Alternative indicators conceptually similar to NFP and private NFP continue to rise.
figure 1: Nonfarm payrolls from CES (blue), from CPS, adjusted to the NFP concept (tan), and total employment from the Quarterly Census of Employment and Wages, seasonally adjusted by the authors using X-13 (green), All in 000, sa Source: BLS via FRED, Bureau of Labor Statistics, U.S. Bureau of Labor Statistics (QCEW).
Note that the CES NFP trend in recent months matches the research series for CPS (tan line). While the QCEW-based series stabilizes in the second half of 2022, early baseline estimates from the Philadelphia Fed for the first quarter of 2023 point to continued growth.
figure 2: Private nonfarm payrolls from BLS CES (blue), ADP (tan), and total employment from the Quarterly Employment and Wage Census, seasonally adjusted by the authors using X-13 (green), all expressed in 000s , sa Sources: BLS, ADP via FRED, U.S. Bureau of Labor Statistics (QCEW).
There has been a great deal of debate as to whether early benchmarks at the Philadelphia Fed using QCEW data would have produced a different picture (Reader Steven Kopits lays out the whole argument for a recession in the first half of 2022 in this series). this latest reading Heralds continued growth in the first quarter of 2023.
source: Philadelphia Fed (March 16, 2023).
Why not pay more attention to the CPS civilian employment series? Because of the variability of this series relative to the CES NFP, as shown in Figure 3.
image 3: First log difference of NFP (blue) and civilian employment (tan). Excludes new population control months. Dates of peak-to-trough recessions as defined by the NBER are shaded in gray. Sources: BLS, NBER, and authors’ calculations.
The standard deviation is 0.0015 for NFP and 0.0023 for clerical employment.
While there is folk wisdom that civilian series are rejected before they are established, the evidence in real time is mixed (as in this article postal). Private employment fell in May, while NFP continued to rise. But civilian employment also fell in April, June and October of last year…
