Freakonomics: Cracking the Real Estate Code

Wired has an excerpt from Freakonomics, which has been making all the usual book-promotions rounds lately. I thought this anecodte was funny:

Consider this true story, related by John Donohue, a law professor who in 2001 was teaching at Stanford University: "I was just about to buy a house on the Stanford campus, and the seller's agent kept telling me what a good deal I was getting because the market was about to zoom. As soon as I signed the purchase contract, he asked me if I would need an agent to sell my previous Stanford house. I told him that I would probably try to sell without an agent, and he replied, 'John, that might work under normal conditions, but with the market tanking now, you really need the help of a broker.'"

But, as much as I'm intrigued by Freakonomics' "math and numbers and some research = astounding party conversation!" model, I think there's a strange danger in "find the correlation" work. I mean, take this quote (again, from Wired):

Or consider these findings of a 1996 medical study: Obstetricians in areas with declining birthrates are much more likely to perform cesarean section deliveries than obstetricians in growing areas - suggesting that when business is tough, doctors may try to ring up more expensive procedures.

But there are dozens of possibilities for this relationship — many that go beyond hucksterism. The smartest one I can think of comes from Freakonomics itself: Did they control for the age of the parent? (As in: Do older moms-to-be have fewer children and more cesareans.)

Maybe they did. I have no qualms with the authors, for all I know, they're impeccable researchers who've put a ridiculous amount of time into this.

Maybe it isn't the "correlation hunt" paradigm I have problems with, but the "find a correlation and dumb it down for a mass market paperback" approach. I can deal with numbers. I'd rather see the data for myself and be led to a conclusion, than simply having a conclusion handed to me "in the raw" — so to speak.

But, maybe that's just me.

Permalink • Posted in: words • Post a comment