Data centres have become one of the fastest-growing segments of nonresidential construction in the United States. The industry frames them as engines of jobs and tax revenue, but a growing number of economists and state legislators frame them as a cost borne by local taxpayers. Data support both descriptions. This paper lays out the figures on each side, then examines the single best-documented build in the country to show what a successful example of the economy changed by data centers actually requires.

The National Picture

Are data centers the most lucrative endeavor for economic growth? The most cited numbers come from a PwC study commissioned by the Data Center Coalition. For 2024, it credits the US data center industry with $926.9 billion added to GDP, 5.5 million jobs supported, and $204.4 billion in tax contributions to federal, state, and local governments.

The year-over-year growth is steep. GDP contribution rose from $768 billion in 2023 to $927 billion in 2024, a 21 percent increase. Jobs supported climbed from 4.7 million to 5.5 million, up 17 percent. Finally, tax contributions increased from $165 billion to $204 billion, up 24 per cent.

The study also reports two multipliers. For every $1 the data-centre industry contributes directly to GDP, its supply-chain purchases and employee spending are estimated to generate an additional $2.10 elsewhere in the economy. Similarly, every direct data-centre job is associated with more than 4.5 additional jobs across suppliers and the wider economy.

Construction activity also reflects the growth of data centers. Data center construction spending reached $59.3 billion at a seasonally adjusted annual rate in May 2026, up 23 percent from a year earlier, according to Census Bureau data. Investment in AI data centers, hardware, and networking amounted to 1.4 percent of US GDP in the first quarter of 2026, up from 0.7 percent.

Read the Big Number With the Alternative Method in Mind

The $926.9 billion figure is real, but it is doing more work than it should. The PwC study arrives at it by counting direct spending, then adding the suppliers that the spending pays. This is too far from the actual impact of the data centre, so it is possible to counterclaim that this consumer spending would exist in some form regardless of whether the data centre was there.

The study was also commissioned by the Data Center Coalition, the industry's own membership association, so it is not without bias.

Other analysts give humbler numbers. A Federal Reserve Board working paper (FEDS 2025-109) modelled import shares directly and found that roughly 70 per cent of data centre investment goes to high-tech equipment such as semiconductors, most of which is imported. That matters because imports subtract from GDP in the national accounts. The PwC model counts a $400 million chip order from Taiwan as economic activity, while the GDP calculation counts it as money that left the country. S&P Global estimated the net GDP contribution at about 0.5 percentage points in the second quarter of 2025, a measurable figure but a fraction of PwC's headline figure. Goldman Sachs put it even lower, at roughly 0.2 percent of GDP growth for 2025.

What Happens on the County Level

The national totals hide most of what matters locally. Research from Georgia Tech tracked what happens to a county after a data center opens. In the first three years, local employment rose about 0.9 percent, wages about 1.1 percent, and the number of business establishments about 1 percent. Over a longer horizon, those effects grew to roughly 3.5 percent, 5 percent, and 4.7 percent.

A single modeled project shows the upper end of the promise. A 2025 study in Ohio found the project could support about 9,700 construction jobs during the build phase, generate $2.4 billion in total economic output, add $1 billion to state GDP, and produce about $84 million in annual peak state and local tax revenue.

Georgia Tech also suggests that location determines the outcome. Metro and higher-income counties capture most of the employment and wage gains. When a data center is located in a rural county, there is just a small drop in unemployment and little beyond it. The characteristics of the host community, more than the facility itself, determine whether local benefits materialise. The factors that matter most are population density, an existing base of technology and professional services firms, proximity to a metropolitan area, and a labor market deep enough to supply both the facility and the businesses that grow around it.

Case Study: Loudoun County, the World's Largest Digital Hub

In the context of economic impact of data centers, it is most interesting to look at the real impact of the data center construction. Let's look at Loudoun County, Virginia, which holds the largest concentration of data centers on the planet, and is known as Data Center Alley. Luckily, there are publications of more than a decade of budget data showing what that expansion did to a local economy.

Data center revenue in Loudon County grew from $60 million in FY 2013 to more than $800 million in FY 2026. Measured against total payments, the Loudoun County Preservation and Conservation Coalition estimates data centers paid about $894.5 million in real and personal property taxes in 2025, up from $700.1 million in 2024, with roughly $1.14 billion projected for 2026.

That revenue now anchors the budget. Data centers generate 38 percent of General Fund revenue and nearly half of all property tax collections. Computer equipment tax revenue alone reached $684.8 million in FY 2026, a 17.8 percent jump. A county study valued the 2024 contribution at about $733 million, equal to 31 percent of local revenue. Data centers added $16 billion in value to the county's real property portfolio in 2024, bringing the data center total to $41 billion.

The expansion of data centers in Loudoun County changed the tax structure for residents. The county has lowered its real property tax rate every year for the past decade, from $1.145 per $100 in tax year 2016 to $0.805 per $100 in tax year 2026, the lowest rate in Northern Virginia. The same revenue allowed the board to cut the vehicle tax rate by 67 cents in FY 2026 and eliminate the $25 vehicle license fee.

A 2026 Mangum Economics report for the Northern Virginia Technology Council found that without data center revenue, Loudoun's residential real property rate would need to rise from $0.805 to about $1.537 per $100, roughly a 91 percent increase that would add about $5,800 per year to the typical homeowner's bill. The revenue also funded services rather than sitting idle. It supported year-over-year budget increases for Loudoun County Public Schools, totalling tens of millions of dollars, and funded the Sheriff's Office and Fire and Rescue. The corridor hosts more than 3,500 technology companies and provides more than 12,000 jobs in the county.

Loudoun County is one of the most successful examples of the data center construction in the county. First of all, it is the concentration of infrastructure. The corridor's fiber, power, and permitting advantages compounded as more operators clustered, which raised the assessed value base faster than any single facility could.

Second, Loudoun did not give data centers a full tax break. It continued to tax their equipment each year, while sales tax exemptions were available only to companies that met investment and job targets.

Third, Loudoun set aside part of its data center tax revenue to protect against an industry downturn. The county's Revenue Stabilization Fund targets 10 percent of this revenue, including a proposed $47.7 million contribution for FY 2026.

There are risks in centring the economy on a single project. Data centers generate substantial property tax revenue for local governments, but their equipment purchases are exempt from state sales tax. Virginia, therefore, gives up an estimated $1.6 billion a year, prompting lawmakers to consider new ways to collect more revenue from the industry.

Data centers bring in tax revenue, but they also consume enormous amounts of electricity. A single large facility can use as much power as roughly 80,000 homes, forcing utilities to revise demand forecasts. In some areas, electricity prices rise by around 5 percent after a data center opens. In Pennsylvania, Duquesne Light attributes $2.18 billion in new capacity costs to existing ratepayers, roughly $225 per customer per year.

The Policy Response

Legislators have noticed the electrical load the data centers have. In many ways, this is a result of the public response to data centres. More than 300 data center bills were filed across 30 states in the first six weeks of 2026, a shift from incentive-focused policy toward regulatory oversight. At least 18 states proposed special rate classes for large electric loads, several of which require data centres to fund infrastructure and demonstrate benefits to ratepayers. Virginia, Georgia, and Oklahoma introduced measures to cut or repeal the tax credits that first attracted these facilities. Oklahoma SB 1488 would pause projects above 100 megawatts until November 1, 2029, while the state studies the effects on water supply, utility rates, and property values.

The Infrastructure Behind the Numbers

US data center power demand is projected to rise from 31 GW in 2025 to 66 GW in 2027, according to Goldman Sachs. Meeting that demand requires major investment in substations, feeders, medium-voltage cable, and distribution systems. Loudoun's data center economy depends on this electrical infrastructure, and on the ability to source and deliver it at scale. Nassau National Cable supplies the wire and cable that powers data center builds across the country, from medium-voltage feeder runs to rack-level distribution. If you are planning a build or scaling an existing facility, request a quote for bulk pricing and availability.

The Bottom Line

The economic impact of data centres cannot be reduced to a single national GDP figure. Large estimates capture construction, supplier activity, and employee spending, but may overstate the domestic benefit when equipment is imported, and indirect spending is counted broadly. The more important question is where the value ultimately lands.

Loudoun County shows what a successful outcome can look like. Data centres can expand the tax base, lower residents' tax burden, fund public services, and attract related businesses—but only where the infrastructure, labour market, and tax structure allow communities to capture those gains. Elsewhere, the benefits may be smaller while the pressure on the power grid remains substantial.

The outcome, therefore, depends on policy and execution: how equipment is taxed, who funds grid upgrades, how incentives are tied to measurable benefits, and whether the required substations, feeders, medium-voltage cable, and distribution equipment can be sourced and installed on time.

Sources

1. PwC for the Data Center Coalition, "Economic Contributions of Data Centers in the United States, 2023-2024," May 2026. https://static1.squarespace.com/static/63a4849eab1c756a1d3e97b1/t/6a04986aac382850dcd9d637/1778686058456/Data+Center+Economic+Contribution+Study+2026_Final.pdf

2. The Center of Your Digital World, 2026 Impact Study, year-over-year growth and multipliers. https://www.centerofyourdigitalworld.org/2026-impact-study

3. Associated General Contractors of America, Data Centers Economic Impact, construction spending and Ohio project model. https://www.agc.org/datacenters/economic

4. Epoch AI, AI data center share of US GDP, Q1 2026. https://epoch.ai/data-insights/ai-datacenter-share-gdp

5. Federal Reserve Board, "Estimating Aggregate Data Center Investment with Project-Level Data," FEDS 2025-109. https://www.federalreserve.gov/econres/feds/files/2025109pap.pdf

6. S&P Global, "Data Center Investments Are Increasingly Moving the Macro Needle," December 2025. https://www.spglobal.com/en/research-insights/special-reports/look-forward/data-center-frontiers/data-center-investment-moves-macro-needle

7. Georgia Tech Scheller College of Business, Yue and Zeng, county-level effects and electricity prices. https://www.scheller.gatech.edu/news/2026/data-centers-are-booming-who-benefits.html

8. The Conversation, on the Yue and Zeng county-level findings. https://theconversation.com/why-better-off-cities-and-towns-see-more-benefits-from-data-centers-than-rural-regions-286750

9. Data Center Frontier, Loudoun FY 2026 budget, revenue arc, and Revenue Stabilization Fund. https://www.datacenterfrontier.com/site-selection/article/55266317/the-future-of-property-values-and-power-in-virginias-loudoun-county-and-data-center-alley

10. Introl, Virginia data center tax analysis, General Fund share and state sales tax figure. https://introl.com/blog/virginia-data-center-tax-exemption-battle-2026

11. Progress Chamber, on the 2026 Mangum Economics NVTC report, homeowner counterfactual. https://progresschamber.org/insights/data-centers-cut-property-taxes-virginia-homeowners/

12. Progress Chamber, Loudoun services and taxes, 2024 and 2025 payment figures. https://progresschamber.org/insights/loudoun-data-centers-better-services-lower-taxes/

13. Loudoun County, VA official site, Data Centers in Loudoun County. https://www.loudoun.gov/6188/Data-Centers-in-Loudoun-County

14. Loudoun County, VA FAQ, Data Centers, Tax Revenues and the County Budget. https://www.loudoun.gov/FAQ.aspx?QID=1790

15. Loudoun County, VA FAQ portal, real property portfolio and tax rate history. https://www.loudoun.gov/m/faq?cat=242

16. Independent Women's Forum, Loudoun assessment structure and tax rate detail. https://www.independentwomen.com/2026/05/19/data-centers-in-loudoun-county-va-created-significant-tax-reductions-for-residents/

17. MultiState, 2026 state data center legislation tracker. https://www.multistate.us/insider/2026/2/20/state-data-center-legislation-in-2026-tackles-energy-and-tax-issues

18. Goldman Sachs, US data center power demand projected to double by 2027, May 2026. https://www.goldmansachs.com/insights/articles/us-data-center-power-demand-projected-to-double-by-2027