The Trouble With A Bubble

Even though the panel I was on last night had nothing to do with housing, the guy up on stage playing the cranky rightwinger role couldn’t resist making an offhand reference to the idea that our problems are grounded in the fact that people bought houses they couldn’t afford. I think this misunderstands the issue in an interesting way that’s worth spelling out in slightly tedious detail.

Start out with the idea that a developer builds a house out of cash that he has on hand. Now you have an asset — a house. Then he sells the house to a person, who buys it with a mortgage. Now you have two assets. There’s a house, owned by a family, and a loan owned by a bank. Then the loan goes through the securitization rigamarole but the loan still exists out there somewhere as a financial asset. And then there’s the house. The homeowner has possession of the house, and he also has an expense in the form of his mortgage payments.

Now let’s say the homeowner takes a negative shock to his income. He had a freelance deal with a magazine that goes bust. With his new lower income, the homeowner needs to reduce his expenses. He can do that two ways. One would be to severely crimp non-housing consumption because he “bought a house he can’t afford.” The other would be to sell the house and move into a cheaper one. Either way, it’s sad for him, but it’s just sad in the general sense that his income has gone down. The economy as a whole trundles along fine. It’s always the case that individual households’ incomes are fluctuating and that consumption levels are therefore fluctuating as well. If you’re eating at restaurants you can’t afford, the solution is to start eating at cheaper places. If you’re renting an apartment you can’t afford, the solution is to move to a cheaper apartment. And if you own a home you can’t afford, the solution is to sell the home and move to a cheaper one. The problem facing the typical troubled household in America today, is that this “sell and downsize” process doesn’t work. That’s because the price of houses has fallen, but the price of already-extant mortgages hasn’t fallen. In other words, if you paid $350,000 for a house that’s now worth only $250,000 then selling the house doesn’t get you out of the need to pay your mortgage.

This is the actual problem. In time one, the value of the assets owned (houses) was proportional to the value of outstanding debt claims (mortgage payments). But now in time two, the value of the assets owned is much smaller than the value of outstanding debt claims. As a result, households can’t scale back their consumption of housing to an affordable level. They have instead two choices. One is that they can strangle non-housing consumption in order to pay off the old debts, and the other is that they can default on their housing debts. Insofar as they do the former, the “real” economy of producing goods and services suffers from reduced consumption. If they default, then what happens is that the value of the other asset created at the time the house was sold — the mortgage — declines, and the financial economy suffers. Banks end up undercapitalized. Consequently, our policymakers are currently in the position of trying to allocate losses between the banking system and non-housing household consumption with an emphasis (it seems to me) on trying to mitigate banking system losses.

There are a few different ways you can categorize this problem. One, à la Dean Baker, is as a wealth effect issue where the vanishing of notional housing wealth is strangling nominal spending. Another, à la Scott Sumner, is an income effect issue where the collapse of nominal spending is destroying the housing wealth. But either way, past overconsumption of housing is not the problem. You could imagine a country where industrial output is collapsing because of a lack of adequate production inputs because all the raw materials have been turned into houses, but that’s not the country we’re living in.