Slate's Bizbox
the best policy: Making government work better.

Spooks vs. SuitsWhy the FBI and CIA don't cooperate, and why they shouldn't.

By Daniel Franklin
Posted Monday, Oct. 14, 2002, at 10:57 AM ET

After investigating the intelligence snafus that preceded Sept. 11, 2001, the members of Congress' Joint Intelligence Committee are offering some homespun wisdom to the FBI and CIA: Learn to share.

The demand for greater cooperation between spooks and G-men is compellingly simple, but it dangerously misunderstands the lessons of the 9/11 intelligence failures. It's true that the FBI and CIA don't communicate well. It's also true that an intelligence operation that crosses international borders as seamlessly as the terrorists do will be a key to preventing future attacks. But if Congress and the Bush administration are going to create an effective response to al-Qaida, they must realize that the communication breakdowns between the FBI and CIA are not evidence of the agencies falling short of their individual missions. In fact, they show just how successful the two sibling rivals are.

While geography represents an obvious difference between the FBI and the CIA—the Bureau takes the homeland, the Agency the rest—it is far from the most important. The FBI is a law enforcement organization that was designed to track down and arrest the crooks that local cops can't. The CIA is an intelligence agency that was designed to tell policy-makers what's really going on in the world. One measures its accomplishment by successful convictions, the other by successful predictions. Both use intelligence to do their jobs, but catching the kidnapper of the Lindbergh baby and determining whether Gen. Badenov still has influence with Khrushchev are two very different goals. And because the work is so different, the two agencies differ in how they collect, analyze, act upon, and share intelligence.

Take, for instance, a piece of information that was shared but not acted upon. In 1998, congressional investigators discovered, the CIA told the FBI that a group of Arab terrorists was planning to fly a plane filled with explosives into the World Trade Center. The FBI dropped the report in its bombing file and forgot about it. And why shouldn't they have? What investigation would it support? Like most CIA intelligence products, it didn't include the sources or the methods used to acquire the information because that might jeopardize the Agency's ability to continue collecting intelligence. For the FBI, a vague prediction without names, sources, or methods is useless. CIA analysts are trained to paint the big picture. "The FBI," says former CIA analyst Larry Johnson, "wants to know about a target: Who are his immediate associates, who is his family, who does he call, where does he eat? Most of that is never disseminated … because it would compromise sources."

The FBI complains that intelligence officers are too untrusting, but the Bureau has earned its reputation for burning sources. In 1985, for instance, the FBI rushed to arrest ex-National Security Agency staffer Ronald Pelton, who was suspected of spying for the Soviets. As the investigation began, then-NSA Director Lt. Gen. William Odom gave explicit instructions to watch Pelton until the case against him was rock solid and investigators learned if others were working for the Soviets. Such patience is a classic intelligence investigation technique that the FBI didn't appreciate. Odom was expecting to wait a year or more, but the Bureau arrested Pelton after only a month of surveillance. The only reason the FBI eventually gained enough evidence for a conviction was that Pelton offered to work for the Bureau as a double agent. The FBI thanked Pelton for his confession, and he was sentenced to life in prison. But, Odom says, "There's no reason he couldn't have just walked. Now, if you've had that experience, would you give the FBI any information?"

FBI agents are similarly skeptical toward the CIA agents who always seem to be getting in the way of their arrests. As Mark Riebling details in Wedge: The Secret War Between the FBI and CIA, the CIA in 1979 told the FBI that fugitive financier Robert Vesco had sought sanctuary in the Bahamas, where he gained protection by bribing Bahaman officials. Just before the FBI arrested Vesco, the CIA station chief objected on the grounds that busting Vesco might scuttle the Agency's attempts to work with those same corrupt officials to find a new home for the recently deposed Shah of Iran. While the FBI and CIA were going back and forth, Vesco figured the decidedly un-Rasta-like men following him for FBI agents and fled for communist Nicaragua and eventually Cuba.

Seen through each agency's lens of experience, the mistakes leading up to Sept. 11 begin to make slightly more sense. One can perhaps understand why the CIA might have failed to tell the FBI that two suspected al-Qaida operatives who would eventually become hijackers had entered the United States. Why risk the FBI arresting two guys who may eventually lead them to Bin Laden's cave? So, too, can one understand why the FBI failed to give the CIA the names of al-Qaida informants who were helping the investigation of the 1998 Nairobi Embassy bombing. Why risk the prosecution by involving an agency that knows little and cares even less about law enforcement? Had either been more forthcoming, the other might have been able to do more to prevent the attacks. But had either done so, it wouldn't have been doing its job, at least not in a strict sense.

That's why the reformers' calls for intelligence-sharing, personnel exchange, and token increases at the Joint Counterterrorism Center at Langley are, in the end, so empty. Ultimately, the people of each agency are going to abide by the goals of the very different organizations for which they work. They have done so for more than 50 years, despite every director of the FBI and CIA's well-meaning promises to improve coordination. And as much as Congress tries, rhetorical whippings or budgetary enticements will not fix what is essentially a problem of structure.

It's tempting, then, to change the structure. And, indeed, there have been plenty of proposals flying around the Hill that would do just that. There are two lead possibilities likely to be considered when the Joint Inquiry Committee completes its report next year. The first is separating the coordinating role of the Directorate of Central Intelligence from the CIA director's office. The second is creating a domestic surveillance agency, distinct from both the CIA and FBI, that could follow suspected terrorists once they enter the United States. The first sensibly elevates the DCI above bureaucratic bickering but will most likely not have an appreciable effect on the midlevel analyst-to-analyst coordination that counterterrorism requires. The second would address the CIA's fears by divorcing law enforcement from domestic surveillance but would probably run into too much interference from civil libertarians and the turf-protective FBI.

So here's something that might work. Create a dedicated counterterrorism agency not from scratch but by yanking the counterterrorism divisions out of the FBI and the CIA. For good measure, they should also add the counterterrorism departments of the NSA and Justice Department. By putting the string-'em-up FBI agents and the string-'em-along CIA officers in the same room, reporting to the same bosses, following the same rules, you would allow them to hash out strategies on specific information and cases. The goals of intelligence collection and criminal prosecution will still interfere with each other, but decisions of how to prioritize can be made by a single, top-level official, rather than by innumerable agents and analysts deciding on their own what to pass along and what to keep to themselves.

Admittedly, such a move would be politically impossible if it were designed to be permanent. But why would it need to be? There's no reason to believe that al-Qaida and international terrorism will be the No. 1 intelligence priority five years from now. (It barely is now, just one year after the attacks.) By sunsetting the new collaboration in, say, five years, Congress could, with presidential muscle, get around the bureaucratic resistance. After all, even J. Edgar Hoover, who tried to smother the infant CIA in its crib in the late '40s, was able to tolerate its forerunner, the OSS, when he thought that it wouldn't survive the end of World War II. If, after five years, a centralized counterterrorism agency is still needed, it can be renewed. But if terrorism returns to its previous position as just one of many potential threats, it can be disbanded or reduced. Then the FBI and CIA can go back to doing what they do best: driving each other nuts.

Print This ArticlePRINTDiscuss this in The FrayDISCUSSEmail to a FriendE-MAIL
Share on FacebookPost to MySpace!Share with MixxDigg ThisShare with RedditShare with del.icio.usShare with FurlShare with Ma.gnolia.comShare with SphereShare with Stumble Upon
Daniel Franklin is a Washington writer. You can e-mail him at .
Join the Fray: our reader discussion forum
What did you think of this article?
POST A MESSAGE | READ MESSAGES

Remarks From The Fray:

I'm sorry to quibble, but Daniel Franklin's article today doesn't say what its headline suggests it says. He does give an explanation as to why the CIA and FBI don't cooperate well, but doesn't say why they shouldn't. Actually, after shooting down a couple of ideas for making these agencies cooperate better he offers one of his own.

This is no criticism of Franklin, and I've heard worse ideas. Both agency bureaucracies would see through his little subterfuge instantly, of course, and the question of what information from those parts of each agency not part of his "temporary" terrorism group would be shared with the group would be a vexing problem. But the idea has possibilities, under the following condition: that the President insist on genuine cooperation between the agencies not just at the beginning of the project but for its duration.

As Franklin recognizes, what we're dealing with here is inertia, the driving force of more than half a century of objectives, habits, and prejudices. Inertia can be overcome, but it requires a very strong force to do it, a force the President can provide but rarely does. He does not because only rarely does a problem like this one loom large enough to demand his personal attention. A President who knows his own mind and the agencies he is dealing with can overpower any bureaucratic resistance if he decides the objective requires it. Better interagency cooperation in fighting terrorism certainly should be considered important enough to elicit sustained Presidential involvement, though I'm less than optimistic that the current occupant of the White House will see it that way.

-- Zathras

(To reply, click here.)

Based on his lengthy analysis, I see two distinct and probably separate problems. First, there is a dichotomy between the information the CIA shares and the FBI wants. The information the CIA gathers is predictive and deals with general trends while the FBI desires information that is deductive and specific. Second, there is a dichotomy in jurisdiction but it once again deals with the interest of the CIA on future trends versus the attention of the FBI on the here and now. There is a definite need to share information despite this dichotomy but not necessarily any need to share missions, structures, or jurisdictions. Rather than force the latter together to allow the former to flow more smoothly, I would suggest that finding ways to divorce them might make more sense, such that CIA and FBI operatives would be less likely to understand that they are in situations of potential mission conflict and avoid the tendency to withhold information from each other.

In the case of information gathering, the CIA would provide information as they do today but perhaps in even greater volumes to remove the need for the FBI to request information tied to specific cases. That raises the classic relevance versus retrieval problem in information management but might be solvable with a better cross-referencing system of data by the FBI that allows specific information to be tied to more general information of which they are part. Using the example Mr. Franklin provides, an FBI field agent notices a lot of Arab men are getting flight training at their local airport but cannot figure out why. Feeding the names of some of those men into the system might have turned up: 1) At least some of them were suspected/known members of al-Qaida; and 2) al-Qaida had known plans to fly an airplane into the World Trade Center. I do not mean to minimize the complexities of designing such a system but I believe it is within the ability of existing indexing and searching algorithms to accomplish.

The other advantage of this setup is that the FBI query would also become part of the database (i.e. known/suspected members of al-Qaida are currently taking flight training). First, this prevents the FBI from asking for specific enough information to allow the CIA to perform filtering on what they release. Second, it might provide the CIA additional data to understand that what they saw as a long-range plan was, in reality, only months or weeks from fruition. This would hardly solve all cross-jurisdictional clashes but, as Mr. Franklin hypothesizes, that may be an acceptable byproduct of the very real strength of keeping the two organizations separate. He seems to take the CIA-centric view that a long-term approach is generally desirable because it will detect and allow apprehension of larger crimes/criminals. That viewpoint is meritorious but it has to be counter-balanced with the need for sometime immediate action, lest analysis paralysis set in. Consider if the FBI agent who asked about the Arab men taking flight training in, say, July 2001 was told by the CIA that they were aware of the situation but great harm could be done if the Bureau did not wait at least six months to let it all play out. Now THAT would be a true intelligence disaster.

-- The Bell

(To reply, click here.)

(10/15)

Washington Post