Collaborative Research: NSCC/SA: Behavioral Insights into National Security Issues

该奖项由国家安全、冲突和合作竞赛的社会和行为维度资助，该竞赛是国家科学基金会和国防部之间的合资企业。在最近的恐怖活动之后，围绕恐怖主义、反恐、预防和安全发展了许多博弈论模型。然而，这些理论中很少有人在实地或在实验实验室的风洞环境中对行为的准确性进行了测试。本研究使用实验经济学的方法来检验捕捉安全决策关键特征的三种类型的理论。第一组实验研究的是攻击和防御游戏，两名玩家将有限的资源分配给一组固定的目标，拥有最多资源的一方获胜。第二组测试各国在决定是追捕恐怖组织、保护自己的领土，还是什么都不做时的合作理论，这种情况被建模为嵌套的囚徒困境游戏。通过实验，研究人员将研究这些游戏是如何进行的，并探索有可能加强合作的制度（如贸易制裁或成本分担规则）。第三组实验针对自杀式袭击。此类袭击的动机往往是对国家间经济不平等的报复。研究人员将在实验室中创造不平等的群体，并改变不平等的程度和群体认同的强度。弱势群体的参与者可以自愿牺牲自己的收入来提高自己群体的表现。研究人员将评估牺牲行为如何随着群体属性的变化而改变。总之，本研究解决了这些理论是否（以及何时）能很好地描述人们的行为，并提出了改进和领域限制以增加其说明性的问题。拟议活动的智力价值源于旨在模拟国家安全战略形势的游戏的开发和测试。在开发方面，本研究对现有游戏进行了改进，以更好地捕捉特定的潜在战略情境，并在第三个类别中开发了新游戏，以捕捉团队间竞争的志愿动机。在测试方面，这些实验代表了这些游戏的第一个数据，因此有助于了解个体在战略情境中的行为。目标是确定可能适用于大部分人口的稳健的心理规律。拟议活动的更广泛影响源于对个人在这些战略情况下如何行动的更多理解。通过识别和避免导致个人做出次优决策的常见错误，这使美国及其盟友能够改进决策。它还可以通过识别敌人可能犯的错误来改进决策，并帮助描述利用这些错误的策略。

This award was funded through the Social and Behavioral Dimensions of National Security, Conflict, and Cooperation competition, a joint venture between NSF and the Department of Defense. In the wake of recent terrorist activities, a number of game theoretic models have been developed around terrorism, counterterrorism, prevention and security. However, very few of these theories have been tested for behavioral accuracy, either in the field or in the wind-tunnel environment of the experimental laboratory. This study uses the methodology of experimental economics to test three types of theories that capture key features of security decisions. The first set of experiments investigates attack and defense games, where two players allocate limited resources over a fixed set of targets, and the side with the most resources on a target wins it. The second set tests theories of cooperation among countries in the decision whether to pursue terrorist organizations, protect their own territory, or do nothing, a situation that has been modeled as a nested prisoner's dilemma game. Through the experiments the investigators will examine how these games are played and explore institutions (such as trade sanctions or cost sharing rules) that have the potential to enhance cooperation. The third set of experiments addresses suicide attacks. Such attacks are often motivated by retaliation for economic inequality between nations. The researchers will create unequal groups in the lab, and vary the extent of the inequality and the strength of group identity. Participants in the disadvantaged group can volunteer to sacrifice their earnings to improve the performance of their own group. The researchers will assess how sacrificing behavior changes in response to variation in the group properties. In summary, this research addresses the question of whether (and when) these theories provide good descriptions of how people act, and suggests improvements and domain limitations to increase their prescriptive power. The intellectual merit of the proposed activity stems from the development and testing of games designed to model strategic situations in national security. On the development side, this study refines existing games to better capture specific underlying strategic situations, and, for the third category, develops new games to capture the incentives to volunteer in between-group competition. On the testing side, the experiments represent the first data on any of these games, and thus contribute to knowledge of how individuals act in strategic situations. The goal is to identify robust psychological regularities which are likely to apply across a large segment of the population. The broader impacts of the proposed activity stem from increased understanding of how individuals act in these strategic situations. This enables improved decision making by the US and its allies, by identifying and avoiding common mistakes that lead individuals to make sub-optimal decisions. It also improves decision making by identifying the mistakes enemies are likely to make, and helping to describe strategies to exploit them.