Marcus Nunes has a nice post The Stabilizing Properties of a Stable Rate of NGDP Growth (along a stable level path) analyzing unhappy outcomes for inflation and real output when the rate of increase in nominal GDP departs from trend. He concludes that the Fed in its next target-setting strategy should abandon its current Flexible Average Inflation Targeting (FAIT) and adopt a Nominal GDP rate of increase target in order to achieve its Congressional mandate of stable prices and maximum employment.
I agree with Nunes analysis, but not his conclusion.
Maximum real income (I do not think Congress intended to require more employment of labor than maximizes real income) is an obvious objective. What is the theoretical justification for a central bank to create any inflation at all? If there are some circumstances that justify inflation, circumstances that make some inflation compatible with maximum real income, is it not plausible that said circumstances would change from time to time so that the optimal rate of inflation would sometimes be higher than at other times?
I think there are such a circumstance, normal variations in supply and demand of the myriad of goods, services, and factors of production in a dynamic, growing economy. And to maintain maximum (not constant) real income requires relative prices among those goods, services, and factors of production to adjust constantly to those changes in supply and demand. But if some prices by custom, or by contract, or by law cannot adjust downward (are “sticky”) other prices must rise. If some prices are to remain constant and others are to rise, the average price level must rise.
Thus, a central bank with a mandate to maximize real income and stable prices will attempt to create enough inflation to allow the average price level to rise enough to facilitate adjustment of relative prices. If the central bank assumes that relative the changes in supply and demand are randomly distributed (a Brownian motion) and the degree of stickiness does not change, the central bank will try to create an average rate of inflation that maximally facilitates the changes in relative prices and hence maximizes real income.
If the central bank does not create enough inflation to allow relative pries to adjust, some markets will not clear, unemployment of factors of production – unrented real estate, idle machinery, unused software, or unemployed people -- will develop. At scale this is a recession.
Becuse there are also prices that by contract or law cannot rise, the optimal rate of inflation is the minimum rate that facilitates the maximal relative price adjustment. This is, I think the logic of average inflation targeting that is common among central banks.
There can be, however, extraordinarily large variations in supply and demand affecting some sectors more than others, economic “shocks” that raise or lower either supply or demand -- positive or negative demand shocks, positive or negative supply shocks. An extraordinary shock of whatever kind requires an extraordinarily large change in relative prices to maintain full employment of resources and by the same logic that implies that average variations in supply and demand require an average rate of inflation, and extraordinary shock means the central bank should temporarily create extraordinary, over-the-average-target target inflation.
This to me sounds like a Flexible Average Inflation Target (FAIT).
Of course, corresponding to any FAIT there is a _Flexible_ NGDP rate target, so one could equivalently say that the Fed should pursue either FAIT or FNDGP rate and, shocks aside, FAIT and FNGSP rate targeting woud look the same. But at any given decision point the Fed will have more control over inflation than real GDP and what it will in fact be doing, it seems to me, is trying to increase, decrease or hold inflation constant, targeting inflation, not NGDP. Consequently, I think the Fed should put the label on the can that describes its contents.
[Standard bleg: Although my style is know-it-all-ism, I am aware that I could be mistaken or overstate my points. I would, therefore, welcome comments on these views.]
Image prompt: An astronomer holding telescope looking up at two moons, one labeled “NGDP” and one labeled “Inflation.”
The actual or “administered” prices (oligopoly, monopsony, and monopoly elements) would not be the “asked” prices, were they not “validated” by M*Vt (money X’s velocity), i.e., “validated” by the world’s Central Banks.