Oh, lord. If your theories of economic growth don't even try to explain Imperial Germany, Imperial Japan, Cold War South Korea, or Cold War Taiwan, what the hell good are they?
Acemoglu's issue is that if he follows his thesis to its conclusion it leads to right wing politics in a way the left wing Acemoglu doesn't want. (This is not speculating on his political allegiances: he's called himself a social democrat.) If you really believe that extractive elites are the issue and then define extraction and elites in a principled way you cannot escape that a lot of his favored policies fall under his definition of extraction. You can see this in Brazil, for example, where his left wing politics blinded him to the ways the Brazilian state was extracting and then not using such extracted resources wisely.
Of course, I find this rather easy to resolve: I do think extraction is bad but it's possible the good done with the money overwhelms the bad. In short we should be suspicious of extraction but sometimes the good it supports really is sufficient to justify it. But in that case he would have to admit that his central thesis (elite extraction bad!) is wrong. He would have to rework everything to be about how extraction occurs (and how much damage it causes) vs what determines what the resources created by extraction are spent on (and how effective that is). Which would be a more interesting and better book in my opinion but wouldn't be Acemoglu.
The examples of E/W Germany and N/S Korea are so seductive! It must, must be the institutions!
But doesn’t work elsewhere, mostly. Poverty, corruption, and violence are the norm; it is the peace and prosperity and the rule of law that are really exceptional and fragile, and we just don’t really know why they happen.
> Alert readers may notice a problem here. If you don't know if X causes Y or vice versa, how can you possibly tell if Z causes X and doesn't otherwise affect Y? You can't.
Well, for example, Z may precede X and Y, which excludes reverse causality. In general, Z is a much simpler variable to analyze, and the stories you have to tell in order for Z to affect Y directly get pretty wild. It may not be possible to prove completely that Z is a valid instrument in all cases, but it's probably better than "control for some observables and pray" or "don't even try" which, in many cases, is the alternative.
Maybe the most classic example of IV is the Israeli school experiment, when Israel limited class sizes to 40. A class of 41 would be split up; one of 40 would not. Whether a class has 40 or 41 students effectively random, and shouldn't have any impact on students (or confounders, like neighborhood income) other than through the effect of class size. (Note that we don't look at actual class size, since some 40-student classes will actually be broken up and some won't, which will not be random). While many of AR's instruments are much less clear-cut than this, reducing all of IV to "In reality, what you are mainly hoping for is that the causal effect of Z through X on Y is more intuitively appealing to journal editors and referees than the simple effect of X on Y." is pretty harsh...
> This is why economics is such fun and has such a high reputation among non-economists. Also, the camp-following whores thing.
I’ve actually felt that Why Nations Fail was overly maligned, as its bearish opinion about China seemed more and more wrong as the years went by. Until, fairly swiftly, it didn’t.
Koyama & Rubin’s recent book, How the World Became Rich, tried its best to be fairly balanced across theories, and seemed to mostly be in favour of the Why Nations Fail world view. So it seems to have survived a bit of time now. It doesn’t fit explain everything in world history, and we can definitely critique the vagueness of parts, but the core emphasis on democracy, rule of law, and other “inclusive institutions” seems to be on the right track.
"Political insitutions" are not fundamental. They're downstream of many other things, like genetics and culture. It's absurd to talk about the "nordic model" as independent of nordic PEOPLE. Haiti can not and will not ever have the 'nordic model' implemented in their country, regardless of who's in charge and how much power they have, because such a 'model' depends on a kind of people which categorically do not exist in Haiti. There's a lot of ways in which this is true, but having the raw economic output required to fund such a model (which must necessarily come before such a model is established - you can't pay for anything otherwise) is surely the most basic. But then of course there's having good people who don't like violent crime, people with low future time orientation, people capable of building high trust societies etc, basically everything that was selected for in Northern Europe during it's inadvertant eugenics program (i.e. it's capital punishment heavy justice systems of the past).
But anyway, this is a strong review compared to most others if only for it's brilliant brevity.
Mercifully to the point, on-topic, readable, and no digressive distactions. Understated unobtrusive humour a plus. I don't really mind the lack of "And the moral of the story is" galaxy-brained extrapolative conclusion which is a regular feature of ACX book reviews, since those so often get out over their skis in ways that detract from the main course. Solidly approve.
Very fair review. Why Nations Fail was very disappointing. It was vague and rambling, and while I didn't come away thinking they were wrong, I didn't come away thinking they were right either. It was generally difficult to figure out what their thesis even was, at any detailed level, and whether they had presented any kind of evidence to support it. I wasn't inspired to put in the effort to try to answer those questions, so thanks for having done so. My main take away was "why are these guys so famous?" - so I'm glad you addressed that as well. I'll check out some of their papers, as you suggest.
I did this book as part of a reading group recently, and thought it was pretty much pure sophistry. So much construction of tautological frameworks of technical terms that mean nothing (is “institutions” is mostly just a way to smuggle in “culture” without hurting peoples feelies).
The arguments against geography or other elements were nearly wholly absent and unconvincing where present.
The historical examples seemed a little cherry picked and when there were counterexamples often just hand waved away through definitional high jinks.
The book mostly doesn’t touch very clearly the few examples of actual economic growth/development we have, nor how the “extractive/inclusive” framing is at all helpful other than as a way to smuggle in prior assumptions/values.
Anyway, just reinforces my priors that the people in these (admittedly difficult) fields are still in the absolute wilderness.
Anyway I had a whole giant word document of notes and criticisms, maybe I will find it.
AR measure extractive institutions vs inclusive in very narrative way, that's my biggest critics
As thought experiment imagine:
- In Kolechia all young people are forced to pay 1/4 of their income to old people, weighted by their social status
- In Arstotzka education system sets kids on rigid tracks to certain social class very early, making it hard to change and forging perfect ideology for that
- In Antegria there kids are forced to go at nearby school
- In Obrastan half of state jobs pay basically for burocrats existing
All of that exists in first world as well and AR seems not to make questions why certain institutions
>>> But by any ordinary meaning of "technological change" this statement is patently ridiculous: horses were replaced with tractors, employment shifted from agriculture to industry, the production of steel, electricity and machine tools grew exponentially, and city dwellers moved into highrise apartments with radio, TV and refrigerators.
Soviet manufacturing was heavily dependent on buying/stealing technology from the West. Machine tools were largely purchased from (West) Germany. Soviets and then Russia never tried to wean themselves of Western precision machine technology and create their own machine tools. Soviets were in the same boat with computers. They copied Western designs using Western-sourced parts. They continued using vacuum tubes in electronics long after the rest of the world mostly abandoned them. Starting in 1929, the Ford Motor built and managed soviet car and tractor manufacturing plants. They sent senior engineers and managers over from the US to manage these plants. US companies largely kept Soviet war production up and running after Hitler invaded the USSR. Likewise, the Soviets had to steal the plans for A-bombs and then H-bombs from the US. They also stole the plans for plants that processed the materials for these weapons.
In fact, I can't think of any original technology that came out of the Soviet Union or post-Soviet Russia. Not that they don't have excellent engineering schools and universities. But they didn't/don't have a culture that allowed innovation. That was was an institutional and cultural failure on the part of the Soviets — and then the Russians.
Clear, well-written, straightforward, and understandable. I tend to prefer reviews that are enthusiastic about the book they're reviewing, but sometimes when you set out to review a book, that just doesn't happen. And if we never see those reviews, that way lies p-hacking and the replication crisis. I very much appreciate the constructive approach to disagreement, where the reviewer pointed out papers by the same authors which made the authors' point better than their book. That feels like a behavior that I want to encourage and reward in these book reviews.
I would prefer a bit more of the original text, and relatedly I don't feel I got a visceral sense of what the book would read like to someone who agreed with it. But that feels like a nice-to-have. It certainly would have made the review longer, and I have to respect the brevity of this one.
The book:
Like I said, I don't have a good sense of it. I've read some books that had that "Old Testament" feel that the reviewer describes, so I have to assume that it's like that. The best ones were presenting concepts that currently defy human description, while the worst were propaganda for an ideology. I get the impression that this was closer to the propaganda side, but without a coherent ideology to back it up.
What is "inclusive" and "extractive"? What is upstream of "institutions"? Do good "institutions" ever fail and do bad "institutions" ever succeed, or are they defined in terms of their success and failure? There were some basic questions here that didn't seem to have answers in the review, and the review leaves the impression (or outright states) that the book never answers them either.
For a much more interesting analysis of the causes of growth-enhancing vs. growth-inhibiting institutions, I recommend "Systemic Vulnerability and the Origins of Developmental States: Northeast and Southeast Asia in Comparative Perspective" by Richard F. Doner, Bryan K. Ritchie, and Dan Slater, published in International Organization 59, Spring 2005, pp. 327-361. Their abstract:
"Scholars of development have learned a great deal about what economic institutions do, but much less about the origins of such arrangements. This article introduces and assesses a new political explanation for the origins of "developmental states"—organizational complexes in which expert and coherent bureaucratic agencies collaborate with organized private sectors to spur national economic transformation. Conventional wisdom holds that developmental states in South Korea, Taiwan, and Singapore result from "state autonomy," especially from popular pressures. We argue that these states' impressive capacities actually emerged from the challenges of delivering side payments to restive popular sectors under conditions of extreme geopolitical insecurity and severe resource constraints. Such an interactive condition of "systemic vulnerability" never confronted ruling elites in Indonesia, Malaysia, the Philippines, or Thailand—allowing them to uphold political coalitions, and hence to retain power, with much less ambitious state-building efforts."
It is worth noting that some of AR's empirical work has not aged well. Here is a critical comment of considerable length that was published in a top economics journal.
Geography does matter: all else being equal, it would be a lot harder for Chad to be rich than for France to be rich because Chad is not a very desirable piece of real estate for humans, while France is really nice.
I also enjoyed a number of @pseudoerasmus' tweets on their other work like "The Narrow Corridor", but he locked his Twitter account.
Showing the causal effect of colonialism based on tropical diseases (as if that couldn't have an independent effect) rather than using non-colonized countries like Ethiopia & Thailand as controls always struck me as laughable.
Here's part of my review of "Why Nations Fail" from 2012:
To understand Acemoglu’s professional popularity, you have to grasp how awkward the major features of global economic reality are to careerist economists. If you look naively around the world, you might get the impression that, say, Chinese territories such as Taiwan, Singapore, and Hong Kong have been economically dynamic because they have a lot of Chinese people in them. Moreover, the Overseas Chinese control much of business in Southeastern Asia, so we might assume that the Chinese tend to have a lot on the ball wherever they go.
The epochal conclusion that Deng Xiaoping, urged on by Singapore’s Lee Kuan Yew, drew from this in the late 1970s was that if all the Chinese folks in the world were getting rich except the Maoist Chinese, the problem must lie more in the “Maoist” than in the “Chinese” part. And, indeed, once liberated from Mao’s dogmas and whims, the Mainland Chinese responded with one of history’s greatest economic surges.
To an economist looking for invitations to conferences, however, the danger of adopting the Lee-Deng perspective is its flip side: Some other peoples, such as black Africans, New World Indians, and Pacific Islanders, have tended to lag notably behind Northeast Asians and Europeans, whether at home or abroad, and under all sorts of ideologies and institutions.
Acemoglu’s contribution was to come up with a regression analysis that, he claimed, showed that Third World poverty was the fault of those all-purpose bad guys, European imperialists. In colonies where early Europeans settlers faced low risks of dying from tropical diseases (such as Massachusetts), they set up good “inclusive” institutions. But in colonies where white men died like flies (such as Nigeria), they set up bad “extractive” institutions.
Institutions are (practically) everything, you see. If, say, the Central African Republic is poor, it’s not because it’s a republic in Central Africa (or because poverty is the default condition of humanity), but because it has extractive institutions. And that’s because Europeans didn’t set up inclusive institutions for the Central Africanese.
If Australia or New Zealand or Canada are richer than the Central African Republic, it’s not because Australia or New Zealand or Canada are full of Europeans, it’s because the Europeans hogged the inclusive institutions for the places they colonized. Or something. Acemoglu wrote:
"These results suggest that Africa is poorer than the rest of the world not because of pure geographic or cultural factors, but because of worse institutions."
According to Acemoglu, that’s pretty much all you need to know.
I find Mancur Olson's classic work 'The Rise and Decline of Nations' to be valuable here, even though he is focused upon economic rather than primarily social issues.
FWIW, I found the book quite convincing. Obviously it's long on narrative and short on proof, but how else could it possibly be? This is not a subject amenable to rigorous mathematical demonstrations.
I can't help but notice that many of the loudest and most passionate critics are coming at the book from a right-wing, and occasionally straightforwardly racist, point of view. You have to suspect that they're looking for reasons to find fault with a thesis that undermines their own worldview.
Then again, the thesis is flattering to my own politics so I might well be committing the same error in reverse. Not sure how to tell.
Part of the problem, in judging the book, is probably that there's too much agreement. Everyone knows that good political institutions with minimal corruption will help a nation succeed. So the Why Nations Fail thesis boils down to "this obviously-important factor is even more important than people give it credit for". It's pretty hard to test the difference between "moderately important factor" and "very important factor".
Focusing on political institutions inside a "failed nation" also conveniently ignores outside forces like say, colonialism (pre WW1) or neocolonialism (post WW2) or the British attempt to take over the Suez Canal or the various Arab Israeli wars subsequent to the fiat formation of Israel after World War 2, color revolutions, "freedom loving insurgents" that later become terrorist groups a la Syria, or the creation of the failed status of Libya thanks to outright Western warfare, etc etc.
Poor review (and comments!). Seems to set out disagreeing with the authors theories and, from there, conclude the book is bad. If you disagree with a thesis and expect scientific evidence in a book intended for a popular audience- of course the book is bad and academic papers are better - but that's not the point of a book like this!
Also doesn't reflect well on ACX or this commenting audience that a similar book, WEIRD, is getting glowing reviews even though it's extremely similar in style just containing different conclusions.
That aside, for those here looking for other great books like Why Nations Fail or WEIRD that making sweeping historical arguments and summarize vast amounts of research without provide bullet-proof data/evidence... worth checking out Escape From Rome. I found it to offer an interpretation of history that transcended and included both WEIRD, GGS, and Why Nations Fail.
The other issue is that AR are not very original. Most of these points were made by Douglass North and coauthors. The question that remains is why is Acemoglu so revered by academic economists. It seems his fame is based on his incredible ability to produce top-5 publications based on shady data or dubious IV strategies.
I agree with the author of the review that the academic papers are much better, but the book is not all that bad if regarded as popular science. The main problem with the book is that "institutions" is an umbrella term for a lot of different stuff, plus that monocausal explanations for something as complex and vague as "desirable development" is likely to seduce you rather than enlighten you (a problem it shares with the WEIRD book).
To break the growth code and avoid the middle-income trap, a country needs three things:
1. Some system to generate a surplus that can be invested
2.Some system that ensures this surplus is not mainly invested in idiotic things
3. Some system that ensures that those who invest in non-idiotic things, can be reasonably confident they will be allowed to benefit from their investments.
...these three things can be achieved/created in several ways, and there are also several ways they can break down. The ways that work is to a varying degree context-dependent, so ways/strategies that worked in one country do not necessarily work somewhere else. The context may also change across time. Also, "system" in this context is a historically derived amalgam of institutions, culture, traditions and ways-of-doing-things that to an unspecifiable-in-advance degree can be influenced by political decision making.
Coincidences and good luck/bad luck also play unspecified, but probably significant, parts. The vikings, which is a historical/cultural export from my part of the world, were deeply impressed by - and likely to follow - a chief that had shown "he was favored by luck", even if he was otherwise not very impressive. Probably because according to Norse mythology, being in luck could be a sign you were favored by the Norns (Nornir), the female supernatural beings that spin the threads of fate not only of humans but also of Gods (Aesir) - thus arguably being more powerful than Gods. Perhaps countries, like Gods and humans, are favored to a different degree by the Norns, spinning the threads of fate on their forever-spinning wheels.
The last paragraph was a digression, but is not that what a comment section is for...
Good Review! To the author: have you read "The Birth of Plenty"? It's currently my favorite 'big idea' macroeconomic history book. Curious what you think about it.
"Each society functions with a set of economic and political rules created and enforced by the state and the citizens collectively. Economic institutions shape economic incentives: the incentives to become educated, to save and invest, to innovate and adopt new technologies, and so on. It is the political process that determines what economic institutions people live under, and it is the political institutions that determine how this process works."
I don't see how that's a definition either. It states what economic institutions do and that they are determined by the political process which are determined by the political institutions. It does not state what institutions are at all!
"To be inclusive, economic institutions must feature secure private property, an unbiased system of law, and a provision of public services that provides a level playing field in which people can exchange and contract; it also must permit the entry of new businesses and allow people to choose their careers. ... such rights must exist for the majority of people in society."
As a couple of other comments have pointed out, Douglass North got there first. In his Nobel lecture, he wrote "if the institutional framework rewards piracy then piratical organizations will come into existence; and if the institutional framework rewards productive activities then organizations – firms – will come into existence to engage in productive activities."
I will have more to say about North on my own Substack soon.
why do A&R insist that institutions are the ur-cause of economic growth? why not "culture geography institutions policy all play a role to varying degrees, with institutions playing a slightly higher role"?
my answer:
i think their framing is a product of the context in which the book was written. during a period when institutions and politics were given very little attention, and development was thought of as a top-down policy project to be implemented by global institutions like the world bank, UN, and IMF. this paternalistic technocratic view which placed very little weight on the internal politics of developing countries themselves was the consensus in the 90s and early 2000s, and to the extent to which it is no longer a consensus is in part due to the immense success of A&R in changing the narrative. i'm almost certain i heard or read acemoglu say this in a podcast or interview, but can't find it now.
I liked that review; I hardly ever read the books reviewed, now I can do so without self-guilt. :D Fine puns and fine links to explore further - oh and about institutions/politics: sure, they must be important, as bad ones can and will screw up badly. But ok-ones do not cause development by default, can't push with a rope. Infrastructure and literacy are fine aims, high taxes and red tape: less so. People's lives in Haiti may be improved fastest by leaving. Only: We won't let them in.
I think you can just strip away all the talk about institutions. The core message, that a powerful minority abuse their power and curtail development because things like the rule of law are against their interests, holds up pretty well imo.
Most of the successful growth stories (South Korea, Singapore) feature authoritative admins who wanted growth and weren't only interested in material wealth and power for themselves. Maybe China is like this too?
Talk about institutions is just a distraction from this core message and doesn't add much imo.
I think you can just strip away all the talk about institutions. The core message, that a powerful minority abuse their power and curtail development because things like the rule of law are against their interests, holds up pretty well imo.
Most of the successful growth stories (South Korea, Singapore) feature authoritative admins who wanted growth and weren't only interested in material wealth and power for themselves. Maybe China is like this too?
Talk about institutions is just a distraction from this core message and doesn't add much imo.
In looking at N dimensional problems, using N-X variables can't work. Think Conic Section or real Flat Land by Abbot.
In looking at population density one variable is the control of enteric diseases (food and water born diseases). The great invention of the toilet and disposal system allowed high population densities in the West, but we lost the nitrogen fertilizer and if we didn't create synthetic fertilizer (ammonia) we would have failed. The East (China, etc.) developed the wok for stir frying, blanching, boiling, or otherwise killing the pathogens on the food from using night soil for fertilizer.
The sanitary engineers have saved more lives that all the doctors in the world, but we can't even see our own culture.
AJR intentionally stymied the career of David Albouy for the accurate criticisms he made of AJR. In https://www.aeaweb.org/articles?id=10.1257/aer.102.6.3059 Albouy he correctly showed that AJR had cherry-picked or fabricated a significant portion of their data, but Acemoglu had already built an entire reputation on that work so pressured colleagues to hamper Albouy's advancement.
An admittedly beyond the pale joke that made the rounds in development economy-circles many years ago may none the less be, well, a bit entertaining to re-tell in this forum:
An Asian and an African student both attend the London School of Economics, and become friends. Some years later, the African fellow gets an email from his friend: "Hey, I've become Minister of Transportation in my country. Want to come for a visit?" He does, and at the airport is met by a man diving him cost-free to his friend's house, which turns out to be a large villa in a posh suburb. His host shows him around in all the nice rooms, and he says: "Wow, how could you afford all of this on your salary?" The Asian fellow smiles and takes him to a window at the back of the house. "Tell me what you see". The African fellow says: I see hills, and further away there appears to be a large motorway". The Asian fellow looks at him, smiles and says: "Ten percent". The African fellow leaves, very thoughtful.
Some years later, the Asian fellow gets an email from his friend: "Hey, it was so nice to visit you. Now, I have also become Minister of Transportation. Please come visit me back". The Asian fellow does. At the airport, a big limousine with a private chauffeur meets him, and drives him to a small castle in a large park. The African fellow guides him through room after room filled with paintings of old masters and gilded furniture. "How could you afford all this?" The Asian fellow, astonished, asks. The African fellow smiles, takes him to a window overlooking the large backyard, and asks him what he sees. The Asian fellows stares and says: "Sorry, but I do not see anything except rolling hills and forests". The African fellow looks at him, smiles, and says: "A hundred percent".
Caveat: Asia and Africa are very big places, with lots of different countries in them, and things have changed a lot in most African countries since this joke went the rounds. But it may more generally illustrate that there are differences between corruption which none the less creates development, and totally malign corruption.
I think there is a game theoretic aspect to how any large group functions: neighborhood, society, tribe, nation, supranational group.... This defies any attempt at finding a solution to a question as broad and sweeping as "why do national fail"? Looking at it from that perspective, AR's assertion that "institutions" are what matter hit the mark (which is rather broad) as accurately (and probably more) as any other silver bullet theory, including those based on culture or geography or genetics. Of course, this begs the question of how good (i.e., inclusive) institutions emerge in the first place. It could be because of "culture" (whatever that means). Or, heck, good institutions good beget good institutions until one's geographical luck runs out (e.g., prolonged droughts start to appear).
This review treats WNF as a punching bag, which is fair, but the review itself sheds no more light on the question being asked than AR do.
Reading this review was weird, because it seemed like the author had a vendetta against the book and yet was doing a poor job of making it actually sound bad.
And even if they were right, that just leads to the "conclusion: review not worth the candle" issue as they put it. Most people here are never going to read any of the books "reviewed". The point of the contest is to produce interesting essays, not a shopping guide. If you think a book is bad, you shouldn't write about it in the first place, or else use it a springboard to write something better like Scott often does.
On the bright side, one problem with this year's contest is that so many of the reviews have been strong contenders for the top spot, so it's refreshing to have a really terrible review for contrast every once in a while.
Why is this an unoptimistic take? Is it that hard to limit how extractive elites are in a country?
It's mentioned a full-on invasion would be required to do so, but can't less dramatic intervention do it if you explicitly target solving expropriation risk for a country you're intervening in?
Oh, lord. If your theories of economic growth don't even try to explain Imperial Germany, Imperial Japan, Cold War South Korea, or Cold War Taiwan, what the hell good are they?
Acemoglu's issue is that if he follows his thesis to its conclusion it leads to right wing politics in a way the left wing Acemoglu doesn't want. (This is not speculating on his political allegiances: he's called himself a social democrat.) If you really believe that extractive elites are the issue and then define extraction and elites in a principled way you cannot escape that a lot of his favored policies fall under his definition of extraction. You can see this in Brazil, for example, where his left wing politics blinded him to the ways the Brazilian state was extracting and then not using such extracted resources wisely.
Of course, I find this rather easy to resolve: I do think extraction is bad but it's possible the good done with the money overwhelms the bad. In short we should be suspicious of extraction but sometimes the good it supports really is sufficient to justify it. But in that case he would have to admit that his central thesis (elite extraction bad!) is wrong. He would have to rework everything to be about how extraction occurs (and how much damage it causes) vs what determines what the resources created by extraction are spent on (and how effective that is). Which would be a more interesting and better book in my opinion but wouldn't be Acemoglu.
The examples of E/W Germany and N/S Korea are so seductive! It must, must be the institutions!
But doesn’t work elsewhere, mostly. Poverty, corruption, and violence are the norm; it is the peace and prosperity and the rule of law that are really exceptional and fragile, and we just don’t really know why they happen.
> Alert readers may notice a problem here. If you don't know if X causes Y or vice versa, how can you possibly tell if Z causes X and doesn't otherwise affect Y? You can't.
Well, for example, Z may precede X and Y, which excludes reverse causality. In general, Z is a much simpler variable to analyze, and the stories you have to tell in order for Z to affect Y directly get pretty wild. It may not be possible to prove completely that Z is a valid instrument in all cases, but it's probably better than "control for some observables and pray" or "don't even try" which, in many cases, is the alternative.
Maybe the most classic example of IV is the Israeli school experiment, when Israel limited class sizes to 40. A class of 41 would be split up; one of 40 would not. Whether a class has 40 or 41 students effectively random, and shouldn't have any impact on students (or confounders, like neighborhood income) other than through the effect of class size. (Note that we don't look at actual class size, since some 40-student classes will actually be broken up and some won't, which will not be random). While many of AR's instruments are much less clear-cut than this, reducing all of IV to "In reality, what you are mainly hoping for is that the causal effect of Z through X on Y is more intuitively appealing to journal editors and referees than the simple effect of X on Y." is pretty harsh...
> This is why economics is such fun and has such a high reputation among non-economists. Also, the camp-following whores thing.
Ah, yes, that's useful and productive.
Do AR, in their analysis of Brazil, examine how (unlike South Korea) the ISI policies never expired on time?
I’ve actually felt that Why Nations Fail was overly maligned, as its bearish opinion about China seemed more and more wrong as the years went by. Until, fairly swiftly, it didn’t.
Koyama & Rubin’s recent book, How the World Became Rich, tried its best to be fairly balanced across theories, and seemed to mostly be in favour of the Why Nations Fail world view. So it seems to have survived a bit of time now. It doesn’t fit explain everything in world history, and we can definitely critique the vagueness of parts, but the core emphasis on democracy, rule of law, and other “inclusive institutions” seems to be on the right track.
"Political insitutions" are not fundamental. They're downstream of many other things, like genetics and culture. It's absurd to talk about the "nordic model" as independent of nordic PEOPLE. Haiti can not and will not ever have the 'nordic model' implemented in their country, regardless of who's in charge and how much power they have, because such a 'model' depends on a kind of people which categorically do not exist in Haiti. There's a lot of ways in which this is true, but having the raw economic output required to fund such a model (which must necessarily come before such a model is established - you can't pay for anything otherwise) is surely the most basic. But then of course there's having good people who don't like violent crime, people with low future time orientation, people capable of building high trust societies etc, basically everything that was selected for in Northern Europe during it's inadvertant eugenics program (i.e. it's capital punishment heavy justice systems of the past).
But anyway, this is a strong review compared to most others if only for it's brilliant brevity.
Mercifully to the point, on-topic, readable, and no digressive distactions. Understated unobtrusive humour a plus. I don't really mind the lack of "And the moral of the story is" galaxy-brained extrapolative conclusion which is a regular feature of ACX book reviews, since those so often get out over their skis in ways that detract from the main course. Solidly approve.
Very fair review. Why Nations Fail was very disappointing. It was vague and rambling, and while I didn't come away thinking they were wrong, I didn't come away thinking they were right either. It was generally difficult to figure out what their thesis even was, at any detailed level, and whether they had presented any kind of evidence to support it. I wasn't inspired to put in the effort to try to answer those questions, so thanks for having done so. My main take away was "why are these guys so famous?" - so I'm glad you addressed that as well. I'll check out some of their papers, as you suggest.
I did this book as part of a reading group recently, and thought it was pretty much pure sophistry. So much construction of tautological frameworks of technical terms that mean nothing (is “institutions” is mostly just a way to smuggle in “culture” without hurting peoples feelies).
The arguments against geography or other elements were nearly wholly absent and unconvincing where present.
The historical examples seemed a little cherry picked and when there were counterexamples often just hand waved away through definitional high jinks.
The book mostly doesn’t touch very clearly the few examples of actual economic growth/development we have, nor how the “extractive/inclusive” framing is at all helpful other than as a way to smuggle in prior assumptions/values.
Anyway, just reinforces my priors that the people in these (admittedly difficult) fields are still in the absolute wilderness.
Anyway I had a whole giant word document of notes and criticisms, maybe I will find it.
AR measure extractive institutions vs inclusive in very narrative way, that's my biggest critics
As thought experiment imagine:
- In Kolechia all young people are forced to pay 1/4 of their income to old people, weighted by their social status
- In Arstotzka education system sets kids on rigid tracks to certain social class very early, making it hard to change and forging perfect ideology for that
- In Antegria there kids are forced to go at nearby school
- In Obrastan half of state jobs pay basically for burocrats existing
All of that exists in first world as well and AR seems not to make questions why certain institutions
>>> But by any ordinary meaning of "technological change" this statement is patently ridiculous: horses were replaced with tractors, employment shifted from agriculture to industry, the production of steel, electricity and machine tools grew exponentially, and city dwellers moved into highrise apartments with radio, TV and refrigerators.
Soviet manufacturing was heavily dependent on buying/stealing technology from the West. Machine tools were largely purchased from (West) Germany. Soviets and then Russia never tried to wean themselves of Western precision machine technology and create their own machine tools. Soviets were in the same boat with computers. They copied Western designs using Western-sourced parts. They continued using vacuum tubes in electronics long after the rest of the world mostly abandoned them. Starting in 1929, the Ford Motor built and managed soviet car and tractor manufacturing plants. They sent senior engineers and managers over from the US to manage these plants. US companies largely kept Soviet war production up and running after Hitler invaded the USSR. Likewise, the Soviets had to steal the plans for A-bombs and then H-bombs from the US. They also stole the plans for plants that processed the materials for these weapons.
In fact, I can't think of any original technology that came out of the Soviet Union or post-Soviet Russia. Not that they don't have excellent engineering schools and universities. But they didn't/don't have a culture that allowed innovation. That was was an institutional and cultural failure on the part of the Soviets — and then the Russians.
The whole 'instrumental variables' methodology sounds deeply sketchy.
The review:
Clear, well-written, straightforward, and understandable. I tend to prefer reviews that are enthusiastic about the book they're reviewing, but sometimes when you set out to review a book, that just doesn't happen. And if we never see those reviews, that way lies p-hacking and the replication crisis. I very much appreciate the constructive approach to disagreement, where the reviewer pointed out papers by the same authors which made the authors' point better than their book. That feels like a behavior that I want to encourage and reward in these book reviews.
I would prefer a bit more of the original text, and relatedly I don't feel I got a visceral sense of what the book would read like to someone who agreed with it. But that feels like a nice-to-have. It certainly would have made the review longer, and I have to respect the brevity of this one.
The book:
Like I said, I don't have a good sense of it. I've read some books that had that "Old Testament" feel that the reviewer describes, so I have to assume that it's like that. The best ones were presenting concepts that currently defy human description, while the worst were propaganda for an ideology. I get the impression that this was closer to the propaganda side, but without a coherent ideology to back it up.
What is "inclusive" and "extractive"? What is upstream of "institutions"? Do good "institutions" ever fail and do bad "institutions" ever succeed, or are they defined in terms of their success and failure? There were some basic questions here that didn't seem to have answers in the review, and the review leaves the impression (or outright states) that the book never answers them either.
Interesting read, one of he reviews that was able to be clear without being too preachy!
For a much more interesting analysis of the causes of growth-enhancing vs. growth-inhibiting institutions, I recommend "Systemic Vulnerability and the Origins of Developmental States: Northeast and Southeast Asia in Comparative Perspective" by Richard F. Doner, Bryan K. Ritchie, and Dan Slater, published in International Organization 59, Spring 2005, pp. 327-361. Their abstract:
"Scholars of development have learned a great deal about what economic institutions do, but much less about the origins of such arrangements. This article introduces and assesses a new political explanation for the origins of "developmental states"—organizational complexes in which expert and coherent bureaucratic agencies collaborate with organized private sectors to spur national economic transformation. Conventional wisdom holds that developmental states in South Korea, Taiwan, and Singapore result from "state autonomy," especially from popular pressures. We argue that these states' impressive capacities actually emerged from the challenges of delivering side payments to restive popular sectors under conditions of extreme geopolitical insecurity and severe resource constraints. Such an interactive condition of "systemic vulnerability" never confronted ruling elites in Indonesia, Malaysia, the Philippines, or Thailand—allowing them to uphold political coalitions, and hence to retain power, with much less ambitious state-building efforts."
It is worth noting that some of AR's empirical work has not aged well. Here is a critical comment of considerable length that was published in a top economics journal.
https://www.jstor.org/stable/41724681
Geography does matter: all else being equal, it would be a lot harder for Chad to be rich than for France to be rich because Chad is not a very desirable piece of real estate for humans, while France is really nice.
I liked Broadberry's critique of "Why Nations Fail"
https://entitledtoanopinion.wordpress.com/2017/09/04/stephen-broadberrys-accounting-for-the-great-divergence/
I also enjoyed a number of @pseudoerasmus' tweets on their other work like "The Narrow Corridor", but he locked his Twitter account.
Showing the causal effect of colonialism based on tropical diseases (as if that couldn't have an independent effect) rather than using non-colonized countries like Ethiopia & Thailand as controls always struck me as laughable.
Here's part of my review of "Why Nations Fail" from 2012:
To understand Acemoglu’s professional popularity, you have to grasp how awkward the major features of global economic reality are to careerist economists. If you look naively around the world, you might get the impression that, say, Chinese territories such as Taiwan, Singapore, and Hong Kong have been economically dynamic because they have a lot of Chinese people in them. Moreover, the Overseas Chinese control much of business in Southeastern Asia, so we might assume that the Chinese tend to have a lot on the ball wherever they go.
The epochal conclusion that Deng Xiaoping, urged on by Singapore’s Lee Kuan Yew, drew from this in the late 1970s was that if all the Chinese folks in the world were getting rich except the Maoist Chinese, the problem must lie more in the “Maoist” than in the “Chinese” part. And, indeed, once liberated from Mao’s dogmas and whims, the Mainland Chinese responded with one of history’s greatest economic surges.
To an economist looking for invitations to conferences, however, the danger of adopting the Lee-Deng perspective is its flip side: Some other peoples, such as black Africans, New World Indians, and Pacific Islanders, have tended to lag notably behind Northeast Asians and Europeans, whether at home or abroad, and under all sorts of ideologies and institutions.
Acemoglu’s contribution was to come up with a regression analysis that, he claimed, showed that Third World poverty was the fault of those all-purpose bad guys, European imperialists. In colonies where early Europeans settlers faced low risks of dying from tropical diseases (such as Massachusetts), they set up good “inclusive” institutions. But in colonies where white men died like flies (such as Nigeria), they set up bad “extractive” institutions.
Institutions are (practically) everything, you see. If, say, the Central African Republic is poor, it’s not because it’s a republic in Central Africa (or because poverty is the default condition of humanity), but because it has extractive institutions. And that’s because Europeans didn’t set up inclusive institutions for the Central Africanese.
If Australia or New Zealand or Canada are richer than the Central African Republic, it’s not because Australia or New Zealand or Canada are full of Europeans, it’s because the Europeans hogged the inclusive institutions for the places they colonized. Or something. Acemoglu wrote:
"These results suggest that Africa is poorer than the rest of the world not because of pure geographic or cultural factors, but because of worse institutions."
According to Acemoglu, that’s pretty much all you need to know.
I find Mancur Olson's classic work 'The Rise and Decline of Nations' to be valuable here, even though he is focused upon economic rather than primarily social issues.
FWIW, I found the book quite convincing. Obviously it's long on narrative and short on proof, but how else could it possibly be? This is not a subject amenable to rigorous mathematical demonstrations.
I can't help but notice that many of the loudest and most passionate critics are coming at the book from a right-wing, and occasionally straightforwardly racist, point of view. You have to suspect that they're looking for reasons to find fault with a thesis that undermines their own worldview.
Then again, the thesis is flattering to my own politics so I might well be committing the same error in reverse. Not sure how to tell.
Part of the problem, in judging the book, is probably that there's too much agreement. Everyone knows that good political institutions with minimal corruption will help a nation succeed. So the Why Nations Fail thesis boils down to "this obviously-important factor is even more important than people give it credit for". It's pretty hard to test the difference between "moderately important factor" and "very important factor".
Singapore seems like another counter example although it is partially Democratic.
This is blatantly polemical in a way that immediately biased me against the reviewer's position. They claim to Steelman, but then use analogies like:
> *" if you are willing to accept the underlying premise that" [...] "are determined by the LORD. "*
Nice review.
Focusing on political institutions inside a "failed nation" also conveniently ignores outside forces like say, colonialism (pre WW1) or neocolonialism (post WW2) or the British attempt to take over the Suez Canal or the various Arab Israeli wars subsequent to the fiat formation of Israel after World War 2, color revolutions, "freedom loving insurgents" that later become terrorist groups a la Syria, or the creation of the failed status of Libya thanks to outright Western warfare, etc etc.
Poor review (and comments!). Seems to set out disagreeing with the authors theories and, from there, conclude the book is bad. If you disagree with a thesis and expect scientific evidence in a book intended for a popular audience- of course the book is bad and academic papers are better - but that's not the point of a book like this!
Also doesn't reflect well on ACX or this commenting audience that a similar book, WEIRD, is getting glowing reviews even though it's extremely similar in style just containing different conclusions.
That aside, for those here looking for other great books like Why Nations Fail or WEIRD that making sweeping historical arguments and summarize vast amounts of research without provide bullet-proof data/evidence... worth checking out Escape From Rome. I found it to offer an interpretation of history that transcended and included both WEIRD, GGS, and Why Nations Fail.
The other issue is that AR are not very original. Most of these points were made by Douglass North and coauthors. The question that remains is why is Acemoglu so revered by academic economists. It seems his fame is based on his incredible ability to produce top-5 publications based on shady data or dubious IV strategies.
I agree with the author of the review that the academic papers are much better, but the book is not all that bad if regarded as popular science. The main problem with the book is that "institutions" is an umbrella term for a lot of different stuff, plus that monocausal explanations for something as complex and vague as "desirable development" is likely to seduce you rather than enlighten you (a problem it shares with the WEIRD book).
To break the growth code and avoid the middle-income trap, a country needs three things:
1. Some system to generate a surplus that can be invested
2.Some system that ensures this surplus is not mainly invested in idiotic things
3. Some system that ensures that those who invest in non-idiotic things, can be reasonably confident they will be allowed to benefit from their investments.
...these three things can be achieved/created in several ways, and there are also several ways they can break down. The ways that work is to a varying degree context-dependent, so ways/strategies that worked in one country do not necessarily work somewhere else. The context may also change across time. Also, "system" in this context is a historically derived amalgam of institutions, culture, traditions and ways-of-doing-things that to an unspecifiable-in-advance degree can be influenced by political decision making.
Coincidences and good luck/bad luck also play unspecified, but probably significant, parts. The vikings, which is a historical/cultural export from my part of the world, were deeply impressed by - and likely to follow - a chief that had shown "he was favored by luck", even if he was otherwise not very impressive. Probably because according to Norse mythology, being in luck could be a sign you were favored by the Norns (Nornir), the female supernatural beings that spin the threads of fate not only of humans but also of Gods (Aesir) - thus arguably being more powerful than Gods. Perhaps countries, like Gods and humans, are favored to a different degree by the Norns, spinning the threads of fate on their forever-spinning wheels.
The last paragraph was a digression, but is not that what a comment section is for...
Good Review! To the author: have you read "The Birth of Plenty"? It's currently my favorite 'big idea' macroeconomic history book. Curious what you think about it.
"Each society functions with a set of economic and political rules created and enforced by the state and the citizens collectively. Economic institutions shape economic incentives: the incentives to become educated, to save and invest, to innovate and adopt new technologies, and so on. It is the political process that determines what economic institutions people live under, and it is the political institutions that determine how this process works."
I don't see how that's a definition either. It states what economic institutions do and that they are determined by the political process which are determined by the political institutions. It does not state what institutions are at all!
"To be inclusive, economic institutions must feature secure private property, an unbiased system of law, and a provision of public services that provides a level playing field in which people can exchange and contract; it also must permit the entry of new businesses and allow people to choose their careers. ... such rights must exist for the majority of people in society."
As a couple of other comments have pointed out, Douglass North got there first. In his Nobel lecture, he wrote "if the institutional framework rewards piracy then piratical organizations will come into existence; and if the institutional framework rewards productive activities then organizations – firms – will come into existence to engage in productive activities."
I will have more to say about North on my own Substack soon.
some people in the comments have asked:
why do A&R insist that institutions are the ur-cause of economic growth? why not "culture geography institutions policy all play a role to varying degrees, with institutions playing a slightly higher role"?
my answer:
i think their framing is a product of the context in which the book was written. during a period when institutions and politics were given very little attention, and development was thought of as a top-down policy project to be implemented by global institutions like the world bank, UN, and IMF. this paternalistic technocratic view which placed very little weight on the internal politics of developing countries themselves was the consensus in the 90s and early 2000s, and to the extent to which it is no longer a consensus is in part due to the immense success of A&R in changing the narrative. i'm almost certain i heard or read acemoglu say this in a podcast or interview, but can't find it now.
I liked that review; I hardly ever read the books reviewed, now I can do so without self-guilt. :D Fine puns and fine links to explore further - oh and about institutions/politics: sure, they must be important, as bad ones can and will screw up badly. But ok-ones do not cause development by default, can't push with a rope. Infrastructure and literacy are fine aims, high taxes and red tape: less so. People's lives in Haiti may be improved fastest by leaving. Only: We won't let them in.
I think you can just strip away all the talk about institutions. The core message, that a powerful minority abuse their power and curtail development because things like the rule of law are against their interests, holds up pretty well imo.
Most of the successful growth stories (South Korea, Singapore) feature authoritative admins who wanted growth and weren't only interested in material wealth and power for themselves. Maybe China is like this too?
Talk about institutions is just a distraction from this core message and doesn't add much imo.
I think you can just strip away all the talk about institutions. The core message, that a powerful minority abuse their power and curtail development because things like the rule of law are against their interests, holds up pretty well imo.
Most of the successful growth stories (South Korea, Singapore) feature authoritative admins who wanted growth and weren't only interested in material wealth and power for themselves. Maybe China is like this too?
Talk about institutions is just a distraction from this core message and doesn't add much imo.
Thank you for crystallising why I was so unimpressed with their book when I read it.
In looking at N dimensional problems, using N-X variables can't work. Think Conic Section or real Flat Land by Abbot.
In looking at population density one variable is the control of enteric diseases (food and water born diseases). The great invention of the toilet and disposal system allowed high population densities in the West, but we lost the nitrogen fertilizer and if we didn't create synthetic fertilizer (ammonia) we would have failed. The East (China, etc.) developed the wok for stir frying, blanching, boiling, or otherwise killing the pathogens on the food from using night soil for fertilizer.
The sanitary engineers have saved more lives that all the doctors in the world, but we can't even see our own culture.
AJR intentionally stymied the career of David Albouy for the accurate criticisms he made of AJR. In https://www.aeaweb.org/articles?id=10.1257/aer.102.6.3059 Albouy he correctly showed that AJR had cherry-picked or fabricated a significant portion of their data, but Acemoglu had already built an entire reputation on that work so pressured colleagues to hamper Albouy's advancement.
https://www.econjobrumors.com/topic/david-albouy-vs-daron-acemoglu/page/6
An admittedly beyond the pale joke that made the rounds in development economy-circles many years ago may none the less be, well, a bit entertaining to re-tell in this forum:
An Asian and an African student both attend the London School of Economics, and become friends. Some years later, the African fellow gets an email from his friend: "Hey, I've become Minister of Transportation in my country. Want to come for a visit?" He does, and at the airport is met by a man diving him cost-free to his friend's house, which turns out to be a large villa in a posh suburb. His host shows him around in all the nice rooms, and he says: "Wow, how could you afford all of this on your salary?" The Asian fellow smiles and takes him to a window at the back of the house. "Tell me what you see". The African fellow says: I see hills, and further away there appears to be a large motorway". The Asian fellow looks at him, smiles and says: "Ten percent". The African fellow leaves, very thoughtful.
Some years later, the Asian fellow gets an email from his friend: "Hey, it was so nice to visit you. Now, I have also become Minister of Transportation. Please come visit me back". The Asian fellow does. At the airport, a big limousine with a private chauffeur meets him, and drives him to a small castle in a large park. The African fellow guides him through room after room filled with paintings of old masters and gilded furniture. "How could you afford all this?" The Asian fellow, astonished, asks. The African fellow smiles, takes him to a window overlooking the large backyard, and asks him what he sees. The Asian fellows stares and says: "Sorry, but I do not see anything except rolling hills and forests". The African fellow looks at him, smiles, and says: "A hundred percent".
Caveat: Asia and Africa are very big places, with lots of different countries in them, and things have changed a lot in most African countries since this joke went the rounds. But it may more generally illustrate that there are differences between corruption which none the less creates development, and totally malign corruption.
I think there is a game theoretic aspect to how any large group functions: neighborhood, society, tribe, nation, supranational group.... This defies any attempt at finding a solution to a question as broad and sweeping as "why do national fail"? Looking at it from that perspective, AR's assertion that "institutions" are what matter hit the mark (which is rather broad) as accurately (and probably more) as any other silver bullet theory, including those based on culture or geography or genetics. Of course, this begs the question of how good (i.e., inclusive) institutions emerge in the first place. It could be because of "culture" (whatever that means). Or, heck, good institutions good beget good institutions until one's geographical luck runs out (e.g., prolonged droughts start to appear).
This review treats WNF as a punching bag, which is fair, but the review itself sheds no more light on the question being asked than AR do.
Reading this review was weird, because it seemed like the author had a vendetta against the book and yet was doing a poor job of making it actually sound bad.
And even if they were right, that just leads to the "conclusion: review not worth the candle" issue as they put it. Most people here are never going to read any of the books "reviewed". The point of the contest is to produce interesting essays, not a shopping guide. If you think a book is bad, you shouldn't write about it in the first place, or else use it a springboard to write something better like Scott often does.
On the bright side, one problem with this year's contest is that so many of the reviews have been strong contenders for the top spot, so it's refreshing to have a really terrible review for contrast every once in a while.
Why is this an unoptimistic take? Is it that hard to limit how extractive elites are in a country?
It's mentioned a full-on invasion would be required to do so, but can't less dramatic intervention do it if you explicitly target solving expropriation risk for a country you're intervening in?