The Scottish Government

« Previous | Contents | Next »

Listen

4. Confidence in government statistics

4.1 As discussed in the introduction to this report, concern about levels of public trust in official statistics has been a key influence on recent and planned changes in the way government statistics are published and assessed, including the establishment of the UK Statistics Authority. The new authority will assess all current National Statistics against a new 'Code of Practice for Statistics', and is accountable to the Scottish Parliament for its work with regards to Scottish devolved statistics.

4.2 The 2007 Scottish Social Attitudes survey included two new questions about trust in government statistics in order to provide baseline data on public trust in official statistics in Scotland. The first asked people how confident they were that official statistics published by the Scottish Executive are accurate on a scale of 0 to 10, where 0 indicates not confident at all and 10 indicates very confident. The second asked people their main reason for indicating that level of confidence. Interviewers were given a code frame and guidance about how to code responses, drawing on earlier work by the Office for National Statistics ( ONS).

Overall confidence in Scottish Executive statistics

4.3 Overall, the average 'confidence' score with respect to the accuracy of Scottish Executive statistics was 5.43, reflecting the fact that most people fell between the extremes of complete confidence (10) and complete lack of confidence (0). However, there was considerable variation around this average, with 15% giving a 'low' score of 0 to 3, 45% giving a medium score of 4 to 6 and 30% giving a score of 7 to 10, indicating a relatively high level of confidence (see Annex A, Table A.27). Nine per cent were not able to say how confident they felt about the accuracy of official statistics published by the Scottish Executive.

4.4 In their most recent research on trust in government statistics (Lader, 2008), ONS found that levels of trust in whether figures give a true picture of what is happening varied depending on the specific statistics being asked about, from an average of 4.89 (on a scale of 0 to 10) for hospital waiting figures to 7.27 for road casualty statistics. Our overall measure is not directly comparable with these questions, but the findings suggest that levels of confidence in Scottish Executive statistics fall within a similar range to levels of trust in various UK official statistics.

Who has confidence in Scottish Executive statistics?

4.5 Factors associated with relatively higher levels of confidence in Scottish Executive statistics (see Annex A, Table A.28) include:

• trust in other people generally - 42% of those who said most people can be trusted compared with 22% of those who said you can't be too careful dealing with people gave a high score of 7-10 for their level of confidence in Scottish Executive statistics.
• political trust (49% of those who 'just about always' trust the Scottish Executive to act in Scotland's best interests give a confidence score of 7-10, compared with just 12% of those who say they almost never trust the Executive)
• socio-economic class (38% of employers, managers and professionals and 35% of those in lower supervisory and technical occupations give a 'high confidence' score, compared with 23% of those in routine or semi-routine jobs).

These findings reflect analysis of the factors associated with trust in the general accuracy of official figures conducted by ONS (Goddard, 2005 and Kelly, 2005), which also found that socio-economic class, trust in people generally and trust in government institutions were associated with higher levels of trust in statistics. ¹⁶ This analysis lends some support to the notion of a 'virtuous circle' of positive associations between generalised social trust and political trust.

Reasons for confidence or lack of confidence in Scottish Executive statistics

4.6 The most common reasons given for expressing very low levels of confidence in the accuracy of Scottish Executive statistics were personal experience (27% of those who with low scores of 0-3), the belief that official statistics are misrepresented or 'spun' by politicians (27%), and the belief that statistics do not tell the whole story (14% - see Table 1). The notion that politicians misrepresent statistics was also mentioned by 30% of those who gave a 'medium' score (4-6) as a reason they did not have more confidence. As reported in Kelly (2005) and Lader (2008), the notion that the government and media manipulate and misrepresent official statistics is a common finding from across various ONS studies on public trust in statistics. These studies tend to suggest that the public has more confidence in the quality of official statistics than in their delivery, with the potential for political manipulation a key factor associated with distrust in delivery.

4.7 As well as explaining negative views of official statistics, personal experience was a common factor explaining higher levels of confidence in statistics (mentioned by 26% of those who gave a score of 7-10). One in ten of those giving high scores mentioned the belief that the Scottish Executive does not have a vested interest in the results as a reason. Goddard (2005) notes that qualitative work on trust in official statistics has shown that many people's views about official statistics are informed by their own experiences, regardless of the national picture. For example, someone who has recently lost their job might be unwilling to believe national figures which show unemployment falling.

Table 1 MAIN reason for level of confidence that Scottish Executive statistics are accurate, by level of confidence (col %)

« Previous | Contents | Next »