Measuring Aggregate‐Level Ideological Heterogeneity

Abstract

Ideological heterogeneity is a key variable for the study of legislative and electoral politics. Scholars have long recognized that members with more ideologically heterogeneous constituencies behave differently than members with more homogeneous ones. Empirical tests of these theories, however, have typically been stymied by a lack of appropriate measures. We corrected this shortcoming by developing a measurement model for ideological heterogeneity, and we used our method to generate estimates for the 50 U.S. states and 435 congressional districts. Beyond the specific results presented here, a key contribution of our model is its flexibility: our technique can be used to produce similar estimates in a variety of contexts.