Essays in behavioral game theory.

Summary The emergence of behavioral economics tends to show that game theory is based on principles that are challenged both empirically and theoretically. Behavioral economics models have been developed in profusion to account for such deviations, but in an incomplete way: they are often contradictory to each other and they imply a weakening of the formal elegance and unification of the theory (chapter 2). A key question is therefore whether it is possible to grasp these patterns within a unified framework. A scheme that generalizes economic behavior beyond that which can be explained by the preference optimization model is proposed here. It is shown that if these behaviors are determined by consistent decision processes, then there is an equilibrium for all strategic interactions (chapter 3). Many types of behavior satisfy this consistency condition. An equilibrium can thus exist with highly heterogeneous agents - different preferences, different levels of rationality, different "behavioral motives" (chapter 4). It is shown that such a behavioral equilibrium could be the resting point of a general class of learning processes (chapter 5). A frequency extension of the behavioral equilibrium is proposed, to capture non-convex behaviors (chapter 6). The behavioral equilibrium is then extended to abstract economies, and a general equilibrium model is proposed, in particular by considering that agents are endowed with decision process functionals (chapter 7).