US public companies’ inflation-adjusted revenue per worker grew faster after 2019, and profit margins widened alongside it, according to an October analysis by a New York Fed economist. The result raises a question that a productivity headline cannot answer on its own: how much of the gain reaches employees?
Jeremy Pearce’s October 15 analysis puts annual growth in real revenue per worker at 2.2% for 2019 to 2025, compared with 0.7% for 2010 to 2019.
Sales-weighted gross margins rose from about 34% in 2019 to 37% in 2025. Operating margins increased from roughly 16% to 18%. These are increases of three and two percentage points respectively, not 3% and 2% increases in profits.
Revenue, output and hours measure different things
Revenue per worker divides a company’s sales by its employee count. Adjusting those sales for industry price changes removes some of the effect of inflation. It still differs from the output-per-hour measure used in official productivity statistics.
The US Bureau of Labor Statistics explains why hours matter. Two businesses with the same employee count can use different amounts of labor if one relies more heavily on part-time work or longer schedules.
A company can also change its reported revenue per employee by moving work to an outside supplier. Its employee count falls, but people elsewhere still perform the work. Managers comparing businesses need to examine what is done inside the company and what is purchased from contractors.
For a hypothetical calculation, sales of $10 million divided among 100 employees produce revenue per employee of $100,000. If the employee count falls to 80 while sales stay unchanged, the ratio becomes $125,000. The arithmetic alone says nothing about hours, outsourcing costs, product quality or employee pay.
Margins do not reveal every claim on the proceeds
A gross margin is the share of sales left after direct production costs. An operating margin also deducts operating expenses. Our guide to profit margins explains why neither measure is interchangeable with revenue.
A wider margin may give a company more room to invest, reduce debt or distribute earnings. It is not a direct measurement of the amount of cash available after every obligation. Nor does a rising margin establish that wages fell.
Pearce’s sample covers public manufacturing and services businesses, excluding several sectors where measurement is difficult, including finance and real estate. Higher research spending relative to sales was associated with greater subsequent margin growth. The analysis does not establish whether the additional margins are investment returns or windfalls.
Company results also need to be kept separate from an economy-wide measure. Listed firms are not the whole business population, and changes in their average margin cannot be applied to every small employer.
Worker compensation needs its own measurement
The BLS’s September 3 productivity release reported that nonfarm-business labor productivity was 2.2% higher in the second quarter of 2026 than a year earlier. Real hourly compensation was 0.1% lower over the same period.
Real compensation adjusts wages and benefits for consumer-price changes. A worker can receive a higher dollar payment yet have no increase in purchasing power if prices rise faster.
The release put labor’s share of output at 52.8%, the lowest reading in a series beginning in 1947. Labor share measures the part of output paid as compensation; it is not the percentage of companies raising wages or a measure of any individual’s salary.
Those national figures cover a different population and period from the public-company study. Taken together, they show why an assessment of productivity, business returns and living standards needs separate measures of output, costs and compensation.
For employers evaluating a new machine or software system, the BLS measurement framework offers another check: unit labor costs compare compensation with real output. Faster output growth can contain the labor cost of each unit even while hourly pay rises. A bigger sales-per-worker number cannot establish that result by itself.
