Abstract
Traditional understanding of urban income segregation is largely based on static coarse-grained residential patterns. However, these do not capture the income segregation experience implied by the rich social interactions that happen in places that may relate to individual choices, opportunities, and mobility behavior. Using a large-scale high-resolution mobility data set of 4.5 million mobile phone users and 1.1 million places in 11 large American cities, we show that income segregation experienced in places and by individuals can differ greatly even within close spatial proximity. To further understand these fine-grained income segregation patterns, we introduce a Schelling extension of a well-known mobility model, and show that experienced income segregation is associated with an individual's tendency to explore new places (place exploration) as well as places with visitors from different income groups (social exploration). Interestingly, while the latter is more strongly associated with demographic characteristics, the former is more strongly associated with mobility behavioral variables. Our results suggest that mobility behavior plays an important role in experienced income segregation of individuals. To measure this form of income segregation, urban researchers should take into account mobility behavior and not only residential patterns.
Introduction
Traditional urban sociology measures income segregation based on where people live (residential segregation). However, most city-dwellers spend the majority of their waking hours away from home—at work, in shops, or at parks. Residential data fails to capture the "experienced" segregation that occurs during these daily interactions.
The challenge has been the lack of high-resolution data to track these movements at scale. Without understanding how mobility patterns contribute to social exposure, urban planners and researchers cannot fully grasp how economic silos are formed or how to break them down. This paper seeks to bridge that gap by quantifying segregation at the level of individual "places" (venues) and individual "experiences."
Methods and Data
The researchers developed a framework to measure segregation as a dynamic behavioral process rather than a static geographic one. They utilized anonymized GPS "pings" from 4.5 million mobile phone users over six months across 11 major US metropolitan areas, combined with Foursquare API data on ~1.1 million verified venues and US Census Bureau data for income quartiles.
Measuring Segregation
- Place-Level Metrics: They defined a segregation metric ($S_\alpha$) for over a million specific venues. A venue is "integrated" if the time spent there is evenly split among four income quartiles and "segregated" if dominated by a single group.
- Individual Experience Metrics: They calculated an "experienced segregation" score ($S_i$) for individuals based on the diversity of the people they likely encounter at the venues they visit.
The Social-EPR Model
The authors extended the well-known Exploration and Preferential Return (EPR) mobility model. The standard EPR model explains how people choose between visiting new places (exploration) or returning to familiar ones. The authors added a Schelling-style extension called "social exploration" ($\sigma_s$), which represents an individual's probability of exploring a new place where their own income group is the minority.
Key Insights: The Granularity of Experience
A primary finding of the study is the hyper-granularity of urban segregation. Segregation is not merely a "neighborhood" phenomenon; highly integrated and highly segregated places often exist just dozens of meters apart. For example, a museum might be perfectly integrated, while a grocery store across the street is highly segregated.
Furthermore, the study identified a "Catchment" Effect: venues that people travel further to reach (like airports, museums, or stadiums) act as "hubs" for integration. Conversely, places that serve local communities (places of worship, grocery stores) tend to be more segregated.
Key Results and Behavioral Drivers
- Low Residential Correlation: There is only a weak correlation ($\rho \approx -0.17$) between the income of a person's home neighborhood and their experienced segregation. Living in a wealthy area doesn't necessarily mean you only encounter wealthy people.
- Encounters are Distant: On average, 78% of a person's potential encounters in the city are with people who live in a different census region (PUMA), and only 3% are with people from their own immediate block group.
- Lifestyles vs. Demographics: The study found a split in what drives behavior. Place exploration (visiting new physical locations) is mostly predicted by "lifestyle" variables—what you like to do. Social exploration (visiting places where you are the minority), however, is strongly tied to residential demographics like education level and race.
- Model Accuracy: The proposed Social-EPR model, using only two parameters ($\sigma_p$ for place exploration and ($\sigma_s$ for social exploration), was able to accurately replicate the variability in experienced segregation for over a million individuals ($\rho = 0.777$).
Conclusion and Future Directions
The results suggest that mobility behavior plays an important role in experienced income segregation. To be truly "economically integrated," an individual must be both a physical explorer and a social explorer.
However, the study has limitations. It measures "exposure" (being in the same place) rather than actual social interaction. Additionally, the data relies on smartphone users, which may underrepresent certain demographics. Future work could apply this "experienced" framework to racial or ethnic segregation and investigate if urban interventions, such as changing transit routes, can causally shift exploration behavior.
Summary updated