A salient feature of current theories of coalition formation is their parsimony-each theory specifies one guiding principle for predicting coalition formation. In their parsimony, however, these theories have paid little attention to the bargaining process leading to a given coalition and how these negotiations might change as a result of the nature and outcome of prior events. Accordingly, a theory that emphasizes the bargaining process is proposed; some critical comparisons are made between the predictions of the theory and the predictions of two other theories (minimum resource and pivotal power theories) ; and some data are presented that support the validity of the proposed theory.

The experiment was allegedly a simulated bargaining session between a buyer and a seller, who communicated via written offers. Actually, the E substituted standardized bargaining patterns for the offers of both opponents, and conducted simultaneous bargaining sessions with each of the Ss. In a factorial design, the type of initial offer, extreme or moderate, the frequency of concession, moderate or infrequent, and the S's role, buyer or seller, were varied. The major findings were: (a) Sellers did better than buyers, because sellers opened with a more extreme initial offer; (b) for bargaining success, an extreme opening offer, probably coupled with infrequent concessions, was best; (c) more frequent concessions elicited more frequent concessions, but not greater movement.

Journal of Management 28 (2002) 591-610. doi:10.1016/S0149-2063(02)00157-5Received by publisher: 2001-05-11Harvest Date: 2016-01-04 12:21:17DOI: 10.1016/S0149-2063(02)00157-5Page Range: 591-61

No abstract

A number of studies on coalition formation have focused, at least in part, on the triad where resources are divided A>B>C and A<(B+C), because the various coalition formation theories yield different predictions in such a case. According to Caplow's (1956) theory, either the AC or BC coalition will occur. The major assumptions of Caplow's theory are that people will prefer the coalition which enables them to control the maximum number of others and that whenever coalition preferences coincide, that coalition will form.In the A > B > C and A < (B+C) triad, B prefers C over A. With a BC coalition, B controls both C within the coalition, because B > C, and A outside the coalition, because (B + C) > A.In an AB coalition, B controls only C outside the alliance, since A's greater resources' give A control of B within the coalition. C controls the person outside the alliance and is controlled by 1 Caplow actually uses the term power rather than Gamson's term of resources. Although Caplow did not define power in his onginal theoretical paper, his use of the term seems to be similar in many respects to resources which Gamson (1961b, p. 374) defines as "a weight associated with each participant involved such that some critical quantity of these weights is necessary for the decision to be made." In this paper, resources will be used in reference to the weights of the participants, and power will be used solely to refer to the concept in pivotal power theory. the coalition partner regardless of whether C allies himself with A or B. Therefore C has an equal preference for A or B. A prefers either B or C, since either coalition enables A to retain his control over both of the others by prohibiting a union against himself. In summary, A prefers B or C, B prefers C, and C prefers A or B. Preferences between A and C and between B and C are reciprocated, while A's choice of B is not reciprocated. Hence AC or BC should occur, and AB should not. Chertkoff (1967) modified Caplow's theory to take into consideration the probabilities of the various partner choices. Since A's preference for B and C are equal, his probability of choosing either is 0.5. Likewise, C's likelihood of choosing A or B is 0.5, and the probabilities of B's choices are 1.0 for C and 0.0 for A. The likelihood of obtaining any of the three possible coalitions is determined by multiplying together the individual probabilities. The probability of the BC coalition is 1.0 x 0.5 = 0.5; the AC coalition, 0.5 x 0.5 = 0.25; and the AB coalition, 0.5 x 0.0 = 0.0. One-fourth of the time, no reciprocation of choices occurs. If when no reciprocation occurs, the triad members choose as before, the ratio of BC to AC to AB coalitions continues at 2:1:0, with the frequency of no coalitions approaching zero.Although there are various game theory approaches to analyzing n-person games, the rationale explained by Vinacke and Arkoff at UQ Library on June 15, 2015 jcr.sagepub.com Downloaded from