Expectations heading into last month’s employment report were understandably subdued. June and July had produced exceptionally weak initial payroll readings, while the economy continued to navigate uncertainty from the war in the Middle East, ongoing trade tensions, and an approaching U.S. midterm election. Against that backdrop, August was about as close to a clean sweep as an employment report gets. Nonfarm payrolls increased by 162,000, more than five times the average monthly gain of the previous twelve months, while unemployment held at 4.1 percent. Just as importantly, June and July were revised upward by a combined 55,000 jobs, breaking with a pattern that had persisted for much of the past year in which initial payroll estimates were repeatedly revised lower.
The household survey was equally encouraging. Employment increased by 569,000 as the labor force expanded, pushing the participation rate from 61.4 percent to 61.6 percent and the employment-to-population ratio from 58.9 percent to 59.1 percent. Meanwhile, the number of people working part time for economic reasons fell by 414,000 and the broader U-6 unemployment rate declined from 7.9 percent to 7.7 percent. The unemployment rate therefore remained at 4.1 percent even as more people entered the labor force and found work. Wage growth was similarly constructive, with average hourly earnings increasing 0.3 percent in August and a relatively contained 3.1 percent over the past year.
The establishment survey also showed that the improvement was broadly based. Private-sector payrolls increased by 127,000, with gains across much of the economy. Particularly encouraging were two economically sensitive sectors: construction added 22,000 jobs, while manufacturing gained 16,000 and has now added 58,000 jobs since its recent low in December 2025. The manufacturing diffusion index jumped to 61.1, suggesting that employment gains were spreading across the sector rather than being concentrated in a handful of industries. The notable exceptions were information, which lost 23,000 jobs, and financial activities, which declined by 11,000. Overall, the breadth of hiring reinforced the message coming from both the headline payroll number and the household survey.
As the accompanying exhibit illustrates, however, one of the most consequential numbers in the August report may have been one of the least conspicuous: the average private-sector workweek increased by 0.1 hour to 34.4 hours, its highest level since 2024. Six additional minutes of work per employee per week hardly sounds significant. Spread across roughly 135.7 million private-sector payroll jobs, however, that seemingly small increase represents approximately 13.6 million additional labor hours every week. At the current 34.4-hour average workweek, that is the labor-hour equivalent of roughly 395,000 additional jobs. Even measured against a conventional 40-hour workweek, it represents approximately 339,000 full-time jobs worth of additional weekly labor input. This does not mean the economy actually created another 395,000 jobs, but it demonstrates why changes in hours worked can sometimes convey as much information about underlying labor demand as the headline payroll number itself.
The implications extend beyond August. When demand for goods and services strengthens, businesses do not necessarily respond by immediately hiring additional workers. Employers can first increase the utilization of their existing workforce by extending hours. If demand persists, hiring becomes the next margin of adjustment. The increase in the workweek may therefore be telling us something about future hiring as well as current economic activity. It also has important implications for household income. With both hourly earnings and hours worked increasing, average weekly earnings rose roughly 0.6 percent in August, while the BLS index of aggregate weekly payrolls, which incorporates employment, hours and wages, increased 0.7 percent. Households ultimately spend income, not hourly wage rates, and the combination of more people working, longer hours and higher wages provides additional support for consumption.
Exhibit: Average weekly hours worked. More hours equal more income and growth.
Source: U.S. Department of Labor
John Maynard Keynes is famously credited with saying, “When my information changes, I alter my conclusions. What do you do, sir?”1 As our readers know, we have been concerned about deterioration in the labor market for some time, and a report as broadly positive as this one requires us to remain open to the possibility that conditions are improving. August was unquestionably encouraging, and continued strength in employment, hours worked, participation and industry breadth would make us increasingly optimistic about the labor-market outlook. At the same time, one report does not make a trend. Payroll estimates remain subject to subsequent revisions and the annual benchmark process, and some of the current strength may reflect powerful but potentially temporary tailwinds from the OBBB and the extraordinary boom in data-center and related infrastructure construction, with effects extending into manufacturing, utilities and other peripheral industries. For now, the August report deserves to be viewed for what it was: an exceptionally positive change in the information in front of us. We will keep an open mind about what that information is telling us while remaining careful not to extrapolate a new trend just as some of the forces supporting it may be approaching their peak.
As always, stay tuned.
Ardavan Mobasheri
Chief Economist
[1] Attributed to John Maynard Keynes by Paul A. Samuelson in “The Keynes Centenary,” The Economist, vol. 287 (1983), p. 19. The remark is widely attributed to Keynes, though no definitive primary-source record of Keynes saying it has been identified.