01.02.2020 - 31.01.2022
Recent failures to predict landmark elections including the 2016 U.S. presidential race and the Brexit vote have, rightly or wrongly, rattled public confidence in election polls and the survey method at large. Expert committees are regularly convened in the aftermath of such incidents to investigate potential error sources. Due to the inherent limitations of case studies, however, the crucial question why such errors occur in a given election but not in others often remains a matter of speculation. The aim of this project is to develop a contextual understanding of polling errors and their triggers. Unlike most previous studies, we take a cross-election comparative perspective and put the theoretical focus on characteristics of the electoral contest which may encourage polling errors. We build on previous efforts to create and maintain the hitherto largest global data base of poll and election results. We tap media archives and other data sources to collect difficult-to-measure contextual information. We draw on recent developments in the statistical modelling of survey errors which allow us to separate systematic and random error components at the election level. We extrapolate from our results to predict polling errors (and, by implication, election outcomes) at the 2020 U.S. presidential race and the 2021 Bundestag election. Our findings will be communicated through suitable channels to expert and non-expert audiences.