John Loftus includes this statement in his widely posted review of John Beversluis's new edition of C. S. Lewis and the Search for Rational Religion.
In the book The Problem of Pain, coming at the heels of WWII, Lewis deals head-on with the Problem of Evil. How Beversluis tackles Lewis’ argument is probably best summed up by Christian philosopher Victor Reppert, who wrote: “If the word ‘good’ must mean approximately the same thing when we apply it to God as what it means when we apply it to human beings, then the fact of suffering provides a clear empirical refutation of the existence of a being who is both omnipotent and perfectly good. If, on the other hand, we are prepared to give up the idea that ‘good’ in reference to God means anything like what it means when we refer to humans as good, then the problem of evil can be sidestepped, but any hope of a rational defense of the Christian God goes by the boards.” (dangerousidea.blogspot.com)
Now Loftus' use of my statement is a tad misleading, because what I was doing was summarizing Beversluis's argument from his 1985 edition, not endorsing it. I have defended Lewis's treatment of the problem of evil in response to the first Beversluis edition, both in the context of The Problem of Pain and A Grief Observed, in my essay "The Ecumenical Apologist", in the four-volume Lewis encyclopedia that came out a couple of years ago from Praeger Press.
I am just mentioning this to set the record straight.