September U.S. Paycheck Test: Prices Outpaced Hourly Earnings
Released October 14, 2026 • Observations: September 2026 • United States
Average U.S. hourly pay rose in September, but it bought less. Nominal hourly earnings increased 0.1% while consumer prices rose 0.6%, leaving real average hourly earnings down 0.5%. The useful question is whether pay, prices and hours are moving together, rather than whether a nominal raise is positive.
These monthly figures are seasonally adjusted. The pay series covers all employees on U.S. private nonfarm payrolls.
A purchasing-power bridge
A hypothetical August hourly-pay index and price index both begin at 100.
September pay becomes 100.1; the price index becomes 100.6.
100.1 ÷ 100.6 × 100 ≈ 99.50 in August purchasing-power units.
The arithmetic explains why “wages rose” and “purchasing power fell” can both be accurate. Subtracting inflation from wage growth gives a useful approximation; dividing the growth factors is the exact mathematical operation when the underlying rates are exact.
Open the monthly calculation and its rounding limit
Real change = [(1 + nominal pay growth) ÷ (1 + price growth) − 1] × 100. With the published rounded inputs: (1.001 ÷ 1.006 − 1) × 100 = −0.4970%, or approximately −0.50%.
BLS calculates its series from underlying data, not from these one-decimal headline changes. This reconstruction is a reader aid, not a substitute for the official estimate.
Hours change the annual picture
Over the year to September, real hourly earnings fell 0.7%. The average workweek rose 0.6%, cushioning the decline in real weekly earnings to 0.1%. That is a different comparison from the monthly result: September's average workweek was unchanged from August.
On a narrow screen, scroll the table sideways. Keyboard users can focus the table region and use the arrow keys.
| Measure | September vs August 2026 | September 2026 vs September 2025 |
|---|---|---|
| Real average hourly earnings | −0.5% | −0.7% |
| Average workweek | Unchanged | +0.6% |
| Real average weekly earnings | −0.5% | −0.1% |
Open the hourly-to-weekly calculation
Using the rounded annual changes, 0.993 × 1.006 = 0.998958. That implies an approximate weekly real-pay change of −0.1042%, which rounds to −0.1%.
More hours can support weekly purchasing power without improving purchasing power per hour. These aggregate averages do not establish that the same people worked longer or received particular raises.
What the inflation detail adds
Gasoline prices rose 9.1% in September on a seasonally adjusted basis. BLS attributes more than half of the monthly all-items increase to gasoline. Core CPI, which excludes food and energy, rose 0.3%. The energy shock matters for the headline, but it does not mean every price category or household budget moved by the same amount.
The 12-month headline CPI increase was 3.7%, measured before seasonal adjustment. Do not insert that rate into the annual earnings bridge above: the earnings release uses seasonally adjusted annual changes. Matching periods and adjustment conventions is as important as choosing the formula.
How to use this result
For a household comparison, separate hourly compensation, paid hours and the prices of what the household actually buys. A commuter, a remote worker and a renter may face different spending patterns. For an employer, the same report can inform a purchasing-power discussion, but it cannot establish the appropriate raise for a particular occupation or location.
These aggregate averages are influenced by the mix of high- and low-wage employment. They are not the pay history of a typical individual, a household disposable-income measure, or a forecast of future inflation. A single monthly decline is insufficient evidence for a lasting trend.
Sources and measurement
- BLS, September 2026 CPI release, October 14 (current release page): opening discussion and Table A.
- BLS, September 2026 Real Earnings: all-employees results, released October 14 at 8:30 a.m. Eastern Time.
- BLS Real Earnings technical note: coverage, price deflators, adjustment and composition limits.
Source vintage checked October 14, 2026. Calculations and diagram are original. Current-release links may advance to a later month; the figures here refer to September 2026.
