It has more to do with people getting degrees for the wrong reasons: being educated doesn't mean you will always have economic security.
And that's where the housing bubble analogy comes in: there are many good reasons to buy a house, but doing so because you expect its value to increase constantly is not one of them.
The Education and Housing markets are complex systems. The bubble analogy works as a vary simplified analogy to describe how some markets can see sustained prices increases for a while followed by a sudden drop.
This works with Housing, at a fundamental level if people decided to more houses one year the price increases, and if people buy fewer houses one year the value of your house decreases which can quickly cycle though to dramatic shifts. However, if fewer people decided to get a degree the value of your degree increases.
Both systems have a constant: massive government subsidies that distort the true supply, demand, and economic ROI for the good. There could not have been a housing bubble without Fannie and Freddie, the FHA, and the protected-by-regulation class of bond ratings agencies.
Similarly, there would not be an education bubble without federal loan guarantees -- certainly they have escalated cost of college, and most of the shady for-profit colleges thrive on student subsidies.
What is really interesting is how little the price could increase before market forces brought things back to reality. Housing is such a large percentage of the worlds wealth that we never saw the sort crazy multiples over value that other bubbles get to. EX: http://en.wikipedia.org/wiki/Japanese_asset_price_bubble
Bubbles require rapid credit expansion. There may be varied reasons for what fuels a particular bubble, but the necessary precondition is easy money/credit.
When the government regulates the currency & credit markets tightly, then government should share in the responsibility for the bubble.
I'd recommend looking at a chart of GDP from before the Fed was established for the express purpose of mitigating boom/bust cycles.
It looks like a jigsaw. If you look at a chart of actual numbers, it's pretty clear that the countercyclical tools available to the fed diminish the effect of bubbles.
The largest and least stable bubbles are often driven by credit but there is a wide range of causes. For example, one of the largest and less talked about bubbles comes from the shift from defined benefit plains to 401k style investing. In the mid-1980s there were fewer than 8 million participants with less than $100 billion of assets in 401(k) plans.[3] By 2006 there were seventy million participants with more than $3 trillion of assets in 401(k) plans. Now, what happens to the US stock market as baby boomers retire and there is a significant shift between people buying and selling stocks?
PS: Many bubbles are simply money looking for somewhere to hide. Assume the US cut it's military budget by 80% and paid of the debt in 20 years, where do you think that money would end up?
"There could not have been a housing bubble without Fannie and Freddie, the FHA, and the protected-by-regulation class of bond ratings agencies."
I beg to differ. You could have pulled the full cost of every defaulted mortgage out of the bottom line of a single big-5 investment bank and not even bankrupted it. But we had all these crazy hyper-leveraged instruments that turned a 50 billion dollar problem into a 5 trillion dollar problem. The issue wasn't the mortgages, it was the leverage and gambling.
Hasn't this always been true? Universities have always offered plenty of degrees in fields that are hard if not impossible to find good paying jobs in.
And that's where the housing bubble analogy comes in: there are many good reasons to buy a house, but doing so because you expect its value to increase constantly is not one of them.