ABSTRACT
In Chapters 7 and 8 we illustrated how to solve quantitative causal models in which one or more of the model's variables are under the control of a single decision maker (hence, the label “single-agent intervention”). An implicit assumption in such models is that the states of those variables outside the control of the decision maker are safely assumed to occur according to some fixed stochastic process. This assumption is consistent with a wide range of managerial problems. In a large class of decision problems, however, managers find themselves in situations where the states of some variables are set independently by other agents — agents who are likely to set the variables they control according to their own objectives.
