Abstract

Summary A collective action (CA), i.e., a group of individuals jointly producing a resource to be shared equally among themselves, is a common interaction in organizational contexts. Ancestral humans who were predisposed to cooperate in CAs would have risked being disadvantaged compared to free riders, but could have overcome this disadvantage through ‘greenbeard ’ reciprocity, that is, by assessing the extent to which co-interactants were also predisposed towards cooperation, and then cooperating to the extent that they expected co-interactants to reciprocate. Assessment of others ’ cooperativeness could have been based on the direct monitoring of, and on reputational information about, others ’ cooperativeness. This theory predicts that (1) CA participants should monitor accurately, and (2) perceived higher-cooperators should have better reputations. These predictions were supported in a study of real-life CAs carried out by a group of Shuar hunter-horticulturalists: (1) members accurately distinguished ‘intentional’ non-cooperators (who could have cooperated but chose not to) from ‘accidental ’ noncooperators (who were unable to cooperate), and their perceptions of co-member cooperativeness accurately reflected more objective measures of this cooperativeness; and (2) perceived intentional cooperators had better reputations than perceived intentional non-cooperators.

Citations