U.S. Employment Report Even Worse Behind The Headline
It turned out to be just as weak as I expected, indeed slightly weaker.
First of, the household survey showed an even bigger decline in jobs than the payroll survey, 785,000 versus 263,000. While the household survey is less reliable than the payroll survey with regard to monthly fluctuations, the fact that it for the second month in a row showed much weaker numbers suggest that the payroll survey probably underestimates the drop in jobs at least somewhat.
Related to the previous point is that the unemployment rate would have risen a lot more than just 0.1 percentage points if it hadn't been for a big drop in the participation rate. Had the participation rate stayed unchanged, the unemployment rate would have risen above 10%.
Secondly, average hourly earnings increased less than 0.1% on the month and just 2.5%. Combined with a drop in the average work week, this means that average weekly earnings dropped by 0.25%. Combined with a drop in the number of employed this suggests a big drop in aggregate labor income, a coincident indicator.
And the drop in the average work week is also significant because the average work week is considered to be a leading indicator. The slight upswing in that indicator in July was hailed by bulls as a sign of a future recovery, but now that upswing has been reversed.