Using an intergroup version of the trolley problem, we explored participants' willingness to sacrifice their lives for their group. In Study 1, Spaniards whose personal identities were fused with their group identity endorsed saving fellow Spaniards by jumping to their deaths in front of a runaway trolley. Studies 2 and 3 showed that the self-sacrificial behaviors of fused Spaniards generalized to saving members of an extended in-group (Europeans) but not members of an out-group (Americans). In Study 4, fused participants endorsed pushing aside a fellow Spaniard who was poised to jump to his death and initiate a chain of events that would lead to the deaths of several terrorists, so that they could commit this act themselves. In all four studies, nonfused participants expressed reluctance to sacrifice themselves, and identification with the group predicted nothing. The nature of identity fusion and its relationship to related constructs are discussed.

When people are ostrasized (i.e., rejected and excluded) by either an outgroup or an ingroup, they may either withdraw or engage in compensatory activities designed to reaffirm their social identity as a group member. The authors proposed here that individual differences in identity fusion (an index of familial orientation toward the group) would moderate the tendency for people to display such compensatory activity. Consistent with this reasoning, the results of four experiments showed that irrevocable ostracism increased endorsement of extreme, pro-group actions (fighting and dying for the ingroup) among fused persons but not among nonfused persons. This effect emerged when an outgroup ostracized fused individuals due either to their nationality (Experiment 1) or their personal preferences (Experiment 2). Similarly, ostracism by the ingroup amplified the tendency for fused persons to both endorse extreme pro-group actions, refuse to leave the group (Experiment 3), and donate money to an ingroup member (Experiment 4). Finally, compensatory activities emerged even when ostracism was based on being "too good" for the group, suggesting that a desire for self-enhancement does not mediate such activities (Experiment 4).

No abstract

is Lecturer in the Department of Marketing, University of Strathclyde. He is particularly interested in using multivariate statistical tools for interpretation and model-building. He is presently involved in research into internationalisation, gift buying and the effects of doctors' waiting rooms on patients.