Saturday, January 10, 2009
Social agency and rational choice
One of the reasons that rational choice theory (RC) is appealing is that it is an agent-centered approach to social explanation: explain the social outcome on the basis of an analysis of the beliefs, intentions, and circumstances of the individual agents who make up the social setting. What rational choice theory adds to this description is a specification of the decision-making processes that are attributed to the individual agent -- typically, that the agent has a consistent set of preferences among accessible alternatives and that he/she chooses in such a way as to maximize the satisfaction of this set of preferences. This can be paraphrased as a "utility-maximizing" model of decision-making.
Many objections have been offered against rational-choice theory as a basis for social explanation -- for example, that it overlooks social motivations, that it presupposes egoism, that it over-simplifies the logic of practical reasoning, or that it fails to correspond to typical human behavior. (See Green and Shapiro, Pathologies of Rational Choice Theory: A Critique of Applications in Political Science, for a developed set of critiques.)
Two points are worth underlining here. First, rational-choice theory has a major theoretical advantage precisely because it is an agent-centered framework. RC theory is one possible way of articulating a set of hypotheses about how individuals reason and act. This is a major advantage in comparison to explanatory frameworks that essentially assume programmed behavior on the part of participants in a social event. Moreover, the assumption of preference-satisfaction lines up pretty well with a somewhat broader conception of human action in terms of goal-directedness and purposiveness. If we believe that individuals have goals and purposes that underlie their choices and actions, then it is an appealing simplification to represent their actions as the outcome of deliberation about goals, strategies, and circumstances. In other words, RC theory can be seen as a specification of a philosophical idea of human action that is at least as old as Aristotle: the idea of individuals as deliberative, purposive agents. And this is in fact a credible and empirically defensible theory of action.
But a second point is equally important: RC theory and its model of utility maximization is only one out of a range of possible specifications of the idea of deliberation and purposiveness. There are important alternative specifications that can be offered. For example, we might say, along with Kant, that individuals possess a set of moral rules as well as a set of specific goals, and that they deliberate among possible choices of action on the basis of both considerations. How do the various possible actions conform to the moral rules? And how do they do from the point of view of accomplishing my goals? This process of reasoning is "deontological" -- that is, it cannot be subsumed under a simple model of maximizing rationality. It is, nonetheless, an intelligible interpretation of what rational human decision-making involves. (Mark Johnson's Moral Imagination: Implications of Cognitive Science for Ethics is interesting in its effort to bring cognitive science into dialogue with ethical theory.)
Another possible interpretation of the basic idea of deliberativeness that diverges from RC theory is one that illustrates some themes that Amartya Sen (On Ethics and Economics), James Scott (The Moral Economy of the Peasant: Rebellion and Subsistence in Southeast Asia), and Doug McAdam (Political Process and the Development of Black Insurgency, 1930 -1970) have emphasized: that real human social behavior is a complex mix of commitments, loyalties, emotions, solidarities -- as well as purposes and goals. So a theory of action that isolates "goal-directedness" and its associated framework of utility maximizing, is one that already overlooks a set of motivational factors that are crucial to explaining real social behavior. It is as if we imagined modeling wine-tasting judgments by experts but "erased" their sense of smell. Given that smell is a crucial ingredient of the experience of tasting wine, our reduced theory won't do a very good job of explaining discrimination across samples by the experts.
Instead, when we undertake to explain an individual's action in the context of a spontaneous rent strike, we need to ask a series of questions: what does he/she expect to get out of the action? What loyalties does he/she have to the organization or other participants? What principles does he/she endorse that are relevant to the context of proposed action? What forms of social identity does he/she embody, and how are these strands relevant to the decision to participate or not? How do the emotions created by the words and actions of others influence one's behavior?
What makes RC theory useful in spite of these complexities of actual motivation is the fact that there are many important situations of choice where other sources and structures of motivation are of minimal importance. When a person chooses a new toaster, it is likely enough that solidarity, emotion, principle, and identity drop away, and the choice is based on perceived value and price. So the market in toasters behaves pretty much as neoclassical economics predicts. But the market for shoes is probably more complicated: emotion, status, style, identity, and a preference for "fair-trade" products may influence one person to buy the more expensive and less functional pair of shoes, while another person will go for the good buy. Likewise, the decision to join AARP is likely to be a fairly simple calculation -- what are the side benefits of membership, how much importance do I attribute to being part of an organization that represents the public good of people over 50, and how much can I gain from there being a successful AARP? This is simple in comparison to the situation of a rent strike or a street demonstration, where one's face-to-face relations with other potential participants may have a very large impact on the decision to participate or not.
So we might say that RC theory represents a special case of the more general category of deliberative action, which is itself a sub-category of intentional action. And RC theory will be most successful in generating explanations in social circumstances where the other sources of social motivation are largely silent -- for example, anonymous market transactions, isolated decisions about participation or non-participation in collective action, and decisions about portfolio investments.
What we really need is a more developed and adequate theory of social agency -- a better account of how the various factors mentioned above fit together into one scheme of deliberation and decision-making. This might be called a full theory of practical rationality. And it would be a more general specification of the situation of agent-based social action than RC is able to provide.