Studying Dynamic Responsiveness: How Vote Advice Applications Enable Cross-Temporal and Cross-Institutional Research on Party Positioning

Proposal for panel 4 (Dumont/Linhart): Voting advice application data and coalition studies Party responsiveness to shifts in public opinion is a long-standing concern in the study of political representation. Yet, despite continuing interest in the dynamics of party position-taking, empirical research has not lived up to theorizing on this important question. Progress has been hampered by the challenge of establishing ideological spaces that (a) ensure comparable party positions across time and context and (b) do not rely on predefined conflict dimensions. One common approach for extracting actor positions without prior definition of conflict spaces builds on information from roll call voting. However, as the set of legislative proposals is idiosyncratic for each legislative session, positional shifts might either be attributed to shifting actor preferences or to shifts in substance. We argue that vote advice applications (VAAs) provide an excellent source of data to overcome this problem and to ensure that positional shifts are indicative of changing party preferences. As VAAs frequently include identical issues in multiple electoral contexts, these overlaps allow researchers to build bridges between conflict spaces across institutions and time. We assess this proposition by estimating conflict spaces using two possible solutions. First, we link spaces at the level of individual items. As common items might not always be available, we go on to investigate whether bridging links can be established at the level of policy fields. For the empirical analysis, we study national and subnational elections in the German Wahl-O-Mat over a period of ten years. Both solutions yield promising results. In a second step, we relate the estimated spaces to public preferences to probe several propositions from the responsiveness literature in a preliminary analysis. The study closes with several recommendations for employing VAAs in party research.