Ethnic Diversity and Social Capital at the Community Level: Effects and Implications for Policymakers

Research on the impact of ethnic diversity on ‘social capital’ is relatively new in the field of political science (Stolle et al., 2008: 57). Reinvigorating a prominent and interesting debate among scholars, Robert Putnam awakened the subject in a 2007 paper entitled “E Pluribus Unum: Diversity and Community in the Twenty-first Century.” His main hypothesis was that ethnic diversity at the neighbourhood level, in the short term, leads to people ‘hunkering down,’ or being less social, trusting, and altruistic. This paper, however, challenges Putnam’s ‘hunkering down’ hypothesis.

Instead, it argues that ethnic diversity does not, with any certainty, erode social capital; rather, it has mixed effects that are contextual and dependent upon measurement methodologies and specific circumstances of historical as well as socio-economic factors. In this essay, a response by Sturgis et al. (2010) is used as a counterbalance and a criticism to Putnam’s paper, since the former uses a wider range of indicators to compare the effects of ethnic diversity. Although this paper largely contests Putnam’s proposition, it also credits his ideas in accordance with the research conducted by other scholars. Additionally, this essay examines the most recent findings relating to the distinction between diversity and segregation, and the importance of ‘contact.’ Most of this paper is devoted to analysing the impact of ethnic diversity on social capital, while the last section of the paper describes implications for policymakers. First, the definition and theoretical approaches to “social capital” are explored.

"Social Capital:" Definition and Approaches

Social capital has many forms and elements in the scholarly literature, and as many definitions. This essay, however, uses Putnam’s succint definition, wherein social capital “refers to connections among individuals – social networks and the norms of reciprocity and trustworthiness that arise from them” (Putnam, 2000: 19). By implication of this definition, social networks, and thus, social capital, similarly to physical and human capital, benefit both members of these networks, and bystanders (Putnam, 2000: 20; 2007: 137-138). There have been various approaches to study social capital, given how difficult it is to quantify and measure. ‘Contact’ and ‘conflict’ theories were used commonly by an earlier generation of scholars to deal with ethnic diversity and its implications. Contact theory suggests that more diversity causes more inter-ethnic tolerance and solidarity, whereas conflict theory ‘fosters out-group distrust and in-group solidarity’ (Lancee & Dronkers, 2008: 79). It is also useful to distinguish ‘bonding’ and ‘bridging’ social capital. Simply put, bonding social capital ‘ties to people who are like you in some important way’, whereas bridging – ‘to people who are unlike you’ (Putnam, 2000; 2007: 143).

Further, Letki (2008: 106) suggests that ‘neighbourhood social capital consists of two major components: attitudes towards fellow neighbours and interactions with them.’ These components help to separate the cognitive and structural sides of social capital, making it more convenient to measure it using determinate variables, such as membership (and participation) in formal and informal organisations (associations, networks), social trust, and civic participation, among others. However, trust, and the many interpretations of it that scholars use, is the core focus of this paper.

Putnam and the "Hunkering Down" Hypothesis

To reiterate, Putnam argues that ethnic diversity, at least in the short term, has a negative effect on social capital at the community level (Putnam, 2007). He addresses the contact and conflict theories, which were more prominent among scholars of an earlier generation, which he then conflates to a ‘constrict’ theory, proposing that both in-group and out-group solidarity may be negatively affected by diversity (Putnam, 2007: 144). More importantly, although the research he conducted was in the context of the US, his arguments are compelling and suggest the applicability of the ‘hunkering down’ hypothesis cross-nationally.

However, the trajectory of later research on this topic has produced mixed results. On the one hand, some scholars have, indeed, confirmed Putnam’s arguments. Shaeffer (2013), for example, finds in his study of 55 German cities that ethnic diversity does erode social capital. Stolle (2008) and Stolle & Harell (2012) also confirm the negative effects of diversity on white majority trust and friendship networks in neighbourhoods in the US and Canada, whereas Lancee & Dronkers (2008) reach a similar conclusion in their study of the Netherlands. There are a number of other scholars confirming Putnam’s hypothesis or, at least, parts of it. Therefore, it might seem that ethnic diversity does have an erosive influence on social capital.

On the other hand, however, as later proposed by a multiplicity of scholars, Putnam’s work is, at the same time, limited and discounts a variety of factors in his analysis of the effects of ethnic diversity on community level social capital. Portes & Vickstrom (2011: 463) maintain that Putnam finds a strong negative relationship between ethnic diversity and social capital, whereas ‘many studies find a relationship that is weak and contingent on various individual and contextual factors’. Indeed, further research by scholars indicates a battery of mixed results of the ‘hunkering down’ hypothesis, especially when tested in different countries and using adjusted or more sophisticated methodologies. The research conducted by Sturgis et al. (2010) contradicts and criticises Putnam’s hypothesis and methods on several aspects, two of which are discussed here.

Two Limitations of Putnam's Argument

First, his argument does not hold true in most other countries except for the US and Canada, and produces spurious evidence in, particularly, the European context. Indeed, according to Sturgis et al. (2010: 59), ‘the nature and implications of ethnic diversity is highly historically contingent, and we cannot assume, without evidence, that associations observed in one context generalize in a straightforward manner to others.’ This criticism is verified by a variety of studies. For example, Hooghe et al. (2008: 218) find in their cross-national study of EU-15 countries that ‘even paying attention to a variety of diversity measures and methodological safeguards, the fullblown negative relationship between ethnic diversity and generalized trust does not hold’. Gesthuizen et al. (2008: 135), too, determine ‘no evidence at all for what we consider to be Putnam’s core claim on the relationship between ethnic diversity and dimensions of social capital: these relationships turned out to be spurious in Europe.’

Interestingly, in Australia, Leigh (2006) finds a stronger relationship between ethno-linguistic diversity and trust, rather than specifically ethnic diversity, which suggests that social issues not always come about due to ethnic reasons per se. However, the study by Letki (2008: 120), for example, still reveals a small degree of empirical confirmation to confirm Putnam’s argument in Britain. From this evidence and studies by other scholars it is not entirely clear whether Putnam’s thesis is applicable outside of North American context. More importantly, ethnic diversity does not seem to have a significant negative community level effect on social capital in terms of trust, yet one cannot claim that there is no effect either.

Second, Putnam and many other scholars do not take into account structural and compositional differences between neighbourhoods (Sturgis et al., 2010: 78). More also, this shortcoming is amplified by the fact that ‘he erroneously did not acknowledge the multilevel structure of the hypotheses’, which would have allowed for more accurate results (Gesthuizen et al., 2008: 128). Certainly, both of these create space for error in measurement and interpretation of results, therefore undermining Putnam’s conclusions, which are, indeed, contradicted by a number of scholars. Letki (2008: 101) in her study of four dimensions of social capital, finds that ‘when the effects of diversity and neighbourhood deprivation on social capital are modelled simultaneously, and the relationship between neighbourhood status and racial diversity is accounted for, diversity has a negative effect on only one – attitudinal – dimension of social capital’.

In addition, once they controlled for neighbourhood differences, Sturgis et al. (2010: 78) found the association between ethnic diversity and trust explaining for just an insignificant 1% of neighbourhood-level fluctuation. Fieldhouse & Cutts (2010: 308) also reach a conclusion that ‘the differential effect of diversity on neighbourhood norms is less apparent once other factors—particularly neighbourhood poverty—are taken into account’, although they find a slight negative impact of diversity on norms. Therefore, clearly, ethnic diversity is not a factor that can work independently, without accounting for specific contextual variables. This is certainly a disadvantage of Putnam’s study, which, together with his single-level neighbourhood analysis, amounts to flawed unrealistic outcomes of research. Nevertheless, these and other existing studies find mixed effects of ethnic diversity, be they positive, negative, or insignificant, which yet precludes us from claiming that diversity does not impact social capital.Continued on Next Page »