Languishing Labor Force

Is there a data series more highly anticipated and thoroughly parsed than the monthly nonfarm payroll release by the BLS? Probably not. But we’ll throw a few graphs onto the pile anyway.

The headline Establishment Survey number came in at negative 95,000 jobs for the month of September. The “new” headline number that everyone watches, private jobs, grew by 64,000. This number became more popular during the Census hiring distortions, which are gradually coming to an end. However, it’s tough to get excited about private sector hiring when the government is more than offsetting its gains. There were 77,000 Census jobs shed during the month, which were temporary to begin with, but this was actually eclipsed by losses in state and local governments, which cut 83,000 jobs. Presumably, these jobs were permanent. The ex-Census job losses come to a negative 18,000 jobs.

The number of people unemployed for 27 weeks or longer fell for the 4^th straight month and is now at the same level at which it began the year. However, the number of people newly unemployed (for 5 weeks or less) seems to have stopped its decline and has been on an upward trajectory for most of the year. Much like initial jobless claims, this number is much lower than its previous peak, but still above levels considered “normal.”

The U-6 unemployment rate, a more comprehensive view of labor force utilization, ticked up substantially to 17.1% from 16.7%. The increase was due to marginally attached and discouraged workers, which both increased in the month, and a record number of people working part-time due to economic reasons. As a percentage of the labor force, this group of underemployed workers has nearly touched its 1982 peak:

The headline U-3 unemployment rate stayed flat during the month at 9.6%. The internals of this rate were mixed, with the number of employed people rising (a positive) and the number of unemployed people falling (a positive). However, the number of people no longer in the labor force increased by 175,000, which is worrisome. The labor force managed to eek out a gain of 48,000 (thanks to population growth), but it’s the recent trend that worries us the most. The labor force has stagnated, and for the first time since 1952, it’s below where it was 2 years ago.

This bears watching very closely, because a stagnant labor force is a headwind to economic growth, and a declining labor force would be terrible. Lower labor force growth yields lower potential GDP growth, plain and simple. “To sharpen the point, observe that real GDP is, by definition, the product of the total hours of work in the economy times the amount of output produced per hour” according to William Baumol and Alan Blinder in their textbook Economics: Principles and Policy.

For a real-world/real-time example of labor force stagnation and its effects, look to Japan. Their labor force growth peaked in the early 1990s, while the number of workers peaked in the late 1990s and has been declining ever since.

The labor force decline in Japan presaged a decline in the total population by roughly 10 years. It is taken as a given that the US has better demographics than Japan, and we do. But the total population matters less to GDP growth than does the labor force, and our labor force participation rate has fallen from a high above 67% in 2000 to 64.7% now. Japan’s participation rate was around 64% in 1997, and declined to just below 60% in 2010. A return to the pre-boomer participation rates of the 1950s would take us below 60% as well. Back of the envelope math: assuming population growth remains at current levels, a decline to a 59% participation rate that took 20 years to occur would give us a total of 5.9% labor force growth during that period. That’s only 0.29% a year, well below the 1.52% annually that the labor force has grown since 1949.

Let’s hope that an eventual job market turnaround includes increasing labor force participation rates. Either that or another 1990s-style “productivity miracle.”