Saturday, May 10, 2014

Academic racism has a K=N problem


Let's use the term "academic racism" to mean "“a belief that race is the primary determinant of human traits and capacities and that racial differences produce an inherent superiority of a particular race" (the Merriam-Webster "full definition" of "racism"), in order to differentiate it from bigotry (the common-use definition).

Anyway, science writer Nicholas Wade has a new book out making the standard case for academic racism. Andrew Gelman, a statistician, has a review of that book in Slate. The review is good, and you should read it, but I thought I'd try to restate Gelman's point in a slightly more compact way.

Basically, academic racism has a problem, and that problem is overfitting.

Here's how academic racism generally works. Suppose you see two groups that have an observable difference: for example, suppose you note that Hungary has a higher per capita income than Romania. Now you have a data point. To explain that data point, you come up with a theory: the Hungarian race is more industrious than the Romanian race. But suppose you notice that Romanians generally do better at gymnastics than Hungarians. To explain that second data point, you come up with a new piece of theory: The Romanian race must have some genes for gymnastics that the Hungarian race lacks.

You can keep doing this. Any time you see different average outcomes between two different groups, you can assume that there is a genetic basis for the difference. You can also tell "just-so stories" to back up each new assumption - for example, you might talk about how Hungarians are descended from steppe nomads who had to be industrious to survive, etc. etc. As new data arrive, you make more assumptions and more stories to explain them. Irish people used to be poor and are now rich? They must have been breeding for richness genes! Korea used to be poorer than Japan and is now just as rich? Their genes must be more suited to the modern economy! For every racial outcome, there is a just-so story about why it happened. Read an academic-racist blog, like Steve Sailer's, and you will very quickly see that this kind of thinking is pervasive and rampant.

There's just one little problem with this strategy. Each new assumption that you make adds a parameter to your model. You're overfitting the data - building a theory that can explain everything but predict nothing. Another way to put this is that your model has a "K=N" problem - the number of parameters in your model is equal to the number of observations. If you use some sort of goodness-of-fit criterion that penalizes you for adding more parameters, you'll find that your model is useless (no matter how true or false it happens to be!). This is one form of a more general scientific error known as "testing hypotheses suggested by the data", or "post-hoc reasoning". It's a mistake that is by no means unique to academic racism, but instead is common in many scientific disciplines (cough cough, sociobiology, cough cough).

Gelman explains all this in layman's terms.

None of this is to say that academic racism is completely wrong, or is a useless way of looking at the world! It might get certain things right. But we don't really know (yet). We don't have a good understanding of the proposed causal mechanism, and we don't have good natural experiments to test any of the hypotheses. Suppose academic racism gets half of its hypotheses right (suppose Romanians really do have a "gymnastics gene" that Hungarians lack), but gets half of them wrong (suppose Hungarians have higher per capita GDP because of better institutions and proximity to German markets, not because of an "industriousness gene"). How would we tell the difference between the right parts and the wrong parts? 

We couldn't.

Academic racism is very alluring, for at least three reasons. First, it tells us that all our stereotypes and prejudices are basically right - and we humans like to be told that all our preconceptions are right. We suffer from confirmation bias. Second, academic racism feels cool and edgy and rebellious, because political correctness still often banishes it from the realm of acceptable discourse. It's fun to feel like the scientific rebel, fighting for The Facts against the thought control of The Establishment. And third, academic racism provides a convenient excuse for racism of the non-academic kind. Scared that a big, masculine black guy will take your girlfriend? Worried that hard-working, intelligent (but "uncreative") immigrants will take your job? Academic racism provides convenient stories to justify policies that protect you against threats like these - at the expense of the black guys and the immigrants, of course.

This is why even though academic racism could be right about some stuff, I just roll my eyes whenever I see it. "Race" is a kind of phlogiston that can be invoked to explain pretty much anything (kind of like "culture" and "technology" and "institutions" in economics!). Until we have a much better understanding of genetics, the infinitely proliferating hypotheses of academic racism will be neither proven nor disproven, and people will go on believing in them (or disbelieving them). The theory that Jeopardy champion Arthur Chu called a "maggoty corpse" will shamble onward, never dying, never truly alive.

218 comments:

  1. The slightly paradoxical combination of rejecting mainstream beliefs while still confirming what you already believed seems to motivate a wide variety of questionable logic. It seems to often manifest itself in a conservative way, since the paradox can happen when institutional beliefs progress faster than "common sense," but it happens at all sorts of positions on the political spectrum in addition to apolitical issues. A certain amount of slate-pitchy contrarianism is motivated by a similar principle.

    ReplyDelete
  2. A couple of key points on why this makes very little sense:

    1. We know that all these traits (indeed all human traits) are heritable.

    2. Genetic evidence clearly distinguish between human groups on the level of ethnicity or even region.

    3. There are measurable differences in stable physical and psychological traits between groups.

    4. It does seem to appear that the genetic-environmental architecture of behavioral traits is similar for groups as it is for individuals.

    5. We have historical evidence for some selective pressures that could be behind many of these group-wide differences.

    These points, especially points 1-3, hardly make the idea of group-wide differences "just-so" stories. Indeed, quite the opposite, as opposed to the various "cultural" and "environmental" explanations that have been put forward to explain them. The idea that biology is behind these differences makes several predictions that do hold up with the evidence (e.g., clustering of traits by groups, stability of traits over time, recurring patterns world wide).

    ReplyDelete
    Replies
    1. You still run into the standard social science problems of reflexive relationships between individuals and environments, and "stable" behavioral expressions that represent more complex mixtures of underlying factors. The "Just-So" story comes in when a Charles Murray treats awareness of race as a neutral factor.

      Delete
    2. We know that all these traits (indeed all human traits) are heritable.

      What an alien thought process.

      Delete
    3. "What an alien thought process."

      Well, that's persuasive!

      Delete
    4. Anonymous10:36 AM

      "What an alien thought process."

      Well, that's persuasive!

      [Maybe it was a compliment. I have often suspected that Noah is an alien.

      OK, seriously! BF Skinner went overboard in the other direction a bit much perhaps, but his work on group behavioural dynamics as well as operant conditioning up to the point of the battery and the alligator clips was relevant. Even Pavlov's dogs would bark at "all human traits are heritable." But I guess it is just another case of monkey see, monkey do.

      Delete
  3. I roll my eyes at it, too, especially since I also read a lot of American history books and thus see all of the old racial descriptions of the past. Remember when Jews were supposedly genetically disposed to being awesome at basketball? And Japanese were sneaky and predisposed to night attacks, and would thus never do something like attack Pearl Harbor in broad daylight?

    ReplyDelete
  4. Anonymous8:17 PM

    Scared that a big, masculine black guy will take your girlfriend? Worried that hard-working, intelligent (but "uncreative") immigrants will take your job? Academic racism provides convenient stories to justify policies that protect you against threats like these - at the expense of the black guys and the immigrants, of course.

    Territoriality is adaptive, period. Regardless of "academic racism". That's why all genetic organisms exhibit it. Bears aren't territorial because they've been reading "academic racism" online.

    Organisms that aren't territorial are quickly replaced by those that are. That's why when we see an organism with suppressed territoriality, that is behaving strangely, we generally infer that it has suffered a parasitic infection of some kind that is suppressing its territorial instincts.

    ReplyDelete
    Replies
    1. Anonymous11:23 PM

      Oh, so the womenfolk are your territory are they? Way to go, racism and sexism all in one bundle. Sheesh!

      Delete
  5. So Noah do you have the same criticisms of cross-country regressions in the empirical growth literature too, or are your criticisms based on overfitting confined to what you call "academic racism" ?

    "Here's how academic racism generally works. Suppose you see two groups that have an observable difference: for example, suppose you note that Hungary has a higher per capita income than Romania. Now you have a data point. To explain that data point, you come up with a theory: the Hungarian race is more industrious than the Romanian race.

    Actually this does not happen -- at least not with actual academics. They are far more likely to treat IQ as the go-to metric that explains a lot of the variance in social phenomena, not come up with ad hoc theories. For example, this contains cross-country regressions using IQ and just a few controls. But Garett Jones does a much better job treating it as a metric of human capital.

    Of course in labour economics, where Hanushek for exmaple has used IQ in modelling educational & related outcomes you have more than countries as data points.

    "This is one form of a more general scientific error known as "testing hypotheses suggested by the data", or "post-hoc reasoning"

    But IQ is the very opposite of this. IQ wasn't constructed to test any pattern suggested by GDP per capita or household incomes.

    "We don't have a good understanding of the proposed causal mechanism, and we don't have good natural experiments to test any of the hypotheses.

    Have you never heard of twin & adoption studies ?

    "You still run into the standard social science problems of reflexive relationships between individuals and environments"

    Have you never heard of twin & adoption studies ?

    "The "Just-So" story comes in when a Charles Murray treats awareness of race as a neutral factor"

    The only really fair criticism I've heard of Charles Murray was James Heckman's, that Murray shouldn't have proxied skill with g in his regressions in the book. By calling it fair I'm not saying Heckman was right, just that it was a reasonable objection.

    ReplyDelete
    Replies
    1. "IQ wasn't constructed to test any pattern suggested by GDP per capita or household incomes."

      No, IQ was constructed by Social Darwinists to give an academic sheen to their theories. IQ tests of the Stanford-Binet type test "verbal ability" (which to the original designers meant understanding of written Standard English) and mathematical reasoning (which even if you are a born polymath depends upon continuous, high quality education from an early age). They applied these tests to urban educated whites, poor rural whites, African=Americans, and recent immigrants. Guess which group came out on top! And I love the propaganda value of calling this system "Intelligence Quotient" rather than "Experience with the American Urban School system", which is what it really tested.

      Though they deny it, the makers of the SAT modeled their test on the Stanford-Binet IQ testing system. And thus they perpetuated the Social Darwinists argument by using their system.

      Delete
    2. One word: Chinese. Their existence and median IQs completely refute your assertions here.

      Delete
    3. The above comment of mine was in reply to Shantanu Saha.

      Delete
    4. "Have you never heard of twin & adoption studies?"

      I have. Have you heard of how few examples of twins adopted and raised separately, and then subjected to methodical behavioral analysis as adults, there actually were? The time window between "hey, these twins who were adopted out and raised separately can provide a valuable scientific resource" and "we shouldn't separate these twins when we're adopting them out" was really quite small.

      Delete
    5. "So Noah do you have the same criticisms of cross-country regressions in the empirical growth literature too, or are your criticisms based on overfitting confined to what you call "academic racism" ?"

      Cross country growth regressions did have a very similar problem (not quite as bad, but of the same kind) which is why Sala-i-Martin went and rightfully killed that whole literature in one master stroke. With some help from Dani Rodrik, who kicked it a few times as it was collapsing to the ground, albeit from a different angle.

      Delete
    6. You wouldn't get any objections from me on this score.

      Delete
    7. @pithom I'm 1st generation Indian. My IQ probably blows yours away. Just because recent Asian immigrants have managed to game the system (since the majority of them come from the higher-educated segments of their home countries and they emphasize education to their children) doesn't make the critique any less valid. And having 2nd and 3rd generation Chinese and Indians in my classes, I don't see much of a difference between them and the average dumb American. 1st generation immigrants, yes they are more dedicated to doing well in school, and perform well on standardized tests (including IQ and SAT).

      Delete
    8. marcel11:03 AM

      Well, I'm of ashkenazy jewish ancestry, so in all lkelihood, the size of my schlong blows yours (i.e. SS's) away... jes saying.

      The rest of your comments were interesting and on point, but the one about your own IQ... well, the less said, the better.

      Delete
  6. Owen Arluine9:42 PM

    So why not apply this same critique to the cultural hypothesis?

    "The reason Africa did worse than Europe is because of colonialism".

    But Brazil and India were colonized and did much better than Africa.

    "The reason Latinos make less money than whites is because they are recent immigrants who arrived with nothing and look very different from the majority."

    But Asians are recent immigrants of that sort and they make more money than whites.

    "The reason there are few blacks in mathematics is that the mathematical community is racist."

    But there are far more blacks in literature. Why isn't the literary community racist?

    I'm not necessarily saying those are the only or the strongest arguments the cultural hypothesis can throw up, just that the cultural hypothesis is doing exactly what you accuse the biological hypothesis of doing here - seeing a difference and throwing up a story to explain it.

    It's true that every time two groups are different, you can say "Well, I bet there's a genetic theory that explains the difference. What about Gene X?". It's also true that every time two groups are different, you can say "Well, I bet there's a cultural theory that explains the difference. What about Social Factor Y?"

    Remember, people think of reasons they are allowed to reject theories they don't like, reasons they are allowed to believe theories they support. You've come up with a good reason for rejecting any story of intergroup differences, yet you only use it to reject the one you don't like, thus making the one you like proven by default. Isn't that interesting?

    I'm also intrigued by your assumption that everyone is just running around looking for reasons to justify their racism. Aside from the sort of people who read economics blogs (I don't understand it either) there seems to be a COLOSSAL PREPONDERANCE of people who hate racism and will jump at any straw to discredit it. According to the last survey, 90% of college professors in elite universities identified as liberal. Shouldn't that make you at least a little concerned about whether there's straw-grasping-at to defend the universally-accepted pillar-of-our-civilization disagree-with-and-get-fired hypothesis, in addition to your Extreme Concern that maybe there is some straw-grasping at to support the super-controversial universally-loathed hypothesis?

    But all that having been said, you're kind right. Wade is grasping at intriguing but unproven possibilities. That's because the facts he could have used to build an obvious, incontrovertible case are too controversial even for him. If you want to argue against your strongest opponents - instead of the deliberately weakened form that can get published - I challenge you to write a blog post on your reactions to http://occidentalascent.wordpress.com/2012/06/10/the-facts-that-need-to-be-explained/

    ReplyDelete
    Replies
    1. So why not apply this same critique to the cultural hypothesis?

      Not a regular reader, eh? ;-)
      http://noahpinionblog.blogspot.com/2012/04/in-which-i-shoot-down-culture-fairy.html
      http://noahpinionblog.blogspot.com/2013/04/can-culture-predict-economic-development.html

      I'm also intrigued by your assumption that everyone is just running around looking for reasons to justify their racism.

      Not everyone. Didn't I give three different reasons why academic racism might be appealing above and beyond its informational content? ;-)

      there seems to be a COLOSSAL PREPONDERANCE of people who hate racism and will jump at any straw to discredit it

      Why do you think that is the case? :-)

      Shouldn't that make you at least a little concerned about whether there's straw-grasping-at to defend the universally-accepted pillar-of-our-civilization disagree-with-and-get-fired hypothesis, in addition to your Extreme Concern that maybe there is some straw-grasping at to support the super-controversial universally-loathed hypothesis?

      Oh, I'm not concerned about anything in this area. Yet.

      I challenge you to write a blog post on your reactions to http://occidentalascent.wordpress.com/2012/06/10/the-facts-that-need-to-be-explained/

      Sorry, but no. This is just not concisely written. Apologies if this sounds harsh, but I don't have time to go down this rabbit-hole and immerse myself in stuff that I think is almost certain to be beside the point. My advice: Write something short and pithy, and if it's intriguing, THEN I'll be enticed to read and think about the long thing.

      Delete
    2. Owen Arluine8:29 PM

      You're right, I haven't been reading your blog for two years and I hadn't seen the culture fairy post before. But that doesn't seem to be getting at the same thing.

      Look. There are definitely different outcomes for people in different races and countries. There are two proposed explanations. One is cultural - whether in terms of things like "blacks have a culture of poverty" or "whites have a culture of oppressing blacks". The other is genetic. Unless you know some third option, you can't dismiss both as just-so stories.

      Or, rather, you can if what you're trying to do is say that there's no good evidence for either and you remain agnostic on the point.

      But I read Wade as saying not so much "It is proven that the explanation is biological, everyone go home now" as "everyone so far has only considered culture, but biological also deserves a lot of investigation". If you're ambivalent between the two possibilities, I don't think you and Wade have any difference.

      And unfortunately, while I agree the cultural explanations are just-so stories, there's a lot of research really strongly supporting the biological side. This is the link you don't want to read. If you insist upon a summary, the summary goes:

      "Every time we measure racial IQ, black people are about 15 points below white people. This is true even when we use tests that couldn't possibly be biased, even when we use black people adopted by white parents, even when we control for the effects of social class, and regardless of things like stereotype threat that are sometimes used as explanations. When we try a lot of clever tricks to tease out the genetic versus cultural influences on the gap, they come out as mostly genetic. This IQ gap predicts very well the observed disparities of outcomes between blacks and whites, and so is probably the explanation for such."

      You are probably smart enough to generate at least twenty objections to such a claim. At one objection addressed every two pages, that explains why it's a forty page document ;-)

      Delete
    3. Every time we measure racial IQ, black people are about 15 points below white people.

      In America, or globally?

      This IQ gap predicts very well the observed disparities of outcomes between blacks and whites, and so is probably the explanation for such

      Are you sure? The crowing of the rooster perfectly predicts the rising of the sun, and yet it is not the explanation for such.

      Delete
  7. Anonymous9:45 PM

    "Scared that a big, masculine black guy will take your girlfriend? Worried that hard-working, intelligent (but "uncreative") immigrants will take your job?"

    No genetics relationship provability is necessary at all to come to those conclusions. Merely large data-sets of last years crime and employment statistics could make these practical behaviors logically justifiable.

    ReplyDelete
    Replies
    1. Since when is taking someone's girlfriend a crime? ;-)

      Delete
    2. Anonymous6:03 AM

      Crime statistics? A huge majority of blacks crimes are blacks on blacks crimes. There is much higher probability that white person will be killed by white person than that it will be killed by black person.

      So you do not talk statistics, you talk you racial prejudices straight from moocher ranch in Nevada.

      Delete
    3. Anonymous9:14 AM

      The relevant statistic would be the probability of a white person being killed by a black person divided by black population, versus the probability of a white person being killed by a white person divided by white population. I'd suspect that this would be a negative for blacks.

      Delete
  8. The only difference is that there is no cultural model to explain anything. That is, there isn't a working model of cultural transmission of traits, at all. Behavioural geneticists long ago concluded based on twin & adoption studies that the effect of "shared familial environment" basically amounts to zero. Parents matter primarily through their genetic contribution to children. The residual referred to as the "nonshared environment" is just a grab bag of unexplained effects, could be just pure noise. Immigrant self-selection, the most widely used factor to explain domestic disparities, isn't a cultural model, either. You just throw back the explanation to the immigrant's home country.

    ReplyDelete
    Replies
    1. Anonymous4:24 AM

      @pseudoerasmus: "The only difference is that there is no cultural model to explain anything."

      Wait for the next installment of Deirdre McCloskey's Bourgeois Trilogy. ;-)

      We'll have to complement it with human differences facts, of course (McCloskey's a woman of certain age).

      But it'll rock like Iron Maiden from the 1980s!

      :-)

      --Anônimo

      Delete
  9. Alex H.10:18 PM

    Why do liberals have to get so uncomfortable at any mention of race? For whatever reason they forget their beloved theory of evolution when it comes to personality and behavioral traits.

    The whole matter is really straightforward:

    1. No one will deny that there are systematic *physical* differences between individuals on the basis of genetic origin.

    2. Personality and behavioral traits, like physical traits, are also the product of genes. This has been substantiated by numerous experiments.

    3. Taking this one further, it would be extremely odd to believe that while there *are* differences in the distribution of physical traits across gene pools, there *aren't* differences in the distribution of behavioral and personality traits across gene pools. One would have to take the complete "blank-slate" view; that is, that behavioral and personality traits are entirely independent of genes. Again, this view has been entirely dis-credited.

    4. If there *are* differences in the distribution of behavioral traits across gene pools it would be similarly odd to believe that group outcomes are independent of those distributions.

    The hard part about 4. isn't figuring out *whether* there are such effects, it's figuring out *what* those effects are. It's an extremely complicated matter of figuring out the aggregate consequences of different distributions of traits (there are complicated feedback loops, everything is non-linear, etc.). But it's an extremely interesting subject, and appears to me to be more than worthy of exploration.

    Lastly, the N=K argument is really pretty weak; you could apply it to just about any scientific inquiry. It should go without saying that we need to be worried about selection bias. Any legitimate inquiry needs to take account of it. Claims about white superiority in sports 50-100 years ago were entirely unsubstantiated, and that's because, among other things, they weren't accounting for selection bias ("barrier" is a more appropriate descriptor than "selection" here but the term describes same the underlying statistical problem). Today if you say that individuals of Western African descent dominate sprinting, and individuals of Eastern African descent dominate long-distance running (both for genetic reasons), I think most would (and should) agree. We realize at an intuitive level that in this specific instance the issues of selection probably aren't that relevant.

    ReplyDelete
    Replies
    1. Anonymous6:20 AM

      "For whatever reason they forget their beloved theory of evolution when it comes to personality and behavioral traits."

      Seriously, this one sentence is enough to conclude that you are intelligently designed moron.

      Delete
    2. "Why do liberals have to get so uncomfortable at any mention of race? For whatever reason they forget their beloved theory of evolution when it comes to personality and behavioral traits."

      Because race is cultural, not genetic. Valiant effort, young squire. Do try again.

      Delete
    3. Daniel10:32 AM

      Because race is cultural, not genetic

      Asserting something does not make it true. Valiant effort, young squire. Do try again.

      Delete
    4. Anonymous2:06 PM

      Would anyone care to take the time to reply to Alex's post rather than nitpicking at his use of the word race. It seems like you guys are avoiding a pretty good argument

      Delete
    5. @Daniel: Would scientists convince you? Or are you a young-earther?


      http://www.pbs.org/race/000_About/002_04-background-01-03.htm

      http://www.aaanet.org/stmts/racepp.htm

      http://news.wustl.edu/news/Pages/184.aspx


      While asserting something may not make it true, that assertion does not make truth less-true.

      Delete
    6. @206: It's just not clear there is an argument to be refuted.

      "No one will deny that there are systematic *physical* differences between individuals on the basis of genetic origin."

      Possibly, conditional on the operative definition of "physical."

      "Personality and behavioral traits, like physical traits, are also the product of genes. This has been substantiated by numerous experiments."

      Citation needed. (Charles Murray doesn't count.)

      "Taking this one further, it would be extremely odd to believe that while there *are* differences in the distribution of physical traits across gene pools, there *aren't* differences in the distribution of behavioral and personality traits across gene pools. One would have to take the complete "blank-slate" view; that is, that behavioral and personality traits are entirely independent of genes. Again, this view has been entirely dis-credited."

      The distribution of genetic variation within any given "race" is sufficient to moot this. Clearly, people who look different from each other have genetic differences. The uncomfortable fact for the Murray crowd is that these differences are a tiny fraction of the sum of genetic differences between groups. What are the chances that it's the genes in which we can observe a physical difference that account for all of these differences they claim to see in intelligence, ability, initiative, etc?

      "If there *are* differences in the distribution of behavioral traits across gene pools it would be similarly odd to believe that group outcomes are independent of those distributions."

      Or, they would rather believe that the differences they want to quantify are in genes we can observe, since that allows Cliven Bundy to be a legitimate applied social scientist.

      Delete
    7. Daniel5:38 PM

      Do you honestly expect me to be impressed by people parroting the politically popular line on race ?

      Are you really that sheltered from people of different opinions that you've lost your critical thinking skills ?

      Razib Khan tells it so much better here

      http://www.unz.com/gnxp/off-topic-comments-and-nick-wades-book/

      The problem is that people move from this non-controversial point, that some racial categories are social constructs, to the assertion that all racial categories are social constructs, and that phylogenetic clustering of human populations is irrelevant or impossible. It is not irrelevant, or impossible. Human populations vary, and that variation matters. Human populations have specific historical backgrounds, and phylogenetics can capture that history through methods of inference.

      Delete
    8. @ Jonas Feit,

      Those organisational statements are about as convincing as the Seville Statement on Violence :-)

      This comment about sums up this debate:

      "Dr. Stephen O'Brien, a geneticist at the National Cancer Institute, said that the conclusion that race was not a valid concept ''comes from honest and brilliant people who are not population geneticists.''

      ''That doesn't mean they are insincere,'' Dr. O'Brien said. ''It's just that they haven't really looked at it. What is happening here is that Neil and his colleagues have decided the pendulum of political correctness has taken the field in a direction that will hurt epidemiological assessment of disease in the very minorities the defenders of political correctness wish to protect.''

      http://www.nytimes.com/2002/07/30/science/race-is-seen-as-real-guide-to-track-roots-of-disease.html?src=pm&pagewanted=2

      I'd recommend you read some of BGI Cognitive Genomics member Steve Hsu's posts on the subject too*. Note that it has real implications in terms of medical research, in terms of GWAS you have to correct for population structure as a possible confound for the real effect you are looking for.

      http://www.nature.com/ng/journal/v38/n8/abs/ng1847.html
      * http://infoproc.blogspot.co.nz/2014/05/nicholas-wade-interview.html

      Delete
  10. AlexH's 1-3 remind that anti-hereditarians routinely invoke the technically correct formula "the fact that X is heritable does not imply that group differences in X are also heritable" -- as though the large heritability estimates found through twin & adoption studies would magically crash to zero between races. (Of course there are transracial adoption studies too.)

    ReplyDelete
  11. Race doesn't make much sense. Classifying people by appearance is really stupid, because it doesn't tell you anything useful about all the other genes that people may have.

    Africa is the most genetically diverse continent and it has populations that produce both the tallest and shortest people on average, the best runners in both long and short distances and all that variety is reduced to calling people "black".

    Try guessing what is the probability a person from a particular area is lactose intolerant- skin color or eyelid shape isn't going to tell you.

    Academic racism can be of any use in only a small minority of isolated populations - say Pygmies or certain small island nations. In fact it is somewhat beneficial - instead of prejudice - it brings understanding and opportunity since those groups will often have similar health needs that could be addressed (diabetes in Pacific Islanders for example).

    ReplyDelete
    Replies
    1. @ Dimitar,

      It's not so much classifying people by appearance, but by continental ancestry. This has implications in terms of medical research across groups. It is a nuisance for GWAS -- you have to correct for population structure as a possible confound for the real effect you are looking for.

      http://www.nature.com/ng/journal/v38/n8/abs/ng1847.html

      http://genomebiology.com/2002/3/7/comment/2007

      Delete
  12. ..

    A roundup of the reviews of Wade's book is at:

    http://occamsrazormag.wordpress.com/2014/05/06/roundup-of-book-reviews-of-nicholas-wades-a-troublesome-inheritance/

    ..

    ReplyDelete
  13. Anonymous2:10 AM

    The alternative hypothesis is Gelman on this issue is being an idiot trying to cover up a previous debate where he lied to the public to try to hide the fact that Harvard is approximately 20% Jewish (among other similar statistics).

    Why his motivation is that strong is unknown. There is possibly real fear or conspiracy type thinking that those trying to observe facts without malice are enabling some sort of political ascendency of truly harmful nuts.

    ReplyDelete
    Replies
    1. Anonymous6:14 AM

      Anonymous: "There is possibly real fear or conspiracy type thinking that those trying to observe facts without malice are enabling some sort of political ascendency of truly harmful nuts."

      +1

      American "liberals" need to understand that basic agreement with facts about human differences does not imply agreement with Goebbels's ethics. There is something called "Hume's Law" or is-ought distinction.

      "Liberals" need to abandon this stupid rhetorical move of denying the existence of population structure and differences in cognitive and behavioral characteristics among distinct human groups. You need to keep up with people like Garrett Jones or Bryan Caplan (or, please just shut up: you're hurting the case we're making in defense of even low IQ immigration. Our rights shouldn't be predicatd on our IQs (or stature or physical attractiveness).

      Delete
    2. Anonymous6:19 AM

      Also, read this so you can keep up with us, true liberals:

      http://openborders.info/iq-deficit/

      --Anônimo

      Delete
  14. There is good evidence that height has been under different amounts of selection in different human populations: http://dx.doi.org/10.1038/ng.2368. It is thus perfectly possible that the same could be true of other traits.

    ReplyDelete
    Replies
    1. @ Noah Carl,

      I couldn't get your link to open, but I take it you were referring to this?

      "Genetic Inference: ... Europeans differ systematically in their height, and these differences correlate with latitude. The average Italian is 171cm, whereas the average Swede is a full 4cm taller. Are these differences genetic? Have they been under evolutionary selection in recent human history?

      Michael Turchin gave some pretty convincing answers to these questions, using genetic data from the 129 thousand individuals in the GIANT consortium. He compared the frequencies of alleles that are known to increase height, and found that they are more common in Northern Europe. Interestingly, he found the same relationship for alleles that have weaker evidence for height association, showing that there are still a large number of common height variants hiding in the genome, which are also more frequent in Northern Europe.

      Height differences are thus heritable, but have they been under evolutionary selection? Or are these differences merely down to genetic drift? This can also be tested using the GIANT data, which shows significant statistical evidence of selection on height variants in recent history. On top of that, the magnitude of the selection is correlated with the effect size of the height variant, providing strong evidence that these variants are being selected specifically for their impact on height.

      This is a textbook example of how an evolutionary study should be done; you show a phenotypic difference exists, that it is heritable, and that it is under selection. This opens the question as to why height has been selected in Northern Europe (or shortness in Southern Europe). Could the same data be used to test specific hypotheses there?"

      http://infoproc.blogspot.co.nz/2011/05/height-breeding-values-and-selection.html

      Delete
    2. Another interesting paper in this vein is: Patterns of Ancestry, Signatures of Natural Selection, and Genetic Association with Stature in Western African Pygmies.

      Delete
  15. Daniel7:02 AM

    When you feel the need to argue against things nobody says, you know your belief system is threatened.

    The bigger question is - why are liberals (in the US sense of the word) so dead-set against acknowledging biological differences between human populations ?

    Is it because it exposes their project as utopian ?

    ReplyDelete
    Replies
    1. When you feel the need to argue against things nobody says, you know your belief system is threatened.

      That's a fun line to say, but it might not be true.

      The bigger question is - why are liberals (in the US sense of the word) so dead-set against acknowledging biological differences between human populations ?

      Now that's an interesting question. See if you can come up with a plausible answer. Hint: It's not because they are communists who want to take over the world.

      Is it because it exposes their project as utopian ?

      No, I don't think so...

      Delete
    2. Daniel10:26 AM

      That's a fun line to say, but it might not be true.

      Gee Noah, seeing as how you try to shut down the discussion by yelling "RAAACIIISM" ... it's almost like there's some uncomfortable facts you're like to prevent from getting known.

      Also, gotta love the guilt by association.

      Wade speaks about biological differences between populations.
      Steve Sailer speaks about biological differences between populations.
      Steve Sailer is a racist idiot (yes, he actually is a racist idiot)

      Therefore all people who talk about biological differences between populations are racists.

      Nope, no faults in that line of reasoning.

      Delete
    3. Gee Noah, seeing as how you try to shut down the discussion by yelling "RAAACIIISM" ... it's almost like there's some uncomfortable facts you're like to prevent from getting known.

      This is why it's hard for smart people to have productive discussions with dumb people (and ironic when the discussion is about smartness itself!).

      Nowhere do I try to shut down discussion. Nowhere do I yell "RAAACIIISM".

      That's just an off-the-shelf response you cut-and-paste from a million other discussions. You didn't even bother to think about, or carefully read, what I said. You just took the same dumb derpy shit that you blurt out in a million other discussions, assumed this one was more of the same, and slapped it down here too.

      I'd say "get a brain", but you probably think that's impossible, so until we invent intelligence-boosting technology, you're kinda screwed.

      Delete
    4. Daniel4:03 PM

      Nowhere do I try to shut down discussion. Nowhere do I yell "RAAACIIISM".

      The article has "racism" in the title and the word itself appears 15 times in the text.

      Smart you may be, but intellectually honest you are not.

      Delete
  16. The troublesome inheritance is an inconvenient truth for the left. The only acceptable view is that race is a social construct and and the differences between individuals are only attributable to their environment. If some people fail to thrive in the economy, it’s because we’re not doing enough to help them, says the left

    ReplyDelete
    Replies
    1. Surely if some people are genetically inferior, that means we have no choice but to help them, since they are not capable of being equal by their own effort.

      Delete
    2. Anonymous10:38 AM

      @UserGoogol, they are not inferior, jyst different. *On average* (important), west Africa blacks will be more strong, east Africa blacks more resilient, pigmeus will have better olfacts, Australian aborigenes better visual acuity (greatest visual cortex of all), sardinians better vascular system health, nepaleses better altitude resistance., jews will have greay stand up artists (higher IQ helps), etc.

      There's no *absolute* superiority, got it?

      -Anônimo

      Delete
    3. Damien1:55 PM

      That'd be true if we considered that all characteristics are equally valuable. But that's clearly not the case: what did it look like last time people were convinced that West African blacks were stronger and Anglos smarter? Hint: not really "Kumbaya, everyone's different and equally valuable" but systematic oppression of those who were perceived as incapable of taking care of themselves and of voicing valid opinions.

      It's a bit of a cop-out to say that it's all about different abilities and not superiority. If we could choose whether our children would have high IQ or great physical strength and better visual acuity, no-one would choose the latter. Because that would destine them to a life of low-paid, back-breaking toil that would put them at the bottom of the social hierarchy.

      Functionally, saying that some groups are more intelligent is the same as saying that they're superior in virtually every way that truly matters in a modern society.

      Delete
    4. Anonymous5:38 AM

      Using the word "Enlightenment" like a security blanket isn't going to make your reasoning any more sound.

      Delete
  17. First time perusing this blog. Everyone seems quite intelligent here, which is rare! You're not considering the crucial factor of whether the "groups" being identified are real biological units. Races, if treated as synonymous with modern parlance of populations, are not real evolutionary units. In fact, above the individual, only species are supposed to have the spatio-temporal cohesion to be real. Moreover, genetic diversity, assuming a number of individuals possess unique traits, which they probably do, actually exist in mere genetic droplets versus deep pools, especially for traits we tend to fixate on. The problem with "Human Biology" studies is that they're being conducted by people who've missed key debates in biology. For example, human ancestry trees are likely ALL wrong. They're phenograms based on the distance differences between allele frequencies! Frequencies arent inherited! If you wanted to ID your POTENTIAL units properly, you'd need to find the ACTUAL unique genetic differences universally held by some individuals you think have shared a cohesive inbred history for "y" amount of time. BUT then it might not even be race. If you looked at it objectively, it might arguably be a different SPECIES. And wouldn't you know it human history is actually a story of constant Species Competition. This explains war, genocide and poverty. But what about admixture you cry? Well there's Homoploid Hybrid Speciation and Hybrid Zones, reinforcement speciation.... If you guys want to end scientific racism, help a Chinaman out and support the NEW Synthesis, which is really just me casting the evolutionary lens I'd use on all life on people. Let a Phylogenetic Systematist or "Cladist" inject some new energy into old debates. Sonsofmonkeys.com

    ReplyDelete
  18. Anonymous10:10 AM

    Re: "building a theory that can explain everything but predict nothing." Sort of reminds of the various religious beliefs.

    ReplyDelete
  19. When Charles Murray reviewed Wade's book in the WSJ, he predicted that critics will dwell on the second part of the book that Wade himself admitted was more speculative, than the first part of the book which contains the solid science regarding population structure. And Murray was prescient ! Look at Gelman's review and Noah's comments. They don't discuss the scientifically solid part at all and focus exclusively on the speculative bits. They even call Wade "racist" whilst rationalising they are just being etymologically pedantic, as though the word "racist" would not be understand by 99.9% of their readers in the conventional, ultrapejorative sense.

    All the same, even the solid science part of the book is underestimated by Wade. Which brings me to :

    "You're not considering the crucial factor of whether the "groups" being identified are real biological units. Races, if treated as synonymous with modern parlance of populations, are not real evolutionary units.

    You should go to Steve Hsu's blog on Wade. He's clearly of the opinion that because of the history of race one must subject claims about racial differences in behaviour to a higher standard of proof & scrutiny, but he has absolutely zero doubt about the validity of race as a population construct. In fact he finds Wade's discussion too primitive and out-of-date. The evidence is far stronger than found in Wade.

    ReplyDelete
  20. I'm not sure I understand MarsQ's point. Individuals with certain advantageous traits have more children, and this changes the population frequency of those traits, and the population frequency changes over time until the population as a whole is in equilibrium with the environment, right ? If you have populations which remain separate because they don't intermarry for cultural reasons or because they never run into each other because of geographical separation, then you have identifiable & distinguishable populations, right ? (And even with limited intermarriage, the distinguishability remains through clusters.) So why do we need to dwell on the semantics of race, subspecies, species, etc.? As long as we understand the shorthand "race" is a term for population at any level of aggregation ?

    ReplyDelete
    Replies
    1. Anonymous12:21 PM

      Excellent points by @pseudoerasmus. Let's just call it "population structure is real" and call it a day...

      http://en.m.wikipedia.org/wiki/Population_stratification

      Population stratification is the presence of a systematic difference in allele frequencies between subpopulations in a population possibly due to different ancestry, especially in the context of association studies. Population stratification is also referred to as population structure, in this context.

      Delete
    2. Because it isn't semantics. Species are inherently sloppy buckets but they're real units of evolution. Populations are even more poorly defined, and can go extinct without the species going extinct so long as its not the last one. if you want to make an evolutionary adaptation argument about a particular group vs another, you should first make sure you're comparing actual groups AND then whether differences in the trait even matter. It's INCREDIBLY difficult to make evolutionary arguments for traits exhibiting continuous variation, especially when there is significant overlap. The racism component comes into play when we fixate on things like IQ because lets be honest, being the smartest is not a goldmine of reproduction. It never has been. Humans only need to be smart enough. If you sample and test with bias and create unnatural groups like Black and White, it's easy to come with statistical differences. It doesnt mean you're making an evolutionary argument though. The fact that people think species vs populations is a mere semantics debate is proof of a huge divide between naturalists (ie. Ethologists, systematists, paleontologists, anatomists) vs the new "life sciences" (i.e geneticists, ecologists, anthropologists).

      Delete
    3. Anonymous6:08 AM

      @MarsQ: If you sample and test with bias and create unnatural groups like Black and White,"

      Wrong.

      Genetic clusters almost always afrees with self identified continental races. Of course dirty cheap genetic sequencing allows us to be ever more precious with data: you can identify population (sub)structure even among "white" Europeans or "black" Africans.

      @MarsQ: being the smartest is not a goldmine of reproduction. It never has been. Humans only need to be smart enough.

      Wronger.

      In malthusian societies (virtually everybody lived in one pre-1800) small differencs in fitness matter more than in our "Great Enrichment" societies. Read Gregory Clark's study of medieval wills (testaments). They show that higher middle class had more children than poor people. If you don't like this data, you must sue it (or show us better data).

      Also, it's conceivable that prevident and organized people (let's not use the "I" word) will thrive better in societies (most of North Europe) that experienced "spring famines" until the XIX century. Some environments seem to be harsher and demand more... organization. Is it that surprising that Esquimos do better in IQ tests than any other hunter gatherers, or that the only large populations with consistent higher IQs are Europeans (less so) and East Asians (mire so)? Long history of agriculture plus high latitudes seem to matter on cognitive performance.

      If you dont like this hypothesis, please give us a better one, that won't be eviscerated by Occam's Razor. Simply calling people "racist" is just bad rhetoric.

      -Anônimo

      Delete
  21. John Soriano11:00 AM

    "None of this is to say that [insert field] is completely wrong, or is a useless way of looking at the world! It might get certain things right. But we don't really know (yet). We don't have a good understanding of the proposed causal mechanism, and we don't have good natural experiments to test any of the hypotheses."

    Reading this again, this sounds similar to some of your criticisms of macroeconomics!

    ReplyDelete
  22. Anonymous1:09 PM

    Lies, Damn Lies and Statistics can explain anything. The fancy term is post-hoc reasoning which nice and academic but not as pithy as my first phrase

    ReplyDelete
  23. Your only problem with this enterprise is over fitting? The solution to over fitting is usually either more data, robust/simpler models or more through training. Academic racist just need more funding for bigger data sets and to start using better methods (say lasso) to get more reliable results. Good advice!

    ReplyDelete
    Replies
    1. Your only problem with this enterprise is over fitting?

      No, but that's the main thing this post was talking about...

      Delete
  24. Anonymous2:12 PM

    I haven't read the book (I don't think anyone commenting on it has really), but here are a few reasons why I'm incredibly skeptical about Noah's arguments. Firstly, I'm not even sure this constitutes academic racism. Your own definition explicitly uses the word "superiority". Now that's partly the definition you used, but even then I think it disqualifies Wade from racism. He mentions it's possible that there are differences between races. I don't think that's really too ridiculous to think. He doesn't make any argument that this makes any race "inherently superior". If I see someone and notice they have a different skin color, and I point it out, that's not racist. If I claim superiority because of it, that is. His own description says:

    "Wade believes deeply in the fundamental equality of all human peoples. He also believes that science is best served by pursuing the truth without fear, and if his mission to arrive at a coherent summa of what the new genetic science does and does not tell us about race and human history leads straight into a minefield, then so be it. "

    So basically he acknowledges the possibility that there are differences, and wants to look into it basically to better understand ourselves. He doesn't think any race is inherently superior, he actually claims the opposite. Noah I think you're jumping to wild conclusions. Giving the idea that he dislikes immigrants or blacks out of fear and inferiority is a bit disingenuous. Another quote:

    "Also controversial are his findings regarding the genetic basis of traits we associate with intelligence, such as literacy and numeracy, in certain ethnic populations, including the Chinese and Ashkenazi Jews."

    So he's racist because he thinks non-whites are inferior because they are....smarter.

    I think Wade is doing exactly what Noah recommends. "Until we have a much better understanding of genetics, the infinitely proliferating hypotheses of academic racism will be neither proven nor disproven, and people will go on believing in them (or disbelieving them)." Well, wouldn't trying to get a better understanding of it put a rest to it all? I believe that's what Wade is doing.

    Here's something to prove my point:

    http://circ.ahajournals.org/content/118/13/1383.full

    Ethnic differences in cardiovascular drug response. They found that there are differences by race, and by acknowledging these differences they save thousands of lives. I think this is what Wade is trying to find, a better understanding of race that can prove to be helpful. By jumping to the conclusion that he must be a white supremacist because he thinks there could be differences, well, that may not be racist in intent, but if his research (or anyone's research) led to something equivalent to how helpful cardiovascular medication has been, that would be racist in effect. It may mean well, but the actual result would be a lower welfare for different races. So really I'd argue Noah is the one being more harmful in this situation.

    ReplyDelete
  25. candid_observer2:17 PM

    This "review" is breathtakingly stupid.

    Look, while Wade himself doesn't push this point, what virtually all "race realists" argue is that there are actually a relatively small number of genetically based factors that may explain the vast majority of differences between countries. In particular, IQ as the most important single most important factor, explaining a very great deal of the differences between nations in terms of economic success. Wade suggests that Openness to experience, one of the 5 components of a standard analysis of personality, may explain differences in curiousity between Asian and Western cultures. Wade also suggests levels of trust may figure in.

    How do these small number of components turn into a K=N problem as this review suggests?

    Of course, it is always the cultural explanations that are plagued by "just so" stories and K=N problems. Why do HIspanics do badly in the US? It's the prejudice and the second-language problem. Oops, except that immigrant Asians do great, and they have had prejudice in the past and a second-language problem, so that with Hispanics its prejudice and second-language combined with the barrio culture. Etc., etc.

    ReplyDelete
    Replies
    1. In particular, IQ as the most important single most important factor, explaining a very great deal of the differences between nations in terms of economic success.

      When you look at the cross-section, it sure looks like that. But when you look at the time-series data, weird things emerge. Nations change places in the income hierarchy surprisingly quickly. That doesn't rule out IQ-based causal explanations, but it requires some explaining. Do populations just evolve really quickly?

      Also, we know that the huge income differences between, say, North and South Korea are not due to racial differences, nor to any IQ differences that existed before 1950. Why should we assume that the (much smaller) difference between Hungary and Botswana is due to completely different factors than the difference between the Koreas?

      In other words, even this simple theory ends up not being parsimonious. It keeps having to be excused and patched up to explain contrary data. That looks to me like a degenerative research program, and that's where the overfitting comes in.

      Of course, it is always the cultural explanations that are plagued by "just so" stories and K=N problems.

      Yes, I agree.

      Why do HIspanics do badly in the US...except that immigrant Asians do great

      Hispanics don't do badly, they do well. Asians do better, but that's obviously due to selective immigration - we grab the best and the brightest.

      In Japan, they think Chinese people are dumb and Indians are smart. In Malaysia they think exactly the opposite. The reason is obvious - just look at which type of immigrants go to each place from each of those countries.

      Delete
    2. Anonymous3:03 PM

      "Also, we know that the huge income differences between, say, North and South Korea are not due to racial differences, nor to any IQ differences that existed before 1950."

      Er, malnutrition?

      Delete
    3. candid_observer3:15 PM

      Noah,

      Yes, of course there will be some exceptional cases that demand further explanation after IQ, and the genes it is based upon, does its explanatory work.

      But the point is just how much work IQ can do in general to explain economic success. That's how such explanations might ever be expected to work. Obviously, there are some unusual conditions under which the simple IQ/genetic model will fail, and it's more than reasonable to find something unusual -- such as perhaps the most totalitarian government in the world -- to account for these exceptions.

      The problem with your world view is that you can't acknowledge the enormous success of IQ/genetics to explain so very much of what we see in terms of economic outcomes. If it's not really important, why does it do such a bang up job in so many instances?

      Delete
    4. Yes, of course there will be some exceptional cases that demand further explanation after IQ, and the genes it is based upon, does its explanatory work.

      But why should we assume that race explains the GDP difference between Hungary and Botswana, and assume that North/South Korea is a special case in terms of cause? Maybe Korea is just a convenient natural experiment that very cleanly shows how different institutions can lead to huge outcome differences. And so what if it's institutional differences that cause the difference between Hungary and Botswana, or between Europe and Africa in general?

      Run a regression, and you'll find that Europe has much better institutions than Africa. Why not assume that institutions - not anything genetic - are the fundamental cause of income differences, especially in light of the convenient natural experiment of Korea?

      The problem with your world view is that you can't acknowledge the enormous success of IQ/genetics to explain so very much of what we see in terms of economic outcomes. If it's not really important, why does it do such a bang up job in so many instances?

      Oh, I don't deny that it could be a large part of what's going on. But I'm not convinced. When you say "explain", I know that you just mean "correlate with". But you can find just as good a correlation for a lot of other things, including some that are obviously not causal. So I remain unconvinced, but I don't rule out the possibility.

      Delete
    5. ***and you'll find that Europe has much better institutions than Africa. Why not assume that institutions - not anything genetic ***

      Well, that is the orthodox view isn't it? What if a pre-requisite for maintaining those institutions is a certain critical mass of human capital and that can't be developed locally? I don't know if that's correct or not, but if it isn't it suggests that those countries require more external assistance.

      Delete
    6. edit: I mean if it is correct.

      Delete
  26. >When you look at the cross-section, it sure looks like that. But when you look at the time-series data, weird things emerge. Nations change places in the income hierarchy surprisingly quickly.

    Relative income (gdp per capita relative to US for example) is remarkably stable over time. Outside Europe only East Asia has converged to any meaningful extent -- unless they had huge carbon or mineral deposits per capita. everyone else has either stagnated or even diverged.

    chart

    Why should we assume that the (much smaller) difference between Hungary and Botswana is due to completely different factors than the difference between the Koreas? "

    because what makes Botswana different from most other SSA is in the deposits of minerals (and a European elite)

    ReplyDelete
    Replies
    1. That chart compares single countries to whole regions.

      Look at countries within Europe, or within Asia.

      Delete
    2. Anonymous5:53 PM

      ". . . what makes Botswana different from most other SSA is in the deposits of minerals . . ."

      Is there a fact checker in the house? Botswana is the only country in sub-Saharan Africa with substantial mineral wealth? Really?

      And since when does a country's raw natural resource wealth correlate with a higher standard of living? I seem to remember some work showing just the opposite (think various petro-states, or Congo, or Afghanistan). Heck, I even seem to remember a guy named Noah Smith writing about this.

      Delete
  27. Anonymous3:20 PM

    Yes, Noah, there are no differences between the races because your politics say so.

    ReplyDelete
    Replies
    1. Yes, Anonymous, there are huge differences between "races" because your politics say so.

      Delete
    2. Anonymous3:59 PM

      I never implied that they're huge just that they're real.

      Delete
    3. What do you know about my politics? ;-)

      Delete
    4. Anonymous5:33 PM

      For starters, that you are a jewish liberal. And that says it all.

      Delete
    5. What does it say? ;-)

      Delete
  28. This is one of the best criticisms of global-warming hysteria I've seen.

    Question: What is the equivalent of lukewarmism?

    ReplyDelete
  29. noah, citing "institutions" doesn't address the difference between exogenous institutions (like North Korea) and institutions which are culturally or genetically endogenous. why do institutionalists always believe everyone has the same ability to build pro-growth institutions ? clearly different peoples have different abilities to build institutions, and different peoples living in the same society respond differently to the same institutions as you do see in multicultural societies.

    African institutions are relatively homogeneous and the outliers tend to be like Botswana or South Africa which has simply preserved the European institutions handed to them.

    ReplyDelete
    Replies
    1. noah, citing "institutions" doesn't address the difference between exogenous institutions (like North Korea) and institutions which are culturally or genetically endogenous.

      Of course not.

      why do institutionalists always believe everyone has the same ability to build pro-growth institutions ?

      Do they really? I offer the institutionalist hypothesis as an "alternative" to the academic-racist hypothesis simply to illustrate some points about evidence, correlation vs. causation, etc.

      clearly different peoples have different abilities to build institutions

      That's not clear to me. Sounds like some derp.

      Delete
    2. Sigh. You don't get the scientific method, do you? I think Noah has said more than once that he is not taking the opposite side of claims about genetic inheritance. He has said -- quite rightly -- that's he's skeptical about such claims. Thus he has no burden of proof requirements to satisfy. You, otoh, have made a lot of claims and the burden of proof is on you to supply evidence and do your best to convince skeptics. But what are you doing instead? You're trying to get people to prove you're wrong, and if they can't make you say you're wrong, why, you win. Uh uh, doesn't work that way.

      So if you want to actually persuade people, I suggest you try to be a little more convincing. Trust me, accusing them of being 'political' is not the way to go about it.

      Delete
  30. as for your reactions to the chart of relative incomes... it makes no difference.

    relative incomes within Europe ? why ? NW Europe is ~70% of the US, and Southern Europe ranges from 50-60%. Eastern Europe exhibits a clear east-west gradient, ranging from the best in the west (the Czech Republic & Slovenia at ~50%) to the least in the east & south (Ukraine 17%, Albania 15%). there's more institutional disability further south and east you go in Europe.

    Outside Europe, Japan, South Korea, Taiwan, and a handful of small city- and island-states are the only ones to be in the range 60-75% of US income level. The only still-poor country that’s clearly converging with the rich countries is China. India might also be on a long-term path toward convergence, but it’s not nearly as evident as China. Everyone else’s relative income has either stagnated, or has even diverged from the rich countries — they aren’t catching up, or at least we can’t tell from the evidence. (as for the oil-rich pseudo-economies I’m not counting celestial-manna accumulation as a modern growth process.)

    No non-European country outside East Asia has achieved 50% of US income. No country in Sub-Saharan Africa has ever achieved more than 10%, unless it possesses oil or minerals combined with a very small population, or a sizeable European & Asian population. No country in the Middle East and North Africa without a large, consistent per capita production of oil has ever achieved more than 25-30% of US income, and even that average is boosted by outliers like the non-oil producing Turkey and Lebanon. In Latin America, outside the Southern Cone, one-third of US income has yet to be exceeded. Chile, Argentina and Uruguay have done it, but back in the late 19th or early 20th century when they were already some of the richest countries in the world. Then they spent most of the 20th century diverging from the rich countries.

    ReplyDelete
    Replies
    1. Anonymous5:20 PM

      Let's not forget that the North Koreans, despite everything totally wrong, are still a nuclear power...

      Delete
    2. relative incomes within Europe ? why ? there's more institutional disability further south and east you go in Europe.

      You're looking only at cross-sectional evidence. I'm talking about time-series. Ireland used to be much poorer than England but is now richer. Spain has caught up a lot too.

      No non-European country outside East Asia has achieved 50% of US income.

      I assume you're excluding countries rich in oil or other minerals. I assume you also exclude tax havens like Trinidad and Tobago. I assume you count Southeast Asia as part of East Asia. Well, when you exclude those, and East Asia, there's just not a hell of a lot of other countries left on the planet!

      Again, this is post-hoc reasoning. First you look at income differences that you know exist, which include income differences between regions. Then you observe that the people in these regions are genetically distinct (which again you already knew). Then you assume the latter causes the former. Voila!

      But it might not help you predict the future. We could make some bets and see how things play out, but it'll be a couple decades before either of us gets to collect!

      Delete
    3. Let's not forget that the North Koreans, despite everything totally wrong, are still a nuclear power...

      As are the Pakistanis...

      Delete
  31. Dear Noah:

    Rather than make up out of whole cloth strawman examples that you impute to me , why don't you try linking to some of my articles so that your readers can see whether your denunciation of me is fair?

    For example, here's one on sports and race from 17 years ago:

    http://isteve.blogspot.com/2013/01/track-and-battlefield-sailer-and-seiler.html

    ReplyDelete
  32. "You're looking only at cross-sectional evidence. I'm talking about time-series. Ireland used to be much poorer than England but is now richer. Spain has caught up a lot too "

    First, let me deal with this cross-sectional nonsense. The only countries ever to converge with the global income leader and the technological frontier at even the modest criterion of 50% are located in Europe and East Asia. OK ? Everybody else's relative income shows divergence or stagnation relative to the income leader as the very long-run pattern. (The exceptions noted below.)

    I assume you're excluding countries rich in oil or other minerals. I assume you also exclude tax havens like Trinidad and Tobago. I assume you count Southeast Asia as part of East Asia. Well, when you exclude those, and East Asia, there's just not a hell of a lot of other countries left on the planet! Again, this is post-hoc reasoning. First you look at income differences that you know exist, which include income differences between regions. Then you observe that the people in these regions are genetically distinct (which again you already knew). Then you assume the latter causes the former. Voila

    No. You only need to exclude the super-oil-rich countries, and only those of the Middle East plus Brunei. No need to exclude the mining ones in Africa. The only countries not populated by Europeans and Northeast Asians that have surpassed 50% of US income are super-oil-rich ones (Qatar, Brunei).

    There is only 1 non-oil exception : the Bahamas. A few others come close : Seychelles, Barbados, and even smaller ones.

    Maybe you can become an adviser to Benin and advise them to emulate one of these countries. You can then tell them to reduce their population by 90-95%, bring in Europeans and South Asians equivalent to 5-10% of the population, and detach and move their country closer to a country full of potential beach bums.

    Trinidad & Tobago ? It's a middle-income country but it's a pretty poor tax haven compared to, say, the Caymans.

    ReplyDelete
  33. I can't believe anyone would bring up Ireland as an example. Yes, it used to be poor. But if there was one single data point that's egregiously not in favour of the institutionalists, it has got to be Ireland. Here is a country who suffered 800 years of colonial exploitation and oppression, but somehow managed to achieve a higher per capita income than the oppressor country itself. Yet you've got ~50 data points in Sub-Saharan Africa, most with less than a century of colonial experience, and the overwhelming majority of them can't get their shit together enough to go beyond 10% of the US income level. One can keep coming up with ingenious models of multiple-equilibria institutional traps or other gobbledygook. Isn't there a statute of limitations on blaming colonial path dependence ? Never mind that African countries uniformly performed better in the 1950s, when most weren't even independent, and in the 1960s, when the wave of independence began. But it's not clear to Noah that people have different capacities for building institutions !

    The institutionalists concoct some proxy, throw a few controls in there, demonstrate the strong positive association between income & "institutions", ignore the fact that African countries cluster near the origin on the scatter, and still convince themselves all that's required is hard work to emulate the rich countries' "institutional quality".

    As for Botswana, some guy other than Noah said this :

    "And since when does a country's raw natural resource wealth correlate with a higher standard of living? I seem to remember some work showing just the opposite (think various petro-states, or Congo, or Afghanistan). Heck, I even seem to remember a guy named Noah Smith writing about this "

    This guy truncated the 2nd difference between Botswana and most other SSA countries : 3% of the local population who are Europeans run the country's business sector. Botswana and DeBeers share the revenue from the country's diamonds. The same arrangement exists in other minerals & gems with other western companies or local European-owned firms. Yes, Botswana has somehow managed to get "good" institutions : a body politic not foolish enough to kill the goose that laid the golden egg. For this wisdom they have managed to get to 20% of US income. Of course if the identical country of Botswana were populated by, say Iranians or Venezuelans, they would have been able to take complete control of their own resources rather than rely on the foreign goose

    It's really the same with South Africa. The ANC was wise enough not to expropriate or drive out the Europeans -- sort of like what Uganda did with South Asians or like Zimbabwe. Otherwise South Africa might have become as poor as Zimbabwe.

    ReplyDelete
    Replies
    1. "Here is a country (Ireland) who suffered 800 years of colonial exploitation and oppression, but somehow managed to achieve a higher per capita income than the oppressor country itself. "

      I got another data point to support your thesis (not). A country which came to be called United States suffered couple hundred years of colonial oppression yet it ended up the richest country on the planet. Clearly it's not colonialism! Except...

      These three guys called Mr. Acemoglu, Mr. Johnson, and Mr. Robinson wrote a paper that a person here or a person there might have read, which sort of addressed that issue.

      Delete
    2. Yes, isn't it funny some extractive institution countries do better than other extractive institution countries, and isn't even funnier where the inclusive institution countries cluster...

      Delete
    3. Translation of Acemoglu et al : "Where Europeans were able to settle and form a demographic majority, inclusive institutions got established. Where Europeans couldn't settle, they established extractive institutions and these just persist persist persist and persist to the present day".

      Delete
    4. I think you're sort of missing the point.

      Delete
    5. Yea I must be. Why don't you explain. Your original point about the USA as a colony in rejoinder to my comments about Ireland was also totally obscure. Maybe you just don't realise that Ireland had the ultimate in extractive colonial history.

      Delete
    6. I agree with Acemoglu politically on a lot of things, but he's embarrassingly ill-informed about history and is drawn as if by magnetism to the kind of unfalsifiable explanations that would drive Karl Popper crazy.

      Delete
  34. My favorite racial comment came from an old guy (Stanford '58) after this year's Super Bowl. " Where were all these black quarterbacks when I was young?" I told him that they were likely playing for Grambling and Southern and were considered biologically incapable of leadership positions in football.

    ReplyDelete
  35. It's not like Africans can't do better than how they fare in Sub-Saharan Africa. We have a perfect natural experiment that demonstrates that they can : the 42+ million descendants of West Africans who live in the United States. They demonstrate that under near-optimal institutions they can achieve ~2/3 of US per capita income. The trouble is, there's no evidence they can build institutions even half as good, all by themselves. The bottom of the range is Haiti, of course. Somewhat better and possibly more comparable with African-Americans is Jamaica, but it's still quite poor. Noah, you are back to the Bahamas when you advise Benin.

    ReplyDelete
    Replies
    1. The trouble is, there's no evidence they can build institutions even half as good, all by themselves.

      It's thoughts like this that I was talking about in my post.

      Delete
    2. And Barbados is better than Jamaica, just as Atlanta has been better governed than Detroit, and, over the last generation, Ghana better than Equatorial Guinea and Botswana better than Zimbabwe.

      Delete
    3. Noah, as you said we will see. I myself would say absolute income will keep growing in SSA, as it has done over the long run, but convergence will be very limited and dependent on the speed of technological innovation in the core. Technology transfer can make up for some of the institutional deficiencies.

      Delete
    4. This comment has been removed by the author.

      Delete
    5. This comment has been removed by the author.

      Delete
    6. Saying that African-Americans have "near optimal instutitions" is rather questionable, since American instutitions have very explicitly discriminated against blacks through much of their existence. That discrimination has been cut back over the last few decades, but still, "near optimal" is just wildly myopic.

      Delete
    7. "As Noah noted before, one does not need a high IQ to understand things necessary for modern productivity, such as programming, linear algebra, or calculus."

      Ouch!

      Delete
    8. Anonymous8:25 AM

      Ouch X 2.

      I still carry my A+ in calculus as a battle strip. Few managed, though: grade inflation is mostly a Humanities issue.

      -Anônimo

      Delete
  36. Anonymous9:13 PM

    Is there a single country where black people do well and Ashkenazi Jews don't?

    Just asking...

    ReplyDelete
    Replies
    1. I guess not. Ashkenazi Jews must be really smart, huh?

      Oh wait...I am one! So if you think my post is wrong, shouldn't you second-guess that conclusion? Maybe my special Ashkenazi super-smartness has given me a subtle understanding of the issues at hand... ;-)

      Delete
    2. It finally dawned on me 20 years ago during The Bell Curve brouhaha that the reason guys like Gould, Lewontin, Kamin, and Rose were so driven to broadcast brilliant disinformation against IQ and heredity is that they worried that if the peasants ever figured out that they tended to inherit higher IQs on average, the goyim would come after them with pitchforks and torches.

      The reality is that the average American is well aware that Jews tend to be smarter, and -- guess what -- they like Jews for being smarter and wittier on the whole. There's no need to try to cover it up with misleading complicated arguments.

      Delete
    3. OH MY GOD, people are going to find out Jews are smart!! And then...they're going to come after us!! Quick, delete the whole internet so no one knows!!!

      ...oh wait.

      Delete
    4. Steve, I must say, even for one such as yourself who was not blessed with my Ashkenazi super-intelligence, that hypothesis is startlingly dumb...

      Delete
    5. @ Noah Smith,

      I thought Steve was part jewish?

      I agree that I don't think that is the reason Kamin & co were so keen to bury the hereditarian hypothesis. A major factors is that they are all communists, in fact Kamin moved to Canada as his refusal to name names made him unemployable in the US (until it became cool again to be a communist & he was welcomed back :P ).

      Delete
    6. Anonymous5:21 AM

      @ Noah Smith

      You clearly are a smart guy but you've been blinded by your own PC politics into believing there are no innate differences in intelligence between populations.

      The massive over-representation of Askenazi Jews in the academic fields is clear proof that they are, as a group, more intelligent than the norm. Once we accept that there are some populations that are smarter than the average we have to logically accept there could be populations that are less intelligent.

      Now, what candidates are there? Let's think about it logically. We'd need to find a population of people that under achieve academically regardless of the country and culture they live in. A group of people that would be likely to have earned the stereotype of being dumb and resistant to education.

      Any ideas?

      Delete
    7. You clearly are a smart guy but you've been blinded by your own PC politics into believing there are no innate differences in intelligence between populations.

      Oh, I don't believe that at all. Suppose there were no innate differences between populations. Well, it would be easy to artificially create some, with selective immigration! Here in America we've let in mostly high-skilled Indians and low-skilled Mexicans; thus, even if there's no average population IQ gap between Indians and Mexicans on the planet, there certainly will be in the U.S. So it's obviously possible.

      The massive over-representation of Askenazi Jews in the academic fields is clear proof that they are, as a group, more intelligent than the norm.

      Of course - though I suspect that selective immigration played a larger part in that than anyone is willing to admit. And I also suspect there are other factors at work besides average population IQ:
      http://noahpinionblog.blogspot.com/2013/08/how-successful-are-jews-really.html

      Now, what candidates are there? Let's think about it logically. We'd need to find a population of people that under achieve academically regardless of the country and culture they live in. A group of people that would be likely to have earned the stereotype of being dumb and resistant to education.

      Any ideas?


      ...Nope, no ideas. Do you have any ideas?

      You're probably trying to imply that Africans are low-IQ, but check this out:
      http://en.wikipedia.org/wiki/African_immigration_to_the_United_States#Educational_attainment

      So you see, African immigrants to America do better than whites, educationally. :-)

      Delete
    8. "Suppose there were no innate differences between populations. Well, it would be easy to artificially create some, with selective immigration! Here in America we've let in mostly high-skilled Indians and low-skilled Mexicans; thus, even if there's no average population IQ gap between Indians and Mexicans on the planet, there certainly will be in the U.S. So it's obviously possible.

      But no one has ever said that demographic shifts can't change the population frequency of traits. In fact there is a book out by Botticini and Eckstein (economists) who argue that Jewish intelligence is largely a product of attrition by out-conversion of those who could not afford to maintain the rabbinical injunction to educate their sons in scriptural interpretation.

      There are sizeable Indian diasporas in over 50 countries, in some of them a majority of the population. Trinidad for example is nearly 1/3 South Asian. The entire Indian diaspora has got a huge variation in incomes & achievements and without knowing something about the characteristics of the original emigrants you can't really relate them back to the home population. Self-selection is always going to be just throwing back the chain of explanation to the home country.

      The self-selection argument really doesn't work for Ashkenazi Jews, though. Or if there is an effect, it must be pretty small and can't be captured by the available data. I mean, Ashkenazi Jews are overachievers eveywhere. Where aren't they ?

      "So you see, African immigrants to America do better than whites, educationally. :-)"

      I'm absolutely sure you know how to calculate what percentage of the population of Africans in the entire continent + diaspora who would have 1, 2, and 3 SD above the mean IQ. When you've got 1 billion people and every incentive for brain drain, there is nothing mysterious about high-achieving African immigrants.

      Delete
    9. But no one has ever said that demographic shifts can't change the population frequency of traits.

      My point is that the certainty that demographic shifts CAN change the population frequency of traits, combined with the fact that demographics have a random component, ENSURES that there will be mean differences in IQ and other traits between populations.

      See?

      The self-selection argument really doesn't work for Ashkenazi Jews, though. Or if there is an effect, it must be pretty small and can't be captured by the available data. I mean, Ashkenazi Jews are overachievers eveywhere. Where aren't they ?

      I think Ashkenazi Jews are a bad example, because of the small overall number, the difficulty in counting (since "Jewishness" is not strictly defined by proportional heredity, but is partially self-defined and partially matrilineal), and because the group suffered large demographic disruptions recently (the Holocaust).

      Leaving Jews aside, it's difficult to find an example of a group that is successful everywhere it goes. Those groups that do fit the bill tend to be small religious minorities.

      When you've got 1 billion people and every incentive for brain drain, there is nothing mysterious about high-achieving African immigrants.

      Of course not. So do you attribute the low population IQ of the African-American descendants of slaves to selective immigration? Or are you also willing, in that case, to toss out diasporas as data points and focus exclusively on home countries?

      Delete
    10. Anonymous6:46 PM

      Chua's World on Fire makes a similar point...also, pretty sure Sailer's part Hebrew.

      Delete
    11. also, pretty sure Sailer's part Hebrew.

      We can't *all* be geniuses...

      Delete
    12. Anonymous1:17 AM

      @Noah 10:01 Am

      "So you see, African immigrants to America do better than whites, educationally."

      Here's part of that link:
      "However, in recent years there has been an increase in the number of African immigrants interested in gaining permanent residence in the United States. This has led to a severe brain drain on the economies of African countries due to many highly skilled professionals leaving Africa to seek their economic fortunes in the United States and elsewhere."

      It's funny, you've mentioned selective immigration multiple times and yet didn't consider that. In fact, you actually disregarded actual IQ levels by nation (which were way more relevant) in order to put that. That's a bit disingenuous, Noah. I would imagine developed countries wouldn't have a tendency of brain drain emigration, and therefore the average immigrant would be exactly that, average. In fact pseudoerasmus explicitly pointed this out, but I imagine it went over your head:

      "When you've got 1 billion people and every incentive for brain drain, there is nothing mysterious about high-achieving African immigrants.

      Of course not. So do you attribute the low population IQ of the African-American descendants of slaves to selective immigration? Or are you also willing, in that case, to toss out diasporas as data points and focus exclusively on home countries?"

      You missed his point, Noah. The educational attainment of African immigrants has selection bias (not to mention education varies and simply attaining an education doesn't say anything about its quality or the actual IQ of the people who attained it). Therefore, it's not useful.

      Delete
  37. Noah makes up a strawman argument about how those of us interested in human biodiversity supposedly reason. In reality, here's what I wrote in VDARE.com twelve years ago in reviewing Lynn and Vanhanen's ground-breaking book "IQ and the Wealth of Nations." The state of the art has progressed well past this point, but it's worth pointing out how sophisticated the HBD discussion already was a full dozen years ago:

    https://vdare.com/articles/a-few-thoughts-on-iq-and-the-wealth-of-nations

    ReplyDelete
    Replies
    1. Daniel4:51 AM

      Quick Steve, tell everyone how immigration leads to increased land prices and stupid regulations.

      Delete
    2. if you have massive immigration to let's say London or Manchester in the 1800's or New York in the same period ...that leads to increased land prices

      and silly regulations or bylaws or ....

      Delete
  38. M_Young3:35 AM

    "Academic racism is very alluring, for at least three reasons. First, it tells us that all our stereotypes and prejudices are basically right - and we humans like to be told that all our preconceptions are right. We suffer from confirmation bias"

    Uh, doesn't that apply to all the academic blank-slaters cum egalitarians out there.

    ReplyDelete
  39. Noah said : "But it might not help you predict the future. We could make some bets and see how things play out, but it'll be a couple decades before either of us gets to collect!"

    My prediction is that relative incomes tend to persist -- far more likely between countries than within countries. In the latter peoples tend to intermarry and their descendants will regress to the national mean. Between countries I have little doubt that in 30 years, what ever their absolute levels of income, Sub-Saharan Africa will still be at the bottom of the international relative income hierarchy. In descending order of relative income :

    Western Europe & Northeast Asia
    Southern & Eastern Europe
    Southern Cone (Chile, Argentina, Uruguay)
    Latin America exclusive of Southern Cone
    Southeast Asia
    Arab Middle East & North Africa (exclusive of super-oil-rich, like Kuwait)
    Southern Africa
    West Africa
    East Africa

    Of course some regions have subregions with very distinctive characteristics, like Central America or Anglophone Caribbean or "Orthodox" Eastern Europe, which coud be ranked separately. Also Southeast Asia is more heterogeneous than almost all other regions, so the individual countries' relative incomes are more dispersed than in the other regions.

    ReplyDelete
    Replies
    1. There are very very few countries which are truly mobile in the relative income hierarchy in the long run. Countries populated by NE Asians comprise most of them. China will join the top rank at some point. India is the only one I'm not sure about, except to say it's unlikely to be in the top tiers of the hierarchy.

      Delete
    2. So are you willing to make only a qualitative prediction (that this "region ranking" will persist), or do you want to make a quantitative prediction (that the gaps will not narrow)??

      Delete
    3. Shleifer has done some work showing that within country variation in development, productivity, income is much larger than cross-country variation. This casts doubt on both the institutional and the hereditary views of development.

      Delete
    4. Noah : I think the regional ranking will hold. I think the quantitative gaps could be a little more fluid, because, as I said, institutional deficiencies could be made up for partially by more innovative, user-friendlier technology transferred from the core. So I don't believe there are hard maxima to convergence. Rather I find it perfectly plausible that there are soft maxima periodically reset by shocks to the world technological endowment (which is outside the control of most developing countries).

      Delete
    5. Dan : That's an effect of the fact that economic growth in nearly all developing countries is extremely unequally distributed. So that even with rapid GDP growth the median household in the Chilean distribution has not necessarily experienced that much faster income than slower-growing economies in the same region.

      Delete
    6. I should have been clear. The within country variation is not based on individuals but regions/states (e.g. new york vs alabama)

      I suppose some thing you can still tie to institutions or race but not the majority. If you look at education, for instance, Massachusetts will beat out most countries while US on average lags behind. State level education policies may have something to do with it or the low % of black population compared to Alabama.

      Delete
    7. Not sure that changes much. Different regions can have different demographic characteristics. Just think India, north vs. south Italy, France still shows numerous divides based on the old langue d'oil and langue d'oc division. Even Japan.

      Delete
    8. So, that's a pretty weak prediction, since switches will tend to be very rare when the number of regions is small, and since most alternative theories (cultural, institutional, and economic-geographical) also predict rare switches (since they, like academic-racism, were invented after observing the rarity of such switches).

      But OK, I'll bite. I predict that during the next 20 years, there will be at least one regional order switch. Want to make a bet? :-)

      Delete
    9. "My prediction is that relative incomes tend to persist -- far more likely between countries than within countries. In the latter peoples tend to intermarry and their descendants will regress to the national mean. Between countries I have little doubt that in 30 years, what ever their absolute levels of income, Sub-Saharan Africa will still be at the bottom of the international relative income hierarchy. In descending order of relative income :"

      I once saw rankings of the 30 countries with the highest per-capita GDP in 1900 and (1990?); they were almost identical.

      Delete
    10. You can check that here: http://en.wikipedia.org/wiki/List_of_regions_by_past_GDP_(PPP)_per_capita

      There are a number of changes in the ranks. Sweden, for example, has climbed a lot. Of course, NOW, academic-racists will say "Of course Sweden climbed the rankings; Swedes are naturally smart." Which is an illustration of the K=N problem.

      Delete
    11. I'm speaking in terms of regions only for simplicity's sake in a comments section. It's also possible to enumerate by country and even by region. Think Brazil, India, Mexico, Italy, Spain, Turkey etc. Regional disparities within countries can also quite be persistent. In a country like the United States with a lot of internal mobility and high rates of immigration the regionalism can be blurred, of course.

      "since they, like academic-racism, were invented after observing the rarity of such switches)."
      IQ was invented for purposes in educational psychology. It was not fashioned at all to account for data in cross-country data sets, unlike most of the qualitative proxies that have been used in them !

      "since most alternative theories (cultural, institutional, and economic-geographical) also predict rare switches "

      These alternatives predict roughly the same things as the genetic theory. "Cultural differences" are usually social phenotypes arising from genotype in interaction with the environment -- like IQ and micronutrient deficiency. "Institutional differences" I've already spoken about : with the exception of some very clear exogenous cases like Soviet armies imposing a regime, they usually reflect innate differences in social capacity.

      Someone brought up Jared Diamond. He is hated by the "human biodiversity" crowd because he wrote the book to exclude theories of genetic differences between populations. I've never understood the hate. If you acknowledge that evolution has occurred since the dispersal of the human species out of Africa -- and that's now undeniable -- then Diamond did such an excellent job of detailing the environmental differences which probably caused the genetic divergences within the human species in the first place... Those divergences may be small quantitatively but in phenotype they make the difference between the range of data that we see in the world. People who embrace Diamond are weakening their anti-genetic arguments.

      Since the genetic theory tends to subsume, Borg-like, alternative theories, you'll probably say something like the theory explains everything and therefore nothing. When I saw this blogpost of yours my first reaction was, "but when did social science ever convince anybody, especially with regressions, other than those already predisposed to confirmation?" Then I thought (wrongly, as it turned out) you applied this scepticism of yours only to the sociobiological theories of human differences. Anyway my point is that it's not going to be regressions and observational studies that will convince anybody, in the end. Darwinism was accepted well before molecular biologists could actually show evidences of selection. So I say, genomics will either confirm or falsify the genotypical differences between populations, and social models will get new stylised facts.

      Delete
    12. "There are a number of changes in the ranks. Sweden, for example, has climbed a lot. Of course, NOW, academic-racists will say "Of course Sweden climbed the rankings; Swedes are naturally smart." Which is an illustration of the K=N problem"

      Most of those data before 1820, especially outside Western Europe, are fictitious. Besides there is no reason to compare income levels in Malthusian and industrial periods, is there ?

      Anyway, there was a time when everybody was poor, including -- mirabile dictu -- Sweden. The point is that in the 250 years since the start of the industrial revolution, with the acceleration of efficiency growth and so much new technology & institutional models to diffuse -- only a small group of countries has converged with the first industrialiser, England.

      Delete
    13. IQ was invented for purposes in educational psychology. It was not fashioned at all to account for data in cross-country data sets, unlike most of the qualitative proxies that have been used in them !

      But the stability of population IQ across time is an assumption that was not implied by the inventors of IQ.

      , as well as the idea that IQ can be augmented by a "race" variable to explain country outcomes, was not

      These alternatives predict roughly the same things as the genetic theory.

      Wasn't that my point?

      "Cultural differences" are usually social phenotypes arising from genotype in interaction with the environment -- like IQ and micronutrient deficiency.

      Now that's a sweeping, general assertion that A) I'm not prepared to accept, but B) seems like it could be used to justify just about any conclusion.

      Someone brought up Jared Diamond. He is hated by the "human biodiversity" crowd because he wrote the book to exclude theories of genetic differences between populations. I've never understood the hate.

      It's called "motivated reasoning". Those guys have reached their conclusions, and don't want anything to raise doubt about the conclusions.

      Since the genetic theory tends to subsume, Borg-like, alternative theories

      That's a classic sign of overfitting. If your theory can subsume anything (with appropriate addition of assumptions and new ideas), you're almost certainly not going to be able to predict anything.

      When I saw this blogpost of yours my first reaction was, "but when did social science ever convince anybody, especially with regressions, other than those already predisposed to confirmation?"

      Sigh...sad but true, in my experience.

      Then I thought (wrongly, as it turned out) you applied this scepticism of yours only to the sociobiological theories of human differences.

      No one even thought to ask me what I really think is going on with IQ, GDP, etc.! But just for fun I'll tell you...

      My prior is that:

      A) Pretty much any human group large enough to qualify as a "race" can raise its average IQ to the mid-80s by eating right, not getting sick much as a kid, and growing up in a healthy emotional and intellectual environment. I suspect that the floor is closer to 90, actually.

      B) I would not be surprised if urban life raises IQ a bit, through processes not well understood.

      C) I suspect that evolution can change population IQ much faster than is commonly believed. This may be the cause of the mysterious effect in (B).

      D) Therefore, I think the bulk of the GDP-IQ correlation is endogenous - something acts to boost GDP, and then IQs rise.

      E) But those "somethings" rarely act, because low GDP is a stable equilibrium. Low IQ can cause low GDP growth, which keeps IQs low. Combined with (A)-(C), this means that even if low IQ causes countries to be poor, that does not imply that stable population genetics is the fundamental cause.

      F) However, I suspect that long-term-stable IQ-boosting population genetics also accounts for a small part of the effect. Because people tend to breed with people located nearby, this can be interpreted as "racial differences".

      That's my initial guess, but I'm prepared to be convinced (by data) that I'm wrong...in either direction. Since data is hard to procure, I don't expect to be convinced conclusively in the next couple of decades, by which time I suspect that IQ-boosting technology will have made this question less relevant...

      Delete
    14. Anonymous7:54 PM

      So ustedes are willingly admitting to being completely idiotic with almost no understanding of science?

      Have you heard of Bayes' Theorem? It's ok, I know there are wide ranges in what most phds in economics know. Willing to reexamine your conclusions?

      So do you know any mathematics whatsoever behind standard evolutionary theory on allele frequencies? There's not a geneticist out there who doesn't know that human intelligence is extremely polygenetic. That's very strong evidence that your (C), and thus (B) and related assertions that claim to combine those factors are pretty much completely wrong.

      Delete
    15. Anonymous4:07 AM

      @Pseudoerasmus: "Institutional differences" I've already spoken about : with the exception of some very clear exogenous cases like Soviet armies imposing a regime, they usually reflect innate differences in social capacity

      Wrong.

      Japan and China lagged the "European core" for centuries, and only stoped underachieving when adopted western ways (occidentalization). “Soft factors” like ideology and institutional quality seem to be relevant even among the “high IQs” countries. That’s why I think Deirdre McCloskey’s theory is important to this debate. Otherwise, is simply much, much difficult to explain why China and Japan, and nations inside the European core lagged decades or even centuries before converging. Also, it is really difficult to explain why the Industrial Revolution demanded so much time, or why the Songs didn’t attained it, at 1080 AD, when they were already experimenting with water/water steam engines to power mine works and textiles.

      This seems to be a big hole -- the size of one fifth of humanity -- in the otherwise golden Pseudoerasmus's rough model -- the best I've ever seen -- which is on his blog: http://pseudoerasmus.com/2014/05/05/economic-growth-human-biodiversity/ Self recommending, this post really cross the t's and dot the i's of what is in the back of his mind while PWNing almost everybody in this comment section.

      @NoahSmith: C) I suspect that evolution can change population IQ much faster than is commonly believed. This may be the cause of the mysterious effect in (B).

      Wronger.

      Rapid human evolution (at least since the Neolithic) has been generally accepted among the specialists after this watershed paper by Cochran, Harpend, Hawks & Wang: http://johnhawks.net/weblog/topics/evolution/selection/acceleration_embargo_ends_2007.html

      But it will never (never!) be so fast as Noah`s implies, because simply population genetics would preclude dramatic changes in intelligence (v.g. fast rising of average IQ in a population by 1 standard deviation) so short notice. The dispersal of favorable alleles need more than one, or four or eight generations to take hold. And there are very good reasons provided by a hard science (population genetics) for this to be true. After all, biology is a real science, unlike growth theory (which is still so retard and qualitative a science, poor thing!)

      --Anônimo

      p.s.: Maybe Noah should create a post imposing general guidelines about which type of posting he’ll not accept in his – always very interesting -- comment section?

      Delete
    16. Anonymous5:41 AM

      When I say "simply population genetics would preclude dramatic changes in intelligence" I mean "evolved" changes. Of course relationship between (better, rapid changing) environment and genetics (largely unchanged in the short term = few centuries) are needed to explain general rise in traits like stature and IQ (Flynn effect) in post. Great Enrichment societies like ours.

      Don't ever forget, though: better environment incur in diminishing gains, and statures and IQ don't change so much in richer societies anymore: that's why Flynn effect and gains in stature are dead in the Netherlands: they've already lunched the gains possible by better nutrition and other environmental factors. There, we probably need Gaetano scenario to make further ptogress: http://en.m.wikipedia.org/wiki/Gattaca

      -Anônimo

      Delete
  40. Anonymous9:43 AM

    I think Steve Pinker or Jonathan Haidt (two authors you probably like) would agree that
    1) IQ is real and important
    2) all big traits are heritable to some degree, have a genetic component (eg, skin color, the OCEAN aspects of temperament, IQ)
    3) demes exist among humans (eg, 'red' doesn't really exist, as any red color is an arbitrary part of the wavelength spectrum, but it's still meaningful, and you can generalize about red vs. blue)
    4) demes have different allele frequencies. It would be very improbable for drift and natural selection to not affect trait-based genes, and instead only things related to appearance like skin color or oddities like lactose tolerance.

    From this, it follows that human subpopulations will probably have different average IQs. Indeed, you test and find this is so. There are lots of ways to tease out omitted variables biases--controlling for income, parent's IQ, sibling IQ, twin studies, controlling for language.

    But in the end none of this matters, the truth doesn't care what any of us think.

    ReplyDelete
    Replies
    1. all big traits are heritable to some degree

      I don't think there's enough evidence to make this conclusion.

      in the end none of this matters, the truth doesn't care what any of us think

      But what does matter, then? What is the upshot?

      Delete
    2. Anonymous7:19 AM

      @Anonymous all big traits are heritable to some degree

      Right, if you add a weasel word: "[almost] all big traits are heritable to some degree". Twin studies (specially in America and Sweden) have established this fact and we've already have unusual grade of certainty (as compared to other findings in social sciences) about it.

      @NoahSmith: "I don't think there's enough evidence to make this conclusion."

      Define "enough". I recommend to you Bryan Caplan's last book ("Selfish Reasons to Have More Kids"), which makes a very good work (as we'd expect from him) in assessing the State of the Art about this issues.

      http://en.wikipedia.org/wiki/Selfish_Reasons_to_Have_More_Kids

      --Anônimo

      Delete
  41. "Academic racism has a K=N problem"

    Don't we all?

    ReplyDelete
  42. Anonymous1:43 PM

    The problem with Noah's post is that it assumes Wade's book has no scientific evidence re: genes, and is instead just telling fanciful "just so" stories that are overfit to the evidence. Perhaps Wade's book does have some such stories, but it is utterly bizarre for Noah to pretend that population genetics doesn't even exist.

    ReplyDelete
    Replies
    1. "Perhaps Wade's book does have some such stories, but it is utterly bizarre for Noah to pretend that population genetics doesn't even exist. "

      Where does he pretend that?

      Delete
    2. Anonymous2:32 PM

      By ignoring it entirely.

      Delete
  43. Here's a question: for those claiming that 'race' predicts IQ, which then predicts economic status, what precisely is 'race'? I find it hard to believe that 'race' = skin colors. Further, the focus on 'race' seems misplaced; one definition may show between 'race' variability, but how does one explain within-'race' variability? This distinction seems like a line-drawing fallacy. If we're going to break persons down into groups, why stop with 'race'?

    ReplyDelete
  44. I will note sociobiological theories of human differences are held to a higher standard than most other theories. Maybe that's okay. But it should still be noted. People jump to accept a lot more speculative things in genetics or sociobiology when they suit their purposes -- like gays are born gay, or some such thing.

    I frequently see demands to see the genomic evidence for claims about the biological origins of racial or ethnic variations in some behavioural trait, like IQ or time preference or whatever. “What alleles have been isolated and what traces of selection have been identified?” That’s of course a very tall order since, it’s still difficult to tease out the genomic correlates of a complex physical trait like height.

    But do such critics make the same demands for all evolutionary speculations ? The evolutionary literature high & low has always brimmed full of speculations about selection mechanisms. Hell, Richard Dawkins has had a whole career coming up with just-so stories of natural selection. While sex-selective arguments in evolutionary pschology have been a wee bit more controversial, they still don’t meet the kind of heated resistance from scientists and lay people alike that racial & ethnic arguments seem to do. Can you even have theoretical biology without inferences from phenotype ?

    ReplyDelete
    Replies
    1. And, as I've stated several times, the "evidence from phenotype" in behavioural genetics (twin & adoption studies) isn't perfect but is a lot stronger than your typical social science correlational evidence with lots of bogus qualitative proxies. When you can control for genes (with identical twins), that's already a lot.

      Delete
    2. Anonymous8:09 PM

      What you claimed here is pretty much exactly the opposite of true. Dawkins along with many of his colleagues like Hamilton over the years contributed incredibly strong approaches involving observation and scientific empiricism combined with pure mathematics, not story-telling. Those people understood and still understand game theory vastly better than the entire profession of economics, including pretty much all active Nobel Prize winners, for instance.

      Delete
    3. I said there are frequent demands for allele identifications, etc. when it's claimed, based on observational data, that there are inherited behavioural differences between certain groups. Are most of Dawkins reasonings in his many popular books supported by genomic data ?

      Delete
  45. Now that's a sweeping, general assertion that A) I'm not prepared to accept, but B) seems like it could be used to justify just about any conclusion.....That's a classic sign of overfitting. If your theory can subsume anything (with appropriate addition of assumptions and new ideas), you're almost certainly not going to be able to predict anything.


    But didn't I already predict (or, rather, pre-acknowledge) you would say that ? That's why in the end what matters is the genomic corroboration or falsification of population genetic differences in behavioural traits. Our bickering about inferences from phenotype will be rendered irrelevant. Then our social models can be recalibrated with different assumptions and parameters.

    Can you imagine Ricardian equivalence or real business cycle theory or even an economy filled with permanent-income consumers, in a world that has accepted genetically mediated differences in time preference ???

    ( I really think you underrate the possible boost to progressives from accepting the sociobiological theories of human differences, but... )
    A) Pretty much any human group large enough to qualify as a "race" can raise its average IQ to the mid-80s by eating right, not getting sick much as a kid, and growing up in a healthy emotional and intellectual environment. I suspect that the floor is closer to 90, actually.

    But then you are in agreement with the "race realists", except maybe on the number. African-American IQ of 1 SD less than the mean is usually thought the ceiling for Africans.

D) Therefore, I think the bulk of the GDP-IQ correlation is endogenous - something acts to boost GDP, and then IQs rise.

    Well, the correlation between IQ and growth is less than 0.6. So almost nobody disputes there's a lot of endogeneity.

    E) But those "somethings" rarely act, because low GDP is a stable equilibrium. Low IQ can cause low GDP growth, which keeps IQs low. Combined with (A)-(C),

    Also reasonable, in my opinion.

    James Watson did not even make a single comment about the heritability of IQ when he said that African IQ was lower. Yet everyone rushed to assume he meant it was some immutable condition. But I can't think of anyone who believes that about Africa. So even now aid efforts to Africa aren't targetted at boosting IQ through alleviating deficiencies in micronutrients, etc.

    this means that even if low IQ causes countries to be poor, that does not imply that stable population genetics is the fundamental cause.

    That's just semantics.
    which time I suspect that IQ-boosting technology will have made this question less relevant...

    Hoping the same thing !


    ReplyDelete
    Replies
    1. "Yet everyone rushed to assume he meant it was some immutable condition. But I can't think of anyone who believes that about Africa

      By which I mean, almost nobody believes the estimated Sub-Saharan African IQ average of 70 is the best that Africans do in an optimal environment. Most hereditarians put the ceiling at ~85. Environmental optimists who assume some hugely nonlinear effects on IQ from changing the US environment by a tiny amount in comparison with US-African environmental difference, has a much higher estimate of the ceiling….

      Delete
    2. Can you imagine Ricardian equivalence or real business cycle theory or even an economy filled with permanent-income consumers, in a world that has accepted genetically mediated differences in time preference ???

      That's quite a common modeling assumption, actually.

      ( I really think you underrate the possible boost to progressives from accepting the sociobiological theories of human differences, but... )

      Or maybe they have motivations that you either haven't thought of yet (due to spending too much time in the HBD bullshit-amplification chamber) or have chosen to ignore...;-)

      But then you are in agreement with the "race realists", except maybe on the number.

      "The number" strikes me as extremely important!!

      African-American IQ of 1 SD less than the mean is usually thought the ceiling for Africans

      I HEAVILY doubt this is the ceiling for the mean. First of all, many researchers find that the gap has narrowed to 2/3SD, and that seems plausible. Second, I suspect that changes in family life could erase some more of that gap. But most importantly, the current African-American population has been hugely negatively affected by selective immigration!!

      Any of these factors by itself means that the "natural race gap", if it actually exists, is going to be less than 1SD. I suspect all three are true, meaning I suspect that the "natural race gap" is a lot smaller than 1SD.

      That's just semantics.

      But it's the exact same point that you said was "Also reasonable, in my opinion" a few lines above...???

      The point is, even if stable population genetics were 0% responsible for IQ differences, initial IQ could still exert a strong causal effect on GDP, through the non-heritable component. That is a key point!!

      And also remember that the "heritable component" observed in Western societies is not the same as the "heritable component" in a poor country, since poor countries will probably have many more environmental things that affect IQ.

      Most hereditarians put the ceiling at ~85

      I suspect it is quite a bit higher, for the three reasons mentioned above.

      But even so, think about model uncertainty. The "ceiling" that the HBD people believe in is the maximum observed value (actually less than the observed maximum, as I noted above). To think that that an observed value is the theoretical maximum takes a leap of confidence that I can only ascribe to motivated reasoning...

      Delete
    3. I HEAVILY doubt this is the ceiling for the mean. First of all, many researchers find that the gap has narrowed to 2/3SD, and that seems plausible.

      In children, who show more plasticity. Black adults always test lower than black children. I don't know if the same set of children whose scores were reported by Flynn in 2006 have been re-tested as adults.

      "To think that that an observed value is the theoretical maximum takes a leap of confidence that I can only ascribe to motivated reasoning"

      I'm not closed to the possibility that it's 90. I mean, in the Bahamas for example it is supposed to be 90 or so. But let's see it happen first unambiguously in the USA.

      But it's the exact same point that you said was "Also reasonable, in my opinion" a few lines above...??? The point is, even if stable population genetics were 0% responsible for IQ differences, initial IQ could still exert a strong causal effect on GDP, through the non-heritable component. That is a key point!!

      I guess it depends on what you think the ceiling is. If the ceiling is even just 2/3 SD below the mean then the development trap you're talking about is ultimately genetic in origin. I think it's perfectly reasonable to say environmental deprivation has prevented Africans from realising their full IQ potential which prevents them achieving higher GDP which prevents them from improving their environments which...

      And also remember that the "heritable component" observed in Western societies is not the same as the "heritable component" in a poor country, since poor countries will probably have many more environmental things that affect IQ.

      Yes yes yes. Heritability varies by environment and is higher in developed countries than in developing countries. All of that is implied even by the assumptions that SSA IQ = 70 and max. potential = 85 or 90.

      Delete
    4. In children, who show more plasticity. Black adults always test lower than black children. I don't know if the same set of children whose scores were reported by Flynn in 2006 have been re-tested as adults.

      Wouldn't children be a better measurement, in order to get rid of nurture effects? Unless something big happens in puberty...

      I'm not closed to the possibility that it's 90.

      How about 95? Remember the selective immigration effect. Remember the feedback effects of growing up with broken family lives. You're only willing to go up to 90? Remember, we're talking about the maximum possible average here. Don't defend the "85" reference point number to the point of irrationality just because you've heard it a million times!

      I guess it depends on what you think the ceiling is. If the ceiling is even just 2/3 SD below the mean then the development trap you're talking about is ultimately genetic in origin.

      How about 1/3SD? If it was 5 average IQ points we were talking about here, would it be that big of a deal?

      Delete
    5. Wouldn't children be a better measurement, in order to get rid of nurture effects?

      They do that in twin & adoption studies. Here you are talking about getting rid of nurture effects through the testing of 12-year olds. Besides, there are no familial nurture effects. More precisely, they show up in childhood but disappear in adulthood. That's a solid finding. Look up "shared environment", "behavioural genetics". What ever enduring nurture effects exist, they are non-familial or "nonshared environment" and no one has the slightest idea what they are. Changing the environment is almost as difficult as changing genes !

      How about 95? Remember the selective immigration effect. Remember the feedback effects of growing up with broken family lives. You're only willing to go up to 90? Remember, we're talking about the maximum possible average here. Don't defend the "85" reference point number to the point of irrationality just because you've heard it a million times!

      Actually the range for the ceiling for SSA that I'm willing to entertain is 80 to 90. For the following reasons :

      Heritability of intelligence in developed countries is high. Could be as high as 0.80. No one knows what is the rest of the variance.

      There are no family effects to speak of. Only the error term optimistically dubbed "nonshared environment".

      African-Americans are ~20% European on average. Outside the USA, admixture rates vary in the diaspora, except in Haiti which appears to have very little European admixture.

      So if anything, whatever is the maximum for African-Americans, it would probably be a little higher than for Africans.

      Delete
    6. What ever enduring nurture effects exist, they are non-familial or "nonshared environment" and no one has the slightest idea what they are.

      Yes...

      Heritability of intelligence in developed countries is high. Could be as high as 0.80. No one knows what is the rest of the variance.

      The general figure is about 0.5 (source: http://en.wikipedia.org/wiki/Heritability_of_IQ), but in any case, heritability can't explain the Flynn Effect, which is substantial, so I don't think that is strong support. Don't subscribe to irrationality in order to keep from modifying your beliefs...

      African-Americans are ~20% European on average. Outside the USA, admixture rates vary in the diaspora, except in Haiti which appears to have very little European admixture.

      So if anything, whatever is the maximum for African-Americans, it would probably be a little higher than for Africans.


      Ah, but that only matters under the hypothesis that the groups have meaningfully divergent stable population genetic determinants of IQ in the first place. The reasoning is circular.

      What you're saying is: "Africans have a big stable population genetic IQ gap with Europeans --> African-American IQ scores are pulled up by white admixture --> Africans have a big stable population genetic IQ gap with Europeans."

      Don't subscribe to irrationality just to hang onto existing beliefs!

      Also, you haven't even addressed the selective immigration point, which i think is a big one...

      Delete
    7. From Noah's Wikipedia link :

      "A 1996 statement by the American Psychological Association gave about .45 for children and about .75 during and after adolescence.[7] A 2004 meta-analysis of reports in Current Directions in Psychological Science gave an overall estimate of around .85 for 18-year-olds and older.[8] The general figure for heritability of IQ is about 0.5 across multiple studies in varying populations.[9]"

      That accords with Plomin's Behavioural Genetics texbook.

      heritability can't explain the Flynn Effect, which is substantial, so I don't think that is strong support. Don't subscribe to irrationality in order to keep from modifying your beliefs…
      
The factors that are driving the Flynn Effect are not the factors that drive the racial gap in IQ. This is not a causal statement, it's a simple "accounting" statement of the factors. Even Flynn agrees on. Please look up the details.
      Ah, but that only matters under the hypothesis that the groups have meaningfully divergent stable population genetic determinants of IQ in the first place. The reasoning is circular. What you're saying is: "Africans have a big stable population genetic IQ gap with Europeans --> African-American IQ scores are pulled up by white admixture --> Africans have a big stable population genetic IQ gap with Europeans."

      Europeans, Africans, South African "Coloureds", African-Americans, Afro-Brazilians, Dominicans, Haitians, etc. have all been tested separately. IQ correlates with European ancestry, at population level. Where's the circularity ? Ideally you should directly make pairwise comparisons of admixture rates and IQ scores for a sample of individuals, but that's probably not feasible poitically now.

      Also, you haven't even addressed the selective immigration point, which i think is a big deal

      You could always study higher-income African-Americans (but that's been done and the results are not favourable), or move the studies to Brazil.

      Delete
    8. From Noah's Wikipedia link

      You can address me in the second person, I'm right here...

      Europeans, Africans, South African "Coloureds", African-Americans, Afro-Brazilians, Dominicans, Haitians, etc. have all been tested separately. IQ correlates with European ancestry, at population level. Where's the circularity ?

      I refer you to my previous explanation. Yup, it's circular, despite the things you mention. Think about it.

      You could always study higher-income African-Americans (but that's been done and the results are not favourable), or move the studies to Brazil.

      I utterly fail to see how that corrects for the effects of selective immigration.

      African-Americans (and African-Brazilians) were abducted, undoubtedly from the poorest classes of society, to be menial laborers. That seems like it would select for low IQ.

      If you refuse to deal with that effect, or wave it away, then you just seem utterly irrational to me.

      favourable

      This spelling tells me that you are not American, which means I should expect more dispassionate rationality than I'm getting so far... ;-)

      Delete
    9. "African-Americans (and African-Brazilians) were abducted, undoubtedly from the poorest classes of society, to be menial laborers. That seems like it would select for low IQ. If you refuse to deal with that effect, or wave it away, then you just seem utterly irrational to me.

      I've been misunderstanding you. I thought for some reason you've been speaking of the impact of low-skill immigration on African-Americans.

      But I think I already addressed this issue of African slavery : we don't know where in the African IQ distribution the slaves came from. But if the mean IQ of slaves was lower than the African mean, then what you're saying is the slave trade must have improved the mean IQ of remaining Africans. Who knew you might advance a eugenic argument against natural racial disparities !

      Delete
    10. we don't know where in the African IQ distribution the slaves came from

      Isn't it highly likely that they were not the local aristocrats, wealthy merchants, or skilled artisans in the countries from which they were sold into slavery? ;-)

      Sure, we don't know. But highly likely, right? ;-)

      But if the mean IQ of slaves was lower than the African mean, then what you're saying is the slave trade must have improved the mean IQ of remaining Africans.

      Maybe a bit, but how much of Africa's total population was sold into slavery? Not enough to make that much of a difference, except perhaps in certain locales, for a short time.

      You're still avoiding or eliding this very good and important point in order to defend a maintained hypothesis. Stop doing that. It is irrational and unreasonable, and there is no good reason to do it.

      Delete
    11. No, I admit it's a good point. And I've even told you I wondered about it myself. But I was ignorant. There's no evidence that any systematic environmental effects can depress IQ by 25-30 points ! Iodine deficiency or lead exposure doesn't depress IQ by so much. Maybe you can come up with nonsystematic effects, like 300 million children being kicked in the head.

      Delete
    12. Also, it's estimated that 25% of Sub-Saharan Africans are undernourished. "Severely malnutritioned" would be a smaller subset. If we assumed that this rate of undernourishment depressed potential IQ by 30 points, then we should see either a high standard deviation of African IQ (because there are a lot of people at the right tail who aren't suffering malnutrition) or a bimodal distribution created by the division between the undernourished and the nourished. But there's no such evidence.

      Delete
    13. No, I admit it's a good point. And I've even told you I wondered about it myself.

      Thanks! :-)

      But now, because of that good point, we're back to talking only about Africa?

      The farther we get from developed countries, the less confident I become that I have any idea what's going on. The reason is that while we can still make point estimates about the sizes of things, the variances of those estimates will be larger, and thus we should have much less confidence in our off-the-cuff theorizing (which already should not have inspired a huge about of confidence).

      I expect that soon we'll see a couple regions in Africa really start to industrialize. And then the IQs there will start to diverge greatly from the other regions. And academic racists (the whole HBD crowd) will say "Oh, but THAT African subgroup was always a lot smarter than the others!" It will be a just-so story, mad up in order to defend the central thesis that populations have fixed average IQ levels.

      Or maybe not. We'll see. In the meantime, I think you have a good idea now about why I'm so skeptical of everything the HBDers say!!

      Delete
    14. Anyway, I think this exchange has clearly showed the amount and depth of motivated reasoning and confirmation bias in the academic-racist community...if you're the most open-minded they've got, then my criticism was far too mild.

      Delete
  46. "Can you imagine Ricardian equivalence or real business cycle theory or even an economy filled with permanent-income consumers, in a world that has accepted genetically mediated differences in time preference ???"

    On second thought the modelling ingenuity of the crazies knows no bounds.

    ReplyDelete
  47. Anonymous6:15 PM

    It's worth noting that in England, where I live,

    1) Blacks are more likely to attend university:
    http://www.timeshighereducation.co.uk/news/geography-and-race-affect-university-attendance/2011226.article
    2) Women of Black Carribean heritage earn more than white women: http://stumblingandmumbling.typepad.com/stumbling_and_mumbling/2014/04/on-intersectionality.html
    2) There is a black-white achievement gap for kids taking their GSCEs (exams for all pupils aged 15/16), but it's so tiny as to be insignificant: http://analysis.mkiobservatory.org.uk/webview/index.jsp?v=2&mode=cube&cube=http%3A%2F%2Fanalysis.mkiobservatory.org.uk%3A80%2Fobj%2FfCube%2FPE130_C1&study=http%3A%2F%2Fanalysis.mkiobservatory.org.uk%3A80%2Fobj%2FfStudy%2FPE130&context=http%3A%2F%2Fanalysis.mkiobservatory.org.uk%3A80%2Fobj%2FcServer%2FMKi&top=yes

    I'm astonished that these statistics aren't cited more frequently when this debate comes up in the US. Maybe it has something to do with the insularity of American political discourse.

    ReplyDelete
    Replies
    1. Thanks!! Good data.

      Delete
    2. But I believe British black IQs have been higher than US black IQ for some time. One would expect them to perform better than in the US. Also British blacks are overwhelmingly Caribbean origin. Caribbean-origin blacks in the USA also outperform not only African-Americans but depending on subgroup white Americans.

      Delete
    3. But I believe British black IQs have been higher than US black IQ for some time. One would expect them to perform better than in the US. Also British blacks are overwhelmingly Caribbean origin. Caribbean-origin blacks in the USA also outperform not only African-Americans but depending on subgroup white Americans.

      There's that "overfitting" I was talking about...

      Delete
    4. You could always a population-weighted mean of the diaspora....

      Delete
    5. Besides, where's the overfitting ? IQ predicts achievement, especially academic achievement. The end. I think that's your conceptual problem : there's no "race" variable in the regressions involving IQ.

      Delete
    6. Every time you hear about a group difference, you ascribe it to stable population genetics. I guess that's not overfitting, really...it's just making the same assumption to explain any observed data point. But it probable assumes more explanatory power for stable population genetics than actually exists...

      Delete
    7. No, there are two separate inferences here. Every time I hear about a group difference, I tend to think IQ predicts it well. Then a totally separate world generates the heritabilty estimates for IQ. There are reasonable objections to putting the two together, but you haven't named them. (Those would be : heritability estimates come only from developed countries, and within-race studies are far more numerous than transracial ones.)

      Delete
    8. And by "group difference" I mean some group disparity or inequality in social outcomes.

      Delete
    9. Every time I hear about a group difference, I tend to think IQ predicts it well.

      I know. The problem is this. You hear about 2 groups of black people in the Western hemisphere. One group has a higher average IQ than the other. You assume that the high-IQ group is less representative of Africans than the low-IQ group. The high-IQ group is treated as the "exception", and the low-IQ group as the "rule". Makes no sense to me.

      Nor does a population-weighted average, since selection effects probably dominate in both populations.

      Delete
    10. Basically, you still haven't dealt with the issue of selective immigration, and the resultant implication that African-American reality is not a great proxy for African potential.

      There are other reasons it's a poor proxy, but this is a big one, and you should deal with it. Don't subscribe to irrationality just to defend a maintained hypothesis.

      Delete
    11. Anonymous4:52 AM

      It's worth pointing out that there are twice as many Black Africans in Britain as there are Black Carribeans. British Nigerians are basically the master race, only slightly behind the Chinese: http://en.wikipedia.org/wiki/British_Nigerian#Education

      Delete
    12. it's apparently now the same in the USA.

      Delete
    13. "Higher-income African-Americans, who presumably are less affected by low-wage immigration than low-income African-Americans, also have lower average IQs than other groups matched by income."

      Evidence? Source? I could be wrong, but I'm highly suspect of this factoid.

      Delete
  48. Anonymous6:23 PM

    Sorry - obviously, that should read "Blacks are more likely to attend university *than whites*"

    ReplyDelete
  49. Oh no, Noah! Don't poke the beast!

    And you thought you had trolls before. :)

    ReplyDelete
    Replies
    1. Remember, I am Smith, the Untrollable...

      Delete
    2. Anonymous8:44 AM

      I think you've met an equal in @pseudoerasmus: (like you) he has lots of IQ points, knowledge, rhetorical ability and disposition to engage in online discussion:

      http://xkcd.com/386/

      -Anônimo

      Delete
  50. Anonymous2:06 AM

    Thanks Noah for the really entertaining post and discussion. After 100+ comments is there anything you would change in your final "so what" paragraph? Or are you still happy to sit in amusement and watch as this debate continues on without ever concluding? One last question, why do you choose or find it acceptable to engage with topics like "technology"or "institutions" which as you said face similar over fitting issues but not this topic? What is the difference and why is it important?

    ReplyDelete
  51. Anonymous2:38 AM

    Noah, just for fun. Do you think differences in the star wars galaxy are more the cause of "institutions", and by that I mean that empire, or genotypes. On the one hand, genotype differences are waay more pronounced in star wars vs earth but then again the empire does suck. Or is it more likely that geography is the single largest factor.

    ReplyDelete
  52. For K>N you assume that the unit of analysis is each 'race' rather than individuals who take different values for 'race'. It is possible to study, especially in places such as the US where there is enough variation in race within the country to isolate it from other possibly confounding variables.

    ReplyDelete
    Replies
    1. Anonymous8:53 AM

      @AnMailleach: It is possible to study, especially in places such as the US where there is enough variation in race within the country to isolate

      Excellent point! I think Brazil is even better than the US: excellent (for a developing country) free data in IBGE site (our governmental statistics Bureau) + unparalleled ethnic diversity + lots of regional economic disparities + Japanese (biggest Japanese ancestry population outside Japan), Italian, Portuguese, Syrian and German immigration.

      -Anônimo

      Delete
  53. @Anônimo, as usual we are going in circles. There is even a divergence in the successful adoption of ideological "soft factors", and that divergence has a pattern. But you already know this. As for why the industrial revolution took so long to occur, or why the Song did not experience industrial takeoff, or why China in the late 19th century did not take off at the same time as the Japanese -- is there a need for some grand unification theory that accounts for everything from 1200 AD to 2014 ? Can't economic history have scale-specific explanations like everyone else ?

    ReplyDelete
    Replies
    1. Anonymous9:11 AM

      @Pseudoerasmus: "Can't economic history have scale-specific explanations like everyone else ?"

      My messianic fervor with McCloskey' late oeuvre come from the fact that I believe this awesome woman found the factorthat explains Netherlands (almost) and England (full) takeoff! We're living in interesting times for growth theory, my friend! Soft (=ideational) factors we'll be finally hardened and quantitied!

      -Anônimo

      Delete
  54. Wow that is a lot of comments.

    Ignoring what anyone else has said it seems that there is a fallacy of composition problem in many "race" studies. There are certain aspects of groups of people that are certainly heritable. Conventional groupings by race are just too rough a filter to be meaningful. It does mean certain studies of human genetic variation are almost taboo because people are afraid of confirming some racial bias. But it would or at least should be uncontroversial to make a statement like people of native American ancestry are much more likely to be alcoholics. There is some genetic basis that explains why the worlds greatest sprinters all seem to be Jamaican or even specifically a small subset of Jamaicans. (Although I am loathe to say they are the "fastest" people as access to American college sports allows them to perhaps develop that physical skill beyond perhaps other groups)

    The problem arises if we expand the aspect of a group beyond the limited definition to paint an entire arbitrarily defined group with a broad brush. Just because a trait appears within a small subsection of a racial group doesn't mean anything larger about that racial group. It is a perfect example of the fallacy of composition. This is perhaps why it is so persistent a problem where it is easy to confuse the traits of some arbitrary group with a subset of that group.

    ReplyDelete