Mapping the Environmental and Socioeconomic Footprint of Data Centers in the United States

The rapid expansion of artificial intelligence and cloud computing has accelerated the construction of data centers across the United States, raising important questions about the spatial factors influencing their location. This study examines tract-level patterns of data center siting using Geographic Information Systems (GIS) and regression analysis. Data center locations were spatially joined to U.S. census tracts and enriched with demographic, housing, and economic characteristics. Descriptive comparisons indicate that tracts hosting data centers have higher median household incomes, lower poverty rates, lower population densities, and higher home values than non-host tracts. However, regression models incorporating metropolitan (CBSA) fixed effects reveal that these socioeconomic differences largely reflect variation across metropolitan areas rather than within them. Population density remains the only consistent predictor of data center presence, with facilities systematically located in lower-density environments within metropolitan regions. By integrating GIS-based spatial analysis with metropolitan fixed-effects regression, this study distinguishes regional from within-metropolitan determinants of digital infrastructure siting. It provides a reproducible spatial framework that establishes a baseline for future analyses of data center infrastructure and broader environmental impacts.