[jimt67]

While quality of education varies with the professor, I would argue that it also varies with the student...possibly even more so.

Let's assume Instructor A teaches Introduction to Nordic Studies at a non-selective university. Two of the factors that control the design of that course (there are many, I'm just isolating two) would be: -the instructor -the students

In ideal conditions, the course iteratively develops into some dynamic equilibrium where aggregate student learning is optimal (yes, I realize this is fantasy). Replace those non-selective students with students from a highly-selective university, while keeping the same instructor. Imagine how different the course might develop.

From a learning standpoint, if you are a student that is talented enough to be admitted to a highly-selective university, the latter course would likely result in you learning more. In reality, there are a lot of other factors that impact the growth and development of any individual student, but all things being equal, if you are a smarty, you will probably learn more in a class with other smarties.