This article describes a substitution model of states' responses to disside
nt behavior and a statistical test of some sequential hypotheses that are d
erived from the model. It is motivated by an interest in understanding the
sequential response of stales to dissident activity. That is, if dissidents
protest, what will the state do next? Similarly, if dissidents are coopera
tive, what will the state do next? The author argues that the answer to bot
h of these questions depends on the interaction of the state's most recent
behavior (i.e., repression or accommodation) and the dissident's response.
The model produces the hypothesis that states substitute repression for acc
ommodation, and vice versa, in response to dissident protest. Statistical a
nalysis of evidence from Peru and Sri Lanka, 1955 to 1991, suggests that th
e model captures well the sequential responses of the Peruvian and Sri Lank
an governments to dissident behavior during that period.