Hale disputes the notion that ethnicity generates conflict or, in fact, causes anything. He also questions the value of defining ethnicity as the two dominant schools do -- as either primordial or socially constructed. Instead, he proposes, based on recent psychological research, that ethnicity is better understood as a prism through which people make sense of their social world. When ethnicity is activated -- say, in pursuit of separatism -- it is interest-driven, not identity-driven. Taking this view makes some puzzles disappear: Why, for example, did Ukraine, where ethnic factors predicted less inclination toward secession from the Soviet Union, and Uzbekistan, where they predicted more, each do the opposite of what was expected? Ethnicity matters mostly as a fillip to more direct impulses, an argument Hale pursues in an ambitious, painstaking quantitative and qualitative historical comparison of 45 ethnic regions in the Soviet Union. He finds that the relative strength of the economic stakes an ethnic region had in the union (poorer regions had greater stakes) explains better than other factors, including national consciousness, the push toward separatism. This simple proposition emerges from a complex, multilevel analytic framework.