Leaving aside race and IQ (and that last post comes quite a bit closer than I am comfortable with to touching the subject with the proverbial ten-foot pole), IQ matters for social policy. We do need to know whether g, the general intelligence factor that IQ is supposed to measure exists, how much of it is simply genetic, and how much more of it consists of environmental factors that we can reasonably change.

Because, assuming that it exists, but that the biggest problem for low-income children is environment, I don't know what sort of policy interventions this reasonably implies. Contrary to the glowing paeans to early childhood interventions that I frequently hear when I talk about schools, as far as I know the gold standard of early childhood interventions was the Perry Preschool Project, and intensive preschool program for three and four year old children run in the 1960s. The results, while admirable, were extremely expensive (contra that website, I calculate, using the Rand results, that the cost per child was about $18,500 per year in today's dollars). This bought lower poverty rates, less teen pregnancy, and lower incarceration rates. But it helped establish the kids in the bottom tier of the working class, with median incomes in the range of $20K. It did not come close to bridging the gap between those kids and the world of the middle class.

Even earlier interventions might help somewhat, but the earlier you go, the more problematic such interventions become. The younger the kids are, the more individual attention they require, which is why preschool is more expensive than fifth grade. Even if you're willing to pay for it, where are you going to find these millions of highly qualified early childhood experts to become, in effect, the surrogate parents to these children?

Not that I'm against trying--early childhood intervention seems to me, like schooling in general, to be one of those goods that society has an obligation to provide children if their parents are incapable. But as I've written before, good early childhood programs have enormous scale problems; I'm not sure how we overcome them.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.