Empirical Evidence in Political Philosophy

There are calls for empirical evidence to be used more widely in normative political theory. This paper supports those calls, but argues that the nature of the empirical evidence is important. It is not simply any social facts or even empirical generalizations that should be utilized within normative theory. Rather normative theorists need to take account of the mechanisms which underlie known empirical generalizations, and the mechanisms that incentivise human agents in their behaviour. The paper describes two sorts of mechanisms and their interaction. Psychological mechanisms are ones that consciously or unconsciously lead individuals to make choices. Many of these mechanisms are well-known and studied in social-psychology and behaviorial economics – such as framing effects – and are described in the extensive ‘nudge literature’. They also include conscious effects such as strategic behaviour in voting and other social and political contexts, and are well studied in axiomatic-deductive modelling such as social choice and game theory. Automatic mechanisms include the manner in which social institutions translate behaviorial responses into outcomes – most notably through voting mechanisms but also through other institutional processes. Automatic and psychological mechanisms interact as agents come realize through their interactions how automatic mechanisms work and react strategically. I argue that what we have learned in social science about automatic and psychological mechanisms and their interactions are vital to anyone wishes to design constitutions for just societies.