Anna-Sophie Kurella

Combining ideological and policy distances with valence for a model of party competition in Germany 2009

This paper addresses two problems: how can we identify a verisimilar policy space and how can we detect Nash equilibria in this space for parties’ policy positions? We argue that the ideological party positions that voters perceive are fixed during the time span of one electoral campaign and that they constrain the policies parties offer the electorate in search of optimal vote shares. We apply the valence model developed by Schofield to party competition during the German federal election campaign 2009. First three issue scales are combined with a left–right scale to form one homogeneous space in which equilibrium locations of parties are sought. The results show that local Nash equilibria in this combined space depend heavily on the start values and are implausible. Fixing the ideological dimension leads to a stable and meaningful equilibrium configuration in which large parties move to more central positions and smaller parties move to more peripheral positions in the policy space.

Party Activists in the 2009 German Federal Elections

Formal modelers of party competition often have to face the fact that their models predict far too centrist equilibrium positions when compared to empirically observed party positions. Various components have been suggested as extensions for the standard Downsian spatial model, in order to receive more plausible, diverging equilibrium configurations. One important improvement was the inclusion of a valence term that accounts for non-policy related factors that influence vote decisions.The underlying assumption is that valence describes an overall perceived external popularity or competence, that is ascribed to a party and/or its leader and cannot be attributed to the parties, policy position.This valence term is thus assumed to be exogenously and constant among the voters. The model can further be extended by the inclusion of an additional individual specific non-policy element, such as partisan bias or ideological distances to party positions. This stabilizes the formal game of party competition by diminishing the probability of parties leapfrogging each other in equilibrium configurations. Still, the predictions of those models show significant discrepancy to empirical party configurations.

Centripetal and centrifugal incentives in mixed-member proportional systems:

How does competition for first (candidate) and second ballot (party-list) votes affect the strategic positioning of parties in mixed-member proportional systems? We study this question in a simulation study of multiparty competition in the two tiers. In the first step, we use data from elections for the German Bundestag to estimate individual vote function for each tier based on ideology, policy, and valence incentives. We then use these parameter estimates to calibrate a model in which parties compete for either first- or second-tier votes. Results suggest that parties may face a dilemma when adopting a positional strategy. When national parties and their candidates hold significantly different valences, large valence advantages generate centripetal incentives whereas smaller valences exert a centrifugal pull. Overall, centrifugal incentives dominate the German mixed-member system.

The Evolution of Models of Party Competition

This chapter describes the development within the field of spatial theory for analyzing party competition. I chose a historical approach, following the major developments within the area of spatial modeling to present the separate assumptions and components of the model in the order as they were introduced in the literature. The starting point is the standard Hotelling-Downs model and many of that model’s assumptions are adapted in the model that is presented later on in the book. The most important extension to the Downsian model was the introduction of a valence term, as conceptualized by Stokes, which will be described in detail in the second section of this chapter. As we move on to the more recent research on spatial modeling, contributions span a wider range, from specifications of the valence term over the separability of preferences in multidimensional policy spaces to modeling activist influence on parties’ policy positions. Given the richness of the literature, only those contributions will be considered, that are relevant with respect to the development of our model.

A Model of Voting Behavior and Party Competition

The recent literature provides different ways to model voting behavior with variation in decision weights. In this chapter, I will follow up on that and ask how we could detect systematic variation in decision weights and explicitly integrate it into a spatial model. Following the modeling tradition of Lin and Schofield, I formulate a baseline model that combines valence with spatial distance in a multidimensional policy space. Individual voting behavior is described by a utility function, which should allow for different levels of issue salience. The assumptions and implications of the vote model will be described in detail. Following to that, the vote function will be used as the baseline for a model of party competition. Since the goal is to identify equilibrium configurations within the policy space, but the model is analytically hard to track, I will describe an optimization algorithm for analyzing optimal party strategies. Eventually, this chapter will illustrate consequences of variation in issue weights on parties’ ideal positions in equilibrium based on stylized examples, that should help to create an intuition about expectations in empirical cases.

A Theory of Issue Weighting

The objective of this chapter is to present a theoretical framework explaining issue importance among the electorate, and establishing a link between the distribution of issue importance among the electorate on the one, and the parties’ anticipation of this distribution on the other hand. This theoretical framework will be the base for modeling heterogeneity in decision weights in individuals’ vote calculus, such that it can be reasonably integrated into optimal party strategies. I argue that behavioral factors are the key to understand heterogeneity in individual interests, and thus in voting behavior. Since behavioral factors also have a large influence on parties and the evolution of Westerns party systems, implementing them in the formal game of party competition is also a reasonable approach from the party side. The cleavage structure of a society strongly influences the landscape of political parties in many countries, as well as the policy demand and thus the voting behavior within the electorate. Therefore, I identify cleavage theory to be the basis for a theory that combines the demand and supply side in my model of party competition.

Empirical Application: The Religious Cleavage

The first part of the book developed a theoretical argument explaining why different groups of the electorate should place more weight on a certain issue dimension in their vote choice, depending on the empirical context. In this chapter, the aforementioned theoretical framework will be applied to an empirical case of German general elections, namely the issue of an abortion law that was heavily discussed in Germany between 1980 and 1990. This issue addresses the cleavage line on which religious conservatives oppose progressive members of the new social movements. The empirical results of the analysis lends empirical support to the general argument of the theory, showing that cleavage group members tend to put more weight on an issue dimension addressing the content of the respective cleavage line, if they hold bargaining power and policy change is to be expected. Although considering this variation appropriately in the vote models may not increase the explanatory power of the model dramatically, it still helps to improve our understanding of the other side of the coin, by yielding predictions about party competition that closer map empirical patterns.

The Construction of Policy Spaces

The construction of policy spaces is a fundamental part of the analysis of voting behavior and party competition. This chapter is dedicated to the description of the relevance of measurement and operationalization of policy issues. It presents a well-grounded way how to construct policy spaces based on those measures. A survey of the current literature regarding empirical studies of spatial voting models reveals a large heterogeneity in the measures and concepts used. The consequences of different measures for the substantial interpretation of the results of models of spatial voting are often discussed in footnotes or completely ignored. Often, the measures in use rely on very different assumptions regarding the nature of the policy space, which makes a comparison of substantial results difficult. This chapter discusses the characteristics of the policy space that are relevant when it comes to analyzing party competition based on a spatial model of vote choice and critically reviews established operationalization methods. It then describes a measurement technique that allows for the construction of empirical policy spaces in which voter ideal points as well as party positions can be transferred, based on which a spatial model of voting behavior and party competition can be estimated.

Empirical Application: The Class Cleavage

In this chapter, the theoretical framework of the book is applied to the class cleavage line to show that the theory is generalizable not only to various points in time, but also to various cleavage lines, even simultaneously. Two exemplary elections will be analyzed in which labor market issues played a significant role, namely the 1994 election, in which mass unemployment was a major issue, and the 1987 election, where the issue of a recent reform of §116 AFG (Employment Promotion Act), regulating the right to strike, featured prominently on the campaign agenda. The results show that in both cases it is the group with more homogeneous and extreme interests that places more emphasis on the respective labor market issue, thus differing to a larger degree in its issue weighting pattern. In both empirical cases the equilibrium configurations based on the extended vote model predict more divergent policy positions and thus perform slightly better in mapping the empirical configuration of perceived party positions. This indicates that the consideration of the electorate’s cleavage structure contributes to our understanding of optimal party behavior when competing for votes in the multidimensional policy space.

Hätten Direktkandidaten der AfD über die 5 %-Hürde verholfen? Eine Untersuchung des Kontaminationseffekts im Mischwahlsystem

Die AfD kandidierte 2013 erstmalig bei einer Bundestagswahl und scheiterte knapp an der 5%-Hurde. Ein Unterschied zur Kandidatur etablierter Parteien ist, dass die AfD nicht in allen Wahlkreisen Direktkandidaten nominierte. Der Beitrag untersucht die Kandidatur der AfD basierend auf der Hypothese der Kontaminationsliteratur, welche einen positiven Kandidateneffekt auf den Parteistimmenanteil in gemischten Wahlsystemen postuliert. Konkret wird untersucht, ob bei der AfD-Kandidatur ein solcher Effekt nachweisbar ist. Damit soll die Frage beantwortet werden, ob die AfD bei flachendeckender Kandidatennominierung den Einzug in den Bundestag geschafft hatte. Nach Berucksichtigung moglicher Endogenitat und eines heterogenen Kausaleffekts kann jedoch kein Kandidateneffekt nachgewiesen werden.