Social Science Job Candidate

Are the rich better represented than the poor, and if so, how does this under¬representation affect policy outcomes in the United States Congress? In this paper, I combine data from the National Annenberg Election Studies (2004 & 2008), the Cooperative Congressional Election Studies (2006, 2008, & 2010), and a unique survey combining the policy questions from both to scale voters the way Congressional scholars scale members of Congress. The data cover 246,000 respondents in 435 Congressional Districts and 50 states. I leverage the power and coverage of these data to show that the expressed preferences of the rich are better represented in both chambers than the expressed preferences of the poor. "Expressed preference representation," however, is only one form of representation. I introduce the concept of "income group representation," in which legislators represent their constituents with regard to income groupings without taking constituent's expressed preferences into account. I find that legislators in districts with larger proportions of poor constituents tend to be more liberal, controlling for poor constituents' expressed preferences. This pattern offsets the under-representation of the expressed preferences of the poor. I show that legislators' positions are not substantially different from what they would be if the expressed preferences of the poor were equally represented in the House. In the Senate, the policy effects of unequal expressed preference representation are larger, but they are offset by income representation to an even greater degree than in the House. The rich are better represented than the poor, but only by the most obvious form of representation; and the policy implications of this differential representation are minimal.