Originally Posted by Bostonian
Megan McArdle espouses my views at the Daily Beast:

We know from the Flynn effect that average performance on IQ tests is increasing over time. It seems to me that theories like that presented here would lead to increasing variance as well, which should also be noted when re-norming the tests, right?

I suppose that there are factors that would lead to a decreasing variance, such as nutritional and medical advancements bringing up the lower tail. It just doesn't seem possible that the shape of the curve remains unchanged if IQ is significantly hereditary, and mating practices change.