An hourly wage and a weekly paycheck can tell different stories. August’s federal earnings report is a clear example.
Real average hourly earnings for employees on private nonfarm payrolls fell 0.1% from July to August, after seasonal adjustment, the Bureau of Labor Statistics reported September 11. Nominal hourly earnings increased, but consumer prices rose faster. The result was a small decline in average hourly purchasing power. [1]
At the same time, real average weekly earnings increased 0.2%. BLS connected that increase to a longer average workweek, which rose 0.3%. In plain language, the weekly measure improved even though the inflation-adjusted value of an average hour slipped. [1]
The year-over-year comparison had a similar split. Real hourly earnings were down 0.3% from August 2025, while real weekly earnings were up 0.3%, with the average workweek longer than a year earlier. These measures describe different units of work, not conflicting versions of the same statistic. [1]
The methodology matters for anyone trying to compare the headline with a personal paycheck. BLS constructs the earnings measures from its establishment survey of payrolls and hours, then adjusts them using consumer price indexes. The all-employee series uses CPI-U. A separate production and nonsupervisory series uses CPI-W. [2]
These are averages across jobs, including full-time and part-time positions. They are not a survey result saying every worker received the same raise or worked the same additional time. Changes in the mix of higher- and lower-paying employment can also affect the averages. [2]
For a household conversation, the practical distinction is useful. More hours may generate more income, but they also require more time. A pay increase per hour and an increase produced by additional hours should not be treated as interchangeable when discussing what work is delivering.
Our analysis is that both lines belong in the headline. Focusing only on weekly earnings could obscure the slight decline in hourly buying power. Focusing only on the hourly decline would leave out the improvement in the weekly measure.
Neither number establishes what happened to a particular Miami household after taxes, commuting costs, benefits or changes in household membership. Those are different comparisons, and this national release does not resolve them.
The report also does not measure job vacancies or announce layoffs. It is a wage-and-hours update, distinct from a hiring count or a survey of jobseekers’ confidence.
BLS schedules the next real earnings release for October 14 at 8:30 a.m. Eastern. Until then, August’s central finding is the split: slightly less buying power per hour, but more inflation-adjusted earnings across the average workweek. [1]
