Publication - Research and analysis

Scottish Social Attitudes Survey 2011: Core Module - Attitudes to Government, the Economy and Public Services in Scotland

Published: 24 Jun 2012

This report uses SSA data from 1999 onwards to explore changing attitudes to government, the economy and public services. It also discusses findings on who people think should be responsible for providing and paying for particular public services.

Scottish Social Attitudes Survey 2011: Core Module - Attitudes to Government, the Economy and Public Services in Scotland


1. Following five quarters of economic contraction, UK Gross Domestic Product (GDP) began to grow again in the fourth quarter of 2009 (ONS, 2010). Scotland exited recession at the same time as the UK, in Q4 2009 (Scottish Government National Statistics, April 2010).

2. GDP between 2010 and 2011 rose by 0.79% in the UK (inflation-adjusted figure) - Guardian 'UK GDP since 1948' accessed at

3. Survey data for the start of 2012 has been more positive than the end of 2011 and the labour market has also shown signs of improvement at the start of this year. However, it will not be known if Scotland has escaped a second quarter of negative growth and thus avoided a return to recession until data for the first quarter of 2012 is released on 18 July.

4. See

5. The sample size was reduced in 2011, from around 1,500.

6. Thus significance tests on differences reported in the text produced p-values of <=0.05. Cases where differences were on the margins of being statistically significant at this level (where p is >0.05 but <0.10) are identified in the text or in footnotes.

7. 87% of 2007 fieldwork was completed before the September name change.

8. The decline in trust in the UK government between 2010 and 2011 was marginally statistically significant (p = 0.077).

9. Trust to act in Scotland's best interest is measured on a 4-point frequency scale ranging from 'just about always' to 'never'. Trust to make fair decisions relates to the strength of trust, and the answers are on a 5-point scale ranging from 'a great deal' to 'not at all'.

10. The question wording altered slightly between years, as follows:
1999 "Will a Scottish Parliament…"
2000 "Do you think that having a Scottish parliament is going to…"
2001-2009 "Do you think that having a Scottish Parliament is giving…"

11. See Annex A, tables A.12 to A.15 for full figures.

12. Differences in response by newspaper readership on the other three questions were not significant.

13. Party political identification is identified through a series of questions. Respondents are asked if they support any particular party. Those who say no are asked if they feel closer to one party than another. Those who still answer no are asked which party they would support if there was a general election tomorrow.

14. Participants were asked to choose between 4 options: The UK Government should make all decisions; the Scottish Parliament should make all the decisions for Scotland; the UK government should make decisions about defence and foreign affairs, the Scottish Parliament should decide everything else; the UK government should make decisions about taxes, benefits and defence and foreign affairs, the Scottish Parliament should decide the rest; and the UK government should make all decisions for Scotland.

15. As a result, the gap in attitudes between tabloid and broadsheet readers on this measure was no longer significant in 2011. NB similar trends were also apparent in relation to believing the Scottish Government is good at listening and feeling the Scottish Parliament gives ordinary people more say (see Annex A, Tables A.12 and A.13).

16. For example, in 2011, 69% of Liberal Democrat identifiers said they trusted the Scottish Government 'just about always' or 'most of the time' - similar to the 72% of Liberal Democrats who said the same in 2010.

17. For full results for these questions, see Annex A, tables A.16 and A.17. Note that the net balance scores are based on the proportions saying standards had increased 'a little' or 'a lot', minus the proportion saying they had fallen 'a little' or 'a lot'. The proportions saying standards have increased/fallen 'a little' or 'a lot' were combined in PASW 18, to avoid rounding errors. As such, they may vary by a percentage point from the sum of the (rounded) individual figures.

18. NB net scores are calculated using the exact figures, rather than the rounded figures - hence they may differ by a percentage point from the value obtained by subtracting the rounded 'fallen'/'weaker' from the rounded 'increase'/'stronger' figures.

19. As noted in Chapter 1, the UK economy returned to recession in April 2012. However, fieldwork for SSA 2011 took place from June to October 2011, prior to this announcement.

20. This was accompanied by a fall in the proportion who felt that standards of living had stayed the same (down from 28% in 2010 to 19% in 2011). For full figures, see Annex A, Table A.17.

21. The Retail Prices Index includes mortgage interest payments.

22. Scotland's unemployment rate was 8% at the start of 2011 (down from 8.3% in the last quarter of 2010), then fell to 7.6% in quarter 2 of 2011 before increasing again to 7.9% and then 8.6% by the end of the year -
see for detailed figures.

23. Respondents were asked to say whether they were: living really comfortably on present income; living comfortably on present income; neither comfortable nor struggling; struggling on present income; and really struggling on present income. Note that the wording of the answer options were changed slightly in 2010. Prior to 2010, people were asked if they were living: very comfortably on present income; comfortably on present income; coping on present income; finding it difficult on present income or finding it very difficult on present income.

24. Given this finding, it is unsurprising that those aged 65 and older were the age group with the highest level of reported satisfaction with their general standard of living - 8.36, compared with 7.82 among those aged 18-29 and 7.53/7.51 among those aged 30-39 and 40-64.

25. It is worth noting, however, that this may in part be due to the relatively small sample sizes of some sub-groups meaning that a large increase in the mean score would be required to show a significant difference. For example because retired people make up c.30% of the total sample the increase of 0.39 between 2010 and 2011 was significant but as the unemployed only make up 7% of the sample the increase of 0.50 between 2010 and 2011 was not significant.

26. See for example Wilby 'This sentimentality over old people is hitting our young' in the Guardian - Wider issues around intergenerational equity were discussed in a recent book by David Willetts (2011) The pinch: how the baby boomers took their children's future - and why they should give it back, London: Atlantic

27. See Annex A, Tables A.20 to A.22, for full breakdown of findings for these questions.

28. This decrease was only marginally statistically significant (p=0.096).

29. This question was included in the Scottish Government core module for the first time in 2011, though it had been asked in SSA on a number of other occasions in the past (funded by the ESRC or self-funded by ScotCen Social Research).

30. Evidence from the British Social Attitudes survey shows that from 2004 to 2010 the proportion of people saying they were either 'very' or 'quite satisfied' with the NHS increased from 43% to 70%, before falling back to 58% in 2011. For further details see Clery, E (2012) 'NHS: Taking the pulse: attitudes to the health service' in Park et al (eds.) (2012) and Appleby, J (2012) Public satisfaction with the NHS and its services: Headline results from the British Social Attitudes Survey, NatCen Social Research.

31. There was no statistically significant change in the proportion who said the SG ought to have most influence in 2011 compared with 2010. The change in the proportion who said the UK government ought to have most influence over the same period was only marginally statistically significant (p = 0.084).

32. The increase between 2010 and 2011 was not significant, however.

33. In 2004 and 2006 a similar pattern was apparent - a rise on the previous year in the proportion believing local councils had most influence was accompanied by a reduction in the proportion of those saying it was the UK Government. The reasons for this fluctuating trend in the proportion attributing most influence to councils are unclear.

34. Logistic regression is a statistical technique that allows you to examine the relationship between a dependent variable (in this case, believing the Scottish Government has the most influence over how Scotland is run), and various independent variables (like party identification, strength of voice in the UK and who is responsible for standards in public services, etc.). The analysis identifies which of these independent variables are significantly and independently related to the dependent variable, after controlling for the inter-relationships between variables. For more detail and regression output, see Annex B.

35. As measured by a set of questions examining views on the distribution of resources in society, whether government should redistribute income and attitudes to 'big business'. Responses to each question in this set were combined to give an average 'score' on a left-right scale. Respondents were then divided into three groups - left, centre and right - based on their total scores on this scale.

36. Women were no more likely than men to feel that private companies and charities would provide the more cost effective services. Gender differences in viewing charities and private companies as more likely to provide the best quality services were only marginally statistically significant.

37. At the 95% level (p = 0.077).

38. This is because the confidence interval for a survey finding of around 50% is bigger than the confidence intervals for a finding of, say, 10% or 90%. Confidence intervals recognise the fact that, because findings are based on a survey of a sample of the population rather than on a census, there is always a margin of error attached to the results. Confidence intervals are typically expressed as a range - for example, a 95% confidence interval of 50% +/-5% indicates that we are confident that if we repeated the survey with different random samples from the same population, 95 times out of a 100 the finding would fall between 45% and 55%. The fact that confidence intervals are wider for survey findings around 50% than for bigger or smaller percentages reflects the fact that, where a view evenly divides a population, it is easier for a random sample of that population to include many more or less people who hold that view just by chance. In contrast, if most people in a population hold a particular view, it is less likely that a random sample will, by chance, include many fewer people who do so. Similarly, if few people hold a view, it is less likely that a random sample will, by chance, include many more people who hold this view.

39. Difference by gender were not significant (p = 0.113).

40. Changes in the proportions in favour or against charging for individual music lessons were not statistically significant.

41. Difference by gender were not significant (p = 0.113).

42. Like many national surveys of households or individuals, in order to attain the optimum balance between sample efficiency and fieldwork efficiency the sample was clustered. The first stage of sampling involved randomly selecting postcode sectors. The sample frame of postcode sectors was also stratified (by urban-rural, region and the percentage of people in non-manual occupations) to improve the match between the sample profile and that of the Scottish population. For further details of the sample design, see para 6 below.

43. See for details.


Email: Linzie Liddell