This isn't really true. The ivies are bound by the rules that govern private institutions that receive federal funding. For example, they are not free to disregard Title 9 or Obama's "Dear Colleague" letter without jeopardizing their federal funding. Very recently a bill was introduced with pretty bipartisan support to require all universities that take federal funding to cease favoring legacy admissions. The federal gov't can't force these universities to do this, but they can withhold all federal funding, and that is a very powerful coercive mechanism.

I agree with you strongly that elite schools are not necessary in software engineering, but they make a huge difference in fields like Law.

Well, once you stop trying to get people into exclusive clubs, you start asking crazy questions like "why don't we just educate everybody to a high level rather than place arbitrary cutoffs based on century old social clubs to maintain an artificially restricted elite?"

> Well, once you stop trying to get people into exclusive clubs, you start asking crazy questions like "why don't we just educate everybody to a high level rather than place arbitrary cutoffs based on century old social clubs to maintain an artificially restricted elite?"

Pro tip... if you want to be educated well through instruction, don't go to an elite school. The instruction in most of the classes really sucks (with some notable exceptions that are often available for free or cheap online).

Small liberal arts schools tend to do a much better job of educating through instruction.

I guess that as private institutions the ivies and most private universities worldwide receive significant concessions and benefits from governments.

So trying not to build a aristomafia could be intended as part of the deal with society.

Because those institutions produce a disproportionate share of the most powerful people in US society, so the kinds of people they do or don't let in tend to have a strong effect on the kind of society we live in.

I think you might be mixing up causation here...