I just wanted to add a few comments about Murray’s treatment of stereotype threat. He cites Claude Steele’s research about the topic, and concedes that the results of said research have held up. Murray is right when he says that “Whatever causes stereotype threat in the larger society, it is not anything that is fixable in the day-to-day conduct of K-12 education.”

However, I think he could have better explained why this concept took hold so much in academia. It did so because people misrepresented Steele’s research — Murray notes this fact, but in going through the details focuses on differences between the lab and the real world, rather than on the outright media distortion the research suffered.

The common narrative about stereotype threat goes like this: Blacks face the stereotype that they aren’t smart enough to do well on tests. When you give them a test and say it measures their intellectual ability, they get nervous about confirming the stereotype, and as a result actually do do worse. When you give them the very same test and just say it’s for your research, they do just as well as whites do.

Seen this way, the elimination of stereotype threat promises not only to improve the test-score gap, but to solve it completely.

The direction of this narrative is true, and as Murray notes, has held up through other experiments — blacks do worse when you tell them you’re measuring their innate ability. But what Murray doesn’t note, and what many in the press ignored, is that Steele’s research controlled for SAT scores. When told the test doesn’t measure innate ability, blacks don’t do as well as whites, but instead do as well as their SAT scores predict — which is, not as well as whites. Blacks told the test measures innate ability perform even worse.

In other words, the SAT scores correspond to the “right” way of administering a test, even though the popular press took the study as a reason to claim the SAT was racist. Also, the research proved that if you try, you can make someone’s scores lower.