How do voters interpret social class appeals? Lessons from open-ended responses

AbstractRecent research has shown that parties’ group appeals, particularly working class appeals, can raise electoral support. But how do such effects come about? This article explores the mechanisms underlying group appeals by content analysing voters’ open-ended responses to working class and upper middle class appeals in a survey experiment in Denmark. The results show that voters generally connect class-based rhetoric from candidates to issues of inequality with some also reacting emotionally. Furthermore, the effect of class appeals on electoral support is conditioned by the considerations activated by the appeals, with stronger effects observed among voters who focus on the groups involved. Finally, while voters recognise traditional class-party alliances, such stereotypes do not undercut the effectiveness of the appeals; thus, also candidates from right-wing parties can use working class appeals to their benefit. These findings advance our understanding of the role of social groups in party electoral strategies.Keywords: Group appealsclass votingexperimentopen-ended responsescontent analysis AcknowledgementsThe authors gratefully acknowledge the help of Joshua Robison in collecting the data and Lasse Endby Pedersen who coded the open-ended responses. Previous versions of the article were presented at the ECPR General Conference in Innsbruck, 2022, the annual meeting of the Danish Political Science Association, 2022 as well as at the Department of Political Science, Aarhus University. The authors are thankful for all the helpful comments received on these occasions.Ethical approvalThe data for the analysis was collected in accordance with Danish law regarding participant consent and ethical approval and is available at https://osf.io/zh43j/?view_only=9fae5d6ce51749039a47adca497806e9.Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 We define considerations as any type of thought that appears in a voter’s mind when exposed to a certain stimulus – a group appeal in this case – and assume that these can be, at least partly, measured via an open-ended question (see below).2 Robison et al. (Citation2021) find that this result applies for all classes except upper middle class identifiers who tend not to react to the appeals. We return to this point below.3 On a 0 to 10 sympathy scale like the one introduced below, all four groups mentioned in the treatments scored between 7.1 and 7.4 – essentially the same, in other words.4 Note that in Denmark high school teachers are university graduates with a master’s degree, thus representing an upper middle class occupation.5 Respondents were debriefed at the end of the survey.6 In the Danish electoral system, voters have the option of voting for specific candidates, as well as for parties, implying that evaluations of single candidates based on their statements is an externally valid exercise.7 70 respondents gave responses regarding the second vignette revealing that they had noticed the pattern of the vignettes already after seeing the first two. Twenty-two of these did, however, provide a substantial response also regarding the second vignette. The remaining 48 were removed from the analysis. We have rerun all analyses separately for the first and second rounds of vignettes presented to respondents. Overall, these separate analyses reproduce the results presented although some effects fail to reach significance at the .05-level due to the reduced sample sizes. Therefore, we retain both rounds in the main analysis to increase analytical power.8 As described in the online appendix, we base this subdivision on v3 in the coding scheme that registers whether or not the overall valence of the response was positive or negative.9 Our confidence in the results is further strengthened by supplementary analyses in which we control for the length (in characters) of the open-ended responses provided. The variable is insignificant in all models and the results are indistinguishable from those presented. This means that the variables capturing the content of the responses are not merely picking up the overall elaborateness of respondents’ thinking.10 To facilitate comparability with the analyses below, only respondents who responded to the open-ended question about their considerations are included in the analysis. If all respondents are included, the results are similar, although the differences between the treatment categories are slightly smaller, see Figure A3 in the Online Appendix.11 As shown in the online appendix, we also replicate the finding that the effect of class appeals are weaker among upper middle class than working class identifiers.12 Because the effects are similar across the two categories for mentions of, respectively, the groups benefitted or disadvantaged in the appeals (see Table 1) we have combined the two categories in the analysis depicted in Figure 2. Due to the small number of respondents mentioning themselves in relation to the appeals (see Table 2) we did not include this category in the figure.13 We note that the lower sympathy scores among those for whom the otherwise popular working class appeal elicits negative emotions suggest a larger role for emotions as moderators of group appeals than implied by our theoretical discussion. This is a point for future work to explore.14 Recall, the mean evaluation of a Liberal candidate appealing to the upper middle class was around four scale points (see Figure 1), while the evaluation of a Liberal candidate appealing to the working class is around seven scale points among voters noticing the counter-stereotypical nature of the appeal (see Figure 5).Additional informationFundingThis work was supported by the Independent Research Fund Denmark under Grant no 1327-00113 to Rune Stubager.Notes on contributorsRune StubagerRune Stubager is a Professor of Political Science at Aarhus University. His research focuses on political and electoral behaviour in particular on the role of social groups as an influence on attitudes and party choice. He is one of the co-PIs of the Danish National Election Study. His work has appeared in journals like the British Journal of Political Science, Comparative Political Studies, the European Journal of Political Research, and the British Journal of Sociology. He is a co-author of the recently published book The Danish Voter: Democratic Ideals and Challenges (with Kasper Møller Hansen, Michael S. Lewis-Beck, and Richard Nadeau, University of Michigan Press, 2021). [stubager@ps.au.dk]Mads ThauMads Thau is a Senior Research Fellow at the Institute for Social Research, Oslo. He is broadly interested in democratic politics, including voting and opinion formation, party strategies, mass-elite interactions, and representation. His research has paid particular attention to parties’ use of group appeals and their electoral consequences. In addition to numerous policy reports, Thau has published journal articles in Comparative Political Studies, Public Administration, European Union Politics, Political Studies, and the Journal of Politics. [mads.thau@samfunnsforskning.no]