Calendar › Productivity and unit labour costs › 2026-11-05
Productivity 3.6%, Unit Labor Costs −0.7% — Both Beat, Labour Share Hits a Record Low
Fundamentals · 2026-11-05
Nonfarm business productivity rose 3.6% annualized in the third quarter (preliminary), against a 2.8% consensus (TradingEconomics), the fastest since Q3 2025, as output rose 3.7% on hours worked up just 0.1%. Unit labour costs fell 0.7% against a 0.6% rise expected, the first decline since Q2 2025, as hourly compensation rose 2.9%. Over four quarters: productivity +2.1%, unit labour costs +2.4%. The labour share was 53.0%, the lowest in a series that begins in 1947 (BLS). Real hourly compensation +1.1%. Q2 was revised up after the annual national-accounts update: productivity to 2.1% from 1.4%, unit labour costs to 1.9% from 1.2%. Manufacturing productivity +0.3%, manufacturing unit labour costs +2.7%; nondurables productivity −1.7%.
What It Changes
- The wage leg of the inflation argument is gone for the quarter. Productivity at 3.6% against 2.8% (TradingEconomics) outran hourly pay at 2.9%, so the labour cost of each unit of output fell 0.7%.
- That does not take a December hike off the table, it narrows the reason for one. With the funds rate at 3.75-4.00% (Fed) and a hike priced near 79% this morning (TradingEconomics), the case now rests on prices and energy, not on wages.
- The hidden number is the labour share at a new record low of 53.0%. Output grew 3.7% on almost no extra hours, and the gain went to margins rather than pay: unit nonlabour payments rose 5.2% annualized while unit labour costs fell.
- Read the beat with the revision beside it. Q2 unit labour costs were revised up 0.7 points to 1.9%, so the two-quarter picture is softer than the headline drop suggests, not absent.
Impact
- USD — Mixed, lean bearish — falling unit labour costs remove a wage-driven reason for the Fed to keep hiking, while strong productivity argues for a sturdier economy and a higher neutral rate.
- This release shared 08:30 with weekly jobless claims (189,000 against 196,000 expected), which pulled the other way, so no market move can be attributed to productivity alone.
- Before the data, as of about 06:35 ET, DXY was roughly flat after a 0.5% fall on Wednesday, and the 10-year closed Wednesday at 4.90% (US Treasury).
- US Indices (ES / NQ / YM / RTY) — Slight bullish — rising output per hour with falling unit labour costs and unit nonlabour payments up 5.2% is a margin signal; the read is strongest for ES, the broadest corporate-profit index.
- NQ's morning is about the chip selling, not this release; pre-release as of about 06:35 ET: RTY +0.74%, YM +0.63%, ES +0.17%, NQ −0.46% against Wednesday's settles.
- Gold (GC) — Slight bearish — softer labour costs weaken the inflation-persistence case, though a hike priced on energy rather than wages leaves the conflict premium in place.
- Gold was +1.78% overnight before the release, on top of Wednesday's +3.37%.
Inside The Number
Nonfarm business labour productivity rose 3.6% at an annual rate in the third quarter, against a 2.8% consensus (TradingEconomics). Output rose 3.7% and hours worked rose 0.1%, so almost all of the quarter's growth came from more output per hour rather than more hours. That is the fastest quarter since Q3 2025 (4.5%). Over four quarters productivity is up 2.1%.
Unit labour costs fell 0.7%, against a consensus rise of 0.6%. Hourly compensation rose 2.9% and productivity 3.6%, and BLS spells out the mechanism: "increases in productivity tend to reduce them." It is the first quarterly decline since Q2 2025 (−1.1%). Over four quarters unit labour costs are up 2.4%, down from 2.6% in the revised Q2.
The labour share set another record low. BLS put the share of output that accrues to workers as compensation at 53.0% in the third quarter, "the lowest level in the series, which begins in the first quarter of 1947." Our September brief reported 52.8% for Q2 as the low; the annual national-accounts revisions on 30 September have since revised history, and on the new data 53.0% is the bottom. The other side of that ledger is unit nonlabour payments, +5.2% annualized and +6.0% over four quarters, against unit labour costs of −0.7% and +2.4%. Output per hour rose, the labour cost of producing it fell, and the difference went to profit, depreciation and taxes rather than pay.
Real pay did turn up. Real hourly compensation rose 1.1% after a revised −1.9% in Q2, and is up 1.0% over four quarters. The gap between nominal (2.9%) and real (1.1%) hourly compensation implies consumer prices rose at roughly a 1.8% annualized rate in the quarter, against roughly 6% in Q2. That is the arithmetic gap between two BLS figures, not a separate estimate.
The revisions are large. The annual NIPA update moved Q2 nonfarm productivity up 0.7 points to 2.1%, all from output. Q2 unit labour costs rose 0.7 points to 1.9%, because hourly compensation was revised up 1.4 points to 4.0%. Annual productivity was revised up to 2.3% in 2025 and 3.2% in 2024. The labour cost picture over two quarters is therefore +1.9% then −0.7%, not +1.2% then −0.7%.
Manufacturing went the other way. Productivity rose just 0.3% as output rose 1.3% and hours 1.0%, and manufacturing unit labour costs rose 2.7%, up 4.0% over four quarters. Durables held up (productivity +2.2% on output +3.8%); nondurables did not (productivity −1.7%, unit labour costs +5.0%). BLS notes the manufacturing output series is built differently from the nonfarm business series and the two are not directly comparable.
The AI question, one quarter on. Productivity in the current cycle, since Q4 2019, has grown at 2.2% annualized, above the previous cycle's 1.5% and equal to the long-term 2.2% rate since 1947. A 3.6% quarter on flat hours is the kind of print an AI-driven acceleration would produce, but one quarter does not move the cycle average away from its long-run norm.
The Internals
Preliminary third quarter 2026, % change from the previous quarter at an annual rate:
Sector · Labour productivity · Output · Hours worked · Hourly compensation · Real hourly compensation · Unit labour costs
Nonfarm business · 3.6% · 3.7% · 0.1% · 2.9% · 1.1% · −0.7%
Business · 4.1% · 3.9% · −0.2% · 3.1% · 1.3% · −0.9%
Manufacturing · 0.3% · 1.3% · 1.0% · 3.0% · 1.2% · 2.7%
Durable manufacturing · 2.2% · 3.8% · 1.6% · 2.8% · 1.0% · 0.5%
Nondurable manufacturing · −1.7% · −1.6% · 0.1% · 3.2% · 1.4% · 5.0%
The same measures against the year-ago quarter:
Sector · Labour productivity · Output · Hours worked · Hourly compensation · Real hourly compensation · Unit labour costs
Nonfarm business · 2.1% · 2.3% · 0.2% · 4.5% · 1.0% · 2.4%
Business · 2.1% · 2.3% · 0.3% · 4.4% · 0.9% · 2.3%
Manufacturing · 0.4% · 1.3% · 0.9% · 4.4% · 0.9% · 4.0%
Durable manufacturing · 1.9% · 3.8% · 1.9% · 5.0% · 1.5% · 3.0%
Nondurable manufacturing · −0.7% · −1.4% · −0.7% · 3.1% · −0.3% · 3.9%
Headline against consensus and prior:
Measure · Actual · Consensus (TradingEconomics) · Prior as published · Prior revised
Nonfarm productivity, q/q annualized · 3.6% · 2.8% · 1.4% · 2.1%
Unit labour costs, q/q annualized · −0.7% · 0.6% · 1.2% · 1.9%
Nonfarm business, the last six quarters, % change at an annual rate (revised data):
Quarter · Productivity · Output · Hours worked · Hourly compensation · Unit labour costs · Unit nonlabour payments
Q2 2025 · 4.2% · 5.2% · 0.9% · 3.0% · −1.1% · 5.9%
Q3 2025 · 4.5% · 4.7% · 0.2% · 4.6% · 0.1% · 8.2%
Q4 2025 · 1.2% · 1.0% · −0.2% · 6.0% · 4.7% · 1.1%
Q1 2026 · 1.4% · 2.1% · 0.7% · 5.1% · 3.7% · 4.9%
Q2 2026 · 2.1% · 2.4% · 0.3% · 4.0% · 1.9% · 13.2%
Q3 2026 · 3.6% · 3.7% · 0.1% · 2.9% · −0.7% · 5.2%
Second-quarter 2026 revisions, % change at an annual rate:
Measure · Previously published · Revised · Change
Nonfarm productivity · 1.4% · 2.1% · Up 0.7 points
Nonfarm output · 1.7% · 2.4% · Up 0.7 points
Nonfarm hours worked · 0.3% · 0.3% · Unrevised
Nonfarm hourly compensation · 2.6% · 4.0% · Up 1.4 points
Nonfarm real hourly compensation · −3.3% · −1.9% · Up 1.4 points
Nonfarm unit labour costs · 1.2% · 1.9% · Up 0.7 points
Manufacturing productivity · 2.4% · 2.9% · Up 0.5 points
Manufacturing unit labour costs · −0.3% · −0.4% · Down 0.1 points
Nonfinancial corporate productivity · 2.2% · 3.4% · Up 1.2 points
Nonfinancial corporate unit labour costs · −0.3% · −0.4% · Down 0.1 points
Annual averages, nonfarm business, after the annual update:
Year · Productivity revised · Productivity before · Unit labour costs revised · Unit labour costs before
2025 · 2.3% · 2.1% · 1.9% · 2.2%
2024 · 3.2% · 3.0% · 2.0% · 2.4%
2023 · 2.0% · 2.1% · 2.2% · 2.0%
Where The Output Went
Measure · Reading · Context
Labour share of output · 53.0% · Lowest in the series, which begins in Q1 1947 (BLS)
Unit nonlabour payments, nonfarm · +5.2% annualized · +6.0% over four quarters
Unit labour costs, nonfarm · −0.7% annualized · +2.4% over four quarters
Implicit price deflator, nonfarm · +2.0% annualized · +4.1% over four quarters
Real hourly compensation, nonfarm · +1.1% annualized · +1.0% over four quarters
- Prices rose 2.0% while labour costs fell 0.7%. The nonfarm output price deflator rose 2.0% in the quarter; with unit labour costs falling, the room between the two went to nonlabour payments, which is where profits sit.
- Hours barely moved. Hours worked rose 0.1% in nonfarm business and fell 0.2% in the wider business sector, consistent with the low-hire labour market the claims and JOLTS data have shown all year.
- Productivity across business cycles, nonfarm business, annualized:
Period · Productivity · Output · Hours worked
Current cycle, Q4 2019 to Q3 2026 · 2.2% · 2.7% · 0.4%
Previous cycle, Q4 2007 to Q4 2019 · 1.5% · Not stated in this release · Not stated in this release
Long-term, since Q1 1947 · 2.2% · Not stated in this release · Not stated in this release
Against This Morning's Open
- The Open listed Q3 productivity at 2.8% expected against 1.4% prior, and unit labour costs at 0.6% against 1.2%. Productivity beat at 3.6%; unit labour costs fell 0.7%, against a rise.
- The Open warned that a jump in claims would sit badly with a 79% December hike price, and said unit labour costs "feed straight into the same argument." Neither jumped: claims fell to 189,000 and unit labour costs fell 0.7%, so the labour data added nothing to the hike case on wages and took nothing from it on jobs.
What This Sets Up
- Revised Q3 Productivity and Costs: Tuesday 8 December 2026, 08:30 ET (BLS). It will show whether the 3.6% survives revision; this release's Q2 moved 0.7 points.
- What would confirm the read: a revised Q3 with unit labour costs still at or below zero and the labour share holding near 53.0%.
- What would break it: an upward revision to hourly compensation like the 1.4 points Q2 just took, which would put unit labour costs back above zero.
What Is This?
- What it is: The BLS Productivity and Costs report, published quarterly in a preliminary and a revised estimate. Labour productivity is real output per hour worked by everyone in the sector; unit labour costs are hourly compensation divided by productivity. Nonfarm business output comes from BEA's GDP data and covers about 76% of nominal GDP; manufacturing output is built from Federal Reserve industrial production and is not directly comparable.
- Why it matters: Unit labour costs are BLS's own "indicator of inflationary pressure on producers." With the Fed weighing a December hike, this release says whether wages are adding to inflation. It also carries the labour share and the cycle-average productivity rate, the official measure of whether AI investment is showing up in output per hour.
- How to read it: Quarterly changes are seasonally adjusted annualized rates, so a 3.6% quarter is not a 3.6% year; the year-ago figures give the trend. Unit labour costs rise when pay outruns productivity and fall when productivity outruns pay. Preliminary estimates revise, sometimes heavily, when BEA updates GDP.
Want the full explainer? Economic News Events, Explained breaks down this release and every other one we cover, in plain English.
_For informational purposes only. Not investment advice._