Doctor of Philosophy (PhD)


Political Science

Document Type



This dissertation is a story about the divisions that characterize the mass public. Specifically, it explores how Americans think about politics, and, in particular, how citizens connect their attitudes, beliefs, and, vitally, ideological identity to their partisan affiliation—a phenomenon known as sorting. Practically, this project proceeds in two parts. In Part 1, I investigate the nature of partisan sorting in the mass public. Chapter 2 reviews the extant scholarly literature regarding partisanship and ideology, or the raw materials of sorting. Drawing on this research, I operationalize two types of sorting in Chapter 3 and compare how different measurement protocols affect the characterization of public opinion. This distinction culminates in Chapter 4, which provides a series of empirical tests that justify partitioning sorting into identity- and policy-based constructs. The second part of this dissertation is devoted to the study of identity-based sorting. Chapter 5 takes up the question of why individuals’ identities converge and conveys that sorting is related to asymmetric perceptions of out-group dissimilarity rather than relative perceptions of between-group differences. Chapter 6 explores how this sorting affects compromise. I discover that, even in the absence of consistent policy preferences, identity sorting is sufficient to decrease an individual’s willingness to accept bipartisanship. Finally, Chapter 7 examines how identity sorting alters the decisional criteria that voters utilize to select political candidates. Here, I show that sorting produces a disconnect between the perceived and objective ideological congruence between voters and their preferred candidate. Sorting, then, is a sufficient condition for pushing citizens toward more extreme candidates—even when individuals’ issue preferences suggest that their “best” candidate is considerably more moderate. Taken as a whole, this dissertation both refines the extant logic of sorting and pushes this research into new territory. In demonstrating that identity-based sorting constitutes a unique and particularly powerful political phenomenon, I reveal why concern over the systematic coherency of mass opinion is, perhaps, misplaced. Instead, it is this identity sorting that contributes to the intemperate and polarized atmosphere that characterizes the state of American politics.



Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Garand, James C.