Abstract
We estimate the intergenerational correlation (R) in the earnings to be 0.2 or less and the regression towards the mean of sons and daughters with their fathers to be substantial. Using education we test Becker's intergenerational endownments model. Contrary to predictions, we do not find grandparents' education enters negatively once parents' education is controlled nor do we find the bias in the OLS equations predicted by his theory when we use instrumental variables and two-stage least squares.

This publication has 0 references indexed in Scilit: