Measuring Integration in Affordable Housing with Location Data
Housing policy in the United States increasingly funds programs — vouchers, LIHTC, inclusionary zoning — premised on the idea that moving lower-income residents to 'better' neighborhoods will improve their access to schools, amenities, employment, and social networks. Yet remarkably few studies have measured at national scale whether this proximity translates into actual experienced integration. This project uses a novel dataset of anonymized mobility patterns for over 66 million U.S. residents, recorded at 15-minute intervals, to develop three measures of experienced integration — social diversity, distance traveled, and types of places visited — and compare subsidized housing residents to matched control groups nationwide.
A case study component examines 29 subsidized developments across four sub-metropolitan areas (Central New Jersey, Denver, Greater Milwaukee, and San Antonio), analyzing four dimensions of realized opportunity: physical environment, school access and outcomes, amenities, and employment. Findings reveal persistent barriers to the hypothesized mechanisms of integration, including physical isolation from surrounding neighborhoods, low rates of school mixing, poor distribution of affordable amenities, and limited employment access. Results cluster into three neighborhood types — downtowns with influxes, slowly-declining inner-ring suburbs, and outer peripheries — each presenting different policy implications. An ongoing analysis of California's high-opportunity affordable housing program will extend these findings with a focus on state-level siting policy.
This project is funded by NSF Award #2621678.
Publications
Same street, different world: Unpacking high opportunity areas with subsidized housing in four US cities
Baird-Zars, B., Xu, W.
Under review, 2026
Housing policies increasingly prioritize and pay a premium for subsidies to locate lower-income families in higher-income “opportunity” neighborhoods. Yet the evidence on outcomes in the immediate term is conflicting, and studies tend to be either national examinations of census data or in-depth single case studies.
Using an extensive mixed-methods case approach, we query the actually-existing conditions of “opportunity” neighborhoods of 29 subsidized housing developments in four sub-metropolitan areas.
Across neighborhoods, we find persistent barriers to the hypothesized mechanisms of access and integration, including isolation from the surrounding physical fabric, low rates of school mixing and outcomes, poor distribution of affordable and useful amenities, and unremarkable employment possibilities.
We identify three neighborhood types that correspond to variation in our findings: downtowns with influxes, slowly-declining inner-ring suburbs, and outer peripheries. The clustering of results suggests initial steps for strategic disaggregation of future analyses, policy criteria, and incentives to more accurately identify and advance “realized opportunity.”
Draft available upon requestDoes siting matter? Measuring neighborhood integration in California's high opportunity affordable housing
Xu, W., Baird-Zars, B.
In preparation, 2026
Draft available upon request