Wednesday, October 26, 2005

Barro Predicts Taylor Equivalence

Interviewed by the Minneapolis Fed this past summer, Harvard's Robert Barro says:

Region: Is the relationship you have in mind here similar to a Taylor rule?

Barro: Yes, it looks like that, but there are some differences. For example, it's clearly not GDP that the Fed reacts to. Given what's going on in the labor market in the sense of employment growth and the unemployment rate, empirically there's no reaction of the federal funds rate to GDP growth. Hence, when GDP growth goes up because of higher productivity growth, there's no tendency to raise the federal funds rate.

But if the economy is strong in terms of a tight labor market, meaning higher employment growth and a low unemployment rate, there seems very regularly to be an upward movement of interest rates. The reverse holds when the labor market is weak.

On the inflation side, it's a broad inflation index that seems to influence the Fed—something like the GDP deflator or the deflator for personal consumption expenditures. The Fed does not seem to react on a month-to-month basis to the consumer price index or the producer price index. So, in a general sense, it's a Taylor rule but it has some features that differ from [Stanford University's] John Taylor's precise formulation.

A crucial feature is that monetary policy responds to the economy in a way that's particularly strong with respect to inflation. The idea is that when you bring the [federal] funds rate up, inflation is supposed to go down. This mechanism has worked much better than I would have predicted at the end of the 1970s and early 1980s—it's kind of amazing how well this has turned out, and not just in the United States. Many other countries have gone further in terms of formalizing this policy reaction into a rule. I think the more formal rule is a good idea. I'm not sure it's that important because the United States is pretty close to it anyway. However, if Ben Bernanke becomes head of the Fed, he will likely try to implement something that looks more like a formal rule.

No comments: