What Are Unit Labor Costs? Formula, Productivity, and Inflation

Unit labor cost asks how much labor compensation businesses incur for a unit of output. At an economy-wide or industry level, it is commonly expressed as nominal labor compensation divided by real output. The same relationship can be written as hourly compensation divided by labor productivity, where productivity means real output per hour worked. The U.S. Bureau of Labor Statistics (BLS) uses both expressions in its productivity concepts guide and calculation methods.

The word unit matters. A company can pay more per hour without having the same increase in labor cost for each unit it makes if each hour also produces more output. Conversely, compensation can remain nearly unchanged while unit labor costs rise if output per hour falls. ULC therefore combines a price paid for labor services with a measure of production. It is not a worker's hourly wage, a company's total payroll bill, or the price of a finished product.

Published ULC figures are often index numbers. The index tracks change from a selected reference period; an index value of 125 does not mean that labor costs rose 25 percent in the latest year. To discuss inflation pressure, analysts usually focus on the index's rate of change, identify the period being compared, and specify the sector or industry covered.

Let C be nominal labor compensation per hour and P be real output per hour. Then:

Unit labor cost = C ÷ P

If compensation per hour rises 4 percent and output per hour rises 4 percent over the same period, their ratio is broadly unchanged. If compensation rises faster than productivity, ULC rises; if productivity rises faster, ULC falls. For small changes, people often summarize the relationship as ULC growth ≈ compensation growth − productivity growth. This subtraction is a useful approximation in percentage points, not the exact calculation for every size of change.

The exact relationship uses growth factors:

1 + ULC growth = (1 + compensation growth) ÷ (1 + productivity growth)

Use decimals in the equation. If compensation grows 5 percent, enter 0.05; if productivity grows 2 percent, enter 0.02. The ratio compares two growth paths over the same interval. Mixing a quarterly compensation rate with annual productivity growth would not produce a meaningful ULC calculation.

The index form used by BLS has the same economic idea: current-dollar labor-compensation indexes are compared with real-output indexes. Depending on the sector or industry, BLS may use real value-added output or a sectoral output measure. Those definitions affect what counts as output, so two series called “unit labor costs” should not be compared until their coverage and output concepts are clear.

Suppose hourly labor compensation rises from $30 to $31.50, an increase of 5 percent. During the same period, real output per hour rises from 100 to 102 units, an increase of 2 percent. The starting ULC is $30 ÷ 100 = $0.30 per unit. The new ULC is $31.50 ÷ 102, or about $0.3088 per unit. The increase is:

(1.05 ÷ 1.02) − 1 = 0.0294, or about 2.94 percent

Subtracting the two growth rates gives 5 − 2 = 3 percentage points, a close approximation here. It is not exactly 3 percent because the denominator also grew. If hourly compensation and productivity both rose 5 percent, ULC would be unchanged: 1.05 ÷ 1.05 = 1. A faster paycheck increase therefore does not, by itself, show that labor cost per unit rose by the same amount.

The example holds the definition of output and the measurement period constant. Actual official indexes combine information across workers, establishments, and industries; they are not calculated from one worker or one production line. The arithmetic explains the relationship, while the published index supplies the measured aggregate.

In its revised second-quarter 2026 release, BLS reported that nonfarm business unit labor costs rose 1.2 percent from the prior quarter at an annual rate and 1.4 percent from the same quarter a year earlier. For the quarterly comparison, hourly compensation rose 2.6 percent at an annual rate while productivity rose 1.4 percent. Because the published components are rounded, recomputing from the displayed values gives an approximate result: 1.026 ÷ 1.014 − 1 is about 1.18 percent, which rounds to 1.2 percent. The release is a dated data example, not a standing estimate or forecast.

The two reported ULC growth rates answer different questions. The 1.2 percent figure annualizes the change between adjacent quarters as if that pace compounded for a full year; it does not say costs actually rose 1.2 percent over the year. The 1.4 percent year-over-year figure compares the second quarter with the second quarter of 2025. A short-lived quarterly movement can therefore look different from the accumulated change over four quarters.

The BLS release also revised the quarterly ULC estimate from 1.3 percent to 1.2 percent after updating the underlying data, with a 0.1 percentage-point downward revision to hourly compensation. When reading a headline, check whether it is preliminary or revised and whether “annual rate” refers to a one-quarter change. The release tables show both the quarter-to-quarter annualized rate and the year-over-year rate.

Hourly compensation is an employer-cost concept. In BLS measures it can include wages and salaries, bonuses, paid leave and other payroll items, employer social-insurance contributions, and benefits such as health coverage or retirement contributions. The exact components depend on the series and sector. A worker's take-home pay is after taxes and deductions; it is not the numerator in the ULC formula.

This distinction can change the interpretation. If employer health-insurance or pension costs rise while cash wages are steady, compensation per hour may still increase. If analysts substitute an average hourly earnings series for total compensation, they may miss benefits and use a different worker or establishment coverage. The measures can be useful together, but their labels and definitions should remain attached to the comparison.

BLS productivity measures can cover different sectors, including business, nonfarm business, manufacturing, and nonfinancial corporate business. A sector's aggregate compensation per hour is not necessarily the raise received by a typical employee in that sector. Changes in industry mix, self-employment coverage, hours, and source data can affect the aggregate series.

Labor productivity is real output divided by hours worked. It can change because of technology, worker skills, equipment and software, management, how production is organized, scale, capacity use, energy and material inputs, or the mix of products and workers. BLS describes these influences in its methods for labor productivity. A rise in output per hour cannot be attributed to greater personal effort from the productivity statistic alone.

This also explains why one employee's output target is not the same measure as economy-wide labor productivity. Aggregate output may include quality adjustment and value-added concepts, while hours can include different groups of workers. In some sectors, a measure of output per hour may move sharply when production changes faster than recorded hours. The ratio is meaningful only with its statistical boundary and data method.

The denominator matters for ULC because a productivity gain can offset some compensation growth. But that does not mean workers receive no benefit from productivity, or that every productivity gain immediately lowers product prices. How the gains are divided among compensation, profits, investment, and prices is a separate economic question.

When compensation grows faster than productivity, ULC can put upward pressure on a producer's labor cost per unit, all else equal. But labor is only part of total cost. BLS separately describes unit nonlabor costs, which can include capital consumption, energy and materials, taxes less subsidies, interest and other payments. A fall in labor cost per unit can coincide with rising total costs if these other components increase.

Even a change in production cost does not map one-for-one into the Consumer Price Index (CPI) or Personal Consumption Expenditures (PCE) price index. Businesses may absorb some cost in margins, change product mix, use contracts with a lag, improve other inputs, or pass some of the change to buyers. Imports, rents, taxes, distribution costs, and demand also affect consumer prices. The CPI and PCE cover consumer baskets; ULC tracks labor cost relative to output in a defined production sector.

For that reason, rising ULC alone does not prove that a wage-price spiral is under way, that firms will raise prices by the same percentage, or that inflation will accelerate. Falling ULC does not guarantee consumer prices will fall. Use ULC with output prices, nonlabor costs, productivity, compensation, margins, and the relevant price index rather than treating it as a standalone inflation forecast.

The phrase “U.S. unit labor costs” can refer to a particular BLS sector series, not every U.S. business. Nonfarm business excludes farming; manufacturing uses a different output concept and can move differently; nonfinancial corporate business has another boundary. In the second quarter of 2026, for example, BLS reported a 1.2 percent annualized quarterly increase in nonfarm business ULC while manufacturing ULC declined 0.3 percent on the same basis. The different readings are not contradictory: they describe different groups and components in the BLS release tables.

Output itself may be measured as real value added or as real sectoral output. Value added removes purchased intermediate inputs such as materials, energy, and services; sectoral output follows a different boundary. Because the numerator and denominator must refer to compatible activity, changing the output concept changes the measure. BLS's calculation guide documents these distinctions and separately presents unit labor and unit nonlabor costs.

For cross-country comparisons, check currency treatment, industry coverage, hours worked, compensation components, and output definitions. An index rebased to 100 in one country is not automatically an absolute cost comparison with an index rebased in another. ULC growth can help study cost trends, but it is not by itself a complete competitiveness ranking or a measure of worker welfare.

Quarterly estimates use several data sources, and some source series are updated after an initial release. BLS explains that its productivity estimates combine output information from the Bureau of Economic Analysis with employment and hours data from BLS surveys. The productivity and costs release is published first and then revised as more complete inputs arrive. Its revision study describes the schedule and the historical range of revisions; a small first-release change should not be treated as a final result.

Before drawing a conclusion from a ULC headline, identify: the exact sector or industry; whether compensation includes benefits; the output and hours definitions; the period and price basis; whether the rate is quarter-over-quarter annualized or year-over-year; and whether the estimate is preliminary or revised. Then compare ULC with productivity and compensation separately. If the question concerns consumer inflation, also inspect nonlabor producer costs and the relevant consumer-price measure.

This framework helps answer a narrow but useful question: did employer labor compensation per unit of measured output rise or fall over the stated period? It does not settle why the movement occurred, who gained from productivity, what a firm will charge next, or how consumer inflation will evolve. Those require additional evidence and a clearly defined economic scope.

For related context, see the guides to nominal and real wages, headline and core inflation, and Okun's law.