This tweet raises an interesting question:
I believe that it’s a big mistake to use monetary policy to target employment (as we saw in the 1960s and 1970s). The counterargument is that Congress gave the Fed a mandate that included “maximum employment” and thus the Fed is sort of forced to set a target for maximum employment. But that Fed mandate does not force the Fed to precisely define maximum employment for the simple reason that the Fed cannot precisely define maximum employment, nor can anyone else. No one knows how many jobs can be created through monetary stimulus without threatening the other part of the dual mandate, which is price stability (defined as 2% inflation, but that’s another issue.)
That doesn’t mean the Fed should ignore the maximum employment mandate, rather the implication is that the Fed needs a model as to how best to address the dual mandate. And one basic model that the Fed has been working with in recent decades is the Natural Rate Hypothesis, which is the idea that you cannot permanently push employment above its natural rate with expansionary monetary and fiscal policies.
Philosophers tell us that the best way to be happy is to directly aim for some goal other than happiness—perhaps devotion to a cause. Similarly, the best way to promote maximum employment is to aim for some other policy target, such as steady growth in NGDP at a rate that is broadly consistent with expectations of the public. Whatever level of employment results from that policy (in the long run) can be inferred to be the “natural rate of employment”. That doesn’t mean that other policy reforms cannot boost employment further, just that you cannot use monetary policy to produce consistently higher employment than what you’d get with a steady rate of NGDP growth of say 4% or 5%.
In my view, the Fed will define “maximum employment” as the actual rate of employment when two things have occurred:
1. NGDP growth has been pretty stable for an extended period.
2. The actual unemployment rate has stopped falling, and has leveled off for a period of years.
As far as I know, that’s never happened in all of US history (but did happen in the UK from 2001-07). If not for Covid, I believe this would have happened here in the early 2020s. Now I wonder if I’ll live long enough to see a “soft landing”, i.e. whether I’ll live long enough to see a sustained period of steady and low unemployment and steady NGDP growth.
PS. Even if the Fed does eventually figure out the definition of maximum employment, it will be a pyrrhic victory, as this elusive figure continually changes over time.
READER COMMENTS
marcus nunes
Aug 30 2021 at 7:52am
“In my view, the Fed will define “maximum employment” as the actual rate of employment when two things have occurred:
1. NGDP growth has been pretty stable for an extended period.
2. The actual unemployment rate has stopped falling, and has leveled off for a period of years.”
Exactly!
https://marcusnunes.substack.com/p/kernels-of-truth
Thomas Lee Hutcheson
Aug 30 2021 at 8:00am
I think that the Fed just believes that it must dial back QE only gradually to fix markets belief that it really does have an average inflation target rather than an inflation ceiling. TIP have been in the vicinity of 2.3% expectations for a relatively short time (10 year TIPS was 2.39 on Friday with 2.30 being the approximate equivalent of PCE 2.00).
Lizard Man
Aug 30 2021 at 10:32am
What is the evidence for the natural rate hypothesis? I thought that it was the case that unemployment causes unemployment, in part because people who are out of work for long periods of time become increasingly unlikely to return to work, go on disability, take early retirement, etc. Isn’t that what happened in during the Great Recession?
I would assume that the natural rate hypothesis is inconsistent with the idea of multiple equilibria.
I think that both of these points of view would support NGDP targeting, because it leads to less disruptive changes in monetary policy, and would allow for the Fed not to need a crystal ball to predict how employment and inflation will interact in the future.
Scott Sumner
Aug 30 2021 at 1:58pm
The unemployment rate did fall back to 3.5%. Total employment is harder to model, but even if people take early retirement, the effects on employment are transitory.
Of course as with any macroeconomic model it’s just an approximation of reality.
Rajat
Aug 30 2021 at 5:28pm
Late to this, but obviously great minds think alike, Scott… I recently read the passage in Ed Nelson’s book describing Friedman’s view of the natural rate of unemployment as expressed in his 1967 AEA presidential address (pp.284-286):
Nelson notes two key features of this definition: (1) it corresponds to an equilibrium or flexible price outcome, and hence does not correspond to some feasible maximum level of employment attributable to sticky prices; and (2) it reflects prevailing distortions to market forces, such as market power, minimum wages and other labor, product and services market regulations.
Nelson goes on to discuss Friedman’s view of the natural level of output (or alternatively, ‘potential output’) as a logical corollary of his understanding of the natural rate of employment, but notes Friedman’s distaste for focussing directly on real variables (pp.288-289):
Rajat
Aug 30 2021 at 5:37pm
I was going to add, your prediction of how the Fed will define ‘maximum employment’ looks pretty consistent with Friedman’s view of the natural rate to me – stable NGDP growth should in time produce an employment outcome similar to what a flexible price model would produce, and one that reflects prevailing distortions in the economy. But yes, even for no other reason than that the structural features of economies are in constant flux (ie even assuming NGDP growth is perfectly stable), the natural rate of unemployment will be too.
Scott Sumner
Aug 30 2021 at 7:18pm
Yes, I was very much influenced by Friedman’s views on this issue.
And Nelson’s book is great.
Scott Sumner
Aug 31 2021 at 12:49pm
Sven, I disagree. Congress never said the Fed needed to target unemployment. They said high employment is a policy goal, which seems reasonable to me. The Fed should not target employment or unemployment. They should target NGDP in the hope that it leads to high employment.
Price stability is a bad policy goal because we have no reliable way of measuring the price level.
Sven
Aug 31 2021 at 8:09am
Fed must target price stability. Period! Unemployment is not FED’s job. Congress must change this law that says FED is responsible for unemployment.
Scott Sumner
Aug 31 2021 at 12:49pm
Sven, I disagree. Congress never said the Fed needed to target unemployment. They said high employment is a policy goal, which seems reasonable to me. The Fed should not target employment or unemployment. They should target NGDP in the hope that it leads to high employment.
Price stability is a bad policy goal because we have no reliable way of measuring the price level.
Sven
Aug 31 2021 at 4:20pm
Either it is a specific unemployment rate or high employment, I think it is a policy choice beyond price stability.
If NGDP=output growth+inflation, is there a difference between price stability target and NGDP?
Assuming no reliable way of measuring the price level, do you think there is a reliable measure for maximum NGDP? Is it a specific rate of growth or natural unemployment rate?
Floccina
Aug 31 2021 at 2:57pm
A little OT but with a 2% inflation target should we and when should we exchange dollars for new dollars at 10 old dollars for 1 new dollar so the penny is more than an annoyance?
Michael Sandifer
Aug 31 2021 at 10:20pm
Of course the Fed shouldn’t target unemployment, because it can temporarily go below the natural rate as inflation picks up. I’m more optimistic than most that it’s possible to estimate the natural rate at most times, though it does fluctuate, but policy based on those estimates would not be as precise as level targeting NGDP. It would be too risky. Unemployment can be one variable, among many however, that can help determine what the stance of monetary policy should be.
Comments are closed.