The new birth/death model has been in for two months; the new population controls were retroactively applied to January (usually, there are no revisions from month to month). NFP miss of 150K vs Bloomberg consensus of +58K.
Figure 1: CES nonfarm payroll payroll employment, February release (dark blue, left log scale), January release (sky blue, left log scale), CPS civilian employment, February release, incorporating smoothed population controls for 2024 (dark red, right log scale), official series through January release (light red, right log scale), all in 000’s, s.a. Source: BLS via FRED, and BLS.
Maybe the felon-in-chief decided to take the lid off of Russian oil to distract from the jobs report.
The U.S. lost 92,000 jobs in February, after adding 126,000 in January. The 3-month average is now just 6,000. Revisions to December and January subtracted 69k jobs. The household survey reports an increase of 90,000 jobs with the jobless rate at 4.4% vs 4.3% in January – Sahm Rule still calm.
Striking workers led to a 28k decline in healthcare jobs. That’ll be reversed in March. I didn’t see the strike reflected in any forecasting commentary, nor does it show up in ADP data.
The swing in healthcare workers relative to the 12-month average was -64k. Looks like that exaggerates the impact of the strike (more like -37k), but even putting aside a 64k swing leaves the headline jobs number negative. Not good news overall, though not as bad as the headline job loss suggests.