Data Center County Outcomes:
local economies?
The evidence is still emerging.

Background. Data center investment has accelerated sharply since 2018, raising questions about local economic spillovers. Boosters cite job creation and tax revenue; critics point to high capital intensity and limited direct employment. Prior studies rely on event studies without controls or on developer-submitted impact reports with obvious incentive problems.

Methods. We restrict the study to counties where the first data center opened cleanly — no prior DC activity and a single isolated opening event — and only include a county once enough post-opening data exists for a stable estimate. 25 counties satisfy the clean first-opening rule; 9 are mature enough to report as of today, with the remainder auto-graduating as time passes. Each qualifying county is paired with a control group of statistically similar counties that never got a data center, selected by our proprietary propensity-matching model built on pre-opening economic indicators. We then estimate a difference-in-differences: how the treated county's economy changed after the opening, versus how its matched peers changed over the same window. All estimates carry bootstrapped 95% confidence intervals.

Outcomes. The employment metric captures DC-adjacent sector employment — a proprietary bundle of federal labor statistics (BLS QCEW) covering the construction, data infrastructure, IT services, and utility sectors most directly affected by a new data center facility. This is deliberately not total private employment.

Limitations. Null quarterly/monthly values (from BLS disclosure rules or missing Comptroller data) are excluded from observation windows rather than imputed; estimates with high suppression fractions should be interpreted with caution. Fiscal metrics are available for Texas counties only. IRS SOI migration and LEHD/LODES flow results are excluded pending vintage validation.