Measuring District-Level Economic and Moral Ideology

Due to insufficient sample sizes in national surveys, strikingly little is known about issue preferences at the level of Congressional districts in the United States, and existing studies focus only on a single issue dimension. This paper takes a first step toward filling this gap by developing a hierarchical Bayesian model to combine survey and census data to estimate the median of each congressional district‘s economic and social preferences. Our estimates generate an intuitive new two-dimensional map of American political geography with a variety of substantive applications in American Politics. Moreover, our methodological approach could help scholars in both American and Comparative politics combine surveys and census data to develop estimates for the preferences of sub-national geographic units to test formal models of political economy.