UK Output per Hour Rises 2.5% on First Reading of New ONS Measure
The Takeaway: Britain’s statistics office has adopted a new headline measure of labor productivity, and its first quarterly reading puts output per hour 2.5% above a year earlier. Output per worker grew 2.0%, its fastest pace since early 2018 outside the pandemic years. Revised across its full history, the new series shows stronger long-run growth than the old survey-based measure, and the office now ranks UK output per hour growth second among the Group of Seven economies over that span.
UK output per hour rose 2.5% on the year in July to September, the Office for National Statistics said on Tuesday, the first reading under its new productivity measure.
The rise came from a 2.0% increase in gross value added, the economy’s output, while hours worked fell 0.5%. Output per worker also grew 2.0%, its strongest annual growth since January to March 2018, leaving aside the pandemic period.
A new headline measure, revised back to 1997
From this release the office uses what it calls the component approach as its headline measure. It replaces estimates built on Pay As You Earn Real Time Information, the tax records of employees on payrolls.
The office told users in a blog on 16 June that the new measure is superior to its earlier method, which took hours and workers directly from the Labour Force Survey, the household survey behind the unemployment rate.
The switch revises whole-economy estimates for every period back to 1997. Revisions by industry for 1997 to 2024 will follow in a separate publication early next year.
The office also drops its split between flash and overview estimates. Each bulletin will now add one quarter and revise the earlier ones.
The method changes how hours are counted. Rather than relying on the actual hours people report in the survey, it starts from usual hours, benchmarked to the annual earnings survey, and adjusts explicitly for annual leave, sickness and other reasons hours differ.
For the latest quarter, with workforce jobs data not yet available, employee jobs are projected using the payroll records’ growth rate and combined with the survey’s count of self-employment jobs.
The estimates remain labeled official statistics in development, and the output figures now include this year’s Blue Book revisions to the national accounts.
Output per hour gains 1.3% in the quarter
UK output per hour worked, latest five quarters (growth, %)
| Quarter | Vs 2019 | On year | On quarter |
|---|---|---|---|
| Q3 2025 | 6.2 | 4.1 | 0.5 |
| Q4 2025 | 6.3 | 1.6 | 0.1 |
| Q1 2026 | 7.3 | 1.5 | 0.9 |
| Q2 2026 | 7.5 | 1.7 | 0.2 |
| Q3 2026 | 8.8 | 2.5 | 1.3 |
Component approach, seasonally adjusted. “Vs 2019” compares each quarter with the 2019 average level.
Output per hour rose 1.3% from the second quarter, the largest quarterly gain in the five quarters the office published. Annual growth picked up from 1.7% in April to June. On the old payroll-based headline, the previous release had put April to June growth at 0.7%.
That is the fastest annual rate since the same quarter of 2025, when output per hour grew 4.1%.
Against the 2019 average, before the pandemic, output per hour was 8.8% higher. Output grew 7.9% over that span while hours worked fell 0.9%.
UK output per worker, latest five quarters (growth, %)
| Quarter | Vs 2019 | On year | On quarter |
|---|---|---|---|
| Q3 2025 | 3.5 | 1.6 | 0.5 |
| Q4 2025 | 3.6 | 1.7 | 0.1 |
| Q1 2026 | 4.1 | 1.4 | 0.5 |
| Q2 2026 | 4.7 | 1.6 | 0.5 |
| Q3 2026 | 5.6 | 2.0 | 0.9 |
Component approach, seasonally adjusted. “Vs 2019” compares each quarter with the 2019 average level.
Output per worker rose 0.9% on the quarter. It stood 5.6% above its 2019 average, a smaller gain than output per hour because the hours each worker puts in have fallen.
The output side draws on the first estimate of third-quarter GDP, which showed the economy growing 0.7% on the quarter, as set out earlier. The office advises users to focus on longer-term trends rather than single quarters, since productivity is a structural feature of the economy.
The old and new measures part ways
The office still publishes the older measures as comparators, for analysis only. They tell a weaker story.
Output per worker was 5.6% above its 2019 average on the new measure and 5.5% on the payroll-based one. The survey-based measure put it only 2.6% higher, less than half the gain.
The gap is wider for output per hour. The new measure puts it 8.8% above its pre-pandemic average, against 3.4% on the survey-based measure. That is a difference of 5.4 percentage points.
Because all the measures share the same output estimates, the gap comes from hours. Output per hour on the new measure has generally grown faster than on the survey-based one since the 2008 to 2009 financial crisis, the office said, as total hours on the two measures drifted apart.
For workers, the divergence is more recent and dates from 2024. The office ties it partly to improvements in the household survey in 2024 and 2025, which brought its employee count closer to the payroll records.
More recently, employee growth in the survey has outpaced the payroll data, and the office is investigating why. The same split showed in Tuesday’s labor market figures, as set out earlier, where the survey showed employment rising while payrolls fell.
On the new measure, total hours worked leveled off around 2023 and declined slightly in 2025. The office puts that down to weak growth in the number of jobs and a long decline in the hours worked in each job.
Average hours per job stand 12.1% below 1997
By July to September, average actual hours per job were 12.1% below their 1997 annual average. The decline has been steady, with one exception.
In April to June 2020, as lockdowns took hold, average hours stood 30.5% below the 1997 level. Sickness and furlough accounted for 14.6 percentage points of that gap, almost half.
Contributions to the fall in average hours per job, July to September 2026 against 1997 (percentage points), ranked
| Factor | Points |
|---|---|
| Usual hours | -5.7 |
| Employment mix | -3.7 |
| Other reasons hours vary | -2.2 |
| Annual leave and bank holidays | -1.1 |
| Sickness and furlough | +0.7 |
Usual hours combine employees’ usual hours with actual hours in self-employment and the armed forces. Other reasons include unpaid overtime and parental leave. Contributions are against the 1997 annual average.
Shorter usual hours made the largest contribution, about 47% of the fall. Changes in the mix of employment added about 31%, so together the two explain roughly 78%.
Annual leave and bank holidays took off 1.1 points, reflecting larger leave entitlements. Sickness and furlough worked the other way, adding 0.7 points.
In 2019 the shortfall stood at 8.6%. It has since widened to 12.1%, reflecting a 3.9% fall in average hours per job over the period. Other reasons for hours differing from usual played a proportionally larger part in that recent decline than in the long-run fall.
Second in the G7 since 1997
On the revised figures, UK output per hour grew faster between 1997 and 2025 than in every other G7 economy except the United States.
The office said that suggests productivity growth over the past three decades “has been stronger than previously perceived”, though still materially lower than in the US. It urged caution because the countries measure productivity differently.
The ranking holds over shorter spans. In the 10 years to 2007, to 2017 and to 2025, the UK placed second each time, behind only the US.
In The Edge’s reading, the revision matters more than the quarter. Since the financial crisis, output per hour has grown faster on the new count of hours than on the old survey-based one, so part of the weakness the old measure showed came from measurement, although the series is still in development and its industry detail is not yet out.
What to watch: The next release, covering October to December 2026, moves to 4 March 2027 from 16 February 2027. It will add industry-level estimates on the new method and updated quality and methodology information.
Method: figures are from the Office for National Statistics article Labour productivity, UK: July to September 2026, released on 17 November 2026 at 09:30 GMT; growth rates are seasonally adjusted. The April to June 2026 rate on the old payroll-based headline is from the office’s Productivity flash estimate and overview, UK: April to June 2026 and January to March 2026, released on 18 August 2026. The largest quarterly gain among the five quarters published, the comparison of annual rates, the comparison of output per worker on the new and survey-based measures and the 5.4-point gap in output per hour are The Edge’s calculations; rounded figures are used throughout.
Sources: Office for National Statistics, The Edge.



