Daubert and Litigation-Driven Econometrics

Abstract

PROFESSOR D.H. Kaye's paper raises interesting and important questions.1 The central issue that his paper tackles is when expert testimony employing statistics, particularly econometrics, should get to the jury. Based on the case study he discusses, his answer appears to be if it's not up to snuff. Kaye nonetheless emphasizes that the answer also must not exclude too much. In other words, we should admit studies that are less than ideal but still within the range of reasonable debate by experts.2 Kaye argues that the admissibility of evidence based on a statistical theory or technique should turn on whether the statistical method has been subjected to sufficient study to establish its validity as applied to a class of problems that includes the one being investigated in the litigation3 (i.e., the major premise, or overarching scientific theory4). The question [w]hether such a method is being applied properly to the problem at hand5 (the minor premise, or case-specific claims6) would generally be left to the jury. While as a conceptual matter this approach is appealing, we are less optimistic than Kaye is about its prospects for providing trial courts with principled guidance about the admissibility of statistical evidence. He concedes that the line between method and specific application can sometimes be difficult to fix,7 but we are afraid that in the case of econometric evidence and social science statistics, drawing that line will usually be problematic. Of course, there will be extreme cases in which the expert cannot justify the use of one model rather than an apparently reasonable alternative, or where