AJ Rossmiller (of AmericaBlog fame) nailed the results of the 2005 Iraqi election. You might recall that as the election where, after it had long become clear Ahmad Chalabi had little base of support in Iraq, some anonymous sources in the Administration nevertheless had great hopes that somehow Chalabi might end up as Prime Minister.

Though he lacks any mass appeal, some U.S. diplomats even cite the secular Shi’ite as a possible compromise candidate for prime minister in a coalition government.

But Chalabi won just .5% of the vote. Iyad Allawi, in whom the Administration also invested their hopes, won just 8% of the vote. And the Shiite coalition dominated by SCIRI and the Sadrists got 41% of the votes. In his book, Still Broken, AJ describes that he saw this coming.

After Iraq’s winter elections, the results validated the predictions contained in the paper I’d written in the fall. It created something of a stir because the paper turned out to be remarkably accurate, far more so than the forecasts of other agencies and departments. Before the election occurred, a high-ranking official requested a follow-up evaluation of our assessments, and I wrote a memo that described our precision. The memo made its way up through the chain, and a few days later the office got a note from Stephen Cambone, the Undersecretary of Defense for Intelligence, praising both the prediction and the self-evaluation.

Unfortunately, the bulk of the last half of AJ’s book describes how such accurate predictions are generally weeded out by higher-ranking analysts worried that their office’s work product might piss off the Administration. For example, AJ describes some of the conversations leading up to the election (edited to take out classified information), where people argued against his analysis because it didn’t accord with that of other intelligence agencies.

"You’re being too pessimistic. [The secularists] are gaining strength."

"There’s no way Iraqis will vote for [those in power] again. We can’t pass this up the chain."

"[Other agencies] are predicting something totally different and we need to make sure we’re not too far off message with this."

You can’t predict the Shiite parties will win the election, apparently, because that is considered pessimistic, because it doesn’t accord with the Administration’s great hopes that they’ll somehow salvage a secular, moderate Iraq out of their disastrous war.

AJ’s description of events leading up to the 2005 Iraqi election gets to the core of the problem with the Administration’s efforts to invent its own reality. The byzantine vetting process for intelligence virtually guarantees that when you conclude something counter to the Administration’s grand hopes, those conclusions will–at a minimum–be watered down long before someone in power ever sees them.

It was pretty clear to me, leading up to the elections, that sheer force of numbers would guarantee that Shiites would win that election (though I can’t claim to have predicted the actual tallies). But those running this country sustained a completely contrary belief, partly because of the perversion of the intelligence vetting process. In addition to their own refusal to see facts in front of their face, top Administration officials are literally shielded from the most important (and fairly obvious) facts. And the politicization of intelligence ensures that few people are even going to try to present the unvarnished truth to top officials.

When AJ was asked how he got the 2005 election right, one of the things he pointed to, half-seriously, was the open source work of Juan Cole.

I began to write the explanation of our methodology, and I tried to resist the temptation to criticize other agencies while explaining how and why we did things differently. State, in particular, was very sensitive about their screwup, and I didn’t want to piss anybody off.

"Sir, can’t I just say that I copied and pasted Juan Cole?"

You see, those running the most powerful country in the world aren’t reading Juan Cole directly, or at least they weren’t. If they’re lucky, some analyst like AJ will read him and allow Cole’s expertise to influence his analysis. And if they’re lucky, that analysis might bubble up to decision-makers without being censored by the vetting process. But AJ’s book demonstrates that those are two very big "if’s."

AJ will talk about his book at FDL’s book salon today at 5PM ET; as a special treat, former counter-terrorism czar Richard Clarke will host the discussion. I imagine this is one book salon not to be missed. And don’t forget to buy AJ’s book!