Bargaining Processes and Coalition Outcomes

This paper discusses a theory of bargaining processes in three-person coalition games and investigates the relation between these processes and the outcome of bargaining. The bargaining theory uses an information-processing approach to model each of the three player's cognitive processes and interrelates three such models to predict who says what to whom next in the bargaining for coalition partners and payoff allocations. The theory succeeds in predicting the act-by-act communications observed in three laboratory studies of sequential games of status in which each player seeks to maximize the rank of his total score in relation to the other players' total scores. In addition, the process theory succeeds in predicting outcomes. It was used in computer studies to simulate bargaining processes to their resulting outcomes, thus producing artificial data on which coalitions form and what payoff allocations are made to coalition members. These artificial data on outcomes agree with the pattern of outcomes observed in the laboratory studies and also conform to Laing and Morrison's (1973) heuristic model predicting coalitions and payoffs in three-person sequential games of status. These investigations lead to an interesting discovery about the effects of variations in the process on the pattern of resulting outcomes. The duration of the bargaining process has systematic effects on which coalitions form and how payoffs are allocated to coalition members. The results suggest that more theoretical attention to social processes can pay important dividends, even when the primary objective is to predict outcomes.