Social norms permeate society across a wide range of issues and are important to understanding how societies function. In this paper we concentrate on 'bad' social norms - those that are inefficient or even damaging to a group. This paper explains how bad social norms evolve and persist; our theory proposes a testable model of bad norms based on anecdotal evidence from real-world examples. We then experimentally test the model and find empirical support to its main predictions. Central to the model is the role of a person's social identity in encouraging compliance to a norm. The strength of this identity is found to have a positive effect on bad norm persistence. Additionally, while the size of the social group does not have a long run effect, smaller groups are more likely to break a bad norm in the short term. Furthermore, the results suggest that both anonymous communication and increasing information about others' payoffs are promising intervention policies to counter bad norms.
Social norms Experiment Identity Behavioral Economics