Josh Marshall has gone into great detail about the extent to which the hawks within the Bush administration fought a bureaucratic battle with intelligence professionals over questions of interpretation and presentation. Marshall links to this Jim Hoagland essay from October 2002 that puts the issue in stark terms:

As President Bush's determination to overthrow the Iraqi dictator has become evident to all, a cultural change has come over the world's most expensive intelligence agency: Some analysts out at Langley are now willing to evaluate incriminating evidence against the Iraqis and call it just that.

That development has triggered a fierce internal agency struggle pitting officials whose careers and reputations were built on the old analysis of the Iraqis as a feckless, inert and inward-looking bunch of thugs against those willing to take a fresh, untilted look at all the evidence.

To which Marshall points out:

[Y]ou can't separate our failure to find a lot of what we thought we'd find in Iraq from the "war" the administration has been fighting with the intelligence community for the last two years. If the administration spent the previous two years "at war" with the CIA, pushing them harder and harder into a set of assumptions (and in many cases conclusions) that turned out to be wildly off-the-mark, shouldn't there be some political accountability for what turned out to be at best a very poor call?

Marshall makes a serious point here -- the management of the intelligence process matters.

However, there are two points worth considering in response. The first is that this is hardly the first administration to take an active interest in the shaping of intelligence. As Chris Sullentrop obseved last week in his assessment of CIA director George Tenet:

Before critics such as New York Times columnists Paul Krugman and Nicholas D. Kristof lambasted the Bush administration for politicizing the CIA's intelligence analyses, spooks blasted Tenet's agency for doing the same thing during the Clinton administration. An anonymous CIA official told the National Review in October 2002 that he was badgered "for writing analyses that did not jibe with Clinton foreign policy," and another former CIA analyst wrote in 1999 on the Washington Post op-ed page, "Politicization of intelligence estimates continues to flourish under Tenet's leadership."

Now the natural counterargument to this is that "everyone else does it" is a poor defense. However, as Marshall himself acknowledges, "sometimes bureaucracies really do need to be taken on, to be shaken up." Eliot Cohen points out in Supreme Commandthat civilian leaders should intervene in the planning and management of military operations. A parallel case can be made for intelligence -- over time, intel experts become locked into their preconceptions of the raw data, and need to be exposed to rival interpretations. Skillful intervention in the intelligence process can introduce intellectual debate, which in turn can generate sharper analysis.

Of course, there's a difference between skillful intervention, mismanaged intervention, and willful ignorance of brute facts. The outcome of the debate that's currently taking place will rests on which interpretation of events will become the consensus.