Policy responsiveness to all citizens or only to voters? A longitudinal analysis of policy responsiveness in OECD countries

Abstract

A close connection between public opinion and policy is considered a vital element of democracy. In representative systems, elections are assumed to play a role in realising such congruence. If those who participate in elections are not representative of the public at large, it follows that the reliance on elections as a mechanism of representation entails a risk of unequal representation. In this paper, we evaluate whether voters are better represented by means of an analysis of policy responsiveness to voters and citizens in democracies worldwide. We construct a uniquely comprehensive dataset that includes measures of citizens’ and voters’ ideological (left–right) positions, and data on welfare spending in Organisation for Economic Cooperation and Development countries since 1980. We find evidence of policy responsiveness to voters, but not to the public at large. Since additional tests suggest that the mechanism of electoral turnout does not cause this voter-policy responsiveness, we outline alternate mechanisms to test in future research.