# America Doesn't Have an Electricity Shortage. It Has a Permission Shortage.

*Op-ed, ~1,100 words*

America has more proposed generation and storage waiting for interconnection than the country currently operates. That does not mean all of those projects are viable. It does mean the process for deciding which projects can connect has become a major constraint. Roughly 2,061 gigawatts, against about 1,400 gigawatts in service, sits in interconnection queues, and 549 gigawatts of it already holds a signed or draft connection agreement, according to Lawrence Berkeley National Laboratory. The typical project waits more than five years for an answer. Historically, only about one-eighth of queued capacity has reached operation.

Hold that fact next to the others in the news. Data-center campuses that would pay a premium for electricity today cannot buy it, and gas turbine order books now quote deliveries near five years out. The capacity charges embedded in 67 million people's bills across the PJM region rose from $2.2 billion to more than $16 billion a year across four years of auctions, each of the last four hitting a price cap, one of them negotiated directly with a governor. Household electricity bills have become a campaign issue from Harrisburg to Richmond.

The problem is not simply a shortage of generation technology or capital. It is the growing gap between what the system can produce and how quickly institutions can connect and coordinate it. Electrons have never been cheaper to produce; utility-scale solar's cost fell 84 percent since 2009 and battery packs fell 93 percent since 2010. The cheapest electricity ever generated is arriving alongside bills up a third since 2019. American electricity prices have outpaced inflation four times: the oil shocks of the 1970s, the nuclear overruns of the early 1980s, the natural gas run-up of the mid-2000s, and now. The first three had a cost story. This one does not, because what is scarce is permission: permission to connect to the grid, permission to compete on it, permission to be paid what a resource is worth. The bottleneck moved from the power plant to the paperwork, and households are paying the toll: supply that cannot connect, reliability charges socialized from data centers, and delivery spending that utilities are paid to maximize.

The institutions are behaving exactly as they were built to. A regulated utility earns a return on capital it deploys, a formula set in 1907, so it prefers a $40 million substation it profits from for forty years over a $2 million a year flexibility contract it profits from never. The two are not the same asset, and the regulatory question is whether the service a feeder actually needs can be procured for less than the full infrastructure solution. Economists have understood the bias since 1962. It explains why utilities propose expensive new transmission lines when re-wiring existing ones with advanced conductors could double their capacity at a fraction of the cost, a finding published in the Proceedings of the National Academy of Sciences. It explains why interconnection stays slow: the entities running the queue bear no cost for delay. And it explains why the fastest fixes, batteries and flexible loads sitting exactly where the grid is strained, are paid under state programs rewritten every few years.

Run the arithmetic for a single 500-megawatt data center. Building the solar and batteries takes 12 to 18 months. A new gas turbine arrives around 2031. A study-first interconnection queue delivers grid access around 2031 too. Texas-style rules deliver it around 2028, and a flexibility deal on the existing grid serves it in months. The engineering takes a year and a half; much of the remaining delay is institutional rather than equipment-related.

There is a real-world comparison, and there are receipts. Texas connects new generation several times faster than the rest of the country by using a simple rule: connect quickly, manage congestion afterward through the market. In Brooklyn and Queens, Con Edison deferred roughly a billion dollars of substation and grid investment with about $200 million of batteries, efficiency, and demand reduction, once New York regulators built a mechanism letting the utility profit from the cheaper choice. Britain pays its grid companies for measured performance instead of capital spent, and its networks stayed profitable and investment-grade, earning about 9 percent on equity against a 5-to-6 percent national average. In South Australia, real-time connection limits are now the default standard for new rooftop solar. Hawaii rewrote utility ratemaking in 2020. Every major component has a precedent somewhere; none has yet been assembled at US national scale. No core state reform requires new federal legislation.

The fix has five parts. Pay utilities for outcomes, so a battery contract and a substation compete on cost. Move the grid's referee functions, interconnection, data, operating envelopes, out of the companies that compete on the field. Price the grid's real scarcity by time and place, with protections for households that want flat bills. Make connection a rule rather than a queue, wherever technically feasible, with fast-track access for big loads that agree to power down during the few hours a year the grid is tight; Google already signed such agreements with utilities in Indiana and Tennessee. And stage a transition toward obligation-based reliability procurement: new large loads should carry the incremental reliability obligations their connection creates, so a data center's growth lands on the data center's books instead of on every household in thirteen states.

The skeptics deserve straight answers. Yes, dynamic pricing must ship with hedged options for anyone who wants a predictable bill. Yes, utilities face real burdens, wildfires, storms, cyberattacks, and reform must let well-run companies earn more, which the British model demonstrably does. And yes, the transition must honor every contract already signed; the reforms of the 1990s showed how.

But the largest fact in American energy is the one hiding in the queue. One of the country's largest pools of prospective infrastructure is already in development or awaiting connection. Unclogging it would do more for electricity bills, for AI and manufacturing competitiveness, and for the climate than any subsidy now being debated. The reforms require implementation work, but their principal tools are tariffs, regulation, and market design, not federal appropriations.

Regulators wrote the rules that made connection scarce. Regulators can rewrite them, starting with two federal rulemakings and a handful of governors in 2027. The grid's institutions were rebuilt twice before, in 1907 and the 1990s, each time only after visible, expensive failure, each time over incumbent objection, and each time the objectors ended up defending the result. The failure is visible now. The electric bills are the evidence.