On some models of altruistic behavior

The paper surveys several recent approaches to studying, modeling, and interpreting altruistic behavior. In the first section we describe a modification of the system of axioms of classical rational choice theory taking into consideration moral reference points. We study the moral monotonicity axiom, compare the predictions of the standard and the modified system of axioms, and describe series of experiments contradicting the standard system, but agreeing with the modified one. In the second section we discuss the approach of J. Roemer who suggested to consider Kantian optimization in place of the standard Nash one. It is well known that Nash equilibria are not necessarily optimal. It happens, as, for example, in the tragedy of the commons, that all the participants would benefit from choosing a different, non-Nash equilibrium, and in this section we consider mechanisms allowing to realize an optimal choice. In the third section we consider a model in which different parenting styles (based on altruism and paternalism) emerge as equilibrium outcomes and are affected by both parental preferences and the socioeconomic environment. We discuss statistical and experimental data showing historical development of parental styles and allowing to check various predictions of the model.