At http://delong.typepad.com/sdj/2006/08/mix_and_match.html#comment-21732824, Jim Dandy writes, apropos of Danny Okrent's belief that Paul Krugman sought to deceive his readers by mixing-and-matching numbers from the household and the establishment survey:
[DeLong writes:] "None of Okrent's sources--whoever they were--egging him on and telling him that..."
How cute! We certainly know who "they" were [Donald Luskin], but that flying attack monkey shall remain unnamed, having already been confined to "stupidest man" territory.
My guess is that that is probably not the case. We know that Okrent thinks Luskin is creepy. If A thinks B is a creep, A would have to be really stupid, a COMPLETE idiot, totally unfit for paid employment, to take B's word for technical details without verifying them. So I think the balance of the probabilities is that Okrent's source for this was not Luskin.