Some Thoughts on the US Labor Market

The eventful week concludes with the US employment report tomorrow. It comes on the heels of a series of economic reports that were weaker than the consensus expected and points to a loss of momentum in late Q1 that is carrying into Q2. 

Although the weekly initial jobless claims reported today was counter to this pattern, as they recorded new cyclical lows, during the survey week for the national figures, weekly jobless claims were a bit higher. The employment component in the manufacturing ISM was lower. The ADP estimate was also disappointing. The 119k private sector jobs compared with consensus expectations of 150k and the March estimate was cut by 27k to 131k (the BLS initial estimate was 95k for private sector jobs).

Since their methodological revisions there have been six reports before this week's. Three times the ADP estimate was above the BLS initial estimate for an average miss of about 45k. Three times the ADP estimate was weaker than the BLS figure. The average miss was about 34k. 

The Bloomberg consensus calls for a 155k net increase in private sector employment. This would represent an increase from the initially reported 95k for March, which of course is subject to revision.

Economists had generally under-estimated the strength of high frequency data in early Q1 and by the time they adjusted, the US economy began softening. This pattern warns of downside risks to tomorrow's jobs report. Yet, even if the number is in line with expectations, it represents a clear slowing. The 3-month average is 171k and the 6-month average is near 201k. 

Although the March jobs data disappointed, the fact that the work week increased by 6 minutes may have helped mitigate economic impact. The increase in hours worked, which matched the cyclical peak, coupled with the rise in temporary workers, was seen by some as suggestive of positive underling dynamics. Slippage in April aggravate any disappointment headline figures.

One of the big mysteries surrounding the US labor market has been the decline of the participation rate. Last year, the participation rate fell 0.4% to 63.7%. The decline in the participation of 25-54 year olds was only partially offset by the increased participation among workers 55 and older. There seems to be both cyclical and structural aspects. Market participants appreciate in a way they may not have before the crisis that the unemployment rate says more about the participation rate than job creation itself.

Even more broadly, there seem to be cyclical and structural issues with the US labor market. Consider, for example, that in recent years the US has gained manufacturing jobs but continues to bleed clerical jobs. This would seem to be a reflection not of off-shoring, but of technological advances.

There also seems to be some skill mismatch. Reports suggest there are some 3.5 mln job vacancies in the US, yet high unemployment. There also appears to be another shift taking place in the US labor market. Consider that in some 15% of taxi drivers have a college degree now compared with 1% in 1970, according to a recent essay by the FT's Edward Luce, who cited work by the Center for College Affordability and Productivity. A full quarter of sales clerks have a college degree compared with 5% in 1970 and 5% of janitors now have a college degree. And as Luce notes, almost half (46%) of those who go to college fail to complete the 4-year degree in six years.