Well causal inference is considered distinct from statistical inference, and accounts for part of the gap here. (Not sure I would call that "qualitative" though.)