Causation in personal injury: legal or epidemiological common sense?

The approach adopted by epidemiologists when attributing a causal mechanism to an observed statistical association is contrasted with the common law of causation in personal injury cases. By recognising the need to distinguish between probabilistic measures of (1) the strength of an association and (2) the fact-finder's 'degree of belief' in the claimant's causal hypothesis, the verdicts in a number of epidemiology-based cases, mostly in British courts, are shown to be questionable. The argument is then made for a wider application of proportionate liability, extending beyond defective drug cases (where epidemiological evidence is most often found) to medical negligence, occupational injury and tobacco-related litigation. An increased coherence in the common law of personal injury can be achieved without compromising the fundamental aims of tort and, it is argued, by reaffirming the importance of just one 'policy' precedent on liability for increasing risk