I have the deepest respect for Robert Shiller, who has been one of the most serious students of the dynamics of speculative bubbles there is, winning a well-deserved Nobel Prize for his work on this important topic. One of the more significant parts of his work has been on housing bubbles in particular, with his Case-Shiller indices being the most widely watched housing price measures in the US. Furthermore, in the second edition of his excellent Irrational Exuberance he laid out the case that housing prices from 19997-98 onwards into the early 2000s were almost certainly a massive bubble that would crash hard with dire consequences. He was right.
So it is with some trepidation that I question his arguments in a recent New York Times column and in other outlets to the effect that we are again experiencing a speculative bubble in US housing markets (and it should be kept in mind that these are local so varied and not all behaving similarly). The centerpiece of his argument is that the average nominal house price in the US has now reaxhed and seems to be surpassing the peak level in 2006 before the crash that brought on the Great Recession. Needless to say, this is serious stuff that desrves a close look.
Unfortunately these recent public statements by Shiller lack a crucial item that was crucial to the case he made in the second edition of his Irrational Exuberance that there was the biggest housing bubble in U/S history. That would be take account of the best measure of the fundamental, housing rents, or to translate that, to consider the path of the price/rent ratio. That he convinceed many people, including me, that housing prices in the early 2000s were a major bubble was that he showed that the average price/rent ratio had been rising in an unprecedented way, with it having been relatively stable theough most of the 20th century, including during the 1940s period that saw a major increase in housing prices. But that increase paralleled a rent increase thar indicated that indeed there was a major shortage of housing. The fundamental rent indicator triggered after a few years a supply repsonse that stabilized both rents and housing prices.
In his NY Times column, Shiller claims that there have in the last century been now three great increases in housing prices, somehow without ever mentionging rents or incomes, a sometimes alternative to rents fairly strongly correlated with them. One was the 1940s runup, which seems not to have been a bubble because there was a genuine shortage of housing due to low construction during the Great Depression and WW II, with the returning vets from the war wanting housing. So both rents ans prices soared for several years until a new equilibrium was achieved, although Shiller somehow did not fill out those details in this recent column. Then there was the screaming bubble of 1997-2006, which he accurately called. Then he threw in the most recent price increase, emphasizing that it now matches and is beginning to exceed the 2006 peak, but somehow not having a word to say about rents. Frankly, I am mystified at this important lacuna.
As it turns out, this recent price runup is sort of midway once one take into account what has happened to rents. For some reason Cas-Shiller does not track rents, but the Bureau of Labor Statistics (BLS) does. So if one puts these together using Jan. 1998 as the 100 base, then the national peak of the ratio occurred in late 2006 at about 1.87. It fell to about 1.2 at the price trough in 2012, and it is now at about 1.37. So clearly it has risen over the last six years, but that increase is about .25% while the curren level is still about twice that lower than the peak in 2006. Maybe this is a bubble, and certainly some local markets look like it. But this is still a long way from what we saw in 2006.
So what is up? One thing that is up is that there clearly has been a lack of housing construction, given the steady increase in prices over rents since 2012 to a pretty high level. It is not clear why this has happened. It certainly cannot be blamed on high interest rates, even though those have been rising in the last few years. They still remain below historical averages in real terms. The obvious other culprit, emphasized by Ed Glaeser and others with good reason has been the mulifsrious restrictions on construction that are notoriously especially present in some of the hottest housing markets in the nation, such as LA and the Bay Area of California, as well as both New York and Boston. These certainly are worthy of condemnation, but it is not obvious that there has been any substantial increase in their severity in the last decade. This may be the case, but I have not yet seen anybody making that case at all.
So it is not obvious looking at the usual suspects why we have failed to see US housing markets responding with a quantity supplied increase to this clear increase in both rents and prices. Rents have risen, but interest rates remain not too high, and supply has not been forthcoming.
Having poked at the estimable Shiller, I recognoze that while he emitted a vibe of worrying about a housing bubble he held back from outright declaring that we are in one. I am sure he knows the data that I have pointed out here, even if he was not fully forthright about it in his recent columns and other public offerings. Not all is well in the US housing market.
The price/rent ratio has been rising, showing some signe of incipient bubble behavior, if still not too wild in the vast majority of the US. But the failiure of housing supply to respond to the clear increase in rents, let alone the larger increase in housing prices, is something worrisome and to be concerned about. At the bottom end, rising rents drive rising homeleessness more than rising house prices, and we already have more homelessness than we should have.