CBO found some more potential workers?

Update on Feb 28: Dr. Levin has very patiently walked me through how to extract the data without eyeballing the chart.  See the tab for Chart 2-5 here if the subject interests you.

The revision that I attributed to this January vs last January was actually achieved mostly since August, which makes it seem a bit more abrupt still. CBO’s estimate of the potential labor force participation rate for 2017 rises 50 bps or — more importantly for “slack” analysis — 0.8% of it’s prior self.  I will leave my imprecise eyeballing presented below intact for posterity, because I am now such an honest blogger, and because it was not bad.

This make the Blanchflower/Levin measure of the employment gap higher, to the point where I would not quite say that the US economy is “indistinguishable” from being at full employment.  With almost-similar agnosticism, I would say it looks to be a bit below full employment.

I will document this in a follow-up post.  Obviously, people disagree about whether this is even important, and a debate about the relevance of the NAIRU (and presumably allied concepts) has recently heated up on the inter webs. I don’t have a dog in that fight, although maybe I would if I were informed enough to do so.

As I see it, we are close enough to full employment for fiscal stimulus to be unwarranted here, particularly given that the US has escaped liquidity trap, at least for now, which is the relevant horizon in this case. (Yes, this perspective is model dependent, just like yours.)  You don’t need to take a strong view on the slope or existence of the Phillips Curve to come to that conclusion, although it may be wrong on other grounds. More on this once I have crunched some numbers.


I no longer follow street research closely, and I would not be shocked if all of this is already in the broker reports.  This is something I have noticed today, which is different from saying it is news. The report I highlight here was produced in late January.


The top panel of the chart is from the January 2017 CBO Economic and Budget Outlook, while the bottom row is from the January 2016 version of the same report. Note that there is reduced variability in the estimate of the potential participation rate. Weirdly, they also reduce the granularity in the scale of the close-up, and shorten the window by two years. Not some dastardly trick. Just makes it hard to follow.

I wanted to weigh in the debate between Simon Wren-Lewis and the “heterodox” economists slagging the concept of the natural rate of unemployment or NAIRU.

Readers of this blog will know that I am not very interested in debates about “essences”, including the question of the whether the NAIRU even exists.  Of course it does not exist. The concept is emergent, like most constructs in macro.  The question is whether it is helpful.  My sense is that the help it provides is limited, but perhaps not zero.  With greater confidence I would say that the debate will not be resolved by enquiring into whether it exists.

In order to say something practical, I wanted to update the blanchflower-levin-labor-slack-24mar2015 measure of the employment gap, referenced in this post.  I had some minor/unimportant difficulty replicating the metric, which I will get to in a follow-up post. But today, I just want to point about that a big part of the B/L measure of employment gap is the deviation between the labor force participation rate and its estimated “potential” level, that is where it would be with the labor market fully renormalized in all regards.

The estimate is conventionally provided by CBO and — point to the skeptics! — they have recently decided that the potential participation rate is higher than they previously believed/projected.   For several years after the crisis (I am not sure exactly how many because the way CBO presents data drives me nuts) estimates of the normal or potential labor force participation rate were steadily lowered, along with potential GDP. Related, the post-crisis compression of de-trend employment looked more structural and less cyclical.

But in the last go around, CBO has apparently revised up their estimate/forecast of the potential participation rate for 2017 from about 62 3/4% to about 63 1/4%, just eyeballing from the charts above, which have been made more difficult to read. (Why is the scale now less granular than it was? Not to be conspiratorial, but that is frustrating.)

Anyhow, if I am eyeballing the chart correctly, then the potential labor force is now (i.e. for 2017) about 50 bps or — more relevantly for gap analysis — 0.8% higher than projected (for now) last year.

I have already calculated that this drives up the B/L estimate of the employment gap relative to what we previously believed it was.  Spoiler alert: basically the whole think looks like U6.  But dealing with these data in the presence of population control shifts and CBO’s irritating lack of footnoting gives me a headache, so I am going to hold off on the charts for now.

Anyhow, the labor market looks a bit “looser” than it did on the old estimates.  I guess that fits with what wages are doing.

I hope to have more on this later. And to repeat, this is not some huge breaking news. Just something I stumbled on to today while doing something else.