“IQ, Academic Performance, Environment and Earnings”, 2001-05 (; backlinks; similar):
This paper explores the effects of peers, friends, family, IQ and academic performance, observed in the last year of high school, on earnings at ages 35 and 53.
All statistically-significantly affect earnings at both ages. The effects of IQ are much smaller than asserted in, for example, The Bell Curve, and badly overstated in the absence of controls for family, wider context or academic performance.
Aspirations appear to be very important. Socialization and role models may be as well, but not ability spillovers. Feasible increases in academic performance and education can compensate for the effects of many cognitive and contextual deficits.
[This paper exemplifies the fallacy of controlling for intermediate variables—as if all those were measured perfectly and the “controls” had nothing to do with IQ causing earnings!]