Bryan Caplan  

Gladwell on IQ: My Chance to Answer Two Plausible But Over-rated Arguments

Are You Tired... Subprime Daily Briefing, Dec. ...

Robin was very impressed by Malcolm Gladwell's piece on IQ and the Flynn effect in The New Yorker. I wasn't; Gladwell leaves out a lot. (He also falsely attributes a bizarre view to Murray and Herrnstein, which the magazine has already had the grace to correct).

However, Gladwell does repeat two arguments that are persuasive, but don't really prove much. Lucky me: I've been waiting years for a chance to answer them!

Argument #1:

If I.Q. is innate, it shouldn’t make a difference whether it’s a mixed-race child’s mother or father who is black. But it does: children with a white mother and a black father have an eight-point I.Q. advantage over those with a black mother and a white father.
The problem is that there is a simple genetic explanation: In all likelihood, the black man who marries a white woman will be above his group average in success and IQ; the white man who marries a black woman will be below his group average in success and IQ. You might see an offsetting effect in the women (i.e., white women who marry black men may be below their group average in success and IQ, black women who marry white men may be above their group average in success and IQ). But since men's prospects in the marriage market are much more heavily linked to success than women's, the offset will probably be modest.

This doesn't mean that the genetic explanation is right, just that the mixed-race IQ pattern Gladwell highlights does not distinguish between the environmental and genetic stories.

Argument #2:

And it shouldn’t make much of a difference where a mixed-race child is born. But, again, it does: the children fathered by black American G.I.s in postwar Germany and brought up by their German mothers have the same I.Q.s as the children of white American G.I.s and German mothers.
The problem with this famous Eyferth study, which formed the backbone of Flynn's Race, IQ, and Jensen, is that it was a study of children. So? After Flynn wrote this book, behavioral geneticists gradually made the amazing discovery that the heritability of IQ (and many other traits) sharply rises as children grow up, while family effects on IQ fade out. The Eyferth results tip the scales modestly in an environmental direction, but primarily confirm what we already know: Children's environment has a substantial effect on their IQs when they are children. The truly probative study would have tested the IQs of the Eyferth subjects in adulthood. As far as I know that never happened, and now it's probably too late.

Again, none of this confirms the genetic explanation. My point is merely that a popular piece of counter-evidence is quite a bit weaker than it seems.

Comments and Sharing

TRACKBACKS (3 to date)
TrackBack URL:
The author at Tim Worstall in a related article titled Being Vile About the Sutton Trust Report writes:
    So we had the Sutton Trust report. Parental background continues to exert a very significant influence on the academic progress of children: o Those from the poorest fifth of households but in the brightest group at age three drop from the 88th percen... [Tracked on December 15, 2007 6:59 AM]
COMMENTS (21 to date)
Chuck writes:

There is a better explanation than that for argument 1 - intelligence could involve genes on the X chromosone and which would be expressed differently based on gender.

Brad writes:

That's a pretty egregious misstatement on Gladwell's part. It says a lot that they don't note the correction inline. It also says a lot about Gladwell's research on this. Skimming The Bell Curve over the course of a day would quickly reveal that even the tone of the book is entirely inconsistent with Gladwell's recounting. Murray and Herrnstein go so out of their way to not make any of the data or conclusions remotely possibly offensive to any semi-objective reader! Gladwell is fast becoming "big idea porn". His stories are fun to read, but he plays fast and loose with the underlying facts.

dearieme writes:

Neither of the arguments that you cite seem to come equipped with any info about the IQ of the parents. They therefore aren't capable of illuminating heritability.

Dan Weber writes:
There is a better explanation than that for argument 1 - intelligence could involve genes on the X chromosone and which would be expressed differently based on gender.
That would suggest that the different sexes could have different intelligence.

And we know that's impossible.

RICH writes:

"the black man who marries a white woman will be above his group average in success and IQ; the white man who marries a black woman will be below his group average in success and IQ."


Dwight writes:


(a) Please stop YELLING.

(b) What is a "quasi-racist assumption"? Is it something different than a racist one?

(c) Caplan has merely introduced a hypothesis. That is what you do when you don't yet have the data, but want to figure out where it might pay to gather some.

(d) Caplan has offered a suggestion for how people in fact behave, regardless of whether it's moral or immoral. If you think his hypothesis is wrong, please say why, rather than simply dismissing it with ad hominems.

Mason writes:

Thank you Dwight.

As for Caplan statement, it is easily defensible, and nothing to yell about.

1. Whites have higher IQs than blacks - Fact

2. People seek mates with similar IQs - Fact

Therefore we can expect mixed race couples to have IQs that differ from their group's average in the manner he suggested. (I haven’t seen any research on this, but it is a very logical conclusion to draw from points 1 & 2)

Rich writes:

Dwight - You make a fair point about hypotheses. But I think its naive to think that the line between a hypothesis and reifying assumption is anything but blurry.

Mason - I dont want to say anything rude to you. But what you wrote is completely.. ahem. misguided.

1. Whites have higher IQ's - fact? How and why is that a fact? That's precisely the issue in debate.

2. People seek mates with similar IQs - Fact. That's fine, but the reason for my outburst at Caplan's statement is that he presumes that white's already have a higher IQ, and therefore any white that doesn't will marry black? Or because a white man marries a black woman has a lower IQ? Either way, its a ridiculous assumption which presupposes the ultimate debate.


Mason writes:

Try a quick google search on race IQ and gap, and you'll find lots, here is one that I found.

Whether the gap is genetic or the product of the environment is being researched, but either way the same mating patterns would emerge.

If it makes you feel better I was shocked to read about it at first as well.

Psychohistorian writes:

"the black man who marries a white woman will be above his group average in success and IQ; the white man who marries a black woman will be below his group average in success and IQ."

This assumes random pairing, which is pretty absurd. Because of cultural barriers to interracial relationships (ranging from just not knowing any people of a given race to overt racism, and everything in between), there is reason to suspect that pairings will be non-random.

The best of my understanding, and I admit I am not aware of detailed research to back this up, is that a lot of highly educated black women marry white men because that is what is available in the social circle they live in. I do not know if that is true of black men.

Add to this consideration that (generally) the more educated someone is, the lower the likelihood of overt racism and (probably) the higher the willingness to be involved in an interracial relationship. If there's data that controverts this, I've got nothing against it, but I'm fairly sure this is true.

If the probability of being exposed to people of a different race is directly correlated with income or social status, as may be likely, and that is in turn correlated with IQ, then interracial couples may tend to be above average for both of their respective races.

Of course, these effects may not represent a majority of such relationships. But it is somewhat foolish to assume random pairing in such a complicated situation where pairing is very unlikely to be random.

On the opposite end, there is some evidence that IQ shows hybrid vigor, in which case the kids have higher IQs as a result of being mixed-race. If that effect applies, then it is quite possible that IQ could be highly genetic and show racial disparity and yet mixed race children perform higher than that alone would predict.

Patrick writes:

The Eyferth study is crap. Military recruits were and are subjected to IQ tests upon induction. The black G.I.s were not a random sample.

Rue Des Quatre Vents writes:

Despite Gladwell's haphazard play with the truth, his argument rests largely on the Flynn effect, which is something that Bryan has not explained away.

But the problem with Gladwell's argument is that it assumes an either/or conclusion--consider his repeated claims that IQ measures not intelligence, but how modern we are, or the quality of the environment surrounding us. (Naturally, this is the good liberal assumption: the lurking forces in the environment). But he overemphasizes this. You walk away from the article thinking he's arguing for a blank slate theory of intelligence. Which simply isn't true. See twin studies, separated at birth, adopted children and so on. Judith Harris's books should be sent to Mr. Gladwell.

LemmusLemmus writes:

I've heard that another problem with the Eyferth study (which I have never read) is that people of northern African descent were counted as "black".

TGGP writes:

I've linked to this here before, but Karl Smith has an interesting theory why the heritability of IQ increases with age. It has to do with one's environment being constrained by outside forces when young versus being an adult and being able to shape your own environment in a way that your genes predispose you to.

I think I remember something in "Norton Reader" by Gladwell where he said that just as in many traits men have higher variance than women, blacks could simply have higher variance than whites, explaining both why the top athletes and the special education classrooms are disproportionately black, but La Griffe du Lion claims here that whites have higher variance. Anyone have data to settle this? I usually just assume the same variance unless I have reason to believe otherwise.

Steve Sailer writes:

The Eyferth study was published in 1961 and hasn't been replicated since. This shows the degree of grasping at straws that the Gladwell side of the argument must resort to.

bilard writes:

You have very interesting site.Thanks for all.

frank cross writes:

Does anyone else find this discussion bizarre. It seems patently obvious that IQ has a substantial genetic component. It seems similarly obvious that it has a substantial environmental component.

When discussing differences between two groups, research like the mixed race study seem like relevant information. And methodological challenges that begin with unsupported assertions like "in all likelihood" aren't terribly compelling. It's a possible problem with the study but wants a little empirical support of its own

Biomed Tim writes:

"Either way, its a ridiculous assumption which presupposes the ultimate debate."

Rich, it's a ridiculous assumption only if you know it to be false. Do you in fact know it's false, or are you just making your own assumption?

Daniel Klein writes:

Suppose there were a serious study of white men who married black women. Suppose the study assessed their IQ relative to the average of white men. Would you expect their IQ to be lower than average? I wouldn't. For one thing, racial tolerance and marital deviance probably vary with IQ. Smarts are especially valuable to those who buck convention.

Robert Fogel married a black woman. Evidence that he is sub-par in IQ among Nobel laureates?

My point is obvious. How is that Bryan overlooks the obvious?

Daniel Klein writes:

I misspoke, I should have said: Bucking convention is especially viable and appealing to those with smarts.

abe writes:

It sounds like Bryan Caplan is officially an acolyte of Steve Sailer.

Comments for this entry have been closed
Return to top