Judicial Decisionmaking and Litigation Expenditure

Recent years have borne witness to a striking increase in expenditures on legal services.’ Various critics of the U.S. legal system have accordingly argued that amounts currently spent on or in anticipation of disputes are socially excessive, and several writers have attempted to address the issue through economic analysis.’ Much of the economic literature has centered on the choice between litigation and settlement, and most of the models either assume a fixed exogenous level of litigation costs or treat such costs as the reduced-form outcome of litigants’ maximization decisions.? This article abstracts from the possibility of settlement to focus on the parties’ behavior at trial. While much of the recent literature on litigation has emphasized suit and settlement, that research strategy is not the only useful one. Though most disputes do not rise to the level of a lawsuit and though the great majority of lawsuits end in settlement, decisions at earlier stages of a lawsuit depend on the parties’ estimates of the ultimate consequences of litigation. Analysis of the litigation stage is therefore logically prior to any analysis of the decision to sue or of settlement negotiation. Furthermore. the relatively small but absolutely large number of cases that go to trial consume substantial resources in doing so, and many cases that settle do so only after substantial expenditure has been incurred in preparation for trial.4 Accordingly. I developed a model of litigation expenditure based on the assumption that the parties are already engaged in a dispute. Within the model, the outcome of litigation is a probability y(X. v) that the defendant will be found liable. with the probability increasing with the plaintiff’s expenditure on legal resources and decreasing with defendant’s expenditure. More specifically, the parties expend resources on legal research to produce arguments or favorable facts to a court or other decisionmaking body such as a jury or administrative agency. The court finds for the plaintiff if and only if a certain function of the parties’ arguments, plus a random variable, exceed a liability threshold. The framework yields some interesting comparative statics predictions about the judicial process. Section II of this article outlines the basic approach to modeling the litigants’ expenditure decision and describes the model of judicial decisionmaking that is the focus of the analysis. Section III analyzes the behavior of maximizing litigants. Section IV analyzes the Nash equilibrium in expenditure and the determinants of