Is your PhD better than mine?

I have often said, when hiring a new faculty member, that I prefer the best student from a small program to the worst student from the best program. But many universities hire the student from Harvard, or MIT, because of something called statistical discrimination. On average, any one graduate from a top program will have a higher probability of succeeding in an academic career than any one graduate from a lesser program. Hence choosing the Harvard PhD is a better “bet” than choosing one from Washington State University, for example.

The research published in the Journal of Economic Perspectives discussed in this article from The Economist supports my idea, albeit based on a very restrictive measure of productivity (the number of published quality-adjusted research papers). While the best students from the best programs clearly publish a lot of high quality research, the publication rate for the lowest 50% of graduates from all schools is the same. And the publication rate of the best students from lesser ranked programs is higher than the publication rate of the worst students from the higher ranked programs.

So there…