UK output per hour worked was 2.5 per cent higher in July to September 2026 than a year earlier, the Office for National Statistics said on Tuesday.
Output per worker was 2.0 per cent higher over the same period, according to the first quarterly estimate for the three months.
The ONS said the rise in output per hour reflected a 2.0 per cent increase in gross value added, the measure of economic output, and a 0.5 per cent fall in hours worked.
Against the 2019 average, output per hour worked was 8.8 per cent higher in the latest quarter. Output was up 7.9 per cent and hours worked were down 0.9 per cent over that longer period.
Average hours per worker were 12.1 per cent below their 1997 annual average. In April to June 2020 they had fallen to 30.5 per cent below that level, with sickness and furlough accounting for 14.6 percentage points of the gap.
New headline measure replaces payroll-based estimates for 2.5 per cent figure
The figures are the first to use what the ONS calls the component approach as the headline measure of labour productivity. It replaces the estimates based on PAYE Real Time Information, the payroll data held by HM Revenue and Customs.
The ONS said it had told users the new measure is superior to its previous direct approach based on the Labour Force Survey.
The method builds labour input from separate measures of hours, jobs and workers. It makes explicit adjustments for annual leave, sickness, bank holidays, furlough and overtime, and benchmarks hours worked to employer-reported data.
It does not change how output is measured, so differences from earlier productivity estimates reflect updated labour input figures, not changes in output, the ONS said in its methodology article.
That article set out the effect on past years. By 2024 output per worker was about 1.9 per cent above its 2019 level under the component approach, against 2.4 per cent under the Labour Force Survey method and 1.5 per cent under the payroll-based measure.
Output per hour in 2024 was 40.7 per cent higher than in 1997 under the new approach, against 34.0 per cent under the old one. Average annual productivity growth over 1997 to 2024 comes out at 1.3 per cent, compared with 1.1 per cent.
The ONS said the component and payroll-based measures of output per worker have been similar in the latest estimates, and both show stronger growth since 2019 than the Labour Force Survey measure. Divergences between the three have appeared since 2024.
First estimate carries cautions, with next update on 4 March 2027
For this first estimate of a quarter, the ONS used alternative sources because standard workforce jobs data were not yet available.
It applied the year-on-year growth rate in payroll employee jobs to estimate employee jobs, then added the latest Labour Force Survey estimate of self-employment jobs.
Output figures include the Blue Book 2026 revisions, which feed through the whole quarterly series. An operational issue affected Labour Force Survey data collection in April to June 2026, though analysis suggests the effect on headline productivity was small.
The estimates stay labelled "official statistics in development". The ONS said it continues to advise users "to focus on longer-term trends rather than movements in individual quarters".
Flash and overview estimates will no longer be separate. Each later update will add a quarter and revise earlier ones, and industry-level estimates on a one-quarter lag will be added, where previously only the whole-economy figure was available that soon.
The next update, covering October to December 2026, is due on 4 March 2027. It was previously scheduled for 16 February 2027, and will include the first industry breakdown under the new method.
