The European refugee crisis, party competition, and voters’ responses in Germany

Abstract

Studies show that globalisation creates political potentials that can transform electoral competition in Western societies. The specific process of how these potentials become effective is not completely understood. It is argued in the article that attention-grabbing events can trigger the transformation of electoral competition as they force actors to take clear positions and thereby allow citizens to align their partisan preferences and policy attitudes. The article analyses the case of German parties’ reaction to the arrival of large numbers of refugees at Europe’s borders in 2015/16. Using panel data that bracket this event, it is shown how German citizens responded to party behaviour by changing partisan preferences on the basis of prior immigration attitudes. The so-called refugee crisis may thus have been a critical juncture transforming party competition in Germany. As such, the crisis represents a striking example of how events may focus attention on a new policy dimension and catalyse the evolution of new cleavages.

Keywords:

• Focusing events
• party competition
• partisan attitudes
• immigration attitudes
• Germany

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Matthias Mader is a research associate at the University of Mannheim. His research interests include political psychology, foreign policy, and party politics. [[email protected]]

Harald Schoen is Professor of Political Science at the University of Mannheim. His research focuses on political behaviour, political psychology, and public opinion. He has published widely on these topics in articles and books. [[email protected]]

Notes

1 As for partisans of the CSU and the AfD, it is not clear if we should expect a change in their immigration attitudes. Both parties acted in accordance with their pre-crisis stance on immigration – i.e. they advocated restricting the refugee flows in one way or another. The cues thus did not change in the course of the refugee crisis. However, one might argue that the immigration-critical stance of these parties became more extreme, inducing partisans to develop more critical attitudes. Immigration cues also became more frequent so that immigration-friendly partisans might have learned about the party’s critical stance and aligned with it.

2 Not all waves cover all variables, which is why the waves analysed below will vary across steps of analysis. For example, perceived party positions were measured only in waves 2, 3, 7, 8, and 9. Of highest substantive interest, however, are those that bracket the refugee crisis (waves 4, 5) and the most recent, post-election wave (wave 10) – these waves include the immigration item and will thus be used in the models gauging the effect of prior immigration attitudes on subsequent partisan attitudes.

3 Below we use all observations with non-missing values that are available for each individual analysis. This results in varying numbers of observations across analyses, which slightly decreases the comparability of results from different regressions but maximises the statistical power of each regression.

4 Question wording of all items is reported in the supplemental online appendix, section A1.

5 In the Intermediate Inquiry (wave 5), a nine-point scale was used. That is why we do not consider this wave when analysing intra-individual change.

6 As another piece of evidence for the salience of the issue and citizens’ awareness of party behaviour during the crisis, citizens were increasingly able to report the parties’ positions on immigration (supplemental online appendix, section A3).

7 Perceptions about the CDU and CSU were not measured separately and AfD perceptions not at all in 2013. We combine the joint CDU/CSU perception from 2013 with the distinct CDU and CSU perceptions measured in the later waves to calculate change and distance scores for the two parties. This is justifiable because cross-sectional data show that the two parties were perceived to be quite similar before the crisis (as shown in Figure 1(b)), with more than 70% of respondents perceiving no or only a one-point difference between party positions and perceptions correlating with r = 0.80.

8 At the same time, the immigration-friendly citizens remained closer to the SPD, the Greens, and the Left (not shown in tabular form).

9 Similarly, the absolute distance to the CDU increased on average among respondents who had been in favour of relaxing immigration restrictions before the crisis while it decreased on average among respondents who had been in favour of tougher restrictions. Again, the pattern for the CSU is reversed. See the online appendix, section A3 for these results.

10 Tables of these and all other regressions reported in this article are documented in the online appendix, section A5.

11 In addition, we accounted for change in absolute distance with regard to welfare. Given our interest in the (total) implications of the refugee crisis, we omitted potential mediators. For example, while (change in) party identification is an obvious candidate for a confounding variable (as it might influence immigration attitudes, perceived party positions, and approval ratings), we consider it as a potential mediator: citizens might change partisanship in reaction to a growing distance between their immigration attitude and the party position. As we show below, there is indeed evidence of such far-reaching change in loyalties.

12 In these and the following regressions we additionally controlled for change in immigration attitudes, prior welfare attitudes, and change in welfare attitudes.

13 Additional results for changes towards the CDU/CSU and away from the AfD are reported in the online appendix, section A4. It is noteworthy that the welcoming refugee policy apparently had no electoral upside for the CDU and CSU: changes in absolute distance and prior immigration attitudes were unrelated to changes in vote choice and party identification towards the CDU/CSU.

14 We can go one step further with our data and leverage the fact that the identification item included three response options in the 2017 Campaign Panel (‘CDU/CSU’, ‘CDU’, and ‘CSU’). Dissatisfaction with the refugee policy should have induced immigration-critical CDU partisans to switch away from the pure ‘CDU’ option. As reported in section A4 of the online appendix, this is indeed the case: CDU partisans who were critical of immigration before the crisis were less likely to choose the ‘CDU’ option and more likely to report an identification with the CSU, the AfD, and no partisanship after the crisis.

15 In the case of the AfD, we also find small effects of previous immigration attitudes. Accordingly, opposition to immigration at the start of the 2013 campaign increased the likelihood of switching vote choice and partisanship to the AfD. Although citizens in 2013 did not perceive the party to be as immigration-critical as in the following years (Figure 1(a)) – in fact, many citizens were not sure where the party stood on immigration in 2013 (supplemental online appendix, section A3) – the AfD seems to have already been able to attract opponents of immigration during this first phase (see also Mader and Schoen 2015; Schmitt-Beck 2017).

16 The midwife of the AfD was the euro crisis of 2010, which was an economic crisis. But from its founding in 2013, at least some AfD leaders’ opposition to European integration was coloured by nationalist – or cultural – sentiment (e.g. Grimm 2015); early party support was already driven in part by critical attitudes towards immigration (Mader and Schoen 2015; Schmitt-Beck 2017).

Additional information

Funding

This work was supported by the German Research Foundation [SCHO 1358/4-3].