Signaling in the ‘Before’ Model of Final Offer Arbitration

Abstract

We develop a signaling model of final offer arbitration (FOA) in which the informed party makes the final settlement demand to the uninformed party. In FOA, each party submits a proposal to an arbitrator and if no agreement is reached, the arbitrator must select one of the two submitted proposals. We analyze a “before” model of FOA in which all settlement activity occurs prior to the exchange of proposals. We characterize the pure strategy separating, equilibrium associated with this game. Relative to a model of conventional arbitration (CA), the pure strategy separating equilibrium is associated with a higher dispute rate because weak player types have an enhanced incentive to bluff in this setting. This result continues to hold in semi-pooling equilibria. Thus our model is consistent with a higher dispute rate in FOA compared to CA. Results on the incentive to engage in costly voluntary disclosure are mixed across the two models. In contrast with CA, costly discovery can occur in this model of FOA. The analysis of out of equilibrium beliefs and actions is considerably more complex than in the standard litigation model.