The conventional wisdom that economic sanctions are not an effective instrument of foreign policy is based on the empirical analysis of numerous cases in which sanctions were applied. This work may suffer from selection bias in that instances in which sanctions would be effective may be characterized by the absence of their imposition. That is, the target may be able to anticipate that it would alter its policies as a result of sanctions and may choose to acquiesce to the potential sanctioner's threat in order to avoid the cost of the economic dislocation. A game-theoretic model designed to explore this possibility is developed in this project. In equilibrium, any outcome of a sanctions episode is possible with some non-zero probability; but, sanctions that would be effective typically would bring about a change in the target's behavior at the threat stage and would never be applied. Derivations from this model also suggest that a rational, expected utility-maximizing sanctions strategy would produce more cases in which applied sanctions would fail than cases in which applied sanctions would succeed. The model produces expectations that are consistent with empirical observation and provides explanations for some of the puzzles regarding economic sanctions (such as why sanctions would continue to be imposed in spite of strong evidence that they are usually ineffective). The model also leads to a number of new hypotheses associating certain variables (such as the cost of the sanctions and the value of the issue at stake) with the probability that various outcomes would occur. The goals of this project are to continue the theoretical devlopment of this model and to develop a data set with which the hypotheses produced can be tested. Data on instances in which sanctions were threatened, but not imposed, do not currently exist. Such data are gathered as part of this project and are used to subject the theory to a rigorous, empirical test. The research promises to enhance substantially our understanding of the topic. 0137792
