Labour market
Hiring narrows to three sectors as private payroll trackers slow to a crawl
The Verity payroll panel shows private employment growth of 31,000 a month across the summer, with health care, utilities and defence adjacent contracting supplying almost all of it. The quits rate is back to 2016 levels.
In this story
Illustrative. Bank Season is an editorial prototype. This story, its sources, the issuers named in it and every figure it quotes are invented to demonstrate the publication. Nothing here is reported fact or investment advice. Read the disclosure.
A labour market can cool in two ways. Firms can fire people, which is loud, shows up immediately in claims data and tends to force a policy response. Or firms can stop hiring, which is quiet, takes two years to become obvious, and leaves the unemployment rate looking respectable while the flow of new work dries up underneath it. The Verity Research payroll panel, covering roughly 14 per cent of US private employment across about 610,000 employers, describes the second kind in unusually clear terms.
31,000
Average monthly private job growth, June to August
Verity Research panel estimate, against 148,000 a year earlier
Thirty-one thousand a month is not a recessionary number. It is roughly the level at which the employment to population ratio holds flat given current labour force growth, so the unemployment rate can sit still for quarters while this continues. What makes it uncomfortable is the composition, and the fact that the trend has moved in one direction for eleven straight months.
The deceleration is steady, not sudden
Monthly change in private employment, Verity panel estimate
Thousands of jobs, September 2025 through August 2026.
There is no single break in that series. Each month is a little worse than the one before it, by between five and twenty-five thousand, for eleven consecutive readings. Series that decay this smoothly tend not to be measurement noise, and they tend not to reverse without something changing in the cost of capital or in final demand. Neither has changed yet.
Verity panel employment change by sector, three month average to August
| Sector | Monthly change | Same period 2025 | Share of panel employment |
|---|---|---|---|
| Health care and social assistance | +18,400 | +52,100 | 16.2% |
| Utilities and energy services | +5,300 | +6,800 | 2.1% |
| Defence adjacent contracting | +3,700 | +4,900 | 3.4% |
| Construction | +2,100 | +12,500 | 5.9% |
| Transport and warehousing | −1,800 | +9,700 | 6.3% |
| Leisure and hospitality | −2,900 | +31,600 | 11.8% |
| Professional and business services | −4,100 | +22,400 | 14.6% |
Sectors not shown contributed a net 10,300 a month. Panel figures are scaled estimates, not official statistics.
Health care has carried this labour market for three years and it is now carrying it almost alone, at a rate of hiring one third of what it managed last summer. Professional and business services, the sector that historically turns first because it contains temporary help and staffing, has been shedding jobs on the panel since April. Leisure and hospitality going negative is newer and matters more for consumption, since it employs a workforce with very little savings buffer.
The composition is the warning, not the level. When two thirds of net hiring comes from sectors whose demand is set by government budgets and demographics, the private cycle has already stopped.
Dov Ackerman, chief economist at Bellweather Macro
Quits are the cleaner signal
Payroll counts are noisy and get revised. The quits rate is neither. It measures how many workers voluntarily left a job, which is a direct reading of how confident people are that another job is available, and it turns before wages do with reasonable consistency. On the Verity panel it has fallen for nine straight quarters.
Voluntary quits rate, Verity panel
Per cent of panel employment, quarterly. Third quarter is a partial estimate.
1.8%
Quits rate, Verity panel
Lowest reading since 2016
At 1.8 per cent the quits rate is below its pre-pandemic average and roughly where it sat in 2016, a year in which wage growth ran near 2.5 per cent. Panel wage growth is currently 3.6 per cent and falling by about a tenth of a point a quarter. If the historical lead holds, wage growth reaches the low threes by the middle of next year, which would remove the main domestic source of services inflation.
What this does to the rate debate
- A hiring rate near 31,000 is consistent with a stable unemployment rate, so the labour data alone will not force faster easing until layoffs pick up.
- Initial claims remain near 224,000 on a four week average, which confirms that firms are freezing hiring rather than cutting staff.
- Falling quits point to slower wage growth by mid 2027, which supports the disinflation case that the services nowcast has so far refused to confirm.
- The composition means fiscal policy now has an outsized influence on the employment number, since health care and defence adjacent hiring is budget driven.
- Watch the temporary help series on the panel, which has fallen for six months and is the only component that reliably leads the aggregate.
Wednesday's policy decision made no reference to any of this, and the statement language on labour conditions was unchanged from July. That is defensible when the unemployment rate is steady. It is also the position a committee occupies just before a labour market that has been quietly freezing for a year begins to thaw in the wrong direction, and the front end of the Treasury curve has started to price that possibility even as the long end prices the opposite.
The honest summary is that this is a slow puncture rather than a blowout. Slow punctures are harder to respond to, because every individual month looks fine and the cumulative reading only becomes obvious in hindsight. Eleven consecutive decelerating months is now enough data that hindsight is not required.
