Pomeranz vs. Acemoglu: what kind of explanation does the Great Divergence require?
One says the gap was young, material, and partly an accident of where the coal happened to be. The other says it was set centuries earlier, in the rules. Two Nobel-grade frameworks, one event — and they are not even arguing about the same kind of cause.
The same event, two kinds of explanation
“Western Europe and the most prosperous parts of East Asia — the Yangzi Delta, the Kanto plain — had achieved surprisingly similar levels of certain crucial kinds of economic development by 1750.”
— Kenneth Pomeranz, The Great Divergence: China, Europe, and the Making of the Modern World Economy, 2000
“England fueled the Industrial Revolution because its political and economic institutions were more inclusive than anywhere else in the world.”
— Daron Acemoglu & James A. Robinson, Why Nations Fail, 2012
Two sentences about the same event: why Europe pulled away from China and the rest of the world to open the largest income gap in human history. But read them again. They are not disagreeing about a fact. Pomeranz says the leading cores were alike until late, and what split them was material and could have gone otherwise. Acemoglu and Robinson say the cause ran deep and old, in the rules a society lived by, and was settled long before the first steam engine. One frame says the divergence was a contingency. The other says it was a destiny written in institutions. Before you can ask who is right, you have to see that they are answering a question one level up: not what caused the divergence, but what kind of cause it was.
That is the hinge of the whole comparison, so it is worth naming plainly. There are two families of explanation for any historical turn. One is proximate, contingent, and material: a specific thing happened — a resource appeared, a constraint lifted — and without it the outcome flips. The other is deep, structural, and institutional: the outcome was the cashing-out of differences that had been accumulating for centuries, so that whatever opportunity arose, only one society was positioned to seize it. The crucial move is that the two families do not even accept the same evidence as decisive. Pomeranz’s kind of claim is settled by synchronic comparison — were the cores actually at similar living standards around 1750? — and by resource accounting: did coal and colonies relieve a constraint that would otherwise have bound? Acemoglu’s kind is settled by long-run variation and natural experiments: do institutional differences predict prosperity once you hold geography and culture fixed? Hand each framework the same data and they will reach for different parts of it. The reward of putting them side by side is that the empirical record — the real-wage and welfare-ratio series, the institutional natural experiments — can be read against both, and it does not vindicate either one whole.
The convergence machinery Pomeranz’s resource argument operates on — the Malthusian land constraint, the question of why pre-modern economies kept hitting a ceiling — lives in Growth Theory §13.1. The inclusive-versus-extractive apparatus that Stage 3 cashes out has its formal home in Institutional Economics §18.1. This stage only needs the door: two kinds of cause are on the table, and they are about to be argued at full strength, one each.
Two serious research programs, neither a strawman
Pomeranz’s The Great Divergence broke a consensus that had stood for a century. The older story said Europe had pulled ahead by 1500, maybe earlier, on the strength of some deep cultural or civilizational advantage, and that the only real question was what had gone wrong with China. Pomeranz, writing in the tradition that came to be called the California school — alongside R. Bin Wong and Jack Goldstone — reset the clock. Marshal the standard-of-living data, he argued, and the most advanced Chinese and European regions look alike deep into the eighteenth century. The interesting question was no longer China’s deficiency but Europe’s specific, datable escape. This is a serious quantitative program, and it carries real weight.
Acemoglu, Johnson, and Robinson built the institutional-origins program, the body of work that won the 2024 Nobel Memorial Prize in Economic Sciences. Their claim is that the long-run wealth of nations is governed by institutions — the rules deciding who can own, invest, enforce a contract, and keep the returns — and that the divergence was the working-out of institutional differences whose European advantages predated the takeoff. They brought to a historical question the identification tools of modern empirical economics: cross-country variation, instruments, natural experiments. This too is a serious program at the top of the discipline. Neither of these frameworks is the foil for the other. The honest comparison argues each at its own strongest before it judges either.
Take both framings seriously
Whatever else is contested, the thing to be explained is not. The divergence was real and enormous: a gap that ran from perhaps five-to-one in the early nineteenth century to something near a hundred-to-one at the twentieth century’s peak — the modern face of which the development chapter measures at Development Economics §20.1, and the historical opening of which is the spine of Economic History Ch.6 (The Great Divergence). The two frameworks do not fight over whether it happened. They fight over what kind of explanation it requires — and that is a genuine dispute at the framing layer, not a quibble over magnitudes. The only way to adjudicate it is to let each frame make its strongest case first.
Start with the framework that says the gap is younger and shallower than the civilizational story ever admitted — and that the thing which opened it was, of all things, where the coal happened to be. Kenneth Pomeranz, at full strength.
Pomeranz at full strength: the gap was young, and material
“There is little to suggest that western Europe’s economy had decisive advantages before the nineteenth century, either in its capital stock or its economic institutions — the core regions of China and Japan look much like the core regions of western Europe on most of the criteria we associate with such advantages.”
— Kenneth Pomeranz, The Great Divergence, 2000
For a century the standard story said Europe had been the destined leader since at least 1500, and that the gap was a deep civilizational fact. Pomeranz’s data reset the clock to roughly 1750 and reversed the question. Once you stop assuming Europe was always ahead, “what was wrong with China?” stops being the right question. The right question becomes: what specific, datable thing happened next, that one side had and the other did not?
The claim stands or falls on a measurable thing, so measure it. “Comparably developed” cashes out as proxies for the standard of living: real wages — what a labourer’s pay actually bought in baskets of food, fuel, and cloth — alongside caloric intake, life expectancy, urbanization, and marketed output. Robert Allen’s welfare ratios are the workhorse here, and you can watch the series directly in the pre-1820 core-city explorer below: London set against Beijing as a stand-in for the Chinese core, 1500 to 1820, with the gap opening through the eighteenth century rather than the sixteenth. One honest caveat before you read it. The explorer’s Chinese series is Beijing, and Pomeranz’s strongest parity claim is specifically about the Yangtze Delta — Jiangnan, the rich lower-river region around Shanghai and Suzhou — not the imperial capital. Read Beijing as a proxy for the Chinese core, not as the Jiangnan series Pomeranz actually lays beside England. The leading-core parity is his strongest case; whole-China parity is weaker, and the distinction will matter in the verdict.
The full parity series, with its sources and the reconstructions that complicate it, lives in Economic History Ch.6 (The Great Divergence). The mechanism behind Pomeranz’s frame is the second half of the apparatus, and it is one idea: the binding constraint on every pre-modern economy was land. Food, fuel, fibre, and building material all competed for the same finite acreage, and as population grew, that competition tightened into a ceiling no amount of effort could push through. The growth-theory framing of that ceiling — why the classical economists expected every economy to settle into a stationary state — is at Growth Theory §13.1.
In the Malthusian frame, land $T$ is fixed and every use of it competes. Let an acre yield either food or fuel or fibre; its shadow price $\lambda$ — the value of relaxing the land constraint by one acre — rises as population $N$ presses against $T$:
$$\lambda = \frac{\partial Y}{\partial T}, \qquad \lambda \uparrow \text{ as } N/T \uparrow$$A windfall that supplies food, fuel, or fibre from outside the domestic land base is, in this accounting, equivalent to acquiring acreage for free — it pushes $\lambda$ back down precisely when scarcity would otherwise bite.
Every acre of forest is an acre not growing food; every acre of pasture is an acre not growing cotton. In a world where everything you need has to be grown, more people means more competition for the same dirt, and eventually the dirt runs out. Anything that lets you get food, fuel, or fibre from somewhere other than your own soil is like discovering free land. That is the whole trick Pomeranz says Europe pulled — and pulled by luck, not by virtue.
The escape: where the coal was, and the free acres across the sea
Europe escaped the land ceiling through two pieces of good fortune, and Pomeranz names both. The first is geological. Britain’s coalfields sat where the people and the workshops already were — the Midlands, the North, within carting distance of cities and rivers — so coal could replace wood as fuel cheaply, freeing forest acres and powering the steam engines that drained the deeper seams in a self-feeding loop. China had vast coal too, but its best deposits, in Shanxi, lay far inland from the booming Yangtze core, behind a punishing and costly overland haul. Same resource, ruinously different geography. The second windfall is the New World — what Pomeranz calls ghost acreage. Plantation sugar was calories, cotton was fibre, and timber was building material that Europe consumed without growing on a single acre of its own soil. The Americas functioned as a vast tract of land Europe got to use without owning it at home, and the windfall landed exactly when European population was pressing hardest against the ceiling. The resource side of that story — the Atlantic plantation flows and their human cost — is the subject of Economic History Ch.9 (Atlantic Slavery and After), and the dedicated argument over whether the West got rich because of slavery lives in the walkthrough Did the West get rich because of slavery? and the reframe Did colonialism enrich or impoverish?. Here, ghost acreage is one mechanism in Pomeranz’s frame, not the whole case.
The reason this is a serious account, and not a shrug about lucky Europe, is that it makes predictions you can test and lose on. If the trigger was contingent and material, then the pre-1750 gaps should be small — testable against the welfare ratios, and largely borne out for the leading cores. The divergence should track resource access and geology rather than some deep cultural endowment — which is why Pomeranz can point at the coal haul from Shanxi rather than at Confucianism. And earlier bursts of prosperity that hit the same land ceiling without a coal-and-colony escape should have petered out — Song China, Renaissance Italy, the Dutch Golden Age all flowered and then stalled against the limits of an organic economy. The frame even tells you where it would break: find a society with comparable resource relief that failed to industrialize, or find the European gap opening centuries before the coal, and Pomeranz is in trouble. On the central point of this stage he is plainly right and the rival frame is weak: around 1750 the leading Chinese and European cores were at comparable living standards, which means whatever deep institutional advantage Europe is supposed to have had was not yet producing a divergence. The frame that insists the cause was old and institutional has to explain why, on the eve of the takeoff, the outcomes still looked the same. The growth-lineage that the resource-relief mechanism breaks out of — the classical stationary state and the road to endogenous growth — is traced in From classical to endogenous growth.
What the frame gets right — and where it bends
Two things in this account are solid. The gap is younger and smaller than the civilizational story ever claimed — the leading-core parity evidence is real, and it is fatal to any reading that has Europe already ahead by 1500. And the triggers were genuinely material and were not institutionally pre-ordained: nobody’s property rights put the coal under Newcastle or the sugar in the Caribbean. Where the frame bends — and this is for the verdict stage to settle, not this one — is at its strong form. Later quantitative work by Stephen Broadberry and Robert Allen pushes the divergence date somewhat earlier than Pomeranz’s 1750, and the parity that holds for the leading cores does not hold for China or India taken as wholes. The basic California-school move survives; the version that claims zero gap until 1800 does not. Hold that thought. First the rival frame deserves the same full hearing.
If the trigger was coal and colonies, why didn’t every society handed a windfall industrialize? Why did the gap stick, and then accelerate, instead of fading like every earlier boom? The rival framework says the answer was there long before the coal — in the rules. Acemoglu and Robinson, at full strength.
Acemoglu at full strength: the cause was in the rules
“The lands that were relatively rich in 1500 are now relatively poor — a reversal of fortune. This pattern is the opposite of what geographic factors would predict, and it reflects the institutions that European colonialism imposed.”
— Daron Acemoglu, Simon Johnson & James A. Robinson, “Reversal of Fortune,” Quarterly Journal of Economics, 2002
Where Pomeranz points at coal beds, Acemoglu and Robinson point at the rules — who could own, invest, enforce a contract, and keep what they earned. On their reading the divergence was set by institutions whose European advantages were building long before the steam engine, and the reversal of fortune is their opening evidence that geography is not destiny: the places nature favoured in 1500 are not the rich places today, because what compounded over the centuries was institutional, not geographic.
The framework is a single distinction with a long reach. Inclusive institutions — secure and broadly held property rights, enforceable contracts, open markets, real constraints on the people who hold power — give ordinary investors and inventors a reason to build, because they can expect to keep the returns and to face competition rather than confiscation. Extractive institutions do the opposite: they concentrate rents in a narrow elite and suppress the creative destruction that threatens it, because the people in charge have more to lose from disruption than to gain from growth. That apparatus, formally, is at Institutional Economics §18.1.
What turns this from a plausible story into a research program is the identification strategy — the way the program isolates institutions from everything correlated with them. The settler-mortality study (“The Colonial Origins of Comparative Development,” 2001) uses the disease environment Europeans faced when they arrived as an instrument: where they could survive they built inclusive settler institutions, where they died in droves they built extractive ones to ship wealth home, and centuries later those institutions, not the mortality, predict prosperity. The reversal of fortune (2002) shows the rich-in-1500 places are the poor places now — the reverse of any pure-geography prediction. And the vivid version is the within-country discontinuity: the two Koreas, identical in geography and culture and split only by a border drawn through their institutions; Nogales, one town divided by a line that runs between prosperity and its absence. The intellectual ancestry of this frame — the line from Veblen through Coase and Douglass North and Oliver Williamson to Acemoglu and Robinson — is the subject of History of Economic Thought Ch.15 (The Institutionalist Tradition: Veblen to Acemoglu).
Why the rules explain what the coal cannot
Here is the claim where the institutional frame is plainly right and the contingency frame is weak. Windfalls are common in history. Sustained, compounding, self-reinforcing growth is rare. Spain got the richest windfall of all — the silver of Potosí — and it financed an empire’s decline rather than an industrial revolution. The Dutch had the trade and the capital and stalled. What separates a society that converts a windfall into a takeoff from one that has a one-off boom is institutional. Britain’s settlement after 1688 made the Crown’s promises credible, which lowered the cost of capital and let a financial revolution proceed; it secured the property of inventors and investors, so the returns to building a better engine flowed to the builder; and it let creative destruction grind forward against entrenched interests instead of being smothered by them. Hand that institutional configuration a coalfield and a colony and you get a trajectory. Hand the same coal and colonies to an extractive order and you get an episode — a windfall spent, not compounded. The institutional frame predicts exactly the thing Pomeranz’s frame underexplains: not the spark, but why the fire kept burning and spread.
And the advantages were deep, not sudden. Long before 1750, parts of Europe had been accumulating the institutional materials of growth: representative-assembly traditions that could check a monarch, merchants with real political weight, courts that would enforce a contract against a powerful defendant, property that a king could not simply seize. The divergence, on this reading, was not an accident that befell an institutionally average Europe. It was the cashing-out of a difference that had been building for centuries and that determined who was positioned to exploit any opportunity that arose. The coal was incidental; the colonies were incidental; what was decisive was the order that could turn either one into self-sustaining growth. That is the strong form of the frame, and at its best it does work the contingency story cannot touch: it explains the durability and the acceleration of a gap that, on a pure-luck account, ought to have closed once the rest of the world copied the machines.
What the frame gets right — and where it over-reaches
The institutional frame is right about the things the contingency frame cannot reach: why windfall-rich societies so often failed to take off, and why this gap, unlike every earlier efflorescence, stuck and accelerated rather than fading. The natural-experiment evidence that institutions cause prosperity is genuinely strong. Where the strong form over-reaches — and this is for the verdict stage — is in two places. First, the cleanest AJR identification is in the colonial world: settler mortality and the reversal of fortune are sharp instruments for what European institutions did overseas, and their bearing on the Europe-versus-China case is a transfer of the institutional logic, not a direct experiment on China. Second, the synchronic-parity evidence from Stage 2 shows comparable outcomes around 1750 even where institutions already differed — which means institutions were a necessary enabling substrate but were not yet, on their own, producing the divergence. The rules mattered; they had not yet finished their work. The within-Europe version of this institutional question — why the takeoff seated in Britain rather than France — is the work of the comparison Britain vs. France: why Britain first?, not this stage.
So one framework says the trigger was material and contingent, and the other says the cause was deep and institutional — and each lands a blow the other cannot parry. The honest move is not to crown a winner. It is to see which framework owns which layer of the same explanation. Hold both against the data.
The verdict: which framework owns which layer
“The Great Divergence between China and Europe opened up well before the Industrial Revolution. The leading regions of Europe were already ahead of the leading regions of China by 1700, and the gap widened steadily thereafter.”
— Stephen Broadberry, drawing on the Broadberry–Guan–Li reconstructions of Chinese GDP, 980–1850
No new framework enters here. The adjudicating evidence is the body of work that sits between the two frames: the quantitative reconstructions Broadberry, Guan, and Li built for Chinese output back to the Song, and Allen’s welfare ratios. They confirm Pomeranz’s basic move — the leading cores really were close around 1750 — while disciplining its strong form: the gap was opening for at least a century before steam, and China taken as a whole had fallen behind the European leaders earlier than zero-gap parity allows. Read the pre-1820 welfare-ratio series one last time with that in mind: leading-core parity is real; whole-economy parity is not.
The cleanest way to adjudicate two frameworks that make different predictions about the same data is to lay their claims side by side and ask, row by row, what the record actually supports. That is the ledger below. The point is not to total a score — it is to see that the evidence vindicates each frame on a different layer.
| Load-bearing claim | Pomeranz frame | Acemoglu frame | What the record shows |
|---|---|---|---|
| Synchronic parity c.1750 | Leading cores comparably developed | European institutional edge already decisive | Pomeranz, for the leading cores: welfare ratios support comparable outcomes — so institutions had not yet produced the gap. |
| Divergence date | ~1750, with the steam engine | Deep, predating the takeoff | Earlier than 1750 for the leading regions (Broadberry, Allen) — against strong-form Pomeranz, not the basic move. |
| The trigger: coal + colonies | Material, contingent, not pre-ordained | Incidental; the rules were decisive | Pomeranz: the triggers were real and not institutionally ordained — no one’s property rights put coal under Newcastle. |
| Persistence + acceleration | Underexplained by resources alone | Institutions sustained and compounded growth | Acemoglu: institutions explain why the gap stuck while earlier windfall booms faded. |
| What kind of explanation | Contingent and material | Deep and institutional | Layered: contingent triggers acting on pre-existing institutional, market-structure, and factor-price differences. |
Walking the ledger
Take the rows in order and the picture resolves. On synchronic parity, Pomeranz wins the layer he cares about: the welfare ratios put the leading cores at comparable living standards around 1750, and strong-form whole-economy parity does not survive Broadberry. On the divergence date, the reconstructions move the opening earlier than 1750 for the leading regions — a point against strong-form Pomeranz, but the basic California-school claim that the gap is recent and was smaller than the civilizational story said still stands. On the trigger, Pomeranz is right that coal and ghost acreage were real and were not institutionally pre-ordained. On persistence and acceleration, Acemoglu is right that what turned a windfall into a self-sustaining trajectory was institutional, and that this is why this gap, alone among history’s booms, stuck and compounded. And on what kind of explanation the divergence requires, neither frame closes the case alone — the answer is layered. Each row shows the same thing: each framework is decisively right about the layer it was built to explain, and over-reaches when it claims the other’s layer too.
The synthesis is the position
Neither pure framework wins, and saying so is not a hedge — it is a strong, specific claim about how the explanation is built. The divergence required both layers. The triggers were contingent, partly accidental, and material: coal in the right place, free acres across the Atlantic — that is Pomeranz’s layer, and he owns it. The capacity to convert those windfalls into sustained, compounding growth, and to keep the gap widening once the rest of the world could see the machines, was institutional and market-structural: secure property, credible commitment, financial depth, integrated markets — that is Acemoglu’s layer, and he owns it. The counterfactuals make the layering concrete. Pull the triggers, and an institutionally able Europe might not have diverged on this timeline — it might have pressed against the same land ceiling that stalled every earlier efflorescence. Remove the institutions, and a trigger-rich Europe might have had a one-off windfall that petered out like Song China or Renaissance Italy or Golden-Age Holland. Both layers are load-bearing; neither is sufficient alone. The leading-core parity vindicates Pomeranz partway; the persistence and acceleration vindicate Acemoglu partway; the earlier-divergence reconstructions of Broadberry and Allen discipline the strong form of the contingency claim.
It is worth being precise about what kind of disagreement this was, because it is unusual. Pomeranz and Acemoglu were not contesting the weight on a shared set of variables — that is the ordinary shape of an economic-history dispute, where the rival camps agree on the model and argue over the coefficients. The question of whether Britain industrialized first because of culture, coal, wages, or science, or why Britain rather than France, is that ordinary parameter-magnitude kind of dispute, fought inside a shared growth-and-institutions frame. This one is different. Pomeranz and Acemoglu were contesting what kind of cause the divergence was — contingent and material against deep and institutional — and that is a disagreement at the framing layer, not the parameter layer. The honest answer to a frame-level dispute is not to split the difference on a coefficient. It is to recognize that the two frames describe different layers of one explanation, and to say which framework owns which layer. That layered synthesis — triggers on Pomeranz’s layer, the capacity to compound them on Acemoglu’s — is the position. It is what taking the evidence seriously commits you to, not a refusal to choose.
Once you see that the two frameworks explain different layers, they stop being rivals. The next questions branch from here. The civilizational-genius public version — was Europe destined to pull ahead, and was the takeoff inevitable? — is the work of Did Britain have to industrialize first?. The within-Europe case — why Britain and not France, where the science was at least as good — is Britain vs. France: why Britain first?. And the modern policy face of the institutions-versus-geography debate this walkthrough historicizes — why some countries are rich and others poor today — lives in Pourquoi certains pays sont-ils riches et d'autres pauvres ?.
Where this leaves us
We started with two sentences about the same event that were not even arguing about the same kind of thing. Pomeranz reset the clock: the leading Chinese and European cores were alike until roughly 1750, and what split them was material and contingent — coal where it was needed and ghost acreage across the sea, relieving a land constraint that had stalled every earlier boom. Acemoglu and Robinson answered the question the contingency frame could not: windfalls are common, but sustained compounding growth is rare, and what made Europe’s windfall stick and accelerate were institutions that had been building for centuries. The quantitative reconstructions of Broadberry and Allen sit between them — confirming Pomeranz’s basic move while pushing the divergence date earlier than his strong form allows.
The honest verdict is layered, and the layering is the point. The divergence required contingent material triggers acting on deep institutional and market-structural differences — Pomeranz owns the trigger layer, Acemoglu owns the capacity-to-compound layer, and neither layer is sufficient alone. The disagreement was never about a coefficient. It was about what kind of explanation a historical phenomenon requires, and the answer the evidence forces is that it requires both kinds at once. The next time someone tells you the West rose because of its genius, or its luck, or its rules, you have the tools to ask which layer they mean — and to notice when they are claiming a layer the evidence does not give them.