Powerful Polls? The Influence of Public Opinion Polls on Elections in Representative Democracies

• Applicant: Dr. Werner Krause
• Subject Area: Political Science
• Term: since 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 507290550

Public opinion polls have become increasingly pervasive elements in representative democracies. Especially during election campaigns, projections of citizens’ vote intentions attract far-reaching media attention, and many voters consider them important. At the same time, the role of opinion polls is controversial as they are suspected of improperly influencing voters. Consequently, discussions have often turned to how polling results undermine the working of representative democracy by negatively affecting democratic accountability and responsiveness. Relatedly, debates have asked frequently whether polls should be banned, especially before elections.The question of when and how public opinion polls impact election outcomes constitutes a major research gap. Against that backdrop, the proposed project aims to study the influence of opinion polls on people’s voting behaviour. In particular, it will focus on the context conditions and individual-level preconditions that give rise to poll-induced voting behaviour. The project will investigate the role of public opinion polls for election outcomes in two steps. First, it will collect polling estimates for at least 20 OECD democracies for more than 50 years (since 1970). Based on the resulting data set, the project will analyse patterns in polling trends in a comparative perspective. Particular attention will be paid to the context factors that make polling effects more or less likely. More concretely, the analyses will focus on country, poll, and political party characteristics to examine under which circumstances polls can become “self-fulfilling prophecies”. Second, the project will conduct two survey experiments to investigate the causal effect of polls on citizens’ vote intentions. Both of these experiments will be fielded in four West European countries. The first experiment will examine which political parties are most likely to benefit or suffer due to the influence of polling results. The second experiment focuses on the question of which voter groups are most likely to be responsive to polling results. The project will thus combine evidence from cross-national, observational data with survey experiments. In this way, the project will maximise the internal and external validity of the research findings, allowing for robust causal inferences. The proposed project will contribute to a detailed understanding concerning the role of opinion polls for the working of representative democracies. The project’s research findings will be important for the existing literature on voters, political parties, and democratic elections. Given the high levels of media attention that polls attract especially in prior to elections, the empirical results will also be of interest to broader public and party strategists alike.

• DFG Programme: Research Grants