New York “closing statement” data provide unique insight into the distribution of damages and settlements amounts in ordinary tort litigation. The distribution of damages and settlements are remarkably similar, and the average settlement is very close to the average judgment. One possible interpretation of the data is that selection effects may be small or non-existent. Because the existing litigation models all feature selection bias, we develop a simple inconsistent-priors model that results in no selection bias and is consistent with the data. In addition, we show that the New York data cannot be explained by simple versions of screening, signaling, and Priest-Klein models. Nevertheless, those three models can be generalized so as to be consistent with the New York data. Thus, the New York data do not allow us to choose among the various models. One might, however, prefer the new no-selection-bias model set out in this article as the simplest model consistent with the data.
Dispute Resolution and Arbitration | Law | Law and Economics | Litigation | Torts
Date of this Version
Eric Helland, Daniel M. Klerman, and Yoon-Ho Alex Lee, "Maybe There's No Bias in the Selection of Disputes for Litigation" (July 2017). University of Southern California Legal Studies Working Paper Series. Working Paper 249.