Polarization in the US Congress 1952-2016: A Theoretical Explanation

• In Event: Polarization in Legislative Politics

Abstract

We explain the individual-level extremity of Members of Congress, and hence polarization, using a model that relies on district ideology, member sorting, and institutional changes in Congress. Of the three mass-elite linkage models we consider -- proximity, directional, and mixed -- only the directional model is consistent with all three explanations. Because the directional model implies, contrary to the existing literature, that centrist representatives will not be advantaged when they seek reelection, we reconsider that literature. Based on an entirely theoretical argument we suggest that the discipline’s current understanding fails on two fronts: first, it does not adequately distinguish between sympathetic and antagonistic districts; and second, it incorrectly assumes that winning or losing elections responds to the same set of factors as margin of victory. We test these claims and find them supported. Our work suggests that polarization is a natural rather than an aberrant condition in two-party democracies with centrist electorates.

Authors

• Stuart Elaine Macdonald, University of North Carolina, Chapel Hill

• Nicholas L. Pyeatt, Pennsylvania State University, Altoona