Signalling and Caste Mimicry in Urban India : Experimental Evidence – Audience Effects

This field-based lab experiment investigates how people in India detect each others' caste, and with what levels of accuracy they do it. It also investigate the abilities of people to pass off as belonging to caste groups to which they do not belong. Caste-based discrimination is possible when people who discriminate can detect castes of people with accuracy and they can sort their victims from others easily. But people can escape discrimination if they can pass off and deceive those who discriminate. Caste detection does not always involve signalling to a solitary receiver. Caste signalling situations often involve multi-person audiences. For instance, courts and classrooms are places where castes of individuals are assessed in the presence of multiple bystanders. This work measures the impact of multi person audiences on success rates of signaling truthfully and of passing off. It finds that multi-person audiences have an overall effect of decreasing success rates for both signalling of truth and passing off by signallers. If potential victims of caste-based discrimination wish to avoid discrimination in social settings, they could try to pass off as members of caste categories against which there is no discrimination