Thursday, July 15, 2010

Behavioral Economics Behaving Badly

I agree 100% with Loewenstein and Ubel's Op-Ed that behavioral economics is overused. I just think all of their examples are rubbish and reflect poor economics.

Like obesity. Obesity is only a problem if you believe people have self-control issues; thus you need behavioral economics. There is little evidence that obesity causes externalities. And the impact of corn subsides on the price of processed foods is negligible. In fact, it is precisely distortionary tariffs on sugar (that economists also hate) that make sugar (and corn syrup) prices too high relative to what they should be in a market system, not too low.

On sunshine provisions, I agree there is scant evidence that these work well, but it is hard for an economist to believe that they are a bad idea. Also, I don't know the literature on pharmaceutical industry gifts, but I suspect the econometric identification is poor on them. I do know the literature on lobbying, and know that there is scant conclusive evidence that lobbing is a problem.

It is also strange that the only health care reform it proposes, requiring patients to pay out of pocket, seems at odds with the rest of their message. That patients are unable to make smart choices when given information.

I agree with them that a gasoline tax as well but they neglect to think about magnitudes. If you look at the elasticity estimates for the impact of an optimal gasoline tax (maybe 25 to 50 cents more than today according to Parry et al's AER paper), the likely reduction in gasoline use is is also on the order of a couple percent (typical long run elasticity estimate of oil is maybe 0.4, so a 10% increase in price is a 4% decrease in quantity, actually less than that if supply adjusts which it should), very similar to the oPower behavioral economics based information about neighbors experiment, which the authors scoff at.

No comments: