blogue atlântico

Since the time of its introduction into political science, rational choice theorising has hearkened to two contrary impulses: an interdisciplinary spirit that seeks to unify social science explanation and a parochial tendency to interpret all social phenomena through the lens of microeconomics. Downs (1957), for example, explicitly shied away from social-psychological explanations, even in the face of an existing stock of knowledge that pointed him toward them, for fear of losing a distinctive voice: “Empirical studies are almost unanimous in their conclusion that adjustment in primary groups is far more crucial to nearly every individual than more remote considerations of economic or political welfare… Nevertheless, we must assume men orient their behaviour chiefly toward the latter in our world; otherwise all analysis of either economics of politics turns into a mere adjunct of primary group-sociology.”In light of Downs’s remark, it is not surprising that what rational choice theorists regard as their multidisciplinary perspective is perceived by others as a form of colonization. The impulse to defend rational choice models at all costs and against all comers evokes a dismissive respond that, if anything, reinforce disciplinary divisions. If social science were viewed less as prize fight between competing theoretical perspectives, only one of which may prevail, and more as a joint venture in which explanations condition and augment one another, the partisan impulses that give rise to methodologically deficient research might be held in check. The questions would change from “whether or not rational choice theory” to something more fruitful: “How does rationality interact with other facts of human nature and organization to produce the politics that we seek to understand?”

We have shown in this book that to date no empirically credible universal theory has been developed by proponents of rational choice. We do not find it surprising that those rational choice theorists who have grappled with empirical applications have frequently abandoned pure universalist ambitions for more subtle and modest formulations. We have also contended that taking this course need not be thought of as threatening to rational choice theorists’ scientific aspirations; on the contrary, if a variant of rational choice theory is to advance our understandings of politics, it is essential. In conclusion, it may be helpful to reiterate some of the ways rational choice scholarship must change if future research is to overcome the problems that have hampered the progress of this form of political science.

The first is that rational choice theorists must resist the theory-saving impulses that result in method-driven research. More fruitful than asking “How might a rational choice theory explain X?” would be the problem-driven question: What explains X?” This will naturally lead to inquiries about the relative importance of a host of possible explanatory variables. No doubt strategic calculation will be one, but there will typically be many others, ranging from traditions of behaviour, norms, and cultures to differences in people’s capacities and the contingencies of historical circumstance. The urge to run from this complexity rather than build explanatory models that take it into account should be resisted, even if this means scaling down the range of application. Our recommendation is not for empirical work instead of theory; it is for theorists to get closer to the data so as to theorize in empirically pertinent ways.

The injunction to theorize “closer to the data” highlights the tension between theory development and theory testing in any empirical science. On the one hand, the failure of theories to be empirically informed can result in irrelevant theorising and the mushrooming of controversies driven by little more than the theoretical conjectures out of which they emerged. On the other hand, empirically informed theorising threatens to collapse into post hoc theory-mending. The only viable way to deal with this tension is never to rest content with revised theoretical conjectures that are designed in response to previous failures of theory. Theories may be revised when they fail to account for the evidence, but the revised theory must then be tested against new evidence, and so on. In sum, rational choice theorists must discover the necessity for systematic empirical testing in the process of theory elaboration.

Second, rational choice theorists should relinquish the commitment to pure universalism and the concomitant tendency to ignore, absorb, or discredit competing theoretical accounts. The hypotheses that flow from rational choice theory would be more insightful were there a clearer distinction between rational choice theory and other modes of behaviour, and empirical tests would be more convincing and informative if they were designed to probe the limits of what rational choice can explain. This change in perspective would, among other things, encourage rational choice theorists to be more forthcoming about the conditions under which they are willing to give up their explanations in light of empirical observations.

Green and Shapiro, Pathologies of Rational Choice Theory, 1994, pgs. 202-204.