By Fred Burton
It has been nearly a year since we first noted the churn taking place within the CIA under then-new Director Porter Goss. In the life of any organization — let alone a political one — there is bound to be some shakeout within the ranks whenever there is a change of leadership, and doubly so when the outgoing leader has been in place as long as George Tenet was. But rather than reaching a crescendo early on and then dissipating, the turnover at Langley has intensified over the past year, and many of the departures have involved seasoned officials from the Directorate of Operations (DO). Considering that the value of an intelligence officer is realized over the course of decades and entire careers, any churn in the secretive DO that is sufficiently high-level or widespread to attract the notice of mainstream news media is cause for concern. Neither intelligence agents nor senior managers — such as deputy DO chief Robert Richer, who resigned in September — are easily replaced; all require cultivation and heavy up-front investment. The causes behind the problem are numerous, and most have been amply discussed in public venues: personality clashes with Goss or dissatisfaction over his management style; purges that were deemed necessary to induce a cultural shift following Tenet's business-oriented approach
to intelligence collection and analysis; an overall intelligence community restructuring that created a new director of national intelligence (DNI) position, now held by John Negroponte. Though this last issue affects no one but Goss personally — it shifts to the new DNI the daily responsibility for briefing the president — it contributes to low morale at the Agency, where one of the perks for those who usually toil in anonymity has been the reflected glory of having your work reported directly each day to the president of the United States. Add to that the castigation to which all of the nation's intelligence organizations were subjected — though perhaps none so heavily as the CIA — for the failures
leading to Sept. 11 and unreliable intelligence about WMD in Iraq, and it is clear that there is a deep and systemic problem to be solved at Langley. Goss is now fighting back, with at least some public attempts to restore the perceived glamour of intelligence work while driving toward a 50 percent increase in the size of the clandestine service and analyst staffs. One of the strategies he is pursuing is a campaign of unilateralism — an attempt to wean the Agency from any dependencies on foreign intelligence services, rendering the CIA increasingly independent while also expanding and dispersing its agents' presence around the globe. "We are going to be in places people can't even imagine," he told employees in an all-hands meeting in late September. The approach is intriguing on several levels. In terms of resolving Langley's immediate problems — first, halting the churn — it may indeed be just what is needed. Whether the intelligence such efforts produce, and the analysis thereof, ultimately helps mend the Agency's tattered image is a question for the longer term; success on both fronts is needed if Goss is to succeed in his mission. That said, the intelligence world is riddled with interdependencies. National security, particularly on the counterterrorism front, requires a high level of coordination between the CIA (tasked with gathering human intelligence overseas), the FBI (tasked with gathering intelligence within the United States), and the State Department (which helps in protecting U.S. citizens and assets abroad, as well as with collecting intelligence), along with foreign intelligence services and liaisons and the National Security Agency. It is a complex system, unwieldy under the best of circumstances, and we would be hard-pressed to frame it as an ideal. Workable alternatives, however, are difficult to find. In the post-Sept. 11 era, all of these systems (which have existed for decades) now come together under the aegis of the Department of Homeland Security (DHS), which — on paper, anyway — is designed to vet any information about potential terrorist threats and then disperse credible and timely intelligence to the appropriate state and local authorities and the public. Now, there are still many kinks in this system, four years after the Bush administration created the DHS, but this is how it is intended to work.
The problem with a campaign of unilateralism, by the CIA or any other intelligence organization, is that unless it scores a resounding success — and quite rapidly at that (which is unlikely, given the nature of the work) — it is more likely to add to national security problems than resolve them in the near term. Fragmentation has been a feature of the U.S. intelligence system for some time, and for numerous reasons. For example, we noted intelligence from sources in February that John Negroponte — then the U.S. ambassador to Iraq — was setting up his own intelligence apparatus
within Iraq because he reportedly did not view intelligence from the CIA as reliable. The Department of Defense and other branches of government likewise have established their own intelligence channels, which are not subject to congressional oversight — and which also make holistic intelligence analysis difficult, if not impossible. This is not a new problem. Where we are now seeing it play out — often with incredible inconvenience for everyday Americans (and follow-on credibility problems for all the intelligence agencies involved) — is in terrorism scares, such as the recent "threat" to Baltimore's Harbor Tunnel or to the New York subway system, that turn out to be based on bogus intel. The difficulty in these cases was not that someone uncovered rumors of a threat, or that those rumors were passed down the chain to local authorities who took action, or even that the daily commutes of thousands of people were interrupted, with resulting costs to the community — all of these are preferable to failing to report a real threat that is ultimately carried out. Rather, the problem lies in the inability to supply timely
intelligence all the way through the chain, consistently. There are simply too many potential points of failure. The threat to the Harbor Tunnel is a perfect example of the system's increasing fragmentation, and bears close examination. We have noted that it is the responsibility of the CIA to gather intelligence overseas, but the Agency is hardly alone in that endeavor. Either the FBI or the Department of State, through its embassies, might also be present in any given country, and quite often all three can be found together — collecting and transmitting intelligence, jointly or independently, back to their home offices at Langley, Foggy Bottom or the Hoover Building in Washington. This system appears to have been fully in play with the Harbor Tunnel scare, which originated with a foreign source who was questioned by Dutch intelligence, which then passed the information on to its U.S. counterparts. (This, by the way, is the sort of liaison dependency that Goss envisions weaning the CIA from.) The "in-country" teams would huddle and send the information back to headquarters in D.C., launching a flurry of back-and-forth communications: Do you think the source is credible? Can you get more information? What about specific targets? This part of the process is not necessarily always smooth, but it does work fairly well and is common sense. The difficulties — at least for homeland security purposes — usually begin in Washington, where dozens of agencies by now have been made aware of the intelligence and are individually assessing what, if anything, to do with it. The State Department alone has a system that allows it to transmit intelligence to more than 50 government agencies simultaneously, so that all the pertinent officials are reading from the same page. In the Harbor Tunnel example, this might be a quite detailed report in some respects — explaining how Dutch intelligence picked up the human source, who he is believed to be, what specifics he gave during interrogation, and so forth. This report might conclude with what is called a "tear line" — literally, a point at which the page could be torn and a slip of paper with a homogenized message passed on by the DHS to state or local authorities and the general public. It would look something like this:
Begin Tear Line On Nov. 1, 2005, a source of unknown reliability in a foreign country advised a foreign intelligence service that a terrorist attack will take place inside the United States before Thanksgiving.
At that point, local officials would face the decision on whether or not to act in the face of what, for all they know, might be an imminent attack in their city. In the Harbor Tunnel case, roads in and out of Baltimore were shut down for about two hours before someone relayed the latest information, which had been known to intelligence agencies in the Netherlands for some time amid all the flurry: The human source was telling tales, and there was no threat to the tunnels.
There are several take-aways from this discussion. First, as we have just noted, tear-line information often is so watered down as to be nearly useless by local authorities. This is a complicated issue in itself. On the one hand, it is a symptom of all intelligence organizations' need and desire to protect their sources and methods and, at times, to compartmentalize sensitive information. On the other hand, it can be almost impossible to interpret and act upon such vagueness — and all of this is assuming that the original source in Foreign Country A was providing bona fide threat information to begin with, which frequently is not the case. The entire system is rooted in the reliability of the sourcing — a problem that STRATFOR faces as well. All of these practical difficulties have added to the cacophony of questions about the reliability of the U.S. intelligence system and fueled impulses by some agencies and local police departments to, like Goss, go it alone and collect their own intel. Increasingly, metropolitan police departments
and other security agencies have taken to deploying their own agents abroad — without the diplomatic cover afforded to official intelligence agents — due to perceived need and distrust of the existing system. We are not unsympathetic to the problem. It is human nature to prefer one's own sources to another agency's intelligence — which is often second-hand by the time it is translated into English. Acting on information from their own human sources, the NYPD, State Department, FBI or other agencies are better able to judge its reliability: They would at least have an idea of the source's identity, something about his possible connections to terrorist groups, whether he was coerced during interrogation or developed a nervous tic when discussing the reported "threat." Everyone feels more comfortable assessing and acting on the intelligence when they've had a hand in collecting it.
But, ultimately, this "pile-on" effect stands only to increase the level of kludge in the existing intelligence system — and it is questionable whether it actually serves the public, as opposed to the intelligence agencies. Because they bypass what ideally should be the firewall imposed by the DHS to shield the public from questionable intelligence, these outriders can lead to more, rather than fewer, needless panics if the local groups' threat information is not well-vetted or protected. And such scares, in turn, tend to reinforce questions and concerns about the reliability of the entire intelligence community — not just the CIA — in the minds of the public. Identifying the problems in a system with so many moving parts is, while not easy, still much easier than proposing solutions — and, as we noted above, finding viable alternatives to the existing system, imperfect though it may be, is challenging. The missing ingredient is trust, which is not endemic to the intelligence community. The task has now fallen to Goss to find ways of generating that trust in the still-tumultuous CIA, and to Negroponte — who, we note, was only months ago contributing to the fragmentation in the system — to streamline it instead.