Some commentators have focused on the sharp declines in civilian employment as well as the ISM services employment diffusion index as signaling greater weakness than suggested by the establishment survey.
In particular, the fact that the BLS birth-death model is accounting for a large share of total growth is worrying (see Forbes, e.g.). In point of fact, the level of NFP has been revised downward for the last three months.
Figure 1: Nonfarm payroll employment, from various vintages. Source: BLS via FRED.
Tabulation of the numbers from the birth-death model are here.
The argument of overall weakness in the labor market is illustrated by the differing behavior of the NFP from the BLS establishment survey (blue line), where in this case I have incorporated the preliminary benchmark revision into the level of NFP. In contrast, civilian employment from household survey (green line), and the diffusion index component from the ISM PMI services index (tan line, left scale, below 0.5 is contraction), both took a big dive.
Figure 2: ISM PMI services employment diffusion index (tan, left scale, below 0.5 is contraction), NFP preliminary benchmark (blue, right scale), private NFP from ADP (red, right scale), and civilian employment from BLS household survey (green, right scale), all in logs 2023M03=0. Source: ISM via TradingEconomics.com, BLS and author’s calculations, ADP via FRED, BLS via FRED, and author’s calculations.
So, adding the PMI measure into the mix leads to a picture of a weaker labor market than otherwise. On the other hand, the ADP measure, which continued to rise in December, is not based on a survey, but on actual check-cutting data. The fact that it is rising in step with the BLS establishment survey provides additional support to the view that employment growth is continuing.
Further note that the Philly Fed’s early benchmark, which incorporates Quarterly Census of Employment and Wages data shows continued growth (in line with NFP) through June (which, admittedly, is 6 months ago).
Source link
