By JD DuRie, in collaboration with AI Matters
Every month, Eversource customers in Connecticut get a bill divided into four lines: Supply, Transmission, Local Delivery, and Public Benefits. Most of us glance at the total, wince, and pay it. Connecticut’s electricity prices rank among the highest in the continental United States, and with Eversource now asking regulators for its largest distribution increase in state history, the bill is about to become a political story again.
Before deciding what to do about it, it is worth reading the bill closely. Each line has a different cause, a different regulator, and a different fix. Treating them as one number is how bad policy gets made.
Four lines, four stories
Supply is the most volatile line. It is the cost of the electricity itself, reset every January 1 and July 1, and it rises and falls with the price of natural gas, which fuels most of New England’s power plants. In early 2023 the standard rate was about 24 cents per kilowatt-hour. Six months later it was under 14. It has since moved from 11.19 cents to 9.75, up to 12.64 last winter, and down to 11.58 today. Winter is almost always worse, because the region’s pipelines cannot carry enough gas to heat homes and run power plants at the same time.
Public Benefits had the most dramatic swing. It covers state programs: energy efficiency, renewable energy contracts, assistance for struggling customers, and the state’s long-term contracts with the Millstone and Seabrook nuclear plants. It spiked sharply in 2023 and 2024. Since May 1 of this year it has appeared on bills as a credit, thanks to state bond funds and nuclear contracts that are now paying off as gas prices run high.
Transmission is the quiet one, and the steady climber. These are the high-voltage lines that move power across New England, regulated not by Connecticut but by federal regulators. Annual transmission costs in Connecticut have grown 77 percent since 2015 and now make up roughly 15 to 16 percent of a typical bill.
Local Delivery is what comes next. These are the poles, wires, and substations on your street. Eversource has not had a distribution rate review since 2018. It has asked to raise the distribution portion of bills by roughly 35 percent, which would lift a typical residential bill about 13 percent starting in July 2027.
The rate case: mostly legitimate, with one exception
Much of Eversource’s request is the ordinary cost of a long gap between reviews. The company says it has put about $600 million of equipment into service that current rates do not yet cover, and that this accounts for roughly half of the request. Inflation has raised the price of everything a utility buys. Rates frozen since 2018 do not reflect any of it.
The exception is profit. Eversource is asking regulators to raise its allowed return on equity to 10.25 percent. The national average authorized for electric utilities this year is about 9.8 percent. Connecticut’s current allowance for Eversource is 9.25. The company says it actually earned only about 6.5 percent last year, and that is likely true, largely because of the gap between rate reviews. But a higher allowance does not guarantee a higher return, and an allowance above the national average, granted to a monopoly whose parent company earned $1.69 billion last year, is where ratepayers deserve the hardest scrutiny.
For what it is worth, executive pay is not the scandal it is sometimes made out to be. The CEO’s 2025 package of about $15 million sits somewhat above the industry average of $12.3 million and well below the industry’s top earners. It is a fair subject for debate. It is not what drives your bill.
Who pays: households versus businesses
Connecticut households pay an average of about 27 cents per kilowatt-hour. Commercial customers pay about 21. Industrial customers pay about 17. Regulators justify the difference with cost-of-service studies, the standard method in every state. Large customers take power at higher voltage, spread fixed costs over far more usage, and pay demand charges based on their peak. Households pay a flat rate that averages all of that together.
Is the split biased toward business? Partly, it reflects real differences in cost. But the studies themselves rest on judgment calls, especially how to assign the cost of the shared neighborhood wire system, and those choices can move hundreds of millions of dollars between classes. For scale: charging every customer the same average rate would cut household rates by about 13 percent, while raising commercial rates by about 12 percent and industrial rates by roughly 40 percent. That would drive some manufacturers away and push their share of fixed costs back onto everyone else. A uniform rate is not the answer. Transparency about the method is.
My personal agenda
The libertarian instinct is to distrust concentrated power, whether it sits in a state agency or a monopoly utility. The compassionate instinct is to make sure the people least able to absorb a high bill are not left holding it. Applied to the electric bill, those instincts point to five reforms.
1. Make every cost visible and voted on. Public policy costs belong in the state budget, where legislators must defend them on the record, not folded into a utility bill where no one votes on them. At a minimum, every mandated cost should appear on the bill by name, including the regional carbon program, RGGI, which is now buried in the supply price. Moving costs to the budget does not make them smaller. It makes them accountable.
2. Discipline the monopoly instead of pretending it is a market. The wires on your street are a natural monopoly. The answer is not deregulation but accountability. Connecticut mandated performance-based ratemaking in 2020, paying utilities for reliability and cost control rather than for how much capital they deploy. The new PURA commission should finish it. Regulators should hold allowed returns near the true cost of capital, and they should publish cost-of-service results under more than one allocation method, so the public can see how much the choice of method shifts onto households.
This matters for technology, too. Artificial intelligence can already target tree trimming where trees are actually likely to fail and let existing transmission lines safely carry more power on cool or windy days. But under traditional regulation, a utility earns its profit on what it builds, not on what it saves. Software that lets an existing line carry more power replaces a new line the utility would have earned a return on. Fix the incentive, and the innovation follows.
3. Open the supply. New England’s winter price spikes come from a gas bottleneck. Part of that bottleneck is a 1920 federal law. The Jones Act requires cargo moving between American ports to travel on American-built ships, and not one of the world’s LNG tankers qualifies. So New England cannot buy American natural gas by sea. It buys from Trinidad and the world market instead.
The Jones Act has been waived since March, and the waiver expires November 15. Tellingly, no LNG appears to have moved under it, because no supplier signs a winter contract around a 90-day waiver. Congress should replace stop-and-start waivers with a permanent exemption for cargoes no American ship can carry. It would cost the American shipbuilding industry nothing, because no American yard builds these ships.
Removing the barrier is not enough on its own. The region has one onshore LNG terminal, in Everett, MA controlled by a single owner. Real competition means open access at import terminals, more storage filled in summer when gas is cheap, and a serious look at a shared offshore terminal for southern New England. The last such proposal, Broadwater in Long Island Sound, was rejected by New York in 2008 with Connecticut officials cheering it on. A site farther offshore deserves a fresh hearing. Federal permitting reform and an end to tariffs on transformers, steel, and copper, both priorities of the Cato Institute, would lower the cost of every new line and pipeline as well.
4. Give households prices they can act on. Today’s meters force a single flat rate. The neighbor who runs the dishwasher at midnight pays the same as the one running central air at five o’clock on the hottest day of the year, and neither can save by changing habits. Smart meters, now approved for Eversource customers, could change that, but only with optional time-of-use rates and only if the rollout cost is kept in check. The evidence on savings is modest, a few percent for most households, but the larger payoff comes later: a grid that peaks less needs fewer new lines and plants. And because smart meters record what happens inside the home hour by hour, customers should own that data, with strict limits on its use and a real right to opt out.
5. Make compassion targeted, not hidden. Today the cost of unpaid bills and hardship programs is spread across every ratepayer, including affluent households that need no help. Assistance should be means-tested and funded openly, with room for voluntary channels like Operation Fuel. Blanket rate freezes, by contrast, mostly reward the heaviest users.
An honest ledger
It is fair to ask what all of this would save. Cato’s recent affordability handbook offers a useful federal and state list: ease permitting, end the cycle of regulatory reversals between administrations, rescind the federal greenhouse gas endangerment finding, eliminate tariffs on energy equipment, repeal resource-specific mandates and 100 percent clean energy goals, and expand competition. Cato attaches no dollar figures to any of them.
Using A.I. calculations, a rough estimate is that the quantifiable parts would save a typical Connecticut household something like $10 to $20 a month over time, mostly from repealing the renewable portfolio standard and RGGI. The endangerment finding matters little for Connecticut electric bills, because it governs vehicle emissions. Retail competition is already legal here, and in one Consumer Counsel report, 86 percent of Eversource customers using a third-party supplier paid more than the standard rate. Competition without consumer protection has not delivered for households.
That is real money, but it is not the whole answer. The larger drivers of Connecticut’s bills are the winter gas bottleneck, the pending distribution increase, storm costs, and a regulatory system that rewards spending over saving. Those are where the agenda above aims.
The case for the other side
Critics deserve a hearing. Climate advocates argue that repealing clean energy programs trades small bill savings for health and climate costs that families still pay, just not on the electric bill. Moving programs into the state budget exposes them to annual cuts that discourage long-term investment. Some state-arranged deals have worked: the nuclear contracts, once blamed for the 2023 spike, saved Connecticut ratepayers more than $250 million in 2025. Consumer advocates note that the market reform of 1998 promised savings for households that largely never came. And utilities warn that squeezing returns too hard leads to underinvestment, which shows up later as worse reliability.
Those points are serious. None of them argues against transparency, open supply, or paying utilities for results.
The bottom line
Connecticut’s electricity problem is not one problem. It is four, stacked on a single bill. The fixes that would do the most good share a common thread: they take power away from whoever is hiding the cost. Put policy costs where voters can see them. Pay the monopoly for performance, not for spending. Let gas reach New England from American wells. Give families prices they can respond to. And help those who need it openly, rather than through a charge on everyone’s bill.
The Eversource rate case is the next test. PURA’s decision is due by mid-2027. Readers who want a say can submit public comments in the docket. It is worth doing.
Figures are drawn from Eversource and PURA filings, the U.S. Energy Information Administration, Connecticut state agencies, S&P Global, the Energy and Policy Institute, Lawrence Berkeley National Laboratory, and the Cato Institute. Household estimates assume 700 kilowatt-hours a month and are approximations.
This post was written in collaboration with AI Matters, an editorial process combining human judgment with AI-assisted research and drafting.
