Red sport, blue sport: political ideology and the popularity of sports in the United States

ABSTRACT

This exploratory study examines correlations between political ideology and the popularity of 53 sports in the United States using geographical units of analysis. Regional interest in sports was measured using Google Trends search data from Nielsen’s 207 designated media market areas. Voting data from the 2020 Presidential Election (percent Trump vote) was used to characterise each area’s political ideology. A measure of the sports’ cultural dominance was developed through a large-scale, computer-aided content analysis of sports news. Areas with high levels of Trump support tended to favour dominant sports like football, basketball and baseball and showed less interest in fringe sports like ultimate Frisbee, rock climbing and badminton. These correlations generally held for large, mid-sized and small areas. There were, however, sports that contradicted this relationship. Focusing on outliers and inconsistent cases, the study concludes by suggesting next steps for establishing a robust link between political orientation and sport popularity.

KEYWORDS:

• Political ideology
• culture
• sport interest
• hegemony
• non-normative sports
• geography

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. Google Trends searches can be conducted using terms or topics. Searching for specific terms produces volume estimates based only on matches for the given term. In this study, we used topic searches, which provide volume estimates of terms that share the same concept or entity.

2. Echelon Insights is a research, polling and analytics firm. It provides a breakdown of the 2020 presidential election results by Nielsen’s designated media market areas. The data are free, open to the public and can be downloaded here: https://echeloninsights.com/in-the-news/presidential-results-by-media-market.

3. In 1972, the correlation between party affiliation and self-reported political ideology was .32; by 2016, the correlation between the two variables had increased to .70 (Barber and Pope 2019).