Payrolls added 29,000. The pay squeeze is an energy bill.
September added 29,000 jobs and pay rose 3.0%, level with core inflation. Energy, up 16.3% in a year, is what puts US paychecks behind prices.
By the numbers
29,000jobs added in September, against an 84,000 forecast
51,000average monthly gain over July to September after revisions
3.0%rise in average hourly pay over 12 months, level with core PCE inflation
16.3%rise in consumer energy prices in the year to August
21.6%odds of an October Fed hike after the report, down from about 70%
US employers added 29,000 jobs in September, the Bureau of Labor Statistics reported on October 2, against a forecast of 84,000. Average hourly pay rose 3.0% over twelve months, to $37.81. That is the slowest pace since 2021, and it is the number that matters most for what the Federal Reserve does next.
Here is our read. At 3.0%, wages are level with core PCE inflation, the Fed's preferred underlying gauge, and 0.6 points above core CPI. Paychecks are losing to prices for one reason: energy, up 16.3% in a year. The Fed raised rates on September 16 as if pay were pushing prices up. The data now shows an energy shock eating into pay, and rate futures repriced the same day.
The jobs number, after revisions
The headline was weak. The revisions were worse. BLS cut July from +21,000 to -10,000 and August from +162,000 to +133,000, a combined 60,000 jobs that no longer exist on paper (BLS, October 2).
Put the three months together: -10,000, +133,000, +29,000. That averages about 51,000 a month. The twelve months before September averaged 45,000.
August was the month that looked hot. When it came out, we noted that restaurants and local school payrolls supplied 101,000 of its 162,000, one of them a seasonal line waiting to be revised. It was revised, by 29,000.
The rest of September's report was flat. Unemployment read 4.2%, with 7.1 million people out of work. Health care added 17,000, about half its twelve-month average of 33,000. Construction added 11,000 and manufacturing 9,000. Financial activities lost 7,000 and are down 129,000 from their May 2025 peak, most of it at insurance carriers.
Wages, set against each kind of inflation
Average hourly earnings grew 3.0% in the year to September. September's price data isn't out yet; CPI lands October 14. Against the latest readings, for August, the wage figure sits in the middle of the pack.
16.3%3.4%3.4%3%3%2.4%
View as table
| label | 12-month change, latest reading |
|---|---|
| Energy prices (CPI) | 16.3% |
| All prices (CPI) | 3.4% |
| All prices (PCE) | 3.4% |
| Average hourly pay | 3% |
| Core prices (PCE) | 3% |
| Core prices (CPI) | 2.4% |
The headline indexes both read 3.4%: CPI from BLS on September 11, PCE from the Commerce Department on September 30. Strip out food and energy and the picture changes. Core CPI rose 2.4%, its lowest since March 2021. Core PCE rose 3.0%, against an expected 3.3%.
So a worker's raise keeps pace with the price of rent, haircuts, cars and insurance. It doesn't keep pace with the pump. Gasoline is up 27.4% in a year and fuel oil 52.0%, by the August CPI.
The loss is real. Real average hourly earnings, which BLS computes by deflating pay with headline CPI, fell 0.3% in the year to August. They were negative in June and July too. That squeeze lines up with what households told the Conference Board last month, when confidence hit a 12-year low as diesel ran up 75%.
Why this changes the case for another hike
A central bank raises rates when wages and prices push each other up. Workers win raises, firms pass them on, workers ask for more. That spiral needs wages running ahead of underlying prices.
September's report shows the reverse. Pay growth slowed from 3.1% in August to 3.0%. The monthly gain was 0.1%, five cents. Hiring is running near 51,000 a month. None of that describes a labor market bidding up its own price.
The Fed's own statement on September 16 said job gains "have kept pace with the workforce" (Federal Reserve). It voted 12-0 to lift the federal funds range to 3.75% to 4%. Two of the three months behind that sentence have since been revised down.
Futures repriced on the day. CME FedWatch had put an October hike near 70% a week before the report. Afterward it read 21.59%. The same tool still prices a hike by December.
Higher rates do little about oil. They work on demand through borrowing costs and hiring. Applied to an energy shock, that means a household pays more at the pump and then faces a weaker job market on top. That's the cost of the December hike the market still expects, and it falls on the same people twice.
What the bond market did instead
The long end didn't celebrate. The 10-year Treasury closed at 5.24% on October 1, 5.28% on October 2 after the report, and 5.31% on October 5, by the Treasury's daily par yield curve. The 2-year moved from 4.44% to 4.47%.
A weak jobs report with an October hike priced out would usually pull yields down. These rose. The bond market is pricing something besides the next meeting: a December hike, a long stretch of high rates, and compensation for inflation risk. We broke down how much of the 10-year is inflation bets a week ago. For a borrower, the short version is plain. A softer Fed in October doesn't make a mortgage cheaper this month.
The case against this read
Three things could make it wrong.
Energy can leak into core. Fuel feeds freight, airfares and anything delivered by truck. If those costs reach core prices over the next few months, the Fed's worry about inflation sticking gains weight, and the hike argument comes back with better evidence.
One month of pay is noisy. Average hourly earnings shift when the mix of jobs shifts. A month where low-wage hiring rises can lift the average without anyone getting a raise, and the reverse. The 3.0% needs a second reading.
Revisions run both ways. The same process that took 60,000 jobs out of July and August can put them back.
What to watch
October 14 brings September CPI and the real earnings release. If core stays near 2.4% while headline holds above 3%, the gap is energy again, and the case for raising rates into a 51,000-a-month labor market gets thinner. The Fed decides on October 28. The next jobs report, for October, follows in early November.
For a reader with a paycheck, the useful number is the gap between your raise and your fuel and utility bills, since that's where the 2026 squeeze sits. For a reader with a mortgage or a bond fund, the useful number is the 10-year, and it went up.
Sources
- BLS, The Employment Situation, September 2026 (October 2, 2026)
- BLS, Consumer Price Index, August 2026 (September 11, 2026)
- BLS, Real Earnings, August 2026 (September 11, 2026)
- FXStreet, US core PCE inflation, August 2026 (September 30, 2026)
- FXStreet, The October Fed hike just died (October 2, 2026)
- Federal Reserve, FOMC statement (September 16, 2026)
- U.S. Treasury, Daily Treasury Par Yield Curve Rates, October 2026
- Fed Rate Calc, FOMC meeting schedule, October 2026
- Yahoo Finance, September 2026 jobs report (October 2, 2026)
This is not financial advice.
More from Markets
The 10-year hit 5.29%. Inflation bets added 12 basis points of it.
The 10-year Treasury rose 85 basis points in Q3 to a 2002 high. Treasury's own TIPS curve puts 73 of them on real yields and only 12 on inflation.