The previous chapter has shown that patterns of Soviet-era separatism cannot be explained as a simple resource grab or as the pure expression of national consciousness, so what can it be? At least three possibilities remain. Ethnicity-as-conflictual theory may still hold that separatism is an expression of national consciousness while pragmatically making one of two arguments. Capabilities arguments posit that groups or regions can lack the capability to engage in this expression. Countervailing Incentives arguments anticipate that the ethnic separatist impulse can be overridden by important nonethnic interests that a group might have in a union. The third argument is that the relational theory of separatism is correct, that the impulse behind separatism is an ethnically charged collective action problem. This chapter weighs these three theories against each other through a comparative, process-tracing study of variation in Ukrainian and Uzbek separatism over time.

This strategy is effective because the explanatory factors central to each theory do not vary together over time during the Gorbachev era. The relational theory holds that where levels of national consciousness are equal, variation in secessionism will depend first and foremost on the degree to which central state policies demonstrate credible commitments not to exploit minority regions and a willingness to use force to preserve the union, as described in Chapter 4. During 1985–91, the Soviet government changed its policies three times in ways that prominent research on credible commitments suggests should be important.