ONS confirms 2025 labour market figures now fully reflect survey overhaul

Employment and unemployment estimates from January to March 2025 are the first to carry the complete effect of major changes to how the UK’s Labour Force Survey collects and weights its data.

The UK’s main employment survey has, for the first time, produced a full quarter of data built entirely on improvements introduced from January 2024 — and that matters for anyone relying on labour market figures to make decisions. The Office for National Statistics confirmed the milestone via its official channel, stating that estimates for January to March 2025 now fully reflect changes to Labour Force Survey data collection, sampling methods, and weighting introduced over the past year.

The scale of the changes is worth spelling out. The targeted Wave 1 sample was increased from around 16,700 to 25,800 new households per quarter — a rise of over 50 per cent. By January to March 2024, the achieved sample had already reached 50,783 person interviews across 23,143 households in the UK, representing a 13.9 per cent increase in household interviews and a 14.5 per cent increase in person interviews compared with October to December 2023. The overall UK response rate climbed from around 17.4 per cent in July to September 2023 to around 24.6 per cent by July to September 2024.

ONS also reintroduced face-to-face interviewing, recruited additional interviewers, and raised participation incentives from £10 to £50 for some respondents. The entire LFS dataset was reweighted back to 2019 using updated population estimates published in January 2024 — a change that pushed level estimates higher across many series and narrowed the gap between survey-based employment figures and HMRC payroll data, while leaving rates and averages broadly unchanged.

But ONS has been careful not to declare the job done. The figures show that January to March 2025 marks the point where January 2024 improvements are fully embedded — yet the ONS cautions that later enhancement work may still affect future periods.

For Kent, the practical effect runs through regional and local authority-level statistics derived from LFS data. A larger national sample and a reinstated sample boost mean more respondents from Kent and the wider South East, which should reduce sampling error in the small-area estimates that Kent County Council, district councils, and Jobcentre Plus offices use for skills planning, employment support, and funding bids. Some previously published figures for Kent and Medway may have been revised as reweighted data were incorporated — users comparing current estimates with pre-2024 series should account for those methodological changes when reading trends.

Some analysts and parliamentarians had previously questioned how reliable LFS-based statistics were during the period of low response rates, and whether methodological changes complicate long-run trend comparisons. The figures show a clear recovery in sample size and response rates, yet questions remain about whether the survey adequately captures precarious workers, gig-economy participants, and migrants at a local level.