Comparing the Effect of Revealed and Expressed Populism on Populist Voting

A sizable literature has identified individuals’ populist attitudes as key drivers of voting for both left-wing and right-wing populist parties and politicians. Against this backdrop, some recent experimental work has suggested that the effect of such “thin” populist attitudes (i.e. support for referendums, distrust of elites, etc.) is overstated and that more attention should be paid to the ways in which people are attracted to populist parties due to their “host” ideological programs, whereby people vote for populists because of their stances on topics such as immigration or globalization. One drawback of these experiments was that they were concerned with voting for fictitious candidates rather than peoples’ vote intentions for real populist parties, thus creating a gap between research on the antecedents of voting for populist parties with the experimental studies examining voters’ receptivity to fictitious politicians’ populist appeals. This paper bridges this gap by building on a new methodological approach that allows the estimation of individual-level estimates from conjoint experiments. We are thus able to compare respondents’ revealed populist behavior (i.e. how responsive they are to populist appeals in conjoint experiments) to their expressed populist attitudes as predictors of populist vote choice. Using data from Germany, we leverage multiple conjoint experiments paired with respondents’ actual vote intentions to examine how expressed populism and revealed populism affect populist voting. Our results indicate a mismatch between expressed and revealed populism, such that voting for the populist Alternative für Deutschland (AfD) and Die Linke is not driven by voters’ revealed populist preferences but rather by expressed populism and issue positions. Importantly, the gap between revealed issue positions and expressed issue positions is much smaller than the gap between revealed and expressed populism. To explain this divergence we conduct a vignette experiment that tests whether voters project populist beliefs onto politicians and parties that espouse issue positions that are often associated with populism. We thus explore whether the link between populist attitudes and populist voting is a result of voters attributing popular populist positions to parties they like based on their stances on issues, thus inflating the link between respondents’ populist attitudes and populist voting.