« Youth Unemployment and the Apprenticeship Gap | Main | What We Talk About When We Talk About "IQ:" Algebra and School Quality »

May 10, 2013


I'm glad you drew the SAT comparison; I was reminded of it too by this unfolding story.

The New Yorker profiled Stanley Kaplan a little over a decade ago, explaining how the SAT was created to try discern intelligence and ability instead of how much an applicant had studied/crammed, reflecting the concern of the time that certain groups (read: recent immigrants or others lower on the social scale, like Eastern and Southern Europeans and many Jews) were performing well academically simply because they studied more, not because they were as smart as other white groups. By comparison, Kaplan saw the SAT as just another test, and designed a program to prepare students accordingly.

I am a PhD candidate in educational testing and psychometrics. As part of my dissertation, I have had to review this (sorry) chapter in the history of my field. And I completely agree with everything in the original post. The only intelligent things I have to add are: misunderstood and cherry-picked intelligence testing research are really catnip for lots of self-identified conservatives, and there’s basically no good-faith explanation for why, and early psychometricians have a touching faith in the idea that one can measure an unobservable thing (“a construct” to use modern terminology); that the test measure exactly what the psychometrician choose it to measure, neither more nor less.

For broader and deeper historical treatment of the ways in which disability has been used as a barrier to immigration, try the work of Douglas Baynton. I'd recommend either "Defectives in the Land" (2005) or "Time and Disability in the Age of Eugenics" (2011).

Well, other than every single one of your facts being incorrect, this was a pretty thoughtful and well reasoned post. Shall I count deconstruct your pitiful list of errors?

1) Barring the mediocre performance for some immigrant groups on tests of verbal ability (which may be dismissed by their lack of familiarity with the language of their host country), the consensus among psychometricians is that IQ gains from practice are hollow, resulting in a temporary boost in a narrow range of cognitive abilities, and do not reflect any corresponding increase in overall intelligence.

Practicing endlessly on the Raven's Progressive Matrices may temporarily increase your ability to rotate objects in your head and recognize patterns, which should increase your performance on that "IQ" test by a couple of points -- but this documented increase never translates into a greater facility with numbers, or an easier time learning physics or chemistry. The same goes for tests of digit span, vocabulary, pitch recognition, et al. You can read and re-read Dostoevsky until your eyes bleed from strain, but that's just too bad -- all else held equal, that kid who actually scored 148 on the WISC will kick your ass not only in English, but also in mathematics, chemistry, physics, and every other subject you can think of. Which brings me to point two:

2) What we call "IQ testing" is the reflection of the astonishing finding that all cognitive abilities which are valuable in the context of a modern, industrialized society are positively correlated. Seemingly unrelated skills such a backward-digit span, pitch recognition, and the ability to understand a passage of written prose all tend to rise or fall together -- and these correlations are not trivial, but the highest ever documented in the social sciences. Statistically speaking, Sandra who scores high on physics is also likely to write better essays than Mindy who scores low. Occasional exceptions to this rule do not disprove this overall trend. A narrow, unidimensional measure of a "linear" sort of intelligence, IQ surely is not.

As for the notion that other cultures have definitions of intelligence that are just as valuable as our own, spare me your chicanery. In the 21st century, there is absolutely no alternative to industrial civilization. The fruits of modern science and engineering have produced a standard of living that is without parallel throughout the entire history of our species. We owe our long lifespans today to immunology, modern sanitation, physics, chemistry, and the fruits of modern medicine -- and it is just too bad that IQ tests do an excellent job measuring your ability to succeed in any and all of these fields. Emotional intelligence alone does not build skyscrapers or bridges; neither will it find the cure cancer or sanitize your water supply.

It is indeed that primitive cultures do honor alternative definitions of intelligence that are perhaps better suited to their environment. For example, the Waripari of Australia teach their children to search for water in the middle of the desert, otherwise they will die from thirst during the dry season. But we don't. We've got geologists and engineers for that. Third world immigrants from hellholes like Haiti and Guinea-Buissau are busy trying to emigrate here, and nobody ever tries to relocate themselves there -- so much for their native definitions of intelligence. Once you discount the three Rs (reading, 'riting, 'rithmetic), what's left?

3) The twin studies and longitudinal adoption studies demonstrate unequivocally that heredity, not the environment is responsible for the vast majority of the variation in IQ scores. If you have ever studied psychometrics, you may recall that the narrow-sense heritability of IQ is roughly 0.5, meaning that heredity is responsible for 50% of the variance in IQ, within the range of environments sampled by the twin studies. This estimate is a _floor_, not a ceiling -- people conveniently forget that the "environmental" component also factors in things like dominance, epistasis, and other genetic effects that usually add up to a total of 75% by adulthood.

And if you have heard somewhere that twin studies are racist, evil, and irredeemably flawed (all of these are lies), we have multiple GWAS studies that largely replicate the same heritability figures for IQ. None of these are biased by the confounding variables that are traditionally blamed for inflating the genetic component to IQ.

By the way, that 25% that is environmental has almost nothing to do with parenting, school quality, choice of neighborhood, or any other variables that children in the same household share in common until they leave the nest. The longitudinal adoption studies are especially compelling on this matter. (See: the Colorado adoption study) Developmental psychologists (who, by the way, are wrong about just about everything) are beside themselves trying to explain away the pesky fact that the children of low IQ parents, who are adopted by the wealthy, score no higher on IQ tests at age 21 than their siblings who were never given up for adoption.

If you are incredulous, consider that IQ scores are no higher among whites in egalitarian, Nordic Sweden than they are in socially stratified America, and may even be slightly lower. How's that for the power of the environment? And if you don't believe my words, controlling for race, Americans also score higher on tests of actual academic achievement than every single country in the world. White Americans scores higher on the PISA than any European country. Asian Americans too sweep the board on these tests, scoring higher than even Chinese in Shanghai and Singapore. Latin Americans, too, score higher than any Latin American country that has been tested, although they still score significantly lower than the former two groups.

4) In order for egalitarian well-wishers to dismiss away the success of immigrant groups who succeed in our supposedly racist society, in spite of all social odds, you are left in the awkward position of having to argue that Asian and Jewish immigrants have been *more* privileged than their white counterparts throughout the entirety of American history. And of course, we know this just isn't the case. Sociologists and historians have documented just exactly how Jewish immigrants from Russia lived in places like New York City at the turn of the 20th century -- they were wretchedly poor, crowded twenty to thirty a room in ramshackle urban tenements, and chronic sufferers of "crowding" diseases such as tuberculosis and whooping cough. And yet today their descendants are responsible for over 25% of our Nobel Prizes in science. Ditto for Chinese Americans, who are over-represented in most fields that require scientific or mathematical ability. You try explaining *that* away through nurture.

Self-selection by immigrants doesn't explain away this phenomenon, either. Japanese Brazilians were imported to the sugar cane fields of Brazil as manual laborers. And yet mysteriously, they too face affirmative action "quotas" restricting their numbers at elite universities in Brazil. There isn't a single place on *earth* where east Asians do not constitute a model minority, and incidentally, these ethnic groups also tend to score higher than the white mean on IQ tests. It's true in Malaysia, where the Chinese are the descendants of illiterate tin miners. And in Canada, where they were hired to work the railroads.

By the way, the idea that Jewish people once scored as low as African Americans on IQ tests is a complete lie, and it originates from a discredited article by the Marxist psychologist Leon Kamin. European Jewish people have always been high academic achievers in the United States and elsewhere. The 1920s was when Ivy League universities first instituted quotas restricting the enrollmment of Ashkenazi Jews, because they scored too high on their exams of cognitive ability. Contrary to popular belief, no politician in America ever cited the supposed "cognitive inferiority" of Jews as a rationale for the Immigration Act of 1924. Henry H. Goddard, that architect of American eugenics, himself recorded a higher documented IQ for European Jewish immigrants -- in 1926, he oversaw a thesis by a Jewish graduate student named Irma Cohen documenting years of psychometric research proving this trend.

The funniest thing about this hysterical brouhaha over race and IQ is that nowhere in his thesis does Richwine ever posit that ethnic differences in IQ are fixed and hardwired. I challenge you to find a single passage where he does. (Incidentally, neither do Herrnstein and Murray in the notorious Bell Curve.) Hispanics indeed have a lower documented IQ than white Americans, and the gap persists throughout the third and fourth generation, closely following their poor record of academic achievement. (See: "Generations of Exclusion" by sociologists Telles and Ortiz.) Whether or not genetics plays any role is moot. If there were *ever* such a thing as an ethnic group in America that demonstrates minimal progress up the socioeconomic ladder after a hundred years, it would be pretty damn stupid to skyrocket their numbers by rewarding millions of lawbreakers with US citizenship.

Fact of the matter is, whether or not IQ is real, you people simply do not have an intellectual leg to stand on -- your only recourse left is to smear and tarnish the reputations of psychometricians like Eysenck or Jensen, silence any and all debate through hysterical cries of "racism", or permanently exile any researchers in the social sciences who dare to espouse politically incorrect points of view. Deconstructing Richwine's thesis through actual facts is simply too complex for some people. No -- it's better concoct fish tales about magical gains in IQ through Head Start and early preschool, the pernicious impact of "stereotype threat" on minority performance, and novel teaching methods that mysteriously close any and all racial gaps on test performance.

This is what you charlatans in the educational industry are all but good for. And I count myself lucky that I see your chicanery for what it truly is. High IQ or not, the vast majority of people in this country are fools, and it is no credit to the chronic dishonesty from "experts" such as the ignoramus who penned this blog entry.

Look, empirically there are 3rd, 4th, and 5th generation Mexican-Americans, and their educational and other outcomes have been studied (by left-leaning researchers who attribute gaps to culture/discrimination, not genetics):


Convergence stops after the 2nd generation, leaving the large gap for the 3rd, 4th, and apparently 5th generations. Richwine discusses this work in his dissertation.

It doesn't matter why for Richwine's argument, the observed fact is that differences in education, taxes paid, welfare use, crime, and poverty have not converged for later-generation Mexican-Americans as they did for later-generation European immigrants. And so there is little reason to expect this to suddenly change for new immigrants.

"Yet in the life of the average, individual human, those "innate" genes are vastly, vastly overpowered by the effects of environment: decent nutrition; an emotionally stable, vocabulary-rich home life;'

Most of the variation in fluid IQ in rich countries is attributable to additive common genetic variants, a subset of all genetic variation in DNA-based studies (a number of which have come out in the last 2 years largely confirming the hereditarian claims):


" If two race-similar individuals are compared, the person with the higher IQ will often have superior social outcomes. He is more likely to graduate high school or get a high-paying job. Yet the evidence suggests that IQ itself--as opposed to all the other social factors correlated with IQ, like parental income--is responsible for only a small fraction of this difference in achievement. Not to beat a dead horse or anything, but correlation does not imply causation. "

If you look at sibling pairs, with the different siblings having different IQs (which controls for the shared home environment) you get effects on income large enough to explain group differences.


"IQ is one predictor of success on the labor market, but it is not the only or even the most important factor."

This is true, but irrelevant. There is a lot of variation in outcomes at each level of IQ based on other factors. Likewise, there is a lot of variation in education and income within ethnic groups. But the differences in IQ are large enough to predict labor market differences by ethnic group among the native-born (first-generation immigrants do lack language skills, access to education, etc).

"Not to beat a dead horse or anything, but correlation does not imply causation."

Which particular genes you get from each parent (you only get a random 50% of each parent's genes on average, and different ones from your siblings) is random, and can be used to show causation. Likewise adoption studies, which show only temporary effects on IQ and income that mostly fade out in adulthood, with children winding up resembling their biological rather than adoptive parents, can show causation.

And again, it's not as though the entire Mexican-American population is going to have its children put into middle-class adoptive homes. For purposes of prediction, Richwine is on solid ground that performance gaps have been stable over multiple generations of native-born Mexican-Americans, and we don't have reason to expect that to suddenly change now.

And why did Christopher Jencks, renowned liberal social scientist and editor of the liberal magazine The American Prospect, sign off on Richwine's dissertation at Harvard?


I suggest you contact Jencks, or James Flynn (of the Flynn Effect), or another leftist scientist in the field and ask them what Richwine said that was false or empirically unsupported. You'll come up empty-handed.

The book I usually recommend to people who think that IQ differences between "racial" groups (whatever that means) are meaningful is "The Mismeasure of Man" by Gould. He points out the obvious: Southern Europeans and Jews were regarded at the beginning of the 20th century the way that Richwine regards Hispanics today. Yet, somehow miraculously, the children and grandchildren of those early immigrants have greatly improved their average IQ scores, as they've been integrated into American society.

Richwine and his supporters are just the latest in a long line of racist fools.

It's quite hilarious that anybody would still recommend the "Mismeasure of Man". In the age of Google, how intellectually lazy and incurious do you have to be to actively that old and discredited book? Go read the literature PghMike4, you ignorant fool.

NYTimes article on Gould's misrepresentation of data and false accusations:




Paul Krugman on Stephen J. Gould:


What I encountered were quite a few references to Stephen Jay Gould, hardly any to other evolutionary theorists. Now it is not very hard to find out, if you spend a little while reading in evolution, that Gould is the John Kenneth Galbraith of his subject. That is, he is a wonderful writer who is beloved by literary intellectuals and lionized by the media because he does not use algebra or difficult jargon. Unfortunately, it appears that he avoids these sins not because he has transcended his colleagues but because he does does not seem to understand what they have to say; and his own descriptions of what the field is about - not just the answers, but even the questions - are consistently misleading. His impressive literary and historical erudition makes his work seem profound to most readers, but informed readers eventually conclude that there's no there there. (And yes, there is some resentment of his fame: in the field the unjustly famous theory of "punctuated equilibrium", in which Gould and Niles Eldredge asserted that evolution proceeds not steadily but in short bursts of rapid change, is known as "evolution by jerks").

What is rare in the evolutionary economics literature, at least as far as I can tell, is references to the theorists the practitioners themselves regard as great men - to people like George Williams, William Hamilton, or John Maynard Smith. This is serious, because if you think that Gould's ideas represent the cutting edge of evolutionary theory (as I myself did until about a year and a half ago), you have an almost completely misguided view of where the field is and even of what the issues are.

God, these takedowns were gorgeous. I have just one thing to add: Dana, if you were so fussed about Jason Richwine and IQ, how come you didn't mention all this information when he wrote about teachers and cognitive ability? You should do your research!


Hellfireawaits killed it. Questions:

How can an IQ study done by whites be considered racist when the nonwhite group (Asians) scored higher than whites?

Shouldn't whites score higher than Asians in order for the study to be considered racist?

Did someone fudge the test scores or did hispanics score lower? If they did then whats the argument?

Are there any PhDs in physics that do not have an IQ that is well above average?

Are there any PhDs in any hard science with an IQ of 95?

The comments to this entry are closed.