A post at Free Exchange reminds me of something that I have been meaning to look into for a while. Willem Buiter had a series of posts a while back explaining that housing could not cause a recession because housing wealth was not wealth. The accuracy of his predictions aside, the series always struck me as being off the mark. Now Free Exchange points to a piece at Real Time Economics which revives the case:
A more careful look at the data, guided by economic theory, however, suggests that much of this evidence [for a housing – consumption link] has been misinterpreted and that the reaction of consumption to housing wealth changes is probably very small. First, several recent studies have shown that the logic of the housing wealth effect is faulty. Houses, unlike stocks, are not just assets but are also consumption goods themselves. Theoretical papers by Willem Buiter, as well as by Todd Sinai and Nicholas Souleles, cast doubt on the notion of a large spending effect from housing
There are a lot of very nerdy economics issues to be waded through here – issues that I have an interest in pursuing. For the purposes of this blog, however, let me cut to what I think is the chase. Here is my interpretation of Buiter’s logic: When the price of houses goes up that doesn’t represent an increase in anything real. Houses still do what they have always done, keep people warm and dry. So why should we expect real economic effects from an unreal cause. That would imply some sort of systemic irrationality on the part of consumers.
The problem with this argument is two-fold. First, there is an illusionary effect. Increases in the price of housing can be symptoms of a real change. In the latest case they were symptomatic of a technological change in financing. That change, we believed at the time, reduced the real costs of risk. A lower risk society is a wealthier society and that wealth manifested itself in higher home prices. Greater wealth drives greater consumption. This effect is closely related to the “permanent income” effect that Calomiris, et al report at Real Time. However, it is important to note that housing prices signal real changes in the economy. In some sense this holds true for stocks as well. Much of the inflation adjusted gain from holding stocks stems from the dividend, but not all of it.
Where does the rest of the gain come from? After all, a non-dividend gain can only be realized by selling the stock and for every seller there is a buyer so isn’t society as a whole neither long nor short that gain? Indeed, the gain comes from future expectations. One could say that once you account for future expectations you will see that there is no real effect of stock price increases on consumption but this misses the point. Stock price increases are important because they are a signal of changing expectations about the future.
There is second, intellectually more interesting effect, however. That is, that home prices provide buyers will a real option that is socially valuable. Everyone, on net, really is better off in a world with higher home prices. Why?
Well, rising home prices allow an owner to sell for more than he or she owes. This in turn lowers the systemic probability of default. Systemic default serves as a tax on financial markets. There are wealth creating transactions that will not occur in a high default environment. There are willing buyers who could be matched with willing lenders if only there wasn’t the systemic problem of default. The important thing to note is that nothing has changed about the buyers or sellers individually but in an environment of rising home prices both can do more business.
This is more than the second-order effect of relieving constrained households. I am not simply arguing that there are some people who exogenously want to consume more but are constrained. I am arguing that society is poorer when housing prices are falling because society is facing an implicit tax. Thus the optimal level of consumption should rise when housing prices are rising.
Interestingly, this effect depends heavily on expectations and so could be spuriously ruled out by Calmiris et al’s technique. That is, when people expect housing prices to rise they feel better. They feel better because they have just received an implicit tax cut (sometimes made explicit by falling mortgage rate spreads.) If we rule out the expectation effect we are missing a real effect of rising housing prices. Said another way, rising housing prices and positive attitudes don’t happen to go together and it isn’t just that home prices are rising because expectations are rising. Rising housing prices are driving the positive attitudes.
Much more needs to be thought, researched and written about this, but at this point I am inclined to believe that housing price can and should have strong effects on consumption.