Electricity prices have gradually moved from a background role in daily life to a focal point of political debate. What used to be an inconspicuous household expense has now become a visible burden, symbolizing policy failures. The rise in electricity prices is not due to corporate greed or market failure, but rather the collision of regulation, politically driven investments, and top-down energy planning with the rapid development of Artificial Intelligence (AI). Inflation, supply constraints, and government directives have transformed the once stable and reliable power grid into an arena of intense economic, technological, and political conflict.
Americans have traditionally seen egg prices as a barometer of inflation. Now, a new impact is coming from electricity bills. Over the past year, electricity prices have risen by about 4.5%, nearly double the overall Consumer Price Index (CPI), primarily driven by the surging electricity demand from AI data centers and advanced manufacturing, while electricity supply remains limited.
“When demand increases and supply is limited, people will inevitably pay more,” said Calvin Butler, CEO of Exelon Corp., the largest nuclear power company in the United States. The company recently allocated $50 million to help low-income families pay their summer electricity bills. This impact is particularly evident in the largest U.S. grid operator, PJM Interconnection, where regulators estimate that the growth of data centers alone has added $9.3 billion to electricity costs.
Data confirms this pressure. The U.S. Energy Information Administration reports that by 2025, retail electricity prices in the U.S. are expected to be about 13% higher compared to 2022, with average monthly household electricity bills reaching $178. In Virginia, where data center density is highest, residential electricity and power delivery costs are projected to increase by up to 26% within this decade and potentially reach 41% in the next decade. In climate-targeted regions like ISO-New England, wholesale electricity prices have doubled since early 2024. Regulatory regimes aimed at stabilizing energy transition have paradoxically exacerbated market volatility.
For two decades, data centers have been small, nondescript, warehouse-like structures. The rise of generative AI has fundamentally changed this status quo. Training large language models requires massive computing power, transforming ordinary warehouses into massive data centers – consuming electricity on par with a mid-sized city and water in the millions of gallons. The scale of energy consumption has shifted the electricity issue from a technical matter to a political debate.
In Virginia and New Jersey, critical battlegrounds in this year’s gubernatorial elections, the political maneuvering over AI infrastructure has become a microcosm of national energy policy debates. Abigail Spanberger, the newly elected Democratic governor of Virginia, advocates for tech companies to pay a “fair share” for grid upgrades needed to operate. During the gubernatorial race, her Republican opponent Winsome Earle-Sears criticized clean energy policies for driving up costs and reliability risks. In New Jersey, some proposals seek data center developers to fund grid modernization, while Republican gubernatorial candidate Jack Ciattarelli calls for building more facilities and natural gas power plants to meet electricity demand. Public discontent over rising bills has blurred party lines: even local candidates from both parties are beginning to call for a pause on new data center construction.
States like New Jersey, which have prioritized offshore wind power development for years while allowing base load power and grid capacity to stagnate, as exemplified by the closure of the Oyster Creek Nuclear Generating Station in 2018 amidst growing demand, now face escalating and continuous electricity price hikes.
Today, the rapid growth of AI data centers clashes with these constraints: PJM predicts an additional peak load of about 32 gigawatts by 2030, with around 30 gigawatts coming from data centers, bringing about “price pressure” and concerns about resource adequacy. Recent reports show that consumers in the PJM region have borne billions of dollars in transmission costs due to data center growth.
Offshore wind power faces renewable energy certificate (OREC) costs above market levels, with regulators and consulting firms pointing out that these costs will be passed on to users. In regions where electricity transmission infrastructure upgrades severely lag, further exacerbating electricity price pressure. In essence, areas that prematurely retired nuclear power, lagged in grid construction actions, and heavily developed offshore wind power, are now inadequately prepared for the era of data centers, either scrambling to catch up or resorting to significantly raising electricity prices to fund the addition of generation capacity and grid upgrades.
Discontent is spreading. Republican Senator Josh Hawley of Missouri publicly criticizes these “monstrous energy consumers,” accusing Silicon Valley of pushing expensive electricity projects while having ordinary residential users subsidize the costs. A controversy has unfolded around the Georgia Public Service Commission, where data center operators enjoy a five-cent per kilowatt-hour electricity rate while ordinary households have to pay four times as much. These disputes reflect an economic truth: fixed industrial electricity contracts and tax breaks do not truly reduce costs, but simply shift costs onto ordinary consumers.
Electricity demand is not only unusually strong but also rising and unpredictable. The International Energy Agency forecasts that global data center electricity consumption will nearly double by 2030. However, in a truly competitive market, this demand would attract private capital for large-scale investments in building new generation and transmission facilities. Yet, the reality is that lengthy approval processes, domestic content requirements, and endless litigation severely inhibit supply.
Utility companies owned by investors plan to invest over $1 trillion in capital projects by 2029, mainly driven by regulation rather than market demand, with these costs inevitably being included in rate bases and reflected in monthly bills.
The political reshuffling around AI infrastructure reveals a deeper lesson: energy cannot be achieved through central planning without cost. States like Texas, which respond directly to price signals in a competitive market, maintain electricity prices about half of Massachusetts even with rapid demand growth, while ensuring stronger power supply reliability. In contrast, states relying on administrative schemes find themselves trapped in a cycle of higher costs and slower innovation.
Both parties keenly recognize the stakes. The Trump administration’s “AI Initiative” aims to expedite project approvals, expand grid capacity to ensure U.S. dominance in competition with China, slow down the retirement of fossil fuel power plants, and gradually phase out tax incentives for wind and solar energy. Democrats driving the “abundance agenda” praise data centers as the cornerstone of a digitized and decarbonized future. However, economic contradictions remain: Who should foot the bill for the electricity costs powering these machines?
Electricity is no longer a neutral input energy, but has become a commodity that drives economic growth and sparks social discontent. When markets are fully liberalized to adjust supply and demand freely, they can efficiently meet the growing electricity needs. Unfortunately, cumbersome government interventions, subsidy competitions, and political appeals lead to shortages in supply, which, in turn, are exploited for political gains. The current solutions – increased regulation measures, subsidies, and planned economies – are, in themselves, detrimental.
A truly market-based approach should involve letting prices reflect resource scarcity, opening up competition in generation and retail supply markets, and canceling cross-subsidies that mask the true costs. Decarbonization and innovation can coexist with affordability, but only if capital flows are allowed towards places where investments yield sufficient returns to justify the risks undertaken.
When Americans open their electricity bills, they find – in some cases rediscover – that electricity is ultimately a market commodity, not a political entitlement, and that efforts to lower costs through regulation will only delay and amplify these costs.