Summary

  • Larry Elliott attributes the global economy’s better-than-expected performance through wars, energy shocks, and inflation to three named supports — AI investment, drawdowns of oil inventories, and Chinese export pricing — and reports each support as already showing strain.
  • Elliott describes the WTO as moribund and reports the IMF and World Bank as reluctant to confront the Trump administration’s “erratic behaviour,” against Richard Kozul-Wright’s argument that those institutions have departed from the New Deal principles on which they were founded in 1944.
  • On Elliott’s reporting, the IMF conditionality framework frontloads adjustment onto borrowing countries, defining success narrowly as avoidance of repeat programming rather than improvement in living standards.
  • Elliott’s closing judgment names the present moment a crisis postponed rather than avoided, contingent on which named support gives way first.

Finance ministers and central bank governors met in Bangkok for the International Monetary Fund’s annual gathering at a moment that has, on Larry Elliott’s reading, refused to become a crisis — but whose refusal looks contingent. Elliott, a Guardian columnist writing from the meeting, attributes the global economy’s better-than-expected performance through two regional wars, a serious oil shock, and rising inflation to three named supports: massive investment in artificial intelligence, drawdowns of stored crude, and China’s strategy of exporting cheap goods abroad. Each, he reports, shows strain. The IMF and World Bank, on his reporting, remain reluctant to confront what he calls the Trump administration’s “erratic behaviour,” and the World Trade Organization he describes as “moribund in this new age of protectionism.” Richard Kozul-Wright, a former UN official, argues in a paper due for publication next month that the Bretton Woods institutions have lost their connection to the New Deal principles on which they were founded in 1944, an institutional critique Elliott reports without endorsing the prospect of restoration. Elliott’s judgment: “This looks more like a crisis postponed than a crisis permanently avoided” — “a lull before the storm, in all probability.”

Three supports, three strain indicators

Elliott offers three reasons growth has remained robust and inflation muted through a period in which oil shocks of the magnitude recorded since the start of the Iran war have historically had a “marked and immediate impact.” The first is what he calls “the massive amount of investment in the development of artificial intelligence,” which he describes as “particularly important in the United States, comfortably the fastest-growing member of the Group of Seven year on year.” The second is the drawdown of stored crude: “had stockpiles not been run down,” Elliott writes, “oil prices would have risen even more sharply.” The third is China’s economic strategy of “exporting cheap goods to the rest of the world, bearing down on import prices elsewhere.”

Each of those supports, on the same column’s reporting, exhibits a strain indicator. The AI investment that has propelled US growth has, Elliott reports, “pushed up the cost of borrowing and resulted in shortages of capital for non-AI sectors,” a pressure that compounds with the deficit-reduction obligation Elliott attaches to governments: “Governments are under pressure to rein in spending or raise taxes in the hope that bond markets will be impressed by their efforts to cut budget deficits.” The inventory drawdown that has absorbed the energy shock is finite — “Oil stocks are running low,” Elliott writes — and the supply disruption the drawdown cushions is widening, with “the conflict spreading to Saudi Arabia.” The Chinese export pricing that has kept import prices down is producing political friction at the receiving end: cheap Chinese goods “threaten manufacturing jobs in the West, with Germany especially vulnerable,” the location Elliott singles out as where the strain would most likely bite.

The three supports are not independent in their fragility. Elliott reads each as running into a constraint at the same time — capital allocation tightening against non-AI sectors, oil stocks running out against a widening conflict, Chinese export pressure registering against manufacturing employment in a vulnerable economy. Whether the cycle turns, whether inventories are exhausted, or whether a counter-response from a vulnerable economy materializes — Elliott names Germany as the probable site — is, on the evidence he offers, the question that decides whether the moment remains a lull.

The multilateral architecture Elliott describes as reluctant or moribund

The frame Elliott’s column applies to the present institutional position of the multilateral order is one of accumulation rather than rupture. The WTO is “moribund in this new age of protectionism.” The IMF and World Bank are “reluctant to call Trump out for his erratic behaviour.” None of the three institutions, on his reporting, currently provides the multilateral cooperation that “is needed to cope with the strains” of an in-retreat globalisation. The dysfunction, he writes, “ran far deeper than that”: “the liberal world order the IMF and World Bank have presided over since the end of the second world war was dysfunctional long before Trump arrived in office.”

That sentence locates Elliott’s reading of the current moment against a longer institutional history. Richard Kozul-Wright, a former UN official, makes the case in a paper due for publication next month. The IMF and World Bank, “created at the 1944 Bretton Woods conference,” Kozul-Wright argues, “may still be operational” but “have lost their connection to the New Deal principles laid out 82 years ago — state intervention in the running of economies, active industrial policy, the pursuit of full employment and help for poor countries to develop.” What they have replaced those with, on his account, is a different agenda: “an agenda which encourages footloose capital, bolsters (and protects) private profit, and reduces the policy space available to governments.” The result, in Kozul-Wright’s words, is that the IMF and World Bank “have acted as hand maidens to an age of rising inequality, debt distress and heightened anxiety.”

Elliott illustrates that argument about policy space with the IMF’s own lending programmes. Help comes with strings attached: an insistence on tax increases and spending cuts that are frontloaded — “half the measures” are implemented, he reports, “in the first year and two-thirds within two years.” The Fund, Elliott reports, argues this “tough love” has a higher chance of success, although its view of what constitutes success is limited and primarily means “that a borrowing country does not need another IMF programme in the next three years” — “not that living standards are improving or that fewer people are in poverty.”

Reading the institutional situation

Several interpretive traditions converge on the same substrate-attributed description — three supports with named strain indicators, multilateral institutions whose reliability he describes as reluctant or moribund, a contingent postponement rather than avoidance of crisis — and supply vocabulary for it. The architectural reading associated with Christian Norberg-Schulz treats a place as a synthesis of natural and man-made conditions that gives it its identity. On this reading the natural place is the post-war reconstruction landscape and the bipolar-to-unipolar geopolitical transition it grew into; the man-made place is the institutional architecture constructed at Bretton Woods in 1944. The orientation function of the original order — embedded liberalism, state action bounded by multilateral discipline — has, on this reading, been replaced by something the original designers would not recognize. Kevin Lynch’s concept of imageability distinguishes places whose elements invite cognition from places whose elements do not. On this reading the orientation landscape has lost several of its landmarks: the WTO is “moribund” by Elliott’s account; the IMF and World Bank, on Kozul-Wright’s account, no longer point to the cardinal directions they once did.

Jay Appleton’s prospect-refuge theory reads resilience as prospect — from positions of relative safety (stable oil inventories, continued Chinese export supply, sustained AI investment), observers can see across the order’s terrain. Refuge is unevenly distributed, surfacing most clearly in the IMF conditionality framework as exposure without shelter: the borrowing country is placed in a position of prospect (visibility to the lender) without corresponding refuge (policy space for its own adjustment path). Christopher Alexander’s pattern language, applied to the institutional ecology, identifies patterns of activity nodes (the IMF/WB/WTO cluster once concentrated the multilateral activity of states), an intimacy gradient violated when one institutional posture is imposed uniformly across structurally different borrowers, and a floor-surface pattern describing the institutional floor as having flattened. James Reason’s Swiss-cheese model treats defensive layers as each having holes; on Elliott’s reporting the holes in IMF lending capacity, WTO dispute-settlement, World Bank development programming, central-bank cooperation, and diversified energy supply have aligned enough to push the system toward crisis without yet reaching it. Charles Perrow’s analysis of complex tightly-coupled systems describes the order as in a regime where interactions exceed any single operator’s mental model, and no operator exists in any case (the relevant actor is a system of states and institutions, not a single controller). The IMF’s “tough love” frontloading — “half the measures” in the first year and two-thirds within two years — describes the tight coupling between lender demands and borrower response.

The negotiation space

Read through Fisher and Ury’s framework of principled negotiation, the multilateral order’s reform dispute has been conducted largely in positional terms. The reform position (Elliott’s prescription) and the status-quo position (the IMF’s “tough love” plus the Bretton Woods institutions’ “reluctance to call Trump out for his erratic behaviour”) are positions. The interests behind them are more compatible than the positions suggest. Developing countries want policy space to design tailored strategies for their own needs; the original 1944 design granted such space within a multilateral discipline. The IMF and World Bank want repayment and macro stability; under the original New Deal principles, those were pursued through state intervention and industrial policy, not through frontloaded fiscal compression. The shared interest is a stable multilateral order that neither collapses into fragmentation nor imposes uniform conditions on structurally different borrowers.

Objective criteria both parties might in principle accept are the original Bretton Woods and New Deal texts, which Kozul-Wright invokes directly. The disagreement is not over what the criteria say but over whether they apply now as they applied then. The BATNA reading is more constrained. Elliott reads the developing-country alternative as continuing dysfunction; the Fund reads its own alternative as continued lending under current terms. Elliott judges the zone of possible agreement to have been narrowing. The institutional reluctance the column documents operates, on this reading, as a constraint on the agreement space independent of the merits of any reform proposal.

What a return to New Deal principles would deliver, and why Elliott judges it won’t

Elliott sets out what he judges a return to the founding principles would deliver: “more policy space for governments to develop strategies tailored to their own needs, a boost to public investment, a new Marshall plan to help poorer countries cope with the climate crisis, and a serious attempt to tackle debt distress.” His judgment is that “none of that has a chance of happening any time soon.” The consequence, in his words: “the world will become hotter, more fractured, more prone to conflict, less equal and a lot angrier.”

That closing judgment is the column’s load-bearing claim. The institutional reluctance Elliott reports — the IMF and World Bank’s “reluctance to call Trump out for his erratic behaviour” — operates as a constraint on the agreement space independent of the merits of any reform proposal. The idea that a new era of partnership will flower under a new US president, he writes, is “for the birds.”

What happens next on the substrate’s evidence

Elliott’s named strain indicators and the dependency surfaces he describes together point to which named support would give way first. The AI-investment support, when non-AI capital shortages compound with bond-market pressure for deficit reduction. The inventory support, when oil stocks run out. The export-pricing support, when a counter-response from a vulnerable economy — Germany per Elliott — materializes. Each is a way in which the present moment could cease to be a lull.

The interface fragilities the column reports are the joints between current and next: the IMF-borrower country interface where “tough love” conditions are transmitted; the WTO-trade interface where protectionism is registered or absorbed; the central-bank-government interface where fiscal discipline is being demanded against voter pressure. Each is currently stressed. The state fragilities Kozul-Wright names — decades of “rising inequality, debt distress and heightened anxiety” — are stock variables, not transient states; once accumulated, they raise the cost of further shocks.

Warnings from Kristalina Georgieva, the head of the IMF, and Andrew Bailey, the governor of the Bank of England, on the need to “run a tighter fiscal ship” come at a moment the substrate reports voters want politicians to do more, not less. “There is little appetite for a new era of austerity, and finance ministers know it.” Elliott reads that mismatch as the precondition for backlash: “If they follow through on the tough talk and deliver painful measures… the backlash is likely to be brutal.” The IMF, on his reporting, is worried that “the limits of this resilience may soon be tested.” The column does not specify what would test them.

Elliott’s closing frame is the most news-pyramid-significant sentence in his column: “This looks more like a crisis postponed than a crisis permanently avoided. A lull before the storm, in all probability.” Whether that “lull before the storm” is short or long depends, on the evidence the column offers, on which named support gives way first and on whether multilateral institutions remain reluctant or moribund — described, again, in Elliott’s words.

Source attribution and verification

The substrate’s anchor on the Bretton Woods date — 2026 minus 82 — checks arithmetically and historically. The Bretton Woods Conference convened 1–22 July 1944 in New Hampshire, where the IMF and World Bank were negotiated. The substrate’s publication context is the IMF annual meeting in Bangkok, with a UK-specific item: the Fund upgraded its growth forecast for the UK and praised the British government’s deficit-reduction plan. The substrate’s publication date is self-reported as 2026-10-15; open-web verification returned no independent corroboration against that date, and the date stands as the substrate’s self-report. Several claims are carried as substrate-attributed rather than independently verified from this prompt: “two big regional wars” and “the conflict spreading to Saudi Arabia” (the “Iran war” referenced from the column headline); the IMF conditionality formula “half the measures in the first year and two-thirds within two years”; the identification of Kristalina Georgieva as head of the IMF and Andrew Bailey as governor of the Bank of England. The Kozul-Wright quotation — that the IMF and World Bank “have acted as hand maidens to an age of rising inequality, debt distress and heightened anxiety” — is preserved verbatim across the substrate and the column.

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Genius Loci — Sense of Place
Reads the character and felt quality of a place.
Pre-Mortem (Fragility)
Imagines a system has already broken and traces the structural fragilities that let it.
Principled Negotiation
Works a negotiation from interests, options, and objective criteria rather than positions.