Mark Carney’s Davos doctrine names the illusion. It doesn’t escape it.
On January 20, 2026, Mark Carney received a standing ovation at the World Economic Forum. That almost never happens at Davos — the room is too full of people who’ve heard everything before and make a point of showing it. The speech earned it. Carney invoked Thucydides, Václav Havel, and the slow death of the rules-based international order, and he said, explicitly, what most world leaders only mutter in private: the system was always partially rigged, the strongest always exempted themselves when convenient, and everyone who kept performing compliance anyway was doing what Havel’s shopkeeper did — putting the sign in the window not because they believed it, but because it was easier than taking it down.
“It is time,” Carney told the room, “for companies and countries to take their signs down.”
The response was electric. The New Yorker called it “a charter for the bleak years ahead.” Carnegie Endowment scholars wrote appreciative analyses. The WEF itself ran a piece framing it as the opening salvo of a new era. Middle powers, the consensus said, have finally found their voice.
Here is the problem. Three days before Carney stepped to the Davos podium, he was in Beijing — the first Canadian Prime Minister to visit China in eight years. He left with a 28-initiative bilateral roadmap, a $4 billion canola deal, a 49,000-unit electric vehicle quota, and a renewed 200-billion-yuan currency swap with the People’s Bank of China. Then he flew to Switzerland and told the world that bilateral deals with hegemons are how middle powers “negotiate from weakness” and “accept subordination.”
That’s not hypocrisy exactly. It’s something more interesting: it’s evidence that even the person making the argument doesn’t believe political will can actually resolve the structural constraints he’s describing. The question matters beyond Davos because the international order Carney is eulogizing — the trade rules, financial systems, and security arrangements built after World War II — is the infrastructure that keeps goods affordable, currencies stable, and small countries from being bullied by large ones. If the replacement is just rhetoric, everyone outside the great powers pays.
The Speech Deserves Its Acclaim
Before taking the thesis apart, give it its due. The Carney doctrine got three things genuinely right.
First, the diagnosis. The rules-based international order — the WTO, the UN, the G20, the COP — was never as neutral as its marketing suggested. The strongest countries shaped the rules, exempted themselves when inconvenient, and enforced compliance asymmetrically. Middle powers benefited from the system’s predictability while privately knowing the story was partial. Carney was right that naming this is a precondition for dealing with it. You can’t adapt to a world you’re still pretending exists.
Second, the structural shift. Over the past decade, great powers have moved from bending the rules to weaponizing the infrastructure. Tariffs as leverage. Financial systems as coercion. Supply chains as vulnerabilities to be exploited. This is qualitatively different from the old order’s hypocrisies. The old order let middle powers free-ride on American hegemony while complaining about it. The new reality is that the infrastructure itself — SWIFT (the messaging system underpinning most international bank transfers), dollar-denominated trade finance, semiconductor supply chains — has become the battlespace. Carney’s observation that “you cannot live within the lie of mutual benefit through integration when integration becomes the source of your subordination” is the clearest formulation of this shift I’ve seen from a sitting head of government.
Third, the framing. “If we’re not at the table, we’re on the menu” is obvious once someone says it. Nobody said it that clearly before. The speech gave middle powers a vocabulary and, more importantly, permission to stop performing optimism about institutions that no longer function as described. That’s genuinely useful.
The speech is correct about where we are. The question is whether it knows where to go next.
The Turn
Carney’s answer to the structural problem is, at its core, a bet on political will.
His prescription has three parts: stop invoking the old order as though it still works; build strategic autonomy in energy, food, critical minerals, finance, and supply chains; and act collectively — “different coalitions for different issues based on common values and interests,” what he called “variable geometry.”
The first part is essentially psychic. The second is something every country has been doing anyway, on its own, for its own reasons. The third is where the argument needs to carry real weight — because without collective action, the first two parts are just every country hedging privately while using Carney’s vocabulary.
And here’s what the last eight weeks have actually looked like.
Canada signed a bilateral deal with China. Then bilateral deals worth $5.5 billion with India — uranium, critical minerals, AI, defense. Then bilateral deals with Australia — clean energy, critical minerals, defense cooperation, pension fund investment. Then Carney flew to Japan for more bilateral discussions. Every stop produced bilateral agreements. None produced coalition architecture. The vocabulary was consistently “middle-power solidarity.” The mechanism was consistently national interest, negotiated one-on-one.
This is not a criticism of Carney’s diplomacy — the deals are good deals. The criticism is of the framework. He has defined the pathology precisely: bilateral negotiation with a hegemon means accepting subordination. Then he spent two months doing bilateral negotiation with hegemons.
The Lowy Institute published its response the same day Carney delivered his Australian parliamentary address. The verdict: “His vision renders middle-power strategic autonomy a question of political will rather than power, or even diplomatic possibility.” They’re right. But it’s worth understanding why — because it’s not mainly about Carney’s personal inconsistency. It’s about what the incentive structure actually allows.
Three Structural Problems the Speech Can’t Solve With a Better Vocabulary
The dependencies don’t dissolve when you name them.
The dollar is the clearest example. De-dollarization is real — the US dollar’s share of global foreign-exchange reserves has fallen from around 72% in 2000 to under 60% today. China’s cross-border yuan settlements have grown dramatically, with CIPS — China’s alternative to the Western SWIFT payment network — processing 175 trillion yuan in 2024 alone — a 43% jump year-on-year. India has 30-plus countries doing bilateral trade in rupees and has cut the dollar’s share of its own trade from 85% to 72%. Canada just renewed a 200-billion-yuan currency swap with China — a practical mechanism for reducing dollar intermediation in bilateral Canada-China trade.
But notice what’s happening here. Every one of these moves is bilateral or national. India explicitly denies that its rupee settlement push is a de-dollarization move — because naming it as such would provoke Washington. Canada’s currency swap is with China specifically. There is no “middle-power monetary coordination.” There is a set of countries independently building optionality, for their own reasons, while carefully managing the optics so the dominant power doesn’t punish them for it.
That’s not coalition strategy. That’s what structural dependency looks like when it’s managed intelligently.
The same pattern holds in security. Most Western middle powers — Canada, Australia, Japan, South Korea — remain deeply dependent on US security guarantees. Australia’s entire strategic posture is built around the alliance. Canada has no serious military capacity to defend its own Arctic without American support, a gap the Carney government acknowledged at Davos itself. The speech calls for building “sovereignty grounded in the ability to withstand pressure.” For most middle powers, that sovereignty is still largely leased.
The defection incentive is structural, not solvable by speeches.
The Greenland episode is the most instructive recent test case. In mid-January, Trump announced 10% tariffs on eight European countries — Denmark, Norway, Sweden, France, Germany, the UK, Netherlands, Finland — explicitly tied to opposition to American designs on Greenland. The coalition held. European leaders were unified. France and Germany and Norway deployed troops to Greenland under “Operation Arctic Endurance.” The EU was preparing €93 billion in counter-tariffs.
Four days later, the tariff threats were dropped. Trump met NATO Secretary General Rutte at Davos and announced a “framework of a future deal.” The framework has never been disclosed. Rutte said sovereignty “did not come up.”
The coalition held for four days. Relief came through NATO — a US-led institution — not through any middle-power mechanism. The coercive leverage was demonstrated and then withdrawn without any structural change. And here’s the part that matters for the scorecard: the coalition would have held as long as the costs were hypothetical. The harder question — would it hold if the 25% tariffs had actually been implemented? — was never answered. Foreign Policy has already documented the game theory: each player is now calculating when defection yields better bilateral terms. Canada, the country whose Prime Minister just gave the solidarity speech, was threatened with 100% tariffs shortly after Davos. The incentive structure punishes coordination and rewards accommodation.
Carney’s framework has no answer for this. “Variable geometry — different coalitions for different issues” is a description of what countries already do when they share interests. It’s not an enforcement mechanism. Without one, the first middle power offered a materially better bilateral deal has every incentive to take it, and calling that a betrayal of the coalition won’t change the arithmetic.
The “middle powers” category is doing too much work.
The speech treats “middle powers” as a coherent bloc: Canada, Germany, Australia, Japan, the EU, and by implication the Global South. These countries have radically different threat environments, different US and China exposure, different levels of military and economic dependency, and — critically — different views on what the new order should look like.
India is the clearest case. India signed $5.5 billion in bilateral deals with Canada in March, including a $2.6 billion uranium supply agreement. India is also a major recipient of Chinese foreign direct investment in certain sectors, doing rupee settlement with dozens of countries including China’s partners, and explicitly positioning itself as a non-aligned swing state rather than a member of any Western-aligned coalition. India’s game is to be indispensable to everyone. That’s not the same game Carney is playing.
The EU’s talks to align with the CPTPP (the Comprehensive and Progressive Agreement for Trans-Pacific Partnership, a trade bloc spanning 40 countries) through supply chain interoperability — is the most structurally interesting development since Davos, and worth acknowledging honestly. Aligning rules of origin across blocs creates a form of de facto integration that doesn’t require a new institution. Carnegie’s Barbara Weisel is careful to note it’s not an attempt to build a new organization, and it’s explicitly not framed as anti-American. But it is real architecture, and it’s the best empirical version of what coalition-building actually looks like in practice. The honest version of the Carney thesis might be: this, not speeches, is where the work happens.
The Working Hypothesis
The middle-power moment is real as rhetoric and largely fictional as strategy — not because middle powers lack will, but because the speech substitutes political clarity for structural power that middle powers don’t have and can’t conjure by naming the deficit.
The specific tell is this: every leader who applauded at Davos went home and did bilateral deals. Not because they’re cowards or hypocrites, but because the incentive structure of the current moment rewards exactly the behavior the speech diagnosed as the problem. Carney himself did it first, in Beijing, three days before he described it as subordination.
The most honest version of the doctrine isn’t “variable geometry” or “coalitions for different issues.” It’s closer to what’s actually happening: a set of countries independently hedging their structural dependencies — in currency, in supply chains, in critical minerals, in defense — using a shared vocabulary that creates the impression of coordination without the mechanism.
That impression might matter. Shared vocabulary is a precondition for shared institutions. The EU started with coal and steel. The CPTPP started with trade alignment. If the Carney speech marks the moment when middle powers began using the same conceptual framework, something real might eventually be built on top of it.
But “something real might eventually be built” is not a coalition strategy for the rupture that is happening now.
What Would Change My Mind
The thesis is that the middle-power coalition is rhetoric without mechanism. Here’s what would falsify it, with a 24-month window from January 2026:
1. A formal multilateral institution explicitly structured around middle-power solidarity — not CPTPP expansion, not G7 reshuffling, not EU mechanisms that predate Davos — is created or has a binding founding document. I’m not aware of anything on the horizon that qualifies.
2. A documented case where a middle power declined a materially better bilateral offer from the US or China, citing coalition commitment, and the coalition held. So far, the pattern runs entirely in the other direction: Canada signed with China the week of the speech; every post-Davos diplomatic trip produced bilateral agreements.
3. Middle powers demonstrably reduce a specific structural dependency — dollar settlement, US military basing, Chinese rare earth exposure — in a way traceable to coordinated coalition action rather than independent national hedging. The de-dollarization trend is real, but it’s a set of countries independently managing risk. Coalition logic would look different: joint reserve diversification, shared payment infrastructure, coordinated supply chain redundancy built to common standards.
Check back in early 2028. If even one of those three has happened, I’ll update the scorecard. The Carney doctrine deserves that much.
If you found this useful, the best thing you can do is forward it to one person who would push back on it. I’d rather be wrong in public than right in private.