The psychology of cosmopolitan behaviour: emotions, norms and social identification

Ethical cosmopolitanism has been the subject of substantial theoretical elaboration over its long history. However, until recently, very little attention had been given to the question of how individuals might be encouraged to behave as cosmopolitans in practice. Political theorists have recently identified a small number of factors – including certain social identities, collective guilt, and prosocial norms – that may increase cosmopolitan behaviour, but whether those factors actually do increase cosmopolitan behaviour has not been empirically validated. Additionally, cognate work in the field of social psychology that may provide both new hypotheses about the motivators of cosmopolitan behaviour, and tentative tests of their empirical validity, has not been integrated into cosmopolitan theorists’ work. Thus, drawing from political theory and social psychology, this dissertation identified a series of factors that may increase cosmopolitan behaviour, and tested their efficacy in a series of six experimental studies (total N = 1562). To examine causal effects on cosmopolitan behaviour, the experiments manipulated: human-, dual- (i.e. national and human) and overlapping-identity salience (Study 1); collective guilt (Studies 2, 3 and 5); empathy (Studies 4 and 5); ingroup-directed anger (Study 5); and, national prosocial norms (Study 6). Results indicated that empathy (Study 4) and, for strong national identifiers, prosocial national norms (Study 6) significantly increased cosmopolitan behaviour. Contrary to existing theory and research, manipulations of collective guilt and ingroup-directed anger did not increase cosmopolitan behaviour (Studies 2, 3 and 5) because such manipulations increased not only collective guilt and ingroup-directed anger but also increased dehumanisation of outgroups (Study 3), reduced outgroup empathy (Study 5), and marginally reduced identification with all humanity (Study 5). The effects of identity salience on cosmopolitan behaviour were not clear because the identity salience manipulations were unsuccessful. Overall, the studies reported in this dissertation represent the first experimental evidence on the causes of cosmopolitan behaviour, and demonstrate that experimental methods can be fruitfully used to develop and test theories about individual-level cosmopolitan behaviour.