Party Competition And Voter Priorities In Multi Issue Contexts PDF Download

"This dissertation uses formal and statistical methods to analyze the nature of voter priorities and party competition in multi-issue contexts. The first chapter, 'Choosing Your Battles Wisely? Activist Preferences, Party Size and Issue Selection', seeks to explain why parties emphasize particular positional issues in their campaigns. I find that the policy preferences of activists are an important influence on party platforms, and therefore, party emphasis decisions on positional issues. However, my analyses reveal party size to be a more important determinant of parties' emphasis strategies than whether a party is 'mainstream' or 'niche'. Large mainstream parties - termed 'major parties' - de-emphasize issues on which their activists are relatively extreme, whereas both small mainstream and small niche parties - 'minor parties' - emphasize issues on which their activists are relatively extreme. Further, large niche parties appear to behave more like large mainstream parties than small niche parties in this respect. Using a variety of empirical approaches, I show that these findings can be explained as the consequence of vote-maximizing choices made by parties responding to different electorates. Conversely, they cannot easily be explained by strategic error or dogmatism on the part of some parties, or by the selection of activists into parties. These patterns hold across Western and Eastern Europe, suggesting that, in a variety of information environments, the appearance of policy moderation may be viewed as advantageous by major parties, and as potentially disadvantageous by minor parties. The second chapter, 'Policy Bundling: A Model of Party Competition in Multi-Issue Contexts', formally models party strategy in contexts where voter priorities may be influenced by campaigns. The model advances an explanation for two puzzling aspects of parties' issue emphasis decisions. First, much empirical research has established that although parties do focus disproportionately on 'preferred issues' in their campaigns, they also spend much of their time focusing on the same issues, and in particular, on issues already salient to voters. However, formal models have struggled to explain why a party might address an issue on which it is widely perceived as incompetent, or on which its policies are generally unpopular. Second, existing theories of issue selection by parties do not explain what binds certain parties to unpopular positions and others to popular ones - even though positional issues, rather than valence issues, are a substantial focus of parties' campaigns. Building on evidence gathered in the first chapter, I suggest that parties' issue positions may be constrained by the policy preferences of their activists. Thus, if able to influence the importance of issues for voters, parties will prefer to emphasize issues on which the preferences of their activists are, on average, more popular. However, I argue that the extent to which a party emphasizes an issue has two potential consequences: it may not only influence the importance of the issue for voters, but also voters' certainty regarding the party's position on the issue. Consequently, parties must balance two competing incentives in their issue emphasis decisions: emphasizing the issues on which their activists' policy preferences are more popular, so as to increase the electoral salience of these issues, and emphasizing issues that are already salient to voters, in order to ensure that potentially sympathetic voters observe their positions on these issues. In the model, the relative strength of these two incentives determines whether, in equilibrium, we observe parties 'talking past each other' or, instead, 'policy bundling' - when at least one party 'bundles' popular and unpopular policies, while disproportionately emphasizing the former. To the best of my knowledge, this constitutes the only formal model of issue selection which finds that parties may frequently campaign on the same issue in equilibrium, and will campaign on unfavorable issues if these are especially salient to voters - as is consistent with the empirical evidence. The third chapter, 'The (Non) Separability of Policy and Valence in Voter Preferences', investigates the nature of voter priorities in contexts where both valence assessments and policy preferences may be important influences on vote choice. Specifically, I consider whether and how the importance of different valence attributes to voters may vary with their ideological proximity to the party or candidate - or, variation in the separability of valence from policy in voter preferences. Using survey data on British voters between 1997 and 2005, I estimate the separability of three valence attributes from policy for voters: party integrity, leader popularity and party strength. I find that how much a respondent likes the leader of a party is more important to proximate voters, but the integrity of a party is of similar importance to all voters. On the other hand, how capable a party is thought to be of strong government is of more importance to less proximate voters. The findings in this chapter bolster the work of researchers who have argued that the assumption of additive separability does not generalize to all types of valence, and also have implications for when parties might invest in particular valence attributes"--Pages ix-xi.