Electricity is getting more expensive, and across much of the country the fastest-growing part of the bill is not the part most people would guess. It is not the cost of generating the power, but a separate charge, one your utility passes on to you, that pays power plants and other resources to promise they will be available years from now. The payment buys their presence on the hottest summer day or coldest winter night, when the grid is under the greatest strain. That promise has a price, and in one large region it has climbed to a level no one had seen before.
A nonprofit grid operator called PJM coordinates the flow of electricity for nearly 70 million people across 13 states, from New Jersey to Illinois. Each year it runs an auction to secure enough future capacity to cover expected demand. In its most recent auction, held late last year, the price cleared at a record, and for the first time in PJM's history the market still fell short of securing all the standby power the grid is meant to keep. A record price that still leaves the grid short is a strange result. That puzzle is the heart of this story.
An Auction That Sells Promises
Like any market, this one has buyers, sellers, and an overseer. The sellers are generators, the companies that own the power plants. The buyers are utilities, which need that power for the homes and businesses they serve. The overseer is PJM, a nonprofit that runs the auction and does the buying on the utilities' behalf: it owns no power plants, sells nothing, and sets no policy. Its job is to make sure enough power is lined up for the years ahead.
Here is what that auction is actually for. It does not sell electricity. Generators and other qualified resources offer commitments to be ready on the worst day, the January cold snap or the August heat wave when every furnace fan and air conditioner pulls at once. The auction matches those offers against what the grid will need, accepting enough to cover the expected peak, plus a cushion.
It is, in effect, insurance against the worst day.
One feature sets this apart from the auction most people picture. Only one side bids. The generators compete to be chosen, but no buyer haggles back. Instead, PJM decides in advance how much capacity the grid should buy at each price, then buys on the utilities' behalf according to that fixed plan.
Here is the part worth slowing down for. The auction does not set the price of power. It discovers it, the way any auction does, by finding the level at which the supply offered is enough to cover what the grid needs. When power is plentiful, the clearing price is low and barely makes the news. When power is scarce, the price climbs, and that climb is not the disease but the thermometer.
Scarcity Drove the Price Up
The cause is not greed but arithmetic: demand is arriving faster than supply can be built. After years of flat forecasts, PJM's projected peak demand has surged, driven mostly by data centers, which run on banks of servers that draw power on a scale the grid was never built for. The growth now expected in a few short years rivals what the region once spread across an entire decade. Supply cannot answer that fast, because a new power plant takes years to permit, finance, and build, and the transmission lines to carry its power take longer still.
The auction read that mismatch and returned a high price, exactly what scarcity produces. It saw a region where the demand for reliable capacity was outrunning the supply available to meet it, and it priced what it saw. The message was simple: power in this region is scarce, and whoever can build more of it will be paid well for doing so. Yet even at that record price, not enough supply came forward to meet the target.
Picture a town hiring lifeguards for a dangerous beach. It raises the wage, then raises it again, to the highest it has ever paid, and still too few show up. A record wage does not mean the town is being greedy or foolish. It means the shortage is real, and that no wage fills it overnight.
The Cap Held It Down
Scarcity is not the whole story. One more thing made the auction fall short: a price cap. To shield households from rising bills, a legal challenge had led to a settlement, approved by federal regulators, that limited how high the auction price could rise. It did lower bills, but it also muffled a signal. When a price cannot rise, sellers offer less than they otherwise would, and without the cap the clearing price would have been about 60 percent higher, enough to coax out more supply. Because that signal was blocked, the auction ended up roughly 6,600 megawatts short of the reserve the grid is meant to keep, a gap close to the peak power demand of a city the size of Philadelphia.
In April, federal regulators extended that cap through 2030, covering the next two auctions. The decision drew rare unanimity, backed by all 13 state governors and the White House, and its aim is the same as before: hold down the bill. But it also locks in the muffling for years, just as the shortage is set to deepen.
None of this means the auction is beyond debate. It is a designed market, and reasonable people argue about every one of its settings, from the size of the cushion it aims to secure to the way each power plant is scored. That argument is worth having. But even an imperfect market can reveal scarcity, and hiding the signal does not eliminate the shortage.
Build More, Don't Hide the Price
The grid does not need its prices muffled. It needs more power plants and transmission in a few hungry regions, and a price signal honest enough to call them forth.
The grid will not go dark tomorrow. PJM has a managed sequence of stopgaps, from follow-up auctions to paying aging plants to stay open past their planned retirement. But these are inefficient fixes, and the cap made them work harder than they should. Letting the price rise would not have conjured power overnight, yet it would have summoned more of the supply that does exist, leaving a smaller gap for the patches to cover.
Watch the Shortfall, Not the Price
The next auction closes on July 7, the first to run under the extended cap. But the clearing price will reveal little, because the ceiling has already done much of the pricing. The final number will sit near the cap, shaped as much by policy as by actual scarcity.
Instead, watch the measure the cap cannot conceal: whether the auction secures the reserve margin the grid is supposed to maintain, or falls short again as it did last year. If the market once again fails to procure enough capacity while the price remains pinned to the ceiling, policymakers should take the hint. The cap is no longer containing costs; it is suppressing the signal. At that point, it should be raised or removed.
Because the shortage is real, and it is growing.
A hidden price merely postpones the day it must be confronted.
