[Estimated time: 12 mins]
The central thesis of Why Nations Fail is that it’s a country’s institutions — rather than geography, culture, or ignorance — that determine its success or failure. Acemoglu and Robinson (A&R) lay out what appears to be a compelling case, supported by dozens of historical examples.
But the book has several serious problems:
- Slippery definitions. A&R never attempt to define “inclusive” or “extractive” institutions with any precision. As such, they can explain almost everything with hindsight, but very little in advance.
- Western classical liberal bias. A&R’s idea of “inclusive” institutions are all centred around constraining the state. This leads them to overlook forms of inclusivity that work through the state.
- Suspicion of state interference in the economy. A&R seem to view all state involvement in the economy as a sign of “extraction”.
- Strawmanning rival theories. A&R dismiss geography and culture as explanatory factors by accusing their proponents of “determinism”. This is unnecessary, as all these factors can coexist and play some part in explaining success or failure.
Slippery definitions
A&R’s theory can explain almost everything in retrospect because they never nail down their definitions of “inclusive” or “extractive”. This leaves them a lot of wiggle room to squeeze almost everything into their framework. Now, I understand that inclusive–extractive is something of a spectrum, and it’s going to be hard to measure exactly how “inclusive” any given society is. But others trying to answer similar questions have at least tried to come up with indices to consistently measure a country’s level of “social development” (Ian Morris) or “relative wealth and power” (Ray Dalio), however imperfect they may be.
Without such indices, A&R’s theory is unfalsifiable. Any time a country with ostensibly inclusive institutions fails to grow, they can say the institutions weren’t inclusive enough. Any time a country with extractive institutions manages to grow, they can either say it won’t last, or that the institutions are not that extractive.
For example, when Spain failed to industrialise in the 17th century despite having somewhat inclusive institutions, A&R argue it’s because they weren’t inclusive enough.
Example: Industrial Revolution in England, not Spain
A&R: England successfully industrialised in the 17th century while Spain went into economic decline because England had inclusive institutions.
Me: Wait, but you told me that Spain also had a parliament (Cortes) at this time that represented different groups? And you said the Cortes’ assent was needed for new taxes.
A&R: Yes, but the Cortes didn’t represent as broad a set of groups as the English Parliament did, and the Spanish monarchy could sideline it more easily.
On the other hand, when Japan industrialised in the 19th century under what looked like pretty extractive institutions, but China failed to, A&R say this was because absolutism wasn’t as strong in Japan.
Example: Absolutist rule in Japan vs China
A&R: Both Japan and China were poor nations, languishing under absolutist regimes in the mid-19th century.
Me: Are you sure about this? I thought the whole reason the British started smuggling opium into China (which later led to the Opium Wars) was because they had a massive trade deficit with China. Like, the British had bought many Chinese goods such as silk, porcelain, and tea, but the Chinese didn’t want to buy many British goods. So that was why the British decided to sell opium into China in the first place. A “poor [nation] languishing under [an] absolutist regime” just doesn’t seem very accurate here.
A&R: Never mind that. The point is, Western threats in the 19th century were a “critical juncture” and China and Japan responded in different ways. China continued its absolutist path after the Opium Wars, while Japan went on to develop more inclusive institutions under the Meiji Restoration.
Me: Hmmm… so why did they go down different paths?
A&R: Political power in Japan was more fragmented. Absolutism in China was stronger, and the emperor wasn’t overthrown until 1911.
Me: Really? That’s the best you’ve got? Isn’t it possible that it was something to do with the fact that Western powers militarily threatened China before Japan? Because then Japan, having had the benefit of seeing how badly the Opium Wars went for China, decided to capitulate. Might this not explain why the two countries went down different paths, rather than just one country being more “absolutist” than the other?
A&R: No, because it all comes down to institutions. Even though there were peasant rebellions and civil strife, absolutism in China was stronger, and the opposition less organised and autonomous.
Me: How can you even tell whether China or Japan was more absolutist at the time? At the time, China’s population was around 400 million people, spread over a territory larger than even China’s current borders. It’s pretty hard to maintain absolutist control over such a large and spread out population, especially without modern communication technologies like the Internet. So I imagine that in practice, any ruler would’ve had to rely on a bunch of lower-level officials and local administrators, not too dissimilar from Japan.
Me: As for “peasant rebellions” — I assume you’re referring to the Taiping Rebellion. But I’m not convinced that the opposition there was “less organised”, since they formed an entire fricking kingdom with nearly 30 million people, and waged a 14-year war against the Qing dynasty! In fact, I’ve heard scholars suggest that the Taiping Rebellion got so big was because the Qing’s control was too decentralised, with the Qing rulers handing over fiscal and military control to local governments.
A&R: …
Me: Also, after the Meiji Restoration, it wasn’t like Japan developed its military-industrial capacity via “creative destruction” under free markets. Instead they built capacity mostly by absorbing technology from Western powers through state-led, foreign fact-finding missions like the Iwakura Mission. China actually attempted the exact same thing but failed. One reason suggested for the failure is that while Japan led a unified national project, China’s attempt was led by provincial governors. So how do you know it was “stronger absolutism” rather than excessive fragmentation that made the difference?
A&R: …
Western classical liberalism bias
A&R’s view of inclusive institutions has a distinct ideological flavour. Both their political and economic criteria share a common logic: constraining the state. Parliaments, constitutions, private property rights, the rule of law, and free markets are all readily seen as “inclusive” because they all constrain state power. This is recognisably a Western classical liberal perspective.
The problem isn’t that Western classical liberalism is wrong, per se. It’s that it creates a blind spot for A&R, and causes them to overlook forms of inclusiveness that work through the state, such as: a meritocratic civil service open to all classes, publicly funded education, and market regulations that protect against abuses of private power.
A clear example of this blind spot is in A&R’s treatment of Song Dynasty China.
Example: Song Dynasty China
A&R: During the Song Dynasty (960 to 1279), China led the world in many technological innovations. The Chinese invented clocks, the compass, gunpowder, paper money, and blast furnaces to make cast iron before Europe did. China had also had a meritocratic examination system for recruiting civil servants for centuries, and living standards were probably at least as high as Europe.
Me: That sounds good. So I guess you’d say the Song Dynasty’s institutions were pretty inclusive then?
A&R: No, the Song Dynasty still had extractive institutions. Their inventions were not spurred by market incentives but came about under the auspices, or even the orders, of the government. And little of it was commercialised.
Me: Huh? Why does it matter if the inventions weren’t spurred by market incentives? It’s not like US or European institutions become “extractive” when innovations emerge from DARPA, CERN, the NIH, or NASA.
And historians have said the Song Dynasty had a booming economy. They had arguably the most sophisticated credit markets up to that point in history with paper money, deposit banking, and even joint stock companies.
A&R: There was no political representation for groups other than the monarchy in society, nothing resembling a parliament or a Cortes.
Me: But surely that can’t be the only criterion for whether or not a society’s institutions are inclusive. As you point out, China recruited its civil servants under a meritocratic examination system, open to all males of any birth. Which, combined with state-funded public education, increased social mobility. So in that particular sense, China was arguably more inclusive than England, who didn’t develop a meritocratic civil service until the 1800s — which, I should point out, was inspired by the Chinese system!
Me: Besides, you credited Venice during the Middle Ages with “the most advanced set of inclusive economic institutions underpinned by nascent political inclusiveness”, even though it was run by a hereditary aristocratic oligarchy. Isn’t that a double standard?
A&R: Ah, but Venice didn’t monopolise overseas trade, whereas the Chinese did. In fact, between 1377 and 1397, Emperor Hongwu banned all oceangoing tribute missions.
Me: But that ban was two dynasties later! None of the sources I’ve found suggest that the Song state monopolised overseas trade. They just regulated and taxed it. By most accounts, maritime trade flourished during the Song era.
A&R: …
Suspicion of state interference in the economy
Related to the previous point, A&R treat state interference in the economy as a sign of “extractive institutions”. They clearly believe that innovation and creative destruction are only possible in free markets, and that state-led development does not work
I believe reality is a little more complex than this. I don’t dispute that there are examples of states intervening in economies in extractive ways. A great illustration from the book is the marketing boards in colonial Africa. Ostensibly set up to absorb price fluctuations, the boards ended up taking substantial margins, with farmers sometimes receiving as little as 10% of gross income.
But it seems like the real problem here was with the underlying extractive political institutions. When state economic involvement occurs in societies with more inclusive political institutions, it can benefit the broader society. See, for example, the East Asian state-led development success stories. A&R will undoubtedly counter that this is just “catch-up growth”, not true innovation. However, as Mariana Mazzucato points out in The Entrepreneurial State, many groundbreaking technologies including the Internet, GPS, touchscreens, solar and wind technology, and mRNA vaccines, did not emerge naturally from “free markets”. Instead, they were the result of significant state investment and direction.
This makes intuitive sense. The most beneficial innovations are those with large positive spillovers. But those very spillovers mean that, without government intervention, the innovations are unlikely to be made. Basic research probably creates the biggest spillovers, yet it’s the riskiest form of research and hard to commercialise.
That’s not to suggest state-led development under inclusive political institutions is guaranteed to succeed, either. There are plenty of examples of well-intentioned, state-led economic schemes that failed due to poor planning and/or execution. So I think it probably looks more like this (though still oversimplified):
Strawmanning other causal factors
Now, I do think institutions matter. I just think other causal factors like geography and culture are important, too. Rather than engaging with the strongest versions of these arguments, A&R attack strawmen versions.
The historical account we have presented so far indicates that any approach based on historical determinism—based on geography, culture, or even other historical factors—is inadequate.
— Acemoglu and Robinson in Why Nations Fail
… in contrast with the geography and culture hypotheses, Peru is not condemned to poverty because of its geography or culture.
— Acemoglu and Robinson in Why Nations Fail
I’ve never heard proponents of the geography or culture hypotheses claim that any country was “condemned to poverty”. Nor have I ever heard any suggestion that only geography or only culture determines whether a country is rich today. I don’t think anyone seriously suggests that geography or culture can fully explain the wealth differences between North and South Korea.
The actual arguments are far more modest. Jared Diamond looked at geography to explain differences across continents. The particular question he focused on was why New Guinea, a very mountainous island north of Australia, wasn’t as rich as Europe (a question that, frankly, seems overdetermined). Meanwhile, Joseph Henrich examines cultural factors such as increased mobility and impersonal trust that preceded the Enlightenment and Industrial Revolution, which probably helped build the “broad coalitions” needed for inclusive institutions to arise in the first place.
Causation is complex. None of these explanations — geography, culture, institutions — are mutually exclusive. They may simply operate at different levels. For example, A&R describe how, when the British arrived in Virginia, USA, they tried to copy the Spanish’s extractive institutions in Central and South America, but the lack of precious metals and low population densities made that impossible. So geography and culture may just be upstream of institutions, explaining why different countries developed different institutions in the first place.
Conclusion
To be clear: I think institutions matter and can have self-reinforcing tendencies. And, as I said in my summary, A&R deserve credit for taking seriously the relationship between political and economic institutions, which economists have often overlooked.
But “institutions matter” is a more modest claim than the one A&R make. In explaining why North Korea is poorer than South Korea, institutions clearly play a bigger role than geography or culture. But when considering why England and Spain diverged in the 17th century, or China and Japan in the 19th century, I’m not convinced that institutions were the deciding factor — or even a particularly important one. Geography, culture, pure happenstance: all of these seem at least as plausible, and they probably interact in ways that resist any monocausal explanation.
A possible piece of good news from all this. When I first read A&R’s account of virtuous and vicious cycles, I came away thinking that institutions must be exceedingly hard to change — that the only chance to escape a vicious cycle was at a critical juncture. After a closer read, I now think history is just more random and contingent than A&R portray.
Am I being too harsh? Let me know what you think in the comments below!
If you enjoyed this post, you may also like: