Site Meter

Thursday, January 27, 2005

One of Tilman Ehrbeck’s Good Ideas.

As I will mention above, Tilman Ehrbeck has had many very good ideas including the following. He began writing his dissertation in 1989. At the time, the rational expectations hypothesis was the hypothesis which must be maintained. Like all people who I have met except someone named Adriano something, he found it implausible. However, he neither decided to accept it as the price of being an economist or to generally insult it as an alternative to doing actual research. Instead he thought that, if people use rules of thumb, really smart policy makers might try to design policy so that, in the event, the rule of thumb corresponds to rational expectations. This is a very hard challenge, since the policy maker must understand both the way in which people form expectations and the law of motion of the economy (Tilman likes hard challenges).

Just minutes ago (more because blogger ate the first draft of this post) I thought of an application of this good idea in economics. It is a case for social insurance. The argument is as follows: People tend to assume that permanent income is current income times a constant greater than one (we have noticed that income tends to rise with labor market experience). This means that policy makers should try to design taxes on labor income and unemployment insurance so that permanent income is current income times that constant greater than one. This is a bit challenging since they have to estimate that constant (what is a reasonable estimate of this constant? Hell if I know. Here is the PSID. Calculate it yourself). The application of Tilman’s idea is that, if people use a rule of thumb to forecast future incomes, policy should be designed so that it is close to an optimal forecast. Otherwise people will make systematic errors in their consumption/savings decisions which imply dead weight losses.

Not bad for a G1 no ?

No comments: