Cody Arlie ReedResearch + data
WorkDo racial patterns correspond with how people move?

Mobility networks · racial segregation

Do racial patterns correspond with how people move?

ProblemResidential segregation tells us where groups live, but not whether the same racial-demographic divisions also organize everyday travel.

What I builtI combined 2020 Census data with 95.4 million mobile-device location observations across 100 U.S. regions, estimated travel communities 100 times per region, and tested whether their boundaries aligned with sharp local changes in racial composition.

OwnershipSole researcher and developer: study design, network construction, modeling, visualization, and writing

Result

Composite alignment was statistically positive in 54 of 100 regions; Black-gradient alignment was positive in 51.

Stacked bars showing the number of 100 U.S. commuting zones with statistically positive, inconclusive, or statistically negative associations between movement boundaries and racial-gradient intensity
Statistically significant alignment across 100 U.S. regions

What I built

Build each travel network, then estimate its communities 100 times.

Boundary propensity

The share of 100 model runs in which a Census block group lies on a travel-community boundary.

New York City block groups shaded by how often they fall on a movement-community boundary across repeated estimates
How often each place appeared on a travel boundary · New York City

Darker areas appeared on a travel boundary more often across repeated model runs.

What I compared

Test whether travel boundaries fall where racial composition changes sharply.

A racial gradient measures how quickly a population group’s local density changes across nearby places. I modeled whether Census block groups that repeatedly appeared on travel boundaries also had sharper racial gradients, while accounting for population density and local racial composition.

Four New York City maps showing local White, Black, Hispanic, and Asian racial-density gradients
Group-specific racial-gradient intensity · New York City

Brighter values indicate a sharper local change in the density of the named population group.

Cross-city result

Composite and Black racial gradients corresponded with movement boundaries in roughly half the regions.

The relationship was statistically positive in 54 of 100 commuting zones for the composite racial gradient and 51 for the Black racial gradient. Positive relationships were less common for Asian, Hispanic, and White gradients. Composite and Black correspondence was stronger in more residentially segregated regions.

How cities differ

Residential segregation and local boundary intensity are not the same city characteristic.

Chicago combines high residential segregation with high average racial-gradient intensity. Detroit is also highly segregated but has more modest average gradient intensity. New York City has unusually intense composite, Black, and White gradients, partly reflecting its population density.

ChicagoHigh segregation

High average racial-gradient intensity

DetroitHigh segregation

More modest average racial-gradient intensity

New York CityUnusually intense gradients

Composite, Black, and White measures stand out