Mapping Reverse Commuting in the United States: A Network-Based Analysis of Urban Structure, Labor Market Access, and Spatial Mismatch.

Eli Kochersperger

Co-Presenters: Bernice Torres Santamaria, Catherine Anukam, Zilu Zhang, Haoxiang Liu

College: College of Business and Public Management

Department: Marketing, Global Bus. & Econ

Abstract:

Reverse commuting—travel from central neighborhoods to outlying job centers—is often cited as evidence of job decentralization, spatial inequality, or shifting residential preferences. Yet most empirical definitions rely on distance-from-downtown rules or simple inflow/outflow ratios that assume cities revolve around a single central business district. In increasingly polycentric metropolitan areas, such measures can misclassify commuting direction and shape misleading conclusions about urban change.This project introduces a network-based definition of reverse commuting that better reflects how modern cities function. Using tract-level Origin–Destination data from the Census Bureau’s LODES program, we model commuting flows as a connected network and compute a measure of “commuting centrality” based on the long-run distribution of movement within that system. Reverse commuting is then defined relative to this endogenous centrality structure rather than a fixed geographic anchor.Comparisons with conventional measures show meaningful differences in how neighborhoods and flows are classified, particularly in polycentric metros. We then link changes in this network-based reverse-commute measure to tract-level shifts in workforce composition, industry structure, and neighborhood amenities to evaluate competing explanations rooted in spatial mismatch and consumer city dynamics.By reframing commuting as a systemwide network process, this approach underscores how measurement assumptions embed urban theory and provides a portable tool for analyzing commuting patterns across cities and over time.