How Do Political Elites Respond to What the Public Wants? Results from a Survey-Experiment with Belgian MPs

Studying the connection between what the public wants and what political elites do, is an old but still major concern of political scientists (Bingham-Powell 2004). Departing from the idea that in a healthy democracy, political decision-making is congruent with the public’s wishes and demands, a great deal of studies have examined to what extent reality meets this democratic ideal (for an overview, see Soroka and Wlezien 2010). These studies generally demonstrate that public opinion—whether of all citizens, or of those from a specific constituency or party—is congruent with public policy. However, the congruence between public opinion and policies may result from the interplay of at least three causal pathways which are hard to disentangle (Butler and Nickerson 2011). First, it could be a consequence of citizens’ successful election of representatives who share their preferences and beliefs and act accordingly (Miller 1999). Second, citizens sometimes adapt their policy positions to the stances of their preferred party or candidate (Hill and Hurley 1999; Jacobs et al. 1998). Finally, policy congruence can not only be driven by ‘correct voting’ or ‘opinion adaptation’ on the side of voters, but also by responsiveness on the side of the elites. Politicians may initially not share voters’ preferences, but may be motivated to inform themselves about public opinion and design policies according to (their perception of) this public opinion (Stimson, Mackuen, and Erikson 1995). Whereas all three causal pathways are legitimate routes to representation, only the latter route implies that politicians are ‘delegates’ (rather than ‘trustees’) in the true sense of the word. This paper focuses on the latter linkage between elites and their voters. It tests whether and how political elites are responsive to information about the preferences of the public (and more precisely: their own party electorate) on concrete policies. To that end, a survey experiment with political elites is conducted. Hence the paper responds to recent calls for experimental research on the topic of representation, which—complementing the large bulk of observational studies—is well-suited to make strong causal claims, and which can help unravel the mechanisms producing elite responsiveness (Grose 2014). Concretely, a survey-experiment with approximately 160 Belgian MPs is scheduled in March-April 2018. In the survey, political elites will be asked about own preferences, and their intended representational behavior, on a number of concrete policy statements. Preceding these survey questions—and this is the experimental manipulation—we present a randomly drawn half of the legislators with information about the opinion of the majority of their party electorate regarding these policy statements (obtained via a representative survey with Belgian citizens; N=2000). We specifically select statements where the electorate’s opinion opposes the viewpoint of the politician’s party. The assumption is that the preferences and intended behavior of the politicians who received information on the opinions of their party electorate, will be more in line with this party electorate, than those of the politicians who did not receive the information. Additionally, we explore whether characteristics of issues and individual MPs moderate elites’ responsiveness to public opinion.