Wednesday, October 05, 2011

Do you believe in magic?

In a recent post, Scott Sumner claims that the Fed can easily credibly commit to a nominal GDP target. He further claims that the act of choosing the target in itself would allow the Fed to hit the target without any great amount of monetary expansion:

"The Fed has plenty of credibility, that’s not the problem. The problem is that they are using the credibility to assure investors that low inflation is here to stay. With the right target, there would probably be no need for massive quantitative easing, or other extraordinary policies.

The punch line is that the problem isn’t the Fed’s unwillingness to do enough QE, twists, or cuts in IOR, the problem is the Fed’s inadequate target, just like in Japan."


In other words, the Fed can credibly commit to arbitrary future policies so well that a simple announcement of a new policy path will put the economy on that path without any heavy lifting required.

In other other words, magic!

People, the Fed has no ability to make credible commitment to future policies that might conflict with their period by period preferences! Read Kydland & Prescott. Read Barro & Gordon. Repeat after me: time inconsistency, time inconsistency, time inconsistency.

All the Fed can do is act each period according to its own preferences (yes it's weird to treat the Fed as a unitary actor given all the dissents we've seen recently, sorry).

Many societies have responded to this dilemma by invoking the Rogoff gambit; we appoint conservative central bankers. They deliver low inflation not because they are following some policy rule or pre-determined path, but because their preference is for low inflation!

The idea that there's a magic bullet out there that solves our economic problems, that the Fed could fix things by putting out a press release with a couple sentences about their plans for future NGDP growth, is wrong and somewhat dangerous.

3 comments:

Tom said...

You are too hard on Scott Sumner. He wants the Fed to make grave pronouncements and the do nothing. I don't care about pronouncements as long as the Fed does nothing. We may have gotten to our positions be different paths, but the important thing is that we agree on policy.

Anonymous said...

Didn't the market react to the announcement of QEII? It's probably true that the Fed can't just make the announcement and then let the market turn it into a self-fulfilling prophecy, but it's pretty likely that once they make the announcement and act on it for a reasonable fraction of time interval, the market will start take them at their word. I don't think this is irreconcilable with Kydland and Prescott.

Norman said...

"All the Fed can do is act each period according to its own preferences (yes it's weird to treat the Fed as a unitary actor given all the dissents we've seen recently, sorry)."

Isn't this essentially Sumner's point, though? Isn't the stated purpose of Sumner's blog to influence those preferences? He wants to change the views in the profession in general, but also to change the views of economists with a voice in setting the Fed preferences in particular.

This might be more ambitious than is reasonable for a blogger, but as Alex says I don't think it's in conflict with the time inconsistency literature. The Fed doesn't have to lie (which wouldn't work) to get markets to believe in a 5% GDP level target if the central bank *actually* wants a 5% GDP level target.