Let me respond briefly to some interesting comments. The main point of my comment on Posner was that I would leave the decisions about admission of men and women to universities and professional schools. If they want to take many women even though they are more likely to drop out, I do not believe we know enough to challenge that decision. I suggested that schools may be taking into account the advantages of co-education, the contribution of female alumni, and other considerations. So the critic who said I only looked at earnings did not get my main point-I do not feel enough is known about the total contributions of men and women, monetary and non-monetary, to be able to tell professional schools what criteria they should use.
I only suggested the use of "profiling" through using sex as a relevant variable. Perhaps one can analyze within the group of say female applicants and try to figure out which ones would drop out. But that is I agree too difficult, and in my approach not even necessary.
On the mainly different issue raised by one commentator, I do believe that colleges and schools should try to "max out"in their admission policies. But by that I do not mean they should try to maximize tuition revenue since they are clearly interested in the quality of students, their subsequent achievements, etc. However, they should engage in even greater price discrimination than they do, and not have a "needs blind" admissions policy since tuition revenue is one of the relevant variables, especially for private universities. After all, most of them, alas including the University of Chicago, do not have an endowment that is anywhere near Harvard's.