Research constructs new neighborhood geography from migration flows
A new working paper from the Federal Reserve Bank of Philadelphia constructs a novel neighborhood geography. Researchers Evan Mast and Alaina Barca define 'districts' based on migration flows, using a revealed preference approach.
Mapping urban areas through movement
Researchers Evan Mast and Alaina Barca introduce a novel method to delineate neighborhood boundaries, termed 'districts,' by analyzing residential migration patterns.
Their approach leverages a revealed preference intuition: if people disproportionately move within certain areas, those boundaries can be inferred from migration flows.
The study utilizes comprehensive address history data from the U.S. Census Bureau, converting migration flows within each county into a network where census tracts are nodes.
The Infomap algorithm then partitions these networks into modules with high rates of cross-migration, forming the new districts.
These districts average 8-10 census tracts, corresponding to roughly 40,000 people, and aim to fill the spatial information gap between individual tracts and counties.
The resulting tract-to-district crosswalk and interactive maps are publicly available.
Beyond census tracts: Real-world alignment
The reliance on census tracts often limits the understanding of neighborhood dynamics, as important drivers of segregation and neighborhood effects can operate at broader scales.
This new 'district' geography addresses this gap by providing a resolution between tracts and counties.
The study demonstrates that district boundaries align strikingly with physical barriers like interstates, rivers, and railroad tracks, as well as with municipal and school district borders.
These alignments suggest that the revealed preference method effectively identifies salient features that influence migration and define cohesive local areas.
The districts often resemble popularly known neighborhoods in cities and coincide with towns or school districts in suburbs, offering a more intuitive and empirically grounded definition of local areas for social science research.
A practical leap for urban analysis
This paper offers a significant methodological leap for urban analysis, providing a scalable and data-driven approach to a long-standing problem.
By constructing 'districts' from actual migration patterns, the research moves beyond the arbitrary nature of census tracts, offering a more realistic lens for studying neighborhood effects.
The practical applications for improving the precision and relevance of spatial data analysis are substantial, making this a valuable tool.