Productivity measures how much output is produced relative to the inputs used. Labour productivity, the measure often meant in macroeconomic releases, is real output per hour worked. For a quant analyst, that distinction matters: more production can come from more working hours, more output from each hour, or both. A rising output series alone cannot tell you which occurred.
The BLS productivity concepts distinguish labour productivity from total factor productivity, also called multifactor productivity. The denominator determines the question being answered.
Measure output per hour
Let be real output during a period and the total hours worked to produce it. For , labour productivity is
Use matching periods and coverage. Whole-economy GDP and hours for one industry do not form a meaningful productivity ratio. Nor does dividing nominal revenue by headcount reproduce real output per hour: revenue includes price changes, while headcount conceals differences in hours worked.
For aggregate goods and services, statisticians use real-output volume measures rather than adding unlike physical items. BLS uses real value added for large sectors such as nonfarm business, and real sectoral output for manufacturing sectors and detailed industries. Published productivity series are often indices; an index level is not a count of physical units per hour. See the BLS calculation method.
Consider a fictional producer making one unchanged product in two equal-length periods:
| Measure | Earlier period | Later period |
|---|---|---|
| Real output, units of the same product | 100 | 110 |
| Total labour hours | 50 | 50 |
| Labour productivity, units per hour | 2.0 | 2.2 |
The productivity change is
Fictional equal-period example. The bars show output per hour, not total output; labour hours are unchanged.
When both numerator and denominator move, use the exact growth identity:
Subtracting the hours growth rate from the output growth rate is only an approximation. The ratio above is exact for these matched measures.
What the increase does—and does not—show
The producer now obtains more output from each labour hour. That does not establish that each worker is exerting more effort. Better equipment, software, skills or production organisation can raise output per hour. More capital services per hour is capital deepening: labour productivity can increase because each hour works with more or better capital. The OECD definition explicitly warns against interpreting the measure as personal effort alone.
Composition also matters. If a larger share of total hours moves to activities with higher output per hour, aggregate productivity can rise even with no improvement within either activity. A change in the workforce's education and experience can also affect measured labour input. These are reasons to investigate a headline increase before describing it as a technological breakthrough.
Multifactor productivity (MFP), or total factor productivity (TFP), compares real output with a combined input measure. Depending on the output concept and statistical framework, those inputs include labour and capital, and may include energy, materials and purchased services. Its purpose is to separate output growth from the contributions of measured inputs. It is not obtained simply by replacing hours with a headcount, and its residual is not a pure observation of technology. The RBA explainer provides the distinction and its measurement limits.
Reading a macro release
Productivity helps interpret economic growth, but a short-run rise need not establish a lasting improvement in potential output. Ask whether output grew, hours fell, or the mix of production changed. The same headline rate can accompany very different economic conditions.
For a research dataset, retain the country, sector, output concept, hours coverage, frequency, seasonal-adjustment convention and publication date alongside the value. Check the rate convention: BLS quarterly labour-productivity growth is expressed at an annual rate, whereas its annual growth compares annual periods. Those are different comparisons. Use the release vintage available at each forecast date, rather than silently giving an earlier decision access to later revisions. The BLS calculation documentation explains its period and seasonal-adjustment conventions.
Check the denominator
A second fictional producer's real output falls from 80 to 72 units, while total hours fall from 40 to 30 over equal-length periods. Did labour productivity fall?
Answer. No. Output per hour rises from to units per hour. Its growth is . Output fell by , but hours fell by . Fewer goods were produced; each remaining hour produced more. Neither result by itself identifies the cause or proves that workers became more skilled.
Sources
- BLS — Office of Productivity and Technology: Concepts: labour productivity, TFP, input contributions and measurement framework.
- BLS — Office of Productivity and Technology: Calculation: real output per hour, output concepts, growth-rate conventions and seasonal adjustment.
- OECD — Labour productivity and utilisation: other inputs, composition and the limits of interpreting personal effort.
- RBA — Productivity: labour versus multifactor productivity, measurement limits and economic interpretation.