During the period from 2021 to 2023, inflation was far higher than the Federal Reserve would have wished, and also far higher than forecast by the markets. Does that mean we can excuse the Fed for allowing inflation to overshoot its target by a large amount?  The answer is no.  I’ll try to explain why using an example of how things would look under both inflation targeting and price level targeting.  We will assume that the Fed’s inflation target is 2%.

Let’s assume that the price level is 100 in March 2021.  The Fed would like prices to rise by 2% per year, or 0.5% per quarter (three months.)  Here’s how they would like to see the price level rise each quarter over two years (for simplicity, I am ignoring compounding effects):

Case A:  100, 100.5, 101, 101.5, 102, 102.5, 103, 103.5, 104

Now assume that for 8 consecutive quarters, the Fed underestimated quarterly inflation by 1%.  They expected 0.5%, and got 1.5%.  Also assume that the Fed was doing inflation targeting, letting “bygones be bygones”:

Case B:  100, 101.5, 103, 104.5, 106, 107.5, 109, 110.5, 112 

Over two years (8 quarters) the price level rose by a total of 12%, much more than the 4% rise desired by the Fed.  Annual inflation was 6%, much higher than the target of 2%.

Now suppose that for 8 consecutive quarters, the Fed underestimated inflation by 1% per quarter.  But now assume that the Fed was doing price level targeting, rather than inflation targeting.  That means that at each and every point in time, the Fed was trying to achieve the price level path shown above in Case A:

Case C:  100, 101.5, 102, 102.5, 103, 103.5, 104, 104.5, 105

Notice that even though the Fed made exactly the same size errors in cases B and C, during every single quarter, the price level path in Case C is far closer to the ideal path shown in Case A.  Under price level targeting, you have an extra 1% inflation in the first period, but after that time the inflation rate is 0.5% per quarter, or 2%/year.  As a result, in Case C the inflation rate averages 2.5%/year between March 2021 and March 2023, not 6% as in Case B. 

In real life, there was roughly an extra 8% worth of inflation in the two years after March 2021.  This occurred even though under “average inflation targeting” the price level path should have been much closer to Case C than Case B.  In other words, the Fed did not adopt the policy regime that it advertised to the public; it had no intention of targeting the average inflation rate. 

Were the Covid supply problems and the Ukraine War a valid excuse?  Not at all.  NGDP growth overshoot the 4% growth path by even more than inflation overshot the 2% trend line.  Policy was far too expansionary under any reasonable criterion.  Nor can you blame the mistake on the fact that even the markets missed the size of the inflation surge.  Under either level targeting, or a true “average inflation targeting” regime, those missed market forecasts would have only caused a small overshoot, the sort we see in Case C.

PS.  I started the clock at March 2021, as by this time the price level had recovered from the initial drop during the early stages of Covid, and was back on trend.