Should we be worried about social capital decline?


Fifteen years ago Robert Putnam, in his book Bowling Alone, drew our attention to social capital in a big way by convincingly demonstrating its decline in the USA. But just what is social capital and is it also in decline in the UK? And if so, should we be worried about it?

Bowling Alone covers a great many different measures under the umbrella term ‘social capital’, but overall it embodies what we might call the civic tradition. Here, the definition of social capital can be thought of as strongly linked to civic virtue. It is theorised that the social trust and reciprocity that arise from our social interactions help our society to run smoothly. Evidence supports the idea that high-trusting societies are probably nicer places to live, at least in some respects. (But we ought to exercise caution with our assumptions about the direction of causality – living in a place with a nice low crime rate will also foster trust).

Social capital, in the vast body of research and policy attention following Bowling Alone, has often been taken to be synonymous with pro-sociality reflecting a position in which (moral) value is attached to helping, getting involved with the community, being trusting and so on. All of these things are worthy of pursuing and measuring in their own right. However, the academic literature is more complex. The use of the term social capital has developed from more than one origin, with quite different roots and applications. The French sociologist, Bourdieu, for example, used the term to explain how class advantage is passed along from one generation to the next. The familiar saying “it’s not what you know but who you know” is a way to capture part of what Bourdieu was talking about: those with well-connected professional parents (as an example) may well get a ‘leg up’ into a good school or a first job. Studies in the 1970s and 80s went on to show that this is more than a theoretical advantage: ‘weak ties’ (i.e. having acquaintances) provide a real and measurable way to improve one’s position in the labour market, and the odds improve if those acquaintances are from higher socio-economic classes. Another seminal study showed how parents’ social capital can influence educational outcomes for their children.

Figure 1: Social Capital is unevenly distributed by ethnicity and education Source: UKHLS 2012 (error bars show 95% confidence intervals)

Figure 1: Social Capital is unevenly distributed by ethnicity and education
Source: UKHLS 2012 (error bars show 95% confidence intervals)

In this conceptualisation, social capital is not about community at all – it does not (and cannot) increase the number of desirable jobs or university places available but rather it divides the well-connected from the not well-connected and is thereby a means by which high status groups can stay ahead of the rest. (“Social capital can divide as well as unify” as one key article puts it). This tradition of social capital research can be labelled the instrumental tradition, reflecting the key point that the acquisition of social ties may not be an end in itself but is a means to an end – a way to access resources.

One of the ‘classic’ indicators of social capital is to take a count of associations with voluntary organisations, typically including Parent-Teacher Associations, community groups, Trade Unions and so forth. While there is probably a good argument to consider this a civic type of social capital, in our latest research report we considered its instrumental value to be of particular importance. This is because we know from previous research that activity in these organisations increases network diversity, allowing greater reach into society bringing opportunities for social mobility.

And why does this conceptual distinction matter so much? Following the civic tradition might suggest that the key barometer of concern is how much social capital there is in society, in total. On the other hand, thinking of the instrumental value draws our attention towards the way it is distributed between groups in society. If some economically disadvantaged groups are lacking in social capital, how will they ever catch up? Particularly if the already-advantaged might be getting a leg up in the labour market.

In our study we looked at the level of social capital by ethnicity and education and we found large disparities – see Fig 1. Educational level matters for all four of our measures, but the gradient for voluntary organisations is particularly steep. Further, we examined how this distribution is changing over time – see Fig 2. Our statistical model showed that those from Asian, black and ‘other’ ethnic groups are dropping out faster than whites. Similarly, those with qualifications below degree level are dropping out faster.

Figure 2: The gap between the participation in voluntary organisations by educational qualification (right) and by ethnic background (left) is growing Source: Community Life Survey – predicted values from binary logistic regression models with age, sex, and ethnicity controlled.

Figure 2: The gap between the participation in voluntary organisations by educational qualification (right) and by ethnic background (left) is growing Source: Community Life Survey – predicted values from binary logistic regression models with age, sex, and ethnicity controlled.

So, should we be worried about the decline? We think that our evidence for polarization is a cause for concern as it suggests that existing economic inequalities by ethnicity and education may become more entrenched over time. Even among types of social capital that are not in decline, such as informal social support, we find evidence of polarization. In the case of social support ethnic minorities appear to be catching up with white groups but our data again suggest growing disparities for those with low educational attainment, particularly for men – see Fig 3 for some examples based on our statistical model. Whether this suggests that white uneducated men ought to be considered a ‘vulnerable group’ in society remains an open question that demands further research (but recent research in the USA suggests we ought to take the question seriously).

Figure 3: Our statistical model shows that social support is increasing for some groups but decreasing at a very slow rate for white men with no qualifications Source: Health Survey for England –predicted values from a linear regression model with demographic controls.

Figure 3: Our statistical model shows that social support is increasing for some groups but decreasing at a very slow rate for white men with no qualifications Source: Health Survey for England –predicted values from a linear regression model with demographic controls.

It’s not all bad news. In our new report on social capital, which coincides with an event on The state of Social Capital in Britain, we suggest that civic forms of social capital (i.e. social trust) seem resilient in the UK. You can read or download the new report here: CSI 15: The uneven distribution and decline of social capital in Britain

 

 

 

This blog post was written by Lindsay Richards, based on (currently unpublished) research with Anthony Heath. Comments to Lindsay.richards@nuffield.ox.ac.uk are welcome. Publication: 10th November 2015