2026 Theses Doctoral

Space in Time: Three Essays on the Daytime Experience of Space by Race, Class, and Partisan Affiliation

Chae, Jongsun Joanna

How do neighborhoods manifest and perpetuate racial and socioeconomic inequalities in urban America? An expansive literature in the social sciences explores a wide range of ways neighborhoods matter. Prior work has primarily studied space in narrow terms by place of residence, schools, or workplaces, although people visit a larger scope of spaces for a broad range of purposes, such as health visits, leisure, and shopping. In addition, neighborhoods are likely to change due to the travel patterns of residents who leave and the visitors who come into these spaces. For these reasons, this dissertation revisits the above question by first demonstrating how neighborhoods are dynamic spaces that change throughout the 24 hours of a day, 7 days a week, and how these changes relate to mechanisms which are known to perpetuate inequalities, specifically voter turnout and cultural consumption.

All three papers in this dissertation use mobile phone location data from October 2018, which has over 15 million unique mobile phone owners for the 49 most populous metropolitan areas in the US. I supplement this data with other large-scale data, including voter file data, business data, and the 2018 American Community Survey. The three essays of this dissertation use several common computational techniques, including clustering algorithms for the location data, imputation of demographic characteristics for mobile phone owners, and pooling of imputed model results to account for uncertainty in the imputed demographic characteristics.

This dissertation finds that residential demographics are insufficient to capture people’s daytime experiences of space. Specifically, in the first chapter, I show that Black-White segregation ebbs and flows and reaches its nadir around 12PM, and is higher on average over the weekend. In the second chapter, I find people are exposed to out-group members by race and party at higher rates than the demographics of their neighbors would suggest. Hence, residential segregation underestimates the exposure people have to people not only by race, but also by race and party.

I explore an implication of the discrepancy between daytime and residential exposure to diversity in the second chapter of this dissertation, in which I find exposure to diversity has different within-race, between-party associations with voting behavior in the 2018 midterm elections. Most notably, White and Black Republicans are more likely to vote with greater daytime exposure to Black Democrats. Black Democrats are more likely to vote with greater residential and daytime exposure to Hispanic Republicans.

In the last chapter, I find that space matters in new ways, such that I find the number of cultural venues in one’s residential neighborhood is a stronger predictor of cultural consumption than a college degree. Moreover, white college graduates not only consume culture most frequently, but also live in neighborhoods with the largest numbers of cultural venues, thereby reinforcing their dominant position to consume culture. Yet, I also find that racial minorities with less than high school degrees consume culture more frequently and live in neighborhoods with greater numbers of cultural venues than their white counterparts. The heterogeneous trends that I find encourages future work in cultural consumption to study the intersection of race and class.