An aside on the participation rate

JW Mason has revived a debate about the size of the US output gap and the role that depressed labor force participation might play in it. See here for Mason’s paper along with his discussion of its reception.  And then see here for a somewhat testy exchange over the issue.

Just by way of a very brief history, the consensus initially held that the lower participation rate after 2007 was mainly cyclical.  Then some orthodox types at the Fed etc. convinced us that the decline was largely demographic and therefore structural. And most recently, the pendulum has begun to swing back again, in the sense that the doves have become louder. I am not sure if the more orthodox types have really conceded.

Screen Shot 2017-08-21 at 8.26.22 AM

(There was also a side debate about whether the high unemployment rate itself was structural or cyclical. The structural guys in that debate were mostly an embarrassment to the profession, so the less said about that motivated reasoning the better.  That “debate” is over.)

What I find odd about this discussion is the confidence with which many of its participants express themselves.  There is a good chance that one of these highly-confident analysts will end up having been right, but it is hard for me to know which one.  And I doubt I am alone.

A few years ago, when the orthodox types were ascendant, I remember seeing some analysis “showing” that the decline of the participation rate was almost entirely structural, with a big contribution coming from demographics.

One element of that work – sorry, no link, have forgotten – that really struck me at the time was how the economist(s) put trend-lines on cohort-specific participation rates and just breezily assumed that those trend lines were structural.

But trends change, and unexpectedly if you don’t know their source.  I would be willing to bet you a pizza that JW Mason saw that work and just rolled his eyes.  Maybe he can give you the link.

The strong-form of this mistake is to call changes in cohort-specific behaviors “demographics”, on the grounds that identifying them requires looking at cohorts. The great thing about that take is that it makes literally everything demographics. Yeah, if you define demographics as everything, then that sure was demographics!

If I recall correctly, the paper I am having trouble recalling was not doing quite that.  It separated demographics from structural forces applying to various cohorts, presumably independent of the size of those cohorts.  But those trend lines were still pretty brazen.

On the flip, side it is supposed now to be obvious that the labor market is not at full employment until the employment population ratio returns at least to 62%. Do we really know that?

A couple years ago, Blanchflower and Levin presented (not anew) a comprehensive measure of labor market slack that took into account not only the deviation of U3 from its estimated natural level, but also the participation rate shortfall and the number of people working part time for economic reasons. When they produced the metric, it gave a dove signal.

Screen Shot 2017-08-21 at 8.26.36 AM

I have roughly simulated it through today and find that it now shows that the labor market is at full employment.  Probably that is wrong. No estimate is precisely correct  — and who knows if the concept of labor market “slack” even necessarily refers to anything.

But I was struck by Blanchflower insisting that he now realizes he needs to update that metric, presumably to show some more slack.  I am guessing — guessing only — that Blanchflower may no longer be so confident outsourcing estimates of the natural participation rate to CBO, particularly now that the CBO-estimated structural rate has declined to where the part rate actually is.

Not really knowing what it is doing, and realizing that it does not know what it is doing, the Fed is probably going to “probe” for an inflation trigger.  They still have some faith in the concept of full employment (and its effects) and they suspect we are probably somewhere near it. So they would really prefer to avoid having demand growth accelerate here.  This is why tightening is even on the agenda.

But they seem to act like they are resigned to seeing the labor market tighten further, if that is the right word, and I would guess they will remain so until they see some effect on the price side.