Strong Jobs Report Includes a Very Healthy Unemployment Shift

There is almost nothing to nitpick about the latest jobs report, though we do have some caveats.

There was a big beat on jobs, with 130,000 versus the 65,000 expected. Private jobs crushed it, adding 172,000 (yes, public sector jobs shrank).

The downward revision from the prior two reports was “only” -17,000.

The benchmark revisions were -862,000. That's a big number but -825,000 was baked in, so it's kind of a rounding error at this stage on “old” data.

The unemployment rate dropped to 4.3%. Not only did the household survey add 528,000 jobs, but we got this drop even while labor participation increased to 62.5% — a very healthy shift in unemployment.

The birth/death model showed job losses of 69,000. Since I do think birth/death had an outsized influence on the revisions, it is good to see a negative number here. It gives me more confidence in the print.

What is there to complain about?

The NSA (not seasonally adjusted) had a drop of 2,649,000 jobs.

We have been complaining (for years) that the seasonal adjustments have a lot of issues and this year’s might be worse than usual in that respect.

We still add a lot of jobs in the winter and take them away in the summer, because that is how the weather worked (there was slowing in the Northeast), but we no longer believe that is accurate as so much construction has moved to the South.

The report added back a lot of jobs that were added for the holidays. It is unclear how many jobs were really added for the holidays. It does not help that the government shutdown(s) made the data even less reliable than usual.

In 2025, the largest downward revision was in February where they took away 167,000 from the prior two reports.

These two factors are why I will take this payroll data with a “grain of salt."