Unrest, Shifting Power, and Death-Watch War

Domestic unrest often creates problems of both shifting power and shifting preferences. We analyze a model in which a foreign state considers a “death-watch” attack against a domestic state temporarily weakened by unrest in which (1) two polarized factions compete for the power to set policy in the future and (2) the foreign state’s favored faction may support an invasion. Military success allows the foreign state to seize policy control, but defeat may hasten the rise of its disfavored faction when the favored faction links the outcomes of the two conflicts by bandwagoning with the attacker. In equilibrium, decisions to attack and bandwagon depend on the interaction of shifting power and shifting preferences. We show first that temporary weakness causes war, but only when the favored faction is unlikely to survive unrest. Second, increasing factional polarization can encourage or discourage foreign attack and local bandwagoning, conditional on how domestic preferences are shifting. Third, extensions show that (1) death-watch attacks are possible at any distribution of power; (2) the threat of attack may not be sufficient to deter domestic unrest; and (3) attack may be profitable even when alternative options of aiding a favored faction are available. We also use the model to explain why similar episodes of unrest in Ukraine prompted Russian invasion in 2014 but not 2004.