“He’s not my prime minister!”: negative party identification and satisfaction with democracy

ABSTRACT

Across Europe and the United States, negative party identification – an abiding aversion to a particular opposing political party is on the rise. At the same time, satisfaction with democracy is down worldwide. In this article, we explore the potential connection between these two phenomena. We develop a theory that posits a relationship between negative party identification and satisfaction with democracy where institutional structures and outcomes can mitigate the deleterious effects of negative party identification. We test our theory by drawing on data from 25 parliamentary and semi-presidential countries in the third wave of the Comparative Study of Electoral Systems (CSES) from 2006–2011 to assess the cross-national effect of negative party identification on levels of satisfaction with the democratic process in one’s own country. We find that satisfaction with democracy declines among voters in multiparty systems when an “out-group” party holds the prime ministership. At the same time, larger vote margins and the nature of the governing coalition can mitigate the effects of negative party identification. Our findings have important implications for how institutions can reduce the potential effect of negative party identification on satisfaction with democracy.

Acknowledgements

We thank Fanghui Zhao for her excellent research assistance and the reviewers and JEPOP editor for their suggestions and comments. All errors remain our own.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Jae-Jae Spoon is Professor of Political Science and Director of the European Studies Center at the University of Pittsburgh. Her work has been published in several outlets including the British Journal of Political Science, Comparative Political Studies and the European Journal of Political Research. Her research focuses on party competition and party strategy in Europe.

Kristin Kanthak is Associate Professor of Political Science at the University of Pittsburgh. Her work has appeared in the American Political Science Review, the American Journal of Political Science, and other venues. Her current research focuses on diversity and representation in politics and the effect of political institutions on diversity.

Notes

1 Although this project is largely motivated by the potential effects of negative partisanship in the US, we do not include the US in our analysis. Party systems in the majority of presidential systems in the world do not have long-term stability and their institutions are not necessarily consolidated, thus making comparisons difficult. Studies that include the US alongside other parliamentary systems, moreover, typically run separate analyses by country (see, e.g. Medeiros and Noël 2014). To explore how the US compares to other countries, we have, however, run additional analyses that examine the effect of having one’s least preferred party in the presidency. These analyses include the US and the semi-presidential countries in Europe (Elgie 2017) and control for presidential powers (Doyle and Elgie 2016). The results differ when the US is/is not included. We have therefore decided to limit the empirical tests of our hypotheses to parliamentary and semi-presidential systems.

2 According to Linz’s conception, fixed terms and ideological parties fuel this instability. The US, he claims, has avoided the inherent instability of presidential democracies partially due to “the uniquely diffuse character of American political parties” (Linz 1990, 53). American parties in the mass public were, at the time of Linz’s writing, ideologically undisciplined (Fiorina, Abrams, and Pope 2010), and evidence indicates that they remain so today (Kinder and Kalmoe 2017). Yet, the rise of negative partisanship could represent a change to that dynamic.

3 Yet, it is important to note that recent research has found that voters may be less able to identify the positions of coalition partners (Spoon and Klüver 2017) because the lines of responsibility are blurred (Powell and Whitten 1993; Whitten and Palmer 1999). One could argue, however, that negative partisans may be more likely to identify positions of coalition partners given that they can identify distinct differences between their least preferred party which holds the prime ministership and other cabinet parties. This relationship should be explored in future research.

4 In the US case, for example, losers who claim that Donald Trump is “not my president” can legitimately point to the fact that he garnered fewer votes than did his opponent, despite having won the Electoral College.

5 We only use the third wave of the CSES as the question about whether there is a party that the respondent would never vote for was not asked in the other waves of the survey.

6 Five countries appear in the dataset twice, as there were two elections during the time period covered by wave 3 of the CSES. These are the Czech Republic, Finland, Germany, Iceland and the Netherlands.

7 The question wording is as follows: “On the whole, are you very satisfied, fairly satisfied, not very satisfied, or not at all satisfied with the way democracy works in [COUNTRY]?”

8 When we run our analyses with only parliamentary countries, the results remain largely unchanged.

9 Most negative partisans in the dataset are from the mainstream party families (measured by the party the respondent feels closest to). Furthermore, when we look at the negative partisans whose out-group party holds the prime ministership by the party family the respondent feels closest to, we find that they are distributed across party families. See Tables A1 and A2 in the Appendix for the distribution for each country in the dataset.

10 Given cabinet re-shuffles, we used the government that was in place at the time the election was called.

11 Following Donovan and Karp (2017), we also estimated models that included a static Gini Index measure and a measure for the change in the Gini Index. Neither of these variables were significant and the main results remained unchanged.

12 Interestingly, however, satisfaction with democracy is only slightly lower (1.62) for positive partisans – those respondents who do not have a party for which they would never vote – which we do not include in the analysis.

13 We know too that voters in parliamentary systems will vote with government outcomes in mind. Kedar (2005), for example, demonstrates that moderate voters may vote for more extreme parties recognizing that power sharing in coalition governments may water down policy and that more extreme parties may pull policy in their preferred direction.