An update from the Interim Chair: How we make decisions about the Transformed Labour Force Survey
September saw two major milestones for the UK Statistics Authority. We announced that André Loranger, currently Canada’s Chief Statistician, would be our next National Statistician. We’re looking forward to his arrival next month, bringing fresh energy and an abundance of experience to the statistical system. The Public Administration and Constitutional Affairs Committee (PACAC) has published its report on the work of the Authority, and we will respond in early November. After a year focusing hard on the Office for National Statistics’ (ONS) recovery, these two events mark an opportunity to enter a new phase for the ONS, and equally importantly, to harness opportunities across the statistical system as a whole. Finally, the Cabinet Office will soon launch the recruitment of a permanent Chair, helping focus on the longer term strategy as well as the more immediate priorities.
As we move into this next phase, I am conscious that one particular issue is currently seen as a litmus test for the ONS: a successful transition to the Transformed Labour Force Survey (TLFS). So it’s a good moment to share the Board’s perspective on this, particularly since the ONS has just published its latest update on economic statistics and the surveys improvement plan. The ONS continues to operate transparently in its public communications; I want to add the Board’s perspective and how we are engaging on the substance. A quick recap: at the readiness assessment in July, the decision was confirmed that although there were encouraging improvements to TLFS, it was too early to transition to TLFS.
We know that the many people working hard on the programme across the ONS are focused on a successful outcome. And I should be clear that the Board certainly wants to be able to agree in July 2027 that it is safe to start transitioning to the TLFS in November 2027. After all, the strategic reasons for the new survey remain the same: an online survey with lower respondent burden, higher participation and improved production processes is better for respondents, users and decision makers. Reduced costs per respondent allows for a bigger sample, improving the accuracy and granularity of estimates. We are also conscious of a Catch 22 situation. Dual running TLFS and the Labour Force Survey (LFS) comes at a price: financial of course, but also the opportunity cost. While the ONS is running and maintaining LFS alongside TLFS, its ability to get the best out of TLFS is more limited: it puts pressure on methodological, testing and digital capacity. It also reduces capacity to improve broader surveys and statistics.
But it would be a mistake to transition at any price, and we will give proper consideration to whether the programme is sufficiently ready. The Board is focused on the agreed acceptance criteria: statistical and data quality, user acceptance and readiness, and operational readiness. TLFS is making good progress, in that, for example, the implementation of data rotation and increased incentives are making a positive impact on response. It is the ONS’s responsibility to execute the TLFS programme successfully; our distinctive role as a Board is to provide support and challenge in the areas where we conclude there is more work to be done, and to support the ONS in taking users with us.
We are particularly focused on the quality of the new statistics: users need sufficiently precise measures of change; and there is further to go on producing sufficiently reliable industrial and occupational classifications in an online self-completion environment. Users will also want to have differences between the TLFS, the LFS and administrative sources clearly explained. The Board will take account of stakeholders’ critical needs, any trade-offs and how these will be managed. Dependencies across the statistical system must be resolved, including coordination with the Northern Ireland Statistics and Research Agency (NISRA) on the consolidated UK picture. Finally, transition will not end the work: as noted already, stopping the dual run will release capacity for further improvement, so the ONS will need to set out a credible post-transition programme.
As the incoming National Statistician, André’s statistical advice to the Board will be invaluable as we approach key decision points. And in due course, the Office for Statistics Regulation will report to the Board, and publicly, its conclusion on whether the new TLFS statistics comply with the Code of Practice for Statistics. PACAC has also asked the Authority to update it on TLFS each quarter; and of course the ONS will continue to publish regular updates for stakeholders.
We are confident that as we approach key decision points in 2027, we will be in a good place to make sensible and rounded judgements.