STRATEGIC BARGAINING FOR THE CONTROL OF A DYNAMIC SYSTEM IN STATE-SPACE FORM

The partition of a pie model is integrated into a two-player differenc e game in state-space form with a finite horizon, in order to derive s trategic bargaining outcomes in the framework of difference games. It is assumed that agreements are binding. In contrast to the model for t he partition of a pie, the outcomes are not necessarily Pareto-efficie nt. For one-dimensional, linear-quadratic difference games, the subgam e perfect bargaining outcome is unique, Pareto-efficient, and analytic ally tractable. However, for higher dimensions the linear-quadratic st ructure breaks down and one has to resort to numerical methods.