When Threats Succeed: A Formal Model of the Threat and Use of Economic Sanctions

认为经济制裁不是外交政策的有效工具的传统观点是基于对实施制裁的许多案例的经验分析。这项工作可能会受到选择性偏见的影响，因为在制裁有效的情况下，其特点可能是没有实施制裁。也就是说，目标国可能预见到它将因制裁而改变其政策，并可能选择默许潜在制裁者的威胁，以避免经济混乱的代价。一个博弈论模型，旨在探索这种可能性，在这个项目中开发。在均衡中，制裁事件的任何结果都有可能具有非零概率;但是，有效的制裁通常会在威胁阶段改变目标的行为，并且永远不会实施。从这一模型得出的结论还表明，一个理性的、预期效用最大化的制裁战略将产生更多的实施制裁失败的案例，而不是实施制裁成功的案例。该模型产生的预期与经验观察相一致，并为经济制裁方面的一些困惑提供了解释（例如，尽管有强有力的证据表明制裁通常是无效的，但为什么制裁会继续实施）。该模式还导致一些新的假设，将某些变量（如制裁的成本和所涉问题的价值）与各种结果发生的可能性联系起来。该项目的目标是继续该模型的理论发展，并开发一个数据集，可以测试所产生的假设。目前没有关于威胁实施制裁但未实施制裁的数据。这些数据是作为这个项目的一部分收集的，并用于对理论进行严格的实证检验。这项研究有望大大提高我们对这一主题的理解。

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