Mr. Shingo Murakami, MA

Abstract:

When harmful levels of pollution are caused, victims should be entitled to compensation from the polluter. For people to win a lawsuit claiming they were harmed by pollutants in the environment, they have to prove that there was a causal link between their contact with those pollutants and the harm they suffered, such as respiratory damage or other illnesses. While epidemiological evidence often supports these plaintiffs, it typically addresses type-level causation, which diverges from the courts' demand for token-level causation. This conceptual distinction carries significant practical implications, especially for the burden of proof.

Judea Pearl demonstrated that even if type-level causation is established through statistical adjustments for confounding variables, these methods aren't enough to make probabilistic claims about token-level causation [1]. Nevertheless, legal professionals often misapply findings from population-level studies, assuming they can be directly applied to individual cases without making the necessary additional assumptions. Should this discrepancy arise, victims who are legally entitled to compensation may be denied a compensatory award by the courts, or, in turn, parties who contributed to the pollution may incur disproportionate liability.

We'll begin by clarifying the decision criterion for assessing type-level causation in legal practice. We'll then explore why this criterion doesn't correctly determine token-level causation, which is an essential step in assigning liability among those allegedly contributing to pollution. Finally, we propose an improved framework for assessing the probability of token-level causation. We believe this framework also offers guidance in areas such as drug-related harms and industrial accidents, where epidemiological evidence is frequently employed.