Paradoxes of bridging and bonding: explaining attitudes of generalized trust for participants of mixed ethnically and Turkish voluntary organizations in Amsterdam.

Abstract

Recently in the Netherlands and in Amsterdam, policymakers have started to see generalized trust as an indicator of societal cohesion, which is taken to be endangered by participating in ethnically homogenous or bonding organizations. However, there is no study that supports this negative socialization effect. Existing surveys in the Netherlands and in Amsterdam lack either appropriate data on ethnic minorities or do not allow this question to be properly addressed. They do not contain the relevant variables or do not have a multilevel structure, since the latter requires one to sample many responses from the same organization rather than collect data that is representative of individuals. This thesis addresses this gap in the literature by juxtaposing Turkish (bonding) associations with organizations whose membership consists of different ethnic groups (bridging). I surveyed 40 non-profit organizations in Amsterdam and collected responses from around 450 participants. I subsequently describe different bonding and bridging practices within and between organizations, and demonstrate that Turkish, as compared to mixed organizations, are internally focused on their own group, but externally are more involved in bridging networks. Thus contact within Turkish organizations is confined to fellow ethnics and this allows for testing the contact hypothesis. This thesis employs a multilevel model and distinguishes individual attributes from organizational factors (ethnic composition). However, the variance in generalized trust at the organizational level is only 4%, which indicates that the context of voluntary organizations has not much influence on it. Secondly, I test an interaction effect between the mixed ethnic composition of an organization and the length of participation in years in order to test for a socialization effect (the contact hypothesis). However, this interaction effect is not statistically significant. Finally, I test for another interaction effect, namely the effect of having a close tie in a mixed organization, in order to test for a sufficient but not necessary condition of the contact hypothesis, which might turn contact into attitude change. Again, this interaction is not statistically significant. Beyond bridging and bonding, there are complementary mechanisms which might have affected generalized trust. I, therefore, extend my model to include cognitive evaluations about one’s humanitarian values, negative life experiences and socio-economic factors. Three theoretical frameworks are tested: psychological; norm driven; and social success. The findings suggest that differences in generalized trust are best explained by individual processes rather than contact between ethnically diverse groups in voluntary organizations. Optimism has the strongest effect size on generalized trust. Other key factors are educational levels, and to some extent older age as well as having been widowed or lost one’s partner due to divorce. Younger people who adhere to humanitarian values are also among the high generalized trusters. The effect of education, age and the experience of divorce or separation is also found in other Dutch representative national samples and support the consensus around social success theories in explaining generalized trust.