What does a continually improving labor market imply about monetary policy?

Not as much as many people think.  As you may know, yesterday’s job market was quite good (NYT) and it is now many years of labor market recovery.  Does this mean the Fed should have been easier with money say two or three years ago?  No, that doesn’t follow.  Let’s talk through a few points:

1. Aggregate demand shocks are major causes of recessions, and when they come you want the central bank to lean against them very, very hard, even if that means higher than average price inflation.

2. The early problems are mostly nominal.  For whatever reasons (morale? long-term implicit contracts?), firms lay off some workers rather than cutting their nominal wages.  This is a big reason why downturns involve so much unemployment, but to the extent the central bank can keep up nominal demand, at least some of this unemployment can be avoided or at least smoothed out over time.

3. Once those workers are unemployed, nominal stickiness starts to cease to be the major problem.  Real rigidities and stickiness become progressively more important with the passage of time.  First, the unemployed don’t even have a nominal wage to be sticky in the first place, and yes some of them are excessively stubborn with their reservation wages for accepting new jobs but that looks suspiciously like voluntary unemployment, albeit with some behavioral irrationality mixed in.  But no, the main problem still is not voluntary unemployment, I am saying if the only rigidities were nominal it would be a problem of voluntary unemployment, a very different claim.  I’ll come back to this shortly.

4. In the very early stages of a recession, there might simply be no jobs available, period, due to uncertainties and liquidity shocks.  But usually within a year or two, a whole host of jobs open up, they just may not be good jobs for many of the unemployed.  Often they are bad jobs, for reasons which relate to real rigidities, not nominal rigidities.  Individuals need to be rematched to new jobs, and that process may or may not go well.

5. Here is a typical real rigidities story (but not the only one): you aspire to be an upper to mid level manager, and you are offered a job as a cashier at Walmart, or you could get such a job if you tried.  You don’t take the job, because you fear its presence on your resume will shunt you onto a permanently lower career track.  That is indeed a problem, and it is a real rigidity, not a nominal one.  It can keep you unemployed, even if you might otherwise prefer to have the work on a temporary basis.

6. As the economy grows in real terms, the quality of jobs available will be upgraded, and eventually the unemployed are offered jobs which are worth taking.  This process can be fast or slow, but in the recent recovery it has been relatively slow.

7. As more people are taking jobs, yes demand is rising but supply is rising also.  The two are rising together.  It is not wrong to say “greater demand is reemploying people,” but it is misleading.  It is more accurate to say “real demand is rising, coincident with growing output, job quality is improving, uncertainty is being resolved, and the economy is doing a successively better job at solving the matching problem in its labor markets.”

8. In that same setting, simply boosting nominal demand with lower interest rates and higher price inflation won’t necessarily help employment much.  You might get a slight labor supply increase through the not very impressive Lucas supply curve (people confusing nominal and real changes).

8b. To be sure, there is likely no harm from the easier monetary policy since price inflation has been slightly below target.  I do not hate these proposed monetary easings, I just don’t think they are likely to matter much.  Telling me that “higher demand has been reemploying workers” doesn’t impress me.  Telling me “the labor market recovery has been underway for a long time” also does not impress me.  Neither claim implies that a nominal push to demand will do the trick, if anything they might imply the contrary.

9. If the labor market recovery has been underway for a long time, that is a sign the coordination problem has been a real one rather than a nominal one.  It is a sign that earlier Fed easing simply may not have mattered much.

10. You might notice that outside of emergency situations, such as 1929 or 2008 (see point #1), economists struggle mightily to demonstrate that money matters at all.  I think Christina Romer has shown that surprise deflationary shocks do matter and are bad.  After that, it is still up in the air, as indeed this analysis implies.

Everything I am writing is consistent with mainstream research economics, and the best and most sophisticated versions of Keynesian economics.  I know it is not usually what you read on the internet.  As a side note and exercise, it is worth pondering what this same framework might imply for the efficacy of fiscal policy, noting the answer will be “under what conditions?” rather than yes or no.