National Statistical

The Office for National Statistics (ONS) has today published its second progress update on its Economic Statistics Plan (ESP) and Survey Improvement and Enhancement Plan (SIEP). The two plans were launched in June 2025 to improve the quality and long-term sustainability of the UK’s core economic statistics. James Benford reflects on a period of continuing tangible improvements, the challenges we have encountered and how we are responding across the breadth of work. 

The past quarter has seen further tangible improvements to the quality of our economic statistics and a fall in the number of major errors made.  We have implemented some of our most impactful changes, introducing scanner data into consumer prices, further improving responses to our household and business surveys, laid the groundwork to adhere to the highest tier of statistical dissemination standards and moved more of our processes off legacy systems.  We have also implemented the final planned set of design changes for the Transformed Labour Force Survey (TLFS). 

A recent external review confirmed that the ONS is operating from a more stable footing, with stronger leadership, governance and grip on delivery.  An agreement has been reached with trade unions on a set of principles for hybrid-working, resolving a 2-year dispute on the approach to returning to the office.    

Stepping back, the breadth, scale and complexity of what we set out to achieve in the ESP and SIEP was considerable — and delivery experience has sharpened our understanding of that. Over two-thirds, 65 of the 93, milestones planned for 2025/26 have been delivered. The remaining third have proved more complicated and difficult to achieve than anticipated, reflecting a mix of system integration challenges, the time taken to mobilise required resource, the complexity of sequencing transformation at pace, and competing demands including preparation for Census 2031. 

It is not unusual to encounter challenges for programmes of this scale and ambition — we are being open about them, learning from our experience and taking deliberate action by more tightly focusing our work.  We are introducing a ‘waiting room’ to sequence improvement work to protect progress on the critical improvements we have in train.  

The update we have published today responds to recommendations and feedback from the Office for Statistics Regulation (OSR) on the previous publication, as well as their original recommendation of quarterly reporting on the plans.  

Surveys and economics statistics are improving across the board 

The improvements we have been developing in recent years are now feeding through into published statistics. Labour Force Survey (LFS) response levels have recovered to close to their pre-pandemic levels, significantly strengthening the quality of, and confidence in, core labour market outputs. On TLFS, the short Core Survey continues to improve response quality by reducing partial household responses and increasing the number of fully responding individuals. We are also seeing improvements in business survey response, with expanded case management now covering nearly 500 Reporting Units. 

The number of major errors has fallen from four in the previous quarter to one in the last. The number of errors overall remains too high, but we are changing our approach to strengthen how we learn from them.  We have introduced a new process to manage errors, started a monthly retrospective to reflect on learnings from each production round and are routinely setting out the actions we are taking when we correct an error.  

The standout achievement this quarter is the incorporation of grocery scanner data into consumer price statistics, covering around half of the UK grocery market from February 2026 data onwards. This concludes approximately four years of work and represents a step-change. Measured prices will be more representative of the experience at the till and less volatile. We have reduced risks by modernising our production process and built capability to bring in further large datasets in future. We have also doubled the sample size for hotels and computer games, reducing volatility in those price components, and improvements to house price statistics have reduced revisions and moved production further away from legacy systems.  

We have moved to publishing our weekly real time indicators in a dashboard, providing a clearer picture for users to see trends on consumer behaviour, business, workforce, transport, housing and energy at a glance, and freeing up internal resource. The dashboard will be complemented every month with a more detailed statistical bulletin and allows us time to focus more on new developments, including scoping out new indicators to add value to our existing suite.  

Elsewhere, we introduced a new system for processing central government data in March, significantly reducing our reliance on legacy technology for public sector statistics. Through migrating some of our key monthly surveys to modern platforms we have been able to validate and process an additional 15% of business survey responses each period, improving the quality of economic outputs. Improvements to Labour Productivity methodology has increased international comparability. 

We have made significant progress on the updated UK Standard Industrial Classification (SIC), a critical building block for future improvements across surveys, the business register and economic statistics. A finalised version has been published. Work on the Standard Occupational Classification (SOC) is also advancing. 

We also anticipate being recognised as fully compliant with all the requirements in the IMF’s Special Data Dissemination Standards Plus (SDDS+) by the end of this month, having completed the requirements for adherence to this highest tier of statistical standards.  

The last planned design changes for the Transformed Labour Force Survey have now been implemented 

All agreed major design changes planned for the TLFS are now implemented, as our update today shows. This includes data rotation — so respondents are no longer asked the full questionnaire at every wave — alongside improvements to pay and earnings questions and an increase in the conditional voucher incentive for respondents from £10 to £20. We expect these changes to further improve data quality and survey performance. 

We are working closely with users to refine criteria and conditions for transitioning from the LFS to the TLFS ahead of the first evidence-led readiness assessment in July 2026.  

At the readiness assessment, only one calendar quarter of data reflecting the latest set of design changes will have been collected, allowing for only a limited assessment of their impact. Given that limited data availability, further data collection and assessment will be needed, meaning the earliest transition of headline labour market statistics has shifted from November 2026 into 2027. A 2027 transition has always been part of our scenario planning and is the most likely outcome.  

The July 2026 readiness assessment will inform final design decisions, future publication plans and the timing of the next readiness review, which will take place no later than January 2027. 

We recognise that running two surveys in parallel for a longer period has implications for our capacity to progress wider improvements and we are actively managing that. 

Focusing on what matters most 

As we conclude our corporate business planning for 2026 to 2029, we are re-balancing our portfolio to protect delivery of the work that matters most. We are introducing a structured “waiting room” approach — a transparent way of sequencing change activity and holding it safely until we have the capacity and readiness to deliver it well. This is a deliberate, strategic choice. We are choosing realism over ambition to protect delivery confidence in our most critical improvements. 

Our immediate focus is on five priorities: sustainable delivery of our economic statistics, setting up our people for success; addressing quality issues, drawing on a new tiering framework that prioritises improvements to statistics underpinning critical decisions; transitioning to the TLFS; delivering the Statistical Business Register, which is critical to improving coverage of smaller businesses and laying the foundation for updated industry classifications; and establishing our programme to implement new international macroeconomic statistical standards. 

What comes next 

Our next quarterly update in July 2026 will reflect the outcomes of ONS business planning for 2026 to 2029. It will set out more detail on our priorities, including a refined set of milestones, a transparent account of what improvement initiatives are being held in the waiting room, and a new suite of Key Performance Indicators to track progress. We will include further detail on our programme of quality reviews and our approach to managing legacy systems, in line with OSR recommendations. 

We are operating from a more stable footing. Momentum is building and tangible improvements are feeding through. We will keep reporting openly on our progress, our challenges and how we are learning from them. 

Feedback on the Economic Statistics Plan can be submitted to: econstatsplan@ons.gov.uk