Abstract

According to instrumental proceduralism, political power is justified when it is the output of a reliable procedure. In this paper, I examine how procedures are supposed to confer normative properties. Based on this assessment, I conclude that many proceduralists set the reliability bar too low. Next, I motivate two additional requirements for instrumental procedures. I introduce the notion of “predictable” procedural failure and argue that in order for a procedure to confer legitimacy or other normative properties on its output, it must not have failed predictably. Finally, I argue that even when procedures are highly reliable, it must not be the case that their failures fall disproportionately on certain people. The goal is to develop an instrumental proceduralism that is more sensitive to the failures of real institutions.