Statistical Methodologies and the Treatment of Risk in Industrial Disease Disputes

Abstract

Workers are often exposed to hazardous industrial contaminants in the course of their employment and go on to develop serious diseases. Yet, when bringing claims in negligence, they face significant challenges in proving a causal relationship between the harm suffered and the employer’s breach of duty. These challenges arise from the methodological gaps between scientific and legal approaches to causation, and the difficulty of translating statistical evidence into findings that can aid the assessment of individual responsibility. To address these issues, the House of Lords developed a special rule of causation in McGhee v National Coal Board and Fairchild v Glenhaven Funeral Services under which claimants may discharge their burden of proof by establishing a material increase in the risk of harm. The article addresses how that rule operates in industrial disease disputes and how courts treat different empirical methodologies in applying it. Specifically, it focuses on the legal problems that arise when parties present competing statistical models. To this end, the article compares cases where courts rely on statistical models that extrapolate data from epidemiological studies to estimate changes in population-level disease incidence rates. It then contrasts this with cases where courts accept models that assess the relative risk of harm by reference to exposure incidence rates drawn from an individual’s personal exposure history. By situating these methodological choices within the broader context of industrial risk and the employment relationship, the article argues that certain choices are at risk of covertly shifting the burden of epistemic uncertainty from employers to the workforce. It also demonstrates why the choice of a given statistical model cannot be regarded as being normatively neutral as these decisions can narrow the scope of the employer’s liability in ways that are difficult to reconcile with the justificatory basis of the special rule of causation where they are not made carefully.