There's a new National Intelligence Estimate on Iraq you might have read about. And it puts me in mind of the one written in 2002.

Dafna Linzer points out in today's Post that CIA estimates done in January 2003 and summer 2004 hold up rather well in hindsight: the 2003 estimate warned of an insurgency, and the 2004 estimate -- whose time frame ran into mid-2006 -- said the spectrum of political-security outcomes ran from "tenuous stability" to "civil war." The 2002 NIE on Iraqi WMD was an embarrassment to the agency and to the United States. Linzer writes that:

After no such weapons were found, the intelligence community -- particularly the CIA -- significantly altered the way in which it would conduct future analyses, highlight uncertainty and acknowledge dissent.

But that's not really right. After all, the January 2003 estimate was completed months before the invasion, let alone the acknowledgment of phantom WMD. Linzer is, of course, right that the estimate process has changed significantly since the 2002 NIE, and she's also right that the specter of that NIE has driven those changes. But it's worth highlighting the differences behind the 2002 NIE and the 2003, 2004 and 2007 NIEs.

In short, what the 2002 NIE doesn't have is instructive. What it doesn't have, of course, is any assessment of Iraq post-invasion. That was by design. In the summer of 2002, Bob Graham, chairman of the Senate Select Committee on Intelligence, requested and received a letter from then-Director of Central Intelligence George Tenet spelling out, in balanced fashion, an assessment of both the threat from Iraq and what would happen during and after a war. Graham, knowing the debate over war was about to heat up, learned that there was no NIE on Iraq, and requested one be drawn up, hoping to inform the congressional debate. Tenet consented -- but he informed Graham that the NIE wouldn't cover Iraqi politics, and would only cover the Iraqi WMD question.

Why? Tenet understood what his bosses wanted, and understood it very well. An NIE that assessed Iraqi politics before the congressional vote on the war would be an NIE that predicted a fragile sectarian politics and Iraqi hostility to a U.S. occupation -- in short, an NIE that looked much like the January 2003 NIE would look. That, in turn, would jeopardize the prospects for the administration winning the war vote; and quite possibly jeopardize George Tenet's job. It was a pattern that had repeated itself throughout 2002. On the question of Iraq's relationship to al-Qaeda, the Directorate of Intelligence's Middle East analysts found no such evidence for collaboration, and a host of reasons to explain that case, but the counterterrorism analysts were more open to the idea. So, when it came time to write an assessment for the White House about Iraq and al-Qaeda in the spring of 2002 -- and compete with Pentagon analysts who insisted on a connection -- Jami Miscik, the head of the Directorate of Intelligence, simply gave the job to the counterterrorism shop. (Their product, still constrained by the facts -- it was subtitled "Assessing A Murky Relationship" -- was rejected out of hand by Doug Feith's analysts as being insufficiently alarmist.)

It's good that the intelligence community isn't bending over backwards to tell the administration what it wants to hear anymore. But no one should pretend that the 2002 NIE represented an honest, good-faith effort at understanding the truth about Iraq.