Why AI Data Centers Boost Cities But Leave Rural Areas in the Dark

Artificial intelligence is driving an unprecedented wave of data‑center construction across the United States, prompting many municipalities to weigh the promised economic upside against potential drawbacks.

Researchers from Georgia Tech’s Scheller College of Business—faculty member Daniel Yue and postdoctoral fellow Yiyang Zeng—have compiled facility‑level data and county‑wide economic indicators to assess how new data‑center openings influence jobs, wages, business activity and electricity costs. Their analysis highlights geography as a decisive factor in determining who actually benefits.

Capital surge yields modest but measurable gains

More than 2,700 data‑center sites are either operational or under development nationwide, with many hyperscale projects exceeding $1 billion in construction costs and consuming electricity comparable to a small city. Over the first three years after a center becomes active, host counties experience an average increase of 0.9 % in employment, 1.1 % in wages, 1.0 % in business establishments and 0.7 % in household income. Extending the horizon to longer‑term effects, those figures rise to roughly 3.5 %, 5.0 %, 4.7 % and 1.9 % respectively, accompanied by a sharp uptick in building permits.

While these outcomes are statistically significant, the researchers note that they remain modest relative to the scale of the underlying investments and are not evenly distributed across the nation.

Urban clusters capture the bulk of spillovers

Metropolitan counties reap the greatest portion of the economic upside. Dense labor markets and established networks of contractors, engineers, equipment suppliers and professional services create an environment where the indirect spending generated by data‑centers can circulate rapidly. High‑skill technical staff boost demand for local restaurants, retail outlets and other services, while suppliers can scale operations swiftly, amplifying the initial capital injection.

Rural locales see limited trickle‑down effects

In contrast, non‑metropolitan counties often host facilities that employ fewer than one hundred permanent workers, relying heavily on external providers for specialized services. Consequently, the broader community experiences only modest benefits, such as a slight reduction in unemployment rates and occasional tax‑revenue gains, but the pronounced job and wage growth frequently advertised during site‑selection campaigns rarely materializes.

“It’s the characteristics of the host community, not just the facility opening there, that determine whether the local benefits show up,” Zeng explains.

Energy demand and rising electricity costs

Data‑centers are power‑intensive; a single large site can consume enough electricity to supply roughly 80,000 homes. In regions where price impacts can be isolated, electricity rates climb by about 5 % following a center’s entry. This increase can affect households, businesses and the broader utility infrastructure, underscoring the need for transparent cost‑allocation mechanisms.

Yue and Zeng advise communities to scrutinize who bears the expense of new grid upgrades and how utility tariffs are structured, recognizing that cost‑sharing models differ widely among states.

Key considerations for local policymakers

For cities and counties eager to attract data‑center projects, the study suggests focusing on concrete terms rather than headline‑grabbing incentives. Critical factors include the design of tax‑abatement agreements, the specifics of electricity tariff arrangements and the allocation of infrastructure costs.

“In ten years, communities will likely wish they had pressed harder on the quality of the decision itself, including whether the debate was evidence‑based, whether their local economy was equipped to capture the gains, and whether the fine print aligned with residents’ long‑term interests,” Zeng notes.

As the landscape of American data‑centers continues to evolve, discerning where these facilities generate shared value—and where they fall short—will remain essential for informed community leadership and policy formulation.