Two independent studies published in May 2026 put new numbers on the economic footprint of the US data center industry and push back on a persistent claim: that data centers are directly responsible for rising household electricity costs. The reports - one commissioned by the Data Center Coalition from PwC, the other from energy consultancy E3 - arrive as AI-driven infrastructure investment continues to accelerate across the tech industry.

The scale of economic contribution

According to PwC's report, titled "Economic Contributions of Data Centers in the United States," the industry supported 5.5 million jobs nationwide in 2024 - up 17 percent from 4.7 million in 2023. That figure includes direct employment within data centers themselves, which held roughly steady at just over 1 million jobs in both years, plus the indirect and induced employment that flows through supply chains and household spending.

Data+Center+Economic+Contribution+Study+2026_Final
Data+Center+Economic+Contribution+Study+2026_Final.pdf
16 MB
download-circle

The distinction matters. According to PwC, each direct job in the data center industry is estimated to support 4.5 additional jobs elsewhere in the US economy in 2024. That generates an employment multiplier of 5.5, meaning that for every worker directly employed inside a data center, roughly four and a half more are employed in adjacent sectors - construction, utilities, professional services, logistics, and consumer-facing industries.

On labor income, the numbers are equally large. According to PwC, direct labor income in the industry rose from $148.5 billion in 2023 to $164.4 billion in 2024, an 11 percent increase driven primarily by rising compensation across key occupational categories. Total labor income - including indirect and induced effects - grew 22 percent, from $431.1 billion to $525.3 billion. The labor income multiplier was 3.2 in 2024, meaning that for each dollar earned inside the data center industry, an additional $2.20 was generated elsewhere in the economy.

GDP contribution crosses $900 billion

The contribution to US gross domestic product grew faster still. According to PwC, the data center sector's direct value added - its contribution to GDP - rose 10 percent from $274.7 billion to $301.7 billion between 2023 and 2024. Including the full chain of indirect and induced activity, total value added reached $926.9 billion in 2024, up from $768.0 billion the previous year, representing 21 percent growth. The value added multiplier was 3.1: for every dollar of direct GDP contribution, an additional $2.10 was supported across the wider economy.

Capital spending drove much of the growth in indirect and induced effects. According to PwC, jobs supported by capital spending increased 56 percent year-on-year - from 1.1 million to 1.71 million - while labor income from capital spending rose 59 percent, from $102.9 billion to $163.2 billion. The pattern reflects the construction boom in large-scale data center facilities during 2023 and 2024, which the report links to increased AI workloads and continued hyperscale expansion.

Tax contributions reach $204 billion

Government revenues from the industry also climbed sharply. According to PwC, total tax contributions - including direct, indirect, and induced effects at the federal, state, and local level - increased from $164.7 billion in 2023 to $204.4 billion in 2024, a 24 percent rise. That was the fastest growth rate among all the economic indicators examined in the study.

Breaking down the $204.4 billion: social insurance contributions accounted for $58.6 billion, sales and use taxes for $55.0 billion, corporate income taxes for $28.9 billion, property taxes for $24.9 billion, personal income taxes for $24.7 billion, and other payments for $12.3 billion. Property taxes and corporate income taxes were the fastest-growing categories, both up 28 percent.

State-by-state distribution

The economic footprint is unevenly distributed, as one would expect given that data center operations cluster in states with abundant power, favorable land costs, and established telecommunications infrastructure. According to PwC, California led all states in direct employment in both years, with roughly 150,000 direct jobs in 2023 and 146,000 in 2024. Direct labor income from California's data center industry was $33 billion in 2023 and $38 billion in 2024; direct value added reached $64 billion and $72 billion respectively.

Other states with at least 30,000 direct jobs in both years include Texas, Florida, New York, Virginia, Georgia, Washington, New Jersey, and Illinois. When cross-state spillover effects are added to the picture - recognizing that economic activity regularly crosses state lines through supply chains and labor markets - California supported 818,160 total jobs in 2024, followed by Texas at 548,110 and Florida at 400,100.

The data center industry operated across all 50 states and Washington DC in 2024. According to PwC, 49 states and DC each had at least 1,000 direct data center jobs, and all but four states supported at least 10,000 total jobs once spillovers were included.

How PwC defined the sector

One methodological point is worth noting for readers who compare this study to earlier figures. PwC used an expanded, function-based definition of the data center sector, rather than limiting measurement to establishments classified under NAICS code 518210 (Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services). The rationale is that a growing share of data center service delivery occurs within firms classified under adjacent IT services and infrastructure-management industries - firms that may perform functionally identical roles, including responsibility for uptime and operational control - but are captured under different industry codes.

This approach captures data center services wherever they are produced. It does not introduce new economic activity into the analysis; it reallocates existing activity across industry classifications. The result is a broader measure of what the data center sector actually does in practice, particularly given the industry's shift toward hyperscale facilities, colocation environments, edge deployments, and software-defined systems.

The electricity question

The second study, produced by Energy and Environmental Economics (E3) and also commissioned by the Data Center Coalition, addresses a separate but politically charged question: are data centers pushing up electricity bills for American households?

Understanding-the-Drivers-of-Rising-Electricity-Rates-and-the-Role-of-Data-Centers_E3-2026
Understanding-the-Drivers-of-Rising-Electricity-Rates-and-the-Role-of-Data-Centers_E3-2026.pdf
2 MB
download-circle

According to E3, the available quantitative evidence does not support that conclusion - at least for the historical period examined. The report reviewed 11 studies on electricity rate trends and the role of data centers, combining backward-looking empirical analyses with forward-looking projections. It also conducted interviews with four industry experts.

The core finding, as stated by E3, is that "there is no quantitative evidence to date that data centers have historically been subsidized by other customers, and historical load growth has been found to be only one driver out of many others contributing to retail rate outcomes."

That nuance is important. Rising electricity prices are real; the LBNL 2026 study cited by E3 found that nominal electricity prices rose 29 percent from 2019 to 2025 nationally. But E3's analysis of the Bates White research found no clear correlation between load growth and rate increases at the state level. According to E3's review, states such as Texas and Virginia - which saw the largest increases in electricity load, largely driven by data centers - had some of the smallest rate increases over the 2019-2024 period. California and New York, by contrast, saw the largest price increases while load actually declined. The LBNL study found broadly similar results, with PJM as the main exception.

What actually drives electricity rates

E3 identifies five primary drivers of rising electricity costs that operate independently of data center load growth. Inflation has significantly increased the cost of labor, materials, and financing, directly raising utilities' revenue requirements. Natural gas price volatility has exposed regions reliant on gas-fired generation to input cost swings. Wildfire mitigation and grid hardening now represent a growing share of utility capital spending. Grid modernization - replacing aging infrastructure and complying with policy requirements - has added further costs. And in wholesale markets such as PJM, capacity price volatility has been driven by market design changes, power plant retirements, reduced accreditation of some resources, higher reliability targets, and interconnection backlogs.

On the PJM market specifically, E3 conducted its own analysis of the change in capacity prices between the 2024/2025 and 2025/2026 delivery years. According to E3, load growth - primarily from data centers - accounted for approximately 50 percent of the capacity price increase, while the other 50 percent reflected market design updates and power plant retirements. The finding places PJM as a regional exception to the broader national pattern, where the relationship between data center growth and rate changes is less clear.

Data centers as a potential cost offset

A separate E3 study examining Amazon data centers across four utility territories - Pacific Gas & Electric in California, Umatilla Electric Cooperative in Oregon, Dominion Energy in Virginia, and Entergy in Mississippi - found that each facility generated, on average, $3.4 million in net surplus revenues above the incremental cost to serve. According to E3, this surplus arises because large, high-load-factor customers like data centers spread fixed infrastructure costs across a larger volume of kilowatt-hours, reducing the per-unit cost burden on other customers.

The mechanism was described by Xcel Energy CEO Bob Frenzel in a 2026 remark cited by E3: "if you put more units of production through a fixed asset, the rate for everybody comes down. A large data center load on your system should, all things being equal, drive prices down for electricity, not up." In Georgia, growth from data centers is expected to place downward pressure on residential bills by $8.50 per month from 2029 to 2031, according to the LBNL 2026 study cited in the E3 paper. In Michigan, residential costs are expected to fall by 8 percent from recently approved data center contracts. PG&E has measured the benefit directly, with large load growth enabling four electricity rate reductions over a two-year period.

At the same time, E3 is careful about forward-looking risks. The pace and scale of new data center construction have accelerated significantly, and the evidence base is largely historical. Forward-looking studies - including a Federal Reserve Bank of Dallas paper and an analysis from the Electric Power Research Institute - highlight conditions under which rapid load growth could put upward pressure on electricity prices, particularly in regions where new generation and transmission cannot keep pace with demand growth.

Tariff design as a regulatory tool

One concrete development highlighted in the E3 report: at least 38 new large load tariffs have been established by utilities between 2018 and 2026 specifically to assign customer-specific costs to data centers and reduce risk for other ratepayers. Thirty of those 38 tariffs were implemented in 2025 and 2026 alone. According to E3, mechanisms include minimum monthly demand payments, contribution in aid of construction, exit fees, minimum load factor requirements above 50 percent, and long-term contract terms exceeding 10 years.

State responses vary. Texas developed Senate Bill 6 to manage large load growth without producing disproportionate rate increases. Georgia regulators strengthened long-term contract requirements and minimum billing protections. Arizona refined rate design and contractual protections to align more closely with cost causation. Missouri proactively established large load tariffs and long-term contractual requirements following extensive stakeholder involvement.

Context for the marketing industry

The scale of data center economic activity has direct relevance for digital advertising professionals. The compute infrastructure that runs programmatic auctions, AI-powered bidding systems, and real-time analytics is physically housed in the facilities measured by these studies. Microsoft's Fairwater datacenter in Wisconsin, described as the world's most powerful AI datacenter when announced in September 2025, houses hundreds of thousands of NVIDIA GB200 GPUs and is directly connected to the infrastructure that runs large-scale AI model training. Google's first US nuclear energy deal, announced in August 2025, reflects the same underlying pressure: the compute demand driving these economic contribution numbers requires long-term, stable energy supply.

The advertising technology industry's own economic contribution - which a separate 2025 study found supports 29 million US jobs - depends in part on the physical infrastructure layer these studies measure. As AI inference and training workloads continue to expand inside digital advertising platforms, the capital spending captured in PwC's data center multipliers will likely reflect activity that advertising technology professionals interact with daily.

The Data Center Coalition, which commissioned both studies, represents the US data center industry's interests through public policy and energy advocacy. PwC conducted the economic modeling using the IMPLAN input-output framework, drawing on 2023-2024 data from the US Census Bureau, Bureau of Economic Analysis, and Bureau of Labor Statistics. E3, which describes itself as an independent economic consultancy focused on the power sector, reviewed 11 quantitative studies and conducted four expert interviews. Both reports note their findings reflect historical economic activity and should not be read as forecasts.

Timeline

Summary

Who: The Data Center Coalition commissioned PwC and Energy and Environmental Economics (E3) to produce the two studies. Meta's Chief Global Affairs Officer Joel Kaplan shared the findings on LinkedIn.

What: PwC quantified the US data center industry's total economic contribution as 5.5 million jobs, $525.3 billion in labor income, $926.9 billion in GDP contribution, and $204.4 billion in government revenues in 2024. E3 reviewed 11 quantitative studies and found no historical evidence that data centers have shifted electricity costs onto residential or small commercial customers, while identifying five primary drivers of rising electricity rates unrelated to data center load growth.

When: Both studies were published in May 2026, covering data from 2023 and 2024 for the PwC study, and a broader historical and forward-looking period for E3.

Where: The PwC study covers all 50 US states and Washington DC, with state-level detail in appendices. The E3 study draws on research from multiple US markets including Virginia, Texas, Georgia, Arizona, and the PJM region.

Why: The studies matter for the marketing and ad technology industries because the data center infrastructure measured in both reports is the physical layer that runs programmatic advertising, AI-powered bidding, and digital media delivery. Rising energy costs and policy questions about cost allocation affect the economics of that infrastructure. The findings also add evidence to a regulatory and political debate about whether data center growth causes household electricity price increases - a debate that affects the operating environment and public perception of the technology sector broadly.

Share this article
The link has been copied!