Several figures emphasize the fact that the number of non-agricultural employment in November should be understood in the context of continuous revision of the data (see press release) here).
figure 1: The non-agricultural employment numbers in November (black), October (red), September (blue-green), August (pink) and the Bloomberg consensus 12/1 (lilac square) are all on a logarithmic scale. Source: BLS, various versions, Bloomberg.
Therefore, overall, in the summer and recent editions, the revisions have been moving upwards.this Average absolute correction of employment change From January to September this year, the first release to the third release is 107K. This means (using 1.96) that, taking into account the average absolute correction shown in the first nine months of the year, the implied employment change for the third release may be zero or may be as high as 420K. (Of course, there will be a benchmark revision in the new year).
The somewhat confusing result is different paths for different employment measures. Household surveys show that civilian employment has risen sharply.
figure 2: Non-agricultural employment (black), Bloomberg’s total working hours index for non-agricultural employment (lilac square), civilian employment (over 16 years old) (blue-green), non-supervisory personnel, production workers (red), all in the log It means the same, 2021M01=0. Source: BLS and author’s calculations.
The annual rate of civilian employment increased by 8.8%, while non-agricultural employment increased by only 1.7% (the total working hours surveyed by the agency also increased slightly faster than the NFP, which was 2%). The average wage of non-agricultural and non-supervised/production workers rose by 5.5% (annualized), which was basically the same as in October. This fact further shows that the intensity of the labor market is slightly higher than that indicated by the NFP. Whether this will still be the case in December is definitely an open question.
Furman and Powell Publish more information, and so does Bill McBride Calculated risk.
