That question may not make sense in the US or the UK, or some other European countries that face very serious public budget constraints. It also does not make much sense given the peak in student attendance in many OECD countries. But where student numbers will keep increasing, it is a good question whether one should increase the size of universities or their numbers.
Berardino Cesi and Dimitri Paolini consider the question from the angle of the students' mobility constraints. Suppose students differ by ability and by location, and mobility is costly. A monopolistic university will only attract the ablest. Adding a new, local university is then welfare improving. Suppose now the mobility costs are rather low. Adding a second university is now welfare decreasing because of a peer effect. Indeed, high ability students now get pooled with low ability ones, and they suffer through adverse peer effects. And given that some students simply do not belong in a university, we are better off with fewer universities and fewer college students.