michael_nielsen

It's a good question. The effect is remarkably strong - it's like there's a systemic force holding things in place.

I'm sure part of the answer (probably only a small part) is just that in the tail differences tend to be large. So e.g., by the Shanghai metrics, Harvard has a long way to fall to get to number 2. But in the bottom half of the list I don't believe this effect wld be that strong

Rankings self-reinforce. Employers more likely to hire students from top-ranked universities, so best students go to top-ranked places, making employers even likelier to hire from there and so on. Same effect in terms faculty recruitment, funding, grad student recruitment, etc.

If that's the only axis for competition, sure. But this effect didn't save DEC from being beaten by Google, Myspace from Facebook, etc.

I don't know much about DEC, but I was an active user of Myspace/Friendster at the time Facebook began, and both were unusable (page loading times >40 secs). Perhaps the moral is that large self-reinforcing effects persist in the face of large screwups...

Blockbuster / Netflix. Toyota / name your favourite defunct car company. Etc x 1,000. The point is flux is common among industries where there's actual competition. The chart suggests research universities aren't really competing.

I'm a fan of Bryan Caplan's argument that education is primarily about signalling. By contrast, choosing between Blockbuster/Netflix is about watching a movie. So if school X does figure out a better way to teach than Harvard, it won't draw many students away from Harvard...

That was my original point: real competition seems to have largely been moved out of the system. (Education/Caplan is not relevant: most of the Shanghai ranking is based on research. Regard it as a ranking of research institutions.)