Does Congress Represent Public Opinion as it is or as it Might Be?

Abstract

Recent research has demonstrated that Congress is far representing public opinion as it expressed in surveys, but has neglected alternative benchmarks for evaluating representation. In this paper, I show that senators may act as roughly representative draws from the distribution of issue opinion that would exist if citizens were more uniformly informed about political issues. Using data from the 2006 Cooperative Congressional Election Survey, I show that the subset of Americans who demonstrate knowledge of the political alignment of seven political issues hold positions on those issues that are comparably polarized to the Senate and far more polarized than the public as a whole. The association of this particular form of political knowledge with holding partisan profiles of issue positions is consistent with a large and polarizing causal effect of political knowledge on issue positions via cue-taking. To the extent that the disparity in measured polarization between the public and Congress is due to many members of the public being unfamiliar with the partisan alignment of many issues, the quality of Congressional representation is higher than previous studies have suggested.