In their 1984 article, “The Selection of Disputes for Litigation,” Priest and Klein famously hypothesized a “tendency toward 50 percent plaintiff victories” among litigated cases. Nevertheless, many scholars doubt the validity of their conclusions, because the model they relied upon does not meet modern standards of rigor. This article updates the Priest-Klein model by considering three modifications. First, we raise a novel critique of the Priest-Klein model—that it is non-Bayesian—and show that most of the results of Priest and Klein (1984) pertaining to limits nevertheless remain valid under a modified model in which parties use Bayes’ rule to refine their estimates of the plaintiff’s probability of prevailing. Second, we show that even when an incentive-compatible mechanism is imposed, many of the results remain valid for symmetric Nash equilibria. Finally, we show how the Priest-Klein model can be modified to analyze asymmetric information, show that most results are false under this modification, and compare the modified Priest-Klein model to standard asymmetric information models.


Civil Procedure | Courts | Law | Law and Economics | Litigation

Date of this Version