Nate Silver had a tweet this morning that’s relevant to a debate that went on here a month or so ago.
The median American’s non-household wealth declined by 14% between 2001 and 2007. So when household wealth evaporated, guess what happened?
I’m not sure of the source of this, so take some of this with a grain of salt. But if it’s true, it is relevant to something Daniel Davies claimed and Brad DeLong rejected, namely (to quote Daniel) “we are in a recession basically because of the disppearance of a huge amount of household sector wealth”.
I basically think Daniel is right on this, and Brad wrong, for reasons I’ll go into below the fold. And I take it Nate is endorsing Daniel’s line, namely that the recession was brought about by a huge collapse in household wealth.
Brad offers, I think, two arguments against the wealth effect explanation of the recession. One is that “wealth had disappeared before—remember Black Monday on the stock market in 1987, or the collapse of the dot-com boom?—without it triggering a Lesser Depression.” But both of those collapses led to a huge loss of wealth among people with substantial stock holdings, i.e., among people with the lowest marginal propensity to consume in the economy. That’s very different to a general fall in the value of houses; that tends to affect people who have a very high marginal propensity to consume, and hence lead to recession.
The other argument is a timing argument. Brad says that new construction fell well before the recession hit. And it’s true, it did. And that would be a good argument against the claim that the recession was caused by job losses to construction workers. But it’s not a good argument against the claim that of household wealth caused the recession. Let’s look at when house prices started to fall (at least in the US).
That looks like like house prices collapsed just before the recession hit. So the timing works.
It’s true that house prices start falling in 2007, and there’s no recession until into 2008. But I think (a) you’d expect a wealth effect to percolate slowly through the economy, and (b) people didn’t think house prices had collapsed in late 2007, early 2008. It’s true by then that people knew that house prices in Phoenix and inland California were on their way down. But you don’t see much rapid movement downwards in the prices in non-bubbly areas until much further into 2008. If you like you can look through the data——to see that. Or you can just remember what that time was like. Certainly around New York, the prevailing wisdom was that a bubble in desert properties had quite properly popped, but that was no reason to doubt that New York properties would be affected. And lots of building went on in 2008 on that basis.
So the timeline works out for the theory that loss of household wealth is what did it. On the other hand, I don’t understand Brad’s positive theory of the recession. Here’s what he says.
It was because people recognized that banks that were supposed to have originated-and-distributed mortgage-backed securities had held on to them instead, that as a result a large chunk of the $500 billion in subprime losses had eaten up the capital base of highly leveraged financial institutions, and that you were running grave risks if you lent to a bank. The run on the shadow banking system that followed was the source of the crash as financing for exports and for equipment investment vanished, and then the whole thing snowballed.
But if banks had followed the originate and distribute model, and we found out that all the MBS’s were lousy, you would have had risks on lending to anyone in the financial system anywhere. That would seem to generate an even bigger run on the shadow banking system. Frankly, a pure originate-and-retain model, followed by the bankruptcy of a few dedicated mortgage brokers, would probably have been less complicated to unwind than the mess we have. Certainly it seems that the fact that so much of the market was orginate-and-distribute is causing big problems for doing mortgage adjustments. (I think that SemiEcon makes basically this point in Brad’s thread.)
The dispute between Brad and Daniel is caught up in a bigger dispute about whether the banks are to blame for the crash. I wonder whether this is going to be too hard to figure out without doing some tedious work on defining just what we mean by “the crash”. Here’s one rough take on the Bush era economy. It basically was awful; a long recession in the style of 1920s England. This fact was disguised because house prices kept going up, and people were able to turn rising prices into consumption via HELOCs and the like. In turn, the easy availability of just this financing helped keep house prices rising. But this couldn’t last forever, and when it stopped, there was a crash.
That suggests that bad, or at least odd, behaviour by the banks contributed to the crash. But it also suggests that a more responsible banking sector would have left us so badly off all through the 2000s, that the economy would have been just as bad by the end of Bush’s term. (Assuming, in the relevant counterfactual world, that Bush even won in 2004; without the housing bubble, he might not have.) If that’s right, the banks aren’t to blame for the crisis; Bush is. And the solution to the crisis isn’t to fix the banking sector, either through regulatory reform or continuing to bail out the banks, it’s to stop Bush-era economic policies.