Onpode
Cover art for Imperial collapse follows patterns independent of geography or century—a structural logic

Imperial collapse follows patterns independent of geography or century—a structural logic

November 20, 2026 · 15 min

Iris Holm & Cyrus Reed

Cyrus, long week — tell me you have something genuinely weird for today. Oh I have something weird. Here — react to this: the Western Roman Empire falls in 476 CE. And then, seventeen centuries later, the British Empire starts its rapid collapse after 1945. And the proximate cause — in both cases — is…

Empires across different centuries and continents—including Rome, Han China, the Mongols, the Spanish Empire, the Ottoman Empire, and the British Empire—exhibit strikingly similar structural collapse patterns despite vast differences in technology, culture, and geography.

0:0014:59
Get the next episode on History →

Follow it free — new episodes land in your feed.

Or make your own — any topic, in minutes

More Onpode episodes on History →

About this episode

Empires across different centuries and continents—including Rome, Han China, the Mongols, the Spanish Empire, the Ottoman Empire, and the British Empire—exhibit strikingly similar structural collapse patterns despite vast differences in technology, culture, and geography.

Grounded in 8 sources
Peter Turchin: Cliodynamics and the Search for the Mathematical Laws of History | by NBM House | Medium ↗ · medium.com
THE COLLAPSE OF COMPLEX SOCIETIES ↗ · medium.com
Why do empires collapse? ↗ · medium.com
Medium ↗ · raybwilliams.medium.com
The people's history of collapse — EA Forum ↗ · forum.effectivealtruism.org
Day of Empire: How Hyperpowers Rise to Global Dominance ↗ · mreriksen.weebly.com
Current Research Projects | Peter Turchin ↗ · peterturchin.com
Fall Of Rome Dbq Documents Answers ↗ · sanandres.uep.edu.py
Read transcript

Iris Holm: Cyrus, long week — tell me you have something genuinely weird for today.

Cyrus Reed: Oh I have something weird. Here — react to this: the Western Roman Empire falls in 476 CE. And then, seventeen centuries later, the British Empire starts its rapid collapse after 1945. And the proximate cause — in both cases — is basically the same three-part failure. Spending on the military outruns revenue. The currency loses real value. And capable people stop taking the hard postings at the edges.

Iris Holm: Hm. Seventeen hundred years and the same failure mode.

Cyrus Reed: Right — and here's the plain-language version of what's actually happening. Every empire, no matter how big or rich, is basically a franchise operation. You're trying to enforce standards in locations that are very, very far from headquarters. And at some point the distance problem becomes fatal — the cost of holding those edges exceeds what those edges produce. That's imperial overstretch. That's not a metaphor, that's the structural mechanism.

Iris Holm: So the question isn't why empires collapse. It's whether collapse is structurally inevitable — or just a pattern we're fitting to everything after the fact.

Cyrus Reed: Yes — and that's where, wait, two specific people tried to actually answer that. Joseph Tainter — his book 'The Collapse of Complex Societies' — argues that the more complexity you pile on to solve problems, the worse your return gets on each new layer. Eventually collapse is literally the most efficient path forward. And Peter Turchin, who founded something called cliodynamics, he's trying to run actual mathematical models on this, to find whether there are real structural laws underneath the history.

Iris Holm: Laws. That's a strong word for history.

Cyrus Reed: It is — and I think that's exactly the right place to be uncomfortable. Because if Turchin's cliodynamics holds, then the Roman Empire's third-century crisis, kicking off in 235 CE, and Britain's post-World War II contraction aren't just similar stories. They're instances of the same loop. And that loop — territorial loss cuts revenue, weaker revenue cuts defense, weaker defense causes more territorial loss — it compounds until reform becomes structurally impossible.

Iris Holm: Frankly — that's either the most important pattern in political history, or we're just very good at telling the same story twice.

Cyrus Reed: And figuring out which one it is — that's what today is actually about.

Iris Holm: And to know which one, you have to actually run the loop. Not describe it — run it. Third century, 235 CE. Walk me through it mechanically.

Cyrus Reed: Wait, no, let me start with the pressure, not the response. The Rhine and Danube frontiers are bleeding soldiers constantly. That's real military expenditure, every year, escalating. And the emperor — whoever the emperor is that week, because there were something like twenty-plus claimants during this period — he has to pay those legions or they switch allegiance. So what do you do when the treasury is short?

Iris Holm: You shave the coin.

Cyrus Reed: You shave the coin. The antoninianus — that's the standard coin — starts at roughly 50% silver under Caracalla and by the 260s it's down to about 5% silver. Five percent. It's basically a bronze disc with a silver wash.

Iris Holm: Hold on. Five percent — that's not debasement, that's just lying.

Cyrus Reed: It is lying! But here's where the loop closes and it gets brutal — prices rise because merchants aren't fooled, right? They can weigh a coin. So trade volume drops, the tax base contracts in real terms because the currency collecting the taxes is worth a fraction of what it was. The treasury gets nominally the same number of coins and can buy dramatically less with them.

Iris Holm: So the fix — debase to pay the army — actively shrinks the revenue that pays the army.

Cyrus Reed: That's exactly the loop. The cure is the accelerant. And you can't stop mid-loop — if you stop debasing, you can't pay the soldiers, they mutiny, another claimant rises. If you keep going, fiscal exhaustion just arrives faster.

Iris Holm: And the twenty competing claimants — that's not incidental. That's structural. Each one needs his own military faction funded, so now you have fragmented resources and no one who can run a coordinated reform even if they wanted to.

Cyrus Reed: Right — and that's when the fourth pressure hits, which I think is the one people skip. Frontier recruitment failure. Because once prestige and pay collapse, capable people stop accepting the hard postings. The administrator who might have stabilized the Danube border takes a posting closer to Rome. The edge decays first, fastest, and worst — exactly where you need competence most.

Iris Holm: Four pressures feeding each other. Military overextension, fiscal exhaustion, ideological fracture — nobody believes in the emperor when there are twenty of them — and frontier recruitment failure.

Cyrus Reed: And Han China ran the same sequence. The Spanish Empire — silver from the Americas actually masked it for a generation, but the same loop eventually ran. The Ottomans, the British. Different interfaces, same feedback structure.

Iris Holm: Which is exactly the assumption I want to stress-test. Because 'same structure' is doing a lot of work there.

Cyrus Reed: And 'same structure' is doing a lot of work — but I think the thing that actually generates the loop, the thing underneath all four pressures, is one constraint. Distance. Like, every single mechanism we just named is a distance problem wearing a different costume.

Iris Holm: Unpack that.

Cyrus Reed: Okay — picture a logistics manager at a multinational right now, 2024. She's got Slack, real-time GPS on every shipment, video calls across six time zones. And her farthest regional office is still hemorrhaging talent, decisions arrive late, and the local manager — wait, he stopped actually following HQ directives eight months ago but no one noticed. Distance did that. Not the technology failure. The technology is excellent.

Iris Holm: So satellite communication doesn't solve the problem — it just makes the problem more expensive to ignore.

Cyrus Reed: That's exactly it. The Mongol Empire is the brutal proof case. They conquered — I mean, militarily they were extraordinary, right? Half the known world in decades. But they had no overarching political identity to hold Persians and Chinese and Russians together once the momentum stopped. No imperial glue. And without that, distance just reasserted itself the moment the cavalry stopped moving.

Iris Holm: So military dominance without ideological infrastructure — that's not an empire, that's a raid at scale.

Cyrus Reed: That's — yeah, that's actually a really clean way to put it. And contrast that with the British. The 'civilizing mission' — that ideology was doing real structural work. It reduced the cost of voluntary compliance. Local elites in India, in West Africa, they participated because the narrative gave them a reason to. That's not soft power, that's infrastructure. When that collapsed after 1918, peripheral elites stopped cooperating — not all at once, but the subsidy was gone.

Iris Holm: Hold on. The Spanish Empire — same problem or different?

Cyrus Reed: Same problem, different — wait, actually no, the Spanish case is interesting because the silver from the Americas masked the distance cost for a generation. You could fund overextension because Potosí was printing money for you. But the underlying constraint never went away, it just got deferred. And when the silver flow wobbled, fiscal exhaustion hit fast.

Iris Holm: So the loop doesn't care whether you're solvent. It just waits.

Cyrus Reed: It waits. And Han China, same architecture — the structural feedback ran on a longer timeline than Rome's third-century crisis but the sequence is identical. Distance from the center, frontier recruitment collapses, fiscal base erodes.

Iris Holm: But the Byzantine Empire ran for 800 years after Rome fell. By accepting territorial loss — redefining legitimacy downward rather than trying to hold the whole perimeter. Doesn't that break the determinism argument?

Cyrus Reed: That's — yeah, that's the tension I can't fully resolve. Because if Byzantium arrested the loop for eight centuries by shrinking, then the loop isn't a death sentence. It's a decision point. Which means somewhere between 'structural inevitability' and 'pure contingency' there's a real question about whether leaders have a window to actually reform — and whether anyone ever takes it. And that's honestly where Peter Turchin's modeling gets unsettling in ways we haven't touched yet.

Iris Holm: Right. What the math actually shows — and what it misses — that's the next layer.

Cyrus Reed: Okay but — Tainter is the reason Turchin is even possible, right? Because Tainter's move is so weird and so specific. He's not saying collapse is tragic. He's saying — wait, actually he's saying collapse is *rational*. The system ran the math and simplified itself. That's not failure, that's optimization.

Iris Holm: Rational for the system. Catastrophic for the people in it.

Cyrus Reed: Yes — exactly that gap. And Turchin takes the verbal argument, Tainter's verbal argument, and tries to make it numerical. Cliodynamics. Mathematical modeling. He's looking for the moment empires cross the overextension threshold, and his models suggest roughly fifty to sixty years between crossing that threshold and visible collapse becoming — I mean, it becoming *inevitable*.

Iris Holm: Fifty to sixty years. That's a *window*. That's not a cliff.

Cyrus Reed: That's what gets me. Because if that window is real — if you could hand a Roman emperor in, say, 220 CE a Turchin model and say, you have fifty years to reform before the loop locks — that changes everything about how we read the third-century crisis. It wasn't destiny. It was a missed appointment. Fifty years of missed appointments.

Iris Holm: But that's the assumption I want to press. Does converting Tainter's verbal theory into equations add predictive power — or does it just make the argument look rigorous?

Cyrus Reed: And historians push back *hard* on that. The mechanisms — wait, this is the actual objection — the mechanisms differ profoundly. The Ottoman devşirme system, slave soldiers recruited by levy, that is not the same recruitment ecosystem as Roman volunteer legions. So are we finding a law or just — I mean, are we naming a category so broad it swallows everything?

Iris Holm: "Defense spending exceeds revenue" describes every complex state ever. That's a tautology.

Cyrus Reed: Right — but then the 4.2ka megadrought hits and the whole framing breaks open. Around 2200 BCE, this sustained climate shock collapses the Akkadian Empire and Old Kingdom Egypt simultaneously. Not the feedback loop. Not fiscal exhaustion. A drought disrupts food production and hierarchies just — dissolve. No structural law predicted that.

Iris Holm: So external shock as a confound. The model can't absorb weather.

Cyrus Reed: And then there's the selection bias problem, which I think is actually the scariest objection. We remember Rome. We remember the Akkadian Empire *because* they collapsed in ways that left records and patterns. The polities that faced identical structural pressure and — I don't know, stumbled into a reform that worked — maybe we just don't have their historiography. They don't generate the same kind of dramatic textual record.

Iris Holm: We're sampling from the collapses. Not from the population.

Cyrus Reed: Which means — and this is what I can't fully shake — maybe Turchin's fifty-to-sixty-year window is real and maybe it's just the average lag in a biased sample. I genuinely don't know how you tell those apart from inside the data.

Iris Holm: So the honest position is: the feedback loop is documented. The structural law is not established. And the window Turchin identifies is either the most important reform signal in political history — or an artifact of which stories survived.

Cyrus Reed: And that's — wait, that's actually the thing that settled for me across all of this. Roman Empire, Spanish Empire, Han China, the Ottomans, the British — they span totally different technologies, religions, geographies. And the loop still runs. Which means the loop isn't about any of those variables. It's about distance. It's always been about distance.

Iris Holm: And the debate between Tainter and Turchin and the contingency historians — it's not really about whether those pressures exist. It's about whether they're deterministic.

Cyrus Reed: Right — but then there's that phrase about 476 CE. 'More exhaustion than defeat.' No single battle ended it. No decisive moment. Just — accumulated small failures until there was nothing left to self-correct with. And I keep thinking, that's not a dramatic ending. That's a systems failure that nobody clocked in real time.

Iris Holm: Which reframes the question entirely. It's not whether modern states face the same structural math. It's whether instant communication and financial abstraction have genuinely changed the cost of distance — or just moved the feedback somewhere harder to see.

Cyrus Reed: And I genuinely don't know the answer to that. I want to say yes, we've changed it. But I'm not sure I can distinguish between 'changed' and 'deferred.'

Iris Holm: You started this by asking if seventeen centuries could produce the same failure mode. That still feels like the right question. We just understand now why it's so hard to answer.

Cyrus Reed: Yeah. Good place to stop, I think. Thanks for pushing on the determinism piece — I needed that.