The evidential role of economic models in legal proceedings

Abstract Scientific models are increasingly used as sources of evidence in legal proceedings, particularly in competition cases, in which economic experts apply theoretical and empirical models to provide quantitative insights into the competitive dynamics of specific markets. Despite being routinely presented in courts, however, model-based evidence often has limited influence on judicial outcomes, undermining its intended role in the courtroom. This paper illustrates this phenomenon and then argues that the dismissal of model-based evidence in court is largely driven by what I call models’ large “attack surface”, which makes model-based evidence particularly vulnerable to challenge in adversarial proceedings. I then show that such dismissal is not without consequences. When model-based evidence is excluded, it is typically replaced by market concentration indices, which are more accessible forms of evidence but also less closely related to the target phenomenon. I conclude by arguing that it is epistemically problematic when courts systematically privilege simpler but less target-relevant sources of evidence.