This is (yet another) great piece from Fred Kaplan, but it's got an error:
A National Intelligence Estimate is not an ordinary report. It marks the one occasion when the Central Intelligence Agency warrants its name, acting as a central entity that pulls together the assessments of all the myriad intelligence departments, noting where they agree and where they differ. Most NIEs are produced on an annual basis. Occasionally, the CIA is asked to produce what used to be called a "special" NIE. The 2002 estimate in question, titled "Iraq's Continuing Program for Weapons of Mass Destruction," was such a document. It was ordered so that the president could decide, in an informed manner, whether to go to war. The president is the main consumer of the NIE; it is written entirely for his benefit. To shrink the thing into a single page--to remove all distinctions between certainty and guesswork--is to evade the whole point.
As I was noting yesterday that's not what happened. After a bunch of hinting around in August, the president started pushing for war in early September. Then Carl Levin asked the CIA to do an NIE and Langley agreed. The president then officially asked the congress for a use of force resolution before the NIE was completed and released. Now it would still be nice to know what the president thought the NIE said, but it can hardly be said that the NIE was driving the decision-making process here.