Ray Gifford is former chair of the Colorado Public Utilities Commission and managing partner of Wilkinson Barker Knauer LLP’s Denver office. Matt Larson is a partner in WBK’s Denver office.
For an industry where investments are measured in decades, utility policy sure is susceptible to short-term trends.
“Affordability” has become the watchword for utility rates and a weaponized term for policy interests seeking desired outcomes. Persistently high prices in the PJM capacity auction, which have failed to induce enough capacity in the market, have governors across the PJM region clamoring for reform. Off-year gubernatorial elections in New Jersey and Virginia made utility rate freezes and data center-driven cost increases major issues as part of an “affordability” agenda.
When public anxiety turns to utility rates, politicians and regulators follow.
At the same time, demand for electricity is growing considerably. Data centers, large industrial users and electrification more broadly are driving demand to levels utilities have not seen in decades.
In theory, demand growth should be good news for everyone. Demand growth spreads fixed costs, supports capital investment and — if managed correctly — lowers per-unit costs for ordinary consumers.
Yet the politics of electricity pricing are increasingly shaped by the fear that growth will come at the expense of households and to the benefit of data center developers.
That fear is now front and center. In New Jersey’s recent gubernatorial race, calls for an electricity rate freeze became a defining political signal. Whatever one thinks of rate freezes as policy, they reflect a hard political reality: Governors understand that energy affordability is a kitchen-table issue and they will act preemptively if regulators and utilities appear unable to manage it.
The lesson for the industry is not that growth must stop, but that growth must be disciplined and clearly beneficial to existing customers.
Examples of how to do this well are emerging. Georgia Power announced it would expand its current generation capacity by nearly 50%, adding up to 10 GW of new generation. In order to win regulator approval, the utility promised to file its next base rate case in 2028 “in a manner that will ensure incremental revenue from large-load customers will provide benefits” of approximately $102 per year for the typical residential customer.
A planning model where a large load is deliberately integrated into the system can work for all customers. The utility is pairing new demand with new generation and transmission investment, ensuring reliability while expanding the rate base in a way that dilutes fixed costs — hence the savings. That is the key point often missed in public debate: When large customers pay their way, they can help with affordability rather than undermine it.
The same pattern is visible in less obvious places. North Dakota, Arizona and New Mexico — hardly caricatures of hyperscale excess — have all seen data-center announcements tied to local generation, long-term contracts and infrastructure investment. These projects are not speculative bets dumped onto captive ratepayers. They are structured to bring jobs, tax base and incremental demand that support system economics. In states long accustomed to flat or declining load, that matters.
Contrast this with the alternative. When regulators delay investment, cap returns or send mixed signals about whether utilities are allowed to build, they do not stop demand from materializing. They simply push it elsewhere or force utilities into reactive, high-cost procurement. That is how affordability crises are born — not from growth itself, but from growth unmanaged.
“Big Tech” is showing it can meet the political moment, too. In early January, Microsoft President Brad Smith announced its “community-first AI infrastructure plan.” Microsoft promised to pay its own way, limit water use and add to the local community and tax base.
Microsoft is the first — but surely not the last — data center developer to address communities’ concerns forthrightly. Managed and regulated properly, the AI-driven electric demand can be a net benefit to the country, to the communities where data centers locate and to other customers on a utility’s system.
The political warning lights are flashing. Rate-freeze rhetoric is not an aberration; it is a response to perceived failure.
The power sector needs to show — clearly and repeatedly — that growth can lower bills, not raise them. The industry’s task, then, is not to resist data centers or large loads, but to integrate them transparently. That means clear cost allocation, upfront commitments from large customers and regulatory frameworks that reward timely investment.
It also means regulators being honest with the public that affordability is sustained by building, not by pretending demand will disappear, and by appropriately supporting the investments needed to meet growth.
Energy abundance, competitiveness in artificial intelligence and consumer affordability are not in conflict. They are mutually reinforcing when policy gets the sequencing right. Data center growth done badly socializes risk and privatizes reward. Done well, it expands the system in a way that benefits everyone.
