ArticlePDF Available

Risks, Costs and Labour Markets: Explaining Far Right-Wing Party Success in European Parliament Elections

Authors:

Abstract

What is the impact of the economy on cross national variation in far right-wing party support? This paper tests several hypotheses from existing literature on the results of the last three EP elections in all EU member states. We conceptualise the economy affects support because unemployment heightens the risks and costs that the population faces, but this is crucially mediated by labour market institutions. Findings from multiple regression analyses indicate that unemployment, real GDP growth, debt and deficits have no statistically significant effect on far right-wing party support at the national level. By contrast, labour markets influence costs and risks: where unemployment benefits and dismissal regulations are high, unemployment has no effect, but where either one of them is low, unemployment leads to higher far right-wing party support. This explains why unemployment has not led to far right-wing party support in some European countries that experienced the 2008 Eurozone crisis.
Risks, Costs and Labour Markets: Explaining cross-national patterns of
far right party success in European Parliament Elections
Daphne Halikiopoulou and Tim Vlandas1
University of Reading
** This paper has been accepted for publication in JCMS.
Please contact authors for citation details**
Abstract
Does the economy affect patterns of far right party support across countries? This paper
reconceptualises micro-level analyses that focus on the effect of unemployment through a
framework of costs, risks and the mediating role of labour market institutions. It then derives
several hypotheses and tests them on the results of the last three EP elections in all EU member
states. Findings from multiple regression analyses indicate that unemployment, real GDP growth,
debt and deficits have no statistically significant effect on far right party support at the national
level. By contrast, labour market institutions influence costs and risks: where unemployment
benefits and dismissal regulations are high, unemployment has no effect, but where either one of
them is low, unemployment leads to higher far right party support. This explains why
unemployment has not led to far right party support in some European countries that experienced
the severity of the 2008 Eurozone crisis.
Key words: European Parliament, far right parties, labour market institutions, economic crisis,
immigration, welfare state.
1 The authors’ names have been listed in alphabetical order. Both authors have contributed equally to this
work.
1
What is the effect of the economy on far right party support? On the one hand, economic
factors are often seen to be shifting the attitudes and preferences of voters towards such
parties (Lipset 1960; Betz 1994). These explanations differ in scope and in particular in
whether the preference for the far right is the product of objective characteristics or
subjective attitudes. But all these theories suggest that indicators such as unemployment,
low income, and more generally deprivation and expectation of deprivation create
conditions favourable to the rise of far right parties. This framework has been used to
explain both the rise of the fascist parties that emerged during the interwar period, and the
radical right variants that have re-emerged in Europe since the 1990’s. On the other hand,
other work has shown that economic stress is not necessarily causally linked to far right
party support (Mudde 2010; Art 2011). An increasing number of studies has found that
ethnic threats have a greater impact than economic threats (Ivarsflaten 2008; Lucassen
and Lubbers 2012) indicating that national identity is more of a driving force of far right
party support than economic interest (Sniderman and Hagendoorn 2007).
Most of these studies, however, either employ small N methods, focusing on one or few
cases thereby limiting the potential for generalizability or large N methods focusing on
individual level voting patterns. But what is the impact of the economy on cross-national
patterns of far right party support in the EU? While much attention has been given to the
effect of unemployment in previous literature, few studies have addressed the question of
variation at the country level. An exception is Arzheimer (2009). This study investigates
cross-national variation in Western Europe before Eastern enlargement, combining
system-level variables with individual sociodemographic and attitudinal data in the period
between 1980 and 2002 and including in the analysis cases where the far right performed
2
poorly.
In this paper, we test the effect of the economy on macro-level patterns of far right party
support in European Parliament (EP) elections. The latter may described as ‘second-order
national contests’ (Reif and Schmitt 1980; Van der Eijk and Franklin 1996), i.e. secondary
to national elections, but still national rather than European contests (Hix and Marsh 2007).
While we may expect the second order effect to inflate support for smaller parties, this
does not explain the significant variations we observe in terms of support for far right
parties across the EU. Given standardization in terms of time and electoral system, EP
elections offer a good platform from which to explore these macro level patterns of far right
party support from a comparative perspective: how may we explain cross-national
variations; and do existing explanations apply at the EP level?
EP elections show interesting cross-national variation in far right party support that is not
prima facie consistent with a simple demand- side model, which therefore calls for a
systematic comparative study. In the 2014 ‘earthquake’ EP elections, the peripheral
countries that experienced the most severe economic crises such as Portugal, Ireland,
Italy, Greece and Spain clearly displayed varying far right party success: while in Greece
the far right Golden Dawn (GD) experienced a significant rise, in Spain National
Democracy (ND) and Spain 2000, and in Portugal the National Renovator Party (PNR) did
not; in Ireland there is no far right party; and in Italy the Northern League (LN) declined.
There are also significant variations within and between countries that did not experience
comparable crisis conditions, including France, the UK, Sweden, Denmark and Finland.
Unemployment levels vary across Europe (see Figure 1), showing again no correlation:
countries with highest unemployment rates including Spain did not witness a rise in far
3
right party support, while countries with low unemployment levels did. Economic malaise is
not necessarily present in all cases where we observe the rise of the far right (although it is
correlated in some cases such as Greece). The correlation between unemployment and far
right party support for the 2014 election is low and not statistically significant.
This calls for a more thorough test of the impact of economic factors on far right party
support. Building on existing demand side literature that focuses on the economy (Swank
and Betz 2003; Kitschelt with McGann 1995; Kriesi et al 2006), this paper conceptualises
the effect of labour market insecurity that unemployment creates in terms of costs and
risks. Costs refer to the material penalty that unemployment entails while risks capture the
greater fear of unemployment that may arise in the employed population when national
unemployment rises. Both costs and risks are partly determined by the national
unemployment rate and partly by labour market institutions. Cost most obviously depends
on the economic condition – the national unemployment – but also on the extent to which
unemployment is ‘compensated’ by labour market institutions. Similarly, unemployment
risks determine the likelihood of bearing that cost and are therefore dependent on labour
market regulations. Thus, we expect unemployment to be associated with higher far right
support only in countries where unemployment is very costly because of low
unemployment benefits or in countries where unemployment increases the risk of
employed workers becoming unemployed because of low dismissal regulations.
We test these expectations using panel data regression analysis on a dataset comprising
three EP elections covering 28 European countries and capturing three time periods:
before, during and after the financial crisis. Our results confirm that unemployment by
itself is not statistically significant when controlling for all other relevant factors. However,
4
demand side economic factors do matter, but in a more complex way: where
unemployment benefits and dismissal regulations are high, unemployment has no effect,
but where either one of them is low, unemployment leads to higher far right support.
This paper contributes to existing literature in two ways. First, empirically, this paper
provides a macro level test of the expectations from the demand side economic literature
and explains the variation in EP election results. Second, we show that far right party
support is fuelled less strongly by national unemployment levels in case of specific welfare
state institutional arrangements.
This paper enfolds as follows. In the next section, we classify far right parties and review
previous literature on demand-side explanations of far right party support. We then derive
a number of hypotheses concerning the impact of unemployment on far right party support.
Using statistical analysis we then test these hypotheses and present our results. The last
section concludes with some implications for future research avenues.
5
Figure 1: Far right parties in 2014 EP elections and 2013 unemployment
6
Classifying and explaining the success of far right parties
Classification
The ‘extreme’ (Ignazi 1991), ‘radical’ (Rydgren 2007), ‘populist-radical’ (Mudde 2007;
Immerzeel et al 2015) or ‘far’ (Lucassen and Lubbers 2012; Vasilopoulou and
Halikiopoulou 2015) right party family has been characterised as one of the least
homogenous party families (Ennser 2012). Part of the problem in generating a broad
theory of patterns of far right party support is precisely this question of classification: which
parties should we compare? Do these parties belong to the same party family, and are
therefore comparable groups, or are the variations between them so fundamental that a
comparison is not possible? Sharp distinctions may be identified in terms of the social
groups these parties attract, the extent to which their voters have right authoritarian
attitudes, the relationship of the parties themselves with democracy, their association with
fascism, the harshness of their stance towards the European Union (EU) and the extent to
which they employ violence to materialise their goals.
Any attempt at classification faces methodological as well as theoretical problems.
Classifying these parties in terms of who votes for them can result in circular reasoning. If
these parties have a different voting base because they are different types of parties, then
the causal logic is skewed. Classifying these parties in terms of their ideology poses a
problem for theorising insofar as ideology may not create clearly defined categories.
Another problem refers to the government-opposition dynamics: as a party moves closer to
power it may moderate its position shifting between categories; or a party may radicalise in
order to compete with other parties in the system. Nonetheless and despite these
problems, a classification remains useful as it creates a starting point for any comparative
7
study that seeks to understand the varied success of these parties.
The key is that, while diverse, these parties also share a number of core common features
that permit them to be grouped together. First and foremost, this refers to a harsh stance
on immigration driven by nationalism (Eatwell 2000; Hainsworth 2008; Halikiopoulou et al.
2012). This has been conceptualised by Mudde (2010) as nativism, a concept that
describes the ethnic variant of nationalism, i.e. the definition of the nation in accordance to
ascriptive criteria, leading to calls for the maintenance of the homogeneity of the nation
(Halikiopoulou et al. 2012). These parties also share authoritarianism and populism
(Mudde 2007).
This paper favours the overarching term ‘far right’, and within it a distinction between
‘extreme’ and ‘radical’ variants which captures both the commonalities and variations
within this party family. The term ‘far’ allows inclusion of all parties that share a nationalist
anti-immigrant agenda, authoritarianism and populism. The distinction between ‘extreme’
and ‘radical’ captures the parties’ relationship with procedural and substantive democracy
(Mudde 2010) as well as a past association with fascist groups and the use of violence.
This paper therefore adopts the term ‘far right’ to describe parties that (a) belong to the
same party family; (b) are characterised by nationalism, authoritarianism and populism and
(c) differ in terms of their relationship with democracy, the extent to which they have a
fascist past and the extent to which violence forms part of their agenda. This means
including some of the borderline cases, but in a way that clearly captures their specificities.
In line with Immerzeel et al 2015 we include extreme right variants such as BNP, PNR, and
Jobbik as well as more radical variants such as DF, PVV and UKIP. We add the GD,
Independent Greeks (ANEL) (Vasilopoulou and Halikiopoulou 2015) and the Cypriot
8
National Popular Front (ELAM) (Katsourides 2013), as well as the Polish populist radical
Law and Justice Party (PiS) which has moved progressively to the nationalist right of the
political spectrum since mid 2000s (Pankowski 2010:152; Harrison and Bruter 2011;
Pankowski and Kormak 2013). Overall we examine 36 parties in 28 EU member states
(see Appendix 2).
State of the art: what explains demand for far right parties?
How may we conceptualise support for the far right? In the vast majority of accounts,
support for the far right is theorised on the basis of a demand and supply framework
examined at the micro level. The former refers to bottom up theories capturing any factors
that increase the ‘demand’ for far right parties, while the latter refers to top down
institutional factors. The demand framework is governed by an overarching rationale that
rests on the premise that societal conditions determine party success. Most demand-side
explanations emphasise the importance of economic factors. The key is some type of
exogenous trigger factor, for example economic crisis, globalization, internationalization,
that impacts negatively on either the socio-economic status of voters, or their expectations.
Economic interests may underline far right party support through several causal
mechanisms, which are either objective or subjective; about levels or change; and current
or future/expected. While some emphasise the level of the problem, others focus on the
perceptions of the problem or on the change in the problem. The ‘modernization losers’
thesis contends that those more likely to support far right parties are the losers of socio-
economic change (Kriesi 1999). It is consistent with the notion that globalization leads to
the formation of a new structural conflict in Western European countries (Kriesi et al 2006).
The socio-demographic model focuses on the objective current socio-economic conditions
9
of voters (Lipset 1960; Betz 1994; Rydgren 2007). The relative deprivation thesis merges
economics with subjective characteristics of voters focusing on the extent to which a
deteriorating economic condition will worsen an individual’s position, understood in
comparison either with one’s own past or with another social group (see, e.g. Lubbers and
Scheeppers 2002;). Realistic conflict theory posits that people are in competition over
scarce resources, understanding the rise of the far right in terms of a struggle for access to
the collective goods of the state (Wimmer 1997; Lucassen and Lubbers 2012).
These theories expect certain types of social groups who are either experiencing
deprivation in their present situation or who expect deprivation in the future to be the likely
supporters for the far right. Within this context, both high levels of unemployment and/or
sharp increases in the levels of unemployment are expected to be associated with far right
party support (Lubbers and Scheepers 2002). Those who are unemployed, lowly-
educated, involved in manual employment and, more generally, have low incomes are
expected to be the key electoral constituency of far right parties. Studies rarely distinguish
between the likelihood of opting for the far right vis-à-vis the far left among certain groups
such as the unemployed. But overall, theories that focus on the economy tend to associate
unemployment with far right party support because of protest and anti-systemic attitudes,
linkage with unfavourable out-group attitudes and authoritarian attitudes (Lubbers and
Scheepers 2002:134). Another reason is competition with immigrants and outsiders for
jobs, welfare, and more broadly access to the collective goods of the state (Wimmer 1997).
Recent studies, however, have found that cultural factors matter more for far right party
support than economic factors (Sniderman and Hangerboorn 2007; Ivarsflaten 2008;
Lucassen and Lubbers 2012); and that when economic factors do matter they do so in
10
different ways across cases (Lubbers and Scheepers 2002; Vasilopoulou and
Halikiopoulou 2015). These different conclusions drawn in various studies, both within and
across cases, have prompted scholars to note that demand-side economic factors in
themselves may be poor predictors of far right party support (Mudde 2010; Lucassen and
Lubbers 2012).
The EP election results provide a good platform for testing the impact of the economy on
far right party support at the national level across Europe. EP elections are ‘second-order
national contests’ (Hix and Marsh 2007) that favour smaller parties. It could therefore be
possible that the relationship between far right party support and the economy, as
identified in literature focusing on national elections and micro-level analyses, is different in
EP elections. Does the second-order effect, which entails that EP elections are more likely
platforms for the expression of discontent with the national government, affect our theories
about the relationship between unemployment and far right party support? In addition,
given that most studies tend to assume that the political reactions to economic and cultural
globalization are bound to manifest themselves above all at the national level (Kriesi et al
2006: 921-922), the examination of EP electoral results is a contribution to existing
literature. But it is also a methodologically sound choice because EP elections allow us to
sidestep endogeneity, timing and comparability issues. In contrast to national elections
where the economy and far right party support are endogenous, EP elections results are
unlikely to affect the economy. They offer a ‘snapshot’ of 28 different elections with
standardized results at a particular point in time using a similar electoral rule. EP elections
also provide a good reference point for understanding the rise of both radical right variants,
which have been the most popular types of far right parties since the 1990s; and also the
11
most extreme variants such as Jobbik and the Golden Dawn. These extreme variants have
experienced a dramatic increase since 2010 contrary to most theories that have postulated
that ‘old’ far right parties (Golder 2003) with clear links to fascism are unlikely to
experience much support in post war Europe. Finally, EP elections provide a platform on
which to test a model of far right party support across Western and Eastern Europe, which
are rarely examined together.
Risks, Costs and labour markets: Theorising the impact of the economy on far right
party support
This section develops our expectations concerning the effect of unemployment in two
steps. In the first step, we build on the demand side literature reviewed in the previous
section to conceptualise the effect of unemployment on far right party support through a
framework that emphasises societal risks and costs. In a second step, we build on welfare
state literature to stress the role of labour market institutions in moderating these risks and
costs.
Economic factors affect demand through risks and costs
As discussed in the previous section, unemployment is shown to be related to far right
party support in most of the demand side literature. Theorising how economic factors may
lead to support requires reconceptualising how unemployment may affect insecurity
(Kitschelt with McGann 1995; Swank and Betz 2003; Anderson and Pontusson 2007).
Economic problems may affect demand for the far right through two empirically related
12
but conceptually distinct channels. We term these two channels ‘risk’ and ‘cost’:
economic factors may mean higher far right support because of higher cost or higher risk,
or both. In the case of labour market problems, the first cost - channel is that
unemployment may increase support because more people are unemployed and the
unemployed are more likely to vote for extreme right parties (Lubbers and Scheepers
2002; Rydgren 2007) for the various reasons explained above, including protest voting,
authoritarian attitudes, negative attitudes towards out-groups and competition for state
goods with outsiders (Wimmer 1997).
That unemployment imposes a cost on someone is not contested, whether in terms of well-
being, life satisfaction, or other metrics (e.g. Jahoda 1988; Gerlach and Stephan 1996).
There is also widespread evidence in the labour market literature that unemployed and
atypical workers have different policy preferences from those in employment because they
face much higher risks (Rueda 2005; Rueda 2007; Emmenegger 2012; Vlandas 2013;
Marx and Picot 2013; Marx 2014). Labour markets have become increasingly dualized with
insiders in permanent contracts and those in non-standard jobs and unemployment have
seen their risks, entitlements, and policy preferences diverge (Emmenegger et al 2012).
But labour market problems may also have an effect through a second – risk - channel, via
the effects of higher unemployment on the perceived risk of unemployment of those
currently in jobs. Indeed, as unemployment increases, this also affects the well-being of
those currently in jobs (Hartley et al. 1991; De Witte 1999; Clark 2003; Böckerman 2004).
This is because when general market conditions deteriorate, this also affects the
perception of insecurity of employed workers (Chung and Carr 2014; Chung and Van
Oorschot 2011; Mau, Mewes and Schöneck, 2012). In addition, workers that were
13
previously unemployed continue to feel more insecure even after returning to work
(Böckerman 2004; Erlinghagen 2008).
Thus, unemployment may affect electoral choice through its effect on economic insecurity
(cf. Mughan and Lacy 2002). Higher unemployment should ceteris paribus increase
national support for far right parties among both unemployed and employed. As a result,
we expect unemployment to be significantly associated with higher far right support:
Hypothesis 1: Unemployment is positively associated with far right party support
Costs can be compensated
Losing one’s employment has a direct income effect through the loss of wages. However,
in most European countries unemployed workers are entitled to some replacement of their
previous income in the form of unemployment benefits (Van Vliet and Caminada 2012).
The cost of being unemployed therefore depends on the generosity of unemployment
benefits: in countries with generous benefits, the cost of unemployment relative to
employment is lower. As a result, the extent to which we should expect higher
unemployment to lead to more far right support is crucially contingent on labour market
institutions. Consistent with this, various studies find that workers report lower levels of job
insecurity in European countries where unemployment benefits are more generous (OECD
2004; Clark and Postel Vinay 2005; Mau, Mewes and Schöneck 2012).
Labour market policies can also compensate costs by minimising its duration. Using the
European Social Survey, Chung and Van Oorschot (2011) confirm that generous
unemployment benefits provide a sense of security by partly replacing lost income. Labour
14
market policies can also mediate the effect of unemployment on other potentially relevant
facets of welfare. Chung and Carr (2014) analyse employment insecurity and life
satisfaction in 22 countries and highlight that labour market policies may mitigate the
adverse effects of unemployment on life satisfaction. In sum, there is significant evidence
that unemployment affects individuals’ welfare at the micro level and that institutions
mediate this impact.
Moreover, globalisation ‘increases social dislocations and economic insecurity, as the
distribution of incomes and jobs across firms and industries becomes increasingly
unstable’ (Garrett 1998: 7). However, our theoretical framework would suggest that
globalisation does not by itself lead to far right support but is instead mediated by welfare
state institutions. This is precisely what Swank and Betz (2003) find: countries with
generous systems of social protection are less likely to exhibit significant ‘radical right
populist parties’- i.e. the new far right parties that have experienced increasing support
since the 1990’s on platforms that emphasise an ideological and organizational
disassociation from fascism. This is because as we discuss above - generous welfare
state policies may compensate and mitigate economic insecurities brought about by
globalization.
Do these dynamics apply at the national level? While our conceptual framework is
consistent with Swank and Betz’s (2003) findings, we test this logic on a different and
broader phenomenon to identify cross-national patterns in EP elections. In addition, while
Swank and Betz (2003) focus on parties that ‘typically embrace neoliberal economic
programmes, xenophobia and strident anti-establishment positions’ (2003: 218), we want
to test this framework on all far right parties. Our sample includes extreme right variants
15
such as GD that have experienced a significant rise since the onset of economic crisis and
have centred their rhetoric on welfare provision reminiscent of the Nazi winterhilfswerk
(Vasilopoulou and Halikiopoulou 2015). While they focus on globalisation and the welfare
state, we focus on unemployment and labour market institutions. We expect
unemployment benefits to lower the cost of unemployment to the unemployed and
therefore to reduce the cost of unemployment. If this is true, unemployment benefits should
have both an independent effect and a mediating effect on national level far right party
support:
Hypothesis 2: More generous unemployment benefits lower far right support
Hypothesis 3: More generous unemployment benefits mitigate the impact of
unemployment on far right support
Risks can be mitigated
Costs are only one of the two key mechanisms through which unemployment could affect
far right party support. The unemployed are only one casualty of rising unemployment.
Those in employment may also be adversely affected by rising unemployment because
they consider the risk that they will incur those costs and a higher risk negatively affects
them. While this risk is influenced by unemployment, it is also determined by Employment
Protection Legislation (EPL): where EPL is high, dismissal costs are also high, and an
employer is ceteris paribus less likely to dismiss a particular employee in response to a
downturn because the cost benefit calculation shifts in favour of keeping the employee. As
a result, objective and perceived risks in society are crucially determined by the level of
16
EPL. For instance, workers in permanent contracts that are protected by EPL and those in
temporary contracts that are not, exhibit different degrees of insecurities and different
policy preferences (Clark and Postel Vinay 2005; Rueda 2007; Vlandas 2013). Where it is
easy to dismiss them, permanent workers will respond much more fearfully to a given rise
in unemployment (for more on the effect of EPL on permanent workers, see Rueda 2005
and 2007; Emmenegger et al 2012; Vlandas 2013). The crisis is a good example of this
dynamic. Governments’ social policy efforts in response to higher unemployment during
the crisis were lower in countries where workers in permanent contracts were highly
protected from redundancy, because these workers were less concerned about losing their
job and therefore did not push their governments to expand social policy initiatives (Rueda
2014).
In sum, welfare is likely to have a protective effect (cf Arzheimer 2009; Swank and Betz,
2003). Where EPL is high, permanent workers have less to fear from unemployment since
they are unlikely to be made redundant. An increase in unemployment should only lead to
higher far right party support in countries with low EPL. By contrast, our expectations
concerning the relationship between the level of EPL itself and far right party support is
more indeterminate. As the labour economics literature suggests, EPL not only reduces
flows into unemployment but also reduce the probability of finding a job for the unemployed
(OECD 1994; Bassanini and Duval 2006, 2009). This would mean that the two effects may
cancel each other out, and indeed there is mixed evidence concerning the impact of EPL
on unemployment (Baccaro and Rei 2007). Consequently, a higher EPL may make the
employed more secure while making the unemployed more insecure. The average effect of
EPL is therefore uncertain: if the effect on workers dominates, unemployment should have
17
a negative effect on far right support; if the effect on the unemployed dominates, the
reverse should be true; and if both effects are equivalent, the association should be
insignificant. We derive two hypotheses from this discussion:
Hypothesis 4: more stringent job security regulation is not significantly associated
with far right party support
Hypothesis 5: more stringent job security regulation lowers the impact of
unemployment on far right party support
Data, estimation strategy and results
Data
We collect data on several variables that allow us to test these hypotheses while
controlling for other demand and supply factors that have been identified in previous
literature discussed earlier. Note that all independent variables are where data is
available in the year before the election takes place (e.g. 2013 unemployment rate for
May 2014 election). In some cases, data is not available for 2013, and we therefore have
to rely on 2012 data or in very few cases on 2011 data, but this is never the case for fast
moving economic variables. Descriptive statistics, definitions and sources of all variables
are discussed in Table A1 in the appendix.
To test the economic impact of the labour market, we rely on the overall unemployment
rate (as % of total) both because it captures the biggest problem in the labour market but
also because it receives the most attention from the public. Since people vote on the basis
18
of their situation prior to the election, we use data on the unemployment rate for the year
prior to the EP election (e.g. 2013 for the 2014 election). We also test the effect of
alternative measures of labour market problems (e.g. youth unemployment rate and long
term unemployment rate).
Next, we want to measure the risk and cost of unemployment. To capture risk, in addition
to the national unemployment rate, we rely on the index of EPL for regular workers
developed by the OECD. The higher the EPL index, the less likely workers in permanent
contracts will fear losing their job in the face of increasing unemployment. To capture cost,
we rely on the unemployment benefit replacement rate, which captures the size of the
income loss upon becoming unemployed. We expect EPL and unemployment benefits to
mediate the effect of unemployment on the far right, and unemployment benefits to lower
far right party support but have no expectations concerning the independent effect of EPL
(i.e. when it is not interacted).
We include a number of controls for demand side factors. The first set of controls we
include is real GDP growth. We also test for the effect of budgetary stress (debt and deficit
as a % of GDP). A second control concerns the impact of deindustrialisation, which has
received some attention in the far right literature (e.g. Kitschelt 2007). The third set of
controls concerns the role of internationalisation, which previous literature has focused on.
Kriesi (2014) argues that the populist right is more likely to mobilise in defence of the
national identity and the nation-state that internationalisation may be undermining.
Lucassen and Lubbers (2012) argue that perceived cultural ethnic threats appear a
stronger predictor of far-right preferences than perceived economic ethnic threats. Thus,
any international forces that threaten the state (e.g. globalisation) or national identity (e.g.
19
immigration) could be expected to lead to greater far right support. We include three
measures to test these expectations. First, we include total trade (sum of exports and
imports) as a share of GDP. Second, we investigate whether immigration affects far right
support. Third, as a robustness check we construct a variable that captures anti-immigrant
feeling. The European Social Survey asks respondents to position themselves on the
following statement: “Country's cultural life undermined or enriched by immigrants”. They
choose a number between 0 (Cultural life undermined) and 10 (Cultural life enriched). Our
variable sums the share of respondents that have responded 0 to 4 (since 5 is a neutral
position, it is not included in our “immigration is bad for culture” variable).
To account for supply- side dynamics, we consider the role of several factors. First, to
capture the extent to which national electoral systems give space to far right votes in
national elections, we rely on whether there is a proportional electoral system in the
country. We expect proportional systems to exhibit lower far right support in EP elections
because dissatisfied voters can voice their discontents in national elections. Second, we
consider the extent to which the party system is open. Our baseline indicator is an index of
electoral fractionalisation2 but we also test the effect of the effective number of political
parties3 in each country.
Finally, we test two contextual factors. The first contextual factor is whether countries have
a post-communist past because there are variations between Eastern and Western Europe
(Mudde 2007; Halikiopoulou et al 2012). The second contextual factor is whether the last
2 Index of legislative fractionalization of the party system according to the formula [F] proposed by Rae
(1968).
3 Effective number of parties on the votes level according to the formula [N2] pro-posed by Laakso and
Taagepera (1979). The effective number of parties uses the same information as the Rae-Index and is
calculated from this index as follows: effpar_ele = 1 / (1 - rae_ele)
20
election was significantly different from the other two, even when controlling for all relevant
economic factors (for instance because of a qualitatively different effect of the crisis). Both
are captured by including a dummy variable coded 1 where the condition is present, and 0
otherwise.
Because we are interested in explaining variation across countries, the inclusion of country
fixed effects would ‘explain away’ what we are trying to explain. Indeed, as Plümper,
Troeger and Manow (2005: 331) argue, “unit dummies completely absorb differences in
the level of independent variables across units”. Thus, the ‘level effect’ of our key
independent variables (e.g. unemployment benefits) is suppressed when including fixed
country effects (ibid: 333). While the effect of a change in unemployment and
unemployment benefits is also theoretically relevant, our main concern here is about the
effect of the level of these variables on far right support. That being said, we do run one
model with fixed effects and our results are unchanged. We are similarly reluctant to
neutralise time effects that affect all countries homogenously and thus favour not including
time effects (though we do run one model with time effects and the results are similar).
Test and results
We run a series of panel data regression analysis on 28 EU countries in the last three EP
elections. Given missing data, this results in 74 country-election year data points for most
regressions and we always mention the number of observations for each regression result.
We focus on the last three elections because European far right parties have changed over
longer periods of time and because our primary interest is to compare, pre-crisis, early
crisis and end of crisis times (i.e. 2004, 2009 and 2014). Our regressions include robust
errors clustered by country to neutralise the effect of autocorrelation and
21
heteroskedasticity.
Table 1 shows the results for our baseline model. Note that all independent variables are
standardised around mean 0 with standard deviation of 1 which means the coefficients are
directly comparable: they show the effect on far right support of a 1 standard deviation
change in the independent variable. The unemployment rate has no statistically significant
effect in all columns so we find no evidence for hypothesis 1. Real GDP growth and total
trade are similarly statistically insignificant. Other regressions (not shown, see table A2 in
appendix) suggest that long term unemployment, youth unemployment, debt and deficit,
exports and imports do not exhibit a statistically significant association with far right party
support.
The coefficients for immigration and industrialisation are both statistically significant but
only at the 10% level. We test alternative measures and find that both immigration from
non-EU and from EU countries is indeed positively associated with higher far right support
(Table A2 in the appendix, columns 2 and 3). However, these results are not stable when
including additional controls during our robustness checks (see columns 12 to 14 in table
A2 in the appendix): Spending on active labour market policies, OECD unemployment
benefit replacement rate, Inequality ratio between top and bottom quintile, Gini coefficient.
Unemployment benefits has a strongly negative and statistically significant association with
far right support (consistent with hypothesis 2): 1 standard deviation change in the
replacement rate of unemployment benefit (about 14 percentage points change) is
associated with between 6.8 and 3.6 points increase in voting share of the far right,
depending on the specifications. For reference, 2 standard deviations amounts to the
difference between the UK and Germany in 2011. Because our unemployment benefits
22
data stops in 2011, we also carry out the same test with an OECD index of unemployment
benefit replacement rate which stops in 2012, but has less country coverage (table A2 in
appendix). The results are the same. Passive or active labour market spending itself does
not matter, suggesting it is entitlements not spending that matters. EPL has no statistically
significant effect (consistent with hypothesis 4).
We find that electoral fractionalisation and a proportional electoral system are negatively
associated with far right party support but the number of parties does not seem to matter
(column 6). While the effect of the PR system is mostly robust to alternative specifications
this is not the case for electoral fractionalisation (see table A2 in appendix). The dummy
variable for post-communist countries is only statistically significant at the 10% level.
We also test the effect of additional factors in table A2 in the appendix. Inequality between
the top and bottom 20% of the income distribution and the Gini coefficient have no
statistically significant association with far right party support. The immigrant population
and the crisis (dummy variable equal to 1 for 2014 election and 0 otherwise) have no
effect. The share of the population that believe immigration is bad for the country’s culture
has no statistically significant effect (see column 15 in table A2). Finally, using the standard
Eurobarometer number 80 (autumn 2013), we can calculate the bivariate correlation
between far right party support and the percentage of respondents in each country that
‘feel like a citizen of the EU’ to check whether euroscepticism is a likely omitted variable of
our analysis. The correlation coefficient is low (-0.14) and clearly statistically insignificant
(0.46).
23
Table 1: Baseline results for the determinants of far right support
Column (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Unemployment rate 0.37 0.19 0.67 0.52 0.27 0.38 0.29 -0.15 -0.42 -0.63
(0.628) (0.831) (1.097) (0.586) (0.662) (0.662) (0.671) (0.751) (0.689) (0.857)
Real GDP -0.45 0.46 0.25 -0.15 -0.63 -0.68 -0.61 -0.39 0.30 -1.22
(0.554) (0.485) (0.735) (0.668) (0.769) (0.800) (0.770) (0.747) (0.645) (1.162)
Unemployment benefits -4.19** -3.62* -4.13* -5.41*** -4.32** -3.88* -4.33** -4.66** -5.25** -6.80***
(1.564) (1.539) (2.034) (1.635) (1.595) (1.618) (1.596) (1.608) (1.602) (1.982)
Electoral fractionalisation 2.46* 2.32* 4.73* 2.99** 2.62** 2.65** 2.59** 2.74** 2.27t
(0.999) (0.932) (2.012) (1.090) (0.909) (0.923) (0.864) (0.896) (1.307)
Immigration 4.83t1.84 2.08t
(2.819) (1.128) (1.160)
Post-communist dummy -6.15t
(3.280)
Proportional
Representation
-11.63*** -11.89*** -11.75*** -9.36*** -7.01* -7.42*
(2.025) (2.147) (2.155) (2.515) (3.392) (3.769)
Effective number of parties 1.60
(0.988)
Total trade (% GDP) 0.11 -0.62 -0.50 0.67
(0.735) (0.876) (0.977) (1.558)
Share of industry -1.83t-2.41* -2.63*
(1.047) (1.054) (1.309)
EPL 0.80
(1.288)
Constant 7.47*** 5.23*** 18.79*** 9.50*** 18.61*** 18.79*** 18.72*** 16.46*** 14.45*** 14.98***
Year effects No Yes No No No No No No No No
Country effects No No Yes No No No No No No No
Observations 76 76 72 76 73 73 73 73 69 56
Number of countries 27 27 27 27 26 26 26 26 26 21
R-squared within model 0.17 0.29 0.25 0.17 0.17 0.12 0.17 0.16 0.23 0.31
R-squared overall model 0.19 0.19 0.85 0.25 0.35 0.32 0.35 0.40 0.41 0.44
R-squared between model 0.21 0.19 1.00 0.29 0.42 0.39 0.42 0.48 0.49 0.45
Note: Robust standard errors clustered by country in parentheses, *** p<0.001, ** p<0.01, * p<0.05; t p<0.1. All independent variables are
standardised around mean 0 with standard deviation of 1.
24
Hypotheses 3 and 5 are about interactions between labour market institutions and
unemployment. This requires running a regression while including an interaction term
between unemployment and unemployment benefits, as well as between unemployment
and EPL. The interaction effects’ magnitude and significance cannot be evaluated from the
table results (see Brambor et al. 2006). We therefore run two separate models and plot the
average marginal effects of unemployment conditional on unemployment benefits and
EPL, respectively. The top left panel of Figure 2 shows that unemployment does have a
positive association with far right party support in countries where unemployment benefits
replace strictly less than 50% of previous income while working; beyond this level
unemployment is no longer significantly different from 0. The top right panel of Figure 2
shows that unemployment does have a positive association with far right support in
countries where the OECD EPL index is lower than slightly above 2; beyond this level
unemployment is no longer significantly different from 0.4 Since we use the 10%
significance level as our maximum threshold for statistical significance, both panels plot
90% confidence interval. In the bottom two panels, we show the results for more
demanding (95%) confidence intervals: the results are substantively the same but the
impact of unemployment is only positive for slightly lower values of EPL than before.
4 The full regression results are shown in Table A3 in the appendix and suggest that immigration and
electoral fractionalisation continues to be statistically significant and positive.
25
Figure 2: Effect of unemployment rate on far right support conditional on unemployment benefit generosity and EPL
26
In other words and consistent with hypotheses 3 and 5, unemployment does not
increase costs and risks sufficiently in countries with high EPL and high unemployment
benefits. This finding captures what has happened in Spain, Portugal and Italy where
unemployment has increased, but the replacement is higher than 60% and EPL higher
than 2, so that the far right has either remained stable and low (Spain and Portugal), or
has fallen (Italy). By contrast, Greece, the UK5, Hungary and Poland all have
unemployment benefit replacement rates that are lower than 0.5 and have seen their far
right party support increase. France is an interesting outlier, having high unemployment
benefits but also a high score of far right. This could be because of other case-specific,
micro-level or supply-side dynamics, having for example to do with the discourse of
Marine Le Pen, the activity of the party at the local level and party system dynamics.
Our macro level findings are consistent with Arzheimer’s (2009: 274) findings using
Eurobarometer survey data in Western Europe and analysing both individual and
contextual level effects. It is particularly important to compare our findings to this study
not only because it is one of few cross-national study on the topic but more importantly
because some of his analysis investigates macro-micro interactions. Using a different
proxy, Arzheimer (2009: 269) also finds that the amount of immigration in a country has
a positive effect on the probability that an individual votes far right.6 Also similarly to our
results, his results suggest a statistically significant effect of unemployment benefits in
reducing far right party support, though he suggests this effect is overall not particularly
strong (ibid: 271). While he does not investigate the effect of employment protection
5 The UK has both a very low unemployment benefit replacement rate and a fairly high far right party
support which may be driving the results. We therefore rerun the results without the UK as a robustness
check: the results are the same (see table A4 and figures A1 and A2 in the appendix.
6 Specifically he finds that “higher levels of immigration are associated with higher support for extreme
right” (ibid: 272).
27
legislation, some of his findings concerning the interaction between unemployment
benefit systems and unemployment rate in shaping the probability that an individual
votes for the far right are relevant to our analysis and worth quoting at length. Like us he
finds that “a positive effect [of unemployment rates] becomes visible but only in contexts
when either levels of immigration or benefits are very low” and that “at high levels of
immigration, unemployment benefits reduce the impact of unemployment” (ibid: 272).
This corroborates using micro level data that labour market institutions mediate the
impact of unemployment on far right party support.
Conclusion
How may we understand the cross-national variation in the rise of far right parties in EP
elections and what is the role of economic factors? While many theories focus on the
effect of the economy, there are conflicting expectations concerning when this happens
and contradictory findings. Previous literature has increasingly contended that the
extent to which economic factors matter varies across cases; and that ethnic factors
may play a greater role. But these have been primarily examined either in small N case
studies, or at the micro level in national election contests. By contrast, there are few
systematic studies that test the impact of the economy on the macro-level, attempting to
address questions of cross-national variation.
To address this issue this paper has reconceptualised micro-level analyses that focus
on the effect of unemployment through a framework of costs, risks and the mediating
role of labour market institutions. Building on previous work (Arzheimer 2009; Swank
28
and Betz 2003), it developed several new hypotheses and tested them at the macro-
level on the results of the last three EP elections in all EU member states. We found
that unemployment, real GDP growth, debt and deficits have no statistically significant
effect on far right party support at the national level. By contrast, our results confirm that
labour market policies and institutions influence costs and risks: where unemployment
benefits and dismissal regulations are high, unemployment has no effect, but where
either one of them is low, unemployment leads to higher far right party support.
Unemployment benefits – but not dismissal regulations – have a statistically significant
negative relationship with far right support.
Our contribution is twofold. First, we test the impact of the economy on far right party
support at the macro-level through a systematic comparison of far right parties across
the EU 28 including both Western and Eastern European countries. Second, we show
that far right party support is fuelled less strongly by national unemployment levels in
countries with regulated labour markets and generous unemployment benefits.
Our finding that labour market institutions mediate the impact of the economy on far
right party support in EP elections is not only interesting in itself, but also opens
avenues for future research towards a more general theory of the economy’s impact on
far right party support. This paper has tested its hypotheses at the macro-level as an
initial step. A most important second step would be an examination at the individual
level, thus offering support for our hypotheses at both the macro and micro-levels.
Which individuals are more likely to opt for far right parties and to what extent do labour
market institutions influence their choices by impacting on the risks and costs of
employment? Does EPL affect these choices? What happens when people become
29
unemployed? Do fixed or flexible contracts impact far right party support? Applying this
framework to national elections would yield valuable results with broad generalizability
potential. We also posit a micro level mechanism linking insecurity, labour market
institutions and voting preferences for far right parties that could be tested using survey
data. Finally, this framework could also be tested on other party families, most notably
the radical left given its emphasis on welfare and the labour market. While far left and
far right parties converge of issues of ‘welfare chauvinism’ (De Koster et al 2012) and
could be drawing voters from the same pool, they tend to be treated as separate in the
literature.
30
References
Arzheimer, K. (2009), Contextual Factors and the Extreme Right Vote in Western
Europe, 1980–2002. American Journal of Political Science, 53, 259–275.
doi: 10.1111/j.1540-5907.2009.00369.x
Baccaro, L. and Rei, D. (2007). ‘Institutional Determinants of Unemployment in OECD
Countries: Does the Deregulatory View Hold Water?’, International Organization, 6:
527-569.
Bassanini, A. and Duval, R. (2006) 'Employment patterns in OECD countries:
Reassessing the role of Policies and Institutions', in labour and social affairs Directorate
for employment (ed.), Social, employment and migration working paper (Paris: OECD).
Bassanini, A. and Duval, R. (2009) 'Unemployment, institutions, and reform
complementarities: re-assessing the aggregate evidence for OECD countries', Oxford
Review of Economic Policy, 25/1: 40-59.
Betz H.-G. (1994) Radical Right-Wing Populism in Western Europe. New York: St.
Martin's Press
Böckerman, P. (2004). 'Perception of job instability in Europe'. Social Indicators
Research, 67(3): 283-314.
Brambor, T., Clark, W. R. and Golder, M. (2006) ‘Understanding Interaction Models:
Improving Empirical Analyses’, Political Analysis 14: 63-82.
Chung, H. and Carr (2014) Employment insecurity and life satisfaction: The moderating
31
influence of labour market policies across Europe Journal of European Social Policy
October 1, 2014 24: 383-399
Chung, H. and Van Oorschot, W. (2011) Institutions versus market forces: Explaining
the employment insecurity of European individuals during (the beginning of) the
financial crisis Journal of European Social Policy October 1, 2011 21: 287-301
Clark, A. and Postel-Vinay, F. (2005) Job Security and Job Protection. CEP Discussion
Paper No 678 February 2005 http://cep.lse.ac.uk/pubs/download/dp0678.pdf
De Koster, W. P. Achterberg and J. van der Waal (2012) The new right and the welfare
state: The electoral relevance of welfare chauvinism and welfare populism in the
Netherlands, International Political Science Review, 34(1): 3–20.
De Witte, H. (1999): “Job Insecurity and Psychological Well-being: Review of the
Literature and Exploration of Some Unresolved Issues”, European Journal of Work and
Organizational Psychology 8, 155-177.
Eatwell R. (2000) ‘The rebirth of the extreme right in Western Europe?’ Parliamentary
Affairs, 53(3): 407–25.
Emmenegger P. (2012) Job Security Regulations in Western Democracies: a Fuzzy-Set
Analysis. European Journal of Political Research 2012;50:336-364.
Emmenegger P., Häusermann S., Palier B., Seeleib-Kaiser M. (2012) The Age of
Dualization. Oxford: Oxford University Press.
32
Ennser, L. (2012) The homogeneity of West European party families: The radical right in
comparative perspective, Party Politics 18: 151.
Esping-Andersen, G. (1990) The Three Worlds of Welfare Capitalism. Princeton
University Press.
Falzon, M.A and Micallef, .M (2012) Rights, Roots and Routes: Local and transnational
Contexts of Extreme Right Movements in Contemporary Malta in Mammone, A., Godin,
E. and Jenkins, B. (Eds) (2012) Mapping the Extreme Right in Contemporary Europe:
From Local to Transnational, Oxon: Routledge
Garrett, G. (1998). Partisan politics in the global economy. New York: Cambridge
University Press.
Gerlach, K., and G. Stephan (1996): A paper on unhappiness and unemployment in
Germany, Economics Letters 52, 325-330.
Golder, Matt. (2003) Explaining variation in the Success of extreme right parties in
Western Europe. Comparative Political Studies, 36 (4): 432-466.
Hainsworth, P. (2008). The extreme right in Western Europe. Abingdon: Routledge.
Halikiopoulou, D., Nanou, K. and 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.
Harrison, S. and Bruter, M. (2011), Mapping Extreme Right Ideology: An Empirical
Geography of the European Extreme Right’, Basingstoke: Palgrave Macmillan
33
Hartley, J., D. Jacobson, B. Klandermans, T. van Vuuren, with L. Greenhalgh and R.
Sutton (1991): Job Insecurity: Coping with jobs at risk, London: Sage.
Hix, S. and Marsh, M. (2007) Punishment or Protest? Understanding European
Parliament Elections. The Journal of Politics, Vol. 69:2, 495-510
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the
European radical right and other parties on typical radical right issues, Party Politics,
DOI: 10.1177/1354068814567975
Ivarsflaten, E. (2008). What unites right-wing populists in Western Europe? Re-
examining grievance mobilization models in seven successful cases. Comparative
Political Studies, 41: 3-23.
Jahoda, M. (1988): “Economic recession and mental health: Some conceptual issues”,
Journal of Social Issues 44, 13-23.
Kaufmann, E. (2014), ‘It's the Demography, Stupid’: Ethnic Change and Opposition to
Immigration. The Political Quarterly, 85: 267–276. doi: 10.1111/1467-923X.12090
Katsourides, Y. (2013) ‘Determinants of Extreme Right Reappearance in Cyprus: The
National Popular Front (ELAM), Golden Dawn's Sister Party’, South European Society
and Politics, 18:4
Kitschelt, H. with McGann, A. (1995) The Radical Right in Western Europe: A
Comparative Analysis. Ann Arbor: University of Michigan Press.
34
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.
Kriesi, H. (2014) The Populist Challenge, West European Politics, 37(2): 361-378
Kriesi, H., Grande, E., Lachat, R., Dolezal, M., Bornschier, S., Frey, T. (2006)
Globalisation and transformation of the national political space: Six European countries
compared, European Journal of Political Research, 45(6): 921–956.
Lubbers, M., Gijsberts, M. and Scheepers, P. (2002), Extreme right-wing voting in
Western Europe’ European Journal of Political Research 41: 345–378
Lubbers, M. and Scheepers, P. (2002) French Front National Voting: a micro and macro
perspective, Ethnic and Racial Studies, 25(1): 120-149.
Lucassen, G. and Lubbers, M. (2012). ‘Who Fears What? Explaining Far-Right-Wing
Preference in Europe by Distinguishing Perceived Cultural and Economic Ethnic
Threats’, Comparative Political Studies, 45(5): 547–574.
Mares, M. (2012) Right-Wing Extremism in the Czech Republic, International Policy̌
Analysis, Friedrich Ebert Stiftung
Marx, P. and Picot, G. (2013) The Party Preferences of Atypical Workers in Germany,
Journal of European Social Policy, 23(2): 164-178.
Marx, P. (2014) 'Labour Market Risks and Political Preferences: The Case of Temporary
35
Employment', European Journal of Political Research, 53(1): 136–159.
Mau, Steffen, Mewes, Jan and Schöneck, Nadine (2012) What determines subjective
socio-economic insecurity? Context and class in comparative perspective Socio-
Economic Review October 1, 2012 10: 655-682
Mudde, C. (2007) Populist radical right parties in Europe. Cambridge; New York:
Cambridge University Press.
Mudde, C. (2010) ‘The Populist Radical Right: A Pathological Normalcy’, West
European Politics, 33(6): 1167-1186.
Minkenberg, M. and Perrineau, P (2004) The Radical Right in the European Elections
2004, International Political Science Review, 28(1): 29-55
Mughan, A. and Lacy, D. (2002). Economic performance, job insecurity and electoral
choice, British Journal of Political Science 32(4): 513533.
OECD, 'Jobs Study', (Paris: OECD, 1994).
OECD. (2004). 'Employment Protection Regulation and Labour Market Performance'
OECD Employment Outlook 2004 (pp. 61-125). Paris: OECD.
Pankowski, R., (2010) The Populist Radical Right in Poland: The Patriots, Oxon:
Routledge
Pankowski, R and Kormak, M., (2013) ‘Radical Nationalism in Poland: From Theory to
Practice’ In R. Melzer and Serafin, S. (Eds.), Right-wing extremism in Europe: Counter-
36
strategies and Labor-Market Oriented Exit Strategies. Friedrich Ebert Foundation.
Plümper, T., Troeger, V., Manow, P. (2005) Panel data analysis in comparative politics:
Linking method to theory, European Journal of Political Research 44: 327-354.
Pytlas, B. (2013) ‘Radical-right narratives in Slovakia and Hungary: historical legacies,
mythic overlaying and contemporary politics’, Patterns of Prejudice, 47(2): 162-183
Rueda D. (2005) Insider-Outsider Politics in Industrialized Democracies: The Challenge
to Social Democratic Parties. American Political Science Review, 99:61-74.
Rueda D. (2006) Social Democracy and Active Labour Market Policies: Insiders,
Outsiders, and the Politics of Employment Promotion. British Journal of Political
Science, 36:385-406.
Rueda D. (2007) Social Democracy Inside Out. Oxford: Oxford University Press.
Rueda, D. (2014) Dualization, crisis and the welfare state. Socioeconomic Review,
12(2): 381-407.
Rydgren, J. (2007) The Sociology of the Radical Right, The Annual Review of
Sociology, 33:241–262.
Rydgren, J. (ed). (2013) Class Politics and the Radical Right, Oxon: Routledge
Reif, K. and Schmitt, H. (1980). Nine Second-Order National Elections: A Conceptual
Framework for the Analysis of European Election Results. European Journal of Political
Research 8 (1): 3–45.
37
Scheve Kenneth and Slaughte, Matthew J. (2004) Economic Insecurity and the
Globalization of Production. American Journal of Political Science Volume 48, Issue 4,
pages 662–674, October 2004.
Sniderman, P., and Hagendoorn, L. (2007). When ways of life collide. Princeton, NJ:
Princeton University Press.
Swank, D and Betz, HG. (2003) Globalization, the Welfare State and Right-Wing
Populism in Western Europe, Socio-Economic Review, 1: 215-245.
Van Vliet, O. and Caminada, K. (2012), ‘Unemployment replacement rates dataset
among 34 welfare states 1971-2009: An update, extension and modification of Scruggs’
Welfare State Entitlements Data Set’, NEUJOBS Special Report No. 2, Leiden
University
Van der Eijk, C and Franklin, M (eds). (1996) Choosing Europe? The European
Electorate and National Politics in the Face of Union. Ann Arbor: University of Michigan
Press.
Vasilopoulou, S. (2011) ‘European Integration and the Radical
Right: Three Patterns of Opposition’, Government and Opposition 46 (2): 223–244
Vasilopoulou, S. and Halikiopoulou, D. (2015) The Golden Dawn’s Nationalist Solution:
Explaining the Rise of the Far Right in Greece, New York: Palgrave. ISBN
9781137487124.
38
Vlandas, T. (2013) The politics of temporary work deregulation in Europe: solving the
French puzzle. Politics and Society, 41 (3): 425-460.
Wimmer, A. (1997). Explaining xenophobia and racism: a critical review of current
research approaches, Ethnic and Racial Studies, 20(1): 17-41.
39
Appendix
Table A1: descriptive statistics, definitions and sources for variables
Variable Observatio
ns
Mean Std.
Dev.
Minimum Maximum Description Sources
% vote for far right
parties
80 7.410 8.75778 0 38.93 % vote for far right parties. For
our classification see
Appendix 2. Years used for
each country: 2004, 2009 and
2014.
For far right results: EU
Election Database
http://eed.nsd.uib.no/webvie
w/index.jsp?
study=http://129.177.90.166
:80/obj/fStudy/ATEP2004_
Display&node=0&mode=c
ube&v=2&cube=http://129.
177.90.166:80/obj/fCube/A
TEP2004_Display_C1&top
=yes
and
http://www.europarl.europa.
eu/elections2014-
results/en/election-results-
2014.html
Unemployment
rate
84 8.736 4.525451 3.1 27.5 Unemployment rates
represent unemployed
persons as a percentage of
the labour force. Years used
for each country: 2003, 2008
and 2013.
Eurostat
Crisis dummy 84 0.25 0.435613 0 1 1 for 2014, 0 otherwise Author’s calculation
Real GDP 83 1.516 2.709865 -5.4 8.6 Real GDP growth rate –
volume. Percentage change
on previous year. Years used
for each country: 2003, 2008
and 2013.
Eurostat
Post-communist 84 0.357 0.482035 0 1 Equals 1 for 12 countries that Author’s calculation
40
Variable Observatio
ns
Mean Std.
Dev.
Minimum Maximum Description Sources
dummy joined the EU in 2004 and
2007 enlargements (excluding
Cyprus and Malta)
Unemployment
benefits
79 0.583 0.137381 0.26 0.9 Net unemployment
replacement rate for an
average worker, one earner
couple with two children. The
dataset scales the
replacement rate between 0
and 1 (i.e. 0.75 means that
that 75% of the income prior
to unemployment is being
replacement by
unemployment benefits).
Years used for each country:
2003, 2008 and 2011.
Van Vliet, Olaf & Koen
Caminada (2012),
‘Unemployment
replacement rates
dataset among 34
welfare states 1971-
2009: An update,
extension and
modification of Scruggs’
Welfare State
Entitlements Data Set’,
NEUJOBS Special
Report No. 2, Leiden
University
Electoral
frationalisation
81 76.529 8.19055 50.6051 90.0687 Index of electoral
fractionalization of the party
system according to the
formula [F] proposed by Rae
(1968). Years used for each
country: 2003, 2008 and 2011.
Klaus Armingeon,
Romana Careja, Laura
Knöpfel, David
Weisstanner, Sarah
Engler, Pana-jotis
Potolidis, Marlène
Gerber. 2013.
Comparative Political
Data Set III 1990-2011.
Bern: Institute of Political
Science, University of
Berne.
Proportional
representation
dummy
81 0.925 0.263523 0 1 Proportional Representation?
(1 if yes, 0 if no): “1” if
candidates are elected based
on the percent of votes
received by their party and/or
DPI2012 (2012)
Database of Political
Institutions: Changes and
Variable Definitions.
Philip Keefer,
41
Variable Observatio
ns
Mean Std.
Dev.
Minimum Maximum Description Sources
if our sources specifically call
the system “proportional
representation”. “0” otherwise.
Years used for each country:
2003, 2008 and 2011.
Development Research
Group, The World Bank.
Total trade
(% of GDP)
84 113.80 65.85811 0 371.4 Sum of Export and Import of
Goods and Services, % of
GDP. Years used for each
country: 2003, 2008 and 2013.
Eurostat
Industry 83 27.096 6.026859 12.4 40.2 Industry employment as % of
Total Employment
European Commission
Immigration bad
for culture
64 29.959 12.96557 7.1 68 Respondents have to position
themselves on the following
statement: “Country's cultural
life undermined or enriched by
immigrants”. They choose a
number between 0 (Cultural
life undermined) and 10
(Cultural life enriched). This
variable sums the share of
respondents that have
responded 0 to 4 (since 5 is a
neutral position it is not
included in our “immigration is
bad for culture” variable).
Three ESS waves were used:
2004, 2008 and 2012.
European Social Survey,
closest wave preceding
the EP election
Employment
Protection Index
61 2.6069 0.447958 1.55839 4.09524 Strictness of employment
protection – individual and
collective dismissals (regular
contracts). Years used for
each country: 2003, 2008 and
2013.
OECD and Avdagic, S.
(2012a), ‘EPL Index in
Central and Eastern
Europe, 1990–2009’. UK
Data Archive, Economic
and Social Data Service.
Retrieved 31 May 2012
42
Variable Observatio
ns
Mean Std.
Dev.
Minimum Maximum Description Sources
Total Immigration 78 132.23
7
193.341 0.967 768.975 Immigration (1000s). Years
used for each country: 2003,
2008 and 2012.
Eurostat
Effective number
of parties
81 4.784 1.705698 2.024501 10.06917 Effective number of parties on
the votes level according to
the formula [N2] pro-posed by
Laakso and Taagepera
(1979). Years used for each
country: 2003, 2008 and 2011.
Klaus Armingeon,
Romana Careja, Laura
Knöpfel, David
Weisstanner, Sarah
Engler, Pana-jotis
Potolidis, Marlène
Gerber. 2013.
Comparative Political
Data Set III 1990-2011.
Bern: Institute of Political
Science, University of
Berne.
Inequality ratio 75 4.708 1.096895 3.1 7.4 Inequality of income
distribution, top to bottom
income quintile share ratio.
Years used for each country:
2003, 2008 and 2011.
Eurostat
Passive Labour
market policies
68 0.9007 0.679872 0.15 2.662 Spending on passive labour
market policies. Years used
for each country: 2003, 2008
and 2012.
OECD
Active labour
market policies
65 0.4329 0.344162 0.034 1.597 Spending on active labour
market policies. Years used
for each country: 2003, 2008
and 2012.
OECD
OECD
unemployment
benefit rate
71 36.360 14.56975 17.41 65.24 Net unemployment insurance
and unemployment assistance
benefits for an average
worker. Years used for each
country: 2003, 2008 and 2012.
OECD, Tax-Benefit
Models
43
44
Table A2: Robustness to inclusion of additional variables
Column (1) (2) (3) (4) (5) (6) (7) (8)
Unemployment rate -0.63 -0.21 -0.74 -0.65 -0.64 -0.62
Real GDP -1.22 -0.95 -1.32 -1.23 -1.21 -1.19 -1.19
Unemployment benefits -6.80*** -6.74*** -7.02*** -6.57*** -6.82*** -6.77*** -6.78*** -6.82***
Electoral fractionalisation 2.27t2.28 2.19 2.08 2.26 2.28t2.25t2.30t
Proportional Representation -7.42* -6.05 -7.29 -9.70** -7.46t-7.35* -7.84* -7.47*
Total Trade 0.67 1.12 0.78 -0.31 0.93 0.76
Share of industrial employment -2.63* -2.37t-2.82* -2.74* -2.62t-2.64* -2.45t-2.57*
EPL 0.80 0.10 0.63 0.84 0.83 0.77 0.82 0.81
Total immigration 2.08t2.05t2.10t2.20t2.09t
Immigration from EU countries 2.91t
Immigration from non-EU
countries
1.94*
Crisis dummy 0.91
Immigrant population 0.80
Exports 0.64
Imports 0.67
Youth unemployment -0.37
Long term unemployment -0.64
Constant 14.98*** 13.37** 14.71*** 17.32*** 15.02*** 14.96*** 15.34*** 15.03***
Observations 56 52 51 56 56 56 56 56
Number of id 21 21 21 21 21 21 21 21
R-squared within model 0.31 0.43 0.37 0.26 0.31 0.31 0.32 0.31
R-squared overall model 0.44 0.39 0.43 0.44 0.44 0.44 0.43 0.43
R-squared between model 0.45 0.36 0.41 0.44 0.45 0.45 0.43 0.44
Note: Robust standard errors clustered by country (not shown for reasons of space, see: *** p<0.001, ** p<0.01, * p<0.05, t p<0.1). All
independent variables are standardised around mean 0 with standard deviation of 1. Thus the coefficients are directly comparable:
they show the effect on far right support of a 1 standard deviation change in the independent variable.
45
Table A2: Robustness to inclusion of additional variables (continued)
Column (9) (10) (11) (12) (13) (14) (15)
Unemployment rate -1.02 -0.22 -1.22 -1.40 -0.59 -0.40 .14
Real GDP 0.23 -1.70 -1.57 -1.73 -1.06
Unemployment benefits -6.68*** -6.78** -5.90* -6.00* -7.66***
Electoral fractionalisation 2.15 2.07 2.54 2.89t3.16t3.11t-.002
Proportional system -7.17t-6.91t-8.87* -5.19 -9.34 -9.17 -11.92**
Industrial employment -2.46 -3.02* -1.03 -2.34 -2.72t
EPL 0.73 0.76 -1.68 -1.49 -0.26 -0.02 2.21
Total immigration 2.04t2.36t0.44 1.95t1.92 1.80 .848
Total trade (% GDP) 0.85 0.75 -0.98 -2.25 1.51 1.45 1.43
Debt (% GDP) 1.32
Deficit (% GDP) 0.36
Spending on active labour market
policies
0.77
OECD unemployment benefit
replacement rate
-4.56**
Inequality ratio between top and
bottom quintile
1.71
Gini coefficient 1.45
% who think immigration bad for
country’s culture
-2.22
Constant 14.73*** 14.67*** 15.50*** 12.68*** 17.14*** 16.89*** 19.25***
Observations 56 56 45 56 50 49 48
Number of countries 21 21 20 21 21 21 18
R-squared within model 0.31 0.32 0.19 0.30 0.30 0.29 0.21
R-squared overall model 0.44 0.43 0.21 0.25 0.35 0.35 0.66
R-squared between model 0.45 0.42 0.23 0.26 0.37 0.38 0.58
Note: Robust standard errors clustered by country (not shown for reasons of space, see: *** p<0.001, ** p<0.01, * p<0.05, t p<0.1). All
independent variables are standardised around mean 0 with standard deviation of 1. Thus the coefficients are directly comparable:
they show the effect on far right support of a 1 standard deviation change in the independent variable.
46
Table A3: Regression results for interaction effects
Column (1) (2)
Unemployment rate 0.79** 2.10**
(0.281) (0.809)
Real GDP -0.03 -0.23
(0.204) (0.355)
Unemployment benefits -23.46* -39.06**
(10.902) (12.859)
Electoral fractionalisation 0.32** 0.36*
(0.116) (0.177)
Total immigration 0.01* 0.02*
(0.007) (0.007)
Unemployment rate* Unemployment
benefits
-1.30*
EPL 3.92
Unemployment rate*EPL -0.83**
Constant -5.65 -8.80
Observations 72 58
Number of id 27 22
R-squared within model 0.20 0.33
R-squared overall model 0.28 0.26
R-squared between model 0.33 0.25
Note: Robust standard errors clustered by country in parentheses; *** p<0.001, ** p<0.01, *
p<0.05, t p<0.1. The interaction effects’ magnitude and significance cannot be evaluated
from the table results (see Kam and Franzese, 2007). Note that contrary to other
regressions, the variables used in this regression have not been standardised so that
interaction effects can be shown for actual values of each variable (note therefore that the
immigration data is expressed in 1000s so values in this variable can be in the 100s).
47
Table A4: Results for regression model of table 1, column 10, without including the
UK in the sample
Column (1)
Unemployment rate -0.68
(0.879)
Real GDP -1.68
(1.220)
Unemployment benefits -7.43***
(1.980)
Electoral fractionalisation 2.45t
(1.363)
Immigration 2.71*
(1.336)
Proportional Representation -10.18***
(2.713)
Total trade (% GDP) 1.01
(1.601)
Share of industry -2.56t
(1.407)
EPL 0.77
(1.385)
Constant 17.65***
Year effects No
Country effects No
Observations 53
Number of countries 20
R-squared within model 0.35
R-squared overall model 0.34
R-squared between model 0.34
Note: Robust standard errors clustered by country in parentheses, *** p<0.001, ** p<0.01, *
p<0.05; t p<0.1. All independent variables are standardised around mean 0 with standard
deviation of 1.
48
Figure A1: Far right-wing party support, unemployment rate, and EPL (without UK)
-2 -1 0 1 2
Effects on Linear Prediction
1.5 2 2.5 3 3.5 4
EPL
Average Marginal Effects of Unemployment rate with 90% CIs
Figure A2: Far right-wing party support, unemployment rate, and unemployment
benefits (without UK)
-1 -.5 0 .5 1
Effects on Linear Prediction
.1 .2 .3 .4 .5 .6 .7 .8 .9
Unemployment benefit replacement rate
Average Marginal Effects of unemployment rate with 90% CIs
49
Appendix 2: Sources for coding of far right wing parties in Europe
Country Far right wing
party Sources
Austria Austrian Freedom
Party (FPÖ)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Austria
Alliance for the
Future of Austria
(BZÖ)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Belgium Flemish Interest
(VB)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Bulgaria National Union
Attack (ATAKA)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Croatia Croatian Party of
Rights (HSP)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Cyprus National Popular
Front (ELAM)
Katsourides, Y. (2013) ‘Determinants of Extreme Right Reappearance in Cyprus: The National
Popular Front (ELAM), Golden Dawn's Sister Party’, South European Society and Politics, 18:4
Czech
Republic
Workers’ Party of
Social Justice
(DSSS)
Mares, M. (2012) Right-Wing Extremism in the Czech Republic, International Policy Analysis, ̌
Friedrich Ebert Stiftung
Denmark Danish People’s
Party (DF)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
50
Country Far right wing
party Sources
10.1177/1354068814567975
Estonia
Estonian
Independence Party
(EIP)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Finland True Finns (PS)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
France National Front (FN)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Germany
NPD
National Democratic
Party of Germany
(NPD)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Germany The Republicans
(REP)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Greece Golden Dawn (GD)
Vasilopoulou, S. & Halikiopoulou, D. (2015) The Golden Dawn’s Nationalist Solution: Explaining
the Rise of the Far Right in Greece, New York: Palgrave. ISBN 9781137487124.
Greece Popular Orthodox
Rally (LAOS)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Greece Independent Greeks
(ANEL)
Vasilopoulou, S. & Halikiopoulou, D. (2015) The Golden Dawn’s Nationalist Solution: Explaining
the Rise of the Far Right in Greece, New York: Palgrave. ISBN 9781137487124.
Hungary Movement for a
better Hungary
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
51
Country Far right wing
party Sources
(Jobbik) 10.1177/1354068814567975
Ireland N/A
Italy (LN) Northern League
(LN)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Latvia National Alliance
(NA)
Vasilopoulou, S. (2011) ‘European Integration and the Radical
Right: Three Patterns of Opposition’, Government and Opposition 46 (2): 223–244
Lithuania Order and Justice
party (TT) Rydgren, J. (ed). (2013) Class Politics and the Radical Right, Oxon: Routledge
Luxemburg National Movement
(NB)
Lubbers, M., Gijsberts, M. and Scheepers, P. (2002), ‘Extreme right-wing voting in Western
Europe’ European Journal of Political Research 41: 345–378
Malta Imperium Europa
(IE)
Minkenberg, M. and Perrineau, P (2004) The Radical Right in the European Elections 2004,
International Political Science Review, 28(1): 29-55; Falzon, M.A and Micallef, .M (2012) Rights,
Roots and Routes: Local and transnational Contexts of Extreme Right Movements in
Contemporary Malta in Mammone, A., Godin, E. and Jenkins, B. (Eds) (2012) Mapping the
Extreme Right in Contemporary Europe: From Local to Transnational, Oxon: Routledge
Netherlands Party for Freedom
(PVV)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Netherlands List Pim Fortuyn
(LPF)
Harrison, S. and Bruter, M. (2011), Mapping Extreme Right Ideology: An Empirical Geography of
the European Extreme Right’, Basingstoke: Palgrave Macmillan
Poland Law and Justice
Party (Pis)
Pankowski, R., (2010) The Populist Radical Right in Poland: The Patriots, Oxon: Routledge;
Harrison, S. and Bruter, M. (2011), Mapping Extreme Right Ideology: An Empirical Geography of
the European Extreme Right’, Basingstoke: Palgrave Macmillan; Pankowski, R and Kormak, M.,
(2013) ‘Radical Nationalism in Poland: From Theory to Practice’ In R. Melzer & Serafin, S. (Eds.),
Right-wing extremism in Europe: Counter-strategies and Labor-Market Oriented Exit Strategies.
52
Country Far right wing
party Sources
Friedrich Ebert Foundation
Poland Congress of the
New Right (KPN)
Harrison, S. and Bruter, M. (2011), Mapping Extreme Right Ideology: An Empirical Geography of
the European Extreme Right’, Basingstoke: Palgrave Macmillan
Poland League of Polish
Families (LPR)
Harrison, S. and Bruter, M. (2011), Mapping Extreme Right Ideology: An Empirical Geography of
the European Extreme Right’, Basingstoke: Palgrave Macmillan; Immerzeel, T., et al (2015)
Competing with the radical right: Distances between the European radical right and other parties
on typical radical right issues, Party Politics, DOI: 10.1177/1354068814567975
Portugal National Renovator
Party (PNR)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Romania Greater Romanian
Party (PRM)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Slovakia Slovak National
Party (SNS)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Slovakia People’s Party-Our
Slovakia (L’SNS)
Pytlas, B. (2013) ‘Radical-right narratives in Slovakia and Hungary: historical legacies, mythic
overlaying and contemporary politics’, Patterns of Prejudice, 47(2): 162-183
Slovenia Slovenian National
Party (SNS)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
Spain National Democracy
(DN)
Ellwood, S. (1995). The extreme right in Spain. In L. Cheles, Ferguson, R. and Vaughan, M.
(eds.), The extreme right in western and eastern Europe. London, New York: Longman.
Sweden Sweden Democrats Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
53
Country Far right wing
party Sources
(SD)
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
UK British National
Party (BNP)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
UK
United Kingdom
Independence Party
(UKIP)
Immerzeel, T., et al (2015) Competing with the radical right: Distances between the European
radical right and other parties on typical radical right issues, Party Politics, DOI:
10.1177/1354068814567975
54
... Where, on the other hand, these policies are less generous, the risks and costs of unemployment are greater and the far right is more likely to increase its support. 22 This is consistent with the increasing appeal of the far right among both the working and middle classes. 23 The classic economic insecurity argument tends to automatically translate into the erroneous assumption that the economically insecure are first and foremost the working classes: blue collar, manual workers in precarious employment, or the unemployed. ...
... As shown above, unemployment increases insecurity and hence lead to higher levels of far right support through two conceptually distinct channels: because it imposes costs on the unemployed and because it increases the risks for those that are employed. 22 This suggests that, because of their impact on both the employed and the unemployed, austerity policies are likely to intensify support for the far right. For example, in Greece those who have suffered from austerity are not only the marginalised sections of the population -ie, the outsidersbut also the large middle classes -ie, those insiders who found themselves much worse off. ...
Article
Full-text available
Speed and Mannion make a good case that the rise of populism poses significant challenges for health policy. This commentary suggests that the link between populism and health policy should be further nuanced in four ways. First, a deconstruction of the term populism itself and a focus on the far right dimension of populist politics; second, a focus on the supply side and more specifically the question of nationalism and the ‘national preference’; third, the dynamics of party competition during economic crisis; and fourth the question of policy, and more specifically the extent to which certain labour market policies are able to mediate demand for the far right.
ResearchGate has not been able to resolve any references for this publication.