On the same space? Measuring the ideology of voters and political elites

Ideology is a key concept in political science. In studies related to political representation, policy responsiveness, and the quality of decision-making in elections, ideology is a central concept that needs precise measurement to test related hypotheses. Despite its prominence in the political science literature, ideology is often measured differently for elites and voters with indicators that do not correlate among themselves. As a result, it is not possible to assess whether the measures of ideology and the latent dimensionality behind measures of ideology in the population and the political elites align or to what extent they differ. In this paper, I build on existing research and propose a way to measure the ideology of voters in population survey on the same space as the political elites by the same policy questions answered by political elites to voters. Furthermore, I show that researchers only need a rather limited number of questions to reach an accurate measure of ideology. Using Monte-Carlo simulations as well as real-world election data from a voting advice application (VAA) and a post-election survey in a Bayesian Item-Response Theory framework, I show that, using only a limited number of policy questions, it is possible to position voters and political elites in a similar ideological space. In sum, this paper shows that taking a handful of policy questions from elites surveys enables the operationalization of a common ideological space for political elites and citizens. This has large implications for the study of the relationship between citizens, voters, parties, and politicians.