This is a realist strategic memorandum written for the American policy community—and an excellent one.
I find its central argument very insightful and persuasive. The course most consistent with American interests would be to abandon an excessively costly attempt at comprehensive containment and shift toward a narrower strategy with clearer objectives, tighter technological boundaries, and measures that allies are more willing to support.
China today possesses the world’s largest manufacturing system, world-class infrastructure, a deep pool of engineering talent, an enormous domestic market, and rapidly strengthening scientific and technological capabilities. The United States cannot isolate China as it did in the 1950s.
China’s strategic response has also been remarkably sophisticated: use technological self-reliance to compress whatever time advantage American controls can create; maintain economic openness to weaken Washington’s ability to assemble a durable containment coalition; and practise restrained economic diplomacy so that commercial partners are not unnecessarily turned into security adversaries.
I agree particularly on the importance of China's continued economic openness, although I'm less certain I'd characterise its economic diplomacy as uniformly restrained.
Either way, the broader strategic effect is important: if partners perceive disengagement from China as economically costly while alignment with Washington increasingly carries costs of its own, the coalition problem becomes much harder for the US than the Cold War analogy suggests.
The June Anthropic shutdown is this thesis in miniature. One secret letter switched off a frontier model worldwide, and within days the EU was talking technological independence and Canada's PM was warning against reliance on American providers. Every unilateral flip of the off switch makes the next coalition harder to build.
This impressive survey is both illuminating and disappointing. Surely it is time to recognize that, its inconsistent with the human project to fool around with mutually assured destruction. What JFK and Khrushchev sought to achieve is the only viable objective. The idea of aggression and bullying as a means of containment and control is past its use-by date. It's the US military industrial complex that needs to be restrained.
What's striking about COCOM's history is how directly alliance cohesion tracked perceived threat level, not stated principle — allies backed the "China differential" hard during the Korean War, then quietly abandoned it the moment the immediate danger passed, commercial interest reasserting itself almost instantly. It's a good reminder that collective economic statecraft isn't really backed by shared conviction, it's backed by a temporary alignment of fear. The 1949 debate about whether British support was "essential to the effectiveness" of controls is basically the same question every sanctions coalition faces today: does the coercive tool work without unanimous buy-in, or does one major defector quietly unravel the whole thing?
I wonder if the unanimity question matters less than WHICH states defect. COCOM seems to suggest coalition cohesion isn't necessarily binary: the effectiveness of controls depends on whether holdouts can provide meaningful substitute markets, technologies or supply routes. The harder challenge may then be keeping those strategically indispensable states aligned once the immediate threat recedes and commercial interests reassert themselves.
That's a sharper way to put it than I had it. It also suggests the real strategic question for any coalition builder isn't "how do we get everyone on board" but "who are the two or three states whose substitution capacity actually matters, and can we make defection costly enough for just them." Everyone else's alignment is basically decorative if the pivotal holdouts stay in. COCOM's Britain problem was exactly this — Britain wasn't just one voice among many, it was the one voice that could functionally break the embargo through Hong Kong trade alone.
Exactly!!! Which makes coalition-building look much more like a problem of identifying chokepoints than achieving consensus.
A numerically broad coalition could still be strategically porous if the few states with meaningful substitution capacity aren't in it. Really enjoyed thinking this through with you!
One thing I kept coming back to was your point that coercive leverage diminishes as states adapt to it. There seems to be a paradox here: the more frequently economic dependence is weaponised, the stronger the incentive for the target - adversary or ally - to reduce that dependence, potentially weakening the same leverage over time. China’s pursuit of self-resilience seems like the clearest example, but I wonder whether we're beginning to see a similar logic among US allies as well.
This is a realist strategic memorandum written for the American policy community—and an excellent one.
I find its central argument very insightful and persuasive. The course most consistent with American interests would be to abandon an excessively costly attempt at comprehensive containment and shift toward a narrower strategy with clearer objectives, tighter technological boundaries, and measures that allies are more willing to support.
China today possesses the world’s largest manufacturing system, world-class infrastructure, a deep pool of engineering talent, an enormous domestic market, and rapidly strengthening scientific and technological capabilities. The United States cannot isolate China as it did in the 1950s.
China’s strategic response has also been remarkably sophisticated: use technological self-reliance to compress whatever time advantage American controls can create; maintain economic openness to weaken Washington’s ability to assemble a durable containment coalition; and practise restrained economic diplomacy so that commercial partners are not unnecessarily turned into security adversaries.
Thanks Leon!
Both of your articles are enlightening. Thank you!
I agree particularly on the importance of China's continued economic openness, although I'm less certain I'd characterise its economic diplomacy as uniformly restrained.
Either way, the broader strategic effect is important: if partners perceive disengagement from China as economically costly while alignment with Washington increasingly carries costs of its own, the coalition problem becomes much harder for the US than the Cold War analogy suggests.
The June Anthropic shutdown is this thesis in miniature. One secret letter switched off a frontier model worldwide, and within days the EU was talking technological independence and Canada's PM was warning against reliance on American providers. Every unilateral flip of the off switch makes the next coalition harder to build.
This impressive survey is both illuminating and disappointing. Surely it is time to recognize that, its inconsistent with the human project to fool around with mutually assured destruction. What JFK and Khrushchev sought to achieve is the only viable objective. The idea of aggression and bullying as a means of containment and control is past its use-by date. It's the US military industrial complex that needs to be restrained.
What's striking about COCOM's history is how directly alliance cohesion tracked perceived threat level, not stated principle — allies backed the "China differential" hard during the Korean War, then quietly abandoned it the moment the immediate danger passed, commercial interest reasserting itself almost instantly. It's a good reminder that collective economic statecraft isn't really backed by shared conviction, it's backed by a temporary alignment of fear. The 1949 debate about whether British support was "essential to the effectiveness" of controls is basically the same question every sanctions coalition faces today: does the coercive tool work without unanimous buy-in, or does one major defector quietly unravel the whole thing?
I wonder if the unanimity question matters less than WHICH states defect. COCOM seems to suggest coalition cohesion isn't necessarily binary: the effectiveness of controls depends on whether holdouts can provide meaningful substitute markets, technologies or supply routes. The harder challenge may then be keeping those strategically indispensable states aligned once the immediate threat recedes and commercial interests reassert themselves.
That's a sharper way to put it than I had it. It also suggests the real strategic question for any coalition builder isn't "how do we get everyone on board" but "who are the two or three states whose substitution capacity actually matters, and can we make defection costly enough for just them." Everyone else's alignment is basically decorative if the pivotal holdouts stay in. COCOM's Britain problem was exactly this — Britain wasn't just one voice among many, it was the one voice that could functionally break the embargo through Hong Kong trade alone.
Exactly!!! Which makes coalition-building look much more like a problem of identifying chokepoints than achieving consensus.
A numerically broad coalition could still be strategically porous if the few states with meaningful substitution capacity aren't in it. Really enjoyed thinking this through with you!
One thing I kept coming back to was your point that coercive leverage diminishes as states adapt to it. There seems to be a paradox here: the more frequently economic dependence is weaponised, the stronger the incentive for the target - adversary or ally - to reduce that dependence, potentially weakening the same leverage over time. China’s pursuit of self-resilience seems like the clearest example, but I wonder whether we're beginning to see a similar logic among US allies as well.
Do not bring a dragon to protect you from a bear. That is the main lesson one needs to learn.