Dear Grant


We recently completed our review of compliance with the Code of Practice for Statistics of ONS’s Business Enterprise Research and Development (BERD) statistics. We appreciate the positive way the team has engaged with us throughout the course of this compliance check and the clear commitment the team has shown regarding the future development of these statistics.

We chose to review these statistics as aggregate business expenditure on R&D is a key driver of economic performance and will act as an important barometer in assessing the impact of the ongoing process of the UK’s departure from the European Union. These statistics also form the basis of the Department for Business, Energy and Industrial Strategy’s (BEIS) target for the UK’s aggregate investment in R&D, along with the various measures of public sector investment in R&D contained within the Gross Expenditure on Research and Development (GERD) bulletin. Our focus has been on how business expenditure on R&D as recorded by ONS’s BERD survey compares with business R&D expenditure as recorded by other statistical outputs. We also considered whether the statistics generate the level of insight that users need.

I’m pleased to confirm that these statistics should continue to be designated as National Statistics, subject to ONS fulfilling the requirement listed in bold on page 2 of this letter by October 2019.

We found a great many positive aspects in the way that ONS produces and presents these statistics and we found many positive examples of ONS seeking to enhance the clarity and insights offered by these statistics, for example:

  • ONS is currently working with BEIS and HM Revenue & Customs (HMRC) to investigate differences between the HMRC and ONS measures of business R&D, to improve understanding in this area including whether firms are claiming R&D tax credits for the R&D they report to the BERD survey. This demonstrates the high priority that ONS attaches to producing a coherent picture of R&D expenditure within the UK and we look forward to seeing the results of this work in the future.
  • ONS holds regular informal meetings and workshops with those who it has identified as its key users (BEIS, HM Treasury, HMRC and others) and it regularly seeks their views regarding the development of these statistics.
  • ONS published an ad hoc article last year which explores in detail the impact and GVA of expenditure on R&D by UK and foreign owned businesses. This is a great example of ONS responding to user requests and considering how it can contextualise the importance of business expenditure on R&D.

We are also pleased to note that ONS has published two articles which aid users in understanding the quality and comparability issues of these statistics. The Coverage of the Business Enterprise R&D Survey note identifies potential gaps in coverage and areas where these risks originate from and the Coherence of UK R&D statistics note which gave users a clearer understanding of the conceptual and methodological distinctions between BERD and other sources of Business R&D expenditure such as HMRC tax credits and the EU’s R&D scoreboard.

Since 2014 the R&D business expenditure as recorded by HMRC’s tax credit claims has exceeded the total amount of BERD’s total recorded R&D expenditure. These statistics are sufficiently distinctive in terms of their methodologies and coverage for their aggregate expenditure figures to be significantly different. However, ONS has not referred to the difference between the estimates, explained to users the reasons for it and why its growing, nor explained how users can best generate insight from using ONS’s data alongside others. In order to enhance public value and generate appropriate insight ONS should publish information that explains to users the differences between the ONS data and others in a way that enables users to draw maximum insight about R&D in the UK. We consider that it would be helpful if this explanation drew out the conceptual and methodological causes of this widening difference, and we encourage ONS to continue to work with HMRC to achieve a better understanding, including pursuing access to HMRC raw data.

There are several further measures that ONS could consider undertaking to develop these statistics in line with the Code of Practice for Statistics, for example:

  • ONS should consider, in conjunction with developing its understanding of users’ needs, how to add further insight and increase the value of these statistics by referencing other relevant business statistics, such as business demography, insolvency and investment statistics.
  • ONS should demonstrate how it works with users to enhance the value and insights offered by these statistics by publishing details of these statistics’ main users, and the user engagement that ONS has previously undertaken, and plans to do.

Thank you again for engaging effectively with us during this review. We will keep in touch with the statistical team as it undertakes the required action and have agreed with the team that it will report back to us by the end of October 2019 on the actions taken in response to our review.

I am copying this letter to Cecil Prescott as the responsible manager for BERD statistics and to Ole Black responsible for ONS’s R&D statistics.

Yours sincerely

Mark Pont
Assessment Programme Lead


Related Links:

Assessment Report 218 – Statistics on Research and Development (ONS)