On August 17, The Economist ran a piece titled “The world’s most influential economist is oddly unconvincing.” Its subject was Daron Acemoglu, the Turkish-American MIT economist who shared the 2024 Nobel Prize in economics with Simon Johnson and James Robinson for their work on how institutions shape prosperity. The magazine did not dispute that Acemoglu is exceptionally productive, generous to younger scholars, and among the most cited economists alive; several economists interviewed for the piece said as much themselves. But the article went on to raise real doubts about the empirical foundations of his most celebrated work, pointing in particular to the settler-mortality data underlying the 2001 paper on the colonial origins of comparative development. A 2012 critique by University of Illinois economist David Albouy noted that figures for some countries were simply borrowed from other countries. A later Nobel Committee report also described them as sketchy. The piece raised a more theoretical worry too: that “institutions matter” risks becoming circular once institutions are defined broadly enough and identified retrospectively to fit whatever outcome one is trying to explain.
Read more Morning Greatness: DOJ Threatens California With Legal Action Over Election Integrity Issues
That skepticism is overdue, and it should reach further than the settler-mortality regressions. The deeper trouble with the Acemoglu account of development is that it treats colonization as a founding moment, the point at which a society’s institutional trajectory was fixed for centuries, rather than as one episode layered onto histories that already had their own political episodes. Colonies had a history before colonization. Societies are not simply the residue of what was done to them by a colonial power; they carry inherited structures, norms, and constraints on rulers that shape how any subsequent institutional imposition actually lands. A theory that begins its clock in 1500 or in 1800 is bound to miss most of the story that came before.
Historians of Africa have made exactly this point about the continent the reversal-of-fortune thesis leans on most heavily. John Beattie’s work on Africa showed that many pre-colonial polities had well-developed mechanisms that limited what a ruler could actually do, even in the absence of anything resembling a written constitution. Likewise, E. C. Ejiogu traced a parallel pattern in West Africa, describing indigenous checks on centralized authority that long predate European contact. Fitting either finding into a story where “extractive” and “inclusive” institutions are essentially colonial creations takes some effort. If constraints on rulers already existed in many of these societies, the interesting puzzle is not why colonialism destroyed good institutions that had been there all along, but why the growth Acemoglu’s framework would predict from that inheritance so often failed to show up, especially during the pre-colonial era.
Furthermore, the reversal of fortune thesis, which argues that places that were relatively prosperous or densely populated before European colonization subsequently became relatively poorer because colonial rulers imposed extractive institutions, fares no better when its own statistical machinery is tested directly. Here, the trouble is not merely the settler mortality data The Economist highlighted, but the geographic scope on which the entire argument depends. A close re-examination of the evidence suggests the story holds only for a small handful of countries and that the mechanism actually doing the work is not colonialism in general but British colonialism in particular. The strength of the original argument rests almost entirely on the United States, Canada, Australia, and New Zealand, the four colonies that received overwhelming British settlement and that scholars have long called the Neo-Europes.
Once these four countries and the two anomalous trading city-states of Hong Kong and Singapore are removed from the sample, the celebrated relationship between pre-colonial conditions and modern property rights collapses. Urbanization in 1500 and population density in 1500, the two variables said to capture how attractive a colony was for European settlement, lose virtually all explanatory power outside these six cases, with significance levels so weak that the results are indistinguishable from chance. A regression of expropriation risk on those same pre-colonial indicators tells the same story: strongly significant in the full sample and essentially flat once the Neo-Britains and city-states are dropped.
What explains the pattern, then, is not settlement as such but the identity of the settler. Of the 10 countries with the strongest property rights protections in the comparative sample, nine were former British colonies. Spanish colonization in Latin America also brought European settlers in large numbers, yet it produced a durable elite that used its wealth and political standing to entrench extractive rather than broad-based institutions, a dynamic well documented in the region’s economic history. British settlers, by contrast, carried with them a distinct constitutional inheritance rooted in centuries of parliamentary constraint on the crown, and it was this inheritance, not settlement in the abstract, that appears to travel with them into the colonies. Even the scale of British settlement was not a gradual, continuous phenomenon but close to all-or-nothing: in the four Neo-Britains, settlers eventually made up the overwhelming majority of the population, while in nearly every other British colony they never exceeded 5 percent.
The instrumental-variable strategy used to establish a causal link between settler mortality, property rights, and income runs into the same problem. Once the four Neo-Britains are excluded, the statistical strength of settler mortality as an instrument falls well below the conventional threshold for reliability, and in one fully specified model it no longer predicts property rights at all. Much of what little explanatory power remains turns out to be carried by Hong Kong and Singapore alone, two city-states whose rapid, colonially engineered transformation into global trading ports has little in common with the experience of ordinary developing countries.
These findings do not suggest that history is irrelevant to development or that colonialism left no mark on the modern world. Population density five centuries ago still tracks roughly a third of the variation in income today, a relationship worth taking seriously. But the specific channel proposed, that European colonialism engineered strong or weak property rights depending on settlement conditions, does not survive scrutiny once it is tested beyond the handful of cases that inspired it. The lesson is narrower, and in a sense more interesting, than the original claim: it was not colonialism that built the institutions associated with modern prosperity but a particular colonial power, operating under particular constitutional traditions, in a small number of places where it settled almost completely.
These two problems—the failure of constitutionalist inheritance to translate into faster growth after independence and the collapse of the property rights story once it is tested outside a handful of cases—have fed an entire literature the profession may quietly be drifting toward. An emerging body of work on economic freedom, security of contract, openness to trade, and low barriers to entry found it explains cross-country growth about as well as, and in some specifications better than, measures of “inclusive” political institutions of the Acemoglu-Johnson-Robinson kind. According to Guanghua Yu’s fascinating research, Japan and China illustrate how open access in the economic sphere can facilitate development even in the absence of comparable political openness: Japan’s economic success was more associated with institutions that promoted economic access than inclusive political institutions, while China’s post-1978 reforms dramatically expanded private activity, foreign investment, and market access without a corresponding liberalization of the political sphere. Nor does the AJR framework explain why some countries have proven so much better than others at absorbing foreign technology and turning that absorption into indigenous innovation.
Read more The ‘Trust and Safety’ Regime Wants Its Power Back
Conversely, consider the reversal nobody in the field predicted. In the 1960s, development economists expected the West Indies, with their inherited British legal and parliamentary institutions, exactly the kind of constitutional inheritance the property rights literature treats as decisive, to be the next tier of industrial success. Instead, it was Singapore, Taiwan, and South Korea, states with comparatively young and often authoritarian institutions at the time, that became the technological and scientific powerhouses of the following half-century. Singapore’s presence on that list complicates the story further: its own contribution to the cross-country data rests less on ordinary institutional inheritance than on its status as one of the two anomalous city-states whose rapid, colonially engineered transformation into a trading port sets it apart from the experience of typical developing countries. Whatever separated the two trajectories, it was not a simple ledger of inclusive versus extractive rule, and evidently not a simple ledger of British versus non-British rule either.
There is also a case sitting closer to home for a European magazine, Acemoglu’s own framework, and the property rights literature built largely on the experience of British settlers, both of which have surprisingly little to say about. Finland spent centuries as a Swedish possession before passing, in 1809, into the hands of the Russian Empire as the Grand Duchy of Finland. Even under the tzars, though, the Finns retained a considerable measure of self-rule: their own senate, their own courts and legal code, their own currency from 1860 onward, and enough autonomy over domestic law-making that the Grand Duchy was widely regarded, at the time, as self-governing under a distant sovereign rather than administered as an ordinary Russian province.
Finland today is among the most prosperous and least corrupt states in the world. Whether that later success actually vindicates Acemoglu’s story, whether the autonomy Finland kept under two empires, neither of them British, was itself the inclusive institution doing the work, or whether something else in Finnish society explains the outcome, is a genuinely open question, and it is one his framework, built mostly on the tropics, has never had much to say about. What the Finnish case does establish is that the crude binary of colonizer and colonized flattens a great deal of internal variation, and that variation—how much self-governance a subject nation retains rather than the mere fact of foreign rule or the specific identity of the colonizer—may be doing more of the explanatory work than either Acemoglu’s blunter categories or an account pinned narrowly on British constitutional inheritance allow.
Interestingly, the Acemoglu literature has generally underused, and a more granular one than any cross-country regression can offer. Gareth Austin’s comparison of the pre-colonial Asante kingdom and the colonial Gold Coast state shows a society that already had a working, if not Western, set of property rights before the British arrived: individuals owned the cocoa trees, kola stands, and gold pits they had cleared and worked, even where the underlying land belonged to a chieftaincy, and colonial courts consistently upheld those claims. Ghana was itself a British colony, but the property rights that mattered here were indigenous inventions the British recognized rather than colonial creations, a distinction the settler-identity story tends to blur.
The wartime Cocoa Marketing Board, introduced in 1939 and made permanent in 1947, was by any definition an extractive institution: it taxed cocoa farmers well below the world price and funneled the proceeds to the colonial and then the independent state. And yet it was precisely that extraction which handed Ghana, at independence, “buoyant tax revenues and substantial foreign exchange reserves.” What happened to that surplus afterward was not institutional destiny but a political choice: Kwame Nkrumah’s government spent it inefficiently, and the economy slid into absolute decline in the years before structural adjustment finally arrived in 1983. An extractive institution built the capital; a set of post-colonial political decisions squandered it. That sequence resists the tidy story in which extraction dooms a society and inclusion saves it.
The subnational evidence out of the Americas points in the same direction, and cuts directly against Acemoglu’s own reversal-of-fortune thesis, and against a purely settler-identity account of it too. Using regional rather than national data across 18 countries, one recent paper found that pre-colonial population density, Acemoglu’s own proxy for the kind of society that supposedly attracted extractive institutions, is not a curse at the subnational level at all. Within countries, high pre-colonial density predicts higher population density and higher income today, not lower.
Mexico City, raised on the ruins of Tenochtitlán, remains one of the richest regions in the country. Massachusetts and California, both high-density states before Columbus, are richer today than the low-density interior, the reverse of what a simple extraction story would predict, and notable given that the United States is itself one of the four Neo-Britains the property rights literature treats as the strongest case for British inheritance. In Colombia, a textbook case of Spanish extractive colonization, the high-density regions of Antioquia, Santander, and Valle de Cauca remain among the wealthiest in the country. Only Argentina and Chile show a statistically significant reversal, and even there the explanation on closer inspection turns out to be geographic and political accident rather than institutional inheritance: Buenos Aires’s rise owed more to an 18th-century Spanish trade decree redirecting silver shipments than to anything resembling inclusive governance. Wherever the data are fine-grained enough to separate agglomeration and locational advantage from institutions, persistence rather than reversal turns out to be the rule, whether the colonial power in question was British or Spanish.
Acemoglu’s thesis on institutions and colonialism ultimately leaves too many questions unanswered. Its broad claims about the origins of prosperity depend on causal mechanisms that are often difficult to distinguish from the economic, geographic, political, and historical factors that preceded and accompanied colonial rule. The evidence also sits uneasily with cases that do not conform neatly to the predicted relationship between settlement, institutions, and development. Rather than providing a settled explanation of why some societies prospered while others did not, the thesis therefore raises as many questions as it answers. It is, as The Economist rather gently put it, oddly unconvincing.
Read more Delayed Accountability in the American Political Class, Part One
