• Nonfarm payrolls were up 161k with positive revisions that beat estimates
  • It was all due to a rather fishy SA factor…
  • ...absent which payrolls would have been +15k and missed
  • A new Trump-appointed BLS Commissioner should explain…
  • ...and the Fed’s taskforces should weigh in
  • Stalled yearly growth in payrolls remains a warning sign
 
  • US nonfarm payrolls m/m 000s / UR %, SA, August:
  • Actual: 162 / 4.1
  • Scotia: 30 / 4.2
  • Consensus: 55 / 4.1
  • Prior: 21 / 4.1 (revised from –23 / 4.1)

It may be a one in a million coincidence. A strong beat by nonfarm payrolls was entirely due to an abrupt shift in the seasonal adjustment factor just as the new BLS Commissioner assumed his role. Colour me sceptical.

Charts 1 and 2 were instantly shared in internal chat rooms. They indicate that the seasonally unadjusted change in nonfarm payrolls was on the weak side of history for like months of August and in the ballpark of what I had estimated. That should have driven weak payrolls.

What I had not expected, however, was a very abrupt shift in the seasonal adjustment factor. All of a sudden the pattern that had been in place over the full prior decade toward low and falling seasonal adjustment factors applied to August payrolls swung toward being an historically average SA factor. This matters a great deal.

If, instead, we had stuck by the August SA factors from 2025 or 2024, then nonfarm payrolls would have only risen by 15k. The abrupt swing in the SA factor explained away what would have been a miss on nonfarm payrolls. The SA factor is not normally supposed to swing that abruptly given that it is calculated in a manner that applies a recency bias skewed to the experience over the most recent years.

There must be a public explanation here. And it must come from BLS Commissioner Brett Matsumoto (here). Mr. Matsumoto just took the reins at the BLS after the Senate confirmed his appointment by President Trump on August 7th—the day when the prior payrolls report was released—after a period of turmoil that ensued when Trump fired the former BLS Commissioner after weak numbers. This is the first payrolls report that Matsumoto presided over. Maybe there is a good explanation but I’m attaching a very low trust factor to these numbers until I hear one.

Recall that two of Chair Warsh’s five taskforces were mandated to look at data sources and to weigh in on developments surrounding productivity and jobs. These independence taskforces are supposed to be comprised of academics and experts in these areas. They should address very probing questions concerning BLS practices.

Another data quality dent is historic lows in survey response rates (chart 3).

This makes the rest of the coverage kind of moot, but here we go anyway.

In addition to +161k payrolls, there were also upward revisions that carried July from –23k to +21k and also added 11k to June for a net two-month upward revision of 55k.

Most of the gain was in private payrolls (+127k) as government added 35k entirely at the state and local level (+40k). Federal workers fell by 5k.

Chart 4 shows the sector breakdown and moderate breadth. I can’t explain the gain in leisure and hospitality when several readings we track suggest waning momentum in services, such as plunging number of airplane passengers. Chart 5 might be an attempt at explaining it because the BLS suddenly went to an unusually high SA factor for that sector. Education workers at local governments were up by 42k, almost fully reversing the prior month’s drop which is difficult to square with expiring funding and guidance from nationwide school boards especially just prior to the start of a new school year when contracts kick in.

The unemployment rate held at 4.1% (chart 6) and is derived from the companion household survey that registered a 569k gain in employment (the biggest by far since January 2025) and a 683k rise in the labour force (also the biggest by far since January 2025). That arrested the declining trend in the participation rate at least for now (chart 7).

Hours worked are tracking a 1.4% q/q SAAR gain which is quickening but not quick (chart 8). Wage growth picked up to 3% m/m SAAR (chart 9).

See the next page for other charts, particularly the first one that shows 0.4% q/q nonfarm payrolls is a warning sign on the health of the economy.