Arnold Kling  

Is Economic Growth Genetic?

PRINT
Walter Block Reviews The My... Freeman Dyson on Global Warmin...

No, this is not another post on Gregory Clark. Enrico Spolaore and Romain Wacziarg write,


We find that measures of genetic distance have a statistically and economically significant effect on differences in income per capita, even when controlling for various measures of geographical isolation, and measures of cultural, climatic and historical differences. The effect of genetic distance holds not only for contemporary income differences, but also for income differences measured since 1500. Moreover, the effect of genetic distance holds not only for contemporary and historical worldwide income differences, but also for income differences within Europe. In fact, the magnitude of the effect of genetic distance is larger for a sample of European countries than across countries from all continents.

...On average, populations that are more genetically distant have had more time to diverge in a broad variety of characteristics transmitted intergenerationally. These include characteristics that are passed on through DNA, but also traits that are passed on not through biological reproduction but through cultural transmission.4 As long as these cultural characteristics are transmitted to younger generations from genetically related individuals within their population, they will be correlated with genetic distance.

Therefore, one should not view genetic distance as an exclusive measure of distance in DNA-transmitted characteristics, and should not assume that the mechanisms linking genetic distance to economic distance are necessarily genetic.


It's a weird paper. Their thesis is that innovations diffuse more easily among people who are close genetically, because of shared language, attitudes, and so on. What are the interesting alternative hypotheses?

1. Diffusion is related to geographic distance.

2. Diffusion is related to average IQ.

3. Diffusion is determined by institutions.

The authors explicitly take on (1), not so much (2) or (3). Do we observe market-oriented, high-IQ societies with slow adoption of innovation that can be traced to genetic distance from the innovation frontier? Do we observe centrally-planned, low-IQ societies with rapid adoption of innovation that can be explained by genetic closeness to the innovation frontier? Or are institutional characteristics, IQ, and genetic closeness to the innovation frontier too highly correlated to distinguish among these hypotheses?

Thanks to David Warsh for the pointer.

UPDATE: Jason Molloy points out in the comments that he already blogged about the paper. If you go to his post, make sure you check out the comments there. The comment that he makes in response to Wacziarg has a number of helpful links.


Comments and Sharing





COMMENTS (6 to date)
Jason Malloy writes:

We blogged on this in February. Romain Wacziarg even stopped by in the comments and explicitly stated that they found IQ too "scientifically flawed" to even consider.

I responded, of course.

Jason Malloy writes:

Thanks for the update. The linked post was actually written by my co-blogger, Herrick, not by myself.

tc writes:

Does this approach really tell us that much? I think that looking at high-level statistical correlations for explanation can be useful when looking at a "black box" i.e. we don't know much about the underlying mechanisms of something. But when it comes to industrial development, we have plenty of historical, low-level detail on how technologies spread (or don't) - eg Clark's book goes into how cotton textiles didn't take off in India, but did in Japan. Why did Japan react with speed and alarm to the appearance of Western technological power, while China seems to have blithely ignored it? Seems to me that will tell us more than looking at their genetic relatedness.

Garett writes:

tc:

China vs. Japan (or England vs. France, Canada vs. U.S., etc) is an important question, but there other important questions in economic growth as well. The Eurasian 'coasts' vs. the rest of the world is also an important question.

I guess you can summarize the two sets of questions as "Why did neighbors turn out differently" vs. "Why did regions turn out differently."

Wacziarg/Spolaore and Comin/Easterly/Gong's "Was the Wealth of Nations determined in 1000BC?" get us thinking about the latter question--which may not be of interest to everyone. They provide some evidence of path dependence, although they are largely mute as to the mechanism driving the path dependence.

www.nyu.edu/fas/institute/dri/Easterly/File/wealth1000bc.pdf

Romain Wacziarg writes:

Arnold,

Thanks for referring to this paper. A few comments on the 2 alternative stories we do not spend much time on:

1) IQ. It's not clear how IQ differences could act as a barriers to diffusion, unless you think cognitive differences across people impede understanding, etc. We're not fond of IQ as a determinant of economic performance because IQ is largely endogenous (the Flynn effect), and is also culturally/socially constructed. As Jared Diamond famously argued, hunter gatherers in Papua New Guinea are very smart when it comes to hunting and gathering - put one of us in this environment, and we would die quickly. They may be less smart at scoring highly on IQ tests because, frankly, they don't get much practice with rearranging triangles in logical sequence, etc. So yes, Jason Malloy is right, we do view IQ as conceptually flawed for the purpose of our research.

2). We completely agree that institutional differences can be a mechanism whereby cultural differences translate into barriers. It may be less easy for a civil law country to adopt innovations from a common law country for instance. This would be completely part of our story. Our work so far has been quite reduced form, but in more recent analysis we do find that institutional differences may well be an important mediating variable whereby common history (captured by genetic distance) may affect differences in economic performance. In other words, we don't view these stories as mutually exclusive. In our paper, historical separation (captured by genetic distance) is associated with differences in all kinds of vertically transmitted characteristics, institutional features among them.

Hope this helps clarify.

Jason Malloy writes:

In response to Dr. Wacziarg's comment:

We're not fond of IQ as a determinant of economic performance because IQ is largely endogenous (the Flynn effect), and is also culturally/socially constructed.


This is like saying 'I'm not fond of strength as a determinant of lifting heavy things because people have gotten stronger over time and strength is "socially constructed"'. It bypasses the important fact that, regardless of how changeable or not changeable strength is, strength IS the determinant of lifting heavy things.

Either something is or is not a determinant. The changeability of that something is a complete non sequitur.

Now IQ might not be "the" determinant of economic performance, but Dr. Wacziarg did not choose to make that argument. (Although, if he does then it will not be hard to point to a voluminous literature to the contrary - IQ is A very important determinant)

As Jared Diamond famously argued, hunter gatherers in Papua New Guinea are very smart when it comes to hunting and gathering - put one of us in this environment, and we would die quickly.


Note Jared Diamond simply made an assertion; his subjective impression of what would happen to a white infant raised by hunter-gatherers is worthless from a scientific standpoint.

Meanwhile two pieces of actual data contradict Jared Diamond. The first is an intelligence study among hunter-gatherers [*] that found three significant things: first of all the test factor structure matched those of people in developed countries (that is the results were not garbled or unfamiliar: the test showed a g factor of ability). Second there was a normal distribution of scores among the foragers. Third the intelligence results were consistent with rank and opinion within the band: those found with higher scores were also said to be more intelligent by the foragers themselves.

All three of these findings directly contradict the 'social construction of intelligence' narrative and demonstrate the test measured exactly what it purported to, even among the most undeveloped peoples.

Second, if Diamond was correct then forager children raised in adoptive Western homes should become as smart as or, indeed, smarter than their white siblings. They do not. [**]


They may be less smart at scoring highly on IQ tests because, frankly, they don't get much practice with rearranging triangles in logical sequence, etc.

1. Contradicted by both studies (and others)

2. Irrelevant since either way economically relevant human capital is being measured (if you can't play the guitar because you don't practice or because of you are genetically incapable, it doesn't make any difference to the fact that you still won't be paid by anyone to play the guitar)

3. Irrelevant since we are comparing modern nation-states with education systems, not primitive tribes

In conclusion, Dr. Wacziarg has not made a satisfactory case for why he and his coauthor "view IQ as conceptually flawed for the purpose of [their] research".

Of course IQ could be irrelevant to your research for a variety of perfectly justifiable reasons.


[*] Reuning, H. (1988). Testing Bushmen in the central Kalahari. In S. H. Irvine, & J. W. Berry (Eds.), Human abilities in cultural context. New York: Cambridge University Press.


[**] Dasen, P. R., de Lacey, P. R., and Seagrim, G. N. (1973). Reasoning ability in adopted and fostered Aboriginal children. In G. E. Kearney, P. R. de Lacey, and G. R. Davidson (1973). The Psychology of Aboriginal Australians. New York: Wiley.

Comments for this entry have been closed
Return to top