Bernard McNamee is a former commissioner on the Federal Energy Regulatory Commission, former deputy general counsel for energy policy at the U.S. Department of Energy, and served four state attorneys general in Virginia and Texas.
Growing affordability concerns often take aim at electric bills. Though the general narrative is that electric rates are rising quickly because of data center demand in support of artificial intelligence, a study by Charles River Associates shows that residential electric rate increases have primarily occurred in California, the Northeast and part of the Mid-Atlantic.
This is not to say that all Americans are not feeling the pinch of inflation-driven higher prices on all goods and services, including electricity; but we need to look at the data before making sweeping accusations and proposing new policies.
Using U.S. government data, the CRA study compared the retail electric rates paid by residential customers in each of the 48 contiguous states over a five-year period ending in October 2025. The data show that individuals and families in most states have not seen dramatic increases in their electric rates. In fact, 34 states saw rate increases that were less than the national average. This means that significant rate increases in just 14 states are driving much of the narrative about rapidly rising electric bills. Just as important, the study also points out some of causes for these increases.
Californians saw the biggest increase in electric rates over the past five years, at 11.8 cents/kWh. But it has not been data centers or utility profits that drive California’s rate increases; rather, customers are paying for wildfire expenses and the cost-shifting effects of rooftop solar rules imposed by California policy makers.
New York and various New England states also have experienced higher-than-average rate increases, ranging from 4.5 cents/kWh to 10.4 cents/kWh. A substantial portion of these increases can be attributed to utilities no longer owning their own power plants for the benefit of customers; instead, they rely on purchased power from an independent system operator or regional transmission organization.
Unlike in California and the Northeast, the CRA Study determined that data centers are one of the causes of increased electricity prices in the mid-Atlantic states, a region served by PJM Interconnection. This conclusion was supported by PJM’s independent market monitor in a report issued March 12.
Moreover, PJM’s market monitor reported that wholesale power prices increased 54% between 2024 and 2025, from $43.5 billion to $67 billion. But according to the CRA Study, the rate impacts for customers differed among the states. For example, Maryland and Pennsylvania saw electricity rates increase by 6.2 cents/kWh and 5.2 cents/kWh, respectively. However, other PJM states have rates below the national average. The CRA Study identifies one of the reasons for the difference.
Like utilities in New England, utilities in Maryland and Pennsylvania are not allowed to build their own power plants for the benefit of their customers; instead, those utilities must buy their power using PJM’s marginal price auctions. This means that such utilities can’t use the cheapest power plants to serve their customers; instead, they must pay the top price for all the power that clears the auction, even if some power plants are cheaper than other power plants. Unfortunately, these marginal price auctions expose customers to bigger bills. For instance, the past two capacity auctions in PJM saw payments increase by 800% and then an additional 22%.
Counter to some narratives, the local utilities do not earn any profit from the power they purchase from PJM to serve customers. The purchased power costs are passed through to customers with no mark-up. Now, some have been arguing that utilities should not construct or own power plants because the utility can earn a return on their capital investments; but these arguments ignore the fact that utilities that build their own generation can often provide power more cheaply than they can by buying power from the “market.” In fact, utility-owned generation serves as a hedge against rising capacity and energy costs.
Furthermore, RTOs like PJM are not building the new power plants needed to serve all customers. In January, a NERC grid reliability report exposed how much of the country, particularly RTO regions, are facing “elevated” or “high risks” of power shortages. In contrast, the Southeast, a region without an RTO and where utilities can build power plants, was not at risk of blackouts. Moreover, as the CRA study shows, most non-RTO states have also seen residential utility electricity rates rise more slowly than the national average.
As electricity bills rise and reliability is threatened, elected officials are expressing frustration with the lack of accountability or control. The vertically integrated utility model used in the Southeast and some of the lower cost states in PJM demonstrates that it is possible to build new generation, enhance reliability and protect customers from unmitigated rate increases. Building the electric grid to support AI dominance, local economies and families can be achieved, while also limiting impacts to customers and their electric bills.
