Avoiding Another 'Slam-Dunk'

By David Ignatius
Wednesday, May 24, 2006

Amid all the debate about intelligence, there has been surprisingly little focus on the question the average citizen (and average policymaker, too) would probably have at the top of the list: Will these guys get it wrong again? Will they tell the world that something is a "slam-dunk," only to discover later that it didn't exist?

The issue here is analysis, the least sexy but arguably most important part of the spy world. In trying to fix what was so obviously broken, Director of National Intelligence John Negroponte did a smart thing. He went to the agency that came closest to getting it right on Iraqi weapons of mass destruction -- the State Department's tiny Bureau of Intelligence and Research -- and picked its chief, Thomas Fingar, as his deputy for analysis. INR, as it is known, had antagonized many in the Bush administration in 2002 and 2003 by refusing to endorse their case that Iraq was trying to reconstitute its nuclear program -- dismissing the claim about Iraqi uranium purchases from Niger as "highly dubious."

So how can Negroponte and Fingar be sure the analysts are getting it right this time on Iran? How can they remake the analysts' world so that their human judgment is again "the intelligence device supreme," in the words of their patron saint, Sherman Kent, who headed CIA analysis during the Cold War?

Fingar's first job was to figure out who worked for him. He took a census and found more than 15,000 analysts at the 16 intelligence agencies under the DNI umbrella. To sort out everyone's specialties, he compiled an "Analytic Resources Catalogue," a kind of yellow pages for analysts. Not surprisingly, the resources weren't being used very efficiently. "It looks like 8-year-olds playing soccer -- bunched around a cluster of issues, with large areas of the field left uncovered," explains one senior intelligence official.

Though some people (including me) think that Negroponte should consolidate all-source analysis under his DNI structure, he and Fingar have resisted that approach. Instead, Fingar wants to create "virtual teams" of analysts drawn from across the intelligence community. His model is the North Korea team he assembled back at INR, which pulled together a dozen analysts in five offices on three floors at State and linked them electronically. Fingar is applying another lesson from INR, encouraging experts to stay with the topics they know about, rather than moving from job to job as often happened among CIA analysts.

The CIA's Directorate of Intelligence will still have a crucial role, but it is no longer the central and dominant player. At the center of the new structure is the National Intelligence Council, chaired by Fingar, which reports to Negroponte. This group is made up of "national intelligence officers," many of them academic experts from outside government.

Analysts can be like a college faculty, in the best and worst senses. They can be brilliant -- but also petty and pedantic. In the past, the haggling was worthy of a faculty club, with each agency insisting on the primacy of its sources, whose identities it refused to share. Fingar wants to break that culture by demanding more transparency of sources, less watered-down consensus, more focus on alternative hypotheses and a demand that analysts "show their homework" by explaining how they reached conclusions. One metric for change has been the National Intelligence Estimates. The average gestation period for NIEs has fallen from about 400 days to about 80; the average length has shrunk from more than 60 pages to about 20.

The toughest problem may be demographics. The baby boom generation is beginning to retire, and so few analysts were hired during the post-Cold War years of the 1990s that there's a missing generation between the graybeards and the greenies. Half the analysts in the intelligence community have five years' experience or less, and this "newbie" problem will get worse with a planned 50 percent increase in CIA analysts. "You can't just add water and get an instant seasoned analyst," cautions Mark Lowenthal, who used to oversee analysis across the intelligence community and now heads a private training group called the Intelligence and Security Academy.

Will the new structure get it right? I would be happier if Negroponte tightened the reins and coordinated analysis under the DNI umbrella. If America tries to maintain multiple centers for analysis to avoid hurting feelings at the CIA, the result will be a confusing hodgepodge. But the new team has taken steps in the right direction. America may make mistakes on Iran and other intelligence challenges, but at least they will be new mistakes.


© 2006 The Washington Post Company