I posted a Note about this, but perhaps it is the opportunity for taking another crack at what I think is a flawed discussion of macroeconomic policy. I think the flaw is of two kinds and it is hard to tell for any particular commentator which is which. The two kinds are a) a mistaken “model” of how events and policy produce outcomes and b) mistaken/overly simplified policy recommendations/opinion about policy based on trying only to be a bit better than the perceived status quo, rather than trying to actually get it “right.” Of course, I am judging from MY model of what is “right.”
Project Syndicate frames a question this way:
Is 2% Really the Right Inflation Target for Central Banks?
Jul 27, 2023
Both the US Federal Reserve and the European Central Bank appear to be dead set on getting inflation back to their 2% target. But while 2% is viewed as a kind of “sweet spot” for inflation – neither so high that consumers struggle to cope, nor so low that it stifles economic dynamism – it is ultimately arbitrary, and its primacy in monetary policymaking is a relatively recent phenomenon.
In this Big Question, we ask Michael J. Boskin, John Cochrane, Brigitte Granville, and Kenneth Rogoff whether it is time to rethink the 2% target.
I really know too little about ECB monetary policy, so I am going to take the answers given as about the US and the Fed.
Boskin says no because he thinks the 2% target will “discipline” fiscal policy. Reading between the lines he would not be too unhappy if the discipline results in recession and if the fear of a politically unpopular recession were the way the discipline works. Cynical me wonders if he would be as sanguine about recession if a Republican were in the White House. Did he worry about disciplining GWB’s massive deficit creating tax cuts or the Trump deficits that were larger than Biden’s?
Cochrane says yes but from a particular stance. He doe not like targeting “inflation” at all. He thinks the Fed should target the price level. It’s not clear exactly what that means. Does this really mean target an unchanging price level? It could mean setting a target of the price level in the future being x% higher or lower than at present and if that is it, what should be the trajectory of intermediate levels? Rising at 2% p.a or some other percent on average. How different would that be from inflation targeting?
Granville says yes because she is pretty sure that a 2% inflation target is not compatible with stable full employment growth and thinks that inflation could be as much as around 5% p.a. with no ill effects. I have serious doubts about both parts of that opinion, but in principle I agree that the target (if there is a target) should be set on the basis of minimizing both the cost of unemployment and the costs of inflation.
Rogoff says no because, while agreeing that 2% is arbitrary (and by implication may not minimize both the cost of unemployment and the costs of inflation), the uncertainty around a change would be worse than the possibly suboptimal status quo.
I most agree with Rogoff because he seems most explicit about the costs and benefits of a higher inflation target. And he appeals to my “ain’t obviously broke, dong fix it” principles. But if the Fed DOES tip us into recession and in response announced that henceforth it would be using a 3% target, would THAT cause unacceptable uncertainty? [BTW, Rogoff also rejects a supposed reason for higher target, that a higher target gives the Fed more room to reduce interest rates in a recession. In this he is wholly right. The Fed even with a 2% target has plenty of ammunition for fighting recessions and not letting actual inflations fall below target.]
In general, I find all four in failing even to gesture at finding out what the relative costs and benefits of higher unemployment and higher inflation target are. So I will, again.
In thinking about optimal inflation – optimal macroeconomic outcomes in general -- I’d start from a model of an economy subject to real events – shifts in demand, changes in fiscal policy, changes in import and export prices, whatever -- that require market participants to adjust quantities supplied and demanded. Changing the amounts supplied and demanded requires changes in relative prices to maintain/reestablish maximum real incomes, in particular to avoid unemployment of resources, especially unemployment of labor.
The point of monetary policy is to facilitate those changes in relative prices and hence the most efficient re-allocation of and full employment of resources. This concept of policy is consistent with the Fed’s Congressional mandate to achieve stable prices and maximum employment. If all relative prices adjusted equally smoothly, all the Fed policy would have to do is maintain the dollar price of one commodity, say “gold.”
Unfortunately, that’s not the way the economy works. Some prices are set for extended periods – leases and mortgage interest rates for multiple years and wages for seldom less than a year. And even if prices CAN be changed, more can be easily changed upward than downward. So, when things happen -- things are constantly happening – relative prices need to be in continual adjustment. If achieving these adjustments means more prices need to rise and few can fall, there will need to be a constant increase in the average price level, inflation.
The Fed needs to engineer the right amount, the optimal rate of inflation, optimal given a) the size and frequency of the initiating events (shocks) and b) the ease of adjustment (stickiness) of prices. And there are cost to getting it wrong. Too little inflation and some relative prices that need to fall cannot and unemployment of that resource occurs. Too many unemployed resources IS a recession. (Generally, this is thought of as wages and unemployment of labor, but recessions produce rental vacancies, too.) Too much inflation shows up differently with resources fully “employed” but not in their most efficient uses.
In this concept, the Fed should decide on a (optimal) target rate of inflation based on expected shocks and stickiness of prices. Shocks will vary sometimes being a little larger or more frequent and sometime less, so the Fed would not expect to have a totally uniform rate of inflation, but rather its target would be an average over time. But sometimes there could be a very large shock – millions of people stay home from or lose their jobs because of COVID, or switch from spending on restaurant meals to fixing up their homes. In these circumstances the Fed ought to “flexibly” allow for/engineer above target inflation. Hence, we have the Fed’s official philosophy of “Flexible Average Inflation Targeting” (FAIT).
Using this framework I’d recast the debate in Project Syndicate over the optimal inflation target, 2% or higher, as being over whether shocks are too large or prices too sticky to make 2%, going forward to achieve full employment, [Granville] although implicit could be different views of the costs of unemployment compared to the cost of “inefficient” full employment [Boskin and Granville in different directions].
My own view is that the Fed’s FAIT framework is the correct one [Cochrane rejects the framework] and that it exercised it flexibility well in allowing above-average target inflation in early 2021 but was too slow to start bringing inflation back down in late 2021 (the Effective Federal Funds Rate did not start to rise until March 2022). With inflation continuing to fall through July 2023 (inflation expectations that had been below target earlier in the year have returned to near target levels) without unemployment developing*, however, I see no evidence that 2% is too low as FAIT. [Rogoff]
*I was wrong with my earlier fears that the Fed had pushed the EFFR up to fast and too far that that we would by now see the beginnings of labor unemployment and recession.