In an unfortunately not too well edited transcript, John Hartley interviews David Beckworth of Mercatus Center on “Monetary Policy and Central Bank Targets”
Beckworth makes clear that the idea of _targeting_ – adjusting policy instruments so as to affect something that people really care about, something that is not itself a policy instrument 😊, makes a lot of sense. Obviously NGDP targeting is better than “interest rate” targeting or “money supply” targeting. They are instruments.
Where I always get hung up, though, is why say “NDGP” targeting and not “inflation” targeting. After all whatever a central bank does to affect NGDP is mainly just what affects the inflation part of NGDP, not the real part. And there is not really any conceptual doubt about what real GDP part should be: as much as full employment of all resources permit. Beckworth even agrees that that in the long run the two targets would look the same. So why not inflation targeting?
Later in the interview Beckworth mentions the superiority of NGDP targeting to other possible policy rules (such as the Taylor rule), convincingly enough for me. He also answers well enough the objection to NGDP targeting that it is published with a lag and initial estimates are subject to large revisions. “Nowcasting” overcome that problem. Beckworth makes that case that NGDP is _feasible, but Hartley never asks Beckworth to take on inflation targeting head-to-head. Why does Beckworth think NGDP Targeting better than Flexible Average Inflation Targeting (FAIT) which is, after all the status quo?
Hartley does ask about FAIT in the context of NGDP targeting not being an option, but the question seems to take for granted that the “A” in FAIT is backward looking over some period (but in that case in what sense is it “flexible?”)
Beckworth has a wandering reply that ends recommending a price level target which implies temporarily below average inflation when inflation has been above target and viceversa. This would be “flexible,” but what is the economic rationale for below target inflation if the target had been chosen as the lowest inflation necessary to introduce sufficient flexibility in relative prices to maximize resource employment?
But again, this is only an aspect of not deriving the target – NDGP, PL, or FAIT – from an optimizing model. What ultimate objective is monetary policy supposed to be optimizing? What is the loss function from missing the objective? Which target does the best job of optimizing given exogenous variables like resource constraints, more or less upward or downward price flexibility across sectors, the size and frequency of positive and negative supply and demand shocks, and the Fed’s ability to gather and process data?
Image Prompt: One man interviewing another (microphone on table) with dollar signs and questions marks floating around
[Standard bleg: Although my style is know-it-all-ism, I do sometime entertain the thought that, here and there, I might be mistaken on some minor detail. I would, therefore, welcome comments on these views.]
"After all whatever a central bank does to affect NGDP is mainly just what affects the inflation part of NGDP, not the real part. "
Say what? Interest rate policy *only* affects inflation via the real economy. The assumption underlying inflation targeting is that real GDP returns rapidly to the long-run growth path, but that isn't true
NGDP, FAIT? I studied economics but never heard of these acronyms