Ballot Box

C.Y.A. Op

Intelligence failures and 9/11: If you’ve got a problem, blame someone else.

If you’re in the mood for flag-waving, breast-beating, and moaning about foul play, there’s no need to travel all the way to the World Cup. Come to Washington, D.C., where everybody’s blaming everybody else for failing to prevent the Sept. 11 attacks. Members of Congress are holding secret hearings and not-so-secret press conferences to talk about who blew it. Democrats blame President Bush; Republicans blame President Clinton; the CIA blames the FBI; the FBI blames the CIA. Everybody tells the same basic story: If only so-and-so had done this or that, Sept. 11 wouldn’t have happened.

Don’t believe it. It’s doubtful that erasing any one of the admitted failures, by itself, would have averted the attacks altogether. To suggest that it would have is satisfying to the accuser but unfair to the accused. It also gives each of the participants in this collective failure an easy out. None of them had enough information to stop the crime. So if having enough information to stop the crime is the standard of culpability, no one is culpable. If we’re going to learn anything from Sept. 11, we need to talk less about what would have been sufficient to avert the attacks, and more about what would have been necessary.

The standard exchange these days between Bush administration officials and the press goes like this: A reporter points to the latest leak about the failure to share information before Sept. 11 and asks whether, if that information had been shared, it would have prevented the attacks. The official replies, correctly, that it wouldn’t have. “I have seen no evidence that would have led me to believe that we could have prevented the attacks,” Bush said Tuesday when asked about information supposedly disclosed by Egypt before Sept. 11. “And, obviously, if we could have, we would have prevented the attacks.”

Bush’s premise is mistaken. Clinton didn’t understand the meaning of “is”; Bush doesn’t understand the meaning of “could.” Bush talks about stopping terrorism as though it’s a matter of goodwill: He’s a decent person, so if he fails to stop an attack, the reason must be that he couldn’t have done so. But “could” doesn’t imply “would.” The latter conveys certainty; the former conveys possibility. In fighting terrorism, possibility, not certainty, should be the operative principle. It suits the complex nature of investigation, gives agents practical guidance, and is a standard to which politicians and bureaucrats can reasonably be held. The question to ask about each step not taken in the months leading up to Sept. 11 is not whether it would have prevented the attacks but whether it would have kept alive a chain of investigation making that outcome possible.

Investigation doesn’t work the way politicians think it does. They tend to distinguish “information” (also known as “facts” or “evidence”) from “analysis” (also known as “speculation” or “connecting the dots”). In this simplistic model, first you gather all the information, then you analyze it. Bush, Attorney General John Ashcroft, and other administration officials like this model because it implies a straightforward solution to past intelligence failures: “gather as much intelligence as we can” (Bush) and create “a centralized way in which all that information will be transmitted to Washington so that we can see the relationships” (Ashcroft). FBI Director Robert Mueller III says he has “a briefing book that is about two inches thick that I go through that has these pieces of information in there, anything that would lead to a possible scenario such as what happened on Sept. 11.”

Many liberals favor the same model because they oppose Ashcroft’s efforts to give the FBI more authority for domestic spying. They argue that because Sept. 11 was a failure of analysis, the FBI doesn’t need more information. Investigators didn’t “understand the salience of the dots,” Rep. Nancy Pelosi, the House Democratic whip, charged last weekend. “The dots were there. It wasn’t a question of needing more collection on American citizens.”

But in practice, information-gathering and analysis aren’t separate stages. They’re interwoven. When you get information, you have to analyze it in order to figure out what kind of information to look for next. Consider the now-famous “Phoenix memo,” in which FBI agent Kenneth Williams advised headquarters of “the possibility of a coordinated effort by Osama bin Laden to send students to the United States to attend civil aviation universities.” Williams spun that scenario from sketchy evidence based on a few examples. He offered a possibility, not a firm theory.

When asked a few days ago about the Phoenix memo and the near-simultaneous Minnesota investigation of accused Sept. 11 plotter Zacarias Moussaoui, Mueller replied that “there was nothing specifically in either of those instances that gave a direct connection to what happened on September 11.” That’s true. But investigations don’t require specific, direct connections. The Phoenix memo was just a link in the chain. The next link was to check out flight schools, as Williams proposed. If the FBI had done that, it might have found enough information to get a warrant to search Moussaoui’s computer, which in turn might have exposed more of the plot and its participants.

Or consider the equally famous intelligence briefing on Aug. 6, 2001, in which Bush was told about the possibility of hijackings by al-Qaida. Three weeks ago, when word of that briefing leaked to the press, National Security Adviser Condoleezza Rice called a press conference to emphasize that it “was not a warning briefing, but an analytic report.” Rice pointed out that it was “generalized” and “talked about [Bin Laden’s] methods of operation, talked about what he had done historically” but included no “specific information saying that they were planning such an attack at a particular time.”

Rice suggested that because the Aug. 6 briefing was vague, it couldn’t have served as a basis for averting Sept. 11. But two weeks later, CIA Director George Tenet, who presumably participated in the briefing, unearthed two vital leads to the Sept. 11 plot based on similarly vague signs of trouble. According to Newsweek,

Tenet ordered his staff to scrub the agency’s files, looking for anything that might help them thwart whatever was coming. It didn’t take long to discover the file on [Sept. 11 conspirators] Almihdhar and Alhazmi. … On Aug. 23, the CIA sent out an urgent cable, labeled IMMEDIATE, to the State Department, Customs, INS and FBI, telling them to put the two men on the terrorism watch list. The FBI began an aggressive, “full field” investigation …

By then, the conspirators had gone into hiding. You can’t blame Bush or Tenet for failing to make the leap from the Aug. 6 analysis to the Sept. 11 plot. As Ballot Box has pointed out, there were other terrorist threats to consider. But you also can’t assume that just because the warning signs were “analytical,” Tenet couldn’t have taken steps in early August to track down the terrorists involved in the plot—especially since the CIA took precisely those steps in late August and was on the right track, albeit too late.

Nor can you excuse the intelligence community’s conceptual failure. “I don’t think anybody could have predicted that these people would take an airplane and slam it into the World Trade Center, take another one and slam it into the Pentagon; that they would try to use an airplane as a missile,” Rice pleaded last month. She depicted this use of an airplane as unimaginable before Sept. 11. Yet several intelligence reports had described such a scenario. It’s unreasonable to expect intelligence agencies to have known that such an attack would happen, much less where and when. But it’s reasonable to expect them to have known that it could happen—particularly since intelligence officers who haven’t read enough reports to know that a certain kind of plot is possible can’t direct an investigation in the direction of that plot.

In Washington, the blame game is just beginning. First the FBI pleaded that it couldn’t have cracked the case based on what it knew. Then the CIA offered the same plea about what it knew. Now the FBI is saying that it could have cracked the case if it had known what the CIA knew, and the CIA is saying it could have cracked the case if it had known what the FBI knew. Each side hopes to escape scrutiny by framing the other. Nice try, spooks.