Offline Ties Outshine Online in Predicting Votes

Offline social networks, revealed by co-location data, predict US voting patterns more accurately than online social connections or residential sorting. Michele Tizzoni and colleagues analyzed large-scale data on co-location patterns from Meta's Data for Good program, which collates anonymized data collected from people who enabled location services on the Facebook smartphone app. Colocation is defined as two people being within the same map tile, which is less than 600×600 meters, depending on latitude. The political affiliation of each person was inferred from their county of residence. This data was compared with Facebook friendships and residential proximity for all US counties, along with individual survey responses from 2,420 Americans regarding their offline and online social networks during the 2020 presidential election. For the residential proximity measurement, the voter registrations of the closest 1,000 neighbors were used. Overall, the authors found that partisan exposure through physical co-location—measured by people being in the same map tile for at least five minutes—explained 97% of the variance in county-level voting patterns, compared to 85–87% for online connections and 75–80% for residential proximity. Individual-level analysis using data from the 2020–2022 Social Media Study, conducted by the American National Election Studies, showed offline social ties had stronger effects on vote choice than online connections. Partisan segregation was higher in physical spaces than online, with educational attainment structuring this segregation in large part. According to the authors, these findings highlight the fundamental importance of real-world social interactions over digital connections in understanding political behavior and suggest that although the internet has been blamed for American political partisanship, it is also very much an IRL phenomenon.