Everyone's reviewing Dean Baker's False Profits. I contribute.
I strongly recommend the book, agreeing with D^2 that it's important that he names the guilty men, reminds you that they're guilty, and keeps on naming them. (In fact, if anything, he drowns the fish - if you weren't sure if there had been a housing bubble, you'll be sick of hearing about it.)
But I think there's an important point here; so long as there is capitalism, entrepreneurs will every so often become over-extended, chase the price of something up a mountain, and then panic and crash. In fact, perhaps even outside capitalism; aren't the investment disasters of the Soviet Union something similar? Usually, however, it's possible to look at this process philosophically; as J. K. Galbraith said, nothing but money is being lost, and the people who are losing it usually have plenty. Residential property, though, is special - it alone brings the full destructive power of a leverage-fuelled financial crisis within reach of the common man.
As the Despair poster has it, none of us is as stupid as all of us together. A housing bubble's effects are therefore very widespread, and individually catastrophic to a degree that isn't true of financial crises restricted to bankers, investors, and entrepreneurs. They also have a special feature, which is that they interact with the tools of macro policy closely. Once the bubble passes a certain point, a significant chunk of the buyers are no longer able to cope with the normal range of fluctuation in the interest rate, and disaster is certain. This is essentially Minsky stage 2 - when capital gains become a necessity in order to service debt.
Once this stage is reached, for many of the people involved, it doesn't matter whether the final crash comes sooner or later - they're going to be ruined. For society as a whole, it would probably be better if it came sooner, but there are obvious reasons why this is unlikely to happen.
Keynes once said, in Economic Perspectives for our Grandchildren, that he hoped economists would become "humble competent people like dentists". Daniel Davies said not so long ago that economics might, if it was lucky, eventually become a branch of control systems engineering. An important concept in both is stability, but stability is often misunderstood. It doesn't mean the absence of change, but rather, the ability to recover to a normal state. James Lovelock's insight about ecosystems was arguably that they are stable - they eventually recover - but that there was no guarantee that recovery would be a good thing.
In a housing bubble scenario, eventually the berserk run-up in prices does indeed go into reverse - but it does so through widespread bankruptcy, unemployment, and systemic bank failure. This is the same whether the bubble bursts on its own, or the bubble is deliberately burst. The recovery mode of the system is indistinguishable from the worst-case scenario. As the purpose of a system is what it does, the fact we get into these scrapes is telling in itself.
Here, Felix Salmon excerpts a range of fascinating charts from the US National Housing Survey. A critical detail; even people who are currently behind on their mortgages or even in the process of repossession believe, by large majorities, that property is a good investment and that it is less risky than Treasury bills. Clearly, it will need more than just the biggest bubble in economic history, leading to the worst economic crisis since the Great Depression, to convince anyone that there is a problem. (There's also an excellent chart making the point that the market is still dear historically by about 20%.)
An interesting counter-example is here; it seems that specific regulation against some of the worst bubble-generating practices can help a lot. And I think that changing the failure modes is probably a good way of changing the system. Dean Baker answers reviewers here; he also points out that although Spain's banks didn't blow up, the housing bubble there was sufficient in itself to flatten the economy.