Ballot-box vigilantism? Ethnic population shifts and xenophobic voting in post-Soviet Russia

To what extent does voting for anti-immigrant parties relate to long-term changes in ethnic composition within states? Four theoretical models are developed, based on studies of interethnic attitudes, housing segregation, racial violence, and hate crime in the United States. Each model is tested with the data on ethnic composition of the Russian Federation from 1989 to 2002 and voting for the extreme nationalist Zhirinovsky Bloc in the 2003 parliamentary election, using multiple regression and ecological inference methods. Most consistently supported is the "defended nationhood" model derived from the sociology of neighborhood vigilantism and the psychology of the security dilemma. Non-trivial, counterintuitive findings are: (1) xenophobic voting was responsive to changes in the proportion of some ethnic groups more so than others and not necessarily those that were more numerous or more widely disliked at the time of the vote (Chechens), but those that raised more uncertainty about the future ethnic composition and identity of the state (Asians); (2) levels of change, but not the rapidity of change in the ethnic composition of the population related significantly to xenophobic voting; and (3) greater percentage of the nation's dominant ethnic group in a region reduced xenophobic voting by members of that dominant group (the highest share of Slavs voted for Zhirinovsky in the ethnically mixed Volga-Urals area).