Ethnic Identifiability: An Experimental Approach*

We propose a radically contingent notion of ethnic identifiability that treats identifiability as a binary relation between two individuals, conditional upon a given informational environment. We use this notion of individual identifiability to develop a language of group identifiability and group distinctness. Drawing on a sample of ninety-six undergraduate students who self declare into seven different ethnic groups, we explore the determinants of identifiability for a given ethnic demography. We find that subjects are less able to place others into ethnic categories than theories of ethnic politics assume. Subjects miscode in-group members 16 percent of the time and out-group members 33 percent of the time. They miscode out-group members as in-group members 7 percent of the time, a large number in the context of our experiment and one that indicates that errors of inclusion are more common than errors of exclusion. Subjects appear to be especially weak at incorporating new information from signals sent by individuals that may be attempting to pass. The use of signals raises the rate of correct ethnic identification to 89 percent in situations where subjects try to convince players of their true ethnic identity but lowers it to 55 percent when subjects try to pass. The net effect is a decline in identifiability, contrary to expectations from rational updating. The use of signals, we find, has a greater marginal effect when the power of signs is weak. In particular, individuals are especially poor at fooling in-group members into believing that they are out-group members and at convincing non-co-ethnics that they are in-group members. *The authors thank Chris Crabbe for his superb programming work; Dan Young, Donna Horowitz, and Kevin Thelen for their research assistance; the Russell Sage Foundation, the Harry Frank Guggenheim Foundation, the Harvard Academy for International and Area Studies, and the International Institute at UCLA for their financial support; and the staffs of the California Social Science Experimental Laboratory (CASSEL) at UCLA and the Center for International Studies and the Law School Library at USC. Extremely helpful comments were received from participants at the 9th meeting of the Laboratory in Comparative Ethnic Processes (LiCEP), University of Wisconsin, 7-8 May 2004. Protocols for the experiment are available on request from the authors.