In lieu of an abstract, here is a brief excerpt of the content:

Chapter 7 Conclusion In concluding, I will try to present ways in which the theory can be augmented and improved as well as justifications for its basic shape and form. As was stated in the beginning of this book, I do not mean to propose that there is a single model which can predict all of human behavior or that a single actor model should be hegemonic in social science analysis. I simply attempted to provide a deductive model of preference and belief formation and to examine its implications in conjunction with the rational optimization assumption for empirical problems in comparative politics. Given this, it is reasonable to see the model presented here as one that is subject to considerable improvement, and to consider some of the ways in which this may be done. At the same time, it is important to discuss why its shortcomings are a cause for refinement rather than abandonment. Modifying the Assumptions of the Model The changes discussed here are based upon modifications rather than replacement of the rational optimization portion or the coherence portion of the model. Although it is conceivable that alternative decision-making models to rational optimization could be devised, existing alternatives either fail to yield assertive predictions or are based upon modifications of or qualifications (bounds) on rational optimization. Furthermore, as was noted in chapter 2, rationality of some sort corresponds to the way in which individuals intuitively make sense of each other's behavior. The major theoretical competitor to rationality, action based upon norms and procedures, has yet to produce an assertive general model. Moreover, given that most versions of such models imply that norms and procedures are absorbed passively by actors from their surrounding environments, an assertive general model of nonrational action will require an assertive general model of structure. As noted, it will be quite difficult to come up with such a model. The alternative preference and belief formation models to coherence-seeking are less clear, but in order to maintain generality and assertiveness while avoiding the problem of having to determine environment-specific parameters any alternative will have to be based upon preference and belief adjustment that maximizes or minimizes some quantity. As was pointed out earlier, expected utility cannot be this quantity since is always possible to adjust preferences in order to provide arbitrarily high amounts of expected utility. Furthermore, one must also choose a quantity that does not cause beliefs to violate the information. This rules out quantities that can be determined through direct observation, such as one's wealth, fame, or physical strength. Likewise, highly abstract normative quantities such as goodness can be ruled out because the impact of particular preference and belief changes on the such quantities are unclear and the resulting model will 242 Conclusion 243 be unassertive. What remains is some sort of relationship between preferences and beliefs and past actions, and it can be said in support of the model presented here that building a coherent identity out of past actions, preferences and beliefs is perhaps the most straightforward. Furthermore, it corresponds quite nicely with the common notion that individuals are constantly being remade and transformed through the process of action. Given this, three possible to the model in particular are discussed here: two have to do with bounded rationality: nonexpected utility and satisficing. Another has to do with norms: focal points. Finally, there is nonrationality. Nonexpected Utility One major possible addition to the model is the incorporation of framing effects, i.e., effects that transform preferences and/or beliefs prior to decision-making process for a particular choice set, a transformation that remains in effect only for that particular choice set. Framing effects operate in two different ways. Those found in the value portion of prospect theory as well as regret and disappointment theories, describe the effect of frames in transforming preferences. On the other hand, those found in the decision weights portion of prospect theory and other nonexpected utility theories, describe the effect of frames in transforming beliefs. The first type of transformation is compatible with the rational optimization assumption, while the second may not be. The transformation of preferences by framing comes in two main known varieties: endowment effects, in which preferences are transformed as a function of the amount possessed or consumed of desired or undesired goods and choice effects, in which preferences are transformed as a function of the various alternatives available in a choice set. Endowment effects can be found in the value portion...

Share