Finn Brooks: Hey — okay, I have to ask before we start, did you actually sleep last night or were you also reading about the British Navy until two in the morning?
Clara Bennett: I did sleep. You clearly didn't — what pulled you in?
Finn Brooks: Honestly? A single contradiction I couldn't resolve. Britain in 1900 — dominant navy, controls global shipping, basically sets the rules for international trade — and they were doing all of that at 2.3% of GDP in defence spending. The US today is at 3.5% and the trade network that we're supposed to be protecting is visibly under strain. How does that — I mean, more spending, less dominance? That shouldn't work.
Clara Bennett: The number isn't the variable. That's the thing — and this is actually the engine of everything we're talking about today. There's a mechanism here, and the percentage is almost a distraction from it.
Finn Brooks: The feedback loop.
Clara Bennett: The military-economic feedback loop, yes. Define it first: military strength secures trade routes and resource access, that generates wealth, and the wealth funds more military development. Round and round. Now — Britain didn't just stumble into this. Between 1689 and 1823 they built what's called a fiscal-military state. Trade revenues, particularly excise on international commerce, were institutionally channelled into naval spending. The navy protected the trade. The trade funded the navy.
Finn Brooks: So the circuit closes. That's what you mean — the money actually comes back.
Clara Bennett: Right — and that's the distinction that matters more than the percentage. Whether the spending closes the circuit or breaks it open.
Finn Brooks: Which makes me want to throw the 2024 number at this immediately — because global defence spending hit 2.46 trillion dollars last year. Sustained post-Cold War high. So states are still betting everything on this loop. They believe it's running. But then you look at Britain versus the US and you have to ask — are they closing the circuit, or are they just pouring money into something that isn't cycling back?
Clara Bennett: And that is — yeah, that is genuinely the question this whole episode turns on. What makes it self-sustaining, and what breaks it.
Finn Brooks: No but wait — I want to sit with the Britain framing for one more second, because 1689 to 1823 is a really long time to hold that architecture together. Like, what keeps the circuit from breaking during that entire stretch?
Clara Bennett: That's exactly the thread we pull next — because the answer isn't just discipline or good policy. It's structural. And once you see the structure, the US-Britain gap starts to make a different kind of sense.
Finn Brooks: Okay but the structure — that's what I'm stuck on. Like, what even holds it together? Because when I picture the loop, I keep wanting to think of it as two separate things — the navy on one side, the trade revenue on the other — but that's not right, is it.
Clara Bennett: It's not. And here's the plain version — imagine a shipping company that hires a security guard to keep its warehouse lane open. The guard gets paid out of the revenue the lane generates. Cut the revenue, you can't pay the guard. Lose the guard, you lose the revenue. That's it. That is the entire loop. No jargon.
Finn Brooks: Wait, that's — yeah. That actually just clicked for me.
Clara Bennett: And the US Navy in the Red Sea right now — that is not a metaphor. Houthi attacks on commercial vessels, 2023 into 2024, triggered a full multinational coalition response led by the US Navy. They are literally protecting the trade-route economics that fund the military that is doing the protecting.
Finn Brooks: No but seriously — the same mechanism, three centuries later, just with guided missiles instead of cannons. Britain runs the loop in 1700, the US Navy runs it in 2024, and the underlying circuit is identical.
Clara Bennett: Which is why naming the two pillars actually matters here. Military capacity — force projection, logistics, the technology to sustain it — and economic capacity — GDP, the industrial base, integration into global trade and finance. Neither pillar alone closes the circuit. The loop between them is the mechanism.
Finn Brooks: Right — but the part that doesn't fit is, why does the loop ever break? Because if it's self-reinforcing, it should just... keep going. Indefinitely.
Clara Bennett: That's the complication. The loop can feed itself — or it can, mm, start feeding on itself. Military spending that grows faster than the economic base it's supposed to protect stops being a complement and starts being a drain. The pillars aren't just reinforcing. They can cannibalize each other.
Finn Brooks: Oh — so grand strategy complementarity isn't automatic. You have to actually maintain the balance between them.
Clara Bennett: Exactly — and that is what makes the Britain-to-US comparison so loaded. It's not that one spent more and one spent less. It's whether the spending was closing the circuit or slowly decoupling the guard from the revenue.
Finn Brooks: Which — okay, I mean, when you put it that way the 2.46 trillion dollar global number feels way less reassuring than it did at the top. Like, states are betting on the loop. But betting on it and actually running it are different things.
Clara Bennett: Right. And once you see that distinction — Britain's fiscal-military state wasn't special because it spent a percentage. It was special because the taxation of trade was institutionally wired directly into naval capacity. The circuit had a physical architecture. That's the layer most people miss when they stop at the loop.
Finn Brooks: But that's actually the thing that breaks my brain about this — if you need both pillars running simultaneously for the circuit to hold, what happens when you're not Britain, when you don't already have both? Like, you're a rising power, you're trying to build the loop from scratch — do you do military first, then economic? Economic first, then military? Because either way you're leaving yourself exposed during the build.
Clara Bennett: That's the sequencing trap. And it's brutal — the scholarship on grand strategy is pretty clear that sequential investment means neither pillar fully succeeds. You prioritize economic openness to grow, you defer military spending, and during that window you're vulnerable. You flip it, build the military first, and you start hollowing the industrial base that was supposed to fund the whole thing.
Finn Brooks: The influence-capability dilemma.
Clara Bennett: Exactly. Whichever pillar you starve first compounds the problem in the other one. It's not linear — it accelerates.
Finn Brooks: Okay wait — so China's whole military-civil fusion thing, is that literally an attempt to just... collapse the sequencing problem? Like, don't build them one after the other, fuse them so they develop in parallel?
Clara Bennett: That's — yes, that is exactly the architecture. Alexandre Dupont-Sinhsattanak's work on MCF and defence-technological innovation makes this really concrete: China isn't bolting a defence sector onto a civilian economy. The civilian tech stack and the military R&D pipeline are, in principle, the same system. A semiconductor advance in a commercial firm is immediately available for weapons development. The feedback loop doesn't have to wait.
Finn Brooks: That's — okay that's genuinely a different architecture than anything Britain was running.
Clara Bennett: Structurally different, yes. And the Belt and Road connects to this — because BRI isn't just economic engagement. Combined with expanding PLA basing across Africa and Central Asia, it's the two pillars deployed as one instrument. You build the port, you station the ships.
Finn Brooks: Which is Djibouti in physical form — US base, Chinese base, French base, all crammed into one tiny country because that chokepoint is where the military-economic intersection is literally happening in real estate. Five foreign bases. One country.
Clara Bennett: Right — and India's response to all of this is, mm, almost the inverse strategy. Rather than fusing the pillars, India refuses to fully commit to either pillar-holder. Engaging the US, Russia, and China simultaneously — buying Russian military hardware while negotiating trade concessions from Washington, hedging every direction.
Finn Brooks: Strategic hedging. Which — wait, is that a solution to the sequencing trap or just a way to avoid the trap by refusing to play?
Clara Bennett: That's the genuinely interesting read on it. India isn't building the loop — it's extracting value from everyone else's loop without locking in. Whether that's sustainable long-term or whether it just defers the moment you have to choose... that's an open question.
Finn Brooks: Which actually brings me back to China's MCF — because fusing the pillars sounds like the smartest possible move, but I keep wondering, no, actually — does fusing them solve the contradiction or does it just hide it? Like, at some point does the fusion surface its own contradictions?
Clara Bennett: That's the question we haven't answered yet — and it connects to something darker in this whole framework, which is what happens when the loop runs in reverse. We'll get there, and it gets worse before it gets better — especially once you factor in whether military spending is actually driving innovation or crowding it out, and whether the two-pillar model is even the whole picture.
Finn Brooks: Wait — running in reverse. That's the thing that keeps snagging me. Because if it's the same loop just going backwards, there has to be a moment when you cross the line and you don't know you've crossed it yet.
Clara Bennett: That's Paul Kennedy's whole argument — imperial overstretch. Military commitments outpace the economic base that's supposed to sustain them. The base shrinks. The commitments become unsustainable. Decline accelerates. It's not a separate theory. It's the feedback loop with the arrows flipped.
Finn Brooks: So overstretch isn't a different failure mode — it's the same mechanism.
Clara Bennett: Right. And the key variable — the thing that tells you which direction the loop is running — is whether the civilian and military sectors are integrated or siloed. That's the diagnostic. The US Cold War case: integrated. Defence investment seeded the semiconductor industry, GPS, the internet. Dual-use innovation cycled back into the broader economy. The loop ran forward.
Finn Brooks: And the Soviets are the — wait, that's the inversion, right? That's the silo case.
Clara Bennett: Completely siloed. Late Soviet military spending — and it was massive — starved the civilian economy rather than seeding it. The defence sector consumed resources, produced nothing the civilian side could use, and the feedback loop just... collapsed. That's not overstretch in the Kennedy sense of too many foreign commitments. That's the industrial base hollowing out from the inside.
Finn Brooks: No way. So you can overstretch without even leaving home — the loop breaks internally before the commitments even become visible as a problem.
Clara Bennett: That's — actually, that's the part that gets underweighted every time. Kennedy focuses on the external commitments. But the Soviet case says the silo between sectors is just as lethal. RAND's postindustrial power work is relevant here — because once you add innovation, knowledge intensity, and organisational proficiency to your power metrics, you can actually see the integration gap. GDP and force size alone would've made the Soviet military look functional much longer than it actually was.
Finn Brooks: Which means Russia right now — in Africa, in Central Asia, energy as economic coercion reinforcing the military footprint — is that the loop working, or is that overstretch in slow motion because the civilian base still isn't integrated?
Clara Bennett: That's genuinely unresolved. The resource politics are real — energy leverage, commodity access, military presence as a package. But whether that's closing the circuit or just buying time while the industrial base stays siloed... I don't think anyone has a clean answer.
Finn Brooks: Okay but then — hang on — what does that mean for the two-pillar model? Because if integration is the actual variable, military and economic capacity might not be enough. Like, China's doing political warfare now. Narrative control, elite capture, subversion — documented in foreign policy research as a deliberate turn targeting the US. Is that a third pillar, or is it just the same two pillars through a different channel?
Clara Bennett: And that is the blind spot I can't resolve. If political warfare is a delivery mechanism — just military and economic capacity operating below the threshold of open conflict — the two-pillar model still holds. But if it operates independently, produces effects that neither pillar can replicate, then the model predicts less than it claims. We don't actually know which one it is.
Finn Brooks: That's — I mean, that should bother us more than it does. If the model can't see a whole instrument of statecraft, everything we said about the loop being self-reinforcing might be missing a variable that breaks it in ways we haven't accounted for.
Clara Bennett: That's exactly where I'd leave it. The loop is real. The inversion is real. But the model's predictive power depends on whether the toolkit has two instruments or three — and that question is genuinely open.
Finn Brooks: The thing I can't settle — and I've been turning this over the whole conversation — is whether the loop is actually inescapable or just... expensive to escape. Because those are completely different problems. If it's inescapable, the hierarchy is destiny. If it's expensive, it's a constraint. And constraints break.
Clara Bennett: That distinction is real. And India's hedging — refusing to lock into anyone's loop — might be the most honest reading of the system we have. Not weakness. Not indecision. An accurate assessment that the loop is constraint, not destiny.
Finn Brooks: Which means the actual risk isn't a dramatic overtaking. It's quieter — underinvestment in one pillar, unnoticed, until the balance tips and nobody's announced it yet.
Clara Bennett: And 2.46 trillion globally in 2024 — that should answer whether the loop is strengthening. Except I genuinely don't know if it does. It might just be the overstretch clock running.
Finn Brooks: Yeah. I don't know either. Which is maybe the most honest place to stop.
Clara Bennett: Good conversation. Genuinely.
