The July ISM manufacturing employment sub-index printed 52.8, up from June’s 44.2, the first expansion read since May 2024 and the largest single-month jump in the sub-index since June 2020. The line matters because it is the ISM survey’s read on hiring intent inside the goods sector, and Monday’s print broke a fourteen-month streak of sub-50 readings that had run continuously since June 2024. The question the strip is now carrying into Friday’s July nonfarm payrolls print at 8:30 ET is whether the employment sub-index crossover reads through to a positive manufacturing payrolls print, or whether the survey has moved ahead of the payroll data by one release cycle.

What the sub-index measures

The ISM employment sub-index is a diffusion index: purchasing managers report whether their firm’s employment is higher, lower, or the same versus the prior month. The index is calculated as the percentage reporting higher plus half the percentage reporting the same. A read above 50 means more firms reported hiring than firing. A read below 50 means more firms reported contraction.

The critical mechanic is that the diffusion index does not measure the size of the change, only the direction. A firm cutting one worker and a firm cutting 500 both register as “lower.” The sub-index is therefore a breadth signal rather than a magnitude signal.

The historical false-positive rate

Single-month crossovers from contraction to expansion in the ISM manufacturing employment sub-index have preceded a same-month or next-month rise in manufacturing payrolls 55 to 60 percent of the time in the post-1990 sample, according to the paired ISM Report on Business and BLS Current Employment Statistics series. The 40 to 45 percent false-positive rate reflects three recurring patterns:

  • Survey optimism (managers report hiring intent before signed offers)
  • Small-firm bias in the ISM panel (the sub-index over-weights firms that are more responsive to sentiment)
  • Sector composition effects (aerospace and defense hiring can lift the sub-index without a broad goods-sector turn)

The July manufacturing payrolls print in the July NFP report Friday at 8:30 ET is the first confirmatory read. Consensus at the pre-print window sits at minus 5 thousand for manufacturing payrolls, consistent with the fourteen-month contraction streak in the sub-index but inconsistent with Monday’s 52.8 print.

Why a 44 to 52 jump is unusual

The 8.6-point single-month move in the employment sub-index is the third-largest since the series began in 1948 outside the pandemic-recovery months of May 2020 through August 2020. The two prior instances of similar magnitude were February 2021 (from 47.0 to 54.4) and August 2003 (from 46.2 to 53.2). Both preceded a manufacturing payrolls turn inside the following two release cycles: February 2021 with a plus 24 thousand manufacturing payrolls print in February and plus 41 thousand in March, and August 2003 with a plus 5 thousand manufacturing print in September (the first positive read after a three-year manufacturing recession).

The 8.6-point move without a payrolls confirmation would be historically unusual. The prior two instances resolved to the upside within one to two months.

The read through to Friday

A July manufacturing payrolls print at plus 5 thousand or higher confirms the sub-index signal and reads as the first manufacturing payrolls expansion since December 2024. That combination walks the December 10 second cut probability past the 50 percent line and pulls the September 17 SEP dot plot median debate onto the terminal rate anchor rather than the September cut.

A July manufacturing payrolls print at minus 10 thousand or worse flags the sub-index as one-off survey noise or panel composition drift and pulls the December second cut back through the 65 percent line, consistent with the labor-side softening trajectory the June JOLTS print at 7.11 million and Tuesday’s vacancy-to-unemployed ratio at 1.03 confirmed.

The 12-month rolling correlation between the ISM manufacturing employment sub-index and the three-month moving average of manufacturing payrolls has run at 0.71 since 1990. A single-month divergence is not a break in the relationship, it is a lead-lag question. Friday’s print resolves it.

Sources