Wage inflation is the rate at which nominal wages rise over a stated period. It measures the changing price of labour before adjusting for consumer-price inflation. For a quant tracking household demand, company margins or interest-rate expectations, the first question is which pay measure changed: an hourly wage rate, average earnings or total compensation.
A wage level is an amount per hour or another specified unit. Wage growth compares that level with an earlier observation. Slower positive growth still means wages are rising; it does not mean pay has fallen.
The pay measure changes the story
| Measure | What it answers | Main caution |
|---|---|---|
| Wage or salary rate | How much is paid for a specified unit of labour? | State hourly, weekly or annual units and what payments are included. |
| Average hourly earnings | What is aggregate payroll divided by aggregate paid hours? | The workforce and hours mix can move the average without anyone receiving a raise. |
| Wage-price or fixed-composition cost index | How has pay changed after controlling for specified job or employment characteristics? | An index level is not a currency wage level; read the exact component and coverage. |
| Total compensation | How are wages and salaries plus employer benefit costs changing? | Benefits such as health insurance and retirement contributions are not the same as cash pay. |
The US Current Employment Statistics survey publishes monthly average hourly earnings for private-sector workers. BLS defines this as payroll divided by paid hours. It reflects pay rates, overtime and other included payments, as well as shifts between higher- and lower-paid work; it excludes employer benefits. Weekly earnings also depend on hours, so rising hourly pay does not guarantee rising weekly income. See the BLS CES concepts.
The quarterly US Employment Cost Index controls for shifts in employment across industries and occupations. It publishes separate wages-and-salaries, benefits and total-compensation components. Choose the component that matches the question; a compensation headline is broader than cash-wage growth. See the BLS ECI guide.
Australia's quarterly Wage Price Index follows comparable jobs and holds price-determining characteristics constant. Its indexes excluding bonuses remove changes in work quality, quantity and labour-market composition; bonuses-inclusive indexes have a different interpretation. The June 2026 methodology identifies total hourly rates excluding bonuses as the seasonally adjusted WPI series. These are different national measurement systems, so check coverage and components before comparing their growth rates. See the ABS methodology.
A raise can exceed inflation without adding the full difference
Consider a fictional one-year comparison. An unchanged job's gross hourly wage rises from 20 to 21 currency units. Over those same dates, a matching consumer-price index rises from 100 to 103. The inputs are illustrative, with no seasonal adjustment or annualisation applied.
Nominal wage growth is . Consumer prices rise by . To find the change in purchasing power per paid hour, divide wage growth factors:
Here is consumer-price inflation over the same period. The exact real growth rate is ; the final decimal and percentage above are rounded. Subtracting the rates gives a 2 percentage-point gap, commonly used as an approximation to real wage growth. The exact purchasing-power gain is about 1.94%. A consumer price index represents a specified consumer basket, not every worker's personal spending pattern. This calculation also leaves taxes, hours and benefits outside its scope.
Fictional one-year comparison. Bars measure percentage growth over identical dates; real wage growth divides the two growth factors rather than subtracting the bars.
From wages to prices: an economic question
Wages matter for production costs, but wage growth alone does not establish a wage-price spiral. Productivity, the labour share of costs, other input prices and firms' margins affect how higher compensation translates into selling prices. Wages may also respond to earlier price increases. A causal claim needs evidence about timing and mechanisms, beyond two rising series. The San Francisco Fed's labour-cost analysis illustrates why pass-through is an empirical question rather than an automatic one-for-one relationship.
For a backtest, preserve the series identifier, component, jurisdiction, units, reference period, seasonal-adjustment status and release vintage. A monthly rate, a quarterly rate and a year-over-year rate are different comparisons. Use the value available when the signal was formed, including the release timestamp; later revised data can change what the model appears to have known.
Check the composition effect
In a fictional group, one worker earns 12 currency units per hour and another earns 24. Each is paid for one hour. The lower-paid worker then leaves; the remaining worker's rate and paid hour are unchanged. Has everyone received a 33.3333% raise?
Answer. No. Average hourly earnings initially equal . They then equal , a change of after rounding. The remaining worker's wage growth is . The average rose because the composition changed. A fixed-composition pay measure is designed to address this different question.
Sources
- BLS — Current Employment Statistics: Concepts, especially aggregate payroll, paid hours and key earnings estimates.
- BLS — Understanding index numbers and percent changes in the ECI, especially components, comparison periods, fixed weights and constant-dollar estimates.
- ABS — Wage Price Index methodology, June 2026, especially index design, wage-price indexes and seasonal adjustment; released 19 August 2026.
- Federal Reserve Bank of San Francisco — How Much Do Labor Costs Drive Inflation?, 30 May 2023. A dated empirical study, not a universal pass-through coefficient.