The Economic Impact of Data Centers on Communities: Taxes, Jobs and the Evidence

Economic impact studies of data centers measure three channels. Direct effects are the spending and jobs at the project itself: construction crews, then the operations staff. Indirect effects are the suppliers those projects buy from, such as electrical contractors, equipment vendors and security firms. Induced effects are the household spending of everyone employed in the first two groups. Fiscal effects, the taxes a project pays, are usually reported separately and matter most to a county.

The direct footprint of a finished facility is small for its size. Virginia’s Joint Legislative Audit and Review Commission (JLARC) found that a typical 250,000-square-foot data center may employ about 50 full-time workers, roughly half of them contractors, while a single building can have about 1,500 workers on site at the peak of construction.¹ Most of the industry’s measurable benefit to Virginia’s economy came from the construction phase rather than from ongoing operations.¹

That pattern holds wherever large campuses are built. Meta’s agreements for its Hyperion campus in Richland Parish, Louisiana, which began with a $10 billion investment commitment, call for up to 500 full-time jobs by the end of 2032, with interim targets of 100 by the end of 2028 and 300 by the end of 2030.⁸ Those are meaningful numbers for a rural parish, but they are small relative to the capital invested, which is why the tax side of the ledger usually dominates local discussions.

JLARC’s December 2024 study put the industry’s annual contribution to Virginia at about 74,000 jobs, $5.5 billion in labor income and $9.1 billion in state GDP, counting direct and indirect effects.¹ Its briefing split that total by phase: about 59,000 jobs, $4.3 billion in labor income and $6.4 billion in GDP tied to construction, against about 15,000 jobs, $1.2 billion in labor income and $2.7 billion in GDP tied to operations.²

Construction employment is real but conditional. It lasts as long as building continues, and Good Jobs First, a subsidy watchdog, notes that construction work on these projects is temporary and often filled by workers from outside the state.⁶ A county that hosts one campus sees a hiring wave and then a much smaller permanent workforce. A county in a market where campuses keep coming, such as Northern Virginia, sees something closer to a permanent construction industry, because new phases replace finished ones.

Operations jobs are fewer but steadier and tend to pay well: facility engineers, electricians, technicians and security staff. Many state incentive programs set wage floors for them. Virginia’s sales tax exemption, for example, requires 50 new jobs paying at least 150% of the locality’s prevailing average wage, along with $150 million in capital investment.⁹ The practical question for a community is whether those jobs will be filled locally, which depends on workforce training as much as on the project.

For most host counties the larger and longer-lasting gain is property tax. Data centers pay real property tax on land and buildings and, in many states, business personal property tax on servers and other equipment, which is replaced every few years and keeps the assessed value high. The mechanics are covered in our guide to property taxes on data center projects.

Loudoun County, Virginia is the best-documented case. The county reported that data centers generated about $1.2 billion in real and personal property tax revenue in FY2026, about 39% of its overall budget, and its adopted FY2027 budget projects about $1.3 billion, or 40%.³ JLARC’s summary listed data center local tax revenue of $733 million (31%) for Loudoun and $110 million (7%) for Prince William County in its study period, which shows how quickly the figures have grown.⁹

The county attributes part of that benefit to how little the buildings ask of public services, noting that data centers consume very few county resources, so most of the revenue is available for schools and tax relief.¹¹ Loudoun’s real estate rate has fallen from about $1.29 to roughly 81 cents per $100 of assessed value since 2012, which county officials estimate saves the average homeowner $3,000 to $4,000 a year.¹⁰ The county also warns that the revenue is hard to forecast and creates variability in planning.¹¹

The most careful evidence comes from state auditors evaluating tax incentives, because they have to ask what would have happened without the incentive. Their answers vary widely, and the assumptions behind them matter more than the headline numbers.

Georgia’s case shows how fragile these estimates are. Auditors initially reported in December 2025 that the exemption, which cost about $474 million in the fiscal year ending in 2025, had produced 28,350 construction jobs and 5,471 operations jobs. A January 2026 revision cut both figures to less than a third after the University of Georgia’s Carl Vinson Institute of Government concluded that only 30% of the state’s data centers could be attributed to the exemption.⁴⁵ The revised summary still found economic growth larger than the revenue forgone, but direct state tax revenue from construction and operations was about $41.5 million combined.⁴

Across states, Good Jobs First reported in 2026 that Georgia, Virginia and Texas each lose $1 billion or more a year to data center incentives, that 14 states do not disclose their losses, and that states that have calculated returns found losses of 52 to 70 cents per dollar exempted.⁶ Virginia’s exemption alone has saved the industry at least $2.7 billion and is set to expire in 2035.¹² The state-by-state rules are in our guide to data center sales tax exemptions by state.

National studies funded by the industry report much larger numbers because they count every indirect and induced job across the economy. PwC’s 2026 study for the Data Center Coalition estimated that the industry supported about 5.5 million U.S. jobs in 2024, up from 4.7 million in 2023, with direct employment at data center companies holding near 1 million in both years.⁷ By that estimate, each direct job supported about 4.5 additional jobs elsewhere.⁷

These figures are not wrong, but they answer a different question from the one a county asks. A national multiplier credits jobs to the industry wherever they occur, including in equipment factories and corporate offices in other states. A host county captures its construction wave, its permanent staff, some local supplier spending and its tax base. When evaluating a project, local officials tend to get more from a project-specific fiscal impact analysis than from national totals.

Critics do not usually dispute that data centers pay large tax bills. They argue that incentives give too much of the bill away, that the jobs per dollar of subsidy are few, and that the costs land on residents in other forms: transmission lines, noise, water use and, potentially, electricity rates. That last issue is covered in our guide to data centers and electricity bills.

Rural deals often rely on local abatements or a payment in lieu of taxes. Richland Parish agreed to a 30-year property tax reduction of up to 60% for Meta’s project, rising to as much as 80% if enough jobs are created.⁸ Even so, a columnist at The Advocate estimated the parish is likely to collect more than $70 million from the PILOT and sales taxes during construction, almost four times its prior take, while noting that payments shrink once the campus is built and operating.¹³

For a landowner or developer, the debate matters because it shapes approvals. Communities that have studied the numbers increasingly ask for written commitments on tax payments, jobs and infrastructure, and some have paused approvals altogether, as described in our guide to data center moratoriums and local restrictions.

A project’s local economic case can be estimated before anyone files an application. The useful questions are concrete:

• What is the local tax rate on business personal property, and does the state exempt or abate equipment? This sets most of the long-run revenue.
• Which incentives apply, which are statewide and which are negotiated locally, and what job and wage thresholds do they carry?⁹
• How many construction phases are planned, and over how many years? That sets the length of the construction wave.
• How many operations jobs are committed in writing, and what happens if they are not met?⁸
• What public infrastructure does the project need, such as roads, water and transmission, and who pays for it?

Economic developers will find a fuller playbook in our guide for economic development organizations, and developers can compare incentive packages using our guide to data center tax incentives. How we weigh these factors in a site screen is described in our methodology. If you are weighing a parcel, you can get a site reviewed.