A paycheck can rise while its buying power slips. To measure that difference, the Bureau of Labor Statistics adjusts average earnings for changes in consumer prices. The result, called real earnings, helps answer whether pay has kept pace with inflation over a stated period. It is an economic measure built from averages and a price index, rather than a calculation of what any particular household can afford. In September 2026, the BLS reported that real average hourly earnings fell for both employee groups it tracks.
The adjustment behind “real” pay
Nominal earnings are wages expressed in the dollars workers receive. Real earnings recast those dollars in relation to price changes. In plain terms, the BLS compares the change in average hourly pay with the change in a consumer price index. When prices rise faster than pay, the inflation-adjusted measure declines; when pay rises faster, it increases.
For a simple hypothetical example, suppose an hourly wage rises from $20 to $20.40, a 2 percent increase. If the relevant price index rises 3 percent over the same period, the worker’s real hourly pay has fallen slightly. Dividing the new wage by the price change gives about $19.81 in the earlier period’s purchasing-power terms. That does not mean the worker’s paycheck shrank: the dollar amount is higher, but it buys less against the index’s basket.
Two employee groups, two price indexes
The BLS does not use one deflator for every earnings series in its September release. It adjusts the all-employees series with the Consumer Price Index for All Urban Consumers, or CPI-U. For production and nonsupervisory employees, it uses the Consumer Price Index for Urban Wage Earners and Clerical Workers, or CPI-W. The two indexes cover different populations, so their readings and the resulting real-wage measures can differ.
That distinction appears in the September 2026 figures. Average hourly earnings for all employees rose 0.1 percent from August, while CPI-U rose 0.6 percent; real average hourly earnings therefore fell 0.5 percent. For production and nonsupervisory employees, nominal pay rose 0.2 percent against a 0.7 percent rise in CPI-W, and real hourly earnings fell 0.6 percent. Those are month-to-month, seasonally adjusted changes, not annual rates Bureau of Labor Statistics.
What the September figures do—and do not—say
Over the year from September 2025 to September 2026, the BLS reported a 0.7 percent decline in real average hourly earnings for all employees and a 0.6 percent decline for production and nonsupervisory employees. The corresponding nominal hourly earnings increases were 3.0 percent and 3.3 percent. These figures reflect the selected earnings series and price indexes, not a record of each worker’s actual raise or spending pattern Bureau of Labor Statistics.
The earnings averages come from the BLS establishment survey of private nonfarm payrolls. The all-employee average can shift when employment changes across higher- and lower-paying industries, as well as when wage rates change. The series also combines jobs, including part-time positions and jobs held by people with multiple jobs. It is not adjusted for a worker’s occupation, age, education, household, or location Bureau of Labor Statistics.
Hourly pay is not the whole paycheck
Real average weekly earnings add another factor: hours. If hourly pay keeps pace with prices but the average workweek shortens, inflation-adjusted weekly earnings may still fall. Conversely, more hours can lift weekly earnings even when the hourly measure is weak. In September, the all-employee average workweek was unchanged at 34.4 hours, and real average weekly earnings fell 0.5 percent over the month. The production and nonsupervisory workweek also held steady at 33.8 hours; that group’s real weekly earnings fell 0.5 percent Bureau of Labor Statistics.
The September 2026 employment report likewise put all-employee average hourly pay at $37.81, up 3.0 percent over the year, and reported a 4.2 percent unemployment rate. Those are useful measures of labor-market conditions, but neither says how a particular worker’s pay compares with that person’s rent, groceries, or bills Bureau of Labor Statistics. The next BLS real-earnings release, covering October, was scheduled for November 10, 2026. Until then, September is the latest reference period in this series Bureau of Labor Statistics.
Sources & further reading
- September 2026 real earnings ↗Bureau of Labor Statistics
- September 2026 employment report ↗Bureau of Labor Statistics
