Dear Jonathan,
Thank you for your email dated 15th May 2025 raising concerns about the Gross Domestic Product (GDP) per head figures published that day as part of the first quarterly estimate of GDP statistics for January to March 2025. We have addressed each of your concerns in turn below.
You expressed your view that the national population projections used for the GDP per head calculation were not ONS’s best available projection for the population. We agree that over the last few months, new data released by ONS have shown that international migration has been dropping and as a result, the principal projection used in the national population projections has become a less reliable reflection of short-term projections of the population. Instead, the migration category variant appears to have been providing a more likely short-term scenario of the impact of migration on population numbers.
We note that on 24 June, ONS published updated sub-national population projections, which advised the use of the migration category variant in place of the principal projection, as this is a better reflection of short-term population change. This recommendation was also accompanied by a blog explaining the issue for non-expert users. We welcome this response from ONS, and the migration category variant was subsequently used for the GDP estimates released on 30 June.
Following this change, you were concerned that users were not made aware that the migration variant category was being used. We have reviewed the ONS communications linked above as well as the GDP quarterly National Accounts: January to March 2025 publication released on 30 June, and the National accounts revision policy. Overall, our view is that the change to the migration variant category has been communicated to users upon its introduction, including updates to published material and clarity on what population data have been used in those calculations. We recognise that in future cases, and where possible, ONS could consider how the change is communicated in advance of the publication and how the impact of the change is explained to users. We have shared these views with ONS.
You also raised a concern about the use of annual population projections as the denominator for the GDP per head calculations, which use quarterly and seasonally adjusted GDP data as the numerator. We agree that this is a limitation of the calculation. We understand that ONS is considering the best way to reflect the latest available demographic statistics in the production of economic statistics such as GDP per head and that one option being considered is a quarterly time-series informed by quarterly population data rather than a linear interpolation, as is currently used. If implemented, this would allow the GDP per head calculations to be based on quarterly population projections. We understand that this work is in its early stages and encourage ONS to keep users updated.
Finally, you questioned whether ‘estimate’ is an appropriate descriptor for GDP per head, and you noted your preference for the term ‘implied’. We appreciate your view and have taken it into consideration. We agree that ‘estimate’ is not the most technically accurate term; however, unfortunately, we have been unable to identify an alternative that is technically correct whilst still understandable and interpretable by the non-expert user. We have passed your views on this issue to ONS so that that it can consider how it communicates and explains the concept of GDP per head in its releases.
Thank you for getting in touch with us.
Yours sincerely
Ed Humpherson
Director General for OSR