Transparent and trustworthy

Lesson 1

Throughout the pandemic we have advocated for transparency of, and equal access to, data used in government decision-making. Our expectations for transparency apply regardless of how data are categorised. For many who see numbers used by governments, the distinction between official statistics and other data, such as management information or research, may seem artificial. Therefore, any data which are quoted publicly or where there is significant public interest should be released and communicated in a transparent way.

The importance of public confidence in government decisions has been evident throughout the pandemic – to support confidence in, and therefore compliance with, lockdown restrictions, self-isolation rules or the vaccination programme. Government statisticians are uniquely placed to demonstrate independence from the political process and build public trust in data which underpin these decisions. To do this, they must be given the freedom to make independent decisions and to challenge the misuse of statistics. Ministers and officials who understand the value of statistics and the expectations of the Code of Practice for Statistics, are better able to proactively support producers to demonstrate transparency and trustworthiness. The Ministerial Code sets the standards of conduct expected of UK Government ministers and requires them to be mindful of the Code of Practice for Statistics. This requirement is also reflected in the Scottish and Welsh Ministerial Codes and the Northern Ireland Guidance for Ministers. The Public Administration and Constitutional Affairs Committee recently asked for the Ministerial Code to be strengthened to require Ministers to abide by the Code of Practice. In response the UK Government reiterated its commitment to transparency and its aim to publish all statistics and underlying data referenced publicly, in line with the Code of Practice.

We recognise that there are challenges when releasing new data at pace. During the pandemic, systems were introduced rapidly in response to urgent data reporting needs. Producers had to find an appropriate balance between transparent public reporting and demands to provide important internal information to inform government and health body decision-making. The rapid release of information to the public was more challenging in organisations which did not have established publishing processes for some situations, such as releasing ad-hoc data alongside ministerial statements. Some organisations also found that the time taken to check the implications of publication for other producers could lead to delays. More regular data sharing and linking, and closer collaboration between producers, would help them to respond to these situations more quickly.

Overall, we consider that health and social care statistics producers across the UK demonstrated a clear commitment to transparency through their efforts to put a huge amount of valuable information in the public domain – from data on COVID-19 cases, deaths, and hospitalisations to information about public health programmes, such as vaccinations, or contact tracing. The work by analysts to improve data presentation and publish data alongside daily Downing Street briefings demonstrated the importance of allowing experts to use their skills to enhance transparency. The additional data and information workbook published by the Scottish Government allowed for the rapid release of ad-hoc data to ensure equal access to figures used in ministerial statements.

We found that there were several ways in which producers were supported to play this important role. For example, well-established senior statistical roles within the Scottish Government and Public Health Scotland helped to ensure a common understanding of the value of statistics and data, and the standards they should meet. Building strong relationships with policy colleagues was key to developing this understanding. We have often highlighted with producers that sharing their plans for statistics with the public improves transparency and therefore enhances trustworthiness, even if these plans are tentative and later change. We have seen good examples where this has been done, such as for vaccination statistics in Wales and NHS England Test and Trace statistics. We also saw the benefit of introducing formal processes for the rapid release of information at the Department of Health and Social Care, which were developed by learning from practices in other departments.

However, we have often had to intervene, both publicly and privately, when data about COVID-19 have not met expectations around accessibility or transparency. For example, we have intervened when figures quoted by ministers or government officials were not publicly accessible (such as those used at press conferences), or when poor use and presentation of data risked undermining public trust (such as on testing figures). Delays to, or gaps in, published data have often resulted in accusations that governments cherry pick or manipulate data – we have seen this in the media and on social media, as well as through concerns raised with us. We have also heard frustrations from some producers who felt unable to do their jobs effectively and make the final decisions about data collections or publications. There can be many reasons for these issues, including pressures on analytical resource and delays to sign off or publication processes. During autumn 2021 we will be building on our work around the importance of transparency with a series of engagements and outputs to highlight our expectations.

Recommendations

• The principles of the Code of Practice and our expectations on transparency should be followed for any analysis published by governments. Producers should consider voluntary application of the Code of Practice, and consider requesting a rapid review by the Office for Statistics Regulation, for data which are not official statistics.
• Organisations should ensure that processes exist for releasing data quickly and transparently. This will be especially important for new bodies and organisations, such as the UK Health Security Agency.

Lesson 2

The pandemic demonstrated the value of government data and statistics and the importance of the data literacy of senior leaders within governments. We found that senior non-analytical leaders within governments and organisations provided valuable support for statisticians, for example by championing the production and publication of new data. Senior leaders are also vital in developing an organisational culture that values evidence and analysis. This includes ensuring that statisticians are involved throughout policy-making and operational processes, and as performance measures are developed for new programmes. We found that some organisations developed a greater understanding of the need for official statistics and the importance of having embedded processes to produce statistics, for example at the Department of Health and Social Care, and Public Health England.

Our review of Statistical leadership: Making analytical insight count highlighted that senior non-analytical leaders need to know how and when to make the best use of statistical evidence. This includes knowing the right questions to ask analysts about data. The Data Masterclass for Senior Leaders, developed by the Number 10 Data Science Team and now run by the Data Science Campus, is a positive step in building the skills and confidence of senior government leaders. It is good to see a commitment to developing data expertise in the UK Government’s Declaration on Government Reform.

Recommendation

• Senior leaders and ministers within governments must be champions for trustworthy, high-quality, valuable analysis and must have the skills and knowledge to carry out this role.