Can Game Theory Constrain Normative Prescriptions?

Abstract
Game theory can be a useful means toward certain political philosophical ends because it provides a conceptual and technical apparatus that constrains the inferences we can make about (counterfactual) situations involving strategic interdependence. While its usefulness for conceptual, exploratory, or diagnostic purposes seems well-established (at least, among game theory’s defenders), it seems to me an open question whether game theoretic models can be useful for the distinct purpose of disciplining normative prescriptions (e.g., regarding the kinds of institutions required by justice). At a glance, some might think the answer is clearly affirmative: game theoretic models are useful for exploring social causal mechanisms; so game theoretic models can be useful for doing the kind of causal analysis required to determine whether a particular prescription is feasible or effective for addressing extant injustice. This answer presupposes that game theoretic models are useful for making predictions about, e.g., which states of affairs we are likely to realize given the status quo, or the likely effects of implementing certain prescriptions. But suppose we’re skeptical that game theoretic models can make these kinds of predictions; what then? I’ll sketch a tentative work-around: game theoretic models can be useful for determining whether a prescription is consistent with the underlying diagnosis. That is, game theory can help determine whether the prescription is feasible and effective given the model implied by the theorist’s diagnosis of the normative problem to be addressed.