The article seems to be under the impression that elite colleges makes elite graduates. They should expand Berkeley, another MIT, whatever. That seems backwards to me. I think that highly selective schools admit elite students and those students, over generations, build the reputations of their schools. If you doubled enrollment you would not double your Nobel output, and over overtime you might reduce it as the students who will go on to earn one preferentially select more elite institutions to study at.
Nobel output is just the symbol from the article for consequential science output. I'm using it in the same way.
If I run a track program and take the 0.1 percentile people by speed at age eighteen, and then give them average training for four years, and you run a track program taking the top fifty percentile by speed and give them the best instruction, my team is going to win.
Because elite institutions are selective the are pupiled by high capability students. High capability students who get to network with other high capability students and who get elite signaling go on to do consequential things. Students who do consequential things enhance the reputations of the school they matriculated from. The cycle repeats.
Professors and the quality of instruction isn't much different between good and great universities. The difference comes from the selection and if you take that away you'll take away the difference.
