PDF(120 KB)

The Nineties and 9/11
By Joshua Micah Marshall

On December 10 2002, little more than a year after the al-Qaida attacks on New York and Washington, the Senate Select Committee on Intelligence issued a long-awaited report on the American intelligence community and the failure to prevent the 9/11 attacks. The report cited errors and lapses on the part of various agencies. But its sharpest criticisms were reserved for the FBI. And its most ominous recommendation--from the Bureau's perspective--was to consider creating an American equivalent of MI5, the British domestic intelligence agency. By calling on the federal government to consider stripping the FBI of its domestic intelligence function and reducing its role to that of a national police agency, the Intelligence Committee issued the most serious challenge to the FBI's institutional integrity in half a century.

Less widely noticed, though no less important, was the recommendation that the newly-created Department of Homeland Security to move the intelligence community toward greater 'jointness' much as the Goldwater-Nichols Act did with the Armed Forces more than a decade ago. Though the Homeland Security Department legislation makes little provision for such a use, the Committee report recommends using the Department to create what amounts to a national intelligence joint staff. Even if the MI5 recommendation is not taken up, this move toward greater 'jointness' would weave the FBI more tightly into a more unified American Intelligence Community and thus drastically reduce the Bureau's independence and institutional insularity. The proposed creation of a new domestic intelligence agency--on the model of Britain's MI5--surprised even many of the Bureau's staunchest critics. But it should not have come as such a surprise. The proposal came after almost a decade of damaging revelations about the once seemingly untouchable agency. During the 1990s, the FBI suffered a series of public embarrassments and botched operations that brought the Bureau's public standing to perhaps the lowest point in its almost 100-year history. True, thirty years ago the agency's reputation was diminished by revelations of the rampant civil liberties abuses and harassment of left-leaning domestic political organizations. But never before had the Bureau's competence been so deeply called into question.

On their face, the mishaps involving ­Ruby Ridge, Waco, the Richard Jewell investigation, the Wen-ho Lee case, the Hanssen spy case, and the inability to detect or prevent the 9/11 attacks--seem and unrelated. Some foul-ups appear rooted in over-zealousness, others seem tied to laxity. Yet a closer examination of each event reveals an underlying cluster of institutional shortcomings that played a key role in each--ones the Committee's published Findings covered in stark detail.

Going back to the days of J. Edgar Hoover, the FBI fraternity has cultivated a deeply-ingrained sense of its own excellence and elite status within the American law enforcement community. This sense of the Bureau's separateness and superiority was based in part on very real achievements--like rooting out the garden-variety venality and corruption which were endemic to law enforcement in the Bureau's formative years. From the beginning, however, it was also based on a sometimes-comic cultivation of superficialities: even today the Bureau has no agents, only "Special Agents."

Over time the Bureau used its reputation, its position at the center of government and its powerful congressional lobby to branch out into more and more areas of criminal law enforcement and national security. Each new growth in jurisdiction brought more power and-- equally important--more funding. Thus, in addition to the FBI's work in such areas as organized crime, kidnappings and complex multi-state criminal investigations, it now has responsibility for counter-espionage, hostage rescue, counter-terrorism, cyber-security, drug interdiction, intelligence collection, intelligence analysis, to name only a few. This tendency toward mission creep and the Bureau's often excessive pride in its own excellence have often proved a fatal combination as the Bureau became involved in more and more activities for which it had little experience or for which it was simply not institutionally well-suited.

Institutional arrogance often blinded FBI administrators and agents to just how little they knew about given areas of responsibility, particularly in areas like counter-terrorism and intelligence collection and analysis--areas on which the Intelligence Committee report placed great emphasis. This intersection of arrogance and incompetence was a key factor in many of the Bureau's recent failures.

Another result of the FBI's overblown institutional self-esteem and insularity is a curious mix of unaccountability and hyper-accountability. Historically, the FBI has been extremely resistant to control or reform from the outside. And the Bureau's friends on Capitol Hill supported laws and regulatory schemes to undergird that resistance. Until recently, for example, the Justice Department's Inspector General was not allowed to investigate the FBI, even though the Bureau was, in theory at least, part of the Justice Department. In more informal ways too, Bureau personnel have perfected ways of nullifying or blocking interference from their nominal superiors. "They're unmanageable," says one high-ranking former Justice Department from the Clinton administration. "Their word for us was the 'Christmas help,' as in they'll just sort of wait us out."
Bureaucratic slow-rolling is commonplace throughout the sprawling Executive Branch bureaucracy, of course. But few parts of that bureaucracy have proven as capable of resisting outside political control as the FBI, both because of the Bureau's strength on Capitol Hill and Bureau personnel's ability to wait out anyone in the Congress or Executive Branch who would try to bring them to heel. That resistance to outside direction is perpetuated by the FBI's internal structure in which field offices operate with a high degree of autonomy from headquarters in Washington--a factor that the current Director Robert Mueller has had to contend with in his own efforts to redirect the Bureau's resources toward intelligence collection and counter-terrorism and away from the garden-variety law enforcement work many agents seem to prefer.

At the same time that the Bureau has been particularly immune to outside oversight, many of its most high-profile embarrassments have been rooted in an extreme sensitivity to political pressures and the lure of publicity. The breakdowns in the Richard Jewell and Wen-ho Lee cases, for instance, were largely due to FBI agents and administrators jumping the gun in the hopes of quickly solving a high-profile crime. Though the precise chain of events in the Jewell case remains difficult to ascertain, the primary culprits appear to have been high-level FBI administrators in Washington reacting to pressure to solve the case quickly and pressing agents in the field to push the investigation of Jewell prematurely.
The pattern in the Lee case was similar, if far more complex in its particulars. For years the FBI's attention to potential security lapses at the Los Alamos National Laboratory had been uneven and dilatory--much as it had been in the Ames and Hanssen spy investigations. Once the press and the public became aware of the situation, through leaks emanating from Capitol Hill, Bureau agents rushed the case, leading to the indictment of Lee on highly questionable evidence and probably destroying whatever chance there may have been of uncovering the actual guilty party.

As the Intelligence Committee's interest in splitting up the FBI implied, one way to solve the Bureau's problems would be to have it focus on core areas of competence like organized crime, complex financial crimes, and inherently multi-jurisdictional criminal activity. Real reform would also require Congress to make a much more aggressive effort to demand accountability from the Bureau's dispersed bureaucracy and force the Bureau to respect priorities established by competent political authorities.

Much of what is required in reforming the FBI is simply pruning the excess responsibilities and powers that the Bureau's bureaucracy has acquired over the years and having the Executive Branch and Congress end their lax oversight, which has allowed the FBI to serve its own bureaucratic needs rather than those of the public. To accomplish this, it will be critical to divide the FBI and make its intelligence collection and counter-terrorism components into a separate agency.
But if such a step proves a political impossibility--which, unfortunately, remains likely--it will at least be necessary to create a firm division within the FBI. The model should be the CIA's Directorates of Intelligence and Operations--with one division or directorate focusing on domestic law enforcement and the other on intelligence and counter-terrorism. Even if the management and bureaucratic shortcomings noted above are corrected, the mores and institutional cultures of law enforcement and intelligence work are simply too dissimilar to be easily accommodated within a single agency. This is all the more the case since the FBI's law enforcement personnel will almost certainly greatly outnumber intelligence and counter-terrorism personnel under almost every conceivable configuration. And thus the former would almost certainly dominate the latter to the detriment of effective intelligence collection and terrorism prevention.
Ruby Ridge and Waco

On August 21 1992, while preparing to arrest Randall Weaver on federal weapons charges at his home near Ruby Ridge, Idaho, U.S. marshals stumbled into a gunfight that left Marshal William F. Deegan and Weaver's fourteen-year-old son, Sammy, dead in a hail of gunfire. When the stand-off settled into a siege, the marshals requested backup from the FBI. The next day, two C-130 transport planes arrived carrying the FBI's elite Hostage Rescue Team (HRT). After surveying the situation, HRT chief Richard Rogers proposed giving Weaver's group two days to surrender or have their cabin demolished by one of the assault vehicles in the HRT's mobile arsenal. Rogers' superiors in Washington rejected this plan. But they did authorize an aggressive, shoot-to-kill order that governed operations at the site for the next three days until Weaver surrendered. By that time, in addition to the earlier deaths of Deegan and Sammy Weaver, Weaver's wife was dead from a gunshot wound to the face.

Six months later, the HRT was again mobilized to contain a crisis touched off by another failed raid by a federal law enforcement agency--this time the Bureau of Alcohol Tobacco and Firearms (ATF). Even more than at Ruby Ridge, the ATF's initial raid on the Branch Davidian compound in Waco, Texas, was excessively confrontational--with counterproductive results. The object of the raid, David Koresh, took frequent trips into town and likely could have been arrested without incident during one of those outings. Why the ATF chose such an aggressive posture remains uncertain. But available clues led to disturbing explanations. As a subsequent congressional investigation revealed, ATF officials were in the habit of videotaping one or two of the agency's most high-profile raids a year to use as props at congressional hearings in which the agency made its pitches for funding. The raid at Waco was one of those videotaped raids. In any event, the raid was poorly planned and executed, with bad communication and coordination between units contributing to the deaths of four ATF agents in the initial raid. The HRT arrived on the scene and after a 51-day siege, mixing drama and farce, launched a raid that triggered the collective suicide and/or murder of all but a handful of the Branch Davidian members.

Over the course of the ensuing decade, the story of what happened at Ruby Ridge and Waco became tangled in a web of politics and conspiracy theories peddled first by right-wing polemicists and, later, increasingly by mainstream Republican politicians. In their particulars, most of the charges leveled at the FBI were either disproved or found to be insupportable by the available evidence.

For instance, though the FBI agents at Ruby Ridge were operating under a possibly unconstitutional 'rules of engagement' order, which allowed agents to shoot anyone moving in the compound, the shot that actually killed Weaver's wife--a shot actually aimed at Weaver's adopted son Kevin Harris, whom FBI sharp-shooter Lon Horiuchi believed was about to fire on an FBI helicopter then flying near the cabin--would almost certainly have been justified under standing FBI guidelines, which justify lethal force when suspects threaten the lives of FBI agents or civilians. In another example, the FBI withheld key information about Waco for a long time and lied about its use of gas in the final raid. But the Danforth Commission demonstrated conclusively that these incendiary devices were not in fact responsible for starting the fatal fire that engulfed the compound.

Though the FBI was largely exonerated of particular illegalities, the two incidents revealed a pattern of escalations and needlessly confrontational approaches in resolving standoffs. The two events also highlighted persistent institutional shortcomings, which would reveal themselves in other high-profile failures over the course of the coming decade. Some were clear within the FBI from the very beginning.

It took more than three years for the mainstream press to question seriously the actions of the U.S. marshals and the FBI at Ruby Ridge. At the FBI, however, questioning began almost immediately. After he arrived on the scene on August 22 1992, HRT chief Rogers quickly decided that Weaver and his group were unlikely to surrender peacefully, and he devised a plan to storm the cabin after forty-eight hours if they did not surrender. The chain of command in Washington led from Danny Coulson (the founder of the HRT) to then-assistant director Larry Potts (both of whom were later disciplined for their involvement in the siege and a later cover-up of the events that occurred there). At the time, however, Coulson and Potts not only countermanded Rogers' plan, they were troubled by the rashness of his approach and how much it deviated from established methodologies used by Special Operations and Tactics (SWAT) teams and Hostage Rescue Teams (HRT) around the country.

Despite the inherently military and violent techniques SWAT teams employ, a basic premise of SWAT team work is that police authorities never seek to escalate an already combustible situation by issuing ultimatums or deadlines, especially before a detailed investigation of the situation has been made. "Any tactical plan that sets deadlines before an in-depth assessment by hostage negotiators or tactical team leaders . . . would be foolish," says law enforcement expert John Cohen, a law enforcement policy analyst with experience working for local and federal law enforcement agencies, as well as on Capitol Hill. "It's almost impossible to deescalate a situation. So you have to be very careful, as the police officer, that you maintain control, but you're not the person who causes the escalation in conflict. It's very hard to ratchet it back down. It's very easy to ratchet it back up."

Though both denied approving Rogers' aggressive 'rules of engagement' memo, Coulson and Potts were later reprimanded for having done so. But whether they signed off on Rogers' suggestion or not, it was Rogers who was on the scene and made recommendations that were by all accounts far more confrontational than the situation warranted. Though Rogers may have done nothing illegal and may have been in other respects a competent agent, the evidence pointed to the conclusion that he was not at all well-suited to running the HRT. Yet in the aftermath of Ruby Ridge incident no consideration was apparently given to reassigning him to another post. And it was Rogers who remained in charge of the HRT when the team mobilized to Waco in February 1993.

In bureaucratic subcultures there is a natural tendency to collective self-protection. Making too much of an effort to hold colleagues accountable for their lapses can invite similar scrutiny to oneself in the future. But the pattern is particularly pronounced in the FBI's highly insular culture. When considering the role of the FBI as an institution, the issue is less why one man, Richard Rogers, made the decisions he did at Ruby Ridge and Waco than why there was no effort to reassign him after the fact.

Other problems were tied to the very conception and structure of the HRT. SWAT teams specialize in rapid, violent, quasi-military operations. Executing successfully requires precision training and unit esprit de corps more like that of military than conventional police training. Maintaining that tactical edge requires virtually constant training and, preferably, routine mobilization.

The HRT was a prestigious group at the FBI and trained constantly with cutting-edge weapons and tactics. But lack of mobilization affects readiness even when training is kept up on the regular basis. SWAT teams have a delicate and complex anthropology. When preparing for dangerous, fast-paced missions, members build up to a fever pitch of adrenaline and anticipation. That requires a readiness and even eagerness to hurl oneself into a dangerous and unpredictable situation. The anticipation, adrenaline and tension can be--probably must be--intense. At the same time, SWAT members must maintain balance and the judgment to use force only when necessary.

Ideally, members of SWAT and HRT teams should operate in the same way as regular police. But the realities of human nature intrude. After such a build-up, having a suspect simply give up can be a big let down. In combination with pressures to make use of highly-trained and expensive HRTs in order to justify their often exorbitant funding, the lack of routine mobilization creates hard-to-measure but difficult-to-avoid pressures to confront such situations with the full range of available technology and weaponry.

Another problem with the organization of the HRT revealed itself more directly in the denouement of the Waco stand-off. As was widely reported at the time, one of the major reasons for launching the raid when it was launched was that members on the HRT were becoming 'fatigued'. In the public arena, just what this meant was seldom clearly articulated. But in the law enforcement community--and particularly among those involved in SWAT operations--the phrase has a very precise meaning.
Because they are mobilized so frequently, the SWAT divisions in major metropolitan police departments usually have several separate teams operating in rotation. When a SWAT team mobilizes, it confronts the situation ready for action and at a psychological peak. But that edge can be maintained for only so long. When an HRT or SWAT team is on-location for a certain period of time without training or going into action, the team will began to lose that edge. At that point they usually will be cycled out in favor of the next team in rotation.

Yet because the FBI HRT is mobilized so infrequently, the Bureau only had one of them. After more than a month had passed in Waco, the team was beginning to lose its operational edge. But there was no back-up team available. This problem--rather than ill-defined fatigue--is what created the urgency to act in the final days before the ill-fated FBI raid.

The proper question is why no other equally competent team was found to substitute in for the HRT and relieve the need for immediate action. (There were offers from several other Texas police agencies--including the state Texas Rangers.) The answer is that FBI agents on the scene--most notably Richard Rogers--refused to cede control of the scene to a non-FBI unit. Numerous law enforcement experts, including Cohen, say that the SWAT teams maintained by most major metropolitan police departments are equal to or superior to the FBI HRT. But Bureau officials insisted there was no substitute. As then-Attorney General Janet Reno told the House Judiciary Committee in April 1993, the commanders on the scene told her that the HRT would soon have to be recalled from the scene for retraining and that "all advised that there was not a substitute civil force that could secure the extensive area around the compound that had the expertise of the HRT."
Hanssen

Initial press accounts portrayed Robert Hanssen's arrest as a coup for FBI counterespionage. The truth was more sobering. Poor management, institutional arrogance, and simple incompetence long hobbled the Bureau's efforts to root out one of the most serious security breaches in American history. In 1985 and 1986 the FBI and the CIA both suffered a series of sudden and catastrophic losses of intelligence assets in the then-Soviet Union. (Though he is believed to have compromised many others, Hanssen was later charged with selling the Russians names of three Russian spying for the United States--two of whom were later executed.) As the country's lead counterespionage agency, the FBI immediately organized a task force to uncover how the Bureau's assets had been compromised. The evidence pointed strongly to the conclusion that one or more spies within the U.S. intelligence community were responsible for the breach. By 1988 the Bureau's own counterespionage branch had concluded that a mole was the only credible answer. The question was, where?

In 1996, two years after the arrest of Aldrich Ames, Justice Department Inspector General Michael Bromwich began an inquiry to discover why it had taken some eight years for the FBI to arrest Ames when simple bank statements, open high-living, and other readily available evidence seemed to point so clearly to his guilt. (Though Ames was a CIA agent, the investigation fell under the FBI's jurisdiction.) Bromwich found that the FBI's investigation was turned over to relatively low-level agents, given often indifferent attention by Bureau leadership, and broke down entirely at several points without having solved the mystery of why so many assets had been compromised. Equally troubling, when the original FBI task force reported to the Director on its findings, the report made no mention of the fact that the CIA had suffered a similar and almost simultaneous loss of intelligence assets in the Soviet Union as well. That information was critical to evaluating the seriousness of the situation. And it would have dramatically heightened the importance of the investigation and--one hopes--the priority it was accorded. Nor did the FBI make a concerted attempt to commence a joint inquiry with the CIA, despite the fact that breaches of security at the CIA came directly under FBI jurisdiction.

During this early phase of the investigation, one key question was whether the mole was in the FBI or the CIA. One of the key findings of Bromwich's Report, however, was that FBI investigators failed to seriously investigate whether the mole might be an FBI agent rather than someone from the CIA or some other U.S. intelligence agency. According to Bromwich and other intelligence community experts familiar with the investigation, the FBI simply didn't really believe one of its agents could be involved in such treachery. The issue was less one of finger-pointing--the CIA also believed the mole was one of their own.
But there was more. Bromwich found that the FBI had not only been seriously deficient in scrutinizing its own agents. It had also been deficient in even some of the most basic, established methods of counter-intelligence work. Bromwich's investigators found that the FBI's Counter-Intelligence Division practiced little of the compartmentalization which is the sine qua non of serious intelligence work. According his report, "as many as 250 FBI employees at the FBI's Washington Field Office alone likely had knowledge" of the key information in question.

That lack of compartmentalization not only insured that a potential mole would likely have had access to a very broad range of classified information. It also greatly complicated the task of investigating and detecting such a mole. Since so many FBI employees had access to compromised material, there would be few ways to isolate a list of potential suspects. Despite making this discovery, FBI investigators made no attempt to scrutinize the FBI personnel with access to the data, and the investigation itself simply ground to a halt. (Nor was there any attempt to remedy the underlying problem of lax compartmentalization that the FBI's own investigation had uncovered.) The investigation picked up again only in 1991, when the CIA itself uncovered financial information inculpating Aldrich Ames. Then the FBI finally set up a joint task force with the CIA.

As in other cases, the Bureau's emphasis on traditional criminal law enforcement techniques and institutional pride severely undercut its ability to operate outside its areas of core competency--in this case, intelligence collection and analysis and counterespionage. "This is a culture that prized sharing information not compartmenting it," says Gregory Treverton, a former Church Committee staffer who served as vice chair of the National Intelligence Council in the early 1990s. "They thought that moving information quickly was the essence of law enforcement. It also had this agent domination, so that once you'd gone through all the hoops and been admitted to the band of brothers you were trusted. The ideas that you would have 'need to know' and compartmenting didn't fit very well. It wasn't the essence of the organization." Ironically, the FBI was far too loose with information-sharing on the inside, where it did a good deal of damage, and far too stingy with it on the outside, where it could have done a great deal of good.

At the time the Bromwich Report was published in 1997, the common view was that the FBI had gotten lucky. Whatever failures of security there had been at the FBI and however lax the Bureau had been in investigating its own agents, the mole turned out to be at the CIA. Of course, this wasn't the case--or not entirely. As early as 1988 there were clues suggesting that the compromised information came from multiple sources. (The rapidity with which even high-level spies were executed suggested that the Russians were operating on more than the simple say-so of a single spy.) In 1997 Hanssen was only a short time away from commencing the third and final phase of his spying for the Soviet Union and Russia. Had Bureau investigators tightened security in the Washington field office when they made their own findings in the late 1980s--or even after the publication of Bromwich's report in 1997 -- they likely would have severely limited and possibly even prevented the spying Hanssen began in the late 1990s.

Taken together, the FBI's record in the Ames and Hanssen cases demonstrated a mix of institutional arrogance and professional mediocrity. The former, in fact, often abetted the latter. This was a combination that Bromwich himself again observed in an investigation of the FBI Crime Lab begun in1999. In response to charges by an FBI whistleblower and numerous press reports, Bromwich assembled an investigative team that included experts in forensic science from major regional crime labs as well as from academia. The investigation was the first in which outside experts in forensic science were allowed a free hand to investigate the practices and methods the Lab used. The investigators were shocked by what they found.

"The 'exchanges' with other labs were in the FBI's view one-way transmissions of technique and experience from the FBI to the other labs, when in fact those other labs were doing things sometimes at a far more sophisticated level and sometimes just better than the FBI," says Bromwich. "We had some forensic scientists from around the world who a) were shocked by the deficiencies they found in the FBI lab but then looked into what the exchanges had been like and were stunned at how few of them there were and that they were essentially one-sided, that is the FBI sharing their knowledge and wisdom with the others. I think the same thing was generalized across a lot of different subject areas. There was a lack of willingness to learn from topnotch police departments that do certain things well."

Further scrutiny into how the Lab interacted with the rest of the law enforcement and forensic science community revealed part of the reason why. Since its founding, the crime lab's administrators had seen the institution as a repository and laboratory of the latest investigative techniques in forensic science. They routinely sent out circulars to state and local crime labs around the United States outlining best practices and new techniques. Yet, revealingly, the flow of information was almost entirely one way. Few at the FBI believed that there was much the Bureau could learn from advances at major metropolitan crime labs in cities like New York and Los Angeles or in academia. The result of that mindset was that the Crime Lab not only fell behind on key advances but also enjoyed little of the scrutiny that might have alerted FBI officials to how far the Lab had fallen behind. Arrogance abetted incompetence; the two went hand in hand.

As in other cases, the FBI mixed an out-sized estimation of its own abilities with a track record that was painfully poor. And the two phenomena were directly related. Institutional arrogance and a lack of accountability provided cover for mediocrity. The fatal mix of the FBI's institutional arrogance and functional mediocrity came more and more into play as high tech and information technology became increasingly important in law enforcement. This was particularly true in computer technology--another FBI problem area that played a major role in facilitating Hanssen's treachery.

In the late 1970s the FBI began to fall behind in computer technology. By the late 1990s the problem had become acute. On the eve of Louis Freeh's departure as director in June 2001, the standard-issue FBI computer was a 386 or 486 PC, machines that had been out of date for almost a decade. Few government bureaucracies were as up to date in computer technology as the more advanced parts of the private sector. But very few were as far behind as the FBI. And still fewer were in greater need of information technology resources.

The technology the Bureau used for its central investigative database was as outdated as the individual workstations used by FBI personnel. In 1995 the Bureau brought online the Automated Case Support (ACS) system, a centralized database designed to store and aggregate the mass of information FBI Special Agents assembled in the course of investigations. Ideally, the system should have provided an automated tool for agents and analysts to 'connect the dots' between all the disparate pieces of information the Bureau collected. Agents investigating large numbers of Muslim students studying in flying schools in Phoenix, for instance, might hypothetically have found out about related concerns about Zacarias Mousaoui in Minnesota through a series of computer document searches.

But the ACS was plagued by problems from the start--partly because of poor design and partly because the technology it used was more than half a dozen years out of date by the time it came online in 1995. The system lacked the complex cross-referencing tools thar could have made it an invaluable tool for agents doing serious investigative work. "It was a dinosaur," explains FBI historian Ronald Kessler. "You could punch in 'flight' to retrieve millions of documents. But you couldn't punch in 'flight schools.' So it was almost useless. People came up with all sorts of jury-rigged ways around this ACS system. There literally were 42 different computer systems within the Criminal Division alone to keep track of different areas because this thing was so useless. If they wanted to check a name they'd have to literally go to dozens of different databases. No one would ever believe it. But that's the way it was."

These shortcomings were crucial in preventing the Bureau from doing the kind of complex 'data-mining' that might have unearthed connections leading to the 9/11 conspiracy. (By 2001, more than 80% of the Bureau's counter-terrorism data remained on paper, and largely out of reach.) But for Hanssen's spying, the problem was the ACS's shoddy security protocols. For security, ACS relied on a series of cross-referencing codes that agents were supposed to use to classify the security level of particular documents. Data could be categorized by degrees of classification or by which divisions and personnel within the FBI would be able to access them. Unfortunately, FBI employees found the codes tedious and difficult to use. And there was only a limited effort to train FBI personnel in their use. For both these reasons, the security protocols were used only sporadically and when used, often incorrectly. Indeed, ACS's abysmal security created a system of dangerous, perverse incentives: FBI personnel who were the most security-conscious and in possession of the most critical data had very good reason not to use the system at all. As a computer expert, Hanssen knew the ACS's weaknesses and used the system to harvest information that he later sold to his Russian handlers. Because Hanssen turned out to be so reliant on the ACS, those agents and field offices who had ignored regulations and refused to use the system were rewarded by not having their assets compromised. One of the ironic footnotes to the Hanssen story is that the Bureau's New York field office--which was refusing to use ACS through most of the 1990s--emerged from Hanssen's treachery almost entirely unscathed. It was a testament to the FBI's poor organization and dysfunctional culture that the New York field office's persistent insubordination turned out to be a boon for American national security.

The inattention to compartmentation uncovered by Bromwich's 1997 inquiry also showed up in the design of ACS--again, with dire consequences. With his seniority and the high-level of access Hanssen enjoyed as a counterespionage agent, he would likely have been able to access a great stock of sensitive material even if ACS's classification protocols had been perfectly designed and implemented. Had they been properly designed, however, they might have made his treachery much easier to detect.

What ACS lacked was any method of tracking the searches that individual users made--the database equivalent of compartmentalization. Hanssen, for instance, routinely searched for documents in subject areas that had little or no relevance to his area of work. That, in itself, should have raised red flags. He even used the database to track whether any investigations had been opened into his own spying. A well-constructed system--the sort of technology in widespread even in corporate America at the time--could have detected the pattern of Hanssen's usage and sent up red flags. But ACS was so poorly designed that it helped facilitate rather than uncover Hanssen's espionage.

The Bureau's shortcomings on information technology were no secret prior to 9/11 or Freeh's departure in June 2001. But there were few sustained efforts--from within or without--to rectify the problem. On the inside, the FBI's leadership--and much of its rank-and-file as well--were notoriously technology-averse. Freeh himself was known for having removed the computer from his office on taking over as Director. And through his tenure at the Bureau he never adopted e-mail--even as it was becoming commonplace throughout government and business.

But outside efforts to solve the FBI's technology problem were minimal at best. The FBI had always enjoyed strong support on Capitol Hill and a pass on the scrutiny Congress gives to many other government agencies. But there were political dynamics unique to the 1990s that further insulated the FBI from the sort of aggressive oversight that might have forced necessary changes. After 1995, as former Director Louis Freeh became a key player in the battles between the Congress and the Clinton White House, his management of the FBI became increasingly immune from criticism from either side.
Freeh was extremely popular with congressional Republicans for the tough line he repeatedly took against President Clinton. And though they criticized FBI lapses at Ruby Ridge, Waco and in the Olympic Park bombing case, congressional Republicans refrained from criticizing Freeh and thus--by extension--the Bureau's key leadership. At the same time, the White House and the Justice Department were increasingly hard-pressed to exercise control over the Bureau because any such efforts could be easily portrayed as political harassment in the face of the Bureau's on-going investigations of high-level administration officials.

Intelligence, Counter-Terrorism, and 9/11.

Even before 9/11 then-incoming Director Robert Mueller signaled his intention to remedy the Bureau's technological shortcomings. And Mueller's reforms expanded into issues of intelligence collection and analysis in the aftermath of the 9/11 attacks. But according to most counter-terrorism and intelligence experts, the problem isn't that the FBI had or has too few intelligence analysts. The problem is that, before 9/11 at least, it didn't really have any. The Bureau had employees it called 'intelligence analysts.' But these weren't intelligence analysts in the sense the CIA or, for that matter, almost any other western intelligence agency uses the term. Saying the FBI didn't have an intelligence analysis capacity, says former CIA intelligence analyst Larry Johnson, "is like saying there's a lack of police investigating capability at CIA. It's just a statement of fact."

At the CIA, an intelligence analyst focuses on a particular subject area like, say, Islamists in Saudi Arabia or Iraq. The analyst's job is to review daily all the information coming in that relates to this topic. This may include publicly available information in newspapers and other media, signals intelligence like wire intercepts, debriefings of human assets, and more. The analyst's job is to make sense of the data and to write regular, peer-reviewed reports for the president and other U.S. policy makers. Outside of the intelligence world, the closest analogy is to a reporter with a beat or a stock analyst who follows a particular industry.

But though the FBI has had domestic intelligence responsibilities for decades, it has never done business this way. Rather than having a particular beat or area of focus, analysts' work is case-driven. The difference is fundamental. FBI intelligence analysts function more like research assistants, gathering information for specific criminal investigations at the direction of the special agent running the case. As such, FBI analysts working in the Counter-Terrorism Division might develop deep knowledge and insights about some aspect of counter-terrorism from long work on a specific case. But outside of a structure of intelligence collection and analysis, the insights gained are more incidental than fundamental to the endeavor. Those insights are not tapped into a broader structure that would put them to use beyond the single individual or case. Nor are analysts working on related aspects of a given topic--like al-Qaida recruitment in Indonesia--necessarily able to coordinate or benefit from each others' work since they are divided into distinct case investigations. To professional analysts at the nation's other intelligence agencies, the heart of the analytic craft is using a process of compilation and peer review to bring together the work of all a given agency's analysts in order to serve a specific audience of top policy makers. "Even if you call someone who's working [at the FBI] an intelligence analyst," says Johnson, "they still don't have an intelligence mission. In other words, they are not writing on a daily basis to serve any particular audience."

In addition,, FBI analysts also lack status and opportunities for professional advancement. The CIA is divided into two separate Directorates--one for operations, and another for intelligence collection and analysis. This gives the two basic types of agents a rough equality in the culture of the organization. This division of labor has not prevented serious inter-directorate rivalries at the Agency. Officers from the Operations Directorate often look down on those from the Intelligence Directorate. More importantly, they have sometimes kept critical information from their analyst colleagues, much as the country's various intelligence agencies have kept intelligence data from each other. On balance, however, the Operations/Intelligence division at the CIA has kept these perhaps unavoidable rivalries and frictions within manageable limits and prevented either group from dominating the other in the competition for status, resources and professional advancement.

The situation at the FBI is very different. There, analysts have little professional status. And they are firmly subordinated to the Special Agents under whom they work. If one wanted to make a serious career at the FBI one would never become an intelligence analyst. "No intelligence analyst has ever risen to any executive level position in the FBI," notes Bromwich.
Given the low institutional priority given to intelligence analysis and the low status of the analysts themselves, it is scarcely surprising that the FBI's record on intelligence work was so spotty and error-prone. But there were still other ways in which the Bureau's emphasis on the mores and traditions of criminal law enforcement hobbled its efforts in intelligence collection and counter-terrorism. One reason that intelligence work was given such low status and priority at the Bureau was the institutional culture that prioritized gun-carrying and case-work over more sedentary information-intensive forms of law enforcement. "Real men don't type" is the mocking description often used by Justice Department employees.

Meanwhile, the Bureau's lack of outside accountability has frustrated earlier attempts to address the problem. In the late 1990s it became clear at the FBI that counter-terrorism was a winning card for getting more money for the Bureau. And over the course of the 1990s President Clinton repeatedly called for and the Congress repeatedly allocated more funding for counter-terrorism work at the Bureau, particularly in the 1996 Omnibus anti-terrorism act. But, according to FBI observers, only a portion of that money went to the actual counter-terrorism. "The problem," says Livingstone, was that "every guy who needed a secretary or an assistant for some reason over at FBI put them in under the terrorism authority. At the end of that time I bet there weren't a hundred people who were really hired for terrorism. You had all these new hires over there that covered all these bureaucratic sins." In other words, in the absence of political accountability, money that had been earmarked for counter-terrorism was doled out according the Bureau's internal priorities which viewed intelligence analysis and counter-terrorism as second-tier tasks.

The Future of the FBI

In the aftermath of the 9/11 attacks, Director Robert Mueller began a major effort to beef up the Bureau's inadequate intelligence collection and analysis capacities. In this effort, his close working relationship CIA Director Tenet proved to be a major asset. Mueller brought a team of CIA intelligence analysts to FBI headquarters to set up an intelligence analysis unit within the FBI. Mueller also has taken steps to insure that FBI-CIA information sharing does not depend on a narrow nexus of communication between functionaries at the two agencies' headquarters. To facilitate fluid information exchange between the two agencies at the level of rank-and-file agents, he has permitted the CIA to station its own agents at FBI field offices throughout the country. Yet lower down the hierarchy, resistance to change continues. Recent reports in The New York Times and other papers note that field offices are still resisting efforts to turn over routine criminal law enforcement responsibilities to state and local police.

Even under the best circumstances, the FBI as currently structured will probably never be able to build up the sort of domestic intelligence and counter-terrorism capacity that the nation needs. In the aftermath of 9/11, commentators commonly argued that the FBI needed to transform itself from an agency concerned with prosecuting criminals to one concerned with preventing crimes or terrorist attacks from occurring in the first instance. But the transition the FBI would need to make is actually different, and rather more complex. As the investigations into the Hanssen case demonstrate, the emphasis on information-sharing and agency professional camaraderie--which are properly at the heart of criminal law enforcement--are inimical to the mutual suspicion, compartmentalization and secrecy required for intelligence work.

Making that change will be a massive undertaking. The Bureau's institutional culture is deeply rooted in an ethic of gun-toting and crime-solving that has historically proven resistant to the more information-intensive modes of analysis and investigation that are at the heart of good intelligence and counter-terrorism work. Indeed, the Bureau has even resisted the more information-intensive trends that have swept through law enforcement in the last two decades. As was the case with the FBI crime lab, the Bureau is only now beginning to take cognizance of a series of new information-based crime fighting techniques that were pioneered by major metropolitan police departments in the 1980s and 1990s. New York City's COMSTAT system uses innovative organizational techniques and information-gathering technologies to track and focus attention on particular areas of the city with hot spots of specific sorts of crime. That has allowed the NYPD to focus resources and demand accountability from police leadership down through the chain of command. But as in so many other cases, the FBI's dismissive attitude toward other law enforcement agencies blinded it to how far behind it was falling even in cutting-edge policing methods.

Even with the best leadership, changing the basic institutional culture and expertise of the agency with its current personnel might simply not be possible. "One of the main problems," says former DOJ Inspector General Bromwich, "is that there is still a culture [at the FBI] of locking up bad guys. If you talk to people at the FBI academy it is almost always the case that they do it because of the FBI's reputation in the criminal investigation field, not in the field of intelligence and certainly not in the field of fighting terrorism. The manpower they get is not at bottom people interested in analysis or who come in interested in doing intelligence work. At least 80 to 90 percent go there because they want to be criminal investigators."

The tempo, priorities and institutional cultures of law enforcement agencies are profoundly different from those of intelligence agencies. And even with the best practices it is not clear that the two functions--law enforcement and intelligence gathering--could coexist well in a single agency. One specialty would almost inevitably end up distorting and undermining the other. And given the fact that the number of policing-oriented personnel in the FBI would likely always outnumber those involved in intelligence work, it would be the policing branch dominating the intelligence one. "I don't know of any law enforcement agency that is required elsewhere in the world to fulfill this kind of double role," says intelligence expert James Adams. There's little reason to think the FBI, as presently constituted, can or should.

Almost every argument--both from theory and experience--weighs heavily in favor of creating a new domestic intelligence agency--without police powers--and restricting the FBI to a more conventional national policing role. Domestic intelligence gathering raises serious civil liberties issues. But those concerns will be no less serious outside the FBI than within it. The only serious objections to splitting the FBI are not ones of substance but arguments about the political difficulty or infeasibility of accomplishing the feat. The Senate Select Committee on Intelligence report suggesting a possible break-up of the FBI should be taken up as the best--and likely the last--opportunity to use the goad of the 9/11 attacks to make this necessary change.

Joshua Micah Marshall writes the Talking Points Memo. He is a Contributing Writer for The Washington Monthly and a columnist for The Hill.
This article was cosponsored by Understanding Government and The Century Foundation.