If your contribution as an economist is very fundamental, other people will use that contribution whether or not they were your students. Lots of people use, or critique, the assumption of rational expectations. So the inventors of RE don’t need students to propagate their fame. Alfred Marshall’s fame today is mostly independent of the students he had (or did not have).
Having fundamental contributions is correlated with quality but within the top tier of quality there is considerable variation. Arrow and Lucas had relatively fundamental contributions but in contrast Milton Friedman, Larry Summers, Robert Barro, and Olivier Blanchard are all more applied. Their demand for students should be higher. If you are an empirical economist, but invented an econometric technique which is fundamental, your demand for students should be relatively low. Students might also prefer advisors who are less "fundamental," for fear of being overshadowed or from wanting to avoid the winner-take-all tier of the market.
Since at top schools the percentage of "fundamental" economists is declining over time, we would expect the distribution of doctoral students, across faculty, to become more even.
I thank Amanda Agan and some of her friends for a useful conversation on this topic.