Monday, May 9, 2011

The Brad DeLong - Bennet McCallum Debate

Over at The Economist there is an interesting debate taking place between Brad DeLong and Bennet McCallum.  They are responding to the following statement: this house believes that a 2% inflation target is too low.  The idea behind this statement is that with a higher inflation rate the targeted short-term nominal interest rate would be higher and thus less likely to hit the 0% bound.  Brad DeLong endorses this view.  He sees the 0% bound as a real constraint on monetary policy and wants to avoid it.  Bennet McCallum challenges it.  He argues that monetary policy is not powerless at the 0% bound and there are real costs with going to a higher inflation target.

A key issue to resolving this debate  is how binding the 0% bound is for monetary policy.  My own view is that it is not truly a binding constraint, but only a self-imposed one because of the way monetary policy is normally conducted.  Conventional monetary policy targets a short-term nominal interest rate.  So when the 0% bound is reached monetary authorities have to switch over to their "unconventional" monetary policy bag of tricks.  But it doesn't have to be this way.  Imagine if the Fed targeted the price level at a 2% growth rate and didn't use the federal funds rate as its instrument.  It simply adjusted the monetary base to hit the price level target and communicated very clearly its goals to the public. Assume that as part of this communication the Fed said it would do whatever is necessary to hit its target, including buying other assets than just t-bills if the need arose.

In that setting it is hard to imagine why the 0% bound on the short-term interest rate would ever matter.  First, the 0% bond would rarely be reached because nominal expectations would be well anchored.  Second, even if it did, say because of a major aggregate demand shock that caused deflation, the price level target would require significant catch-up inflation that would lower the expected path of real interest rates presumably enough to restore full employment. Over the long-run there would be price level stability as the price level returned to trend and 2% growth.  Thus, the 0% bound would not matter and there would be no need for permanently higher inflation.

On the catch-up inflation scenario above, something similar happened during the 1933-1936 period.  FDR communicated clearly that he wanted the price level to return to its pre-crisis level and backed up his talk with the devaluation of the gold-content of the dollar and deciding not to sterilize gold inflows. (See Gautti Eggertson and this for more.)  Short-term rates were at the 0% bound at this time too. Nonetheless, this monetary easing sparked a remarkably robust recovery that was unfortunately cut short. 

What all this means is that we can have our cake and eat it too.  If the Fed were too adopt an explicit price level target and vow to hit it no matter what (i.e. engage in other asset purchases if necessary), then Brad DeLong would get some higher-than-normal catch-up inflation and Bennet McCallum would not have to worry about a permanently higher inflation rate.  Better yet, if the Fed were to adopt a nominal GDP level, then DeLong and McCallum wold get all the above benefits plus the fact that the Fed would not be responding inappropriately to aggregate supply shocks.

P.S. See the recent posts by Josh Hendrickson and myself on why the 0% bound typically is not enough to create a liquidity trap.

No comments:

Post a Comment