One simple way to examine wage differentials (or discrimination) by gender is to estimate y=β_0+β_1 x_1+β_2 x_2+β_3 M+ϵ, where M is a dummy equal to 1 if male and zero otherwise. Assuming that the model is correctly specified (satisfies CLM assumptions), under what conditions would you conclude that there is gender discrimination in terms of wage earnings against female?

A major difficulty with the simple human capital model given is related to finding a correct functional form. In this regard, the inclusion of the squares and interaction terms of x1 and x2 could be considered for improving the adequacy of the model. How would you test if in deed inclusion of such terms would improve the adequacy of the model. Outline the procedure briefly.

0