The causes and consequences of valence attacks in European multi-party systems
MetadataShow full item record
What are the causes and consequences of valence attacks in multi-party elections? The strategic electoral behavior of political parties in multi-party systems has so far predominantly been studied in the spatial framework, which assumes that political parties compete for votes, seats, and office with ideological, issue, and policy appeals. In this framework, parties appeal to voters with programmatic messages in election manifestos, speeches, and other campaign advertisements. Voters (more or less accurately) evaluate these programmatic messages with regard to the proximity to their own programmatic preferences and cast a vote based on this evaluation. However, spatial competition through ideology, issues, or policy is not the only way political parties can compete in elections. In fact, there is a long-standing literature, predominantly in the United States, that highlights a second dimension of party competition and voter evaluations: valence. Electoral competition based on valence is focused on voter evaluations of inherent traits and qualities of candidates and parties. Valence is more generally seen as consisting of attributes or characteristics that are universally desirable. In this dissertation, I focus on honesty, integrity, competence, unity, and charisma. I argue that the use of valence, especially valence attacks, as an electoral strategy is very risky for political parties in Europe. The party systems of Europe emerged out of the societal cleavages of the Industrial Revolution. Parties have historically build and up to now still present themselves as organizations rooted in ideology. This emphasis on ideology, issues, and policy – a focus on content – makes it more difficult to deviate from electoral strategies that are focused on content. Spatial competition certainly is not the only way parties can and do compete against each other. However, non-spatial competition can more easily be called out by competitors. It is more likely to be branded as out of the ordinary, unfair, or dirty. This means that this type of electoral competition is more likely to be associated with unintended consequences. In this dissertation, I develop the argument that political parties can reduce the risk of valence attacks and minimize their potential for unintended consequences. Valence attacks that are more likely to be perceived by competitors, journalists, and voters as legitimate electoral conduct are less likely to backfire. In the multi-party systems of Europe, this legitimacy for attacks can be borrowed from the predominant way of competing for votes, seats, and office: the spatial competition. More specifically, I argue that there are two types of valence attacks: valence attacks that are connected to issue statements and valence attacks that are not. It is the first type of valence attack that is more likely to be perceived as legitimate, the second type is more likely to be registered as something out of the ordinary or even inappropriate. Political parties should hence be more likely to use valence attacks that are connected to ideology, issues, or policy. They should also be more likely to be rewarded when they use these attacks. Parties should be less likely to use valence attacks that are not connected to spatial competition and if they do use them face repercussions from voters. In order to test my hypotheses, I use data on party campaign discussions from the Comparative Campaign Dynamics Project (CCDP). This data set includes a detailed collection of valence attacks between 60 parties in 21 elections in 10 European countries between 2005 and 2015. I supplement this party-level data with individual-level survey data from National Election Studies and the Comparative Study of Electoral Systems (CSES) project. This cross-national data is supplemented with panel data from the 2014-2017 British National Election Study, where five waves of the internet panel were matched with data from the CCP. I also rely on reports from campaigns in Europe and the U.S. that illustrate that my theoretical argument is reflected in real-world decision making of party leaders and campaign managers. The results of my statistical analyses support the legitimacy argument. Political parties that connect their valence attack statements with issues are less likely to be punished for these attacks and more likely to be rewarded. Political parties are also much more likely to use issue-related than nonissue-related valence attack statements. These findings have important implications for our understanding of electoral competition in multi-party systems. I identify the conditions under which valence attacks work in these systems and how they work as parties intend them to. I also demonstrate that spatial competition and valence are closely connected in elections. These two modes of electoral competition have been treated as separate strategies in extant literature. However, the findings of this dissertation show that they build on each other.