Tuesday, December 9, 2008

Brad DeLong Watch - A Posteori Predictions

My son had a team assignment in school recently. The team had to choose some stocks and then track them. There wasn't any reward for selecting the best performing basket, but still, at the second meeting, the team felt obliged to replace surreptitiously one of their stocks with another, which had gained 15% the previous day. Pride, after all, incites as much as greed.

I am reminded of that when thinking of Calculated Risk's challenge to economists to "do some serious soul searching" and write a post why they were wrong about the housing crisis.  Not too many have accepted the tossed-down gauntlet.  Most economists, it seems, prefer to think that part of the job description is never having to say they're sorry.   Brad DeLong is to be commended for having at least taking the challenge seriously enough to provide a reply.

Sadly, he fell guilty to exactly the same temptation that my son's team did.  He wrote:
Let me say what things I was "expecting," in the sense of anticipating that it was they were both likely enough and serious enough that public policymakers should be paying significant attention to guarding the risks that it would create:
(2) A fall back of housing prices halfway from their peak to pre-2000 normal price-rental ratios.

Wow! A halving of housing prices! That's some foresight. Only, if you check previous posts, Brad DeLong sang a different tune. In December 2007, in reply to a Krugman blog entry which said housing prices had "a long, long way down" to go, he wrote:

"A long, long way down" means, I think, "a fifty percent fall along the coasts."

I would cut that in half for two reasons: not 50% but 25% along the coasts, and much less in the interior. First, the likelihood that savings interested in being invested in the secure-property U.S. will be ample over the next generation is high: we will sell political risk insurance to foreign individuals and governments for quite a while yet. So real interest rates are likely to be lower in the past. Second, the zoned zone is not growing--and America's population still is. The gap between heartland and coastal values is likely to grow over time, and that anticipated capital gain should push up prices now.

Perhaps that was a slip of the tongue? No. Here is he again in April 2008:

That [a decline in home prices, back to more or less their pre-bubble inflation-adjusted levels] strikes me as too pessimistic. The rise of Asia and the resulting demand by the rich and by governments for U.S. assets to hedge political risk is likely to keep savings glutting for decades. We aren't buiding more superhighways, there are no major transportation improvements on the horizon, America is filling up, and so land-value gradients are on the rise. If the income distribution continues to erode, we will wind up with higher prices for scarce positional goods--chief among which is location, location, location.

My guess is that we will ultimately give back half of the doubling...

Half of the doubling means, yep, -25% off the peak.

Am I being too hard on Professor DeLong? Almost surely. He did, after all, qualify his "halfway" remarks by saying he thought it was something that needed to be anticipated, not something he thought was going to happen. And he did, as I said, answer Calculated Risk's challenge; kudos to him for at least realizing the question was important. It is not really him, but the others who are silent, for whom we should ask, "Why oh why can't we have better economists?"

No comments: