Is there a gene for educational achievement?


But recently, a consortium of nearly 100 authors has discovered over 1200 genetic variants that apparently are associated with “educational attainment,” measured as number of years in school. I have read the article, and I confess that much of it may as well have been written in Greek. I have, however, managed to glean a few facts from the article, as well as a Times article by Carl Zimmer and an unsigned article submitted by the University of Colorado.

In particular, the consortium genotyped approximately 1.1 million people and isolated 1271 variants that correlated significantly with number of years of education. The subjects were exclusively white people of European descent; this restriction was necessary to obtain a relatively homogeneous sample. The genes that were isolated are mostly related in some way to the development of the brain or to communication between neurons (see the Abstract and Figure 3 of the Nature Genetics paper).

Using an aggregation of many tiny genetic weights, the authors calculated a polygenic score for individuals. A person with a high polygenic score probably achieves a higher educational attainment than a person with a low polygenic score. A subset of the authors are at pains to point out (FAQ 3.4) that the polygenic score predicts only about 11 % of the variation in educational attainment, and it should not be used to predict the educational attainment of an individual. After all, they say, in the white European population, 89 % of the variation in educational attainment is the result of other factors. Unless I misunderstand, though, some of that variation is very probably due to genes yet undiscovered, in which case the 11 % figure will probably rise (Robert Plomin, a distinguished psychologist and geneticist now at King’s College, London, estimates that half of the variation of intelligence is heritable, but nowhere near all the relevant genes have been discovered. Intelligence is not the same as educational attainment, but, analogously, nowhere near all the relevant genes have been discovered. The present study, incidentally, also predicted 7-10 % of the variance in cognitive ability).

Research into intelligence and particularly into the heritability of intelligence is so controversial, particularly among my fellow leftists, that the day after Mr. Zimmer’s article appeared, the Times ran an article with the forthright title Why progressives should embrace the genetics of education, by Kathryn Paige Harden, an associate professor of psychology at the University of Texas at Austin. Professor Harden’s article raises the specter of eugenics and notes that such “thinking is not safely in the past,” and she is certainly right that we have to be on our toes. Nevertheless, she claims that we should ask, “How can the power of the genomic revolution be harnessed to create a more equal society?” and provides 2 responses:

(1) Genetic studies such as the Nature Genetics paper will demonstrate that success is largely a matter of luck, and they will show “that everyone should share in our national prosperity, regardless of which genetic variants he or she happens to inherit.” As My Wife and Harshest Critic might say, “Should is a bad word.” A modern eugenicist, I am afraid, will draw precisely the opposite conclusion. The modern eugenicist will also ignore the flat statement in FAQ 3.5, “that polygenic scores of individuals from different ancestry groups cannot be meaningfully compared [italics in original].”

Studies of the heritability of educational attainment, intelligence, or “success” can be extremely dangerous.

(2) If we know from the polygenic score how much of a person’s educational attainment is due to genetics, then we will know how much is due to other factors, and we can concentrate on them. I think that is a much sounder argument than (1).

I am afraid that Professor Harden may be whistling past the graveyard when she continues,

Talking about including genetics as a variable in statistical models doesn’t have the same dark allure as eugenic proposals to screen embryos or assign children to schools based on their genotypes,

and I hope she is right when she adds,

But the widespread use of polygenic scoring in research aiming to understand how environments shape children’s lives will yield big payoffs for knowing how to maximize a child’s potential. We can’t change someone’s genes, but we can try to change how she grows up.

At any rate, the cat is out of the bag and the genie is out of the bottle. We can no longer argue about whether to do such research, but we certainly need to discuss what to do about it and avoid the temptation to use it to ill effect.

Acknowledgments. Jonathan Kane first alerted me to the Nature Genetics article. Emily Willoughby, one of the co-authors of the article and a sometime PT contributor, commented astutely on this post.