Social and physical Disorder. How community, business presence and entrepreneurs influence disorder in Dutch neighborhoods. ICS-dissertation,
Utrecht.
178) Miranda Vervoort (2011). Living together apart? Ethnic concentration in the neighborhood and ethnic minorities’ social contacts and language practices. ICS-dissertation,
Utrecht.
179) Agnieszka Kanas (2011). The economic performance of immigrants. The role of human
and social capital. ICS-dissertation, Utrecht.
180) Lea Ellwardt (2011). Gossip in organizations. A social network study. ICS-dissertation,
Groningen.
181) Annemarije Oosterwaal (2011). The gap between decision and implementation. Decision making, delegation and compliance in governmental and organizational settings.
ICS-dissertation, Utrecht.
239
ICS Dissertation Series
182) Natascha Notten (2011). Parents and the media. Causes and consequences of parental
media socialization. ICS-dissertation, Nijmegen.
183) Tobias Stark (2011). Integration in schools. A process perspective on students’ interethnic
attitudes and interpersonal relationships. ICS-dissertation, Groningen.
184) Giedo Jansen (2011). Social cleavages and political choices. Large-scale comparisons
of social class, religion and voting behavior in Western democracies. ICS-dissertation,
Nijmegen.
185) Ruud van der Horst (2011). Network effects on treatment results in a closed forensic
psychiatric setting. ICS-dissertation, Groningen.
186) Mark Levels (2011). Abortion laws in European countries between 1960 and 2010. Legislative developments and their consequences for women’s reproductive decision-making.
ICS-dissertation, Nijmegen.
187) Marieke van Londen (2012). Exclusion of ethnic minorities in the Netherlands. The
effects of individual and situational characteristics on opposition to ethnic policy and
ethnically mixed neighbourhoods. ICS-dissertation, Nijmegen.
188) Sigrid M. Mohnen (2012). Neighborhood context and health: How neighborhood social
capital affects individual health. ICS-dissertation, Utrecht.
189) Asya Zhelyazkova (2012). Compliance under controversy: analysis of the transposition
of European directives and their provisions. ICS-dissertation, Utrecht.
190) Valeska Korff (2012). Between cause and control: Management in a humanitarian
organization. ICS-dissertation, Groningen.
191) Maike Gieling (2012). Dealing with diversity: Adolescents’ support for civil liberties and
immigrant rights. ICS-dissertation, Utrecht.
192) Katya Ivanova (2012). From parents to partners: The impact of family on romantic
relationships in adolescence and emerging adulthood. ICS-dissertation, Groningen.
193) Jelmer Schalk (2012). The performance of public corporate actors: Essays on effects of
institutional and network embeddedness in supranational, national, and local collaborative contexts. ICS-dissertation, Utrecht.
194) Alona Labun (2012). Social networks and informal power in organizations. ICSdissertation, Groningen.
195) Michal Bojanowski (2012). Essays on social network formation in heterogeneous populations: Models, methods, and empirical analyses. ICS-dissertation, Utrecht.
196) Anca Minescu (2012). Relative group position and intergroup attitudes in Russia. ICSdissertation, Utrecht.
197) Marieke van Schellen (2012). Marriage and crime over the life course. The criminal
careers of convicts and their spouses. ICS-dissertation, Utrecht.
198) Mieke Maliepaard (2012). Religious trends and social integration: Muslim minorities in
the Netherlands. ICS-dissertation, Utrecht.
240
ICS Dissertation Series
199) Fransje Smits (2012). Turks and Moroccans in the Low Countries around the year 2000:
Determinants of religiosity, trend in religiosity and determinants of the trend. ICSdissertation, Nijmegen.
200) Roderick Sluiter (2012). The diffusion of morality policies among Western European
countries between 1960 and 2010. A comparison of temporal and spatial diffusion patterns of six morality and eleven non-morality policies. ICS-dissertation, Nijmegen.
201) Nicoletta Balbo (2012). Family, friends and fertility. ICS-dissertation, Groningen.
202) Anke Munniksma (2013). Crossing ethnic boundaries: Parental resistance to and consequences of adolescents’ cross-ethnic peer relations. ICS-dissertation, Groningen.
203) Anja Abendroth (2013). Working women in Europe. How the country, workplace, and
family context matter. ICS-dissertation, Utrecht.
204) Katia Begall (2013). Occupational hazard? The relationship between working conditions
and fertility. ICS-dissertation, Groningen.
205) Hidde Bekhuis (2013). The popularity of domestic cultural products: Cross-national
differences and the relation to globalization. ICS-dissertation, Utrecht.
206) Lieselotte Blommaert (2013). Are Joris and Renske more employable than Rashid and
Samira? A study on the prevalence and sources of ethnic discrimination in recruitment
in the Netherlands using experimental and survey data. ICS-dissertation, Utrecht.
207) Wiebke Schulz (2013). Careers of men and women in the 19th and 20th centuries.
ICS-dissertation, Utrecht.
208) Ozan Aksoy (2013). Essays on social preferences and beliefs in non-embedded social
dilemmas. ICS-dissertation, Utrecht.
209) Dominik Morbitzer (2013). Limited farsightedness in network formation. ICS-dissertation, Utrecht.
210) Thomas de Vroome (2013). Earning your place: The relation between immigrants’
economic and psychological integration in the Netherlands. ICS-dissertation, Utrecht.
211) Marloes de Lange (2013). Causes and consequences of employment flexibility among
young people. Recent developments in the Netherlands and Europe. ICS-dissertation,
Nijmegen.
212) Roza Meuleman (2014). Consuming the nation. Domestic cultural consumption: its
stratification and relation with nationalist attitudes. ICS-dissertation, Utrecht.
213) Esther Havekes (2014). Putting interethnic attitudes in context. The relationship between neighbourhood characteristics, interethnic attitudes and residential behaviour.
ICS-dissertation, Utrecht.
214) Zoltán Lippényi (2014). Transitions toward an open society? Intergenerational occupational mobility in Hungary in the 19th and 20th centuries. ICS-dissertation, Utrecht.
215) Anouk Smeekes (2014). The presence of the past: Historical rooting of national identity
and current group dynamics. ICS-dissertation, Utrecht.
241
ICS Dissertation Series
216) Michael Savelkoul (2014). Ethnic diversity and social capital. Testing underlying explanations derived from conflict and contact theories in Europe and the United States.
ICS-dissertation, Nijmegen.
217) Martijn Hogerbrugge (2014). Misfortune and family: How negative events, family ties,
and lives are linked. ICS-dissertation, Utrecht.
218) Gina Potarca (2014). Modern love. Comparative insights in online dating preferences
and assortative mating. ICS-dissertation, Groningen.
219) Mariska van der Horst (2014). Gender, aspirations, and achievements: Relating work
and family aspirations to occupational outcomes. ICS-dissertation, Utrecht.
220) Manu Muñoz-Herrera (2014). The impact of individual differences on network relations: Social exclusion and inequality in productive exchange and coordination games.
ICS-dissertation, Groningen.
221) Tim Immerzeel (2015). Voting for a change. The democratic lure of populist radical
right parties in voting behavior. ICS-dissertation, Utrecht.
242

Writing a Bachelor Thesis in Computer Science

LSC Minutes 20150519 (Special Meeting)

Facilitaion of Learning 2

Luxury Van Market

more continuity of PC

C HA RL ES VA N DA MM E FE RR Y

Eindexamen Engels havo 2005-II

Is health more important when you have less?

a few businesses with research exemptions state aircraft RPAS

full text