Elissa Philip Gentry
The inherent mismatch between the questions law asks and the answers statistics provides has led courts to create arbitrary rules for statistical evidence. Adherence to these rules undermines deterrence goals and runs the risk of depriving recovery for whole categories of injuries. In response, some courts adopt new theories of recovery, relying on the loss of chance doctrine to provide some relief to injured plaintiffs. These solutions, however, only serve to exacerbate the fundamental misunderstanding of probabilities. While these doctrines largely operate within the context of medical malpractice, the increased ability to capture more statistical data may prompt courts to acknowledge the probabilistic nature of causation in other contexts. It is important to ensure that courts correctly approach this information. This Article presents a simple framework for thinking about probabilistic harm. The framework identifies the “attributable risk rate” as the correct metric for assessing whether a plaintiff belongs to the “avoidable” class—people who would not have experienced harm in the absence of negligence—or the “inevitable” class—people who would have
experienced harm even in the absence of negligence. The Article then proposes a practical two-step (“personalize/operationalize”) process for using attributable risk rates to assess causation. It provides a concrete example of how this process compares to other legal rules. It also demonstrates that this process is compatible with current legal requirements, harmonizes the treatment of causation in probabilistic and non-probabilistic contexts, and ensures that statistical evidence is taken seriously.