The better students, moreover, are at the better schools.  Most important
scholars I have know received their training at major graduate schools, no matter how
ordinary their undergraduate education.  On the other hand, many extremely promising
undergraduates must go to the lesser graduate centers, because of personal
circumstances, ignorance of their quality, better financial aid, or whatever.  Why have
so few of these latter students succeeded in research?  After all, the material that is
taught in the major graduate centers is in print and available everywhere.  My
explanation is that in the leading graduate centers the students learn primarily from
one another.  They learn to impose higher standards upon themselves, both in the
selection of problems to work on and in the adequacy of the solutions they provide to
these problems.  Bull sessions are a more effective method of teaching and learning
than classroom lectures or discussion.  One colleague has said that he considered his
role in the classroom to be that of providing topics for bull sessions.  I don't think
that the successes achieved by the graduates of the major schools are due simply to
an old-boy network.  Anyone who does outstanding work is in strong demand even if,
like the hypothetical mathematician I referred to, he has few other redeeming traits.


				-- George Stigler, Memoirs of an Unregulated Economist