Comment on Price Targeting

October 19, 2010

One idea reportedly being explored at the Federal Reserve would be to target a general price level.  Many central banks target inflation, which is the rate of change, or first derivative, in a general price index.  The Fed is not one of them, but the desired pace of price change has been informally expressed by U.S. monetary officials as lying either between 1.5% and 2.0% or, more narrowly, between 1.75% and 2.0%.  The latter coincides with the ECB’s target of “below, but close to, 2%.  For central banks that target inflation, the time horizon covered by the forecast is a medium-term span of one to two years into the future.

A target for a price level is not the same a targeting a point, and it makes sense only for an economy when inflation has deviated significantly from a notion of optimality, that is price stability.  The targeted price level will be unique for every point in time and not the same for different moments in time.  This is possible because the price level target is in fact a designated path for the price level.  But whereas an inflation target, which also implies a path of price levels, disregards past deviations from the desired path, a price level target critically incorporates that information.  In current U.S. circumstances, officials would go back to the end of 2007 when the previous business upswing ended, declaring the price level at that moment appropriate and applying a 2% rate of price rise from that point to create a targeted path for prices.  In this way, time after end-2007 when prices were rising at less than 2% per annum will be equally offset with time in the future when prices will be allowed to climb faster than 2% so that the actually level of prices can eventually catch up to the vector of 2.0% inflation from end-2007. 

The greater the speed by which prices rise more rapidly than 2%, the shorter the period of time that it will be necessary for the Fed to target a price level.  By assigning different dates in the future when one hopes to reconnect with the 2% inflation vector, it is possible to calculate an average rate of inflation between now and then that will enable the central bank to meet its goal.  If the Fed were to use the consumer price index, it has to climb 3.8% per annum to accomplish the mission by end-2011, 3.0% per annum to do it by end-2012, 2.7% per annum to do it by end-2013, 2.5% per annum by end-2014, and 2.4% per annum for 5-1/4 years to get the job done by end-2015.  With core CPI, which the Fed is more likely to favor than total CPI, the intermediate tolerated rates of inflation would be 2.9% to complete the task by end-2011, 2.5% by end-2012 and 2.2% by end-2015.  These simulations underscore that the Fed will tolerate higher inflation over the next few years by adopting a price target than would be the case with its informal notion of an acceptable inflation ceiling.

It should be pointed out that this kind of logic was not applied in reverse when the back of double-digit inflation was broken in the 1980s.  If that had been the case, the central bank would not have taken its foot off the brake when inflation slowed from 14% to 4%, or even if it had plunged to minus 5%.  A price contraction of 5% would not have neutralized an increase of 14% earlier.  However, inflation and deflation are different phenomena, and they present monetary officials with quite different challenges.  So the fact that price targeting doesn’t make sense in wholly opposite circumstances doesn’t invalidate the concept’s merit in ultra-disinflationary times like the present.

It’s actually healthy for monetary officials to be asking if they should not be aiming to offset lower-than-desired inflation that has already occurred.  It’s called correcting a prior mistake, and it will help put expectations in right relationship with where policymakers want to steer.  Regular readers of this site might even detect a similarity between the idea of a price target and the model I’ve used to define a “normal” labor market and to measure the size of the U.S. jobs deficit.  Jobs grew at virtually identical rates in the 1980s and 1990s but stopped expanding at all after 1999.  By applying the 1.839% per annum rate of employment growth between end-1979 and end-1999 to the period that followed, I derive a long-term normal jobs trend-line for the subsequent 10-3/4 years.  By subtracting actual employment from that trend-line at any point in time, it’s possible to quantify the jobs deficit.  If we are ever to recapture a job market like the one we used to have, that deficit will need to be closed.  Like a price target that climbs 2.0% per annum, the optimal jobs level grows at a pace of 1.84% per annum.  To fill in the jobs deficit as soon as end-2011 would require non-farm payroll employment to rise on average by 1.953 million workers per month over the coming fifteen months.  To normalize the labor market by end-2013 would take jobs growth of 903K per month for the next 39 months, and to fulfill the mission by end-2015, jobs must advance 656K per month over the coming 63 months

Needless to say, none of the above yardsticks will be met.  Indeed, it may be that nobody alive today will see jobs touch a vector of 1.84% per annum growth from the end-1999 actual level.  Even though the Fed could not responsibly stimulate monetary policy as much as a jobs target would dictate, it may find a politically acceptable more expansive stance by adopting a price target.  Unfortunately, recovery is going to be way too slow regardless. 

Copyright Larry Greenberg 2010.  All rights reserved.  No secondary distribution without express permission.



Comments are closed.