Inflation Targeting - Again

2023 10 24

https://www.ft.com/content/38ce1a9c-89cc-4157-992a-f0d4cdd95440?emailId=0f6d959b-b9be-4aa7-88fc-cec14b9f2616&segmentId=7d033110-c776-45bf-e9f2-7c3a03d2dd26

Clearly the author failed to read my earlier post 😊

Optimal Inflation Targeting - by Thomas L. Hutcheson (substack.com)

on this topic and so makes the same mistakes I was trying to correct in that discussion.  So let me try again, being much more didactic this time. 

“That 2% inflation target may not be sacred for much longer” reflects a common but very confused way to thinking about what the best inflation target should be.  Rather than take up specific points of confusion, let’s step back and ask why there should be and inflation target and for that we have to step back again to ask why we need monetary policy at all. [Hint: To allow/engineer optimal inflation.]

Let’s start in a world of flexible prices and, say, a fixed number of cowry shells serving as a medium of exchange.  Such an economy would not need a monetary policy beyond prohibition of counterfeiting or altering the cowry shells.  Economic events – failed harvests, increased or decreased demand by foreigners for whatever the economy export, new machines to turnout more of some things by using less of other things -- could cause relative prices of things to change and make some people richer or poorer, but this would have no significance for the number of cowry shells that were exchanged.  If over time the number of things to be exchanged grew, the ratio, between the number of things exchanged and the number of cowy shells, would grow.

If this economy was truly lucky there would be no way to divert resources from producing exchanged items to increasing the number of cowry shells.  If in addition to being lucky the economy also had a ruler (let’s call the ruler a “Fed”) the Fed could save everybody the trouble of lugging around cowry shells by calling in all the cowry shells and replacing them one for one with pieces of paper called “the Cowry” and stamped “1 Cowry” or “10 Cowries” or “10,000 Cowries.”  If in the process the Fed replaced the cowry shells with too many or too few pieces of paper the total of whose stamped sums were greater or less than the number of cowry shells, this would change the number of items exchanged against each paper Cowry but not the relative prices of the exchanged items.  [Now if the Fed ALSO used some of its new Cowrys to build a new headquarters or go to war with a neighboring Fed, THAT would change relative prices and shift resources from everyone else in the economy to the Fed.  This would be the equivalent of Medieval kings debasing the coinage and was a form of “monetary policy” uniformly condemned by Medieval economists.]

Leaving behind this monetary quasi utopia, let’s add an additional element, relative prices are not perfectly flexible.  There are some things that have a cowry price that can rise but cannot fall.  Now suppose one of those economic events happens, say a particularly good harvest, that sets off a chain of events that increases AND decreases supplies and demands for everything else.  That will mean relative prices need to change, some to rise and others to fall.  Oops!  There are some items whose prices cannot fall, so some of those items cannot be bought in the quantities that sellers wish to sell – unemployment of the fixed price items.  Paradoxically, this will mean that even a “good” economic event can have worse consequences in this economy that it would have had in the flexible price economy.

Enter the benevolent ruler.  Suppose this Fed can (say by increasing the number of Cowries or somehow making them easier to use) engineer increases in the Cowry price of the flexible price items and fall in the relative prices of the fixed-price items, mimicking the relative price effects of the good harvest in the flexible price economy and avoiding unemployment of the fixed price items.  But it did so by increasing the average price of all items – inflation. 

Of course, our Fed is only benevolent, not omniscient, so it might engineer some inflation, but not enough to reduce the relative prices of the fixed price items sufficiently and some unemployment of those items would still result.  On the other hand, it might engineer more than enough inflation; relative prices would fully adjust but the Cowry price of even the (downwardly) fixed-price item would rise. 

Note that bad news, a crop failure, is inflationary, too.  Perhaps the Fed does not need to do anything to the number of Cowries for relative prices to adjust, but there is less stuff being exchanged for the same number of Cowries; the Fed allows inflation even if it does not need to engineer it.  And as before the Fed could allow too much or too little inflation with the same consequences for unemployment. 

And as I explained in “Rise and Fall (?) of the Phillips Curve:

(100) The Rise and Fall(?) of the Phillips Curve (substack.com)

repeated instances of too much and too little inflation when plotted on a graph will show an inverse relation: too much inflation little unemployment, too little inflation more unemployment.

Now its time to add or make explicit a further element.  Events that require changes in relative prices happen all the time and in staggering numbers.  Does this just mean the Fed should give up trying to engineer or allow enough inflation to facilitate adjustment in relative prices and keep unemployment low?  (What would “giving up” even mean?)  No, if the Fed can predict (seat of the pants, sophisticated model) on average the rate of events (shocks) the economy will experience and what rate of average increase in the price level is needed to facilitate adjustments in relative prices, it will attempt to engineer that average amount of inflation.  In other words, it has chosen an average inflation target.

And (wait for it) if there are extraordinary, greater than average shocks that require extraordinary adjustments in relative prices, the Fed will engineer above-target inflation temporarily: “Flexible Average Inflation Targeting.”

Only now can we make sense of the assertion “That 2% inflation target may not be sacred for much longer.”  The issue being raised in a confusing way is whether perhaps the average level of shocks has increased from the level that (presumably) led to the adoption of 2% as the FAIT.  Raised, but without any conceptual framework, like the one I try to provide here, for addressing it.