An economy-wide average wage can move even when no worker gets a raise. That happens because the calculation combines pay across jobs, and the share of jobs in higher- or lower-paying industries can change. The September employment report showed average hourly earnings rising 3 percent over the year, while the separate inflation-adjusted measure fell 0.7 percent. Those figures describe averages across jobs, not a record of what happened to each person’s paycheck. Bureau of Labor Statistics Bureau of Labor Statistics
The average reflects who is working
Think of a deliberately simple, hypothetical economy with two groups: 10 workers in jobs paying $30 an hour and 10 in jobs paying $20. The overall average is $25. Now suppose five more workers enter the lower-paid group, while nobody’s hourly rate changes. The average becomes about $23.33: total hourly pay represented by the groups is $400 across 15 workers in the first group and $300 across 10 in the second, or $700 across 25 workers. The average falls because the workforce has more lower-paid jobs, not because anyone’s wage was cut.
The reverse can happen too. If lower-paid jobs disappear from the count, the average may rise even if the remaining workers’ pay is unchanged. This is called a composition effect. It is a feature of averages built from changing groups, not evidence by itself that workers are gaining or losing purchasing power. The BLS earnings series aggregates payroll and hours across industries, so shifts in industry employment can affect the reported average alongside changes in wage rates. Bureau of Labor Statistics
An individual raise is a different measure
A worker comparing two pay stubs is asking whether their own wage rate changed. The national average asks how the pay represented across a broad set of jobs changed relative to the number of jobs or hours in the calculation. These are related questions, but they do not have the same answer. A person might receive a raise while the national average declines, or see no raise while the average increases because the job mix changed.
The BLS itself describes average earnings as averages across jobs, rather than the earnings of a typical worker. The series does not adjust for differences such as occupation, age, or schooling, and the jobs counted include part-time work and jobs held by people with more than one job. It is useful for tracking a broad labor-market measure, but it cannot identify an individual worker’s raise. Bureau of Labor Statistics
What September’s figures do and do not say
In September 2026, average hourly earnings for private nonfarm employees were $37.81, up 3.0 percent from a year earlier. The real-earnings report, which adjusts the measure using the Consumer Price Index for All Urban Consumers, found that real average hourly earnings fell 0.7 percent over the same period. The nominal measure tracks dollar pay; the real measure accounts for measured price changes. They are distinct readings of the aggregate. Bureau of Labor Statistics Bureau of Labor Statistics
Neither figure says that every worker’s pay rose 3 percent or fell 0.7 percent. The reports also describe different movements over shorter periods: nominal average hourly earnings edged up 0.1 percent from August, while real average hourly earnings declined 0.5 percent, reflecting that small pay increase alongside a 0.6 percent monthly CPI-U increase. Those are economy-wide averages, not estimates of each household’s bills or each worker’s wage history. Bureau of Labor Statistics Bureau of Labor Statistics
Read the labor-market picture in layers
Employment composition is one reason to read average pay alongside other indicators. The September report counted 29,000 additional nonfarm payroll jobs, while the unemployment rate was 4.2 percent. Employment changed little in most major industries, though health care employment continued to trend upward. These data describe jobs and labor-force status; they do not explain by themselves why the average wage moved or what happened to a given worker. Bureau of Labor Statistics
A useful mental check is to ask whether a statistic follows the same people over time or combines whoever is in the measured group at each point. The wage average does the latter. A rising average can reflect individual raises, a shift toward better-paid industries, or both; a falling average can reflect weaker pay growth, a changing job mix, or both. Without a measure that holds the composition constant, the headline average alone cannot separate those forces. It remains a meaningful snapshot of aggregate payrolls, but not a personal pay report. Bureau of Labor Statistics
Sources & further reading
- September 2026 employment report ↗Bureau of Labor Statistics
- September 2026 real earnings ↗Bureau of Labor Statistics
.jpg){kind=link}