ABSTRACT
Since organizational variance in generalized trust attitudes is so low (4%), it is highly unlikely that contact with ethnically diverse others in voluntary organizations socializes one to become more trusting of strangers. However, most research assumes that voluntary organizations socialize their participants into trusting people, but fails to adequately address this question since the structure of their data does not allow for the distinguishing between participants’ and organizational variance in generalized trust. Most studies fail to sample organizations and their participants; instead they rely on surveys that sample individual households. A limitation of the present study might be the design, which only included Turkish
organizations. However, it might be unlikely to find evidence for this mechanism among other voluntary organizations, if we regard Turkish and mixed organizations as critical cases. Insignificant variances at the organizational level might also seem to weaken the findings. Nevertheless, we might also question whether confirming the statistical significance of the results justifies the collection of more data. Finally, the results seem to suggest that high-trusting individuals are selfselecting into mixed organizations. Stolle’s (1998) study reached a similar conclusion when examining generalized trust for participants of different types of voluntary organizations in the USA, Germany and Sweden (see also Uslaner and Brown 2005). Ultimately, however, we need longitudinal panel data, which should show whether careful individuals are not opting out of mixed organizations. As argued before, the data suggest that participants of mixed organizations seem to
have self-selected themselves into those organizations on the basis of their higher trust levels. Controlling for length of participation, participants of mixed organizations are not more trusting. This implies that the context in which interaction takes place, namely the presence of diverse ethnic groups or conversely ethnic homogeneity in voluntary organizations, is unlikely to affect generalized trust. Thus a contact mechanism, through bridging and bonding ethnic ties, in voluntary organizations seems not to explain differences in generalized trust. This is not so surprising, since the voluntary sector literature suffers from the same shortcomings as the earlier versions of contact theory. That is to say, most of these studies only predict that ‘positive contact effects will occur, not how and why’ (Pettigrew 1998, 80).
