Why Does Housing Matter?

8/20/2012 01:39:00 PM
As I'm currently homeless and looking for a new apartment, I can't help but think about housing policies during the recession. Since the beginning, there has been a large and vocal segment talking about the need to "fix" the housing sector before the economy gets better. Ezra Klein makes the argument that the failure to revive the housing market is the primary reason why the economy is not doing as well as everyone hoped it would by now. In his words, the "precise nature of the administration’s misunderstanding was that the key problem was household debt, and until that problem was solved the economy couldn’t recover...it never had a clear strategy for reducing housing debt."

This perspective has always bothered me a little. I would go so far as classify it as a logical fallacy--the idea that to resolve an issue you have to remove what caused the issue. The same fallacy is pervasive in much of psychology, although less so than it used to be. A while back Krugman offered a great analogy that shows just how wrong this fallacy can be: a flat tire. If you have a flat tire on your car, that's a big problem. Think of a flat tire as being an economy in recession, and a non-flat tire as an economy at full employment and output. What caused the flat tire? Well, if it was previously full, then clearly there's a hole somewhere. But, as Krugman says, you don't have to try fill the tire through that hole! That would be idiotic--tires have valves that make it easy and efficient to regulate the air pressure inside. Think of these as being the government. Now, sometimes tires just go flat from time to time just from old age. In that case the best we can do is refill them--the holes are too tiny to be able to "fix." Other times, something has actually punctured the tire and we need to fix that hole, or else the tire will just go flat again after we fill it.

Arguably, the housing crisis actually caused a "rupture" in the economic "tire." So we would be justified to offer up regulations designed to prevent another housing crisis from occurring in the future, which is what Frank-Dodd attempted to do. Of course, plugging the hole isn't enough--we still have to re-inflate the tire, and the question is how. Klein is arguing that we have to pump the air back in through the same hole we just plugged--re-inflate the housing sector.

Unfortunately, that's where the analogy ends. Now back to economic theory--is there a basis for saying that we need to resurrect the housing sector to have a recovery? There are two components to this issue: the balance-sheet story, and the structural change story:
  1. Balance Sheet Recession
    The balance sheet story says that the economy is weak because people are averse to holding debt. This wasn't a problem during the housing boom, because on average homeowners weren't actually going into debt to buy their houses--the price of their house was bigger than the loan they bought it with, so they were on net actually saving money. But the decline in home prices represented a transfer from homeowners to banks, putting homeowners underwater, and thus in debt. They are averse to holding debt, so they've since reduced consumption and tried to increase savings to compensate. In the face of nominal price and wage rigidities, this decline in demand resulted in decreases in output and increases in unemployment. (As an aside, David Andolfatto has a persuasive argument why nominal wage rigidity doesn't matter the way Keynesians think it does)
  2. Structural Change Recession
    Arguably, a lot of our institutions were trained and tooled specifically for building houses prior to the recession, and aren't as efficient doing anything else. That is, we had a lot people who were very good at construction, but not good at any other job. This meant that when there was a change in demand--for whatever reason--that resulted in less demand for housing, the productive capacity of the economy declined. A few of those construction workers left housing and took up their next-best occupations, while a few more construction workers decided to retire early due to falling wages in the housing sector. This resulted in a predictable fall in GDP. But the structural change story is missing something: why did unemployment rise? The fact that construction workers aren't as good at, say, car mechanics as they are at construction is no reason for them to be unemployed--I could certainly collect data on their productivity and come up with a wage low enough at which I could profitably employ them in a mechanics shop. If they got such offers by think that the wages are too low, fair enough, but then they are technically not unemployed--they are out of the workforce.

    Then there is the additional problem that there doesn't seem to have been a structural change--the economy is depressed across the board, and job losses have been distributed across all industries, with no notable increases in wage differentials. But even ignoring the data, I have yet to find a model of structural change that results in cyclical unemployment without resorting to some type of non-structural-change mechanism (like sticky wages).
Ezra Klein actually argues (1), but I've seen (2) a lot. What are the policy implications for each?

I've written before about some implications of balance-sheet models before. Klein claims (or subtly hints, really) that we need policies to write down household mortgage debts. I'm not terribly opposed to this idea--it basically institutionalizes the practice of "strategic defaults," which I think is fair since it was up to banks and investors to recognize that they bear the risks if the mortgage goes underwater. If they didn't realize that the incentive to default depended on the future price of the house, they are idiots. But I'm skeptical of the macroeconomic benefits: one person's debt is another's asset, so while we would be rearranging who has debt and who doesn't, we aren't actually increasing aggregate net savings with such a policy, and hence not necessarily eliminating the balance sheet problems we face, unless the heterogeneity between homeowners and investors is substantial. I can think of two alternative policies that avoid this critique: one is to expand the money supply. Since households can stash hard cash under their mattress, or at least let it sit in their checking accounts (which in turn would sit in reserve accounts at the Fed), this could restore their balance sheet so long as it is not highly inflationary, which until balance sheets are restored it won't be. The second option is to increase government borrowing. This allows the private sector to increase savings in the aggregate, since we have a corresponding agent (the government) going into debt. The fact is that the government has gone deeply into debt, and the money supply has increased dramatically:
It is worth noting that that increase in monetary base had very little effect on the aggregate money supply due to the decline of the money multiplier at the zero lower bound (something that my blog co-author Daniel J Molitor can tell you all about, since he has written a thesis on this). And maybe it is the balance sheet effects of these factors that led to the recovery we did see, but I don't really see it. I think that there are balance sheet effects in the economy, but they are not the primary reason the economy is still depressed.

Now to the structural change story. The decline in GDP due to structural change is regrettable, I suppose, but still optimal. We are not wealthier if we are manipulating the economy to produce things other than what we want, simply because we are better at producing those things. So the real problem is unemployment, not GDP, which the structural change story is incapable of explaining. But proponents of this narrative seem unphased by the existence of cyclical unemployment right now, so lets continue: the policy implication of structural change is to provide resources to retrain workers for new jobs in higher-demand sectors, to make them more efficient at producing what we do want. As above, this does not suggest that we should do anything at all about the housing market.

In conclusion, the fact that the housing market caused the recession, or the fact that the housing market remains depressed, does not imply that we would benefit much from new housing-related policies.