y, which "predicts that the way payoffs are divided, and sometimes even the coalitions that form, will tend to change from initial instances of coalition formation to later instances" (1980, p. 156). Miller's view is that in a game of bargaining, coalition behavior is mainly a function of bargaining strength and not of resources, which means mainly the anticipation of the payoffs available in alternate coalitions. Results show a tendency to focus on receiving more or less payoff when forming a coalition. In other words, as Miller says, "the larger the payoff per member for a coalition, the more likely the coalition is to form" (Miller, 1980, p. 163). In Miller's study, where there was virtually no differential in potential payoff, available resources were found to be a relatively more important variable. Where there was a differential in Miller's research laboratory, payoff was the more important variable. This leads Miller to make a claim for the value of minimum power theory in predicting behavior in interpersonal coalition conditions.
Komorita, Aquino, and Ellis (1989) have a different view of coalition formation in a test of the same three theories that Miller tested. Their conclusion, however, is that bargaining theory is, on the whole, a superior predictive tool where coalition formation is concerned. Part of the reason for this appears to be that the conditions of their experiment involved giving subjects the ability to affect the allocation of resources rather than simply to respond to resources already allocated to
...