Mass Torts and Rational (Bayesian) Optimism

当无法获得暴露于某种产品或行为如何造成伤害的知识（例如，许多治疗药物的作用机制尚不清楚）时，使用基于人群的证据，从而依赖流行病学研究。这类证据与大规模侵权行为特别相关，在这种情况下，使用大规模销售的产品（如药物）会对一小部分使用者造成伤害，从而导致一些人的经历可能允许推断被告的责任。在其他条件相同的情况下，pi已经表明，依赖基于人口的证据会导致原告一方获得更高的预期共同赔偿，更高的和解要求，以及更多的诉讼（例如，损害赔偿较低的案件）被提起，尽管不一定会提高审判的总体可能性。本文运用信息经济学和博弈论来解决法学和经济学中的两个重要问题。首先，pi构建了一个集体诉讼形成的动态模型，该模型解释了原告最初的理性乐观和随后的诉讼浪潮，因为更多的原告观察到先前诉讼提供的公共信号。在最广泛的层面上，一个更好的阶级形成过程的理论可能最终确定改进法律制度如何补偿受害者和在大规模侵权环境中为被告提供预防措施的激励的方法。其次，没有受到伤害的产品用户会合理地对受到伤害的用户比例进行向下修正的估计。虽然这些评估与提起诉讼无关（因为他们没有受到伤害），但大众市场产品的用户可能会在审判中作为陪审员出现。第二个项目将模拟由非用户和未受伤害的用户组成的陪审团在审理此类案件时的行为，并结合各方的诉讼策略，因为他们试图回应陪审团的可能组成。

Population-based evidence is used when knowledge of how exposure to a product or behavior causes harm is unavailable (e.g., the mechanism of action of many therapeutic drugs is not precisely-understood), resulting in reliance on epidemiological studies. This type of evidence is particularly relevant to mass torts, wherein the use of a mass-marketed product (such as a drug) generates harm in a fraction of users, resulting in a population whose experience may allow inference of defendant liability. Holding all else equal, the PIs have shown that reliance on population-based evidence results in a higher expected joint payoff to the plaintiff's side, a higher settlement demand, and a larger set of lawsuits (e.g., cases with lower damages) being filed, though not necessarily a higher overall likelihood of trial. The proposed research uses information economics and game theory to address two important issues in law and economics. First, the PIs construct a dynamic model of class-action formation, which accounts for the plaintiff's initial rational optimism and subsequent waves of filings, as more plaintiffs observe the public signal provided by previous filings. At the broadest level, a better theory of the process of class-formation may ultimately identify ways to improve how the legal system compensates victims and provides incentives for defendant precaution in the mass-tort setting. Second, users of the product who are not harmed will rationally have a downward-revised estimate of the fraction of users harmed. Although these assessments are not relevant to filing suits (since they were not harmed), users of mass-marketed products may appear as jurors in a trial. The second project will model the behavior of a jury composed of non-users and unharmed users hearing such a case, incorporating the litigation strategies of the parties as they try to respond to the likely composition of the jury.