Measuring Strategic Voting in Denmark: A Direct Method

• In Event: Elections and Voting Behavior Virtual Poster Session
  In Virtual Poster Session: Elections and Voting Behavior: Electoral Campaigns, Systems, and Strategy Virtual Poster Session

Abstract

Political science scholarship on strategic voting has expanded in recent decades, with evidence for strategic voting in certain segments of the population reported cross-nationally in coalition systems. However, current measurements of strategic voting are not without limitation. Measuring levels of strategic voting at the aggregate level suffers from the classic ecological fallacy problem, and reveals nothing about individuals voters’ motivation. When surveyed individually, voters may under- or over-report strategic voting behavior, especially the further away in time from Election Day they are interviewed, due to a strong post-election bias of reporting a vote for the winning party. This paper uses data from Denmark to measure the level of strategic voting via the direct method, which has emerged as the preferred methodological approach in recent literature, when using individual-level data. As such, this project represents a straight-forward and accurate test of the extent to which voters in the Danish 2015 parliamentary election engaged in policy balancing and/or threshold insurance voting. Additionally, it provides a clean measure of strategic voting at the individual level, overcoming methodological challenges noted by previous work.

Authors

• Andra Pascu-Lindner, Rice University

• Lie Philip Santoso, Duke Kunshan University