Review of pre-release access in Scotland

Public trust in statistics

In this section, we consider the available data from the British Social Attitudes Survey and the Public Confidence in Official Statistics survey (PCOS) to explore trends and patterns in trust in official statistics. We also consider discussions from stakeholders and the public engagement session as to what affects trust in statistics and the impact that pre-release access has on public trust.

Trends and patterns in public trust in official statistics

To understand current trends and patterns in levels of public trust in official statistics, data from the British Social Attitudes Survey (BSA) (2014, 2016, 2018) and the Public Confidence in Official Statistics survey (PCOS) (2021) were analysed. While specific survey data on public attitudes to official statistics produced in Scotland is not available, data from BSA and PCOS provide robust, representative insight into public trust in official statistics in Britain. As these surveys are part of a time series, this data allows the exploration of trends in these attitudes over time.

BSA and PCOS include questions on people’s attitudes towards official statistics, including whether they trust institutions such as the Office for National Statistics (ONS), and reasons why they do or do not trust these institutions and the statistics they produce, including:

• Please indicate whether you tend to trust this institution or tend not to trust it: The ONS.
• How much trust do you have in statistics produced by ONS? For example, on unemployment, inflation, economic growth, or life expectancy.

When asked the extent to which they trusted the Office for National Statistics (ONS) as an institution, people could indicate that they ‘Trust it a great deal’, ‘Tend to trust it’, ‘Tend to distrust it’, or ‘Distrust it greatly’. The majority of respondents (68%) indicated that they trusted the ONS ‘a great deal’ or ‘tended to trust it’ (Table 1.1, BSA 2018).

When asked the extent to which they trusted statistics produced by the ONS, people could indicate that they ‘Trust them greatly’, ‘Tend to trust them’, ‘Tend not to trust them’, or ‘Distrust them greatly’. Similarly to trust in the ONS as an institution, the majority (69%) of respondents trusted statistics produced by the ONS ‘greatly’ or ‘tend to trust them’ (Table 1.2, BSA 2018). While the proportion of people who said that they ‘Trust them greatly’ or ‘Tend to trust them’ was higher in 2021 compared with other years, differences in the way that the question was asked means that direct comparison between these waves is problematic.

Over time, it appears that the Scottish public has become more trusting of the ONS as an institution. In BSA 2014, people from Scotland (60%) were less likely than those from England (68%) and Wales (74%) to trust the ONS as an institution ‘a great deal’ or to ‘tend to trust it’, compared with figures in 2016 and 2018, and in the 2021 PCOS survey, which show that trust evened out between the three nations (Annex table 1.3). For example, in the 2021 PCOS survey, people from Scotland (82%) were similarly likely to those from England (83%) and Wales (81%) to trust the ONS as an institution.

Similarly, when asked the extent to which they trusted statistics produced by the ONS, BSA 2014 respondents from Scotland (62%) were less likely than those from England (67%) and Wales (68%) to trust these ‘a great deal’ or ‘tend to trust them. These proportions became increasingly similar between Scotland and England throughout 2016 and 2018, and in the 2021 PCOS survey, people from Scotland were more likely to trust statistics produced by the ONS (90%), than people from England (84%) and Wales (82%) (Annex table 1.4).

Demographically, those less likely to trust the ONS were young people aged 18-34 years old (compared with those in older age groups), those who were unemployed or retired (compared with those who were employed), those who were employed in semi-routine and routine roles (compared with those in managerial and professional roles), and those who were renting their homes (compared with those who were buying or owned their homes). These trends hold across trust in the ONS as an institution and trust in statistics produced by the ONS. For example, when asked how much trust they had in statistics produced by the ONS, people aged 18-34 years old (67%) were less likely than those aged 35-54 (75%) to ‘trust them greatly’ or ‘tend to trust them’ (Table 1.5). When asked the extent to which they trusted the ONS as an institution, those who were unemployed (52%) or retired (65%) were less likely to say they trusted it ‘a great deal’ or ‘tended to trust it’ than those who were employed (72%) (Table 1.6), and those who were in semi-routine or routine roles (52%) were less likely than those in managerial and professional roles (79%) to trust it ‘a great deal’ or ‘tend to trust it’ (Table 1.7).

Similar patterns were also seen in relation to a further question where respondents were asked the extent to which they agreed or disagreed with the statement “personal information that is provided to ONS will be kept confidential”. Those who were renting their homes were less likely than those who were buying or owned their homes to ‘strongly agree’ or ‘tend to agree’ that personal information that is provided to the ONS will be kept confidential, both among those who had participated in ONS studies before (86% of those who rent compared with 89% of those who were buying or own) and those who had not participated in ONS studies before (62% of those who rent compared with 71% of those who were buying or own) (Table 1.8).

Those who said they trusted statistics produced by ONS ‘greatly’ or ‘tended to trust them’ were then asked what their main reasons were for trusting ONS statistics. They were able to select up to three responses from a list of eight options (see Table 1.9 for full list). Respondents were then asked which of the three they selected was the most important reason for trusting ONS statistics.

The most common reason people gave for trusting the ONS was that the ONS does not have a vested interest in the results or does not manipulate production or collection of statistics. This was also selected as the most important reason that these respondents trusted the ONS, with between 29% and 41% indicating that this was their most important reason for trusting the ONS (Table 1.9).

Similarly, those who said they ‘tended not to trust’ statistics produced by ONS or ‘distrusted them greatly’ were then asked what their main reasons were for not trusting ONS statistics. They were able to select up to three responses from a list of ten options (see Table 1.9 for full list). Respondents were then asked which of the three they selected was the most important reason for not trusting ONS statistics.

The most common reason people who did not trust ONS statistics gave for not trusting ONS statistics was that the figures are misrepresented or spun by politicians or the media.(Table 1.10).

On this theme, in response to a further question which asked respondents whether they agreed or disagreed with the statement ‘statistics produced by the ONS are free from political interference’, in each year around half of the public said they agreed or strongly agreed (51%-55%), with men more likely than women to agree (59% compared with 51%) (Annex table 1.11, BSA 2018). Those who were employed were more likely than those who were unemployed or retired to agree with this (59% compared with 40% and 50% respectively) (Annex tables 1.12.1 and 1.12.2, BSA 2018), and those who were buying or owned their own home were more likely than those who were renting their home to agree or strongly agree (71% compared with 63%) (Annex table 1.13, PCOS). Respondents were also asked directly about pre-release access in BSA 2014, 2016, and 2018. Respondents were given the statement: Under the current rules, Government ministers are shown official statistics the day before [in England] / five days before [in Scotland / Wales] they are released to the public, and asked which of the following statements came closer to their view:

• The current rules are right; Government ministers alone should be shown official statistics before they are released to the public
• The current rules should be changed; Official statistics should be made equally available to everybody, including the public, at the same time

Overall, the majority (67%-71%) of people stated that the current rules should be changed, with between 25% and 29% of people stating that the current rules are right, with these proportions remaining fairly stable between 2014 and 2018 (Table 1.14). Women were more likely than men to state that the current rules should be changed (Table 1.15), as well as those who were in intermediate or lower supervisory and technical roles, compared with those in managerial and professional roles (Table 1.16).

Public trust in statistics and impact of current PRA on public trust

Participants in the public engagement session were broadly able to identify topics on which official statistics may be collected, and gave examples of the Census, health (including statistics released during the COVID-19 pandemic), crime, employment, nationality, immigration, income, weather and global warming. Most received information on these from news sources – on television, online or via social media. The reasons identified by public participants for the collection and collation of official statistics included to inform policy for the government, to compare different time periods and look for trends in the data and to gather all the relevant information in the same place to make it easier to read and understand.

Participants felt that they would be more likely to trust official statistics which were produced by an independent organisation rather than the Scottish Government, especially if the statistics were collected and published by a reputable organisation whose branding they recognised and if the statistics seemed plausible based on their personal experience. Others expressed distrust in statistics and felt that these could be manipulated or “made up”.

So I suppose I would have a bit more trust given that it's independent and it's not The Scottish Government. (Public participant)

Sometimes I think they just make them up as they’re going along. (Public participant)

There was a perception from stakeholders that the public was generally unaware of pre-release access and that pre-release generally had little impact on the public’s trust of statistics. As discussed in later sections, stakeholders acknowledged the possibility that some officials having prior sight of official statistics could be detrimental to public trust in statistics. However, they felt that the benefits of PRA, which are discussed in later sections, outweighed these and that PRA increases the accuracy and clarity of communication of official statistics to the public and, in turn, increases public trust in both official statistics and the government.

…I would guess that probably most people aren't really aware of what the sort of rules and arrangements are and that the perception of trust is more about the strength of the messages when they are communicated, which is coming from, in the case of Scottish Government anyway, most of the time, ministers. So yeah, probably more advantageous to trust to have that pre-release period and make sure that those messages are accurate and reliable. (External statistics producer)

When given information on pre-release access in Scotland and how it is used, public participants had mixed views on this and the ways in which it impacted their trust in the statistics. Questions were raised as to the extent that the Scottish Government could manipulate or withdraw statistics. It was suggested by one participant that PRA offered the opportunity to spin statistics, and there was a belief that statistics are frequently manipulated by politicians. The risk of statistics being leaked and of politicians being able to take financial advantage of statistics were also highlighted.

It depends what they're going to do with the information. Are they going to then see that the…whatever it says doesn't suit them, or they think it's going to cause an uproar? Would they then withdraw it?…Have they got the power to then do that? (Public participant)

…I think there might be risks about it, such as like leaks to the media and stuff. I don't know if that does happen, but are these trusted people fully trustworthy? (Public participant)

…it gives them time to get their spin on it. Tell us it's not so bad or whatever. Some people actually believe a lot of these things that they tell you when they do the spin. I don't think you should have previous advanced knowledge of these things. Let them work on their feet, think on their feet and react that way… rather than preparing for it because it's not going to change what's there. …I don't see an advantage of letting them having advanced knowledge of these things. (Public participant)

However, other participants identified potential benefits of PRA, including allowing the government to develop policies and responses, make plans and to prepare for any public response to statistics. PRA was also seen as helping the Scottish Government manage workload. The use of PRA to develop media handling strategies was raised, and there were differences in opinion as to whether this was positive or negative.

I do think so because they are just human and they need to process, they need to have time to process things just like anybody does, if they're given new information. So I do think they need some time to process. I don't know how much time, but some time to process things before, as you said, they're put on the spot and ask questions so they can give a more accurate answer, I suppose, and they're not fumbling and thinking on their feet, but I know the other chap said that they need to think on their feet perhaps… (Public participant)

I suppose the only one I can think of is like if the important people that need to see it are like not working on the day it's released and maybe five days gives them more opportunity to begin and see it before it goes out to the public. (Public participant)

For some participants in the public engagement discussion, their trust in statistics was inextricably linked with their trust in politicians and government more generally, and they also rarely distinguished between the Scottish and UK governments. It was clear that, for some, the handling of the Covid-19 pandemic had had a profound negative effect on their trust in government, and participants gave additional examples of politicians betting on the next leader of Conservatives and the recent cuts (and reversal of cuts) to winter fuel payments as impacting their trust in government.

…and it's schoolboy-ish is what we’re looking at. They've done it and they ran away and we're left with it, sort of that nonsense and that's why politics is... very mistrusted now. (Public participant)

Both stakeholders and public participants expressed the opinion that pre-release access in Scotland had a minor influence on public trust in statistics and that other factors, such as the public’s view on government and politicians more generally, had a much greater influence.

There are other things about public trust that probably contribute more. I have no evidence to support this, but my own feeling is that there are other things that we that we do that that that have an impact on trust. (Scottish Government statistics producer)

So the statistical release, is it gonna make a big difference in the big game? I don't think so. (Public participant)