Comparing Arbitration Procedures When Participants Are Risk Averse

We examine a game theory model of arbitration that generalizes previous models of both Final Offer Arbitration and Conventional Arbitration, and thereby enables a direct comparison of the two procedures. The model assumes there is a single quantifiable issue in dispute between two risk averse disputants. We analyze the relative range between the disputants' equilibrium offers and find that final offer arbitration is generally the superior procedure at encouraging compromise, if existing models of conventional arbitration reflect what actually occurs in practice; an alternative conclusion however is that the existing models are incomplete at describing arbitration practice.