The Effect of Data Center Development on Local Housing Prices and Rents: Evidence from Virginia and Texas

Abstract

This paper estimates the effect of data center completion on nearby residential housing prices and rents. To identify causal effects, I use a stacked difference-in-differences design that compares properties located within 1.5 miles of a facility to properties located 5 to 10 miles away. The pooled average effect on housing prices is small and statistically indistinguishable from zero. This null reflects a composition effect, where the sample includes both small facilities with no detectable impact and larger facilities with negative effects. Small facilities, typically co-location sites with limited physical footprint, show no detectable effect, whereas mid-size and large purpose-built facilities reduce nearby housing prices by about 1.3 to 1.8 percent. The pooled rental effect is negative and statistically significant, with declines concentrated around mid-to-large facilities. The comparable magnitudes of the housing and rental effects for larger facilities are consistent with a contemporaneous local disamenity channel, rather than with a sustained local demand boom.