Stop Talking About Race and IQ

Take it from someone who did.

Photo illustration by Slate. Photos by Thinkstock.

The race-and-IQ debate is back. The latest round started a few weeks ago when Harvard geneticist David Reich wrote a New York Times op-ed in defense of race as a biological fact. The piece resurfaced Sam Harris’ year-old Waking Up podcast interview with Charles Murray, co-author of The Bell Curve, and launched a Twitter debate between Harris and Vox’s Ezra Klein. Klein then responded to Harris and Reich in Vox, Harris fired back, and Andrew Sullivan went after Klein. Two weeks ago, Klein and Harris released a two-hour podcast in which they fruitlessly continued their dispute.

I’ve watched this debate for more than a decade. It’s the same wreck, over and over. A person with a taste for puncturing taboos learns about racial gaps in IQ scores and the idea that they might be genetic. He writes or speaks about it, credulously or unreflectively. Every part of his argument is attacked: the validity of IQ, the claim that it’s substantially heritable, and the idea that races can be biologically distinguished. The offender is denounced as racist when he thinks he’s just defending science against political correctness.

I know what it’s like to be this person because, 11 years ago, I was that person. I saw a comment from Nobel laureate James Watson about the black-white IQ gap, read some journal articles about it, and bought in. That was a mistake. Having made that mistake, I’m in no position to throw stones at Sullivan, Harris, or anyone else. But I am in a position to speak to these people as someone who understands where they’re coming from. I believe I can change their thinking, because I’ve changed mine, and I’m here to make that case to them. And I hope those of you who find this whole subject vile will bear with me as I do.


Here’s my advice: You can talk about the genetics of race. You can talk about the genetics of intelligence. But stop implying they’re the same thing. Connecting intelligence to race adds nothing useful. It overextends the science you’re defending, and it engulfs the whole debate in moral flames.

I’m not asking anyone to deny science. What I’m asking for is clarity. The genetics of race and the genetics of intelligence are two different fields of research. In his piece in the Times, Reich wrote about prostate cancer risk, a context in which there’s clear evidence of a genetic pattern related to ancestry. (Black men with African ancestry in a specific DNA region have a higher prostate cancer risk than do black men with European ancestry in that region.) Reich steered around intelligence where, despite racial and ethnic gaps in test scores, no such pattern has been established.

It’s also fine to discuss the genetics of IQ—there’s a serious line of scientific inquiry around that subject—and whether intelligence, in any population, is an inherited social advantage. We tend to worry that talk of heritability will lead to eugenics. But it’s also worth noting that, to the extent that IQ, like wealth, is inherited and concentrated through assortative mating, it can stratify society and undermine cohesion. That’s what much of The Bell Curve was about.


The trouble starts when people who write or talk about the heritability of intelligence extend this idea to comparisons between racial and ethnic groups. Some people do this maliciously; others don’t. You can call the latter group naïve, credulous, or obtuse to prejudice. But they might be open to persuasion, and that’s my aim here. For them, the chain of thought might go something like this: Intelligence is partly genetic, and race is partly genetic. So maybe racial differences on intelligence tests can be explained, in part, by genetics.

There are two scientific problems with making this kind of inference. The first is that bringing race into the genetic conversation obscures the causal analysis. Genes might play no role in racial gaps on IQ tests. But suppose they did: To that extent, what would be the point of talking about race? Some white kids, some black kids, and some Asian kids would have certain genes that marginally favor intelligence. Others wouldn’t. It’s still the genes, not race, that would matter.

This is a rare point of consensus in the IQ debate. In his interview with Harris, Murray notes that in The Bell Curve, race was a crude proxy for genetics. Since the book’s publication in 1994, our ability to assess genetic differences has come a long way. Today, scientists are evaluating thousands of genes that correlate with small increments in IQ. “The blurriness of race is noise in the signal,” Murray tells Harris. “It’s going to obscure … genetic differences in IQ.”


“Race science,” the old idea that race is a biologically causal trait, may live on as an ideology of hate. But as an academic matter, it’s been discredited. We now know that genes flow between populations as they do between families, blurring racial categories and reshuffling human diversity. Genetic patterns can be found within groups, as in the case of prostate cancer. But even then, as Ian Holmes notes in the Atlantic, the patterns correlate with ancestry or population, not race.

The second problem with extending genetic theories of IQ to race is that it confounds the science of heritability. Sullivan and Harris cite research that indicates IQ is, loosely speaking, 40 percent to 80 percent heritable. It can seem natural to extend these estimates to comparisons between racial groups. That’s what I did a decade ago. But it’s a mistake because these studies are done within, not between, populations. They measure, for example, the degree to which being someone’s twin or biological sibling, rather than simply growing up in the same household, correlates with similarity of IQ. They don’t account for many other differences that come into play when comparing whole populations. So if you bring race into the calculation, you’re stretching those studies beyond their explanatory power. And you’re introducing complicating factors: not just education, income, and family structure, but neighborhood, net worth—and discrimination, which is the variable most likely to correlate directly with race.


Murray and others have answers to these objections. They argue that education programs have failed to close racial gaps, that studies haven’t proved that getting adopted has much lasting effect on kids’ IQ scores, and that collective increases in IQ scores are based on factors other than “general” intelligence. These are complex disputes full of nuances about replicating studies, interpreting test questions, and extrapolating from trend lines. But notice how far we’ve drifted from biology. The science here is oblique, abstract, and tenuous. Are you still comfortable speculating about genetics? Are you confident, for instance, that studies that compare black children to white children properly account for family assets and neighborhood, which differ sharply by race even within the same income bracket?

It’s one thing to theorize about race and genes to assist in disease prevention, diagnosis, or treatment, as Reich has done. But before you seize on his essay to explain racial gaps in employment, ask yourself: Given the dubiousness of linking racial genetics to IQ, what would my words accomplish? Would they contribute to prejudice? Would they be used to blame communities for their own poverty? Would I be provoking thought, or would I be offering whites an excuse not to think about the social and economic causes of inequality?


Murray, Sullivan, and Harris try to soften their speculations by stipulating, as I once did, that even if racial differences in IQ are genetic, you shouldn’t make assumptions about any individual. They’re correct that it’s both wrong and irrational to make such inferences from aggregate data. But it’s also easier to treat people as individuals when you don’t start with racial generalizations.

If you’re libertarian or conservative, you might think I’m calling for censorship. I’m not. I’m just asking for precision. Genes are the mechanism under discussion. So talking about the genetics of race and the genetics of IQ is more scientific, not less, than pulling race and IQ together.

Many progressives, on the other hand, regard the whole topic of IQ and genetics as sinister.
That, too, is a mistake. There’s a lot of hard science here. It can’t be wished away, and it can be put to good use. The challenge is to excavate that science from the muck of speculation about racial hierarchies.

What’s the path forward? It starts with letting go of race talk. No more podcasts hyping gratuitous racial comparisons as “forbidden knowledge.” No more essays speaking of grim ethnic truths for which, supposedly, we must prepare. Don’t imagine that if you posit an association between race and some trait, you can add enough caveats to erase the impression that people can be judged by their color. The association, not the caveats, is what people will remember.


If you’re interested in race and IQ, you might bristle at these admonitions. Perhaps you think you’re just telling the truth about test scores, IQ heritability, and the biological reality of race. It’s not your fault, you might argue, that you’re smeared and misunderstood. Harris says all of these things in his debate with Klein. And I cringe as I hear them, because I know these lines. I’ve played this role. Harris warns Klein that even if we “make certain facts taboo” and refuse “to ever look at population differences, we will be continually ambushed by these data.” He concludes: “Scientific data can’t be racist.”

No, data aren’t racist. But using racial data to make genetic arguments isn’t scientific. The world isn’t better off if you run ahead of science, waving the flag of innate group differences. And if everyone is misunderstanding your attempts to simultaneously link and distinguish race and IQ, perhaps you should take the hint. The problem isn’t that people are too dumb to understand you. It’s that you’re not understanding the social consequences of your words. When you drag race into the IQ conversation, you bring heat, not light. Your arguments for scientific candor will be more sound and more persuasive in a race-neutral discussion.

The biology of intelligence is full of important questions. To what extent is it one faculty or many? How do we get it, grow it, maintain it, and use it? If it’s heritable, should we think of it less as merit and more as luck, like inheriting money? To what extent does a class structure based on intelligence duplicate or conceal a class structure based on family wealth? Is intelligence truly supplanting other kinds of inheritance as a competitive advantage? Is it unleashing social mobility? Or is it, through assortative mating, entrenching inequality? These are much better conversations than the one we’ve been stuck in. Let’s get on with them.

Update Consent

Already a subscriber? Sign in here.

Already a member? Sign in here.

Subscribe Now