TellWell
← Misinformation tracker
Partially FalseOther · Science

Did Data Centers Cause Rising U.S. Electricity Bills? The Claim Is Partially False.

Data centers caused the rise in US electricity bills

The argument in brief

U.S. electricity bills did rise roughly 22% between 2020 and 2023, but the primary cause was a natural gas price spike — not data center demand. According to FERC's 2022 Energy Market Assessment, Henry Hub natural gas prices hit a 14-year high of $6.45/MMBtu in 2022, which the EIA and FERC both identify as the dominant driver of the surge. Data centers account for only 4–6% of U.S. electricity consumption and have not been named by any major energy authority as a leading cause of recent bill increases.

The numbersU.S. Average Retail Electricity Price vs. Data Center Share of Consumption (2015–2023)

Data: EIA Electric Power Monthly 2024; LBNL Data Center Report 2024

Why it spread

The claim taps into genuine and growing anxiety about AI and Big Tech's resource consumption. When people receive a higher electricity bill at the same moment they are reading about massive new data centers being built, the mental connection feels intuitive and satisfying. Tech-skeptic media coverage amplified that intuition by pairing the two trends without interrogating whether one actually caused the other — a classic case of correlation being mistaken for causation in a story that already fits what many people want to believe.

The claim is that data centers are responsible for the rise in American electricity bills. The verdict is partially false. The premise — that bills rose — is accurate. The cause assigned to them is not supported by the evidence.

The most concrete rebuttal comes from the agencies that actually track electricity pricing. According to EIA's Electric Power Monthly 2024, average U.S. retail electricity prices climbed from 10.42 cents per kWh in 2020 to 12.72 cents per kWh in 2023. The EIA's own explanation for that 22% jump points to fuel costs, transmission and distribution infrastructure investment, and regulatory compliance — with demand growth from any single sector listed only as a secondary factor. FERC's Summer 2022 Energy Market Assessment is even more direct: the 2021–2022 price spike was driven predominantly by natural gas spot prices more than doubling, with Henry Hub averaging $6.45/MMBtu in 2022 versus $3.99 in 2021 — a 14-year high. EIA's historical natural gas price data confirms this was the sharpest fuel cost shock in over a decade.

The steelman version of the claim is not baseless. Data centers are real, large, and growing energy consumers. Lawrence Berkeley National Laboratory's 2024 United States Data Center Energy Usage Report estimates they consumed 200–250 TWh in 2023, or roughly 4–6% of total U.S. electricity. Goldman Sachs Research projected in 2024 that AI workloads could push that share to 8% by 2030 — a genuinely significant future pressure on grid costs. The concern about Big Tech's energy footprint is legitimate and well-documented.

But the claim breaks down on the question of timing and causation. Data center electricity consumption has grown steadily for years, yet retail electricity prices were essentially flat from 2015 through 2020 — hovering between 10.41 and 10.54 cents per kWh according to EIA data. Prices only spiked sharply in 2022, precisely when natural gas hit its 14-year peak. If data center load were the primary driver, prices would have risen in step with data center growth across the prior decade. They did not. The American Council for an Energy-Efficient Economy's 2023 Electricity Bill Trends analysis found no evidence that data center load growth was a primary cause of residential bill increases through 2023, attributing the rises instead to utility capital expenditures and fuel costs.

What is genuinely true: data centers do consume a meaningful share of U.S. electricity, that share is growing fast, and Goldman Sachs frames this as a forward-looking risk to grid costs. Conceding that point is important. But a future risk is not the same as the documented cause of a past price increase, and conflating the two is the core error in the claim.

The manipulation pattern here is post hoc reasoning dressed up with real statistics. Two trends — rising electricity bills and rising data center power demand — are presented side by side, and causation is implied without any mechanism connecting them. Watch for this structure whenever a claim pairs a cost increase with a politically convenient villain: check whether the proposed cause and the price spike actually align in time, and whether the agencies responsible for tracking that cost have named the same culprit.

Sources

TellWell AI

Related debunks