E-News sign up

Where are the workers coming from: January – February?

Bill Wheaton | March 2019

The strong US January jobs growth figures were actually quite similar to many of the monthly reports over the last year. In fact, over the last year US job growth has averaged about 3m annually (250k monthly).  Over the same time frame, the number of unemployed-and-discouraged workers has been essentially stable, declining a bit during the first half of 2018 and now rising somewhat. The long 10 year downward trend in this series appears done. Note that I have used the broader definition of those out of work, one that includes worker’s not actively looking for employment. Hence the calculated growth in the labor force over the last year has roughly equaled the growth in jobs: 3m.      

These numbers are really very different from the average numbers during the longer 9 year recovery stretching from 2010 through 2018. Over this period annual job growth averaged a little less than 2m of which 650k was contributed from the average decline in unemployed-plus-discouraged workers. This left the yearly contribution from growth in the labor force at an average of 1.30m. This derived growth in the labor force is just a little below what demographers have been predicting based on births and aging – with the assumption of constant labor force participation. So what has happened in the last year that made the labor force grow more than 2x its trend over the last decade? Clearly, it involves changes in labor force participation wherein workers that were previously not even interested in working now for some reason are.

We can get some glimpse of what might have caused this increased interest in working by examining the change in labor force participation over the last decade. These data are from the CPS survey and so can be a little different from calculated data. We highlight the average annual percentage point change from 2010-2018 and then during the last year.  

Annual Change in LFP rate by age (percentage points):

Category ∆ LFP (9 year average) ∆ LFP (2018:1-2019:1)
Total -.17 .35
Male -.12 .0
Female -.11 .76
20-24 .0 1.35
65+ .46 1.42
Black .06 .44
Hispanic -.12 1.10

The first observation is that in total and for all groups, LFP has ticked up quite significantly over the last year relative to the gradual decline over the previous 9 years. Secondly, within the categories examined the longer term trends in LFP are actually quite small except for one – the pronounced increase in participation by those over 65 (without disability). Finally, over the last year, participation by Females, minorities, young and retirement age workers has all increased dramatically. Only for males (and especially white males) has it not. No doubt there are interesting stories behind each of the groups, but when taken collectively the data illustrate that the US is simply running out of workers. To keep job growth in the range that it has been over the last year, is going to require structural changes in the “interest” of people in working. Without this, a secular reduction in job growth seems inevitable.    

And now for February’s surprise. The .1% drop in unemployment should create 140k jobs, but with only 20k jobs actually created, the labor force must have contracted – from a reduced participation rate. While it is never wise to rely on monthly numbers, February perhaps is a warning that US labor force growth is near to exhaustion. With unemployment at its lower bound, expecting job growth beyond that which results from demography relies on some structural change – change that is not well understood and should not be counted on. With such a tight labor market, perhaps wages will finally begin to recover more significantly (as they just did in the February data)!