One UPDATE below.
So you know how during the New Deal they had federal agents go around and tell farmers to overturn crops, etc.? (I haven't actually looked this up myself, but I bet that the famous scene in The Grapes of Wrath where they douse oranges with gasoline, was not really a market phenomenon. And I don't just mean, because the Depression was caused by the Fed.)
Well, Holman W. Jenkins, Jr., in a July 30 op ed, said that one solution to the housing mess would be to blow up some of the excess housing to bolster prices. No, this is not The Onion, and I think Jenkins was being serious:
So far, Washington has put its political capital into trying to refinance salvageable homes for unsalvageable homeowners, when a relevant policy would consist of judiciously [!] buying unsalvageable houses and demolishing them. Fannie and Freddie's strength is software: They could be put to work devising a least-cost, maximum-bang strategy for demolishing unoccupied homes to preserve as much value as possible for the homeowners and mortgage creditors who remain.
OK first of all, what's wrong with this? It's even cruder than the broken window fallacy. Jenkins isn't saying, 'Let's blow up some houses to give work to the homebuilders.' No, he's saying, 'Let's blow up some houses because otherwise the price of houses will continue to fall.'
You don't make a country richer by destroying houses. If you were going to disassemble the houses, and redirect their lumber, nails, wiring, etc. into other lines, then maybe that could make sense. Or, if we wanted to use the land on which these houses stood, in order to hold an apartment unit, a shopping mall, a factory, etc., then that could make sense.
But to simply reduce the housing stock, when those houses are at this point sunk cost, is insane. I can't believe the WSJ didn't at least ask him to put in a paragraph dealing with this type of objection.
Think of it this way: Suppose you personally were given a newly built house in an neighborhood stuffed with McMansions etc. The donor said, "Now look, you're gonna have a tough time selling this. When I put down the money to have this thing built, I was hoping to unload it for at least $250k in 2007. Well it's been on the market for 15 months now, and not a bite. I even lowered the price 3 times, and nobody's is even looking at $200k."
Do you really think you'd tell the guy, "No thanks. If you gave that to me, I'd just end up knocking it down and selling the plot of land. With all the hassle and possible lawsuits from neighbors, I don't want to deal with the headache."
No, I don't think you'd do that at all. You'd take it, and drop the price until it sold. (Or, you'd sit on it for a few years waiting for the market to recover.) If it's in a residential zoned area next to 30 other houses, the only thing that land can be used for is to hold a house. So unless there was some fundamental flaw in the architecture, there's no way the real estate can be worth more w/o the house than with it.
Even if you had to drop the price down to $50,000 to sell it, you would still be up a lot of money from the whole deal. You would want to kiss that donor for giving you a gift of $35,000 (netting out closing costs and your headaches).
Of course, if you sold it for $50,000, that would bring down the official property values in that area. But so what? Market prices communicate information. If the market is so bad that what used to sell for $250k now only sells for $50k, people need to see that and adjust their plans.
Final point: Here's the real interesting question. Suppose that Fannie and Freddie followed his advice and set up a sophisticated computer program with all the variables including time-indexed supply and demand functions for housing, a general equilibrium model of the economy and housing sector, production functions for new housing, etc. etc. Do you think if the programmers were competent, the answer would pop out:
THE OPTIMAL NUMBER OF HOUSES TO DESTROY: 0
UPDATE: I just realized that that's not what would happen. Really the answer would be negative; i.e. the computer would tell them that the economy would benefit from the further construction of houses. (And, as Jenkins notes, that's exactly what's happening right now--building continues in some areas.) The programmers would be stunned, and then if they put in a non-negativity constraint, the answer would then be zero. At that point they would probably start looking for the bug in their code, since they were expecting a positive number.