The intelligence community is now engaged in a flurry of forefinger pointing over the mistaken identification of 81mm aluminum tubes as being uniquely intended for the Iraqi nuclear weapons program. According to Saturday’s Washington Post story, the analysts responsible for this “gross intelligence failure” are George Norris and Bob Campos. The biggest forefingers are pointed down to the working level at the National Ground Intelligence Center (NGIC), my former agency. The story states that both analysts have received several awards for the quality of their work.
Platt’s law gives a clue as to how two excellent analysts could come to be held responsible for this error.
I left NGIC just over nine years ago. I did not know either analyst well – but I have difficulty believing that either one of them would not have known that the tubes in question were suitable for use as conventional rockets. I believe that if they had been convinced of this they would have clearly stated their doubts about Iraqi use of the tubes for uranium purification in gas centrifuges.
With a little imagination, I can use Platt’s law to write a scenario that explains why Norris and Campos are being blamed for what is, in fact, an intelligence community failure.
The intelligence community mindset was shaped by awareness that Saddam had successfully concealed a nuclear program prior to the first gulf war. Efforts were bent on breaking this pattern of deception and proving the existence of the concealed program. Very little effort seems to have been expended on learning the exact status of these programs. This should have been a key question: even if the tubes were suitable for producing weapons grade uranium, the program would have still been years from producing a usable weapon. This conclusion would, of course, have undercut the administration's justification for going to war.
High level administration officials are now known to have been willing to override doubts from the intelligence community concerning the source known as Curveball. Reports from Curveball were used to justify operation Iraqi Freedom. The Robb-Silberman Commission seems to have overlooked this point.
And so we come to Platt’s law, a humorous (to intelligence analysts) aphorism coined by Brigadier General Washington Platt. Writing in 1951, no doubt after many frustrations in getting a National Intelligence Estimate approved, Platt considered what happened to a long, carefully written estimate. The first reviewer would shorten it, claiming that it would never be read. The second reviewer would find the short document not credible and insist that supporting arguments be added. The third level reviewer would, of course shorten it. Thus Platt’s law:
Whether or not the necessary explanatory details and pet phrases of an: intelligence paper appear in the paper as finally published, depends entirely upon whether the number of higher groups which successively review the paper is even or odd respectively.
I have no idea of the arguments involved or the case for/against the aluminum tubes as a component of the Iraqi nuclear program. I can imagine that, at every level, the analysts were suspicious of Iraqi deception and therefore wrote a lengthy and carefully worded statement. This statement would have indicated that the tubes were suitable for ground weapons but might be usable in centrifuges. (The strength of this contention might have been weakened by CIA’s withholding critical information from the NGIC analysts. (See para. 39 of this White House report). Whatever the case, we can assume that the aluminum tubes report went through extensive coordination and review at four or more levels, including the Director of Central Intelligence.
Curiously, none of the reviewers – all of whom shortened/lengthened or otherwise modified the report – has accepted responsibility for this intelligence failure. The advocates of intelligence reform speak of accountability. If accountability means that reviewers and top level officials will do nothing but point fingers at working level analysts, future intelligence failures are guaranteed.
