A constructive version of Condorcet's jury theorem (CONICET, Argentina)

There has been a well-established research program for several decades that uses Condorcet's Jury Theorem (CJT) to provide an epistemic justification for democracy and majority voting. Very succinctly, the idea is that majority voting can be justified because, if certain conditions are met, it tracks the truth: at the limit, if certain conditions are met, very large groups of voters always choose the 'correct' option. In this talk I discuss the applicability of the jury theorem beyond epistemicism. I propose a version of the theorem that does not presuppose the existence of an objective fact on which the members of the group must vote; I will refer to this version as 'constructive CJT'. The possibility of circumventing the epistemic perspective shows that the basic structure of the original theorem is more versatile than we might think. Furthermore, I argue that the constructive version of the theorem reveals a surprising property of the majority rule: for some scenarios, gathering information about what the majority of the group intends to do (for example, through an opinion poll ) can encourage cooperation, assuming certain conditions are met.