The title is a warning to me. The claim that "this time it's different" is supposed to be typical of bubbles, but I write it meaning it. I think that this huge increase in the relative price of housing is different. In the past such an increase has been a reliable sign of a bubble. Such bubbles burst and the bursting is followed by a recession.

This means that a high relative price of housing has been an excellent predictor of low growth including a recession over the following 5 years.

The graph shows real log GDP growth over the following 5 years and the predictions based on a super simple model with a trend and the relative price of housing. The only anomaly is that the great recession was less severe than the super simple model predicts. Interestingly super crude Keynesian calculaions siuggest that the Obama stimulus should have reduced the severity of the great recession by about 50% and the great recession was about 50% as severe as predicted,

If that pattern holds we are in for very bad trouble.

My super simple model predicts a recession markedly more severe than the 2008-9 recession some time in the next five years.

So do I believe in my model ? Not at all. Maybe I jiust don't want to, but I really think that this time it's different. I think the high relative price of housing is caused by a genuine shortage of houses and not be a speculative bubble.

Partly this belief is based on qualitative signals. In the period 200-2006 there were active house flippers who bought and soon sold a house - they were clearly speculators. There was enthusiastic discussion of houses as investments. The shamefully small amount of data on expected returns to investment in houses (collected by professor Housing Robert Shiller) showed huge expected returns. People were buying houses with extremely high leverage (sub prime mortgages). Banks were lending on the assumption that house prices could only go up (I mean this literally they were using models to price residential mortgage backed securities in which one could insert a number for the national average increase in house prices - note increase - the computer accepted only positive numbers).

These signs of mania are not present now.

But more importantly the rent is too damn high for one to conclude that the high price of houses is due to a speculative bubble. One strong sign that there was a bubble during 2000-2006 was that the ration of the house price index to the rent index went up. The point of buying a house are the housing services one gets from living in it plus the possible capital gains from selling it later. The normal price of housing services (based on rent payed by renters) combined with the rapid large increase in prices showed that people were expecting huge capital gains.

The increase is less than the relative price of houses, but that is not really surprising. The numerator is the weird extremely sticky series owner's equivalent rent - which is rent that is not actually paid but rather an estimate of what owners would pay if they rented. Rents are extermely sticky being usually set by a contract with fixed dollars per monthy which lasts at least a year. Krugman fans (like me) will have read a lot about how the persistence of CPI inflation is mainly based on owner equivalent rent in which a large increase in rent on newly signed leases worked its way through the stock of existing leases (he explained it well just as I explained it very poorly). This does mean that I predict that rent will continue to go up just as I don't predict a huge recession (you can't have everything).

I think the solution to the current problem is not to prepare for the bursting of a bubble (not that anyone knows how to do that anyway) but rather to build more housing by reducing restrictions on construction (that is I am 100% YIMBY). A very conventional conclusion, but again you can't have everything.