1. The Code of Practice for Statistics

The call for evidence asked three questions to gather stakeholders’ views about the Code:

1. how they saw the Code currently achieving or not achieving its ambitions;
2. how well or poorly they thought the Code supported analysts and others to deliver and use statistics that serve the public good; and
3. how they thought the Code could be strengthened and improved.

There were 40 responses on these questions.

Ten other respondents used the call for evidence to provide specific feedback on the application of the current principle T3.6, which enables alternative release times for official statistics when this is judged to be in the public good, to market sensitive economic statistics. OSR is looking into the issues raised by these respondents. We will report separately on this matter.

What you told us

A strong message received throughout the Code review, and endorsed in the call for evidence, is that the Code of Practice is highly regarded by those producing official statistics, as well as by other stakeholders. The current three pillars – Trustworthiness, Quality and Value – were found to be a clear framing of fundamental aspects of good statistical practice, that are memorable and easily explained to others to support application of the Code. However, there was also wide recognition of the evolving nature of the data landscape over recent years and consensus that the Code can be enhanced both in its content and presentation.

The following paragraphs briefly summarise key views shared that relate to each pillar of the current Code.

The feedback highlighted that we could enhance standards relating to Trustworthiness by:

• further clarifying independent statistical leadership authority and a Head of Profession for Statistics’ responsibilities
• integrating OSR’s concept and guidance on Intelligent Transparency
• reviewing the orderly release principle
• better addressing the communication of statistics
• expanding the standards regarding data ethics
• updating current practices relating to data governance to be in line with GDPR legislation

Respondents recommended that we strengthen the Quality pillar by:

• more fully addressing different types of data sources (such as administrative data) and the boundaries with other types of government numerical information (such as management information and research)
• clarifying comparability requirements
• addressing the balancing of timeliness and accuracy to meet user needs
• clarifying expectations around quality assurance
• better integrating with international frameworks for quality management
• incorporating learning from our spotlight on quality indicator framework

Stakeholders emphasised that we could strengthen the Value pillar by:

• further prioritising meeting user needs and addressing the balance of different user needs within and outside government
• considering the implications of the RSS’s concept of ‘public statistics’ (in which the official statistics start from identifying the questions that need answering and being clear about the purpose)
• further encouraging and supporting innovation by providing guidance on the use of new technologies such as AI
• emphasising the need to address data gaps
• requiring transparency around business planning and achieving value for money

Beyond changes to the Code itself, a wide range of feedback emphasised the importance of finding more effective ways to support understanding of the three pillars by members of the public and of analysts outside of government and of non-statistical staff in government, to support their use of data and statistics.

While it was recognised that there is a body of guidance alongside the Code already, those producing official statistics asked for more accessible materials. This guidance could range from simple high-level guides about the pillars to detailed material aimed at some of the most challenging current statistical issues, such as the implications of AI and the greater use of administrative data.

OSR’s response

Based on the feedback received, it is clear that our stakeholders have identified various opportunities for us to enhance the content of the Code, as well as how we communicate and present its standards. In addition, we recognise that the Code could better support understanding of how statistics serve the public good by drawing on OSR’s research and our developing definition of this concept and that we could incorporate learning from UN value of official statistics task force.

Outcome and next steps

We will refresh the Code of Practice to produce a third edition.

We anticipate that this refresh will update and extend the content of the Code.

We will also review the presentation of the Code, both online and on paper, to improve the clarity and accessibility of the Code for a range of different audiences.

We will consider how topic specific guidance can be integrated into the Code to address emerging specific statistical practices.

The refresh will commence in February 2024. We are planning to consider the content of the Code in detail over the following three months and aim to begin a period of engagement about the refreshed Code in the summer of 2024. We will run a public consultation on the refreshed Code.