Dear Liz,
Business enterprise research and development (BERD) statistics – Confirmation of accredited official statistics status
We have reviewed ONS’s actions to address the two requirements outlined by our Spotlight on Quality assessment report on ONS’s BERD statistics. On behalf of the Board of the UK Statistics Authority, I am pleased to confirm that ONS’s subregional and sectoral BERD statistics for 2022 and onwards comply with the standards of Trustworthiness, Quality and Value in the Code of Practice for Statistics, and should once again be labelled as accredited official statistics. I can also confirm that ONS’s UK BERD statistics can continue to be labelled as accredited official statistics. The detail supporting our judgement is included in the annex to this letter.
ONS BERD statistics provide insights into how much UK businesses invest in research and development (R&D), where this funding originates and the workforce involved in R&D activities. These figures are used by the government, academics and researchers to inform policy and drive innovation.
ONS conducted extensive engagement with several users of the BERD statistics, and we have heard positive feedback from some of those users on that engagement. Whilst the use of the new methodology (sample design and survey) began in 2022, ONS has also carried out extensive work reviewing different methodologies to produce an improved back series for BERD data to 2014. ONS has committed to publishing an article for users later in the year explaining the improvements made and the work carried out on the back series. We look forward to the further development of an improved back series, which ONS plans to undertake once the new transformed survey has produced data for more years.
I can also confirm that ONS’s statistics on Gross domestic expenditure on research and development (GERD) can once again be labelled as accredited official statistics. These statistics had their accreditation suspended because they rely on ONS’s BERD estimates for data on the business sector, which comprises the largest share of R&D expenditure of the four sectors which make up GERD. The new sample has increased the volume of survey data received and ensured that the BERD statistics better reflect business R&D performed in the UK. We are currently planning to conduct a compliance review of these statistics later this year.
I would like to thank your team for its open and positive engagement throughout our assessment and for its ongoing commitment to improving these statistics. We are particularly impressed by the positive working relationships demonstrated between the R&D teams at ONS and the Northern Ireland Statistics and Research Agency (NISRA).
I am copying this letter to James Benford, Director General, Surveys and Economic Statistics; Mary Gregory, Head of Profession for Statistics; Rachel Skentelbery, Deputy Head of Profession for Statistics; and Hannah Finselbach, Deputy Director, Surveys and Economic Indicators.
Yours sincerely,
Ed Humpherson
Director General for OSR
Annex
Review of actions undertaken in response to OSR’s assessment: Statistics on UK Business enterprise research and development (BERD) produced by the Office for National Statistics
Requirement 1: To provide transparency to users and to add to their understanding of what information is included in the BERD statistics, ONS should publish a sample questionnaire, and the guidance for completing the survey should be made available to users.
Actions taken by ONS: ONS created a page on the ONS website, Questions for Business Enterprise Research and Development Survey, that lists all questions asked on the BERD Survey for both the long- and short- form types. A link to the survey was included in the BERD bulletin published in November 2025.
ONS discussed the challenges around providing the guidance linked to each question with OSR and its concerns that adding the guidance would have made the page less user friendly. ONS considered alternative options to include the guidance, but these would have meant having to use downloadable files, which is against accessibility guidelines, and so it was not a viable approach. However, at OSR’s suggestion it included contact details at the top of the page, allowing users to request any additional guidance if needed.
OSR’s evaluation of evidence: We judge that this requirement has been met. Publishing the Questions for Business Enterprise Research and Development Survey page provides users with the necessary information about what is included in the BERD statistics. The page is clearly linked in the BERD bulletin and will be included in the annual publication in the future. There is also a clear route for accessing the survey guidance.
Requirement 2: To reduce the risk of users misinterpreting BERD statistics, ONS should better communicate uncertainty about these statistics, and in doing so should engage with users to understand what further explanatory information they may require of the strengths and limitations of the BERD statistics. ONS should also make it easier for users to find explanatory information, such as articles, on methodological and other improvements on its website.
Actions taken by ONS: ONS published updated quality and methodology information for BERD statistics in January 2025. The publication includes further information on the strengths and limitations of BERD statistics, as well as recommendations for suitable use of the statistics.
ONS published quality measures for BERD statistics from the 2023 period onwards in the BERD bulletin published in November 2025. This will allow users to assess the margin for error in BERD statistics at the product group level, which is the main area of focus for users of BERD statistics. These measures will be included in all future publications going forward. ONS will provide NISRA with details of the process so they can also start producing quality measures for the Northern Ireland Research and Development (R&D) Survey.
ONS has reviewed and updated the R&D topic page on its website. This page brings together the methodology information for R&D, R&D datasets, publications and the transformation articles published throughout the development period. ONS has added a link to the topic page in the e-mail signature of the R&D inbox. When new R&D releases are published, ONS will send this link to members of the user group to make them aware and encourage them to access the release through this page. ONS will take a pro-active role in updating this page as and when articles are released.
OSR’s evaluation of evidence: We judge that this requirement has been met. ONS used a method and process approved by ONS methodologists to introduce standard errors as quality measures at the product group level in the November 2025 release, which will aid users in understanding the uncertainty of the estimates. ONS welcomed feedback on the approach from users of the statistics. Stakeholders told us that ONS was very receptive to their queries.
We are pleased to see that ONS considered the Northern Ireland figures and shared its knowledge with the R&D team at NISRA so that NISRA can also start producing quality measures.
The new topic page makes it a lot easier for users to access all the information related to the BERD statistics, including articles on methodological and other improvements. ONS has optimised the accessibility of its R&D information, making sure it is available to users when published, and will continue to review this as it develops its new website.
Requirement that the final phase (phase 3) of the methodology improvement plan was implemented: The refinement of methods and systems, including the delivery and integration into National Accounts in 2025 and the production of a back series of the breakdowns of BERD in late spring 2024.
Actions taken by ONS: Following the improvement of R&D data sources, ONS successfully included R&D data in the Blue Book 2025.
ONS intends to publish a back series to 2014 following agreement with key stakeholders. Datasets from 2018 to 2021 have been uplifted. ONS methodologists have suggested that they believe at least the growth rate from the previous data should be trusted, so ONS plans to use the 2018 uplifted data and this growth rate and work backwards. However, it does not believe that taking this method back beyond 2014 would result in a robust series. ONS undertook substantial work to ensure the methodology operated effectively following the changes introduced in 2021, including testing a range of approaches to identify the most appropriate solution. ONS also plans to introduce additional breakdowns in line with user requirements.
ONS has committed to publishing an article later this year to explain key information to users. This will include explaining the options considered for improving the statistics (including for the implementation of the back series) and why options were ruled out.
OSR’s evaluation of evidence: We have heard positive feedback from users regarding the decision to limit the BERD back series to 2014. ONS has carefully considered multiple ways to approach the back series and involved methodology colleagues, dedicated resource and key stakeholders.
We have further assurance that ONS has investigated this thoroughly through its commitment to publish an article, which has been shared with us in draft form. The article sets out clearly what work has been undertaken to develop a back series, including the different approaches explored, the methods used and why these did not result in a credible or robust series. The article explains why NISRA has been able to produce a back series and why, given the constraints of the ONS data, this is not currently possible for the wider UK breakdowns. It also describes the engagement ONS has had with users, including the mixed feedback received when draft data were shared and the concerns raised by stakeholders. Finally, it sets out how ONS plans to revisit backcasting methods in the future, once more years of data from the transformed survey are available.