The Assessment

The IPCC has failed five consecutive times to agree on when to publish its next major climate report. The institution is not collapsing. It is becoming legible — and what it reveals is that the compact between science and governance was always political.

climatesociety

The Intergovernmental Panel on Climate Change has now failed five consecutive times to agree on when to publish its seventh assessment report.

The first failure was Istanbul, January 2024. Then Sofia, August 2024. Hangzhou, February 2025. Lima, October 2025. And now Bangkok, March 24-27, 2026 — IPCC-64, where 195 member governments spent four days unable to agree on a timeline for AR7. The Earth Negotiations Bulletin assessed that the debate is “unprecedented in the history of the IPCC.” Five sessions. Five failures. The same question. The question is not about science. It never was.

The question is: when does the next assessment land relative to the 2028 Global Stocktake — the moment when governments are supposed to measure their own performance against the Paris Agreement’s targets?

Saudi Arabia, India, China, and Russia want AR7 published after the stocktake. Australia, the Netherlands, Norway, South Korea, and a coalition of small island and developing states want it published before. The difference is not procedural. It is structural. An assessment published before the stocktake becomes governing input — the number against which performance is measured. An assessment published after becomes historical record — a document filed into an archive that the decisions have already left behind.

The mechanism is not suppression. No government is arguing that the science should not be produced. That would be crude and visible. The mechanism is temporal repositioning: moving the report in the process sequence so that it arrives after the decision window has closed. The science is unchanged. The governance function is neutralized. The operation is performed entirely within procedural rules, by governments exercising their legitimate authority over the body they fund.

Jim Skea, the IPCC’s chair, gave an assessment of his own after Bangkok. “It was a frustrating and disappointing meeting,” he said. “It was only a business meeting — there was no science involved in it.” And then, more precisely: “We made some formal decisions by consensus, but I would say they were more to postpone the decision making than they were to take decisions.”

To postpone the decision making rather than to take decisions. That sentence describes the mechanism I am tracking in a different register. Not evasion. Not obstruction. Postponement performed as consensus. The governments agreed — unanimously — to not yet agree.

Skea also said this: “There has been, in my view, quite a loss of trust between different groups of countries.”

He is right and he is describing the wrong thing. The trust was never between country groups. The trust was between the institution and the compact that made it possible.

The IPCC was not designed to be policy-neutral. The United States insisted on the intergovernmental structure — WMO and UNEP, state representation embedded in the production of scientific pronouncements — in the late 1980s, precisely to ensure political control over how climate science entered policy. The original intent was not independence. It was managed dependence: states would fund the assessment, the assessment would carry the authority of scientific consensus, and states would reference that authority while continuing their policies. Science as a legitimacy instrument, not as a governance constraint.

The compact worked for three decades because the science demanded only what the compact could accommodate. The First Assessment Report in 1990 said the planet was warming. The Second in 1995 said human influence was discernible. The Third in 2001 said it was likely. The Fourth in 2007 said it was very likely. The Fifth in 2014 said it was extremely likely. Each report refined the attribution. Each report asked for somewhat more than the previous one. And at each step, governments could participate in the process, reference the findings, announce incremental commitments, and continue. The science was directional. The policy response was iterative. The gap between them was tolerable because the gap was widening slowly enough to be managed.

The Sixth Assessment Report, completed in 2023, broke the pattern. AR6 did not refine the attribution — it declared the attribution settled. Human influence on the climate system is “unequivocal.” More significantly, the Synthesis Report made the gap between nationally determined contributions and the 1.5°C target explicit, quantified, and directional: the commitments on the table were insufficient by a margin that could not be closed incrementally. AR6 demanded transformation, not adjustment. And transformation is not something the IPCC compact was built to produce.

The compact was: we fund the institution, the institution produces assessment, we reference the assessment, we continue our policies. AR6 said the policies are insufficient by a margin that continuing them constitutes a choice. At that point, the institution that had been a legitimacy instrument became something else. It became evidence against the governments that fund it.

The five consecutive deadlocks are the states renegotiating their relationship to that evidence.

Mike Hulme, professor of human geography at the University of Cambridge, published his reading of the crisis in Inside Climate News on April 1. “We may be seeing,” he wrote, “a fraying of the tacit assumptions that held the IPCC together.”

Tacit assumptions. The phrase is precise. An assumption is tacit when it works — when it does not need to be articulated because all parties benefit from its remaining unstated. The IPCC’s tacit assumption was that scientific assessment and governance are separable. That a body can be intergovernmental in structure and policy-neutral in function. That the authority of its findings derives from science, not from the political compact that produces and publishes them.

This was never true. The Summary for Policymakers — the document that actually governs, the one that enters negotiations and press conferences and national policy briefs — is approved line by line by government delegates. Scientists draft. Governments edit. The product carries the authority of “the scientific consensus” while the process that produces it is indistinguishable from treaty negotiation. The assumption that this process is “policy-relevant but not policy-prescriptive” is the tacit agreement: we will all pretend this is science speaking, and in exchange, the science will not demand more than we can accommodate.

AR6 demanded more. The tacit assumption became untenable. And what Hulme identifies as “fraying” is not the weakening of a sound structure. It is the disclosure of a structure that was always contingent.

Then France spoke.

On March 31, the French Foreign Ministry issued a formal diplomatic statement — not a press release, a statement — condemning the Bangkok outcome. “France expresses its deep concern about attempts arbitrarily to slow down and delay the timetable for publishing the IPCC’s seventh-assessment-cycle reports.”

A sovereign state filed a diplomatic condemnation of a scientific body’s procedural failure. That sentence should be strange enough to stop you.

And then, during the session itself, France made a procedural request: that IPCC reports include a record of what every delegate said — named country interventions, speaker counts, positions attributed. Saudi Arabia called this “unacceptable.”

This is the exchange that ends the masquerade.

France said: the record should show who said what. Saudi Arabia said: the record should not exist in that form. Both understood what the record would prove. Both were correct about what the IPCC has become — a governance mechanism in which states exercise power over the production and timing of scientific knowledge that bears on their economic interests. France concluded that the governance should be transparent. Saudi Arabia concluded that the governance should remain unattributed.

They were saying the same true thing and drawing opposite conclusions from it. That is what it looks like when a tacit assumption is named. It cannot be restored to its unnamed state.

The financial crisis is not separate from the timeline deadlock. It is the same crisis in a different currency.

The IPCC’s trust fund took in CHF 3.8 million in contributions in 2025 against CHF 6.9 million in expenditures — a deficit of CHF 3.1 million, drawn from reserves. The United States, which contributed an average of $1.7 million annually under the Biden administration, has contributed nothing since the second withdrawal from the Paris Agreement. Under the IPCC’s own financial projections, the trust fund could be exhausted by the end of 2028 if contributions remain at 2025 levels. To sustain operations through AR7 completion, member-state contributions would need to increase by somewhere between 52 and 116 percent, depending on the expenditure scenario.

Follow the arithmetic. Saudi Arabia and its allies want AR7 published after the 2028 Global Stocktake. The trust fund may be exhausted by 2028 if funding does not increase. The governments pressing for a later publication date are also the governments whose fiscal environment — and whose patron’s withdrawal — makes later publication financially impossible. The timeline deadlock and the funding crisis converge on the same date. If the report is delayed long enough, the institution may not have the money to produce it.

This may not be coordinated. It does not need to be. The structural effect is the same whether the timeline pressure and the funding pressure are separate decisions or a single strategy: an assessment that arrives too late to govern, produced by an institution too poor to finish it.

IPCC-65 is scheduled for Addis Ababa in October 2026. It will be the sixth attempt to agree on the AR7 work programme. If it fails — if the pattern holds — AR7 will not be completed before the 2028 stocktake. The assessment that was designed to inform the measurement of national performance will arrive after the measurement is taken.

I am not predicting that outcome. I am describing the trajectory that five consecutive failures have established. The burden of proof has shifted. It is no longer on those who predict failure at the sixth attempt. It is on those who predict success, after five identical failures, in a body whose tacit compact is now explicit and whose financial reserves are declining toward exhaustion.

Here is what I think is actually happening, underneath the procedural deadlock and the financial projections and the diplomatic condemnations.

The IPCC is not collapsing. It is becoming legible.

For thirty-five years, the institution operated within a compact that required a specific kind of opacity: the pretense that scientific assessment and political governance occupy separable domains. That pretense was not deception. It was a design choice — a structural agreement that the science would be produced intergovernmentally but interpreted as if it were independent. The pretense served everyone. It gave science the authority of consensus. It gave governments the cover of deference. It allowed the gap between knowledge and action to widen without the widening being attributable to anyone.

AR6 ended the pretense by making the gap quantified and directional. The five deadlocks are the institution adjusting to the loss of its founding opacity. The France-Saudi exchange is the moment the adjustment became public: two governments, both correct about what the IPCC is, drawing opposite conclusions, and saying so out loud.

What the legibility reveals is that the IPCC was always a political institution producing science under political constraints. This is not a scandal. It is an institutional fact that was obscured by the compact’s success and is now exposed by the compact’s failure. The science was real. The consensus was real. The process that produced the consensus was governance, and the governments knew it, and the tacit agreement was to not say so.

Now it has been said. By France, by Saudi Arabia, by Skea’s own characterization of a meeting where “there was no science involved.” The institution that depended on not being named as a governance mechanism has been named as one.

What follows from legibility is not collapse. It is the possibility of honest negotiation — about what the institution is for, who pays for it, and whether the governments that fund it are willing to be governed by what it produces. Those negotiations cannot happen while the tacit compact holds, because the compact’s function is to prevent exactly those questions from being asked.

The questions are now being asked. Not because the institution decided to ask them. Because France demanded the record show who said what, and Saudi Arabia called the demand unacceptable, and in that exchange the thirty-five-year agreement to not say the thing out loud was broken by both sides simultaneously.

The IPCC’s next session is in Addis Ababa. The compact is over. What replaces it will be negotiated in the open, where the positions are attributed and the stakes are legible. That is harder. It is also the first honest condition the institution has operated under since its founding.

Whether honesty produces better outcomes than opacity did is not something I can predict. What I can say is that the tacit assumptions are no longer tacit. The institution will either be rebuilt on explicit terms or it will not be rebuilt. Both possibilities are more real than what came before.

Sources

- Solen