Public Confidence in Official Statistics

1. Introduction

1.1 Background

This report presents findings on public attitudes to official statistics in Britain in 2023, based on results from a web-postal survey of adults aged 18 and over, conducted by the National Centre for Social Research (NatCen). The report explores levels of awareness, use of, and trust in official statistics in Britain. It also explores changes in attitudes to official statistics over time and variation in attitudes by demographics. 

The research was commissioned by the UK Statistics Authority (the Authority), an independent body at arm’s length from Government with the statutory objective of promoting and safeguarding the production and publication of official statistics ¹ .

Its executive office, the Office for National Statistics (ONS), is the UK’s National Statistical Institute and largest producer of official statistics. The Authority also has an independent regulatory function (Office for Statistics Regulation or OSR), which ensures that statistics are produced and disseminated in the public interest and acts as a watchdog against misuse of statistics. 

The Public Confidence in Official Statistics survey (PCOS) has been run at regular intervals since 2004, most recently as an independent survey in 2021, but also as part of NatCen’s face-to-face British Social Attitudes survey or BSA (in 2014, 2016, 2018) ² .

The full list of questions can be found in Appendix A of the accompanying Technical Report. 

1.2 Methods

The survey was designed to yield a representative sample of adults aged 18 or over in England, Wales and Scotland, using a sample of addresses drawn from the Postcode Address File (PAF).

As in 2021 the 2023 PCOS was run as a standalone push-to-web survey. Respondents were encouraged to complete the 10 to 15-minute survey online, but paper self-completion surveys were offered to all non-responding households to maximise response and sample quality. Fieldwork took place between 4th October and 17th December 2023. Up to two adults per sampled address could complete the survey. Interviews were achieved with a representative sample of 2,364 adults aged 18 and over in Britain from 1,695 households. This represents an adjusted household response rate of 21.2%. The majority of respondents, 1,718, completed the survey online.  More information on the survey methodology and response rates can be found in the accompanying Technical Report. 

Differences between the 2021 and 2023 data presented in the report are statistically significant (at the 95% level). Where differences by age, sex, education, or employment are commented on in the text these differences are statistically significant at the 95% level. When testing for statistical significance, including between years, as in previous years ‘don’t know’ responses were excluded from the analysis (see below for more on the reasons for this). 

Data relating to individual questions does not always add up to 100% due to rounding. 

1.3 Measuring change over time

One of the aims of this report is to compare how people’s awareness of and attitudes towards official statistics have changed since data was last collected in 2021. There are reasons to expect that awareness may be lower in 2023 compared with 2021. The decennial census, which is run in England and Wales by ONS and which all households are required to complete, was conducted in 2021 ³ . In addition, the COVID-19 pandemic, which began in early 2020 and continued throughout 2021, gave additional prominence to official statistics and their role in public policy making with COVID-19 statistics being quoted daily in the media. 

There is a need for caution when making comparisons between the 2021 and 2023 surveys and earlier years of the study given the change in survey methodology and the move from a face-to-face to a push-to-web survey in 2021. 

Changing how a survey is conducted brings a risk both that the profile of people who respond may be different and that the way those that do respond answer the questions may vary. 

PCOS 2023 has been designed to minimise as far as possible the impact of the change in the mode of data collection on the comparability of the data over time. Weighting has been used to ensure that the final sample in 2023 is similarly representative of the population on key characteristics such as age, sex, education, region, tenure and ethnicity as previous years ⁴ .

The impact of methodological change on findings is expected to be relatively small but it is not possible to say for certain the extent to which any changes in attitudes (or lack thereof) observed across time are down to real-world change or methodological change. Where it is thought particularly likely that methodological change may explain all or part of the trend observed over time this is flagged in the report. 

In 2021 and 2023, the proportion of respondents with ‘don’t know’ /’not answered’ responses was significantly lower than in 2018. The analysis in the report, including any comparisons over time, focuses on comparing responses among those able to express an opinion, excluding any ‘don’t know’ or other missing responses unless otherwise stated. Further information on the reasons for and implications of this are discussed in the Methodological Note which can be found at the end of this report. Further details of the proportion of ‘don’t knows’ given at each question can be found in the summary tables accompanying the technical report.

2. Awareness and use of ONS

Key findings 

• Awareness of ONS and the statistics it produces has remained stable. In both 2023 and 2021 around three-quarters of respondents had heard of ONS. 
• Just over a third of respondents (35%) reported that they had used ONS statistics (compared with 36% in 2021).
• More than two thirds (69%) of the respondents reported that they recalled taking part in an ONS survey including 61% who recalled taking part in the census (down from 73% in 2021) and 13% who had taken part in the Coronavirus Infection Survey.
• 92% of respondents who reported that they had taken part in an ONS survey agreed that personal information provided to ONS would be kept confidential (the same proportion of respondent as in 2021). 

2.1 Awareness of ONS

In 2023, three-quarters (74%) of respondents indicated that they had heard of the Office for National Statistics (ONS) ⁵  (compared with 75% in 2021).  Reported awareness of ONS remained lower than awareness of other public sector bodies such as the Bank of England (93%) and the Department for Work and Pensions (88%). A similar pattern was also observed for the other organisations asked about in the survey (Figure 2.1). 

The proportion of respondents who had heard of ONS and who said they knew ONS ‘well’ increased compared with 2021 (from 18% to 20%). There was a corresponding decrease (from 50% to 48%) in those who said they knew ONS ‘somewhat’. Around a third of respondents who indicated that they had heard of ONS reported they had only heard the name (32%); a similar proportion to in 2021 (33%). 

Figure 2.1 Proportion of respondents who reported they had heard of ONS and comparator organisations

2.2 Use of ONS statistics

As well as measuring awareness, the survey also asked respondents whether they had had any engagement with ONS, either through using statistics ONS produced or by taking part in one of their surveys (see section 2.3).

The proportion of respondents who reported using or referring to ONS statistics remained consistent between 2021 and 2023 (35% in 2023 compared with 36% in 2021). This was split between people saying they used ONS statistics occasionally (24%), frequently (4%), or at least five years ago (7%).

Respondents who answered that they had used or referred to ONS statistics ‘frequently’ or ‘occasionally’ were asked how long they had been using these statistics. Almost one in ten respondents (9%) reported they had been using ONS statistics for less than a year, and a further 35% reported using them for two to five years. Over a quarter (28%) of frequent or occasional users used ONS statistics at least a couple of times a month. This has decreased since 2021 where 39% of frequent or occasional users used ONS statistics at least a couple of times a month.

Figure 2.2 Proportion of respondents who reported using ONS statistics

2.3 Participation in ONS surveys

In 2023, more than two thirds of respondents (69%) reported having ever taken part in an ONS survey compared with 78% in 2021. 61% of respondents reported that they took part in a census ⁶  in 2023 compared with 73% in 2021. The lower proportion who reported taking part in the census probably reflects the fact that the decennial census took place in England and Wales in March 2021. The fieldwork for the 2021 PCOS survey was conducted between October and December 2021, which may have been easier for respondents to recall completing a census. The further away the survey is from a census year, the more people are likely to forget having taken part or to have been too young to take part. 

Whilst respondents most commonly reported having taken part in the census, more than one in ten (13%) respondents in 2023 reported taking part in the Coronavirus Infection Survey. This is a higher proportion of the population than was actually sampled for the survey ⁷ . This may be because respondents incorrectly recalled participating or that those that participated in PCOS were also more likely to participate in the Coronavirus Infection Survey. 

Figure 2.3 Proportion of respondents who reported they had participated in an ONS survey

It is important that people taking part in data collection trust that their personal information will be kept confidential. When asked if they thought that personal information provided to ONS would be kept confidential, 90% agreed (with 42% strongly agreeing). Respondents who reported taking part in surveys were more likely to agree compared with those who had not (92% compared with 84%), although even among those who indicated they had not taken part in an ONS survey most still agreed that their data would be kept confidential. 

2.4 Differences by demographic characteristics

Reported awareness of ONS and use of ONS statistics varied significantly by education and by socio-economic status. Respondents with a degree were most likely to report having heard of ONS (83%) and those with no qualification the least likely (64%). Similarly, over half of respondents with a degree (55%) reported that they had used ONS statistics compared with 13% of those with no qualifications. Seventy-eight percent of respondents in managerial or professional occupations had heard of ONS whilst 42% had used ONS statistics. The equivalent figures for respondents in routine or semi-routine occupations were 68% and 26%. 

Men were more likely to have heard of ONS than women (79% compared with 70%) and were also more likely to have ever used ONS statistics (41% compared with 31%). 

Awareness of ONS also varied by age with only 53% of respondents aged 18-24 having heard of ONS. Despite being among the most likely to have heard of ONS, respondents aged 65 and over were less likely than other age groups to have used ONS statistics (27%).

3. Trust in ONS and official statistics

Key findings

• Trust in ONS remained high; 87% of respondents said that they tend to trust it or trust it a great deal. 
• Respondents generally also trusted ONS statistics; 85% of respondents said that they trust statistics produced by ONS. 
• Reported trust in ONS was higher among people who stated they frequently use ONS statistics (98%), but still high among non-users (80%). 
• The most common reason respondents gave for trusting ONS was that ONS did not have a vested interest in or manipulate the results. 
• The main reasons respondents gave for not trusting ONS relate to how statistics were used by the media and politicians.

3.1 Trust in ONS overall

Most respondents trusted the Office for National Statistics (ONS), with 87% of respondents who gave an opinion indicating that they tend to trust it (71%) or trust it a great deal (16%). This is comparable with previous years (Figure 3.1). Trust in ONS was generally high, however respondents who have heard of ONS (91%) were more likely to trust the organisation compared with those who had not (72%). 

Those who have used official statistics were more likely to trust ONS than those who have not used them. However, trust in ONS was high even among non-users. Eighty-two percent of non-users said they trusted ONS compared with 99% of frequent users of ONS statistics.

As in previous years, PCOS 2023 asked about the level of trust in ONS compared to other institutions in British public life (Figure 3.2). Of all institutions included to this question, ONS had the highest level of reported trust of all (87%) among respondents, similar to levels of trust in the courts (82%) or the Bank of England (79%). Trust was lowest in the British media (25%), the Government (31%) and the UK Parliament (36%), a finding which is consistent with previous years. Compared to 2021, there was a noticeable decrease in the level of trust respondents reported in a number of state-led institutions such as the UK Parliament, the Government, the civil service, the police, and the Bank of England. However, trust reported in ONS, the media, the courts and high street banks and financial institutions remained consistent with the answers from the previous survey year.

Figure 3.1 Proportion of respondents that reported trusting ONS 

Figure 3.2 Proportion of respondents that reported trusting different institutions in British public life

3.2 Trust in ONS statistics 

As well as trust in institutions, respondents were asked whether they trust statistics produced by ONS (Figure 3.3). As in previous years, respondents generally reported trusting ONS statistics. In 2023, 85% of respondents said they trusted ONS statistics (with 18% saying they trusted them a great deal and 67% saying they tended to trust them). The level of trust is similar to 2021. As with overall trust in ONS, trust in ONS statistics was higher among respondents who had used ONS statistics (98% of those who used ONS statistics frequently trusted them) but still high among non-users (80%).

Figure 3.3 Proportion of respondents who reported that they trust statistics produced by ONS

Respondents who indicated that they trust ONS statistics were asked their reasons for doing so and could select up to three reasons from a predetermined list (Figure 3.4). The reason most frequently chosen by respondents was that ONS did not have a vested interest in or manipulate the results (63%). The next most common reason given was that ONS were experts in statistics (56%). 

Similarly, the (smaller proportion of) respondents who indicated they do not trust ONS statistics were asked their reasons for not doing so (Figure 3.5). The most frequently chosen reasons were that the statistics were being misrepresented by politicians (49%), the statistics alone do not tell the whole story (45%) and that the Government has a vested interest in or manipulates the results (41%). This is consistent with the low levels of trust reported in the UK Parliament and the Government (Figure 3.1) and concerns about how the government uses statistics (see section 4). When people were asked their most important reason for not trusting ONS statistics, the most common responses were that the Government has a vested interest in or manipulates the results (24%) and statistics alone do not tell the whole story (23%). This is consistent with the results from 2021.

Figure 3.4 Reasons for trusting ONS statistics

Figure 3.5 Reasons for not trusting ONS statistics

3.3 Differences by demographic characteristics

In this survey, trust in ONS and the statistics it produces increased with level of education. People educated to degree level or above were most likely to report that they trust ONS (93%) and the statistics they produce (92%) whilst those with no qualifications were least likely to report that they trust ONS (68%) or the statistics they produce (70%). Similar patterns of trust by education were also found for the Civil Service and the courts. 

People in managerial and professional occupations or intermediate occupations were more likely to trust ONS (both 90%) compared with those in technical or semi-routine or routine occupations (81% and 83%). A similar pattern of trust by occupation was also found in trust in the Civil Service. 

Most respondents trusted ONS regardless of age. There was no difference by age in trust in the government, the Bank of England, high street banks and financial institutions or the media.

4. General attitudes to statistics 

Key findings

• Almost two-thirds (60%) of respondents reported seeing statistics on the news at least several times a week. 
• Most respondents agreed that statistics produced by ONS are important to understand our country (90%) and that official statistics are accurate (83%). 
• Most respondents agreed that statistics produced by ONS are free from political interference (72%). However, they are less positive about the government and the media’s presentation of statistics. 68% of respondents disagreed that the government presents statistics honestly whilst 75% disagreed that the newspapers present statistics honestly. 
• In this survey, users of statistics generally hold more positive attitudes towards statistics than non-users. However, both users and non-users are sceptical about government and media presentation of statistics.

4.1 Exposure to statistics

We estimate that nearly two thirds (60%) of the public  would have reported seeing statistics on the news at least several times a week, including over a quarter (27%) who reported seeing statistics on the news daily. A smaller proportion of respondents reported seeing statistics on social media, with almost half (49%) reporting they saw statistics on social media at least several times a week (Figure 4.1).

Figure 4.1 How often respondents reported seeing statistics on social media and the news

There were significant differences by age in terms of whether respondents tended to report seeing statistics on the news or social media. Respondents aged 44-54 and 55-64 were the most likely to report seeing statistics daily on the news (both 32%). Nineteen percent of respondents aged 18-34 reported never consuming news at all (about statistics or otherwise). Conversely, respondents aged 18-24 were the age group most likely to report seeing statistics daily on social media (41%), while respondents aged 65+ were least likely to report this (8%) and most likely to report not using social media at all (35%). 

Respondents were also asked whether they think statistics have helped them to make decisions about their life (Figure 4.2). The majority of respondents indicated that they disagreed that statistics have helped them make decisions about their life (60%), with 22% strongly disagreeing compared to only 6% strongly agreeing. Compared to 2021, respondents to the 2023 survey were more likely to disagree that statistics have helped them make decisions about their life, with overall agreement dropping from 52% to 40%.

Figure 4.2 Proportion of respondents indicating that they agree that statistics have helped them to make decisions about their life

4.2 Importance and accessibility of official statistics

There was strong support for the idea that statistics produced by the Office for National Statistics (ONS) are important to understand our country. In 2023, 90% agreed this was the case, similar to 2021 and 2018, when levels of agreement with this statement were already very high (Figure 4.3). 

Figure 4.3 General attitudes to official statistics

Respondents were also asked whether they thought official statistics were easy to find and easy to understand (Figure 4.4). The majority agreed that statistics were easy to find (64%) and to understand (66%), similar to 2021.

Figure 4.4 Proportion of respondents indicating whether they agree or disagree that official statistics are easy to find and understand

4.3 Perceived accuracy of official statistics

Similar to 2021, most respondents tended to agree that official statistics are accurate (83%) and most also indicated that they agree that statistics produced by ONS are free from political interference (72%). Despite most respondents reporting that statistics are produced free from political interference, they were less positive about how the government and the media present statistics. The proportion of survey respondents who indicated that they agree that the government presents official statistics honestly remained relatively low (32%, compared to 35% in 2021). Even fewer respondents (25%) tended to agree that newspapers present official statistics honestly. This was the case both among respondents who did not consume news at all (20% of whom agreed that the media presents statistics honestly) and those who saw statistics in the news at least a couple of times a week (26%).

4.4 Do attitudes vary between statistics users and non-users?

Respondents who reported using ONS statistics were generally more positive about official statistics than those who do not use statistics produced by ONS (Figure 4.5). Frequent users of statistics were more likely to agree that statistics are important to understand our country (94% vs. 87% of non-users), that statistics produced by ONS are free from political interference (82% vs. 68%), and that official statistics are accurate (91% vs. 80%). Users of statistics were also more likely to agree that official statistics are easy to find (85% vs. 56%) and easy to understand (85% vs. 60%). However, there were no significant differences between users and non-users of statistics in terms of whether they thought the government presents official statistics honestly when talking about its policies or whether newspapers present official statistics honestly.

Figure 4.5 Proportion of users and non-users of ONS statistics indicating they agree with statements regarding attitudes towards statistics

4.5 Differences by demographic characteristics

Respondents’ attitudes towards official statistics varied significantly by level of education, perhaps reflecting the fact that more highly educated people were more likely to have used statistics (see section 2). Among those able to express an opinion, people educated to degree level or above were the most likely to agree that statistics were important to understand our country (95%), that statistics were free from political interference (79%) and that statistics were generally accurate (90%). Respondents with higher levels of education were also more likely to agree that official statistics were easy to find (71%) and to understand (74%). Conversely, those with no educational qualifications were the least likely to agree with these statements (57% agreed that official statistics were easy to find and 56% that they were easy to understand). 

Respondents aged 25-34 were the most likely to agree that statistics were easy to find (73%) and those aged 18-24 were the most likely to agree that statistics are easy to understand (77%), whilst respondents aged 65 or older were least likely to agree that statistics are easy to find and understand (48% and 52% respectively). This may reflect lower use of ONS statistics among the oldest age group (see section 2) or, perhaps, the lower likelihood of older age groups being internet users, given that statistics are most readily available online. 

5. Attitudes towards specific statistics

Key findings

• Use of official statistics among respondents has continued to rise for all data series except for the census, which saw a five percentage point decline in use since the PCOS 2021.
• In 2023, of all data series asked about, the majority of respondents agreed that official statistics are accurate and reflect changes in the UK. This was similar to previous years. However, the exact level of agreement varies depending on the statistics.
• Respondents had mixed views on whether specific statistics are free from political interference. While over three-quarters of respondents (76%) agreed this is the case for the census, the figures for crime statistics (53%) and employment statistics (56%) are much lower. 
• Most respondents who used statistics agreed that the statistics they have used are released quickly. The exception is the census and crime statistics, with only 71% and 75% of respondents estimated to agree that they are released quickly, respectively.

5.1 Use of specific ONS statistics

As in previous years, the 2023 survey contained an array of questions focusing on specific sets of statistics (data series) produced by the Office for National Statistics (ONS): the census, the Consumer Price Index (CPI), employment and unemployment statistics, Gross Domestic Product (GDP) and crime statistics. Respondents were asked whether they had used any of these statistics and whether this was within the last five years or not. 

Estimated public use of specific data series produced by ONS is relatively low (Figure 5.1). However, overall use of statistics has been reported to be gradually increasing since PCOS 2018. We estimate that all data series excluding the census increased in use by approximately three percentage points since PCOS 2021, and approximately 10 percentage points since PCOS 2018. After the census at 63%, Crime and CPI statistics were most used in 2023, with 31% and 26% of respondents indicating that they were using them, respectively.

In 2023, the greatest change in reported use of statistics was for the census, with 63% of respondents indicating that they had used the census compared to 68% in 2021. While census usage among respondents for 2023 is still high in comparison to other data series, it had reduced in comparison to 2021, when the last census for England and Wales was conducted. This decline since 2021 is likely to be due to a reduction in interest and familiarity that may have heightened response to the 2021 census data collection. It is also possible that the 2021 PCOS data on the census usage was inflated by respondents confusing participation in the census with usage of it, and this may still apply, to an extent, to PCOS 2023. In addition to this, the proportion of Scottish respondents who had reported to have used the census in the last five years increased from 2021 to 2023, although the overall proportion of Scottish census users remained approximately the same. This is likely due to the publication of Scottish census data falling around the same time as the 2023 PCOS data was being collected.

Figure 5.1  Proportion of respondents that have reported to have used statistical series by ONS

5.2 Which statistics reflect what is happening in the UK?

Respondents were also asked for their opinions on different aspects of the data series discussed above. To begin with, respondents were asked whether they thought the series reflected changes in the UK, one measure of accuracy. As shown in section 4, the majority of respondents agreed that, overall, statistics produced by ONS were accurate. There were also high levels of agreement regarding the accuracy of specific data series, although the exact level of agreement varied by data series (Figure 5.2). Based on opinions reported, respondents were most likely to believe the census reflected changes in the UK (85%) and least likely to agree that the crime statistics did so (74%). There was no data series that saw an increase in agreement compared to PCOS 2021. We estimated that agreement in Crime statistics reflecting changes in the UK decreased most significantly by three percentage points, while data series such as the census and CPI statistics saw no major fluctuations.

Figure 5.2 Percentage of respondents reporting agreement that statistics reflect changes in the UK

5.3 Which statistics are free from political interference?

While 72% of respondents reported agreeing that, overall, the statistics produced by ONS are free from political interference (see section 4) there was more scepticism about the independence of some specific data series (Figure 5.3).

For all data series the proportion of respondents who agreed the statistics were free from political interference was higher than the proportion who disagreed. As with previous years, there was still a considerable proportion of respondents who questioned the statistics’ independence. This, however, made an exception of the census, with just over three-quarters of respondents (76%) reporting they view the census as free from political interference. Concerns about political interference were highest for crime statistics, with 47% of respondents reporting they disagree that they were free from political interference. Attitudes towards the independence of statistics were largely unchanged compared to 2021. 

Figure 5.3 Percentage of respondents reporting agreement that statistics are free from political interference

Which statistics are useful and timely?

Respondents who reported using the different statistics were asked additional questions about whether they thought these statistics were useful and timely. Most respondents that reported to be statistic users agreed the data series they had used were useful (Figure 5.4). Respondents were most likely to report agreement that the CPI statistics were useful, while all other data series presented similar levels of agreement at 89%. All data series except for crime statistics presented consistent levels with 2021, with agreement over the usefulness of crime statistics declining by 4 percentage points.

Figure 5.4 Percentage of respondents considered to be statistic users who reported that they agreed that statistics are useful

In addition to agreeing that data series were useful, most respondents who reported to be data users also believed the statistical series asked about were timely, that is they were released quickly to the public. As detailed in Figure 5.5, in 2023, most PCOS respondents reported to agree that the CPI, employment statistics, GDP and crime statistics were released quickly. However, as with previous years, respondents were much less likely to agree that the census was timely, with only around seven out of ten (71%) reporting they agree that it was released quickly, compared to nine out of ten (93%) agreeing for the CPI. This is reasonable given that the census is only conducted every 10 years, and subsequent data is released 1 to 4 years after collection.

Figure 5.5 Percentage of respondents considered to be statistics users who reported that they agreed that statistics are release quickly

5.4 Differences by demographic characteristics

Respondent’s attitudes towards specific statistical series produced by ONS varied by education, occupation. and age in similar ways to what was described for general attitudes to statistics in section 4. 

While the census showed similar usage across all demographics in PCOS 2023, of those that did use the census, respondents who were degree holders were more likely to report agreement that the statistics reflected changes in the UK.

Respondents with a degree were most likely to have reported use of many of the data series (40% used crime statistics, 36% used the CPI, 35% used GDP statistics, and 34% used employment statistics). Respondents with a degree were also more likely to indicate they agree that statistics reflected changes in the UK and were free from political interference.

Respondents in managerial and professional occupations were also more likely than those in other occupations to have reported use of GDP (29%) and crime statistics (35%), while employers in small organisations and own account workers were more likely to have reported use of the CPI (46%). Use of employment statistics was equally popular among both respondents in managerial and professional occupations as well as employers in small organisations and own account workers at 29%.

Respondents aged 18 to 24 years were most likely to have reported use of crime statistics (49%), GDP statistics or employment statistics (both 36%). Broadly speaking, use of these data series decreased with age, falling to the lowest level among those aged 65+.

6. Perceptions of UK Statistics Authority and Office for Statistics Regulation

Key findings

• Less than half of respondents had heard of the Authority or OSR. Around a fifth (22%) of respondents reported to have had some or a great deal of knowledge of the Authority whilst 28% had just heard the name. Similarly, a fifth  (18%) of respondents reported to have had at least some knowledge of OSR whilst 26% had just heard the name.
• Although respondents did not necessarily know about the Authority or OSR there was strong support for their role. Most respondents (95%) reported agreement that there should be an independent body to speak out against the misuse of statistics and 94% agreed such a body should ensure that statistics are produced free from political interference.

6.1 Knowledge of the Authority and OSR

In 2023, respondents were asked about how well they knew both the UK Statistics Authority (the Authority) and the Office for Statistics Regulation (OSR). Respondents were less familiar with the Authority and OSR compared with ONS (see section 2). Only 19% of respondents reported to have known the Authority: 3% said they knew the Authority well and 19% said they knew it somewhat (Figure 6.1). Awareness of the Authority has increased marginally by two percentage points since asked in 2021. More than half of respondents (56%) reported that they had not heard of OSR, similar to previous years. Only 2% of respondents reported knowing OSR well and only 16% knew it somewhat.

Figure 6.1 Level of knowledge of UK Statistics Authority and Office for Statistics Regulation reported among respondents in 2023

6.2 Importance of the Authority's role

Although the Authority was not widely known, respondents agreed that its role is an important one.

Almost everyone (95%) able to give a response to this question asked agreed that it is important for an independent body such as the Authority to speak out against the misuse of statistics (Figure 6.2), including 66% of the public strongly agreeing. 

Figure 6.2 Percentage of respondents reported to have agreed that it is important for UK Statistics Authority to speak out against misuse of statistics

Similarly, 94% of respondents reported to agree that it is important that an independent body ensures that official statistics are produced without political interference (Figure 6.3), including 63% who agreed strongly. 

Figure 6.3 Percentage of respondents that reported to have agreed that it is important for UK Statistics Authority to ensure that official statistics are produced without political interference

6.3 Differences by demographic characteristics

Knowledge of the Authority varied by occupation, with respondents in lower supervisory and technical occupations (26%) reporting to be most likely to be aware of the Authority, followed by managerial and professional occupations (24%) and employers in small organisations or own account workers (23%). Knowledge of OSR showed no significant variations by occupation, nor by education for both OSR and the Authority.

In the same way as they were more likely to have heard of ONS (see section 2) male respondents were more likely than female respondents to say they were aware of the Authority (25% vs 19%) and OSR (20% vs 16%). 

Agreement with the need for an independent body to oversee the use and production of statistics was very high across all demographic groups. However, there were some differences by education and occupation. Respondents with higher levels of education were more likely to agree that it is important for the Authority to speak out about misuse of official statistics (97% of respondents with a degree or higher compared with 92% of those with no qualifications) and that it was important for official figures to be free from political interference (96% vs 90%). Those in managerial and professional (97%) and intermediate occupations (96%) were also more likely to agree it is important for the Authority to speak out against the misuse of official statistics than those in lower supervisory and technical (91%) and semi-routine and routine occupations (92%). 

7. Methodological note

Throughout this report, results are presented exclusive of ‘don’t know’ and ‘prefer not to say’ responses, that is for all respondents able (and willing) to give an opinion rather than all adults 18+. As noted in the report introduction, there was a large difference in the level of ‘don’t know’ responses in 2021 compared with 2018. A similarly low level of ‘don’t know’ responses was also observed in 2023. 

The decision to present analyses exclusive of ‘don’t know’ responses was taken for two reasons. First, the approach is consistent with that used for the majority of the analysis in previous years’ reports. Second, it is not known whether the reduction in ‘don’t knows’ in 2021 and 2023 was driven by a genuine increase in knowledge and awareness across the population or the self-selection of more engaged respondents into the 2021 and 2023 responding samples. Controlling for potential differences in sample composition by focusing on those respondents able to give an opinion in each year represents the most appropriate way to isolate real world change. It should however be acknowledged that ignoring the reduction in ‘don’t knows’ may potentially lead to underestimating the extent of change and the extent to which there has been a hardening of both positive and negative attitudes towards official statistics. 

The summary tables accompanying the technical report provide results inclusive of ‘don’t knows’ for each question.