Economists have been reluctant to interpret as purely causal the relationship between educational attainment and earnings. In an influential paper in which they use quarter of birth as an instrument for educational attainment in wage equations, Angrist and Krueger interpret their estimates as the causal impact of education on earnings. To support this interpretation, they argue that compulsory school attendance laws alone account for the association between quarter of birth and earnings. In this work we present new evidence suggesting that this interpretation may not be well-founded. We document an association between quarter of birth and earnings in cohorts that were not bound by compulsory school attendance laws. Moreover, we find that the association between quarter of birth and educational attainment was weaker in more recently-born cohorts while no similar pattern existed in the association between quarter of birth and earnings. Our results call into question the validity of any causal inferences based on Angrist and Krueger's estimates regarding the effect of education on earnings.
