Starting from the hypothesis, confirmed by numerous past studies, that the rate of return
differs according to the sex, I divided the available data on individuals in two groups and I
estimated the earning function [12] for male and female both separately and jointly.
In a third step I added at the analysis some demographic variables, as, for example, the
living place, and the type of employee, to extrapolate possible effects on the rate of return
of education and on the salaries.