New article in Intelligence explaining why beautiful folks tend to be bright folks.

The conclusion that beautiful people are more intelligent follows from four assumptions. (1) Men who are more intelligent are more likely to attain higher status than men who are less intelligent. (2) Higher-status men are more likely to mate with more beautiful women than lower-status men. (3) Intelligence is heritable. (4) Beauty is heritable. If all four assumptions are empirically true, then the conclusion that beautiful people are more intelligent is logically true, making it a proven theorem.

I am not near as knowledgeable about EP as other folks here, so I will let y'all comment on the minutia, but it appears to me that they are missing a personality component. If you are very smart (IQ > 175), but are also extremely neurotic, psychotic¹, and introverted, I doubt you will be as successful with the ladies as a moderately intelligent person, but whose personality is less neurotic, psychotic and more extroverted. Nonetheless, I am excited about my prospects.

1. This is from Eysenck's work, and does not, necessarily, mean clinical psychosis.

Godless comments:

Actually, I don't think that the conclusion is accurate. Even if all four of the premises are true, it need not be the case that the conclusion follows. Technically speaking, suppose that X, Y, and Z are zero-mean normal variables. E[XY] and E[XZ] can both be nonzero while E[YZ] is zero. You can demonstrate this by construction - simply use a degenerate covariance matrix with a zero covariance for Y and Z. This can of course also be done for non-normal RV's.

In other words, the positive correlations between X (=income) and Y (=intelligence) and that between X (=income) and Z (=beauty) do not imply a nonzero correlation between Y (=intelligence) and Z (=beauty). It is suggestive, but you'd have to do a direct study before you chained the correlations in this fashion.

Intuitively, the fallacy in reasoning was identified by Alex. Using pairwise correlations alone can obscure internal structure in multivariate data sets. For example, the highly intelligent tend to be economically successful. However, it may be the borderline high-IQ (Morgan Stanley banker) rather than the ultra intelligent (Harvard researcher) who tend to get the most attractive women.