Give "A Farewell to Alms" credit

by Ricardo Hausmann, guest blogger

I read Greg Clark's book [A Farewell to Alms] and really liked it. The book is much more effective in destroying previous explanations for the transition to modern growth than it is in proposing a viable alternative. But this is something that Clark acknowledges. He explicitly states that we know very little of the causes behind the transition to high growth. I learned several very important insights from this book. I will list a few:

1- The typical dating of the transition out of the Malthusian equilibrium is probably off by a century or two. This is so because the high productivity growth sectors had a very low weight in output (because productivity increases in the largest sector - agriculture - were low). Weighting growth by the sectoral weights of a later date reveals a break in productivity trends somewhere back in the 17th century. To me this is interesting because I think that what was key was the emergence of activities much less intensive in land and so more scalable. But at low levels of income people spend most of their income in food, thus trapping the economy in an agricultural-centered process, where the Malthusian mechanism of population growth causing declines in income more chance to work. This opens up other explanations for the Industrial Revolution that remain to be explored.

2- It is hard to argue that the lack of diffusion of the industrial revolution in the XIX century was any of the usual suspects in today's most wanted list: poor institutions, lousy finance, lack of human capital. Within the British empire (e.g. in India) property rights were secure, financial markets were pretty open and efficient and there was quite massive transfers of managerial know-how through out-migration of British managers and skilled workers. The slow spread of the industrial revolution in the XIX century is an important puzzle to which the current development debate - which gets most of its intuitions from the post 1960 datasets needs to propose a convincing explanation. Contrary to Dani's opinion, I do find Clark's evidence of the textile industry in the XIXth century interesting, even if today cars in South Africa or textiles in China are produced with world-class productivity. It points, in my mind, to some other missing factor that is not a usual suspect.

3- It would be a pity if the Clark book is dismissed because of his Darwinian argument. I find very interesting the following two facts. First, in olden days incomes were stagnant. Second, in those days, as opposed to now, the rich had substantially more surviving children than the poor. This means that on average there was a rather strong downward social mobility that lasted a pretty long time. I think these are facts and are very convincing as such. I find the idea that this lead to some form of genetic or cultural (why cultural?) selection that triggered changes in something related to the industrial revolution as completely unsubstantiated in the book. Economists use the subjective rate of time preference as a parameter explaining behavior that they take as exogenous. I know of no work that has established whether this is a genetically or a culturally determined parameter or whether it would buy much in terms of triggering a transition to high growth. Clark does show that there was a decline in interest rates, but this was a global phenomenon and not obviously related to the mechanism he has in mind.

Comments

[robertdfeinman]

This book has been discussed at length on Mark Thoma's blog. Here's the link:
http://economistsview.typepad.com/economistsview/2007/08/gregory-clarks-.html

There were lots of comments about the accuracy of his demographic arguments and many of us took exception to his (implied) "white man's burden" or "survival of the fittest" mindset.

[terence]

I don't think it would be a pity at all if Clark's book was dismissed because of it's Darwinian argument. It is very, very tenuous and what's more it is, to be frank, rather repugnant. People who see other arguments of merit in Clark's work, ought to put them forwards themselves free of the Social Darwinism.

[notsneaky]

Ay, this is the best and most accurate review of the book I've seen yet. For one thing, it's obvious that the person writing the review has actually read the book before opinin' on it, unlike many commentators and I suspect even few of the professional reviewers.

"I find the idea that this lead to some form of genetic or cultural (why cultural?) selection that triggered changes in something related to the industrial revolution as completely unsubstantiated in the book. Economists use the subjective rate of time preference as a parameter explaining behavior that they take as exogenous. I know of no work that has established whether this is a genetically or a culturally determined parameter or whether it would buy much in terms of triggering a transition to high growth."

I agree that the link here is very very weak. But it should be noted that there's actually two evolution related arguments.
One is that since the founding of agriculture rates of time preference have fallen. This would apply globally and the explanation for IR would be simply that for whatever reasons it happened slightly faster in England than in other places. As far as work that looks at that parameter in the evolutionary context see:
http://ideas.repec.org/a/aea/aecrev/v84y1994i3p460-81.html

which is cited in Clark's book (at least it was in the web edition).
I find this argument plausible in a general sense.

The second evolution related argument is that in England between 1200 and 1800 cultural or genetic evolution took place which set the stage for IR. There is some supporting evidence - fall in murder rates, increase in literacy - interspersed through out the book but this is where I agree that the link is tenuous.

[Per Kurowski]

I have NOT read the book but on the Darwinian argument and even if I am willing ex-ante to take Ricardo Haussman’s word on that “the idea that this lead to some form of genetic or cultural selection that triggered changes in something related to the industrial revolution as completely unsubstantiated in the book”, were it to be true we should perhaps ask us:

Does the generous procreation of the rich that spills over into the less fortunate development-path also include shortcuts as Angelina Jolie’s adoptions of poor children?

Although I haven’t seen the film that is opening up this weekend titled “The Nanny Diaries” either, would Clark’s book make a case for lifting all restrictions on the immigration of nannies, at least for families over a certain level of wealth, and so as to make it easier for these wealthy citizens to do their social duty?

[robertdfeinman]

I'll just mention one factor about considering evolution as a factor in determining human behavior.

The author makes two fundamental errors. First the speed at which the human race procreates means that there hasn't been enough time since 1200 for a genetic modification to have made a difference. A mutation doesn't occur in a class of individuals. It occurs in one individual. This organism would first need to spread this gene to others in his social circle and then this would then need to be passed on down to the "lower classes".

There has been some work recently on the gene which controls adult ability to digest lactose. In the African tribes studied it appears that this small change in an existing gene took at least 4000 years to propagate.

The second error is that genes which don't affect procreation are not preferentially inherited. A gene that makes a person wealthy (by whatever means) doesn't imply that this individual will have more offspring or that they will survive better. The only thing that matters from a Darwinian point of view is that the gene improve the chances of survival through the period of procreation and child raring. Other genes won't be preferentially selected.

Herbert Spencer didn't understand Darwin and has done a great disservice by his coining "survival of the fittest". Darwin's theory is much more subtle than that and that's why most people misunderstand it.

I recommend reading his "Origin of Species". Not only is it perfectly clear, but it is model of how scientific research is to be done. He not only explains his theory, but goes to great lengths to describe exactly which new discoveries would invalidate it.

[pat toche]

"Economists use the subjective rate of time preference as a parameter explaining behavior that they take as exogenous. I know of no work that has established whether this is a genetically or a culturally determined parameter..."

My understanding is that it is mostly a mathematically determined parameter to overcome the shortcomings of our imagination, as Frank Ramsey would, more or less, have it. Unless of course it is understood as a parameter of the "finiteness" of life, as Olivier Blanchard put it, in which case it is definitely genetic but certainly not exogenous and not constant...

[notsneaky]

robertf

I agree that 1200 to 1800 is too short of a time for evolution to have a meaningful impact (I guess there's some evidence that under some circumstances selection pressures can be strong enough so that it happens that quickly, but "could" is different from "did") and that's why I think this particular argument in the book is speculative at best. But again, it is actually only a small part of the book, perhaps overemphasized because it is both novel and controversial hence good publicity. A good portion of the book is just going into the details of how all pre industrial economies worked, the implications that emerge from certain basic facts about those economies and so on. Another big portion is comprised of the details of the IR and as Ricardo Hausmann notes above, knocking down (pretty successfully) existing theories of how IR came about.

On the other hand the argument that people got more patient since the invention of agriculture (when storage technology became available and made saving possible) is stronger since, it involves a much longer time period (11K + years) and also because there is an explicit argument - in the Rogers paper I linked to above - about why selection would favor more patient individuals (up to a point. Up to that ln 2 discount rate).

Anyway, the point is that there's a lot more to FTA then you'd think if you just read some of the reviews (in particular the NY Times one).

Pat t
The context in which Ramsey made the "failure of imagination" remark in regards to the time preferance was different. Ramsey set up his problem as that of choosing optimal consumption allocations across generations. Hence what he was talking was the SOCIAL rate of time preference (discounting future generation's utility relative to the present) not individual time preference. There did used to be some economists who argued that discounting the future was individually "irrational" but that was a fairly strict and arbitrary notion of rationality. So you can still take individual time preference seriously while agreeing with Ramsey that discounting the welfare of future generations involves a "failure of the imagination" (that position actually has its own problems but never mind ...)

[Arun Garg]

"Within the British empire (e.g. in India) property rights were secure, financial markets were pretty open and efficient and there was quite massive transfers of managerial know-how through out-migration of British managers and skilled workers. The slow spread of the industrial revolution ...is an important puzzle... It points ... to some other missing factor that is not a usual suspect."

I'm not sure why the Indian example seems so compelling to both Clark and Housmann ... surely the incentives of a colonized worker cannot be assumed to be similar to that of a free British worker?? How can property rights and openness of markets be perceived as guaranteed when the Indians saw discriminatory behaviour by the "sahibs" on a daily basis. It seems amazing to argue that conditions in India were essentially similar to Britain except for people's genes/culture, given the lack of democratic nature of the colonial experience.

In fact the AJR paper on colonial origins can be interpreted as implying that whenver the colonizers could not settle as a majority presence themselves they expropriated instead of trasnsplating institutions. The British were a miniscule minority in India, in terms of population (as compared to US or Australian colonies), and hence were much more inclined to expropriate rather than transplant institutions with a few exceptions (railways and the civil service).

[spencer]

I did not read the book, but didn't the same downward mobility occur in France, Germany, Italy, Spain, etc.?

If so, aren't we back to the point we were before, why did the industrial revolution occur in England and not the rest of western Europe?

[Aqdas Afzal (Pakistan)]

Arun Garg --

I would like to echo Arun's contention here. I think if we want to look at why in the presence of certain incentives (property rights etc.) India was not able to develop at a same pace as, say, UK, then we must not control for colonialism.

I think Colonialism is much too important a variable to ignore here. In the sense, that though the preferences of the colonized and the colonizer might be identical, their incentives, definitely, are not. In an environment of institutionalized bias against the natives, it would have been hard for native Indians to make use of the same incentives.

I think if we want to endogenize the lack of productivity (low TFP) in the Indian industries in the 18th and the 19th century then we should compare it with some other British dominated colony (Ghana, e.g.?).

[Bruce Wilder]

I am not one of those inclined to reject an evolutionary or genetic explanation out of hand. And, I'd have to disagree with rdf and notsneaky about 1200-1800 not being "enough time". Clark is not very specific about mechanism, but I don't think Clark has to be talking about a mutation in a single proto-industrial-mogul in medieval England propagating -- it seems to me that he's could be suggesting that a change in the proportions or prevalence of already common genetic characteristics might be a factor. For something characteristic, which is already present in, say 20% of the population, to go to being present in 40% of the population doesn't necessarily take dozens of generations or thousands of years.

The Enlightenment entailed a profound change in consensus reality, which has been well-documented. At various times in the 17th and 18th centuries, all over Europe, a generation prosecuting witches gave way to a generation interested in science. In Britain, this was mostly a 17th century phenomenon. A recent book on the Scottish Enlightenment illustrated it by contrasting a celebrated case of an unfortunate young man, who was executed for blasphemy, for making a joke, on a cold winter night, about wishing he could make a quick visit to hell to warm up, with the exquisite humanity demonstrated a generation later among the leading scholars of the Scottish Enlightenment.

These kinds of momentous changes preceded and laid foundations for the Industrial Revolution. Was there some correlation with something genetic? Or an inherited culture? It would not necessarily require a wholesale change in a generation, just a matter of a proportion reaching a critical tipping point. It is not as if religious fervor and superstition disappeared in the 17th century; they just lost their presumptive grip on the culture and on politics -- one doesn't have to search far to find modern Americans, who profess to believe in fundamentalist Chrisitianity or witchcraft or some combination. The Enlightenment was just a shift in presumptive consensus reality, underlay, perhaps, by a subtle shift in intensity of belief and numbers, whether those numbers count genes or cultural memes or some other analytic atom.

[robertdfeinman]

BW:
Aside from the shaky evolution arguments there is the implicit assumption that the aristocracy was the source of progress. They had more children, many of whom moved into the lower classes and their "superior" characteristics propagated down to the peasants.

This is not accurate demographically (see the comments on Mark Thoma's blog) and is just another form of the moral superiority of the wealthy argument.

Can a society rethink its goals and change its mindset? Of course. That's what Calvinism was all about, but why did the progress happen mostly in England?

I prefer to think it had something to do with the availability of water, coal and iron in easy to exploit forms.

Haven't we had enough of plutocrats yet? Shall we start talking about the divine rights of kings soon as well?

[Bruce Wilder]

Clark's a Tory, rdf, and by definition, hopeless.

That his "evidence" was provocative to others was rather more interesting, as were the feeble efforts to think in Darwinian terms in response. I include my own feebleness, as I tried several times to formulate a comment for the discussion on Thoma's site, and fell short.

Hausmann appears to be as mesmerized as many others: "in olden days incomes were stagnant. Second, in those days, as opposed to now, the rich had substantially more surviving children than the poor. This means that on average there was a rather strong downward social mobility that lasted a pretty long time. I think these are facts and are very convincing as such."

Are these "facts"? And of what do these "facts" convince?

I am at a loss.

I am an old institutionalist, so I like institutional explanations of historical change: change the institution, change history.

So, I naturally wonder how institutions affect this biological pattern.

The biological pattern is older than history -- social dominance and resource provision figure in the reproductive schemes of most primates, as far as I know, as well as many other sexually-reproducing, social species. In general, sexual reproduction in a social species can serve to keep alive and persistant a lot of genetic variation. A social division of labor can make a variety of phenotypes useful; sexual reproduction can bury currently disadvantageous characteristics in recessivity. Social dominance might enhance selective pressure in some circumstances, but it might also encourage, say, hybridization, or, in other circumstances, in-breeding; hybridization and in-breeding have very different implications with regard to selective pressure.

So, being an institutionalist, I tend to wonder about institutional arrangements: primogeniture comes to mind, with regard to wills. But, wills are mostly about men, and, therefore, putative fathers. I'd also be curious about where the women came from -- maybe women married up? Maybe, big daddy was cuckolded? Maybe, the third daughter got herself to a nunnery? Or the second son became an unmarried adventurer at sea, or a priest.

All very curious.

[pat toche]

notsneaky,
yes, you are correct, the context is different, but I still believe the main justification for exponential discounting (almost-clearly the kind of discounting R. Hausmann is discussing) is mathematical convenience. There is a fascinating empirical literature on impatience which rarely, as far as I know, supports exponential discounting -- patience is exercised in some contexts (call that counting, as opposed to dis-counting), non-exponential discounting of the hyperbolic variety has some empirical support, habit formation, Uzawa functionals, etc..

[notsneaky]

pat toche
I think there is pretty broad support for some kind of discounting. It might be hyperbolic but exponential discounting is just a special case of that.
Also there's just plain ol' introspection. Have you ever been impatient? And the fact that discounting has a natural interpretation as impatience means it's not just mathematical convenience. The same natural interpretation does not arise in inter generational discounting so there it's more suspect.

[reason]

I commented elsewhere that this is a very Eurocentric view of things. Why wasn't there a discussion of Roman times and the dark ages. (Suggesting the Bruce Wilder has a large part of the answer with his enlightenment suggestion.)

But what about Jane Jacobs view that it was urbanisation that mattered? The idea of rapid productivity in non-agriculture would surely support this.

[pat toche]

notsneaky,

I agree with you, it's not just mathematical convenience, but my guess is that it's mostly that. Exotic discounting such as hyperbolic discounting introduces difficulties -- such as time inconsistency -- which are conveniently avoided by exponential discounting.

I may be impatient to eat when I'm hungry, it doesn't mean I am discounting the future, but rather because (say) my marginal utility for food is high at that moment.

I'm a lot more convinced by the interpretation of exponential discounting as mortality risk than as an intrinsic human psychological trait, there's an awful lot of evidence that people eat the cherry on the cake last, you wouldn't expect them to if they were discounting exponentially...

anyway best to stop this here as you might get impatient with me...

[notsneaky]

The impatient-when-hungry example is interesting but rather particular.

[Per Kurowski]

Re Pat Toche: "people eat the cherry on the cake last, you wouldn't expect them to if they were discounting exponentially..."

Why not? You could be discounting exponentially and be aware of it and so you leave the cherry for last knowing that doing so you will be able to enjoy it without any exponentially discounting distractions… especially as when you say there is a whole cake under the cherry.

[Aditya]

Nice Post ! Thanks for the information shared.