The consensus take on Warsh’s labor market turnover remarks at Jackson Hole is that he was simply acknowledging a cooling jobs market. The JOLTS data released alongside that speech tells a more specific, and more interesting, story than that framing allows.
What the rematching argument actually says
In his Jackson Hole speech, Fed Chair Warsh said: ‘In my view, the relatively low turnover in today’s labor market is partly a result of the significant rematching between employers and employees that happened at scale in the post-pandemic environment.’ That is not a weakness narrative. It is a structural one, and the distinction matters.
The backdrop is the enormous wave of voluntary quits in 2021 and 2022. Labour shortages, partly driven by people remaining out of work while receiving extra government payments, prompted employers to raise wages and poach workers aggressively. Employees, seeing opportunities across sectors, jumped jobs and even entire industries. The result, in Warsh’s reading, was not chaos but calibration: workers landed roles that better matched their skills and ambitions, and employers found better-fitting staff. The churn settled. So did Warsh’s labor market turnover reading.
The jobless rate currently sits at 4.1%, according to Warsh’s Federal Reserve speech. That number is not screaming distress. Placing it alongside the JOLTS figures makes the rematching thesis look more coherent than a simple slowdown story would suggest.
The JOLTS numbers behind the Warsh labor market turnover argument
The Bureau of Labor Statistics Job Openings and Labour Turnover Survey tracks the mechanics of labour market churn: quits, layoffs, other separations, job openings, and hires. It does not track payroll growth or the unemployment rate. That scope is important context for what Warsh was and was not saying.
Voluntary quits fell by 157,000 in July to 3.06 million, with the three-month average rising to 3.14 million. Quits account for 60% of total separations and are the primary engine of turnover. Fewer quits mean fewer slots left behind for employers to fill.
Layoffs and discharges declined to 1.67 million in July, down 106,000 year-over-year, with the three-month average holding at 1.74 million. These levels sit at the lower end of the pre-pandemic range. Other separations, covering retirements and deaths while employed, rose to 350,000, though their 12-month average of 314,000 remains a small share of total turnover.
Job openings rose by 89,000 in July to 7.27 million. The three-month average dipped to 7.33 million but remains 129,000 above a year ago. It is worth noting how the survey defines an opening: a specific position must exist, work must be available for it, it could start within 30 days, and the employer must be actively recruiting from outside. Internal transfers, promotions, and contractor roles are excluded. The 21,000 business locations surveyed are HR departments, not online job boards.
Hires fell by 278,000 in July to 5.05 million, with the three-month average declining to 5.21 million. Nearly all of those hires filled slots left behind by separations. The number does not speak to net payroll growth; it speaks to turnover replacement, which is precisely what Warsh’s labor market turnover framing is pointing at.
The productivity footnote the consensus tends to skip
There is an underappreciated second-order effect here. During the peak churn years of 2021 and 2022, labour productivity declined. Newly hired workers, however well-matched in theory, still had to climb the learning curve. Once that wave settled, productivity recovered to relatively high levels. That sequence supports the rematching thesis rather than complicating it: the short-term friction was the price of a better long-run equilibrium.
What makes Warsh’s framing worth scrutinising is precisely what it pushes back against. For years, according to Wolf Richter writing for Wolf Street, former Fed Chair Powell used this same JOLTS data to underpin arguments about labour market strength or weakness, depending on the moment. Warsh’s reading treats the current low-turnover environment not as a symptom of fragility but as the settled state after a large-scale structural reshuffle. Whether that interpretation holds is a fair question. What the data does not support is reading Warsh’s labor market turnover remarks as a straightforward recession signal dressed up in academic language.
