Last week, we witnessed a strange paradox. The U.S. economy lost 23,000 jobs in July. At the same time, unemployment... fell.
Yes, you read that correctly. Fewer jobs, lower unemployment.
And the cherry on top? Just one day earlier, on Thursday, weekly jobless claims had shown something we had not seen since 1969.
Let's start from the beginning.
On Thursday, new unemployment benefit claims were released. They increased by 1,000 to 199,000 for the week ending August 1.
That was still below the 201,000 economists had expected.
The important part is that claims stayed below 200,000 for a third consecutive week. That is the longest such streak since 1969.
The four week moving average fell by 4,500 to 198,700. Continuing claims stood at 1.801 million, while the insured unemployment rate remained unchanged at 1.2%.
Analyst Guy Berger also pointed out that claims are running below comparable levels seen in 2023, 2024, and 2025.
In short, on Thursday evening the message was clear:
The labor market was holding up.
Then Friday arrived.
Economists had expected somewhere between 83,000 and 88,000 new jobs in July.
Instead, the economy lost 23,000 jobs.
And this is where things become even more interesting.
Because the truly bad news was not July itself.
It was the revisions to previous months.
May, originally reported at 129,000 jobs, was revised down to 63,000, a reduction of 66,000.
June was revised from 57,000 down to just 20,000.
As a result, the average monthly job gain over the past twelve months fell to only 34,000 jobs. This labor market was much weaker than we thought.
At the same time, the unemployment rate declined to 4.1% from 4.2% the previous month.
So how did unemployment fall?
Here is the key.
Household employment declined by 87,000 people.
But the labor force shrank by 264,000.
In very simple terms, far more people left the labor market than lost their jobs.
And when someone stops looking for work, they are no longer counted as unemployed.
The labor force participation rate fell to 61.4%, its lowest level in more than five years.
If we exclude the Covid period, we have to go all the way back to 1976 to find comparable levels.
The employment-to-population ratio dropped to 58.9%, its lowest reading since May 2014.
As Bill Adams of Fifth Third Bancorp put it, unemployment is falling for the wrong reason.
There simply are not enough workers.
In the years immediately following the pandemic, immigration helped offset the aging workforce. Now it no longer does.
The biggest culprit was local government, particularly education, which lost 50,000 jobs.
Leisure and hospitality followed with a loss of 40,000 jobs, after already losing 43,000 in June.
One possible explanation: the World Cup ended.
Retail lost 19,000 jobs, while financial services shed another 14,000.
On the positive side, healthcare added 22,000 jobs.
However, its twelve month average is 36,000, meaning even one of the strongest sectors is slowing down.
Construction added another 22,000 jobs.
Overall, the private sector remained positive with a gain of 30,000 jobs, while the public sector lost 53,000.
Within retail, another interesting trend emerged.
Supercenters and large department stores cut 21,000 jobs.
Meanwhile, bookstores, hobby stores, musical instrument shops, and sporting goods stores added 10,000 jobs.
Chris Lau interprets this as a sign that consumers are cutting subscriptions and online services in order to cope with higher spending on food, housing, and energy.
Now let's talk about wages, arguably the most concerning part of the report.
Average hourly earnings increased by just two cents.
Two cents.
That translates to only +0.1% month over month, compared with expectations of +0.3%.
On an annual basis, wage growth slowed to 3.2%, down from 3.5% in June, the lowest level in five years.
On one hand, this means wages are not fueling inflation.
On the other hand, it means wages are struggling to keep up with rising prices.
We now have a situation we do not see very often.
The Federal Reserve is considering raising interest rates, not cutting them.
Inflation remains above its 2% target, and several officials have openly discussed a rate hike as early as September.
Last week, the FOMC voted 9 to 3 to keep the federal funds rate unchanged at 3.50% to 3.75%.
After the jobs report, market expectations shifted significantly.
The probability of rates remaining unchanged rose to 55.9% from 45%.
The probability of a rate hike fell to 44.1% from 55%.
For October, markets currently assign a 59.2% probability to rates remaining unchanged.
Analysts are sharply divided.
Olu Sonola of Fitch Ratings argues that the negative payroll figure looks more like a seasonal distortion than a genuine warning sign, and that September remains a 50-50 decision.
Chris Zaccarelli, on the other hand, called the report a "game changer."
Until yesterday, most investors believed the Fed had no choice but to raise rates because the labor market appeared strong.
Now, it doesn't.
