Data Centers Spur Economic Growth, But Who Reaps the Benefits?
Artificial intelligence infrastructure expansion is driving unprecedented capital investments across the United States, yet the economic gains of these massive facilities remain sharply divided between metropolitan and rural communities. According to research published by faculty and fellows at the Georgia Institute of Technology Scheller College of Business, hosting a hyperscale data center delivers measurable boosts to local employment and wages, but these benefits heavily favor urban areas equipped with deep labor markets and dense business ecosystems.
- Metropolitan counties capture the vast majority of economic spillovers from data center openings, seeing quantifiable gains in employment, wages, and new business establishments over a three-year observation window.
- Rural and non-metro host communities experience negligible indirect economic growth despite absorbing localized infrastructure pressure and significant resource consumption.
- Electricity prices rise by an average of approximately 5% in areas where researchers can cleanly measure price effects following a data center activation, highlighting a critical trade-off for local rate payers.
The United States currently hosts more than 2,700 active or under-construction data centers, with individual hyperscale projects routinely exceeding $1 billion in capital expenditure. These facilities consume vast amounts of electrical power, often rivalling the usage profile of a small city. To evaluate the actual localized impacts of this industrial boom, Georgia Tech Scheller College of Business information technology management faculty member Daniel Yue and postdoctoral fellow Yiyang Zeng analyzed detailed facility-level data paired with county-level economic indicators.
Over the initial three-year period following a facility activation, host counties experience an average employment rise of approximately 0.9%, a 1.1% increase in wages, a 1.0% growth in business establishments, and a 0.7% bump in household income. Over extended operational timelines, these metrics expand further. Building permits also surge, reflecting the heavy construction activity required to erect these sprawling structures. However, Yue and Zeng note that these statistical gains are modest when juxtaposed against the sheer scale of multibillion-dollar capital investments.
Geographic clustering determines whether communities capture indirect economic spillovers. Metropolitan areas possess the agglomeration benefits necessary to supply construction contractors, specialized engineering firms, equipment vendors, and professional services. High-wage technical personnel employed by the facilities subsequently stimulate local retail, housing, and service sectors. Conversely, non-metro counties often see specialized services imported from outside the region, leaving local economies with fewer than 100 permanent facility jobs and minimal broad-based wage expansion.
Beyond employment figures, the research uncovers a distinct cost-shifting mechanism tied to public utility consumption. Because a single large data facility can draw as much electrical current as roughly 80,000 households, local power grids experience severe demand pressure. In markets where price impacts are isolatable, utility rates climb by about 5% following activation. Because utility tariff structures vary significantly by state and divide overhead differently among residential, commercial, and industrial classes, communities face complex policy decisions regarding who ultimately subsidizes grid reinforcement. When corporate developments shift infrastructure burdens onto local populations, administrative and regulatory oversight becomes critical.
Policymakers must look past headline-grabbing recruitment packages to evaluate long-term fiscal impacts. As Daniel Zeng emphasizes, host communities frequently overlook the fine print regarding tax exemptions and electricity tariff arrangements. Ensuring sustainable development requires rigorous, evidence-based scrutiny of local economic capacity before breaking ground.