Electricity Affordability — Trump’s Achilles’ Heel?
The economy under President Trump’s leadership has performed brilliantly: economic growth and productivity are up, inflation is subdued, and unemployment is low.
But the fact is, Americans are paying more for electricity these days, and the president — as he promised in the runup to 2024 — needs to show he’s doing something about it. This concern of affordability will play a pivotal role in the upcoming mid-term elections, just 9 months from now. The outcome of the midterms will determine which party controls the House and Senate and, hence, the ability of Donald Trump to continue his domestic economic policies. (RELATED: Drill, Baby, Drill — But What About the Electrical Grid?)
A new poll has found that more Americans think Democrats (37 percent), as opposed to Republicans (25 percent), are the party most committed to reducing energy prices.
Enter Artificial Intelligence. It is no surprise that AI is rapidly and significantly being integrated into the U.S. economy. To support this burgeoning investment in AI, tech companies are increasingly developing data centers that draw enormous power from America’s aging electrical grid, thereby pushing up the cost of electricity to utility customers. (RELATED: Data Centers Are an Easy Scapegoat for Failing Electrical Grids)
The Trump administration is exploring measures to reduce the impact on consumers by requiring technology companies building data centers to absorb all associated utility and infrastructure costs as electricity prices continue to rise across the US. (RELATED: On Venezuelan Oil Tankers, Chinese AI, and American Energy Traitors)
The discussion around data center costs comes as electricity prices have increased significantly, rising 6.9 percent (double the rate of inflation) year over year in 2025, with no signs of remitting. Through the end of the decade, data centers will constitute 40 percent of electricity demand growth.
In January, several states joined the White House in signing an agreement urging PJM Interconnection, the nation’s largest grid operator, to require major technology companies to finance $15 billion worth of new power plants. PJM manages the electrical grid in regions with high concentrations of data centers, including northern Virginia and New Jersey.
But here’s the thing — electricity prices have been rising steadily for more than a decade, and this is before AI meaningfully entered the picture. Putting a cap on AI-generated energy costs will not, by itself, solve President Trump’s electricity affordability problem with America’s electorate.
Since 2021, the average price of electricity across the U.S. has risen by almost 30 percent, with no signs of slowing down. In 2025, electric utilities across the nation requested rate increases totaling $71.2 billion through 2028.
As a result, the public has been pointing the finger at everything from the rise in new AI data centers, grid investments, to growth in “green” energy.
But electricity pricing is too complicated to blame just one thing. A recent study revealed that electricity price changes across the U.S. in recent years were driven by a combination of factors, including location, infrastructure impacts from extreme weather conditions, grid upgrades needed for modernization and resilience, and volatility of fossil-fuel costs.
There’s no single explanation for U.S. price trends across all types of customers in all contexts.
The reason for this lies in a combination of structural pressures that are now threatening energy affordability.
Demand Is Outpacing Supply
In some locations, higher prices can be explained by the law of supply and demand. More electricity is needed than is currently available. But Rome was not built in a day — and neither are new power generation and delivery systems.
In PJM, the regional transmission organization serving the Mid-Atlantic and parts of the Midwest, tight supply and growing demand have contributed to record prices in the auction for future supply. According to PJM’s independent market monitor, new AI data centers are largely responsible for this trend, driving a $7.3 billion, or 82 percent increase, in auction revenues in 2025 alone.
Unless President Trump is successful in his efforts to have “Big Tech” absorb the costs from new AI data centers and their increased energy demand from America’s power grid, that burden will ultimately be paid by utility customers, and the latter are already seeing higher costs on their bills from last year’s capacity auction.
In the face of constrained supply, one way for the president to quickly access affordable sources of energy is by removing barriers that limit demand-side resources. Inconsistent rules and inadequate compensation structures hinder the potential of solutions that address the problem by reducing and managing demand rather than building more supply immediately — such as more flexibility at data centers and grid-enhancing technologies. Longer term, the U.S. will need to build new supply and the infrastructure to support it, which means streamlining permitting and reviewing timelines that can take years to navigate.
Aging Infrastructure and Deferred Investment
The electric grid is the largest and most important “system” in the country, and much of it is old. For decades, electricity demand was flat. Utilities and regulators responded rationally by deferring major investments. That kept rates low, but it also meant infrastructure aged in place.
Transmission lines, substations, generation assets, and gas-electric coordination systems all require large capital outlays at the same time. Natural gas supplies roughly 43 percent of U.S. electricity generation; yet, the infrastructure needed to move gas reliably to power plants has not expanded at the same pace as dependence on it.
We postponed investment during decades of flat demand and are now paying the bill when demand is rising again.
Most of the U.S. electric grid was built in the 1950s and 1970s and is now approaching the end of its expected 50- to 80-year lifespan. At the distribution level, utilities report that roughly 28 percent of their spending on local power systems is driven by a need to replace old infrastructure.
Electricity Is Cheap, Delivering It Is Not
Over the past 10 to 15 years, the cost of generating electricity has fallen steadily. We have become remarkably proficient at producing “cheap electrons.” But moving them over long distances is another issue. In the future, the electricity we consume may be relatively cheap, while the systems that make it available account for a growing share of our utility bill.
Although President Trump is making prudent decisions, it would be too simple to blame AI data centers for the current electricity price hikes. Rising electricity prices are not a sign of sudden market failure. They reflect a system being asked to do more, more reliably, and under harsher conditions, after decades of “kicking the can down the road.” The real question is not whether prices will rise, but whether the president can (over the next 9 months) convince Americans that he is taking concrete steps to fix a problem not of his own making.
READ MORE from F. Andrew Wolf Jr.:
The Cost of Harvard’s Intransigence