For the past year, data centers have become a convenient scapegoat for rising utility bills. State legislatures have even floated construction bans, and local governments have passed moratoriums as a result. The uproar led to the White House intervening in March 2026 with the “Ratepayer Protection Pledge,” an assurance that data centers would not raise consumer electric bills. The economic evidence backs this assurance.
The Conventional Wisdom — And Why It’s Wrong
A new working paper from researchers at EPRI and Watershed set out to answer a simple question with real data instead of assumptions: did the recent boom in data center construction actually raise what Americans pay for electricity? Using an instrumental variables approach, a statistical method designed to isolate cause from mere correlation, the researchers found that data centers actually caused average U.S. residential electricity rates to fall, not rise, between 2015-2024. A doubling of a state’s data center capacity was associated with a roughly 3.5% decrease in residential rates, holding household electricity use constant.
How did more demand lead to lower prices? The answer lies in basic power-sector economics that gets lost in the current political noise. Electricity retail rates aren’t set purely by marginal supply and demand the way, say, gasoline prices are. The grid carries enormous fixed costs due to the poles, wires, substations, and power plants that exist and operate regardless of how much electricity flows through them. Because of this, they operate using a rate structure that recovers those costs across every kilowatt-hour sold. When durable new demand like a data center comes online, it spreads those fixed costs over more usage, and it often justifies newer, more efficient generation that costs less per unit than the aging assets already on the books. The net effect, as the study’s authors describe it, is that new demand can lower average costs for everyone in the system.
This aligns with what NetChoice has argued for months: large data centers shoulder a growing share of utilities’ fixed costs, which reduces demand charges on residential bills. The real culprits behind recent rate increases include aging grid infrastructure, extreme weather disasters, state renewable mandates, and the retiring of natural gas, coal, and nuclear power plants — not the servers powering America’s digital economy.
A Case Study: Virginia v. California
Virginia is the nation’s data center capital, hosting more capacity than any other state, and where data centers draw over 20% of the state’s electricity. If the “data centers spike your bill” narrative were true, Virginia’s ratepayers should be feeling it acutely. Instead, the study finds Virginia’s residential rate increases tracked closely with the national average, with no meaningful outlier effect occurring from being the epicenter of American data center growth. Virginia’s experience is a live, large-scale test of the theory, and the results undercut the fear-based narrative driving opposition in statehouses around the country.
Contrast that with California, which saw the single largest residential electricity price increase of any state in the study period, nearly 40%, despite having comparatively modest data center growth relative to its size. The driver wasn’t AI infrastructure; it was wildfire-related grid hardening costs layered onto the state’s utilities. California is the clearest available evidence that when electricity bills spike, the explanation usually lies elsewhere, not in the presence of a data center down the road.
What This Means for the Policy Debate
Put the two case studies together with the paper’s national-level finding, and a consistent picture emerges: states adding the most data center capacity have not been the states seeing the steepest electric bills. If anything, the correlation runs the other way. The study also specifically tested and rejected the fear that data centers are quietly shifting costs onto residential customers while enjoying discounted industrial rates. The efficiency gains from new demand appear to flow to households, not away from them.
The researchers are careful to note real caveats, like turbine backlogs, transformer shortages, tariffs on solar equipment, and stalled permitting for new generation, which could eventually make new capacity more expensive to build, reversing the price-lowering effect if supply can’t keep pace with demand. That’s precisely why NetChoice has consistently called for smart, competitive state policy: modern permitting processes, interconnection reform, and tax codes built for large capital investment, rather than blanket bans that would only choke off the very grid investment the data shows is good for ratepayers.
Data centers are the backbone of America’s digital infrastructure and economic activity, and now there is rigorous economic evidence to back up the fact that growing, durable electricity demand tends to make the whole system cheaper to run, not more expensive. As more states debate legislation this year that hinders the data center buildout in America, policymakers should look past the fear-driven talking points and toward the data. The next time a headline blames your electric bill on the data center down the street, the receipts probably say otherwise. Lawmakers should let facts, not fear, drive decisions on the infrastructure powering America’s economic future.
Image via Unsplash.