Diversity Galor

Quamrul Ashraf and Oded Galor argue that “the level of genetic diversity within a society is found to have a hump-shaped effect on development outcomes in both the pre-colonial and the modern era, reflecting the trade-off between the beneficial and the detrimental effects of diversity upon productivity.”  Africans have too much genetic variation, Amerindians too little, while Europe and East Asia are just right.

Of course, most genetic variation is neutral, having no significant effect on phenotypes, so the numbers they use are totally irrelevant to the question they’re addressing. One could imagine that it might be better to have more (or less) genetic variation in cognitive or personality traits, but we don’t know enough about the genetic architecture of those traits to say diddly about who has more or less.

Lots of people – not just Ashraf and Galor – seem to think that having more overall neutral variation implies more trait variation.  That isn’t the case.

A population with more total (mostly neutral) variation can easily have less variation in a particular trait.  For example, hair color and eye color ( both genetically controlled) are more variable in Europeans than in sub-Saharan Africa, even though African populations have more overall genetic variation.

There are particular populations in Africa that have extreme phenotypes – for example, Pygmies are shorter than anybody else – but that is a product of the special selective pressures they have experienced over a long time.  It is not as if the standard deviation of height is lots bigger in major African populations than elsewhere.

If we ‘re talking IQ, African-Americans pretty clearly have less trait variation: their standard deviation of IQ is about 12 points, rather than 15 in Europeans.

The difference in mean IQ is far more important – as it is globally.

 

 

 

 

 

 

 

 

 

Advertisements
This entry was posted in Economics, Genetics. Bookmark the permalink.

66 Responses to Diversity Galor

  1. “If we‘re talking IQ, African-Americans pretty clearly have less trait variation: their standard deviation of IQ is about 12 points, rather than 15 in Europeans.”

    Given that they’re an admixed population, shouldn’t we expect more trait variation rather than less?

    • Jim says:

      It amazes me that so many people can believe that a highly complex natural process with a huge random component like human evolution is likely to produce a simple pattern like say equality in cognitive ability between different ethnicities. A pattern that just happens to jibe with our moral sensibilities. The fact that so many, often well-educated, people seem to believe this shows that animism remains the fundamental outlook of even suppossedly sophisticated. people in the modern world. The world must be the result of gods and spirits motivated by good and evil.
      Of course the real world incuding humanity is a huge mess that makes no moral sense whatever. Just what one would rationally expect from a highly complex natural process
      whose fundamental laws are totally devoid of any moral significance.

      • The fourth doorman of the apocalypse says:

        It amazes me that so many people can believe that a highly complex natural process with a huge random component like human evolution is likely to produce a simple pattern like say equality in cognitive ability between different ethnicities.

        And yet pretty much all humans can talk, and those that can’t have pretty obvious problems.

        The fact that not all humans can manage integration in polar coordinates does require an explanation, IMO, even though it very clearly is a fact.

  2. Greg says:

    Ashraf and Galor in a recent document argue that “empirical evidence suggests that the prehistoric “out of Africa” migration process not only generated variation in the measure of neutral genetic diversity, but also created correlated shifts in many other dimensions of “functional” diversity, including diversity in (i) craniometric features (Manica et al., Nature, 2007; von Cramon-Taubadel and Lycett, American Journal of Physical Anthropology [AJPA], 2008), (ii) dental traits (Hanihara, AJPA, 2008), and (iii) pelvic features (Beti et al., Human Biology, 2012), and (iv) contemporary height (Galor and Klemp, mimeograph 2014).

    Thus, the “out of Africa” migration has indeed left its mark on many dimensions of “functional” diversity; a mark that persists to the present day despite tens of thousands of years of differential natural selection, mutation, genetic drift, and inter-ethnic admixtures.”

    • gcochran9 says:

      They don’t have the faintest idea what they’re talking about. Of course, I guess they’re no more wrong than typical economists who assume and even believe that average cognitive ability is the same in every population.

      If you were looking for an explanation of such differences, an ancient bottleneck is not the first thing that comes to mind – that would be different selective pressures, maybe over a long time but quite possibly entirely in the Holocene. It’s not the second possibility that comes to mind – which might be the effects of adaptive archaic admixture, which now has been seen to exist for some traits. It’s not the third idea that would come to mind – which might be differences in genetic load, say generated by systematically older fathers in some populations. There are a number of possible scenarios that might generate differential selective pressures: differences in pathogen pressure; differences in time depth of agriculture, or in the kind of agriculture; cold winters, etc.

      Populations in Eurasia, some of them, have larger brains, score higher on IQ tests than Africans, and have a far stronger record of cultural accomplishment. The ‘sweet spot of diversity’ hypothesis is saying that these higher cognitive abilities are the product of a settlement bottleneck – which is to say, drift. Chance. Well, that could happen, just as it is possible for all the air molecules to rush to one side of the room.

      • The fourth doorman of the apocalypse says:

        Well, that could happen, just as it is possible for all the air molecules to rush to one side of the room.

        However, I wouldn’t hold my breath waiting for it to happen.

      • Greg says:

        Mr. Cochran, unfortunately, it appears from your blog post that you have not understood the Ashraf and Galor hypothesis properly.

        Ashraf and Galor do NOT reject the possibility that variations in the mean level of traits across societies could have had a significant effect on comparative economic development. However, they have chosen to focus on the 2nd moment of the distribution of traits and to document the effect of diversity on comparative development.

        Moreover, the importance of diversity in their thesis is not necessarily due to the effect of diversity of the size of the upper tail of the cognitive distribution. Instead, their argument is partly based on the contribution of complementary of traits to the production of knowledge.

        As to your main thesis in the blog post, I think that it is inconsistent with the existing evidence in the field of evolutionary biology.

        There are two observations that are well documented in the evolutionary biology literature:

        1. Lower genetic diversity in microsatellites among indigenous populations at greater migratory distances from East Africa.

        2. Lower genetic diversity in some human characteristics among indigenous populations at greater migratory distances from Africa (e.g., craniometric features, dental traits, pelvic features, and height).

        Hence, as one would have expected logically, the serial founder effect operated on all genes and not only on microsatellites.

        Of course, unlike microsatellites that were selectively neutral, genes with phonotypical expression were subjected to the forces of natural selection and as you pointed out correctly, in some instances these forces may have dominated the forces of the serial founder-effect. However, a-priori from a probabilistic viewpoint, diversity in microsatellites would be expected to be correlated with diversity in phenotypes. (Note that as is the case in any other correlation, there are traits that would deviate from this pattern, but the correlation overall would remain positive).

        Moreover, suppose that indeed diversity in microsatellites would not be positively correlated with diversity in traits with phenotypical expression. If this would have been the case, one should not have expected to find any effect of genetic diversity on comparative development. But Ashraf and Galor find a highly statistically significant hump-shaped effect that is robust to a large variety of specifications, lending further credence to the hypothesis that diversity in microsatellites is indeed correlated with diversity in phenotypes.

        I hope to benefit for an academic, rather than the rhetorical, exchange.

      • YC says:

        I doubt you have understood their statistical argument. Sadly, you appear like a religious zealot to most readers.

        • gcochran9 says:

          A. The proposed cause – population differences in genetic variance

          B. The phenomenon: differences in the amount of within-population variation in psychological or cognitive traits (high in sub-Saharan Africans, medium in Europe and China, low in Amerindians)

          C. The proposed consequence: too much variation gets in the way of complex civilization, too little inhibits it

          There is no theoretical reason to expect A to cause B. A is a measure of overall genetic variation, not in genes that influence psychological traits. In fact we know of many counter examples (less overall genetic variation, but more in a given trait). .

          There is no evidence that B exists. There is a little direct evidence that it does not (IQ variance in US blacks, Europeans, East Asians is not as predicted). As far as I can tell, no one even has the anecdotal impression that differences of this kind exist. I asked Henry – when you spent time with the Bushmen, did you see anything that looked like greater psychological variation? Were their shy people shyer, their bold people bolder? Not that he noticed. I mention them because they’re probably the most heterozygous population on Earth.

          There is no evidence that B would cause C, even if B were the case.

          So where does that leave us? With a totally worthless hypothesis. But not just that: the economists in question, like most economists, know very little about the relevant disciplines – quantitative genetics, evolutionary genetics, psychometrics, etc. From this example (and, alas, not just this example) they’re not good at figuring out causation, either. That’s bad.

    • Matt says:

      including diversity in (i) craniometric features (Manica et al., Nature, 2007; von Cramon-Taubadel and Lycett, American Journal of Physical Anthropology [AJPA], 2008), (ii) dental traits (Hanihara, AJPA, 2008), and (iii) pelvic features (Beti et al., Human Biology, 2012)

      In terms of these things which Galor did *not* self publish, the context in which all these anthropologists have published is of these as relatively selectively neutral because they track genetic distance. They may be right (in fact I think they are right).

      Essentially there is a difference here between “Variation in selectively neutral traits (irrelevant to economic and technical progress) increases with genetic diversity” and “Variation in traits relevant to economic and technical progress increases with diversity”, and without the latter of these, the theory is a lot of waving hands and saying that *somehow* genetic diversity has an effect.

      Re: Galor (et al)’s thesis, to actually test it, you would need populations with the same selective conditions as Europeans and East Asians (who have rather different degrees of genetic diversity, if not either as high as Africans or low as Amerind/Oceanians), but different within population genetic diversity.

      Which don’t exist* (I don’t think, unless someone can point me out a natural experiment?). And Galor would likely beg the question by saying that they *can’t* exist (because the “optimum” level of genetic diversity would be required for that selective environment).

      In other words, rather an untestable theory, so likely junk.

      *In the world of insane power mad eugenics overlords, perhaps you could mash together pygmies (high diversity) and Papuans (low diversity). But we do not, thankfully, live in such a world.

      • AD says:

        Precisely as you requested, the empirical analysis Ashraf and Galor includes “regional fixed effects” and their estimates therefore are based on variations WITHIN regions/continents (e.g., within sub-Saharan African countries).

  3. Greg says:

    In the same document, Ashraf and Galor states: “Furthermore, in various complementary papers, Ashraf and Galor have presented evidence that neutral genetic diversity (as predicted by migratory distance from East Africa) has highly robust and statistically significant effects on:

    a. The degree of ethnic and linguistic fractionalization of a national population (Ashraf and Galor, AER Papers and Proceedings, 2013)

    http://ideas.repec.org/a/aea/aecrev/v103y2013i3p528-33.html

    b. The level of interpersonal trust within national populations (Ashraf and Galor, AER, 2013; Arbatli, Ashraf, and Galor, Working Paper, 2013)

    http://ideas.repec.org/p/bro/econwp/2013-15.html

    c. Heterogeneity in preferences for redistribution and political attitudes (Arbatli, Ashraf, and Galor, Working Paper, 2013)

    http://ideas.repec.org/p/bro/econwp/2013-15.html

    Overall, there is a wealth of evidence suggesting that migratory distance from east Africa has influenced functional genetic diversity and economic behavior.”

  4. The fourth doorman of the apocalypse says:

    I like the Goldilocks theory though!

  5. a very knowing American says:

    Wasn’t that Oded Galor’s sister or aunt or some relative in “Goldfinger”?

    • gcochran9 says:

      Why can’t you remember the name? Cat got your tongue?

      • The fourth doorman of the apocalypse says:

        How droll! I think it was a surfeit of felines that got his tongue.

      • a very knowing American says:

        Galor is a riot.

        • gcochran9 says:

          His Harvard critics are even funnier. Someday I have to get some of these people stinking drunk and see if I can figure out exactly why they think that emitting a lot of nonsense is such a good thing.

          I’ve seen smart people, boys who should know better, reject twin studies of IQ and personality, as far as I can tell purely because they didn’t like the implications.

      • Mario says:

        Indeed these Harvard critics (mostly anthropologists) reject scientific discoveries that appear politically incorrect. As posted below, Ashraf and Galor do NOT reject the possibility that variations in the mean level of traits across societies could have had a significant effect on comparative economic development. However, in an attempt to capture the trade-offs associated with diversity in the context of economic development they have chosen to focus on the 2nd moment of the distribution of traits and to document the effect of diversity on comparative development.

        • gcochran9 says:

          The problem is, the idea that genetic diversity is the key quantity to consider is just silly. It is not a reasonable way of explaining the observation that some populations in the world are considerably better at getting complex crap done than others. On the other hand, that set of Harvard critics thinks that you shouldn’t believe your lying eyes. Galor suggests a bad explanation, but his Harvard critics deny the facts and condemn anyone who mentions them. They’re part of a big societal movement – not floridly crazy as individuals, like someone who thinks that he’s Napoleon, but still crazy. More like the sort of people who actually believed whatever the current Soviet line was back in the 1930s.

      • Jim says:

        People are very prone to seeing the world according to moral categories. It ssems very difficult for many people to accept a totally amoral universe.

      • Greg says:

        Mr. Cochran , how can you argue that a factor that explains 1/6 of the variations in silly to consider and that only thing that matters is the mean level of IQ?

        Aren’t you committing the same sin that the anthropologist had?

        Ashraf and Galor find that variations in genetic diversity explain 1/6 of the variation in the disparity in income per capita across countries. Their empirical analysis includes “regional fixed effects” and their estimates therefore are based on variations WITHIN regions/continents (e.g., within sub-Saharan African countries). Hence, their findings will not be altered if one would introduce into the analysis variations in cognitive ability across regions/continents. These variations are already fully captured by the “regional fixed effects”.

        Note that in AG regression “regional fixed effects” explain a fraction of the remaining variation in income per capita across countries (i.e., some of the 5/6 that is unexplained by genetic diversity). It is plausible that some of this fraction can be attributed to differences in the mean level of cognitive trait across regions, along with geographical, institutional and cultural attributes. But 1/6 of the variation will still be explained by genetic diversity.

  6. The fourth doorman of the apocalypse says:

    The difference in mean IQ is far more important – as it is globally.

    La Griffe du Lion suggests that what is important is the smart fraction (the fraction over 105) but I guess that relates to the difference in mean IQ.

    • Sesquiterpene says:

      It could also be the stupid fraction: once some critical mass of law breakers or non-functioning individuals is reached, society as a whole is dragged down. This is what usually happened in most of my classes in school.

      But anyway, assuming similar within-group variance and normal distributions, mean IQ is tightly correlated with the fractions of those up or down and there’s no way to tell what exactly the important factor is.

    • Patrick L. Boyle says:

      I’ve been toying with the idea that the relevant figure is a bit higher. It seems to me that the watershed score is around 130. That’s about one or two percent of the population. It is two SDs out from the mean. Most CEOs and Presidents are about that smart. You really aren’t smart enough for most high offices unless you are that ’99th’ man. I don’t think anyone whose IQ isn’t 130 or so can make much of a contribution to most fields.

      This is in a sense an operational definition of ‘The Great Man’ theory of history.

      On the other hand there is really not much point in most professions in being much smarter than that. Other traits predominate. Among rulers good health and longevity were in ancient times paramount. Augustus and Ramses II were probably pretty smart but their main impact on history was through longevity.

      The big exception to the upper limit on the benefit of brains effect, is in the sciences and math. People with IQs about 130 or so don’t presume that they can actually understand how things really work. They learn to rely on the thinking of others. Al Gore is a good example. He’s smart enough to know that there is increasingly more and more CO2 in the atmosphere but he doesn’t pretend to be competent to build a model or understand the math. He quotes experts. Really smart guys quote themselves. (See any Cochran posting).

      If his IQ were another standard deviation above the mean he would assume that he could deal with all the issues of climatology. After all climatology, before all the recent controversy, was always considered a home for the weakest minds in science. If he were a bit smarter he would assume that he was himself an expert. He wouldn’t fear debate.

      • Positive effects of intelligence have no upper limit.

      • Bruce says:

        I think you’re basically right. I am about 2-sigma above the mean and work for a big defense company in a somewhat technical role. I am competent at my job but not an innovator. My friends there who have +3-sigma IQs are innovative types.

  7. Greying Wanderer says:

    Not sure if this is relevant but

    1) If an organism has been getting random mutations in the same selective environment for a long time wouldn’t the proportions of positive, negative and neutral change? For the sake of argument a random mutation in a half-evolved shark-like creature might have a 1/3 each chance of being positive, neutral or negative for sharkish-ness whereas once a perfect shark is fully formed the chance of a positive random mutation that would improve it drops towards zero. At that point a random mutation might be 50% neutral or 50% negative for example.

    So if an organism stays in the same selective environment for a very long time becoming more and more suited to that environment would you expect it to gather a lot of neutral dna?

    2) If (1) is correct (or maybe even if not) a second point that might follow is neutral where? In the same way dna that is positive in one environment might be neutral or negative in another (e.g. dark skin) dna that is neutral in one environment e.g. tropics, might be positive or negative in a different environment. dna that is suddenly negative in the new environment goes into the trash pile to be got rid of and dna that is positive in the new environment gets selected.

    So

    1) You might expect neutral variation to be reduced when moving from one environment to another as some of that once neutral dna suddenly becomes good or bad and is either selected for or against.

    2) If neutral dna in one environment can be positive or negative dna in a different environment then collecting large stores of neutral dna could be seen as stored pre-adaptation. For example imagine a population in one environment that regularly pumps out new forager bands into a different environment but they never have enough time to adapt to that new environment and always die off. However if that source population is continuously generating new neutral junk dna that just sits there in their genes then eventually they might generate the adaptations that would allow survival in the new environment.

    So for example say moving out of the tropics required longevity and menopause adaptations but also required that they exist *prior* to moving because there wouldn’t be enough time to develop them in situ.

    Also, separate thought

    3) If original point (1) above is correct and if neanderthals were adapted to the cold for a long time thus collecting lots of neutral junk dna and hominids in the tropics were doing the same i wonder if you could get a situation where some of the neanderthal junk dna and some of the hominid junk dna only become useful when someone got both?

    • minoritymagnet says:

      1) If an organism has been getting random mutations in the same selective environment for a long time wouldn’t the proportions of positive, negative and neutral change? For the sake of argument a random mutation in a half-evolved shark-like creature might have a 1/3 each chance of being positive, neutral or negative for sharkish-ness whereas once a perfect shark is fully formed the chance of a positive random mutation that would improve it drops towards zero. At that point a random mutation might be 50% neutral or 50% negative for example.

      So if an organism stays in the same selective environment for a very long time becoming more and more suited to that environment would you expect it to gather a lot of neutral dna?
      ==

      In reality, there is never a “same selective environment”. Pathogenes are always evolving generating new selective pressures. Even with no pathogenes etc., there is still intraspecific competition (between the sexes etc.) generating new selective pressures.

      Take living fossils (e.g. tuatura) whose phenotype did not change in centimillions of years. If you could fetch a tuatura from only a few million years ago, it would not be able to breed with the recent ones.

      • Greying Wanderer says:

        “In reality, there is never a “same selective environment””

        Yeah it’s over-simplifying but to the extent that it’s true – say dark skin in the tropics or the key sharkiness traits in a shark – once it’s done it’s done. Once random mutations that are positive for the unchanging bits of the environment are fixed then the chance of any later random mutation being positive must go down slightly and the chance of any later mutation being negative or neutral must go up slightly.

        If negative mutations are constantly selected against while neutral mutations are ignored wouldn’t that mean that a population which stayed in (mostly) the same environment long enough would continuously increase its neutral genetic diversity?

        (I’m not saying this is relevant to anything. Just wondering.)

  8. Pingback: Diversity Galor | Brain Tricks: Belief, Bias, a...

  9. dearieme says:

    When, in world history, was there a “pre-colonial” era?

  10. Greg says:

    Mr. Cochran, unfortunately, it appears from your blog post that you have not understood the Ashraf and Galor hypothesis properly.

    Ashraf and Galor do NOT reject the possibility that variations in the mean level of traits across societies could have had a significant effect on comparative economic development. However, they have chosen to focus on the 2nd moment of the distribution of traits and to document the effect of diversity on comparative development.

    Moreover, the importance of diversity in their thesis is not necessarily due to the effect of diversity of the size of the upper tail of the cognitive distribution. Instead, their argument is partly based on the contribution of complementary of traits to the production of knowledge.

    As to your main thesis in the blog post, I think that it is inconsistent with the existing evidence in the field of evolutionary biology.

    There are two observations that are well documented in the evolutionary biology literature:

    1. Lower genetic diversity in microsatellites among indigenous populations at greater migratory distances from East Africa.

    2. Lower genetic diversity in some human characteristics among indigenous populations at greater migratory distances from Africa (e.g., craniometric features, dental traits, pelvic features, and height).

    Hence, as one would have expected logically, the serial founder effect operated on all genes and not only on microsatellites.

    Of course, unlike microsatellites that were selectively neutral, genes with phonotypical expression were subjected to the forces of natural selection and as you pointed out correctly, in some instances these forces may have dominated the forces of the serial founder-effect. However, a-priori from a probabilistic viewpoint, diversity in microsatellites would be expected to be correlated with diversity in phenotypes. (Note that as is the case in any other correlation, there are traits that would deviate from this pattern, but the correlation overall would remain positive).

    Moreover, suppose that indeed diversity in microsatellites would not be positively correlated with diversity in traits with phenotypical expression. If this would have been the case, one should not have expected to find any effect of genetic diversity on comparative development. But Ashraf and Galor find a highly statistically significant hump-shaped effect that is robust to a large variety of specifications, lending further credence to the hypothesis that diversity in microsatellites is indeed correlated with diversity in phenotypes.

    I hope to benefit for an academic, rather than the rhetorical, exchange.

    • gcochran9 says:

      First, we have no evidence that the variance in psychological traits or cognitive abilities tracks overall genetic variation. There is no theoretical reason why that would have to be the case – and that’s a good thing, because we know several counterexamples in humans, traits (such as hair color, eye color, skin color, cold adaptations, etc) in which trait variation doesn’t track overall genetic variation. I would guess that there are many more such examples in other species.

      Second, we don’t know much about the variance of psychological traits or cognitive abilities in different populations – and the little we do know goes in the wrong direction. There seems to be a smaller variance for IQ in US blacks than in Whites, and about the same in east Asians as in Europeans, even though Africans have more genetic variation than Europeans, while east Asians have less. I mention these two case because they’re the only ones I know: we don’t have a lot of info on what the variance of IQ is in different populations.

      Third, a priori, why would anyone think that differences in variance in, say, IQ, were more important, or comparably important, to differences in the mean? Moderate differences in the mean have drastic effects on the fraction capable of high-complexity tasks.

      I can of course imagine a situation in which differences in variance would drive the story, but that situation is far from anything seen in humans. Imagine a population of aliens, fairly similar to humans, with an average IQ of 120 (higher than any existing population) and zero variance. They ALL have an IQ of 120. There are tasks that they would be better at than any existing human society, and there would also be accomplishments that they would be utterly incapable of. On the other hand, imagine a society with a mean IQ of 80 and a standard deviation of 60. 2% of that population would be smarter than anyone, most would need to live in a sheltered workshop. Some could probably do things we couldn’t imagine, but most would have trouble tying their shoes.

      Are there any human populations like those two examples? Even somewhat? No there are not. Sub-Saharan Africans may have more genetic variation, but do not produce a surprising high number of ultra-geniuses. Judging from measured accomplishments in fields like math, they’ve never produced a single one. Which is about what you’d expect from the difference in the mean.

      I understand that Galor can’t talk about population differences in average IQ, because he’d be fired – but since they are the real story, he should pull a Wittgenstein.

      • Greg says:

        Ashraf and Galor find that variations in genetic diversity explain 1/6 of the variation in the disparity in income per capita across countries. Their empirical analysis includes “regional fixed effects” and their estimates therefore are based on variations WITHIN regions/continents (e.g., within sub-Saharan African countries). Hence, their findings will not be altered if one would introduce into the analysis variations in cognitive ability across regions/continents. These variations are already fully captured by the “regional fixed effects”.

        Note that in AG regression “regional fixed effects” explain a fraction of the remaining variation in income per capita across countries (i.e., some of the 5/6 that is unexplained by genetic diversity). It is plausible that some of this fraction can be attributed to differences in the mean level of cognitive trait across regions, along with geographical, institutional and cultural attributes.

  11. mindfuldrone says:

    Re Pygmies. At least some of the height variance as compared to neighbors appears to be to fast life history trajectories

    Migliano, A. B., Vinicius, L., & Lahr, M. M. (2007). Life History Trade-Offs Explain the Evolution of Human Pygmies. Proceedings of the National Academy of Sciences of the United States of America, 20216-20219.

    • gcochran9 says:

      You know, I have no idea what you mean when you say ‘fast life history trajectories’. I could believe that under conditions of high mortality, pygmies were selected for a fast life history, which involved faster maturation and being shorter. But they would be shorter because of genetic variants that were favored, over the long term, by that selective regime. The difference in height would indeed be entirely the result of genetic factors. And we’re not talking about ‘variance in height”, we’re talking about a difference in mean height. Probably the biggest influence on height variation among the Pygmies is the degree of Bantu admixture.

      I don’t believe that life history argument, but I could, if I saw some more convincing evidence. That would be one of a number of possible selective explanations of short people. I do believe that selection somehow favors the Pygmy phenotype in tropical jungles, because it has evolved independently several times.

      • mindfuldrone says:

        Well, my reading of the Migliano paper was that high infant mortality was the cue for the live fast die young (and leave a short corpse) adaptation in pygmies. Presumably this would only explain some of the variance–but morphological response to early life cues can be pretty considerable. The classic example would be Coho salmon who go down a fast and sneaky small body (Jack Salmon) or large and fighty one (Hooknose salmon). They were thought to be different species but they are genetically identical. Gene expression varies massively between them, of course.

      • mindfuldrone says:

        Sorry–should have linked key conceptual paper for LHT–
        http://www.jstor.org/stable/2825234
        Stearns, S. C. (1976). Life-history tactics: a review of the ideas. Quarterly review of biology, 3-47.

      • The fourth doorman of the apocalypse says:

        But they would be shorter because of genetic variants that were favored, over the long term, by that selective regime.

        What I think Greg is saying here, and he can correct me if I am wrong, which would be a useful service, is that using life history theory as a catch-all explanation does not help.

        There are multiple phases to an organism’s life history (perhaps 5 or so in humans) and the timing of each of these is controlled by genes. There will be variance in the timing of these in individuals, because of genetic variance (different alleles) and specific environmental conditions will select those variants that are more effective in that environment. In additional, I suspect, as I think Greg is saying, that height is somewhat orthogonal to the timing of various life-history stages and will be selected orthogonally to it.

        We need to get beyond generalities and onto specifics, although in some cases I imagine we are yet to understand the genes that cause certain factors.

  12. SocialScientist says:

    The statement “the difference in mean IQ is far more important – as it is globally” is overly bombastic and rather baseless from the viewpoint of serious academic inquiry in the social sciences (particularly, economics) because there are severe issues here with empirical falsifiability. First, as is well known from controversies surrounding the work of Lynn and Vanhannen, the measurement of IQ is extremely noisy and marred by systematic bias that makes the data on different populations incomparable. Second, even if this were not the case, the mean IQ of a population is co-determined with that populations socioeconomic outcomes (i.e., there is bi-directional causality and there could be numerous latent variables that could generate a spurious relationship between mean IQ and economic outcomes at the population level).

    Regarding the statement “if we’re talking IQ, African-Americans pretty clearly have less trait variation: their standard deviation of IQ is about 12 points, rather than 15 in Europeans,” this too is empirically problematic due to the simple fact that African-Americans are primarily descended from a population (those that were subjected to the trans-Atlantic slave trade) that is NOT representative of the original source populations in Africa – i.e., there will undoubtedly be sample-selection bias when considering only a “special” subset of the original source population, a subset that is conditioned on empirically unobserved personal traits that potentially increased the likelihood of being selected into slavery. In addition, as is well-documented in the historical record, African slaves in the Americas were subjected to different socioeconomic conditions (and, thus, different selective pressures) in comparison to their source populations whose descendants continue to reside in Africa. As such, barring the serious issues with IQ measurement for the moment, the claim that lower IQ variance amongst African-Americans (in comparison to Europeans) runs contrary to what we might expect from a serial founder effect is very, very far from being conclusive.

    • So in a way you are acknowledgeing the hereditarian position with respect to African-American IQ, since you believe (a) African-Americans’ descendants were negatively selected out of the African population and (b) slavery produced what amounts to dsygenic effects on the slaves. Is that correct ? I will leave the dysgenic effect to Cochran.

      (1) Even L&V acknowledge environmental deficits lower IQ in Africa. They never argued that their correlations were between outcomes and “genotypic IQ” (loosely defined as the maximum potential IQ under optimal environments).

      (2) If African-Americans represent descendants of a negatively selected segment of the African population in the slave trade era, which is plausible, then that would imply African mean “genotypic IQ” would be higher than African-Americans’. Since African-Americans’ measured IQ is ~1std higher than the SSA measured IQ, then that implies Sub-Saharan Africans must have more than a standard deviation deficit in potential IQ. How much more, do you believe ? Two standard deviations ?

      Short of severe head trauma can something environmental depress IQ by that much ?

      • correction : you believe African-Americans’ ANCESTORS were negatively selected out of the African population

      • Jim says:

        Lynn estimates Sub-Saharan African “genotypic IQ” at about 80. I’ve seen a figure for the IQ of the US Virgin Islands as 78. I assume living conditions in the US Virgin Islands are way above muost of Sub-Saharan Africa.

      • Jim says:

        One would think that the Atlantic passage would have selected for a strong physical constitution.

    • Greying Wanderer says:

      “the measurement of IQ is extremely noisy and marred by systematic bias that makes the data on different populations incomparable”

      So you’d be in favor of collecting the data properly so we could make sure?

    • Anonymous I says:

      “First, as is well known from controversies surrounding the work of Lynn and Vanhannen, the measurement of IQ is extremely noisy and marred by systematic bias that makes the data on different populations incomparable.”

      Sorry, no. The measurement of IQ is sufficiently accurate to compare different populations, and that PISA data and other indicators reveals no important sources of bias that would get in the way of the obvious conclusions. We may very well question whether the number for any one country is accurate, yes, but the aggregate data paint a very clear picture. Sorry if this seems bombastic; it’s just a statement about what is generally known among social scientists (except maybe for you).

      “Second, even if this were not the case, the mean IQ of a population is co-determined with that populations socioeconomic outcomes (i.e., there is bi-directional causality and there could be numerous latent variables that could generate a spurious relationship between mean IQ and economic outcomes at the population level).”

      Of course. Yet, bi-directional causality would tend to magnify rather than confuse differences, particularly when aggregating data from multiple countries. Further. we can see dramatically different environments prevailing in, say, China and the United States, yet Han Chinese score similarly in each country, even generations later. So the arrow of causality doesn’t seem to run very strongly from other variables to IQ.

  13. Steve Sailer says:

    I wrote about Ashraf and Galor’s paper here:

    http://isteve.blogspot.com/2012/10/the-latest-car-crash-in-trendy.html

    By the way, Spolaore is a lot better. He wants to make sure he doesn’t get lumped with A and G.

  14. Mario says:

    Ashraf and Galor responded recently to common criticisms raised by scholars as well as uniformed superficial bloggers

    https://docs.google.com/a/brown.edu/file/d/0Bw0xeVVYEG5XSGxDTms4ZllCb2M/edit

  15. Greying Wanderer says:

    “Ashraf and Galor find that variations in genetic diversity explain 1/6 of the variation in the disparity in income per capita across countries.”

    If variations in genetic diversity are related to distance from east Africa then they’re a partial proxy for latitude.

    • Ehud says:

      Ashraf and Galor account for the potential confounding effect of latitude and other proxies for distance from east Africa, including aerial (rather than migratory) distance from Africa, and placebo origins in London, Mexico City and Tokyo.

      http://ideas.repec.org/p/bro/econwp/2010-7.html

      • Greying Wanderer says:

        so a population with high genetic diversity but no cold adaptations would be better off in a cold region than a population with low genetic diversity but lots of cold adaptions.

        makes perfect sense.

      • AD says:

        Where were you trained Greying Wanderer? The degree of ignorance among the participants of this blog is alarming. Making bombastic statements with literally no understanding of basic statistical methods.

        FOR A GIVEN CLIMATE there is an optimal level of diversity. Ceteris Paribus, my dear.

      • Sandgroper says:

        If you don’t like it, don’t read it, you moron.

      • Harold says:

        “The degree of ignorance among the participants of this blog is alarming.” Maybe, but you should see the degree of ignorance among people who are not participants of this blog. That’s really alarming!

      • georgesdelatour says:

        AD

        “FOR A GIVEN CLIMATE there is an optimal level of diversity.”

        Tell us more. What is their theory then? Low genetic diversity good in Nordic countries, higher diversity good at the equator? I’m trying to understand it.

        Finland has a cold climate. It’s very wealthy. It has the best PISA results of any country outside East Asia. And it’s genetically very non-diverse, but becoming more so. Does Galor make any predictions about how that increased (mostly Somali origin) diversity will affect PISA results?

        Singapore lies on the equator. It’s very wealthy, with excellent PISA results. It’s more diverse than Finland (77% Chinese, 15% Malay, 7% Indian). But seriously, which comparison makes more sense? Comparing it with another Han Chinese majority island at a much higher latitude like Taiwan (98% Han Chinese), or with a nearby equatorial state like Papua New Guinea?

  16. BurplesonAFB says:

    Most people are incredibly sloppy thinkers. For these people, all notions of “diversity” blend into one, from the incredibly diverse genes of the vibrant and wise Africans to multiracial coworkers going for thai food, it’s all diversity and it’s a synonym for good.

    The ultimate aim of Newspeak was to reduce even the dichotomies to a single word that was a “yes” of some sort: an obedient word with which everyone answered affirmatively to what was asked of them. So the ultimate aim of newspeak was duckspeak – articulated speech issue from the larynx without involving the higher brain centres at all.

  17. Greying Wanderer says:

    @AD

    “FOR A GIVEN CLIMATE there is an optimal level of diversity. Ceteris Paribus, my dear.”

    1) If all the actual important stuff is equal maybe so.

    2) When you put the theory like this:

    “so a population with high genetic diversity but no cold adaptations would be better off in a cold region than a population with low genetic diversity but lots of cold adaptations”

    it sounds silly.

    Now it might not be silly if a population with a lot of spare part dna that was neutral in their original environment moved to a different environment because the probability would be higher that some of the spare part dna that was neutral in the old environment might be positive in the new one but then the critical point would be the presence of dna that was beneficial in the new environment not neutral dna itself.

    .

    “Ashraf and Galor find that variations in genetic diversity explain 1/6 of the variation in the disparity in income per capita across countries.”

    How much does latitude explain on its own?

    .

    It seems more likely to me that large amounts of neutral genetic diversity is a sign of relatively low levels of recent strong selection and low levels of neutral genetic diversity would be a sign of high levels of recent strong selection and high levels of recent strong selection would imply being more optimal for the environment.

    So population A strongly selected for their home environment for a long time and then hitting diminishing returns and collecting large amounts of neutral diversity aka spare part dna.

    Various population Bs who move from A into different environments with the spare part dna from A some of which turns out to be beneficial (or deleterious) in the new environment and then get strongly selected for those new environments losing a lot of their spare part dna in the process.

    Interesting question then is what happens if you have a particular population B who are honed for a new environment – losing a lot of the spare part dna they inherited from A in the process – who then emerge into a new environment beyond that.

    A population C like this (e.g. Amerindians?) strongly selected for a harsh environment – e.g. Bering Strait – so a prime example of a population being honed and maybe losing a lot of their spare part dna in the process but then emerging on the other side into a much different environment (or set of environments) without all the spare part dna they started with. So instead of entering a new environment carrying a ton of spare part dna (some of which turns out to be positive in a different environment) they have to re-evolve it all?

    Just a thought.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s