I would think the LSE can do better to discern correlation from causation.