The Populist Radical Right and European Integration: A Comparative Analysis of Party–Voter Links

Abstract

This article analyses the links between populist radical right parties and their voters regarding European integration in 11 European countries. It does so by using data from the 2008 European Social Survey and the 2006 UNC-Chapel Hill Expert Data Base on political parties and European integration. In addition to mapping the Eurosceptic orientations of political parties and their voters, the article examines the degree to which attitudes towards the EU and voting for populist radical parties are connected to each other. The results lend support to the hypothesis that most populist radical right parties have managed to establish links with their voters regarding European integration. The analysis also shows that links between populist radical right parties and their voters tend to be stronger for those parties that adopt more extreme negative positions towards European integration.

Acknowledgements

We would like to thank Ian Down, Meindert Fennema, Aitor Hernández-Carr, Liesbet Hooghe, Gary Marks, Anthony Messina, Patricia Otero, Nina Wiesehomeier, Carole Wilson, and two anonymous reviewers for their helpful comments on previous versions of this article. We also thank the Spanish Ministry of Science for funding the research project (Ref CSO2010-15143) that made this article possible.

Notes

1. We have decided to use the ESS4-2008 instead of the 2009 European Parliament Election study because it was conducted closer in time to the 2006 UNC-Chapel Hill expert data base, and also because the ESS data includes higher numbers of populist radical right voters than the 2009 European Election Study (twice as many on average for the countries considered here). In some cases (such as the French FN) the number of voters in the 2009 European Election Study is particularly low (eight voters for the FN in the last general elections and four in the last European elections).

2. Included are political parties that attained at least 2 per cent of the vote in the election immediately prior to the survey year or that elect at least one representative to the national parliament (Hooghe et al. 2010).

3. Many labels have been applied to this type of parties (new radical right, populist radical right, extreme right, etc.) (see for instance Hainsworth 2007; Kitschelt 1995; Mudde 2007). Following Mudde (2007), in this article we opt for the populist radical right label. We prefer this denomination to that of extreme right because the latter has explicitly authoritarian overtones (Mudde 2007: 23–31) and is often linked to the experience of fascist regimes and neofascist movements (Ignazi 2004). We also prefer it to the quite extended new radical right label because this term denotes an economically neoliberal programme (Kitschelt 1995) that is not shared by most of these parties.

4. Due to the presence of important differences in their origins, ideological profiles, and social bases, considerable attention has been dedicated to describe and explain the heterogeneity of this party family through the identification of different subtypes of populist radical right parties (Carter 2005; Kitschelt 1995).

5. See in particular Mudde (2007: 16 and the following) and Hainsworth (2007: 7). Several authors have also stressed the populist radical right rejection of key elements of liberal-democratic regimes and their anti-establishment populism (Carter 2005: 17–28; Hainsworth 2007: 12; Ivaldi 2004; Mudde 2000: 177; Taggart 1995).

6. On the effects of internal ideological tensions on electoral competition on specific issues, see Odmalm (2011).

7. On the trade-off between definitional precision and empirical problem-solving, see Popper (2002: 20–30).

8. Boomgarden et al. (2011) have revealed the multidimensional character of attitudes towards the EU (see also Hobolt et al. 2011). Due to data limitations, we cannot map the positions of both parties and voters in each of these specific dimensions.

9. This result is not independent of the fact that public Eurosceptic positions are often underrepresented in national party systems, given the overall pro-EU stance of mainstream party families (Taggart and Sczcerbiak 2008).

10. In order to measure public attitudes towards the EU we have used question B34 of the ESS survey, which is phrased as follows: ‘Now thinking about the European Union, some say European unification should go further. Others say it has already gone too far. Using this card, what number on the scale best describes your position?’ Responses range from 0 (‘Unification has gone too far’) to 10 (‘Unification should go further’). We also use question 1 in the 2006 CHES questionnaire: ‘how would you describe the general position on European integration that the party leadership took over the course of 2006?’, ranging from ‘strongly opposed’ (1), to ‘strongly’ in favour (7). As previously mentioned, we cannot locate the positions of citizens and parties in the different dimensions underlying European attitudes (Boomgaarden et al. 2011). We assume, however, that our indicator for public opinion Euroscepticism is especially connected to attitudes towards the strengthening of the EU, the extension of decision-making competencies, policy transfer, and further integration (Boomgaarden et al. 2011: 258).

11. These averages are based on the respondents voting for each populist radical right party in the last national election (question B12). The number of voters for each party responding to this question on European integration (question B34) is as follows: DF (127), LAOS (46), LNNK (82), FN (29), NOA (88), PRM (142), PVV (35), VB (97), SNS (88), PiS (291), and TF (52). Unfortunately, the numbers of respondents for some of these parties are very low. As previously mentioned, using the 2009 European Election Study would lead to a more dramatic limitation on the number of respondents. The average voting positions in the ESS4-2008 and 2009 European Election Study are however highly correlated for these 11 cases (Pearson correlation of +0.88). From 2008 to 2009 party voters have become more Eurosceptic in nine of the 11 cases considered here, the exceptions being the PiS and the SNS voters. The average position of party voters towards the EU (11 cases) moved between these two surveys from 4.99 to 4.10.

12. Among these, only the voters of TF, FN, and DF adopt hard Eurosceptic positions (with around 50 per cent of their voters positioning themselves below point 3 in the 0 to 10 scale of Europeanism we use in this paper). The percentage of PVV voters adopting strong Eurosceptic positions (3 or below) is also high (51 per cent), but in this case the party does not adopt Eurosceptic positions.

13. The Pearson correlation index between these two variables equals +0.53. This index rises to +0.73 when the TB-LNNK is excluded from the analysis.

14. In this article we do not examine whether populist radical right parties cue their voters in the direction of Euroscepticism or whether they appeal to already Eurosceptic voters by adopting anti-EU positions. Assessing the strength of these two causal flows would demand a different kind of analysis.

15. With the only exception of Belgium, we have conducted these multivariate analysis at the national level. In Belgium, where ESS respondents are divided between Wallonia and Flanders, we conducted the analysis just for the voters in the Flanders region.

16. For the wording of this question and the scale used, see note 10. As we mentioned above, we assume that this question is particularly related to one specific dimension of European attitudes, the one that refers to the strengthening of the EU, the extension of decision-making competencies, policy transfer, and further integration (Boomgarden et al. 2011: 258).

17. These dummy variables are based on responses to Question F32 (HINCTNTA). Respondents in the lowest two income groups were labelled as low income, and assigned a value of 1 in the low income variable, whereas respondents in the two highest income groups were assigned a value of 1 in the high income variable. Due to the absence of data for Question F32, these two dummy variables were not included in the OLS models for Bulgaria and Slovakia.

18. These variables are based on responses to question F6 (EDULVL).

19. This variable is based on responses to variable F30 (MBTRU). Respondents who belonged (currently or previously) to a trade union were assigned a value of 1. All other respondents were assigned a value of 0.

20. Question B23, ranging from 0 (left) to 10 (right).

21. In order to assign values to this variable, individual self-placements in the left–right scale (question B23) were standardised (at the national level) and then squared.

22. Question C21, ranging from 0 (not at all religious) to 10 (very religious). However, we could not find any indicator for religious intolerance, which has been shown to have a positive impact on Euroscepticism (Hobolt et al. 2011).

23. In order to map orientations towards immigrants we have chosen variable B39 (IMUECLT), which reveals perceptions regarding the effects of immigration of national culture. Responses to this variable range from 0 (immigrants undermine the country’s cultural life) to 10 (immigrants enrich the country’s cultural life). Our selection of a variable focusing on the cultural implications of immigration is based on the importance that cultural demarcation has for both populist radical right parties and voters.

24. Our results are consistent with Lachat’s (2008) findings for the case of France, the only country in which his study covers a populist radical right party examined here. His study reaches similar conclusions in the Netherlands, but his work focuses on voting for the LPF instead of for the PVV. His results also reveal the presence of statistically significant links between voting for populist radical right parties and European integration in the case of the Swiss SVP, but not in the case of the Austrian FPÖ.

25. In fact, for this group of parties there is no association between the level of nationalism and nativism, as measured by questions 31 and 25 in the UNC-Chapel Hill Expert Survey, and their links with their voters regarding European integration.

26. For the sake of simplicity, and since absolute and relative party positions are strongly correlated (Pearson correlation of –0.85 between the absolute populist radical right party positions and the distance between populist radical right party positions and the weighted average of the party-system positions), we will leave aside from this exploration the connections between these two indicators of party–voter links and the relative positions of populist radical right parties in their respective party systems.

27. The Pearson correlation index between party positions and the summary for party–voter links equals –0.72, and that between party positions and the B-coefficients in the OLS analysis equals +0.74. Certainly, the first two indicators on which we based our summary index already include information on party positions (whether parties are Eurosceptic and whether they are the most Eurosceptic ones), but they also take into consideration the absolute and relative positions of party voters. As for the B-coefficients, they are not connected a priori with the positions adopted by parties regarding the EU.

28. Party dissent and party saliency are already connected to party positions on European integration. The Pearson correlation between populist radical right party positions and party dissent equals +0.94, and that between party positions and party salience equals –0.61 (2006 UNC-Chapel Hill data base).

29. In fact, the Pearson correlation indexes reported in Table 4 showed that anti-immigration and Eurosceptic attitudes are positively related with each other among the populations of all the countries analysed here.