Two headline numbers from two different surveys
The Bureau of Labor Statistics reported that total nonfarm payroll employment increased by 162,000 in August 2026. It also reported that the unemployment rate was unchanged at 4.1%. Those figures appear together, but they do not come from the same survey.
The payroll figure comes from the Current Employment Statistics survey of business and government worksites. The unemployment rate comes from the Current Population Survey of households. Reading them as complementary measures is more accurate than expecting them to move together every month.
The surveys count different things
The payroll survey is designed to measure jobs, hours, and earnings in the nonfarm sector. A person appearing on two payrolls can be counted as two jobs. The household survey classifies people as employed, unemployed, or outside the labor force and includes groups such as the self-employed and agricultural workers that the payroll survey excludes.
BLS says the payroll survey generally provides the more precise estimate of month-to-month payroll change because its sample is much larger. The household survey provides demographic and labor-force information the payroll survey cannot supply. A difference between them is therefore a reason to inspect definitions and trends, not evidence that one number must be false.
Revisions belong in the headline
The August release revised June payroll growth from 20,000 to 31,000 and revised July from a loss of 23,000 to a gain of 21,000. Together, those changes added 55,000 jobs to the previously reported totals for the two months.
Monthly payroll estimates are preliminary. BLS revises them in the next two monthly releases as more sampled employers respond and seasonal factors are recalculated. That planned process means the first estimate is useful but not final; a three-month view can be more informative than treating every first estimate as permanent.
Wages, hours, and industries add texture
Average hourly earnings for employees on private nonfarm payrolls rose 0.3% during August and 3.1% over the year. The average workweek edged up by 0.1 hour to 34.4 hours. These are averages before adjusting for consumer-price changes, so they do not describe the purchasing power or schedule of every worker.
Employment increased in food services and drinking places and in local government education, while the information industry lost jobs. BLS reported little change across several other major industries. That distribution matters because a national gain can coexist with very different conditions by industry and location.
Questions for the next report
- How much will the August payroll estimate change as additional employer responses arrive?
- Do gains broaden across industries or remain concentrated in a small number of categories?
- How do average hourly earnings compare with the matching official consumer-price data?
- Do the payroll and household surveys point to a similar multi-month direction despite monthly differences?
