By Shane Snider | Senior News Writer, Data Center Knowledge
August 19, 2026
Executive Overview
As the explosive growth of generative artificial intelligence (AI) and cloud computing workloads continues to strain power grids across the United States, federal lawmakers are stepping into the fray with a sweeping legislative proposal.
Rep. Andrea Salinas (D-Ore.) has officially introduced the Data Center Community Reinvestment Act of 2026, a groundbreaking House bill designed to impose a federal excise tax of 1 cent per kilowatt-hour (kWh) on electricity consumed by commercial data centers.
Targeting facilities with a maximum rated power capacity or peak load exceeding 1 megawatt (MW), the measure represents the first federal attempt to directly monetize the soaring energy appetites of the digital infrastructure sector. If enacted, the tax would apply broadly across the ecosystem—reaching everything from hyper-scale AI campuses down to mid-sized enterprise, colocation, and institutional facilities—while sweeping up both grid-supplied electricity and self-generated, behind-the-meter power.
Proponents argue the bill ensures that rapid technological expansion does not leave local populations holding the bag for infrastructure upgrades and environmental strains. However, energy market experts warn that a blanket consumption tax fails to address the underlying economic and technical realities of "cost causation" within modern power grids.
Detailed Chronology & Legislative Landscape
The introduction of the Data Center Community Reinvestment Act arrives at a critical juncture for both the technology and utility sectors, reflecting months of mounting political and economic pressure.
The Rise of Digital Infrastructure Backlash
Over the past three years, the commercial real estate and utility landscapes have been completely upended by the generative AI gold rush. Hyperscale operators—including tech giants like Microsoft, Google, Amazon, and Meta—have scrambled to secure massive blocks of power, often pushing regional transmission organizations (RTOs) to their absolute limits.
This unprecedented demand has triggered fierce public debates over utility rate hikes, carbon emissions, and the strain placed on local water and electrical networks. Communities from Northern Virginia to the Pacific Northwest have increasingly pushed back against data center developments, demanding tangible community benefits or enacting outright moratoriums.
State-Level Precedents Set the Stage
Representative Salinas’s federal bill does not emerge in a vacuum. It builds directly upon a state-level regulatory revolution that began earlier this year when Virginia—home to the world’s densest concentration of data centers, known globally as "Data Center Alley"—became the first state in the nation to enact a direct tax on data center electricity consumption.
Effective July 1, Virginia’s state law established a 1.1-cent-per-kWh consumption tax, applying it universally to utilities, retail power providers, and self-generated energy sources. Initial Virginia state budget projections estimated that the policy would pull in approximately $600 million annually for the state’s general fund.
While Salinas’s federal proposal borrows from Virginia’s blueprint, it introduces a slightly lower rate (1 cent versus 1.1 cents) and a fundamentally different revenue-distribution mechanism. Rather than dumping the receipts into a centralized state pool, the federal bill mandates that the proceeds be funneled into five distinct federal trust funds aimed at public welfare and infrastructure.
Committee Referrals and Congressional Path
Following its formal introduction in mid-August 2026, the bill was swiftly referred to the House Ways and Means Committee, reflecting its tax-writing jurisdiction. Given the expansive cross-sections of energy policy and technological development involved, the legislation also received concurrent referrals to the House Energy and Commerce Committee and the House Science, Space, and Technology Committee.
As of late August 2026, the bill remains firmly in the introduced-bill stage, with political analysts debating its viability in a deeply divided Congress. Nonetheless, its mere presence signals a major escalation in federal legislative interest regarding the digital economy’s physical footprint.
Financial Impact & Economic Metrics
To fully understand the weight of a 1-cent-per-kWh excise tax, industry analysts have crunched the numbers across various facility scales. For major operators, the financial exposure is profound.
Cost Projections by Facility Scale
According to metrics analyzed by Neil Osnato, founder of Persistence Analytics Group, the tax burden scales aggressively alongside facility capacity:
- 100-Megawatt (MW) Facility: Operating at a continuous full load, a 100 MW data center would incur an annual tax liability of approximately $8.76 million. At a more typical 90% annual load factor, that figure dips slightly to roughly $7.9 million.
- 500-Megawatt (MW) Facility: Large-scale cloud campuses scaling to 500 MW would face an annual tax bill of roughly $43.8 million under continuous operation, or approximately $39.4 million at a 90% load factor.
- 1-Gigawatt (GW) Super-Campus: Massive next-generation AI training clusters reaching 1 GW of capacity would see staggering annual excise liabilities of $87.6 million (continuous) or $78.8 million (90% load factor).
Percentage Additions to Baseline Power Costs
Osnato notes that the economic friction introduced by the tax depends heavily on local wholesale power prices.
- If a data center’s underlying electricity cost sits at 5 cents per kWh, a 1-cent tax tacks on an immediate 20% premium to the facility’s energy component.
- If the base power cost is 10 cents per kWh, the tax adds a 10% premium.
While these numbers are striking, Osnato suggests that for the largest hyperscale and AI players, the tax may only affect siting decisions "at the margin." When developers evaluate multi-billion-dollar AI builds, access to executable megawatts, transmission line availability, fast interconnection timelines, and overall generation certainty often eclipse the marginal cost of a federal excise tax.
Revenue Allocation: Where the Money Goes
Unlike typical corporate income or equipment taxes, the Data Center Community Reinvestment Act explicitly frames data center energy consumption as a vehicle for broader public investments. Salinas’s office projects that a nationwide 1-cent tax will generate approximately $1.76 billion annually based on current energy-consumption estimates.
The legislation carves this revenue into five equal slices, directing roughly $352 million per year into each of the following federally managed destinations:
- The Land and Water Conservation Fund: Dedicated to preserving federal, state, and local public lands, safeguarding water resources, and enhancing recreational opportunities nationwide.
- The Housing Trust Fund: Focused on building and preserving affordable housing for extremely low- and very low-income families, addressing a crisis exacerbated by rapid tech-sector real estate developments.
- The Hazardous Substance Superfund: Aimed at financing the cleanup of some of the nation’s most contaminated land and responding to environmental emergencies.
- The Highway Trust Fund: Supplying capital for the maintenance, repair, and expansion of federal highways, bridges, and mass transit systems.
- A Newly Established Energy Technology Trust Fund: Designed specifically to support federal loan guarantees for advanced clean energy projects, next-generation nuclear power development, and critical grid infrastructure enhancements.
By spreading funds across conservation, housing, infrastructure, and clean energy, the bill seeks to position data centers as financial contributors to the broader public good.
Expert Perspectives: The Flaw in "Cost Causation"
Despite the populist appeal of forcing tech giants to fund housing, conservation, and transit, industry specialists argue that a blunt federal consumption tax fundamentally misdiagnoses the engineering and economic challenges facing the U.S. power grid.
The Problem of Megawatt Equivalency
Neil Osnato points out that the core policy debate in the power sector centers on cost causation—the principle that customers who drive specific expenses on the electric grid should be the ones to pay for them.
"A flat federal consumption tax does not distinguish among any of those conditions," Osnato explained, emphasizing that "megawatts are not interchangeable."
Under the proposed framework, two distinct 500-MW data centers could consume the exact same quantity of kilowatt-hours while imposing radically different operational and financial burdens on their host electric systems. Variables that dictate true grid impact include:
- Geographic Siting: Whether the facility locates in a transmission-constrained load pocket or an area with surplus generation capacity.
- Load Profiles: Whether the facility’s power demand coincides directly with the utility’s peak system load hours.
- Infrastructure Requirements: The specific transmission line upgrades, local substation constructions, and new generation resources necessitated by the build.
- Operational Flexibility: Whether the data center features demand-response capabilities, allowing it to dial back compute workloads during grid emergencies.
Layered Electricity Costs
Critics of the bill emphasize that data centers are not currently riding the grid for free. Large commercial users already pay through multiple structured layers of the electricity system—including energy charges, transmission tariffs, distribution fees, and capacity markets. Furthermore, many hyperscalers are increasingly locked into customer-specific interconnection agreements and private construction charges to fund the exact grid upgrades they require.
Osnato warns that lawmakers risk committing a policy error by assuming that a federal tax substitutes for rigorous electrical engineering analysis.
"Policymakers could collect money from the right industry while still allocating the underlying grid costs incorrectly," he cautioned.
Rather than relying on flat consumption metrics, Osnato argues that regulatory focus should remain on granular questions: Is the projected load real? When will it materialize? What specific infrastructure does it cause? And how durable is the demand?
Future Outlook & Industry Implications
As the Data Center Community Reinvestment Act sits before its assigned House committees, its immediate legislative trajectory remains uncertain. Passing a novel federal excise tax in a closely divided Congress is an uphill battle, especially given fierce opposition anticipated from major tech trade associations and enterprise coalitions.
However, the introduction of H.R. 10102 marks a watershed moment in the political maturation of the digital infrastructure debate. It proves that the societal externalities of the AI boom—ranging from local utility rate pressures to community land-use tensions—have firmly reached the steps of Capitol Hill.
Even if Salinas’s specific bill stalls in committee, the template it establishes—treating data center energy consumption as an economic engine for public reinvestment—is unlikely to disappear. Whether through state-level taxes à la Virginia, specialized RTO rule changes by entities like PJM Interconnection, or future federal mandates, the era of unhindered, low-scrutiny power consumption for hyper-scale computing has officially drawn to a close.
For data center developers, cloud providers, and utility planners alike, the message is clear: the true cost of compute will increasingly include not just the price of silicon and substations, but a reckoning with the communities and power grids that make the digital world possible.
