- Hypothesis F ("F" for future): During the boom the housing market expected (rationally or otherwise) that housing demand would be high in the future. That is, unless housing construction were far above pre-boom rates, there would not be enough housing to satisfy demand in 2007, 2008, 2009 and beyond. Based on the expectation, housing construction proceeded at a furious pace. Once we learned that the future demand wasn't there, we were worse off because those houses couldn't be unbuilt. This ultimately false expectation cost America -- according to my forthcoming (hopefully by Monday at papers.nber.org) paper "Market Responses to the Panic of 2008" the losses were in the trillions. These losses will cause people to work more and, through that channel, will raise GDP.
- Hypothesis P ("P" for present): Suddenly, people no longer found housing to be an important priority. The scenario has some of the losses as with Hypothesis F (and thus the work stimulus), except that GDP may fall to the extent that actual housing/apartment rental rates are lower (recall that part of GDP is the rental value of housing) and this dominates the work stimulus. In the scenario above, actual rental rates do not fall, it's just that people revised their expectations about future rental rates.
- Hypothesis G ("G" for government distortion): The boom was caused by a distortion (induced by the government, or otherwise) of capital away from the nonresidential sector. It is good news that the distortion finally ended; the housing price crash is just an indicator that things are getting better. Our economy can use its resources more productively. This may ultimately raise GDP, but through different mechanisms: productivity will be higher, work will be induced by a substitution effect (rather than a wealth effect), and we may even be able to obtain the same government revenue with lower tax rates (recall that the housing sector is a major area for tax avoidance).
The housing rent data tell me that Hypothesis P is not important. Notice from the graph that the inflation-adjusted rents paid by persons living in houses and apartments are essentially as high now as they were during the boom, which themselves were not that high.
I don't yet know which of the remaining hypotheses is more important. I do not get the sense that America is saying "Finally! We can do things more efficiently!" so I have started pursuing the logical implications of Hypothesis F. But I have to start somewhere, and will report back when I have given more thought to G and/or have evidence on its relative importance.
[Update: I haven't made up my mind yet, but in the meantime I recommend Professor João Marcus Marinho Nunes' very nice article on Hypothesis G]