The key to credible causal inference is not just the method, but how well the method fits the question at hand.
Work must necessarily precede reward.