There has been a lot of focus recently on labour productivity, which looks at how much is produced for each hour worked across the economy. Richard Heys explains the different measures the ONS calculates as well as which gives the best indication of recent changes in the UK economy.
For many years our sole measure of labour productivity was derived by dividing UK Gross Value Added (GVA) by the number of hours worked in the economy, calculated from the Labour Force Survey (LFS).
In recent years, in line with many other statistical institutes internationally, we have been finding it harder to elicit responses to the LFS, which affected its quality and made its estimate of the number of hours worked more uncertain. We have seen notable improvements to LFS quality in the last year, with response levels now close to pre-pandemic levels. However, productivity mainly focuses on year-on-year changes so these historic lower responses are still impacting annual growth rates in the headline productivity data.
Partly in response to these challenges, in August 2024 we started to publish a second, experimental, approach to measuring productivity. This uses the number of employees recorded on payrolls by HM Revenue & Customs – known as RTI – and is augmented with self-employment and hours data from the LFS. For various reasons, these two measures have sometimes shown different trends, particularly in some recent periods.
We have been making progress on developing methodological improvements to our headline statistics, using lessons learnt from other countries and the OECD. Drawing on this experience, we are developing a definitive measure of labour productivity, which takes account of international best practice to build labour metrics up from their ‘components’ (usual hours worked, overtime and leave). This approach draws on the strongest aspects of multiple data sources, including RTI and LFS, as well as the jobs survey we send to employers, known as Workforce Jobs. Our aim is that this metric will ultimately replace the two currently published series. While we’re not yet ready to publish estimates from this new component measure we are currently aiming to do this around the end of this year.
While we continue to develop the ‘component’ approach, the historic lower response rates for the LFS lead us to recommend, at the current time, that users focus on the experimental RTI approach as the best estimate of how productivity is changing in the UK economy.
This update does not impact other labour market measures, which have different use cases and are calculated separately according to different international conventions.
We will continue to work on the new components approach, and expect to share our findings and methods with users later in the year.
Richard Heys is Deputy Chief Economist at the ONS.