Everybody remember July?
The first print said the U.S. lost 23,000 jobs. Local government education. Retail. The internet wrote recession fan fiction for a weekend.
This morning BLS walked that back. July is now a 21,000 gain. June went from +20,000 to +31,000. Combined, the “jobless summer” is about 55,000 jobs better than the story you were told.
That is not a conspiracy. That is how the establishment survey works.
Why the first number lies a little
Payrolls come from a sample of employers. The first release is whoever got their homework in on time. Late homework arrives for the next two months. Seasonal adjustment — especially around summer school jobs — then takes another swing.
July’s original hole was local government education (−50,000) on the first print. That category is famous for seasonal noise. When August prints a 162,000 headline and July flips sign, you are allowed to say: the first story was too loud.
Revisions do not make August fake. They make July’s panic look expensive.
How to read the next one
Three habits:
- Never marry the first print. Especially in education, retail, and birth-death model months.
- Stack three months. One −23k is a weather report. Three down months in a row is a climate.
- Pair it with the household survey. Unemployment 4.1% with participation up to 61.6% is a different movie than 4.1% with people quitting the labor force.
ADP said +38,000 private jobs Wednesday. BLS said 162,000 total Friday. Different surveys, different weeks, different arguments. Use both. Worship neither.
What this means for you
If you made a life decision off July −23k, you made it off a draft. If you make a life decision off 162k alone, you are doing it again. The Fed will look at the level of the labor market — still tight — and at CPI next week.
Bottom line: the revision is the adult in the room. The headline is the teenager with a megaphone. Listen to both. Quote the second one.
GLHR NEWS explains the news. Not investment, tax, or trading advice. Do your own work before you act.
