Last Friday one question was left open after the consumer data: is the income side - jobs and wages - holding up? The September jobs report answered it the same afternoon, Israel time. The US economy added 29,000 jobs, wages rose at the slowest annual pace in more than five years, and part of the two prior months was erased.
And yet the bond market ended the day with higher yields than the day before.
There may be errors or inaccuracies. Found something that looks wrong? Write to me and I will fix it.
The numbers
| Jobs report, September 2026 | Actual | Forecast |
|---|---|---|
| New jobs (establishment survey) | 29,000 | 84,000 |
| Unemployment rate | 4.2% | 4.1% |
| Average hourly earnings, monthly change | 0.1% | 0.3% |
| Average hourly earnings, annual change | 3.0% | 3.1% |
Source: US Bureau of Labor Statistics, October 2, 2026. Forecasts: Dow Jones survey as published by CNBC; the Reuters poll expected 90,000 jobs.
The revisions added to the picture. August was revised from 162,000 jobs to 133,000, and July from a gain of 21,000 to a loss of 10,000. July has had a long journey: in its first print, in early August, it was reported as a loss of 23,000 jobs; last month it was revised to a gain; now it is back in negative territory. Together the two months are 60,000 jobs lower than previously reported, and the three-month average stands at about 51,000 jobs a month.
Average hourly earnings in the private sector rose five cents, to $37.81. On an annual basis that is 3.0%, the lowest since May 2021. The average workweek was unchanged at 34.4 hours, and the broad underemployment measure, U-6, fell to 7.6% from 7.7%.
Where jobs were added, and where they fell
The private sector added 46,000 jobs, and government lost 17,000 - most of them, 13,000, in local government.
The figure that stands out in the breakdown is actually on the positive side. Health care, the sector that has carried the US labor market for the past two years, added 17,000 jobs - about half its monthly average of 33,000 over the prior 12 months. Construction added 11,000, leisure and hospitality 10,000, and manufacturing 9,000.
On the other side: information lost 10,000 jobs, professional and business services 9,000, and financial activities 7,000.
Two surveys, two pictures
The jobs report is built from two separate surveys, and this time they told very different stories. The establishment survey, which produces the payroll number, showed 29,000. The household survey, from which the unemployment rate is calculated, showed 406,000 more people employed.
| Household survey | August | September |
|---|---|---|
| Labor force participation rate | 61.6% | 61.8% |
| Employment-population ratio | 59.1% | 59.2% |
| Employed | 162.75 million | 163.15 million |
| Unemployment rate | 4.1% | 4.2% |
Source: Bureau of Labor Statistics, Table A-1, seasonally adjusted.
In other words, the unemployment rate rose not because people lost jobs, but because more people entered the market and looked for work: the labor force grew by 485,000, employment by 406,000, and the difference - about 80,000 - was added to the unemployed. The household survey is far more volatile than the establishment survey, so neither one is decisive on its own. But both directions are in the same report.
Cleveland Fed President Beth Hammack said in a PBS interview that day that the report was consistent with the trend of recent months. She cited an average of about 41,000 jobs a month over the past year - close to her estimate of the pace needed to keep the labor market stable - and added that there is plenty of data before the next meeting. Hammack, according to Reuters, was an early supporter of raising rates.
Wages versus prices
This is the part that connects directly to the question we asked last week. Wages rising 3.0% a year are below the pace of inflation: the personal consumption expenditures price index, the measure the Fed targets, stood at 3.4% in August. The comparison is between two different months, but the direction is clear - the average wage is not keeping up with prices.
And that comes after a month in which real disposable income did not move, real spending rose 0.6% and the saving rate fell to 4.1%, the lowest since November 2022.
On the price side, Thursday's data pointed the opposite way. The prices component of the ISM manufacturing index jumped to 77.9, from 71.1 in August - and according to ISM itself, it has returned to where it stood at the start of the war with Iran. The headline index stayed in expansion at 54.5, and the manufacturing employment component actually rose to 52.7. Among respondents' negative comments, price volatility, tariffs and the war stood out.
And today, Monday, the services index joined it: its prices component rose to 74.0, from 72.6 in August - the highest since July 2022. The headline index slipped to 54.9 from 55.4, and the employment component returned to expansion at 50.1, after two months below 50. According to ISM, fuel costs were mentioned by respondents twice as often as any other issue.
What the bond market did
In the first hours after the release, yields fell, as you would expect after a weak report. By the close the direction had reversed:
| US Treasury yields | 16.9 (rate hike) | 30.9 | 1.10 | 2.10 (report day) |
|---|---|---|---|---|
| 2-year | 4.74% | 4.88% | 4.78% | 4.83% |
| 10-year | 5.01% | 5.29% | 5.24% | 5.28% |
| 30-year | 5.35% | 5.64% | 5.61% | 5.63% |
US Treasury closing data.
On report day the 2-year yield rose 5 basis points, the 10-year 4 and the 30-year 2. The 10-year yield, at 5.28%, is close to the September 30 high of 5.29% - the highest since May 2002 - and the 10-year real yield closed at 2.92%, near its highest level since November 2008.
The more important picture is since the September 16 rate hike: the 2-year yield is up 9 basis points, the 10-year 27 and the 30-year 28. The short end of the curve, the part that prices the Fed, has barely moved. The long end has risen almost three times as much.
In the fed funds futures market, according to CME FedWatch as reported by CNBC, the odds of another hike in October fell after the report to 17%, from about 36% a week earlier. The odds of a December hike stayed above 75%. For comparison, in the projections the Fed itself published in September, the median committee member expected a rate of 4.1% at year end - that is, one more hike.
The week ahead
- Tuesday: a $58 billion 3-year note auction; speeches by the Fed's Michelle Bowman and Lisa Cook.
- Wednesday: a $39 billion 10-year note auction, closing at 1:00 pm ET. An hour later, at 2:00 pm, the minutes of the September meeting are released.
- Thursday: a $22 billion 30-year bond auction; a speech by Christopher Waller.
- Later this month: the consumer price index on October 14, and the Fed meeting on October 27-28.
הזווית שלי
דעה אישית של אילן אברמוב - לא ייעוץ ולא המלצה
On Friday I wrote that the question is which side converges to the other - the consumer who spends or the consumer who complains. The jobs report did not settle the question, but it shifted the weight. Wages rising 3.0% while prices rise 3.4% means the money coming in is not keeping up with what it has to buy, and the savings cushion is already at its lowest since 2022.
What interests me more is the bond market's reaction. The short end did exactly what you would expect after a weak report: the odds of an October hike were cut. The long end was barely moved. I read it this way: the market is separating two questions - what the Fed will do in the coming month, and how much it takes to lend to the US government for ten and thirty years. The report affected the first. It barely touched the second.
And it is worth remembering the other side: the household survey showed 406,000 more people employed, and unemployment rose mainly because more people were looking for work. Anyone who reads only the 29,000 is missing half the report. On Wednesday, an hour after the 10-year auction, we will see what the committee itself was thinking three weeks ago.
(To be clear: this is my personal opinion only. Nothing here is a recommendation to take any action, and there is no forecast here of any data release or rate decision.)
Ilan Abramov, the author, or anyone on his behalf may hold some of the securities or funds mentioned in this article.
