Two Labour Markets, One Fed: US Payrolls Fell in July While Small Business Hiring Plans Hit a Four-Year High

The July jobs report was bad. The July small business survey was good. Both are true, and the gap between them is the most important thing happening in the US economy right now.

Start with the headline. Nonfarm payrolls showed a surprise loss of 23,000 US jobs in July — not a slowdown in hiring, an outright decline. Markets read it the way you'd expect: as evidence the Fed's path just got more dovish. Bond yields fell, gold rallied to a seven-week high, and equities broadly liked it.

Then, on 11 August, the NFIB released its small business survey. The Small Business Optimism Index rose 2.4 points to 99.8 — its best reading since August 2025 and above the survey's 52-year average of 98.0. Buried in the detail was the number that should have moved more people: a seasonally adjusted net 20% of owners said they were planning to create new jobs, the highest reading since October 2022 and up 9 points from June.

So which is it? Is the US labour market falling apart or gearing up?

Both, in different places

The payroll survey is weighted heavily toward large employers. The NFIB survey is, by construction, Main Street — the restaurant with 14 staff, the HVAC contractor with six trucks, the regional accounting firm.

What the two datasets together describe is a labour market that is contracting at the top and expanding at the bottom. Large companies — especially in technology, media, and anything touched by an AI-driven restructuring — have spent 2026 cutting. Smaller employers, who never staffed up the way big corporates did in 2021–22, are still filling gaps.

That is a genuinely unusual configuration. In most cycles the small business sector is the first to crack, not the last. Small firms have thinner balance sheets and worse access to credit; when conditions tighten they shed staff early. The current pattern is the reverse, and it makes the standard recession playbook harder to run.

Why this makes the Fed's job worse, not easier

A single interest rate cannot address two labour markets moving in opposite directions.

Cut aggressively to support the headline payroll weakness and you pour fuel on a small business sector that is already planning to hire — into an economy where oil has been sitting in the mid-$80s and where the Strait of Hormuz situation remains unresolved. Hold, or hike, to guard against that inflation impulse and you compound job losses at exactly the large employers that drive the headline number.

This is not a hypothetical tension. It is why nine of eighteen Fed officials have signalled they see a hike rather than a cut as a live possibility in 2026 — a split we wrote about in The Fed's Next Move Might Be a Hike — even as the payroll print argues the other way.

What this means for a long-term investor

Very little, on its own. One month of payroll data is noise, and one sentiment survey is a sentiment survey.

What is worth internalising is the shape of the disagreement, because it explains a market that has felt incoherent all summer: the Dow printing records while the Nasdaq slips, energy leading while technology lags, small caps outperforming. Those are not random rotations. They are the market repeatedly repricing which of these two labour markets the Fed is going to respond to.

The practical takeaway is unglamorous. If your portfolio's performance depends on the Fed cutting three times this year, you have taken a macro position you probably didn't intend to take. A portfolio built to survive either outcome looks boring: broad index exposure, some duration, some real-asset exposure, and a cash buffer you don't have to touch. It also lets you stop caring which survey was right.