We compare standard productivity estimates of UK output per worker derived from the Labour Force Survey (LFS) with newly released estimates based on more reliable "Real Time Information" (RTI) administrative data. LFS-based productivity measures have been flat since 2024 Q3, a continuation of disappointing longer-term trends since the Global Financial Crisis. By contrast, RTI-based productivity estimates point to a meaningful productivity pickup. Between 2024 Q3 and 2026 Q1, our baseline administrative estimates indicate that productivity increased by 2.4 per cent, compared with just 0.3 per cent under the LFS-based measure. A divergence in labour input explains these differences. While the LFS suggests that labour input increased by 1.8 per cent between 2024 Q3 and 2026 Q1, administrative measures indicate a small decline. The difference in the series is explained by employees, rather than the harder to measure group of self-employed. Between 2024 Q3 and 2026 Q1, the LFS recorded an increase of 377,000 employees, while RTI recorded a decline of 133,000 employees. While some of the divergence could in principle reflect definitional and coverage differences between survey and administrative data, we believe that it is due to sampling challenges affecting the LFS, particularly during and since the pandemic. Given that administrative data currently provide the most reliable measure of employee growth, productivity estimates using such data provide a more accurate picture of recent developments.
Niki Barbas, Anna Valero and John Van Reenen
20 July 2026 Paper Number CEPOP74
Download PDF - UK national productivity trends: getting the data right
This work is published under POID and the CEP's Growth programme.