How markets and economies work Lesson 5 of 7
Economics · Foundations
Productivity: output per hour, unit labor costs and living standards
How US labor productivity and unit labor costs are measured, how fast productivity grew in each era since 1948, and why it sets long-run wages and living standards.
On this page
Over the long run, almost everything an economy can afford comes down to one number: how much it produces for each hour people work. That is labor productivity. When it grows 2% a year, output per hour doubles in about 35 years, and with it the economy's capacity to pay wages, profits and taxes. When it grows 1%, doubling takes about 70.
In the US nonfarm business sector, output per hour was 5.3 times higher in the third quarter of 2026 than in early 1948, according to the Bureau of Labor Statistics. The pace has varied a great deal: fast in the post-war decades, slow from the early 1970s, fast again around the turn of the century, slow after 2004. This lesson explains how productivity and its close cousin, unit labor costs, are measured, what the eras show, how productivity connects to pay and inflation, and why the measurement is harder than it looks.
What productivity measures
BLS's headline measure covers the nonfarm business sector. Its handbook explains that business output leaves out general government, private households, nonprofit institutions serving households and the value of owner-occupied housing, activities where output is hard to measure separately from what is paid for it; the nonfarm measure also leaves out farms. Hours worked include paid employees, the self-employed and unpaid family workers.
Labor productivity is not a measure of how hard people work. It rises when workers have more and better equipment (capital deepening), when the workforce is more skilled, and when the economy finds better ways to combine labor and capital (total factor productivity, the part that reflects technology and organization).
Worked example: unit labor costs
What the data show
Our calculations from the BLS indexes, compound annual rates between the quarters named:
| Era | Output per hour | Unit labor costs | Real pay per hour |
|---|---|---|---|
| Q1 1948 – Q4 1973 | 2.67% | 2.48% | 2.54% |
| Q4 1973 – Q4 1995 | 1.50% | 4.42% | 0.73% |
| Q4 1995 – Q4 2004 | 3.12% | 1.31% | 2.02% |
| Q4 2004 – Q4 2019 | 1.55% | 1.14% | 0.67% |
| Q4 2019 – Q3 2026 | 2.22% | 2.77% | 1.04% |
The era boundaries follow the usual reading of US productivity history:
- 1948–1973: the post-war boom. At 2.67% a year, output per hour doubled in about 26 years.
- 1973–1995: the slowdown. Growth fell to 1.50% a year, a doubling time of about 47 years. Economists have debated the causes for decades; the oil shocks of the 1970s, a shift toward services and slower technical progress are the usual suspects, and no single cause has won. Unit labor costs rose 4.42% a year, part of the inflation of the 1970s and early 1980s (see Stagflation: high inflation, weak growth and the 1970s lesson).
- 1995–2004: the IT revival. Growth jumped to 3.12% a year, a period widely linked to falling computer prices and businesses reorganizing around information technology.
- 2004–2019: the second slowdown. Back to 1.55% a year, across a financial crisis and a long recovery.
- Since 2019. Output per hour has grown 2.22% a year through Q3 2026, faster than the previous fifteen years. Whether that lasts is an open question; seven years is short, and it includes the pandemic distortions described below.
The quarterly chart shows why eras matter more than single readings. In the year to Q2 2020, output per hour rose 6.6% because hours worked collapsed even faster than output during the lockdowns. Two years later, in the year to Q2 2022, it fell 2.1% as hours came back. Neither number said much about the economy's long-run efficiency. Unit labor costs, meanwhile, rose as much as 12.7% over four quarters in the year to Q3 1974, and 6.1% in the year to Q1 2022, the highest since 1982. In the latest data, Q3 2026, output per hour was up 2.1% on a year earlier and unit labor costs 2.4%. Note the convention: the BLS headline reports the change from the previous quarter at an annual rate, which for Q3 2026 was 3.6% (our calculation from the index: 3.55%), not the year-over-year change.
Why productivity sets living standards
In the long run, an economy can only consume what it produces. If output per hour does not rise, higher real wages for everyone can come only from working more hours or from taking a larger share of output from profits or the government, and neither can continue forever. That is why economists treat productivity growth as the main source of lasting gains in living standards, and why the difference between the slow and fast eras matters so much: compounded over 25 years, 1.50% a year raises output per hour about 45%, and 2.67% a year about 93%.
The productivity–pay gap
In the BLS data, real hourly compensation grew more slowly than output per hour after 1973: from Q1 1948 to Q3 2026, output per hour rose about 5.3-fold and real pay per hour about 3.2-fold. Part of this gap is real: labor's share of income fell in some periods. Part comes from measurement. A BLS study of 183 industries (Brill and others, 2017) found that deflating pay with consumer prices while output is deflated with the industry's own prices accounts for much of the gap; switching to the output deflator narrowed the gap in 87% of the industries where productivity had outpaced pay. Compensation also includes benefits such as health insurance, and an average can rise while the typical worker's pay does not.
Measurement problems
- Quality and new goods. A smartphone in 2026 does far more than one from 2010 at a similar price. Price indexes adjust for quality, but imperfectly, and errors flow straight into real output and productivity.
- Free and intangible products. Search engines, maps and online services that cost users nothing add little to measured output, whatever their value to users.
- Services are hard to measure. What is a unit of health care, education or financial advice? Where output is hard to count, measured productivity growth is often weak.
- Hours. Self-employed hours, unpaid overtime and remote work make hours uncertain.
- Revisions. Each quarterly figure is revised as GDP and hours data are revised (see Seasonal adjustment and data revisions: why the numbers change).
- Composition. When low-output jobs disappear in a recession, average output per hour rises without anyone becoming more productive, as in 2020.
Productivity and related measures
| Measure | Formula | Tells you |
|---|---|---|
| Labor productivity | Real output ÷ hours | Efficiency of each hour worked |
| Unit labor costs | Hourly compensation ÷ productivity | Labor cost per unit of output |
| Real hourly comp. | Pay per hour, adjusted for prices | What an hour's pay buys |
| Total factor productivity | Output growth not explained by labor and capital | Technology and organization |
| Potential GDP | Sustainable output at full use | See Potential GDP and the output gap: how far the economy is from capacity |
Productivity growth is a key input to potential GDP: an economy's sustainable growth rate is roughly the growth of hours worked plus the growth of output per hour.
How Finabase does it
Finabase does not chart productivity or unit labor costs; the figures on this page are our own calculations from BLS series. In the app:
- The Economy page lists scheduled US data releases with the forecast, the actual and the previous figure, and charts CPI inflation, unemployment and the Fed funds rate, which productivity and unit labor costs feed into.
- The Economy news topic collects coverage when the quarterly Productivity and Costs report comes out.
Frequently asked questions
What is labor productivity?
Labor productivity is real output divided by hours worked: how much the economy produces for each hour of work. BLS publishes it quarterly for the US nonfarm business sector, where it grew 2.14% a year on average from 1948 to Q3 2026.
What are unit labor costs?
Unit labor costs are what businesses pay in compensation for each unit of output, equal to hourly compensation divided by output per hour. They rise when pay grows faster than productivity, which is why they are watched as a sign of inflation pressure.
Why did US productivity growth slow after 1973?
There is no single agreed answer. The oil shocks, a shift toward services that are harder to make more efficient (and to measure), and slower technical progress are the usual explanations. Growth fell from 2.67% a year in 1948–73 to 1.50% in 1973–95.
Does higher productivity mean higher wages?
Over long periods productivity sets the ceiling for real wage growth, and the two moved closely in 1948–73. Since then real pay per hour has grown more slowly in BLS data, partly because of different price deflators and partly because labor's share of income fell.
Summary
Labor productivity is real output per hour worked, and unit labor costs are pay per unit of output. In the US nonfarm business sector output per hour grew 2.67% a year in 1948–73, 1.50% in 1973–95, 3.12% in 1995–2004 and 1.55% in 2004–19, and 2.22% a year since 2019. The pace decides how fast living standards and sustainable wages can rise, and together with pay growth it sets unit labor costs, a source of inflation pressure. Quarterly readings swing with the cycle and measurement is hard, so the useful view is the long one.
Check your understanding
4 questions on this lesson. Each answer comes with the reason, and nothing is scored.
Hourly compensation rises 5% this year and output per hour rises 1%. Roughly how much do unit labor costs rise?
- About 6%
- About 4%
- About 1%
Show the answer
B. About 4% Unit labor costs are compensation divided by productivity, so their growth is about pay growth minus productivity growth: 5 − 1 ≈ 4% (precisely 1.05 ÷ 1.01 − 1 ≈ 3.96%).
In the year to Q2 2020, US output per hour jumped 6.6%. What is the best reading?
- The pandemic made workers far more efficient
- Hours worked fell even faster than output during the lockdowns, so output per hour rose mechanically
- A technology boom began that quarter
Show the answer
B. Hours worked fell even faster than output during the lockdowns, so output per hour rose mechanically Productivity is output divided by hours. When hours collapse faster than output, the ratio rises without any improvement in efficiency, and it reversed as hours returned.
Productivity grows 2.67% a year in one era and 1.50% in another. Over 25 years, how different is the rise in output per hour?
- About 93% versus about 45%
- About 67% versus about 38%
- There is little difference over 25 years
Show the answer
A. About 93% versus about 45% Compounding: 1.0267^25 ≈ 1.93 and 1.015^25 ≈ 1.45. A gap of about one point a year nearly doubles the gain over a generation.
A chart shows output per hour rising much faster than real pay per hour since 1973. Which is a valid caution?
- The data must be wrong, because pay always equals productivity
- Pay and output are adjusted with different price indexes; a BLS study found that using the output deflator narrows much of the gap
- Productivity data include only manufacturing
Show the answer
B. Pay and output are adjusted with different price indexes; a BLS study found that using the output deflator narrows much of the gap Part of the measured gap comes from deflators (consumer prices versus output prices), and part from a falling labor share. Both are real issues; neither alone tells the whole story.
Exercise
One task to do in Finabase. Nothing is graded or saved.
Open the Economy page and look at the CPI inflation chart. The lesson says unit labor costs, pay per unit of output, rose 6.1% over four quarters in early 2022, the most since 1990, and 2.4% in the year to Q3 2026. Find where CPI inflation was in early 2022 and where it is in the latest months. Did consumer inflation move in the same direction as unit labor cost growth?
Open the Economy pageWhat you should find
CPI inflation was high and rising in early 2022, peaking that year, and is far lower in the latest months, the same direction as unit labor costs. That fits the lesson: when pay grows faster than output per hour, firms face pressure to raise prices. Energy and demand drove 2022 too, so the two moving together is consistent with the link, not proof of it.
Sources and further reading
- BLS — Nonfarm business sector: real output per hour of all persons (FRED series OPHNFB), downloaded 2026-11-18
- BLS — Nonfarm business sector: unit labor costs for all workers (FRED series ULCNFB), downloaded 2026-11-18
- BLS — Nonfarm business sector: real hourly compensation for all workers (FRED series COMPRNFB), downloaded 2026-11-18
- BLS — Productivity and Costs program, retrieved 2026-11-18
- BLS Handbook of Methods — Productivity measures: business sector and major subsectors, concepts, retrieved 2026-11-18
- Brill, M., Holman, C., Morris, C., Raichoudhary, R. and Yosif, N. — Understanding the labor productivity and compensation gap, BLS Beyond the Numbers 6(6), June 2017
