An important aspect of these findings that might be overlooked is the relatively low level of income at which the association of cognitive ability and earnings

An important aspect of these findings that might be overlooked is the relatively low level of income at which the association of cognitive ability and earnings plateau. In Figs 3 and 4 the plateau starts at about SEK 600,000, or US$ 57,000. This is below what most people would consider "top earners". So in essence, and as I remember from previous reports, there isn't a very strong correlation of intelligence and earnings at the _very_ top levels, say the top 0.5% of earners. I also find the tight, almost monotonic association nevertheless shown to be a bit suspect. Very few traits have such a tight association. Even genetic variation and a trait such as height doesn't have such a strong association.