Measuring Access to Amenities and Exposure to Risks among Manufactured Housing Units

This study provides examines exposure to environmental risks and access to environmental amenities among manufactured housing units across the contiguous United States. The dataset integrates high-resolution building footprint data with geospatial indicators derived from multiple administrative and environmental sources. Manufactured housing units are identified using a machine learning classification approach applied to building geometry and spatial characteristics. A wide range of environmental and socioeconomic variables are constructed for each building location, including:
Proximity to public transit stops and schools

Exposure to floodplains

Access to jobs and industrial activity

Land cover composition (e.g., green space, water, tree cover)

Road network characteristics and intersection density

Distance to facilities regulated by environmental authorities

These measures are computed using spatial joins, nearest-neighbor distance calculations, and buffer-based aggregations, typically within a 500-meter radius of each structure.
The final analytical dataset incorporates sampling weights derived from a stratified county-level sampling design to ensure representativeness across geographic areas and building types. The data are suitable for analyses of environmental inequality, infrastructure access, and housing vulnerability.
<br><br>