A jobs report can count new positions and still leave a basic question unanswered: how much work are employers getting from the people already on payroll? Average weekly hours offer one clue. In September 2026, the average workweek for employees on private nonfarm payrolls stayed at 34.4 hours, while payroll employment rose by 29,000. Read together, those figures describe a small increase in job count alongside no change in the average hours recorded per job. Neither figure alone tells the whole story. Bureau of Labor Statistics
A headcount is not a measure of hours
The monthly employment report’s establishment survey measures payroll jobs, hours and earnings. The jobs figure counts positions; the average workweek estimates how many hours are worked per week, on average, across private nonfarm payroll jobs. That makes hours a measure of labor input, not another way of counting workers. Bureau of Labor Statistics
The distinction matters because a business can add jobs while reducing the hours of some existing employees, or keep headcount steady while expanding schedules. A payroll gain therefore does not automatically mean the economy is using more labor time in the same proportion. The overall average can also shift when the mix of industries or jobs changes, even if no individual’s schedule changes.
September’s steady reading, in context
The September report put the private nonfarm average workweek at 34.4 hours, unchanged from August. It also reported a 0.1 percent monthly increase in average hourly earnings and a 0.6 percent increase in the CPI-U. The separate real-earnings report calculated that real average weekly earnings fell 0.5 percent over the month, with the unchanged workweek contributing to that result alongside the decline in real hourly earnings. Bureau of Labor Statistics Bureau of Labor Statistics
These measures answer different questions. A flat average workweek says the average hours per payroll job did not change in the month. The real weekly earnings calculation adjusts earnings for consumer-price changes; it is not a direct measure of whether every worker’s purchasing power fell. Averages can conceal differences by industry, occupation, schedule and individual circumstances. Bureau of Labor Statistics
Overtime can reveal a different pressure
Overtime is useful because the total average can hide shifts in particular sectors. In manufacturing, the average workweek was 40.6 hours in September and overtime was 3.0 hours; both were unchanged from August. That does not establish that every manufacturing worker had the same schedule, but it gives a sector-specific signal about hours beyond the regular workweek. Bureau of Labor Statistics
If overtime rises while employment is flat, one possible interpretation is that employers are meeting demand by asking current staff to work more before adding payroll positions. If hours fall, that could reflect reduced demand, tighter scheduling or a change in the composition of jobs. The statistic alone cannot distinguish among these explanations. It needs to be read alongside industry employment, earnings and other indicators.
Read the average as one piece of evidence
A practical way to read a report is to separate three questions: Did payroll jobs increase? Did average hours per job change? Did inflation-adjusted weekly earnings move? In September, payroll employment increased by 29,000, the overall average workweek held steady, and real average weekly earnings declined month to month. Those findings can coexist because they measure different parts of employment and pay. Bureau of Labor Statistics Bureau of Labor Statistics
The workweek is an economy-wide average, not a forecast of an individual’s next shift or a verdict on job quality. It helps economists track the volume of labor being used and compare it with job counts and output measures. Its value lies in adding a dimension to the employment picture, while its limits make it important to avoid treating a single month’s flat reading as proof of a broad hiring or demand trend.
Sources & further reading
- September 2026 employment report ↗Bureau of Labor Statistics
- September 2026 real earnings ↗Bureau of Labor Statistics
.jpg){kind=link}