Exploring Dimensions of Social Capital in Scotland Findings from the Scottish Social Attitudes Survey and Scottish Household Survey

2 Analytical approach

2.1 The analysis in this paper is based on data from the Scottish Social Attitudes (SSA) survey (2009) and the Scottish Household Survey (SHS) (2010). The Scottish Social Attitudes survey is an annual survey exploring social and political attitudes. In 2009, it included a number of measures of social capital for modules funded by the Scottish Government and as part of its unfunded and background sections. The Scottish Household Survey is a continuous survey funded by the Scottish Government measuring the characteristics of households and individuals in Scotland. It includes a number of measures of social capital each year (2010 was the most recent year for which data was available at the time of writing).

2.2 The ten measures included in the analysis were selected in the light of a framework for measuring social capital in the UK, developed by ONS (Harper, 2002). This framework comprises the five different dimensions of social capital set out in table 1. In developing these dimensions, Harper (2002) takes a deliberately broad view of social capital, including aspects that may be important correlates of social capital (like views of the local area) as well as those (like social networks) that are more obviously a source or component of social capital.

2.3 Although this paper focuses primarily on data from SSA 2009, a smaller number of questions from SHS 2010 were also included to ensure that there was a suitable range of measures relating to all five dimensions. These questions are not the only ways in which these dimensions could be measured - it is of course possible that had we included a different set of questions, we would have found different results. The measures included are intended to cover at least one aspect of each dimension, but are not comprehensive. For example, questions around communities focus on geographical communities, rather than communities of interest, while questions about views of the local area focus on perceptions of anti-social behaviour.⁶ Again, it is important to note that if other questions had been included - for example, focusing on more positive aspects of people's neighbourhoods - the results may have differed from those presented here.

Table 1 - Summary of questions included in analyses

1- List of possible answers which respondents selected all that applied were: No, have not done any of these, Contacted an MP or MSP; Contacted a government department directly; Contacted my local council; Responded to a consultation document; Attended a public meeting; Contacted radio, TV or a newspaper; Signed a petition (including online petitions); Raised the issue in an organisation I already belong to; Gone on a protest or demonstration; Attended an event organised as part of a consultation exercise; Spoken to an influential person; Formed a group of like-minded people; Joined an existing organisation; Actively took part in a campaign (e.g. leafleting, stuffing envelopes etc); Given money to a campaign or organisation.

2.4 The analysis conducted for this paper used logistic regression to explore which of a range of key demographic factors are independently associated with various aspects of social capital. Logistic regression is a statistical technique that allows you to examine the relationship between a dependent variable (in this case, various measures of social capital), and various independent variables (like gender, age, income, 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.

2.5 Logistic regression models were created for each of ten measures based on questions from SSA 2009 and SHS 2010 that explored different elements of social capital. Each of these regression models looked at the relationship between the specific aspect of social capital - for example whether people feel they have people in their area to turn to for advice and support - and various demographic, socio-economic and area-based variables that might be associated with different levels of social capital. The socio-demographic and area-based variables included in the analyses were:

• Gender
• Age⁷
• Education
• Household income
• Socio-economic class (as measured by the National Statistics Socio-Economic Classification⁸)
• Marital status
• Tenure (owner-occupier, private renter, or social renter)
• Whether or not the respondent has a disability
• Whether or not there are school-aged children in the household
• Urban-rural (based on the Scottish Government urban-rural classification⁹)
• Deprivation (as measured by the Scottish Index of Multiple Deprivation¹⁰).

2.6 These variables were chosen partly based on evidence from previous work on social capital, and partly on the basis of hypotheses about additional factors that might affect social capital. For example previous analysis of SSA data has shown variations in levels of 'community-connectedness' by gender, age, income and area deprivation (Anderson and Dobbie, 2008), while analysis of Understanding Society data (a longitudinal study of households in the UK) shows that people with higher incomes tend to score more highly than those with lower incomes on various measures of social participation¹¹ (Ferragina et al, 2011). Research by Li et al (2005) on three dimensions of social capital (neighbourhood attachment, social networks and civic participation) suggests that class, education, income, gender, age and social deprivation are all significantly associated with differing levels of social capital.

2.7 Meanwhile, other research has suggested that being in a relationship is positively associated with higher levels of wellbeing,¹² which might in part be explained by the impact of relationships on widening (or deepening) people's networks and social capital. Similarly, having children may broaden people's social networks and may therefore have an impact on levels of social capital. Being disabled could impact on people's social capital in different ways. For example, barriers relating to the accessibility of public spaces and buildings might have a negative impact on people's social networks. On the other hand, experiences of such barriers might encourage disabled people to get more involved in trying to change and influence their communities. Finally, given the diverse geography of Scotland, the Scottish Government had a particular interest in exploring whether or not living in an urban or rural area is associated with higher or lower levels of social capital.

2.8 Differences in social capital by ethnicity and religion could not be robustly explored due to small sample sizes for key sub-groups. Data on sexual orientation was not available in the data used for this study.

Limits of the analysis

2.9 Logistic regression is a useful technique for exploring the relationship between multiple 'independent' variables and a given outcome. However, it can only tell us if a statistically significant and independent relationship exists between each variable and the outcome. It cannot tell us whether or not these 'independent' variables cause this outcome - such conclusions can only be made on the basis of experimental research. In the context of this paper, this means that we cannot conclusively say that, for example, having a higher level of education leads directly to higher levels of social capital. All we can say is whether or not a relationship between the two appears to exist. Assumptions about the direction and nature of this relationship rely on hypotheses on the part of those interpreting the data.¹³

2.10 Other limits relate to the possibility of drilling down to very specific sub-groups. As noted above, differences by ethnicity or religion could not be explored due to the small numbers of people falling into particular sub-groups. It is also worth noting that while this report looks at variations by deprivation and by urban-rural, it does not compare patterns of social capital in deprived urban areas with those in deprived rural areas. Given that rural deprivation is often relatively dispersed, there may be differences in the relationship between deprivation and social capital between rural and urban areas. However, this analysis is beyond the scope of this report.