Spatial Models of Party Competition Applied
The aim of this project is the construction of a common policy space describing the policy and ideological preferences of both voters and parties shortly before or after German Bundestag elections. From these spaces consequences will be derived concerning the development of the German party system.
Input data are voters’ perceptions of policy and ideological positions of parties and the respective policy and ideological preferences of voters. The resulting policy and/or ideological space is relevant for voters allowing the at least partial prediction of their vote intentions by the distances of parties from their ideal points. This same space is also relevant for parties allowing them to communicate with voters provided that the voters agree about a verisimilar picture of the party configuration. To solve this problem, a scaling algorithm developed by Aldrich and McKelvey will be applied; its deficiencies in calculating the variance of the perceptions around the expected values shall be overcome by using Bayesian methods (Bräuninger and Giger 2011).
In this project German election studies from 1980 to 2009 will be secondary analyzed. First analyzes show (Pappi, in print) that the perceived left-right positions of parties are relatively stable over time; parties, however, have room for strategic maneuvering concerning new issues though these also offer chances for new entrants into the party system capitalizing on their reputation as "issue owners".
On the supply side, policy offers and valence or competence reputations of parties are the essential ingredients of party competition. For the single Bundestag elections the expected issue positions of the competing parties will be predicted as local Nash equilibria applying the valence model of Schofield and they will be compared with the empirically ascertained positions. In a further step, the equilibrium model will be used to investigate the dynamics of party competition over time. The aim is here to interpret the repositioning of parties over time as dynamic equilibria, so that the development of the German party system can be explained by the electorate’s demand for and the parties’ supply of policies and competence reputations.
We tested our baseline model which combines policy distances to parties with party valences for the pre-election cross-section survey 2009 of the German Longitudinal Election Study. Distances of voters to parties were both measured concerning position issues and the ideological left-right scale. Only when fixing the ideological distances a plausible party equilibrium within the issue space could be identified.
In a research proposal for external funding by the DFG we focused the research question on the specific effects of the German mixed member proportional electoral system on party competition. Our central hypothesis is that party competition concerning first votes triggers centripetal and concerning second votes triggers centrifugal forces. The DFG will finance the project for 2014 and 2015.