Conflict resolution under uncertainty

Abstract

Rubinstein's alternating offer bargaining model is extended to uncertain situations. When the players do not have complete information on the feasible payoff set, the bargaining is based on the players' own estimations on the Pareto frontier. It has been proved that there always exists a unique stationary fictitious subgame perfect equilibrium (SPE) if the estimates of the Pareto frontier are close to each other. Monotonicity and convergence properties of the stationary subgame perfect equilibria (SPEs) are next examined. It has been shown that the convergence of the disagreement payoff vector and the break-down probabilities implies the convergence of the SPEs as well. The controllability of the resulting dynamic systems is examined and it is shown that by selecting an appropriate disagreement payoff vector and a break-down probability, any desired outcome or maximize payoffs can be reached. The bargaining processes with time-varying Pareto frontiers are also analyzed. Four examples are provided to illustrate how to use the general model to design optimal negotiation strategy. The results of the dissertation provide schemes that can be applied to design and conduct future negotiations.