Christina and David Romers' reply to our article 'Post Hoc Ergo Propter Hoc Once More' misses the point. Our argument was never that monetary policy did not matter, but that their methods could not provide useful evidence that it did. Yet, they offer additional evidence of the same type with respect to the efficacy of monetary shocks without effectively replying to the criticisms of their methods. We show point by point that such responses as they give leave our original conclusion intact: their narrative/statistical approach is a complicated version of the fallacy post hoc ergo propter hoc; and, as such, will not sustain inferences with respect to the direction and strength of the causes of output fluctuations. © 1994.