From the Atlantic
What We Mean When We Say 'Race Is a Social Construct'
In a world where Kevin Garnett, Harold Ford, and Halle Berry all check "black" on the census, even the argument that racial labels refer to natural differences in physical traits doesn't hold up.
TA-NEHISI COATES MAY 15 2013, 12:55 PM ET
Most of the honest writing I've seen on "race and intelligence" focuses on critiquing the idea of "intelligence." So there's lot of good literature on whether it can be measured, its relevance in modern society, whether intelligence changes across generations, whether it changes with environment, and what we mean when we say IQ. As Freddie mentions here, I had a mathematician stop past to tell me I needed to stop studying French, and immediately start studying statistics — otherwise I can't possibly understand this debate.
Walter White. Chairman of the NAACP. Black dude.
It's a fair critique. My response is that he should stop studying math and start studying history.
I am not being flip or coy. If you tell me that you plan to study "race and intelligence" then it is only fair that I ask you, "What do you mean by race?" It's true I don't always do math so well, but I understand the need to define the terms of your study. If you're a math guy, perhaps your instinct is to point out the problems in the interpretation of the data. My instinct is to point out that your entire experiment proceeds from a basic flaw — no coherent, fixed definition of race actually exists.
I constantly hear this line of argument. Millions of people must find it incredibly persuasive, but it doesn't make any sense on two levels.
Consider, as an example of the complexities of racial classification, the President of the United States, Barack Obama. In philosophy debates, the current President of the United States is a stock example, so I'll use him too. The President seems like a reasonable example, right? I don't know what the President's IQ is, but it's obviously above average. Imagine that Obama were part of a study of IQ by race. I'll answer the questions in this FAQ:
Q. What race would Obama be classified under?
A. Black.
Q. Why?
A. Because he says so. (The White House announced that on the 2010 Census, the President checked only the "black" box.) The standard methodology in studies of race and intelligence is exactly the same as in the studies of race and discrimination cited by liberals: self-identification. People check whatever box or boxes they feel like, and that's what the researchers use.
Q. Why rely upon "self-identification?"
A. Because it's easier bureaucratically, for one thing. This isn't apartheid South Africa where bureaucrats told people they are different races from what they claim to be. Here, they just go with the flow.
Q. How accurate is this?
A. It's good enough for government work, evidently.
Q. But what if we gave genetic tests to all subjects in studies?
A. These days, that would be increasingly affordable.
Q. So wouldn't that change the results?
A. Yes. It would likely make the white-black IQ gap slightly worse than under self-identification.
Q. Huh?
A. Consider Obama, who had a sub-Saharan African mother and >99% white mother. Currently, because the President chooses to self-identify as black and only black, his above average IQ would be credited wholly to blacks. But, genealogically, he's half white and half-black. So, if his IQ were split among the white category and the black category, the white average would go up slightly and the black average would go down slightly.
Q. Is this true on average that genetically whiter self-identifying blacks average higher IQs?
A. Probably:
The quants at Human Varieties have been kicking this around and it appears to be modestly true, although the correlation isn't enormous. I could imagine that this might change. If Jason Richwine were put in charge of immigration policy and he only let in high IQ immigrants, over time the average IQ of people who are 100% sub-Saharan African might get pretty high.
Q. La-la-la-la, I have my fingers in my ears, I can't hear a word you are saying.
A. Let's put it another way: leaving out recent African grad school immigrants, nobody, white or black, has ever argued that blacker African-Americans are smarter than lighter ones on average. But let's just assume they are absolutely equal. Then what would be the effect of divvying up subjects in an IQ study by genetic background? C'mon, it's simple arithmetic.
Q. I'm a history guy, not a math guy.
A. Zero. You'd get the same results as now. So, even under the most favorable of assumptions, this entire discussion is a red herring.
Q. Red herring? Is that racist?
A. Or consider your photographic example, Walter White (presumably the ironic namesake of the anti-hero of "Breaking Bad," but that's a whole different story). Walter White was 27/32nd white, the direct descendant of two Presidents. Or consider former NAACP head Julian Bond in the SNL video from the 1970s above. Or consider the current NAACP president Benjamin Jealous, who looks like Mark Ruffalo. (There'a reason the dark-skinned masses called mixed-race elites of the NAACP the "National Association for the Advancement of Certain People.")
Presumably, these are guys with IQs > 100. How in the world would moving them from the black to the white category narrow the IQ gap?
Take your time and think about it. Get back to me when you've done the arithmetic.
——
To drop out of the FAQ format, I'm fascinated by the terrible statistical sense displayed by pundits in the race-IQ debate. Evolutionary psychology suggests that human beings ought to be pretty decent at statistics. And, indeed, we see that intelligent people are pretty good at figuring out which are low crime neighborhoods to buy a home in and which are low violence schools to send their kids to, even if they aren't particularly good at math.
Yet, when they try to talk about the statistics of race and IQ, they seem completely inept. Why?
It's a little like conspiracy theorizing, where the most popular conspiracy theories are almost always wrong, even though conspiracies really do play a massive role in history.
For example, there's a new major history of World War One out that traces the origin of the Great War to one man's conspiracy: Dragutin Dimitrijevic, called Apis, the head of Serbian military intelligence, who arranged the assassination of the Archduke Franz Ferdinand. But I was unaware of that name until Greg Cochran told me it when I was about 45. Nobody is interested in Dragutin Dimitrijevic even though he organized the most catastrophic conspiracy since the assassination of Julius Caesar. Conspiracy theorists would rather pull out their toenails than discuss how Dimitrijevic's conspiracy blew up the world.
Same with race and IQ: pundits seem to have an instinct, a magnetic attraction, for being wrong. Yet, when they try to talk about the statistics of race and IQ, they seem completely inept. Why?
It's a little like conspiracy theorizing, where the most popular conspiracy theories are almost always wrong, even though conspiracies really do play a massive role in history.
Same with race and IQ: pundits seem to have an instinct, a magnetic attraction, for being wrong.
