Ed Humpherson to Liz Mckeown: Historic gross fixed capital formation statistics 

Published:

Sent from: Office for Statistics Regulation

Sent to: Office for National Statistics

Download correspondence (132 KB)

Dear Liz, 

Historic gross fixed capital formation statistics 

Thank you for ONS’s continued work to investigate the quality of the historic pre-1997 gross fixed capital formation statistics, and for the greater transparency provided in the July update. It is helpful that ONS has now set out the main categories of issue identified and has given users clearer warnings about the interpretation of the series. 

The update sets out that parts of the historic pre-1997 published series are affected by significant quality issues that are large enough to affect interpretation. It also makes clear that ONS has further work to do before it can confirm the full impact on historic GFCF data and wider National Accounts aggregates, as well as the nature and timing of any corrective changes. In the interim, users are directed to the Blue Book 2010 vintage, which uses older methods and is not coherent with the current post-1997 series. 

Long running historic time series are important for economic analysis and given the importance of these data for understanding long-term investment and the post-war economy, users need confidence that this work will lead to a satisfactory and timely resolution. I am therefore asking ONS to publish a clear plan for the remaining work. This should: 

  • set out the scope of the remaining investigation, the intended route to a solution, key milestones and a target timetable for decisions and implementation; 
  • commit to publishing indicative historic estimates, ranges or other quantified analysis at the earliest credible point, with appropriate explanation of uncertainty and provisional status; 
  • explain how ONS will assess and communicate any implications for wider historic National Accounts aggregates, and provide coherent interim guidance to users; 

We recognise that the work is technically complex, requires specialist resource and must be properly assured. Our expectation is not that ONS publishes estimates before they are sufficiently robust, but that users are given the best available indication of scale and direction as soon as this can be done responsibly. 

ONS should publicly set out its plans within 6 weeks of this letter. My team will continue to engage with yours as the work progresses. 

Yours sincerely, 


Ed Humpherson
Director General for OSR