D. De Cremer

Social identification effects in social dilemmas: a transformation of motives

Three experimental studies were conducted to examine two alternative explanations for the widely established positive eAect of social identification in promoting cooperation in social dilemmas. We hypothesised that social identification eAects could be either ascribed to (1) an increase in the value assigned to the collective good (i.e. goaltransformation hypothesis) or (2) an enhancement of trust in the cooperation of other group members (i.e. goal-amplification hypothesis). To disentangle these two explanations, we examined the eAects of social identification on the contributions to a public good of people with a diAerent social value orientation (i.e. pre-existing diAerences in preferred outcome distribution between self and others). Following the goal transformation hypothesis, we predicted that an increased group identification would raise contributions, in particular for people essentially concerned with their personal welfare (i.e. pro-self value orientation). Alternatively, following the goal amplification hypothesis it was expected that increased group identification would primarily aAect decisions of people concerned with the collective welfare (i.e. prosocial value orientation). The results of all three studies provided support for the goal-transformation rather than goal-amplification hypothesis, suggesting that ‘selfish’ individuals can be encouraged to cooperate by increasing the salience of their group membership. Copyright # 1999 John Wiley & Sons, Ltd.

Leadership in social dilemmas : The effects of group identification on collective actions to provide public goods

Two experimental studies investigated the role of group identification in the selection of and cooperation with leaders to manage public good dilemmas. The findings of the 1st study revealed that there was a general preference to select leaders with a legitimate power base (i.e., democratic, elected, and internal leaders), but these preferences were particularly pronounced when peoples identification with their group was high rather than low. The 2nd study complemented these findings by showing that when group identification was low, an instrumental leader (i.e., who punishes noncontributing members) was far more efficient than a relational leader (i.e., who builds positive intragroup relations) in raising contributions. Yet, when group identification was high, both leader types appeared to be equally efficient.

Effect of group identification on the use of attributions.

RESEARCHERS OF ATTRIBUTION (Weiner, 1985; Wong & Weiner, 1981) suggest that both failure and success activate emotional responses that, in turn, trigger an attributional process to determine the cause of the event. Investigators of interpersonal attributions for success and failure (Bradley, 1978; Zuckerman, 1979) have found that people have the tendency to attribute failure to external causes and success to themselves. As Heider suggested, “One is inclined to attribute to oneself good things but one suffers when one has to attribute to oneself something that is not so good” (1976, p. 16). Thus, people rarely stay within the bounds of given data (provided by the social environment); rather, they go beyond objective information and attach a subjective meaning to it (Ross & Nisbett, 1991). Drawing on the self-attribution literature just noted, I examined in the present research whether people also use the attribution bias to explain actions of fellow group members. Previous researchers (Hewstone, 1990; Weber, 1994) have reported that attributions about the behaviors of in-group members were often biased in favor of the in-group-that is, group success was attributed to the qualities of the ingroup, whereas group failure was attributed more to external factors. However, there may be individual differences in the ways that group members use groupserving attributions. According to social identity theory (Tajfel & Turner, 1986), group members with an especially strong sense of group identification evaluate other in-group members positively; accordingly, it can then also be suggested