Review of pre-release access in Scotland

Discussion

This review aimed to evaluate how the current processes around pre-release access to Official Statistics are working in Scotland and identify possible options for the Scottish Government. The project included a desk review of relevant publications, analysis of relevant data from the British Social Attitudes survey and the Public Confidence in Official Statistics survey, interviews and focus groups with stakeholders who use pre-release access as part of their role and a discussion with members of the public. This review is somewhat limited in scale as it was not possible within its scope to engage with large numbers of those who use PRA or with a larger group of members of the public. Challenges were experienced in the recruitment of Ministers and members of their private offices, meaning that their views are not represented in this review. The review is also limited in the extent to which it can examine public trust in statistics which are published in Scotland as there is no available survey data which looks specifically at the trust of Scotland’s population in statistics produced and released by the Scottish Government and other Scottish bodies such as Public Health Scotland and National Records of Scotland.

The arrangements for pre-release access in Scotland differ from those in place in other parts of the UK, and across the EU. Scotland currently has a PRA period of up to 5 working days for certain types of statistics. Changes to PRA arrangements in the UK in recent years have been largely driven by the work of the UK Statistics Authority and campaigns by the Royal Statistical Society. While noting that they are not directly able to set the rules on PRA, the UKSA has argued for legislative changes to grant them control over PRA and criticized the provisions in Scotland and Wales as going against the spirit of the Statistics Code of Practice. The Royal Statistical Society (RSS) has long campaigned against PRA, asserting it undermines public confidence and continues to advocate for PRA to be removed. The Independent Review of the UK Statistics Authority by Professor Denise Lievesley in 2024 strongly recommended the recommending alignment with the ONS approach across the UK. Awareness of the external drivers for changes to PRA was low amongst the stakeholders who contributed to the review, and some participants questioned the need for the review in the first place. There was a perception that PRA had been restricted in the UK Government due to issues with statistics being leaked.

The review considered the extent to which public trust was perceived as a driver of how PRA is managed. The survey data indicated that public trust in both the ONS as an institution and in statistics produced by the ONS have remained consistent since 2014, with around two-thirds of people in Britain trusting the ONS. In 2014, Scottish adults had lower levels of trust than those in England and Wales, but by 2021 levels of trust were comparable across all three nations. Those least likely to trust the ONS were younger people, those who were unemployed, those who were in semi-routine and routine professional roles, and those who were renting their homes. The primary reasons for people trusting ONS statistics included the perception that the ONS does not have a vested interest in the results and that they trust the figures from personal experience. Among people who did not trust ONS statistics, reasons given for doing so included the perception that the figures are misrepresented or spun by politicians or the media, and that the government has a vested interest in the results or interferes in production or collection. The public is generally in support of statistics being released at the same time to everyone, rather than Ministers receiving prior access. Survey data on public trust in official statistics produced by the Scottish Government is a gap in the evidence and, if any changes to PRA are implemented, it would be of interest to evaluate the impact of these changes on trust in statistics in the Scottish context.

The reasons for having greater or less trust in official statistics seen in the survey results were echoed by the public engagement group. Participants stated their trust in statistics was greater if they were produced by independent, reputable organisations but expressed concern about potential manipulation of statistics by the Scottish Government. Public participants also highlighted potential benefits of PRA such as allowing the government to prepare responses and manage workload, but expressed concern that data could be manipulated or withheld. Overall, pre-release access was seen by public participants and stakeholders as having a limited impact on trust. Public participants’ trust in statistics was more significantly influenced by their general trust in government and politicians, and the COVID-19 pandemic had a profound negative impact on this.

Experiences of the changes to PRA for economic and market statistics indicated that the implementation had been challenging in some cases due to difficulties in communicating the changes to PRA and the additional workload involved in amending PRA documentation. While the changes to PRA periods for economic and market sensitive statistics were now successfully implemented and most stakeholders understand the rationale behind the different arrangements for different types of statistics, some found that the shorter time had led to higher pressure and workload. It was noted this had led to staff working outside normal hours and that it potentially limits the ability to develop proactive communications, leading to less detailed responses.

Generally, the ways that stakeholders used PRA across the different professional groups aligned with what they perceived to be the purpose of PRA, and with the purpose of PRA as set out in the Pre-release Access to Official Statistics (Scotland) Order 2008, namely, understanding, interrogating, and querying statistics, briefing Ministers, preparing communication materials, and coordinating with local data providers. PRA also allows the Scottish Government to produce publications which combine external and internal statistics, assure quality, and correct errors, though it was noted that these latter activities were not the purpose of PRA.

Participants valued PRA for the time it provides to understand complex publications, brief Ministers, and prepare accurate communications. This was seen as especially beneficial for health statistics due to their volume and complexity. PRA was considered crucial in supporting Ministers to respond accurately and promptly to media and parliamentary queries, ensuring clear and informed communication. It also allows for immediate countering of inaccurate reporting and development of accessible infographics. For these reasons, PRA was generally seen as conducive to, rather than detrimental to, strong public trust in statistics. Challenges identified were generally limited to operational issues such as managing PRA lists and ensuring the right people receive access.

Some participants felt that differences in PRA arrangements across the UK reflected differences in culture and believed they were driven, at least in part, by historic problems with UK government statistics being leaked. This was not a behaviour that they saw within the Scottish Government. Some felt that the differences in PRA are justified by the principles of devolution and the relatively smaller size and staff complement of the Scottish Government, though a few argued for consistent arrangements across the UK.

Stakeholder and public participants were invited to consider three options for pre-release access: retaining the current arrangements; reducing PRA to 24 hours and removing PRA entirely. There was no consensus among the public participants regarding the options, and their views tended to align with their opinions on government and their trust in government and politicians more generally.

Among those stakeholders who participated in the review, there was a clear preference for retaining the current PRA arrangements of up to 5 working days. Notably, this review did not identify any practical or operational issues with the current arrangements. Retaining the current arrangements would leave Scotland out of line with other parts of the UK and with the recommendations from the Independent Review of the UK Statistics Authority by Professor Denise Lievesley, but this was perceived to be an issue primarily for those at the highest levels of governance, rather than those who work directly with statistics and PRA, who were broadly unaware of external drivers to remove PRA.

When discussing the options of reducing PRA to 24 hours and removing PRA entirely, 24 hours was seen as preferable to no PRA and stakeholders saw no practical benefits in removing pre-release access entirely. Stakeholders generally discussed similar potential negative impacts for both options, with these impacts being compounded in the case of PRA being removed entirely. Most participants felt the reduction to 24 hours or removal of PRA would negatively impact the government's ability to understand and communicate statistics accurately. This could lead to less informed responses from Ministers, undermining public trust in both official statistics and the government. While some acknowledged that removing PRA could theoretically increase transparency and public trust, they broadly disagreed with this, arguing that it would lead to greater public misinformation and in fact reduce trust in statistics. The Cabinet Office review of the reduction in PRA from 5 working days to 24 hours^[25] which came into force in December 2018 found that, overall, 24 hours PRA was workable and did not have a significant detrimental effect on the ability of government to do its work. However, it did identify issues related to dealing with large, complex publications and statistics produced by other departments and noted that whilst Ministerial briefings had remained of good quality, because of the reduction in PRA, they were of reduced comprehensiveness and depth. Nevertheless, the Cabinet Office review did not identify any instances of Ministers making inaccurate statements about statistics.

Stakeholders felt that reduced PRA would increase workload pressure, particularly for complex statistics and on busy parliamentary days. Practical challenges such as managing flexible working schedules and ensuring officials' availability were highlighted. Stakeholders felt that the removal of PRA would also increase workload and stress for staff, make roles less accessible to part-time or flexible workers, and necessitate significant cultural and practical changes in how responses are developed and approved. Participants also expressed concerns about potential leaks and unregulated access to statistics if the legitimate route of pre-release access is removed.

The Cabinet Office review noted that some of the officials who contributed to the review were working longer hours during the 24 hour PRA period and just over half of respondents were described as saying the overall impact had been negative but manageable. However, it should be noted that contributions to the Cabinet Office’s review were limited to private secretaries of Ministers, press officers, senior officials who are on pre-release lists primarily to receive briefing, officials who are on the lists to produce briefs for colleagues and Heads of Profession.

There are several considerations for whether or not any changes to PRA arrangements are implemented. Firstly, there is a need for clear communication from the Office of the Chief Statistician regarding the reasons behind any proposed change to PRA. With those working with PRA appearing to be generally unaware of any external pressure to restrict PRA, there is a risk that changes may be interpreted as resulting from leaking or other mishandling of statistics under PRA. Likewise, if Scotland is to retain 5 working days PRA or retain a PRA arrangement that differs from other parts of the UK, the reasons for this should be well evidenced and communicated.

There is evidence that believing statistics are produced independently and free from political interference is an important factor in trusting statistics. It is notable that, despite the independent nature of the ONS, mistrust in the ONS endures. Some believe that ONS figures are misrepresented by media or spun by politicians and that the government has a vested interest in the results or interferes in the production or collection. Given that the production of official statistics in Scotland sits as part of the Scottish Government, it may be even more important to emphasise the independence of statistics production and communicate the safeguards which are in place to ensure that official statistics are robust to increase trust in statistics in Scotland.

The British Social Attitudes survey suggests that a majority of people believe that the rules should be changed, on the basis of a very brief description of the current arrangement. While not designed to be representative of the population, the public participant group showed that people’s views on PRA are more nuanced, with participants changing their minds on PRA following discussion. Reflections from both stakeholders and public participants suggest that public attitudes towards, and trust in, government generally are potentially more important in influencing trust in official statistics than pre-release access arrangements. It may be that these general attitudes towards government are also a factor in people’s opinions regarding whether PRA rules should be changed. Stakeholders identified potential negative impacts of reduction or removal of PRA on public trust in statistics as a result of incorrect or misinformed statements from government. It is clear that public trust in statistics is complex, influenced by a number of factors, and that increasing public trust in statistics may not be easy or straightforward. At a time when both trust in the Scottish Government^[26] and response rates to national surveys are falling,^[27] it may be of value to explore additional ways to increase public trust in statistics.

A suggested change to pre-release access beyond changes to the number of days was the potential extension of PRA to elected representatives of opposition parties who have a remit for particular topic areas, such as members of the shadow cabinet. It was suggested that this has the potential to make PRA more equitable However, it would also further expand the number of people on PRA lists. While the Scottish Government has limited ability to address or rectify breaches of PRA by opposition parties, and this could potentially increase the risk of misuse of PRA, it should be noted that people within external agencies outwith the Scottish Government are included on some PRA lists.

There are also opportunities to improve processes within the Scottish Government. Stakeholders highlighted some challenges in adding people to PRA lists to cover holidays and absences, and in knowing who is on the PRA list for upcoming publications. It may be that a centralised PRA function within the Scottish Government would provide greater visibility of lists internally and streamline processes.

While there is undoubtedly a workload burden associated with the current PRA arrangements, particularly for statisticians in terms of responding to queries from various officials over the 5 working day period, stakeholders did not seem to find this problematic. However, stakeholders identified several significant potential impacts of reducing or removing PRA on their workloads and their wellbeing at work. These included increased pressure and stress, and having to work longer hours outside of their normal working hours. Any changes to PRA should therefore be carefully considered in relation to these issues as well as in the context of the Scottish Government’s commitment to flexible and part time working as changes may disproportionately affect those who work part time or have flexible working arrangements.

Practical and cultural changes would be needed to support reduction or removal of PRA. Stakeholders cited the process by which government responses are approved and signed off by statisticians, directors, policy leads, special advisers and Ministers as constituting a barrier to efficient decision-making, and new approaches, such as live calls involving the relevant decision-makers, would have to be developed to for responses to be developed in a timely way. Changes to PRA, and the reasons for these changes, should be clearly communicated and there would have to be a change in expectations of those who currently receive PRA as to the level of analysis and insight they could receive with a shorter or no PRA period.

It would be useful to follow the example of the Cabinet Office where, if any changes to PRA were brought in, the Office of the Chief Statistician would review the impact of these changes on the quality of responses and briefings and on workload. As noted above, it would also be of interest to evaluate whether these changes had resulted in any change in trust in official statistics produced in Scotland.