That's a quote from an article in the Economist. The article cites a recent work by McCloskey and Ziliak on how economists often use statistics blindly, or worse, misuse them. (Not really related, but this reminds me of John's advice to me about the danger of focusing too much on significance and too less on magnitude -- I remember it, because he cited McCloskey, as well). Most interesting from the article is, McCloskey and Ziliak criticize Becker: "... that addiction is rational, mainly on the basis that people's response to changes in price is statistically significant. This is interesting, but does not really explain much...". Hm, a divison inside the "Sweetwater Camp"?
Thanks to Y. Chang for the link.