Judging Models in Legal Fact-finding

Abstract

Scientific, statistical, and economic models are often crucial in legal fact-finding. Competing models yielding different inferences may represent the same phenomenon. This article studies the judgment of models in legal fact-finding, in particular where competing models and inferences must be judged. Three categories of the use of models are identified: formal models provided by experts, formal models not scrutinized as expert knowledge, and the use of models more informally to aid storytelling. In the first case, the nature of expert knowledge makes it difficult for non-expert fact-finders to judge the conclusions, which emphasizes the importance of judging the reliability of the methods used to reach the conclusions, both on their own terms and in conjunction with the specific circumstances of the case. In the second case, the challenge is to benefit from the logic of models, but avoid tautologies disguised as expert knowledge. In the latter situation, the methods applied may not be transparent, which emphasizes the need to judge the conclusions on the basis of explanatory power. It is argued that it is important to judge models according to the appropriate standard dependent on their use. The study mainly draws on examples from antitrust, regulation, and torts.