IS CAUSAL REASONING HARDER THAN PROBABILISTIC REASONING?

Abstract Many tasks in statistical and causal inference can be construed as problems of entailment a suitable formal language. We ask whether those are more difficult, from computational perspective, for probabilistic languages than pure (or “associational”) languages. Despite several senses which reasoning is indeed complex—both expressively inferentially—we show that satisfiability) systematically robustly reduced to purely problems. Thus there no jump complexity. Along the way we answer open concerning complexity well-known probability logics, particular demonstrating ${\exists \mathbb {R}}$ -completeness polynomial calculus, well seemingly much simpler system, logic comparative conditional probability.