Issue salience and affective polarization

Abstract

ABSTRACTWhich voters hold polarized affects for political parties in Western Europe? We consider distinct characteristics of different political issues that shape political actors’ behaviors and argue that voters are more affectively polarized when they put salience on cultural issues because their stances on the issues are embedded in their deep-seated identity, value, belief, and morality. Empirically, we use measures of affective polarization that incorporate the multiparty systems of Western European countries. Using the Comparative Study of Electoral Systems (Module 3), we find that voters who put salience on cultural issues are more affectively polarized than others. The result implies that rising affective polarization in the past decades might have been related to increasing priorities on cultural issues. It also implies that political parties may potentially weaken voters’ affective polarization by manipulating their issue agenda. Disclosure statementNo potential conflict of interest was reported by the author(s).Supplementary materialSupplemental data for this article can be accessed at https://doi.org/10.1080/17457289.2023.2277429Notes1 Literature on affective polarization distinguishes voters’ positive and negative feelings on political parties, their elites (e.g., party leaders), and their supporters. Affective polarization on one of these does not necessarily lead to that on others (e.g., Knudsen Citation2021). We limit our concept of affective polarization to a voter’s different and contrasting feelings toward political parties.2 Unlike the concept of voters’ ideological polarization, which indicates that voters come to have more different ideologies each other, voters’ affective polarization is basically an individual-level concept because it means that each voter comes to have more different affects for different political parties.3 Consequently, while there is one clear way to measure affective polarization in the two-party system (i.e., the difference in party affects toward two political parties), we need an alternative way to measure how diverse voters’ affects toward multiple political parties are (Wagner Citation2021).4 In this paper, we use the term of “cultural issues” to describe “new politics” or “post-materialism” issues (Flanagan and Lee Citation2003).5 Voters may dislike only political parties that belong to a rival party bloc. Nonetheless, in addition to grand coalitions that include both (left-wing and right-wing) major parties, coalition governments that embrace political parties from both sides of the political spectrum (typically a major party from one side and a minor party from the other side) have occurred in many Western European countries. For example, the Social Democratic Party government in Denmark in 1990 invited the Christian People’s Party, and the Finnish right-wing government in 2007 invited the Green League.6 As far as we are aware, there is no empirical analysis that examines how voters’ affects toward political parties or their supporters determine coalition-building or inter-party cooperation. However, literature supposes such a plausible causal relationship as well. For example, in their research on the effect of inter-party cooperation on affective polarization, Bassan-Nygate and Weiss (Citation2022) acknowledge that political coalition can be endogenous to political attitudes and party affects and therefore employ natural and survey experiment research designs.7 There is no consensus among scholars, media, and political elites about what populism means. There is disagreement even about whether populism is political ideology, political style of communication, or political strategy (Abts and Rummens Citation2007; Moffitt and Tormey Citation2014).8 Multiples studies examined voter-level factors for affective polarization in the U.S. Mason (Citation2016) studies how voters’ identity alignment shapes affective polarization, and Ondercin and Lizotte (Citation2021) find a gender effect. Rogowski and Sutherland (Citation2016) and Webster and Abramowitz (Citation2017) also find that voters with extreme ideologies are affectively polarized.9 While Iyengar and Westwood (Citation2015) find that voters’ party affect is strengthened when their partisan identity is combined with their cultural identity such as race and ethnicity, Webster and Abramowitz (Citation2017) offer a different finding with saying that American voters’ party affects are more closely related to their preferences on welfare programs than those on abortion or gay right. Nonetheless, welfare issues are not purely economic issues because they are “racialized” in the U.S. (Brown Citation2013). In addition, abortion and gender issues are still very minor issues in the U.S., and only 0.8 percent of people indicated either issue as the most important problem in the American National Election Studies (Citation2012).10 Average moralization scores are 3.90 (cultural issues), 3.22 (neither cultural nor economic issues), and 2.95 (economic issues) in the 1 (not at all based on moral values) ∼ 5 (a great deal based on moral values). We categorize budget deficit, economic recession, and unemployment as economic issues and abortion, same-sex marriage, and environment as cultural issues. Issues of the War of Afghanistan, education, health care, and illegal immigration are not categorized into any issue group. We do not add “illegal immigration” to any issue group because it has both economic and cultural implications. Nonetheless, adding the issue to any issue group does not change the result on moralization by issue.11 Scholars do not completely agree on whether people’s issue salience determines their personal political behavior (e.g., Johns Citation2010). Nonetheless, it is found that issues on which voters put salience are a major determinant of their views on political parties and governments (Singer Citation2011; Han Citation2022).12 The CSES (Module 3) includes the following Western European countries: Austria, Denmark, Finland, France, Germany, Greece, Ireland, the Netherlands, Norway, Portugal, Spain, Sweden, and Switzerland. However, Austria and Ireland are excluded because responses to the issue salience question are not available for these countries. Each of the following countries has 2 election years in the CSES (Module 3): Finland, Germany, the Netherlands, and Norway.13 likei¯=∑p=1N(vp×likeip)14 The correlation between the two measures is 0.8131.15 We use the CHES wave prior to the election year in the CSES. See the supplementary appendix for the list of cultural issues in the CHES.16 At the same time, voters’ salience between different issues is influenced by macro-level experiences such as income inequality and immigration. See the supplementary appendix for data sources.17 We also performed diverse robustness checks. See the supplementary appendix for the discussion and results of the robustness checks.18 It may be plausible that there is spurious relationship between issue salience and affective polarization. For example, populist party supporters may both hold a high level of affective polarization and put salience on cultural issues. Their party affects may be polarized because of their disapproval of established political parties. They may put salience on cultural issues because many of their parties (the populist radical right) mobilize primarily on these issues (e.g., immigration). Nonetheless, our data shows that though they demonstrate a statistically significantly higher level of affective polarization than other people in the majority of country/elections, they do not put more salience on cultural issues than others in a statistically significant way in the majority of country/elections.19 Nonetheless, we acknowledge that we did not precisely tested the causal relationship between issue salience and affective polarization in this paper.20 For example, major political parties in Ireland delegated issues on abortion and the Catholic Church to referenda to diffuse the issues and prevent minor parties from successfully campaigning on these issues (McGraw Citation2014).