When higher pay does not mean higher unit labor costs

The question behind a wage headline

Can pay rise without an equally large increase in the labor cost of producing goods and services? Yes. The missing variable is output per hour. A productivity gain can absorb part of a compensation increase, but it does not automatically establish stronger demand, lower consumer inflation or higher profits for a particular company.

This is a guide to interpreting the relationship, not a new economic release. The Bureau of Labor Statistics' September 3, 2026 revision of second-quarter productivity provides a dated example: nonfarm business productivity rose at a 1.4% seasonally adjusted annualized rate, while hourly compensation rose 2.6% and unit labor costs rose 1.2%. The four-quarter increase in unit labor costs was separately reported as 1.4%. Keep those comparison periods attached to the numbers.

The practical task is to construct a worksheet in which compensation, hours and real output share the same population and period. Then test what the result can support before using it in a company research note.

Start with a ratio, not a subtraction

BLS defines unit labor cost as labor compensation divided by real output. An equivalent expression divides compensation per hour by output per hour. Compensation includes benefits as well as wages. The denominator is real output, not nominal sales inflated by a higher selling price.

Consider an explicitly hypothetical operation with hourly compensation of $30 and production of five identical units per hour. Its labor cost is $6 per unit. In the next matching period, compensation rises 5% to $31.50, while output per hour rises 3% to 5.15 units. Labor cost becomes approximately $6.1165 per unit, an increase of 1.9417%.

The exact calculation is 1.05 divided by 1.03, minus one. Subtracting 3% from 5% produces a useful approximation of two percentage points, but not the exact unit-cost growth rate. Keep full precision in the worksheet and round only the displayed answer. Do not change an official rounded statistic because your calculation from other rounded headline numbers differs slightly.

This small model holds product quality and the definition of a unit constant. It is an illustration of the identity, not a reconstruction of BLS's aggregate index methodology.

Ask what happened to both output and hours

A larger output-per-hour ratio does not require total output to rise. In a second hypothetical example, output falls from 100 identical units to 99, while hours worked fall from 20 to 18. Productivity moves from five units per hour to 5.5, a 10% increase, even though output declines 1%.

If compensation remains $30 per hour, the total labor bill falls from $600 to $540. Unit labor cost falls from $6 to approximately $5.4545, or 9.0909%. That is a real change within the invented example, but it is not evidence of expanding demand. A note celebrating only the productivity increase would conceal the output contraction.

Write three separate lines: output growth, hours growth and their implied productivity growth. This makes the explanation inspectable. Ask whether the observed improvement reflects a growing business using its existing workforce more effectively, a contraction in hours faster than output, or another combination. The ratio alone cannot identify the cause.

BLS's methods also distinguish hours worked from hours paid and adjust the latter when constructing its productivity measures. Replacing that denominator with a company's year-end employee count creates a different measure. Revenue per employee can be useful, but label it separately rather than presenting it as the official output-per-hour series.

Keep annualization and purchasing power separate

The quarterly changes in the September 3 release are seasonally adjusted annualized rates. The 1.4% productivity figure is therefore not a claim that output per hour rose 1.4% within those three months. Using the displayed rounded rate, the implied one-quarter change is approximately 0.3482%: take the fourth root of 1.014 and subtract one. This conversion is illustrative because the published annualized figure is rounded.

The same release reports a 2.2% productivity increase from the corresponding quarter a year earlier. That is another comparison, not an interchangeable description of the quarter's pace. A worksheet should give annualized quarterly and year-over-year readings separate columns.

Do not silently swap inflation-adjusted compensation into a calculation defined with current-dollar compensation. BLS's unit-cost methodology uses nominal labor compensation against real output; real hourly compensation answers a purchasing-power question. A worker's purchasing power and an employer's labor cost per unit can move differently because their denominators differ.

Before comparing any two numbers, record the sector, reference quarter, release date, adjustment status and growth convention. If a later revision changes one component, refresh the matching components together instead of combining vintages.

Build a company sensitivity without inventing a margin forecast

Return to the first hypothetical operation. Suppose its selling price is $10 per unit and other costs are $2 per unit in both periods. Initial profit is $2 per unit. After the compensation and productivity changes, profit becomes approximately $1.8835 per unit, a decline of 5.8252%, even though unit labor costs rose only 1.9417%.

This is a deliberately simplified arithmetic example, excluding taxes and financing. Its purpose is to show why a cost growth percentage is not the same as a profit growth percentage. The result depends on the initial profit cushion and on the assumption that selling prices and other costs remain fixed. Changing any of those assumptions changes the answer.

For an actual company, begin with disclosed business conditions rather than inserting the national productivity rate into an earnings model. Identify which operations, countries and labor arrangements the company describes. Look for evidence about production volumes, pricing, compensation and other expenses. If comparable hours or output are unavailable, leave the productivity bridge unresolved rather than manufacturing a precise estimate.

Use the macro release to formulate a question for further research: can this business offset compensation pressure through additional output per hour, and what evidence would demonstrate that? It supplies context, not a company-specific answer.

A repeatable reading checklist

Keep a source row for each input and a calculation row for each inference. Label hypothetical assumptions explicitly. Check that compensation and output cover the same interval, that hours mean the same thing in numerator and denominator, and that annualized rates have not been mixed with simple quarterly changes.

Then write the strongest conclusion supported by the evidence and one condition that could overturn it. For example: the modeled increase in productivity offsets part of the assumed compensation increase; weaker output, a different cost mix or a selling-price decline could still hurt profit. This is more useful than treating a single favorable ratio as proof of an improving economy.

On the next productivity release, compare the revised history before interpreting the new quarter. Revisit output and hours independently, preserve the release's own precision, and update the worksheet's vintage. The related benefits-and-compensation guide explains the separate distinction between benefit coverage, cost levels and compensation-index growth.