this Employment status release January 2024 incorporates annual baseline revisions for the establishment survey series and reports on population controls for the household survey series. NFP is +353, while the Bloomberg consensus is +187K. (This time last year, NFP surprised again, at +517,000, beating Bloomberg’s estimate of +115,000.)
This is based on nonfarm payrolls based on a number of indicators.
figure 1: Nonfarm payrolls in January 2024, including baseline revisions (bold red), NFP in December 2023 (light red), and implied baseline revisions calculated by the authors using preliminary March 2023 baselines (blue ), January changes in the author’s preliminary benchmark and Bloomberg consensus implied levels (light blue squares), the Philadelphia Fed early benchmark (light green), and the civilian employment household survey series adjusted to the reported NFP concept (pink ), and the January 2024 adjusted household series that removes the impact of revised population controls (purple squares). The adjustment to remove the revised population control assumes that overall civilian employment increases by 0.957 of the total control for all civilian employment. Source: BLS (released in January 2024, December 2023), Author's calculations, Bureau of Labor Statistics.
To my surprise, the series after the benchmark revision far outperformed both the preliminary benchmark revision I calculated (blue line) and the earlier benchmark calculated by the Philadelphia Fed (light green line).
The CPS employment series (research series, unofficial) adjusted for the NFP concept shows a sharp decline, reflecting the decline in the civilian employment series. A similar decline was seen a year ago as new population control measures were implemented. I tried to get a feel for the possible impact of the new population controls by adjusting for the -270,000 that January's population controls brought in, using the ratio of adjusted employment to total employment (about 0.957). This is shown as a purple square in Figure 1. So, unlike last year, the implied decline has not been eliminated. Having said that, it is important to remember that sampling error is much larger for demographic survey series than for establishment survey series. To track changes in the business cycle, series based on household surveys generally provide less information than series based on business surveys.
Typically, January has quite a seasonal impact. Therefore, it is necessary to consider how large an impact possible misestimation of the seasonal component may have. One way to look at this is to compare 12-month changes in seasonally adjusted and non-seasonally adjusted series. as shown in picture 2.
figure 2: Twelve-month nonfarm payroll employment change, January release (bold red), not seasonally adjusted, January release (lilac) and December 2023 (light red), both as logarithmic differences. Source: BLS and author's calculations.
Note that the year-over-year growth rates between the seasonally adjusted series and the January 2024 seasonally adjusted series are not very different. On the other hand, the seasonally adjusted series for January 2024 and December 2023 do differ significantly.
