Far right parties and ‘Europe’: societal polarization and the limits of EU issue contestation

ABSTRACT

Europe has experienced both marked growth in public support for far right parties and growth in public Euro-skepticism. While far right parties have become associated with particularly strident negative orientations towards the European Union (EU), there is considerable variation in the extent of their antipathy, and, in the extent to which they emphasize the EU. Here we examine the conditions under which far right parties can successfully leverage the EU issue to win votes. Specifically, we argue that the extent of societal polarization on the EU constitutes an important intermediating condition. Our findings indicate that individuals are more likely to vote far right when these parties contest the EU, but only if they do so in a context of societal polarization on the issue.

KEYWORDS:

• Far right parties
• societal polarization
• European Parliament election
• European Union
• issue contestation

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed here.

Notes

1. Far right parties in this paper are those that hold extreme positions on the overall left-right ideological dimension. More specific aspects of the concept, the criterion, and the list of these parties are presented in subsequent sections of the paper.

2. Hooghe and Marks 2009.

3. Down and Wilson 2008.

4. de Vries 2010.

5. Marks and Wilson 2000. Parties on the far left are frequently anti-EU in orientation, but there is not the same degree of consistency or hostility across far left parties as there is amongst those on the far right (Hooghe, Marks, and Wilson 2002).

6. Kriesi et al. 2006.

7. See Lubbers, Gijsberts, and Scheepers 2002.

8. See van der Brug, Fennema, and Tillie 2000; Lubbers, Gijsberts, and Scheepers 2002; Arzheimer 2009.

9. Steenbergen, Edwards, and de Vries 2007; Gomez-Reino and Llamazares 2013.

10. See Lindberg and Schiengold 1970, and Hooghe and Marks 2009.

11. Carruba 2001; de Vries 2007; de Vries 2010; Down and Wilson 2010; de Vries, Edwards, and Tillman 2011; de Vries and Tillman 2011; Tillman 2004.

12. de Vries 2007.

13. See, for example, Carruba 2001.

14. Hix and Lord 1996; Hooghe, Marks, and Wilson 2002; Steenbergen and Scott 2004.

15. Werts, Scheepers, and Lubbers 2012; Gomez-Reino and Llamazares 2013.

16. See Kopecký and Mudde 2002; Hooghe, Marks, and Wilson 2002; de Vries Catherine and Edwards 2009. Regarding the mobilization of the EU issue as a party strategy, the political parties that tend to politicize the EU are those on the periphery of the party system (Taggart 1998) and those out of government (Sitter 2001).

17. As such, far right party behavior may not always reflect simple strategic considerations. For example, our data shows that public opinion is statistically significantly more negative and divided on the EU issue in Austria than in France. Yet, Vasilopoulou (2018) finds that while the French National Front, particularly until its recent modernization of party programs (Mondon 2014), had supported withdrawal from the EU. In contrast, the Freedom Party of Austria has not opposed European integration, despite its critical stance on EU enlargement and the practices of EU institutions (see also Almeida 2010). And indeed, our issue contestation score for the National Front is statistically significantly greater than the issue contestation score for the Freedom Party of Austria.

18. See, for example, Akkerman, de Lange, and Rooduijn 2016.

19. Redding and Viterna 1999.

20. Carter 2005; Art 2007; Mudde 2007.

21. Huchfeldt, Johnson, and Sprague 2003.

22. Selek 2006.

23. Art 2007; Karp and Banducci 2008.

24. DiMaggio, Evans, and Bryson 1996.

25. Downs 1958.

26. Jacobson 2005.

27. Though Down and Wilson (2010) have shown that small parties have played a role in the development of EU issue contestation.

28. Bale 2003, 76.

29. de Vries 2007.

30. See Hooghe, Marks, and Wilson 2002; de Vries Catherine and Edwards 2009; Hobolt, Spoon, and Tilley 2009.

31. Endogeneity may be a problem if the extreme positioning of far right parties itself generates party system polarization. However, this hinges on the behavior of other political parties. For example, as Bale et al. (2010) note, other political parties respond in a heterogenous fashion to far right party success (i.e. some move closer, some move further away, and yet others hold their position). Accordingly, in the absence of a reason for expecting a common response on the part of other political parties to the adoption of an extreme position by the far right, we do not have a basis for expecting such positioning to simply cause polarization and endogeneity should not present a problem. Empirically, there is a lack of consistency across countries in the change over time in far-right party positions on European integration and party system polarization. For example, far-right party positions and party system polarization did move in the same direction (either a far-right party became more Eurosceptic and the system more polarized, or, a far-right party became less Eurosceptic and the party system less polarized) in five party-cases (the Freedom Party of Austria, Alliance for the Future of Austria, Northern League, Slovak National Party, and Slovenian National Party). However, the opposite occurred (a far-right party became more Eurosceptic but the party system less polarized, or, a far-right party became less Eurosceptic but the party system more polarized) in six party-cases (Flemish Interest, Danish People’s Party, National Front, National Alliance, For Fatherland and Freedom, and Law and Justice). In the remaining three party-cases, we could not make a judgment because the two measures of polarization (i.e. standard deviation and kurtosis) moved in opposite directions over time, one indicating an increase in polarization and the other a decrease. In short, there is little evidence that party system polarization is a product of the Eurosceptic positions of far-right parties. Nonetheless, to obviate any possibility of endogeneity we exclude far right parties from the measurement of party system polarization.

32. de Vries Catherine and Edwards 2009. The list of parties is presented in the supplementary appendix. We also test a broader and a narrower set of far right parties in the supplementary appendix.

33. Electoral support for far right parties is not substantial in some countries. As such, depending on the sample size, the vote propensity data from surveys may not correctly represent the actual electoral support far-right parties receive. However, in our survey data, the average vote propensity for far right parties in European Parliament elections (9.5 percent) is quite similar to the average vote share of the same parties in the corresponding actual elections (2004, 2009, 2014) (10.1 percent).

34. See de Vries and Hobolt 2012.

35. The most recent round of the CHES prior to each EES wave is used. The CHES 2002 is used for the EES 2004, the CHES 2006 for the EES 2009, and the CHES 2010 for the EES 2014. Party position and salience scores in the CHES data are averages of experts’ estimations. Therefore, there can be concern over the reliability and consistency of the data. However, experts’ estimations on party position and salience are normally distributed in a statistically significant way for almost all far right parties in the dataset – 83% of all such parties in the CHES data used in the analyses. In addition, in the CHES 2010 data, the average within-a-party standard deviation (among experts) of far right party’s position and salience is smaller (0.81 and 0.49, respectively) than the standard deviation (between parties) of the average position and salience scores of the parties (1.32 and 0.71, respectively). As such, we have no reason to be concerned with the possibility of a nonstandard distribution of experts’ estimations or with the possibility of excessive noise in the estimations.

36. While a greater value of standard deviation indicates a more dispersed distribution a smaller value of kurtosis indicates a more bimodal distribution. We employ both measures because they indicate different aspects of polarization: while the dispersion of opinions (standard deviation) captures the average distance from the mean, the modality of opinions (kurtosis) captures the peakedness of the distribution.

37. The precise question wording is ‘Generally speaking, do you think that (OUR COUNTRY’S) membership of the European Union is … … a good thing, a bad thing, neither good nor bad, don’t know.’

38. See, for example, Lubbers, Gijsberts, and Scheepers 2002; van der Brug, Fennema, and Tillie 2000. The inclusion of an individual’s position on European integration is needed in order to examine whether societal polarization has an intermediating effect on far right party vote propensity, even after controlling for its spatial model effect (i.e., reducing the position distance between voters and parties).

39. We also include the interaction between migration and unemployment, though the literature is not consistent on the expected effect: Golder (2003) finds a positive effect (i.e., migration and unemployment strengthen each other’s effect on party support) while Arzheimer (2009) finds a negative effect and suggests a ‘ceiling’ (i.e., that each variable has a support-increasing effect only when the value of the other variable does not exceed a certain threshold).

40. The horizontal axes in Figure 2 range from the 10^th percentile to the 90^th percentile of each of the variables.

41. Predicted probabilities are calculated with all other variables fixed at their mean.

42. See, for example, de Vries 2007; de Vries and Tillman 2011.

43. Some work on EP elections has highlighted the wide variance in the substantive impact of the EU issue on voting behavior (de Vries and Tillman 2011).