Marginal Revolution and Matt Yglesias both take swipes at the reason behind huge cost increases in higher education, and what can be done about it. I agree with Matt that a big part of the problem is the so-called Baumol effect, in which teaching a 90 minute lecture has, at least historically, not really been subject to any large productivity gains at least since the chalkboard -- it still takes 90 minutes. It might turn out that there could be future productivity gains here -- if Paul Krugman decides to offer his lectures, via webinar, to students at universities all over the US, and then to test the students electronically (or via local TAs), there could be some cost reductions. However, I don't see this type of thing catching on in any big way anytime soon, although Brad DeLong has/is reportedly trying something similar at the UCs, in part because most college faculty have a vested interest in seeing that this does not happen. Yglesias' idea of credentialing sounds like a really good one to me. I remember when I was first considering applying to Econ Phd programs, I really needed to take lots of math classes as quickly as I could. I probably should have considered University of Phoenix, but instead I was just forced in to taking a bunch of math classes one semester (since there were few offered in the summer), but was severely limited in how many courses i could take by the fact that the finals for these courses were limited to a five-day finals period. With 3-4 days to prepare for each final, I might have been able to take 7 or 8 courses over a six month period, but instead what happened was that I had to take three finals in a 24 hour sequence, got really exhausted midway, and didn't do as well I might have had I had more time. Obviously, this is inefficient. However, once gain, it's not clear that changing this state of affairs would be a good idea for most universities. If students can take more classes per semester, they could then graduate sooner, pay less tuition, and feel less of a bond to the school.
One under-rated factor in the cost of college tuition, especially at the higher end of the market, is that students tend to be fairly insensitive to prices, and there is no perfect competition among schools. There is only one Harvard, and thus it has market power to charge more than Yale. At the higher end, this cost insensitivity must arise in part due to trends in inequality in America. The sons and daughters of hedge fund managers will just choose the "best" university they can get in to, and it will be an investment that will be worth it, largely due to the signalling value. Hence, it would now likely make sense for most prestigious public schools such as UC Berkeley, to simply raise their tuition in order to compete for better faculty, better athletic coaches, and better facilities in order to rival Stanford. And given the state of inequality and the American quest for status, top universities could probably raise their tuition substantially before having to worry about reduced enrollment. At some point, of course, students will go to cheaper options (especially in a recession), but I'm quite curious what the data say about what happens to enrollment when tuition rises. I suspect schools find that raising tuition almost always raises revenue, and on a 1 for 1 basis (schools like Berkeley are going to fill their enrollments...)... Until this ceases to be the case, there will likely continue to be rapid appreciation in the cost of higher-ed, and no real productivity gains. What's more, due to the signalling value, ever more Americans will spend huge chunks of money to collect even more MBAs and other degrees of dubious value.
It's a prisoner's dilemma scenario, and the only solutions are likely to be heavy-handed centralized intervention. On the positive side, education, when done well, really is priceless, and so an ever-increasing share of GDP spent on it is not exactly the world's greatest evil.
20 minutes ago