… In epidemiology, causation is usually a population-level inference. Researchers ask whether an exposure reliably changes disease risk across groups of people. That requires replication, statistical association, dose-response evidence, biological plausibility, and efforts to rule out confounding. Even then, conclusions remain probabilistic. A risk factor can increase disease likelihood without proving why one person became ill.
The legal system has a different task. In civil litigation, the plaintiff generally must show that the specific injury would probably not have occurred without the defendant’s conduct. That standard does not exist to produce scientific consensus. It exists to resolve disputes.
Sometimes law and science converge. Sometimes courts assign responsibility in individual disputes while the broader scientific evidence remains mixed, unsettled, or difficult to translate into individual causation.
Public confusion arises when those situations are collapsed into one category.
Large jury awards can make scientific causation appear settled when it is not. Settlements can make legal exposure appear equivalent to admitted causation when it is not.
Related on the SLP
…
… A hazard classification is not the same thing as proof that a product caused one plaintiff’s injury. A settlement is not the same thing as admitted causation.
