The former dean of Georgia Tech's College of Computing has a series of posts arguing that many (most?) universities actually lose money on research, contrary to the grants argument:
http://innovate-wwc.com/2010/07/05/why-universities-do-resea...
http://innovate-wwc.com/2011/05/18/if-you-have-to-ask-ten-su...
The need is very high (everybody wants to get a good job!), while supply is limited to established institutions.
Partly because it's so hard to prove to the world that your education program is excellent, it's very difficult to disrupt the market. Also, having money correlates with success later. Only well off people can pay for the best universities, but the best universities produce the best students partly because those students were well off in the first place.
"Alas, today’s full-time professional administrators tend to view management as an end in and of itself. Most have no faculty experience, and even those who have spent time in a classroom or laboratory often hope to make administration their life’s work and have no plan to return to teaching."
Um, any data _at all_ to back this up? In my experience at a public university, this is not at all true. And when administrators make a career out it, it is often in roles like working on technological infrastructure of a campus.
Also, many universities are growing their research programs. From what I've seen, it takes a lot more staff to support the research side of things than to support the educational side of things. And research is paid for (in theory) from grants and such, not tuition.
Anyway, there are a variety of problems with the way higher education is run and funded, but this article doesn't cover them. Instead, it uses a few statistics and a lot of bold, inaccurate, unsubstantiated, sensationalistic hand waving about that darn old wasteful ivory tower.
"There are lies, darn lies, and statistics" applies quite well in this case.