Legible Normativity: the Value of Silly Rules

Abstract: In this paper we model two important, we argue related, features of human normative systems: 1) that the enforcement of rules is routinely dependent on the voluntary enforcement actions of individual agents other than official enforcers; and 2) that human systems of rules frequently include rules with little or no discernible direct impact on welfare ("silly rules"). We show that agents in environments with dense normative structure (lots of silly rules) are able to more accurately and quickly determine whether important rules with consequences for welfare are effectively enforced by other agents. As a result, groups with dense normative structure are more robust to shocks to beliefs about enforcement and adapt more quickly to changes in the sustainability of enforcement. We argue that some norms, rather than directly impacting social welfare, may play a legibility function, assisting agents in their understanding of what are the active rules in a community.