Situational Ethnic Identity, Attitudes, and Political Violence

• In Event: Identity and Attitudes in Times of Political Upheaval

Abstract

We have long known that identity is situational, but surprisingly little research has probed the extent of this phenomenon and its consequences for research, especially survey research. Our study investigates the extent to which different randomly assigned social settings can lead individuals to represent their own identity differently in the context of war, a hard case for a study of identity flexibility since war is often regarded as hardening identities. We do so by exploiting survey methodology itself, in particular the assignment of interviewers to respondents. Specifically, we derive the hypothesis from theories of situational identity that individuals interviewed by members of different ethnic groups are likely to identify their own ethnicity (including on the dimensions of nationality, native language, and language spoken in daily life) differently. We do so through an original survey of Ukraine’s population in early 2020 in which we randomize assignment of a significant portion of the interviewers by ethnicity. This, we believe, will be the first effort to use random interviewer assignment in a study of ethnic identification; while it is well documented that the identity of an interviewer can influence how respondents answer questions on attitudes (Davis 1997), existing studies either do not study ethnic identity or do not actually employ random assignment (e.g., Adida et al. 2016). We expect to confirm our own suggestive findings from a 2014 study that did not employ random interviewer assignment, but found a strong correlation between the interviewer’s own ethnic identity and (a) the self-reported ethnic identity of the respondent and (b) the respondent’s self-reported preferences and behaviors. For example, when people are interviewed by enumerators who spoke Ukrainian as a child (which we posit would correlate with being identifiable today by interlocutors as a “Ukrainian-speaking Ukrainian”), we find that respondents are over 20 percent more likely to report using Ukrainian in private life, are about 20 percent more likely to report supporting NATO membership for Ukraine, and are about 15 percent less likely to want to compromise with Russia in the conflict. If confirmed, these findings have at least two major implications. For one thing, they indicate ethnicity identity is more dependent on the social situation, and hence more politically manipulable, than commonly posited—even in the context of identity-charged war as with today’s Russia-Ukraine conflict. Moreover, these findings indicate that previous survey-based findings on identity’s importance in driving outcomes may reflect omitted interviewer effects rather than identity effects. It is possible, then, that we may be seriously overestimating ethnicity causal role in political violence.

Authors

• Volodymyr Kulyk, Institute of Political and Ethnic Studies, National Academy of Sciences of Ukraine

• Henry E. Hale, George Washington University

• Olga Onuch, University of Manchester