This paper uses the framework of modal logic to show how certain empirical discoveries ought to constrain political argument in logically decisive ways. Social scientists identify patterns and probabilities in the social world. While such patterns are of course real, they often appear of little consequence to philosophers dealing with modal intuitions about possible worlds. A pattern does not exclude the possibility of outliers. The recent attention to the concept of feasibility in political philosophy attempts to address some of these issues. I argue these attempts will fail insofar as they continue to rely on the use of counterfactuals and their interpretation using Lewis’ (1973) “similarity relation”. There is an alternative approach – associated with the causal inference of Judea Pearl – that is semantically equivalent to Lewis’ counterfactuals, but more empirically tractable. It is a technical branch of empirical analysis that suggests the problems associated with the feasibility literature may dissolve once we treat feasibly simply as possibility.
