Why Nations Fail: A Contrarian Take

Bryan,
Here’s my take. These are excerpts from my review in Regulation. There’s a lot I liked and a lot I disliked. I talked about both. Smith makes some good points above, but he seems pretty rigid on terminology. I don’t think it’s “bizarre,” for example, to talk about inclusive economic institutions.

I am yet to read Why Nations Fail as well, but I don’t know if you’ve seen this exchange between Jeffrey Sachs on the one hand and Acemoglu and Robinson on the other regarding the book: Round 1, Round 2 and Round 3.

I thought “Violence and Social Orders“, by Douglas North et al, was much better all around, more rigorous, fact filled, and with better explanations of the processes and courses that have led a few societies but not most to the advanced modern level of development. WNF has its ups and downs.

WNF covers the same ground overall and to that extent is OK. Like North & Co it does explain the initial great net social benefit provided by the first warlords, kings, etc, who arise to suppress violence generally so as to have a monopoly of it for themselves. (An idea to which libertarians seem to strive to remain totally oblivious). North & Co cover this better too, IMHO, but WNF has some good, different examples of the benefits still being visible hundreds of years later in locations in Africa and elsewhere. That’s OK. And WNF’s examples of national leaders suppressing economic development to maintain their own power — such as the Hapsburgs — no matter what the current cost to the populace and ultimate cost to them, are also pretty good. Again, I think North & Co are more rigorous and have much better detailed explanations of the processes involved, but the more examples one sees of such things the better.

But in my mind WNF really jumps the shark when it starts talking about events and issues that are in any way related to things close to home (the USA) or which might have implications for contemporaneous events. Yes, then the “politically correct views” that Mr Nathan talks about start infecting the whole thing — to the point where Acemoglu and Robinson seem to forget their own reasoning.

Until then, their reasoning is that the richest elites in undeveloped and failing nations gain their wealth “extractively” through rent collection and preserve it through political and social *power* that eliminates competition — thus descendents of the conquistadors remain today among the richest and most powerful families in South America — and that this is *bad* for economic development and society. Fair enough.

But then they somehow jump to how in modern capitalists states with highly competitive markets there are billionaires too, so somehow they must have obtained their wealth through rent collection and suppressing competition too. (They name Bill Gates, though how he managed to accumulate all those rents while a kid sleeping on the floor of his startup’s office while competing with IBM is beyond me.)

At this point howlers start abounding. The one that really grabbed me was how they invoked Cornelius Vanderbilt as the prototype US capitalist “Robber Baron” rent collecting exploiter. This shows absolute ignorance of the historical reality of what they are talking about, combined with totally naive gullibility regarding politically correct cliches.

The truth is exactly the opposite. Vanderbilt started off literally as a kid with a rowboat, and worked his way up to riches by breaking extractive state-sanctioned monopolies, which were the norm at the time as politicians granted lucrative monopoly corporate charters to favorites to operate transport businesses. The landmark US Supreme Court case of Gibbons v Ogden, which established free inter-state commerce as we know it, was about Vanderbilt running a ferry from New Jersey to Manhattan to under-price and break a highly lucrative ferry monopoly granted by the NY State legislature to a single person. The Court ruled for Vanderbilt’s side, and this broke rent-collecting state monopoly charters everywhere across the USA for all time, with immense competitive benefits for us all. No small thing!

That historians of economic development like Acemoglu and Robinson are totally oblivious to this, one of the most fundamental events of economic development in US history, and actually got it totally backwards, calling Vanderbilt the extractive monopolist, I found really appalling.

(The tale of that ferry is also hugely entertaining. Demand for its service was so great that it could take people to near anywhere on the Manhattan shoreline and they’d jump off and be happy to be there — Vanderbilt had to operate that way because if he went to a pre-announced location he’d be arrested and his boat seized. The boat itself was built with trap doors and hidden passages so the bridge crew could get to a hidden “auxiliary control room” if it was boarded by NY police. At one point it was boarded by them, and when they got to the bridge they found a 10-year-old girl at the wheel and the boat headed straight back to a line of New Jersey police boats at mid-river. The NY police jumped overboard to escape being arrested by the NJ police when it got there. Thus state-chartered monopolies functioned. How can historians of this very subject like Acemoglu and Robinson not know this? I mean, it is even fun to know!)

Vanderbilt operated by seeking and breaking over-priced monopolies like this his entire career. See the Pulitzer-winning biography of him, The First Tycoon. He didn’t do it to as a community service to be sure, but to make money. But this shows the process by which economies develop, how markets compete away monopoly rent collectors — how social orders evolve from “extractive” to “inclusive”.

Yet here at home A&R are totally blind to it. The slur “Robber Baron” was hurled at Vanderbilt by monopolists fighting to keep their monopoly state charters — how wrong could Acemoglu and Robinson be?? By their own analysis they got everything exactly backwards here. To preserve the PC cliche.

This and a few other things like it in the last sections of WNF made me wonder about the whole book. But I think it is just standard, mandatory university faculty political correctness contaminating anything that could have anything to do with current events. The sections about earlier times and distant places are much better, and that’s most of the book.

My recommendation: First read Violence and Social Orders. Then read Why Nations Fail for info and examples that supplement V&SO, and maybe some other relevant ideas. Forget WNF for anything that has to do with the USA or contemporary politics or events. Just close the book when you get there.

One person’s opinion, FWIW.

“inclusive economic institutions” means protection of property rights, even though property rights consist precisely in the right to exclude others.

What makes property rights an ‘inclusive institution’ is who has them, how they go about getting them and how well those rights are respected by all, including those with political power.

Bashing WNF seems to have become fashionable. Henderson’s approach is more thoughtful and more civil than most. I’ve been perplexed by the straw men I’ve seen others throw out. Those made me wonder if I read the same book, because they seemed to be trying too hard not to understand it. Almost, as if they made their critique by reading the back cover and riffing off the other bashings.

I read the book and found it worth while. Perhaps that’s because I wasn’t one of anybody’s most brilliant students.

Also, I find the lack of econometrics a plus.

I’ve read the Acemoglu and Robinson book. The counterexamples to their theory are everywhere. All political and economic systems sometimes produced economic growth: slavery, feudalism, capitalism, fascism, Bolshevism, you name it. The main problem is that inclusive institutions, by definition, require that all people are included. And how about society in which, say, only 66% of people are included and others are excluded? Economy of such society should grow, roughly 66% as fast as the same but inclusive one. Its institutions might not even remind on anything inclusive.

I agree with Shane L above. I did not finish the book yet, but I keep up with the authors’ blog now and then, and it does seem to me like they have some valuable insights to offer, which they then over-apply and oversell their explanatory powers.

WNF presents a very simplified, almost anecdotal, version of the Acemoglu, Johnson and Robinson argument. For the rigorous research that lies behind the book, look at their journal publications, starting with

Acemoglu, Johnson and Robinson “Colonial Origins of Comparative Development” AER 2001

Many casual and dubious sounding assertions in the book are backed by solid empirical work in articles by them and others. I believe many economists who wrote positive reviews of HNF based their evaluations on the previously published empirical work. A particularly nice piece of research on long term institutional consequences within a country, looks at a region in Peru: See Dell “Persistent Effects of the Mining Mita” Econometrica 2011:

There are clearly problems with their arguments eg circularity in concept of “inclusive” but don’t dissmiss them on basis of HNF alone.

In response to David on “inclusive institutions”…

My quarrel with the term “inclusive institutions” is partly semantic. Yes, if the government protects everyone’s property rights, rather than just those of an elite, it could make sense to call that “inclusive.” Everyone is “included” in the regime of property rights protection. But what about the later Soviet Union, which didn’t recognize a whole lot of property rights for anyone, but more or less guaranteed you a job and subsistence. Might we call that “inclusive institutions?” It doesn’t seem any less “inclusive” to me than a generalized protection of property rights. Is universal conscription an “inclusive” institution? Everyone is included in it. And there’s a sense in which universal property rights protection is rather, well, exclusive. If you don’t happen to own anything, if you’re homeless, all doors are barred to you. Property rights are, at bottom, about exclusion. “Inclusive economic institutions” has a warm-and-fuzzy feel that seems ill-fitting when they consist in protecting the property rights of all the haves against a handful of have-nots.

Maybe you reply, “Well, ‘inclusive institutions’ could mean a lot of different things, but the authors defined it in a certain way. Go with it.” OK, but it’s still a bad term. It’s vague and misleading. And we don’t really need a new term here. We can just say “protection of private property rights.” Acemoglu and Robinson want their term to mean a bit more than protection of property, a kind of value-added property rights concept. But it isn’t. It’s a value-subtracted version of property rights protection: never properly defined, full of wishy-washy and arbitrary nuances, and much less able to serve the purpose of explaining prosperity than the simple version of property rights is.

To the extent that “inclusive” is indirectly defined by being set in opposition to “extractive,” the objection to it is more fundamental. Inclusive/extractive is not a good dichotomy. How, for example, are we to classify the welfare state– inclusive, or extractive? Well, both, really: everyone (citizens at least) has access to it in bad times; but it subsists by extracting resources from the productive. That seems like a paradox in the Acemoglu and Robinson conceptual framework, but that’s because the conceptual framework is dumb. There’s no reason a regime shouldn’t be both inclusive and extractive. Again, what if a society excludes outsiders without exploiting them, as might in some degree be a fair characterization of the Puritans’ treatment of Indian tribes, or of the Cold War United States to the developing world? It’s not “inclusive”: groups are being excluded. But it’s not “extractive” either: resources aren’t being extracted from the excluded groups.

That “inclusion” and “extraction” do not amount to a mutually exclusive and exhaustive categorization of economic relationships shouldn’t be surprising. Logically, there is no inconsistency between including someone in a club and extracting resources from them. If anything, they might be expected to go together, but either can occur with or without the other. Society’s relations with slaves are probably best described as extractive and not (very) inclusive; with foreigners, typically neither extractive nor inclusive; with low-income individuals, or ancien regime French aristocrats, who pay no tax, inclusive but not extractive; and with affluent American taxpayers, inclusive and extractive.

But what really annoys me about the terms “inclusive institutions” and “extractive institutions” are that they are used to describe political and economic institutions in a misleading parallel way. The big story in WNF is that prosperity results from inclusive economic institutions, which in turn result from inclusive political institutions, at least as a rule, a tendency, a general pattern, from which there are doubtless some exceptions. But not only is this not true in every case; it is not true as a rule, a tendency, a general pattern. It’s not just that there are counter-examples. The counter-examples are as common as the examples, if not more so. WNF is a rambling history that selects anecdotes to support a certain theory. It would have been just as easy, or more so, to collect anecdotes that support an opposite theory.

English history may illustrate. According to WNF, it got inclusive political institutions in the Glorious Revolution of 1688, which established inclusive economic institutions. Actually, the 1688 constitution fell far short, if not of “inclusive political institutions” since that’s Acemoglu and Robinson’s pet phrase and means whatever they want it to mean, then of something more definable, democracy. Only a very small share of the electorate was represented at that time. Not until the Reform Act of 1832 did England begin to have something like a democratic franchise, and universal manhood suffrage came only in 1881.

Parliament did curtail some royal monopolies, but the basic institutional framework it established wasn’t a late 17th-century invention, but grew out of the common law tradition that went all the way back to the Norman Conquest in 1066. The essential features of the English economic institutions in which the Industrial Revolution take place had mostly been in place since the High Middle Ages. And note that England after the Norman Conquest didn’t have “inclusive political institutions” at all. It was a conquest regime in which the rulers were transplanted from France and still spoke French, while the common people spoke Anglo-Saxon dialects.

Again, there’s no reason to be surprised here. For one thing, by respecting property rights, English kings might promote prosperity, enriching themselves as well as their subjects. But I wouldn’t emphasize that. It would be truer to say and easier to understand that English kings might simply care about justice sometimes. This is the kind of consideration to which economists are often blind becomes of their addiction to *homo economicus.* *Homo economicus* is a reasonable approximation of human behavior in situations where they have no particular positive moral duties, e.g., when shopping. Its predictive power deteriorates greatly when people do feel they have some moral obligations. This is a weakness of the public choice school of economics, but at least public choice tends to aspire to a rigorous methodological individualism. Acemoglu and Robinson are constantly reifying the people as against the rulers, and treating these large classes as if they had some sort of collective self-interest. The truth is that the emergence of a rights-respecting regime in Britain has a lot to do with the scribblings of medieval schoolmen, the monasteries, the confrontations of various kings and archbishops, England’s peculiar geography that allowed it to do without a standing army, and the pervasive influence of Christianity, and not much to do with “inclusive political institutions” in any sense.

Not long after British institutions became really inclusive in 1881, Great Britain’s respect for property rights began a sharp decline. Regulation, social legislation, nationalization of key industries, and so forth, displaced laissez-faire liberalism. This seems paradoxical if we think of democracy as “inclusive political institutions” and property rights as “inclusive economic institutions,” and expect the former to cause the latter. But the paradox is artificial. There’s really no reason to expect democratic majorities primarily to demand protection of their property rights, and in practice they don’t.

In a war of anecdotes, Acemoglu and Robinson could doubtless offer stories on the other side, and I’d fire back, and it would all be rather meaningless. Acemoglu and Robinson’s categories are elastic enough that their macro-level thesis is not really falsifiable. The burden of proof is on them to define “inclusive institutions” and “extractive institutions” (political and economic) clearly enough that their hypothesis could be tested. I’m pretty sure that they couldn’t come up with measurable ways of defining their terms such that they would pass statistical tests so as to confirm their theory. That is, if they were forced to provide clear criteria to classify “inclusive” and “extractive” political and economic institutions, such that global and historical datasets could be constructed, it would turn out not to be the case that inclusive economic institutions are generally explained by inclusive political institutions. In that sense, I think their theory is false. In the absence of such an exercise in clarification, it would be truer to say that the theory doesn’t rise to the level of truth or falsity. It’s too vague for that.

WNF is supposed to provide an overarching framework for understanding the success and failure of nations. But the framework is worthless. I can’t even say that “What is true is not new and what is new is not true.” Even the well-known truths they reiterate, e.g., that protecting property rights is good for prosperity or that totalitarianism is bad, are rendered gratuitously obscure by conflation into larger, vacuous generalizations.

I’ve heard it said that “I’d rather be vaguely right than definitely wrong.” Acemoglu and Robinson manage to be vague and wrong at the same time.

Nathan,

Thanks for adding to the discussion– yours are some of the best comment posts I’ve seen in some time on here. Bryan, please coax Nathan into making some regular contributions to this blog!

Your critique of A&R’s ill-defined concepts of inclusive/extractive was my principle problem with WNF. I actually couldn’t finish the book. Believing that their argumentation rested on such a flimsy base, I just kept dreaming up counterfactuals for every chapter and it became a bore.

Yes, if the government protects everyone’s property rights, rather than just those of an elite, it could make sense to call that “inclusive.” Everyone is “included” in the regime of property rights protection. But what about the later Soviet Union, which didn’t recognize a whole lot of property rights for anyone, but more or less guaranteed you a job and subsistence. Might we call that “inclusive institutions?”

@Nathan – You seem to be equivocating ‘government’ and ‘institution’.

A definition of institution is ‘an established law, practice or custom’. Government protecting property rights for everyone is not the institution.

In an interview in 2012, Acemoglu described the differentiation between inclusive and extractive:

At the root of it, it’s actually quite simple. Inclusive institutions are those that have economic institutions that encourage innovation and investment. They provide secure property rights and eliminate significant entry barriers. They create a level playing field so that the majority, or ideally all, of the citizens of a given nation can take part in economic activity.

In contrast, extractive institutions do the opposite. The elite are able to grab things from others or distort allocations and are far from creating a level playing field. A few control political power at the expense of the rest of society. And these institutions are not there by mistake. They are designed, as the name extractive suggests, for the interest of those who have the political power in the society and can do the extracting at the expense of the rest