The geography of partisan homophily in the 2020 US presidential election

Mastrosavvas, Andreas ORCID: https://orcid.org/0000-0002-1354-5523 2024. The geography of partisan homophily in the 2020 US presidential election. Applied Geography 171 , 103371. 10.1016/j.apgeog.2024.103371

PDF - Published Version Available under License Creative Commons Attribution. Download (57MB)

Abstract

Partisan segregation in the United States is often interpreted as evidence of limited social interaction among out-partisans, or partisan homophily. In this paper, I draw on 2020 US presidential election results and data on the pairwise density of social ties between the populations of 22,537 zip code tabulation areas (ZCTA) to examine how different areas are socially connected to politically similar others. Using the local Moran index, I first identify clusters of ZCTAs where there is evidence of partisan homophily or heterophily. In a series of multinomial logistic regressions, I then also examine differences in the probability of each cluster across different settlement types and regions, and across areas with differences in the relative connectedness and geographic distance to others. I find that partisan homophily is the norm across areas, broadly tracking partisan segregation along the urban-rural continuum. However, the populations of Democratic-leaning areas, which are most likely to be in cities and suburbs, are on average likely to have more of their co-partisan social ties in relatively distant areas when compared to the populations of Republican-leaning areas. This highlights the prospect of partisan differences in the role of non-local context in local political outcomes.

Item Type:	Article
Date Type:	Publication
Status:	Published
Schools:	Social Sciences (Includes Criminology and Education)
Publisher:	Elsevier
ISSN:	0143-6228
Date of First Compliant Deposit:	25 September 2024
Date of Acceptance:	5 August 2024
Last Modified:	10 Oct 2024 11:27
URI:	https://orca.cardiff.ac.uk/id/eprint/172378

Actions (repository staff only)

Edit Item