Intelligence-gathering is something of a square peg in the round hole of contemporary political morality. It is about unearthing that which is willfully concealed, an enterprise that necessarily calls for invading privacy and inducing betrayal—discomfiting acts in an age that exalts the individual and his liberties above community and country. It is about assuming and preparing for the worst in an era that sees “bad” as an outmoded adjective for “different,” another dash of enlivening spice in a rich social stew. Intelligence is gimlet eyes in a world of rose-colored glasses.
Now, however, that foreign pathologies long denied have visited their excesses upon us, many among the benignly tolerant have turned overnight into the equivalent of ambulance-chasers. In particular, they have confidently laid at the door of America’s intelligence apparatus the success of America’s enemies on September 11, 2001. Even as investigators in the CIA and FBI were unable to “connect the dots,” it is said, nineteen al-Qaeda hijackers cavorted for months in this country before carrying out the atrocities of that day. Nor was this catastrophe—“by definition, the worst intelligence failure in our country’s history,” in the words of the Reagan-era intelligence expert Herbert Meyer—a singular phenomenon. Less than a year earlier, a billion-dollar battle ship, the U.S.S. Cole, had been bombed and nearly sunk, causing the deaths of seventeen servicemen, because we unwittingly berthed it in the al-Qaeda-infested port of Aden, Yemen. This, after our embassies in Kenya and Tanzania were turned to rubble in August 1998 by the very same al Qaeda, which had already attacked numerous times previously, and which no less often had expressly declared war on the United States.
Nor is that all. Thanks to our failed intelligence services (the indictment continues), the Bush administration grossly overestimated the stockpiles and production capacity of chemical, bacteriological, radiological, and nuclear weapons of mass destruction (WMD) in Iraq. In the meantime, in North Korea, construction of nuclear weapons seems to have ensued for years right under our noses. And Pyongyang’s mischief marked only a single strand in a web of proliferation woven by our ally Pakistan, a web that may have spread into as many as seven nations, including Iran, where the mullahs now harbor the remnants of al Qaeda’s leadership.
How did this wide wreckage in our intelligence capacities come about? One incisive answer has been given by Mark Riebling in his gripping history, Wedge: How the Secret War between the FBI and CIA Has Endangered National Security (1994, reissued in 2002 with a new epilogue). Riebling’s thesis is that the problem is longstanding, that it has a single “root cause,” and that this root cause is institutional. In his telling, a full half-century’s worth of national disasters—from Pearl Harbor through the Bay of Pigs, the Kennedy assassination, Watergate, Iran-Contra, and 9/11—can be traced directly to intelligence failures, and those failures were proximately caused by turf-battling between our two great rival agencies.
This has now become conventional wisdom, accepted on all sides. And one can see the apparent sense in it. A ramified system of multiple agencies having similar missions and chasing the same budget dollars will inevitably produce rivalry; rivalry begets pettiness, and pettiness begets failure. Such, indeed, is the reasoning behind virtually all of the proposals now under consideration by no fewer than seven assorted congressional committees, internal evaluators, and blue-ribbon panels charged with remedying the situation.
One proposed fix, supported by, among others, Senator John Edwards and James B. Steinberg, a deputy national security adviser in the Clinton administration, would create a new entity, analogous to Britain’s MI-5, to assume the FBI’s domestic-intelligence mission. Decoupling that agency’s information-gathering from its law-enforcement duties would allegedly result in a specialist agency that would more resemble, and be less likely to rumble with, its foreign-intelligence counterpart, the CIA. These hoped-for efficiencies would, it is (naively) supposed, compensate for the loss of the FBI’s critical power to leverage intelligence-gathering with the ready hammer of prosecution.
Steinberg and Senator Dianne Feinstein are also among those who would solve the pitfalls of conflicting bureaucracies by . . . adding another bureaucracy. This new National Intelligence Directorate would oversee the full spectrum of relevant entities, compelling the likes of the CIA, the FBI, the National Security Agency (NSA), the National Geospatial-Intelligence Agency, the Defense Intelligence Agency (DIA), and the State Department’s intelligence branch to play nice with each other. Presumably it would also render obsolete the Terrorist Threat Integration Center, another new entity (under CIA direction) created by President Bush a year ago to promote harmony.
But is it true that inter-agency rivalry is the problem everyone claims it is?
That rivalry exists is indisputable; likewise, that its effects can be pernicious. One of my first encounters with the CIA a decade ago occurred when I and other prosecutors preparing the conspiracy case against the organization responsible for the 1993 World Trade Center (WTC) bombing asked the agency for a much-needed briefing. The CIA was perfectly willing to come to New York for that purpose—but not if our FBI case agents were going to be in the same room.
Nevertheless, like many facts that appall at first blush, internecine warfare is only, at best, half the story. For one thing, intelligence professionals are correct (if occasionally disingenuous) when they complain that the public has a skewed perception of their operations: while catastrophic lapses are always notorious, intelligence successes are more numerous. These, however, must typically be kept secret in order to preserve sources of information and methods of gathering it. The unfortunate result is a portrait of ceaseless “failure” that, aside from giving intelligence-gathering an undeserved bad name, also obscures other verities.
First, day-to-day cooperation among agencies, and particularly between the FBI and CIA, is actually far better than people have been led to believe. In terrorism cases, in the decade after the 1993 WTC bombing, teamwork improved in leaps and bounds. To be sure, there are occasional breakdowns, usually due to personality conflicts. But this is an unavoidable function of the human condition—which no legislation on earth can repeal—and it is just as frequently a factor in intra-agency disputes as in those between agencies. Today, agents who fail to compare notes are generally acting in violation of information-sharing protocols; it is hard to imagine additional directives improving the situation.
Second, intelligence-gathering is not monolithic. Domestic intelligence is radically different from the foreign variety, and both differ critically from the needs of the military. So polysemous an imperative requires a variety of skills to meet widely divergent situations and assumptions. As both a practical and a political matter, it is inconceivable that the task could be accomplished by a single agency, and proposals that suggest otherwise are certain only to reshuffle, rather than eradicate, natural rivalries while damaging the quality and quantity of information collection.
Third, and most misunderstood, rivalry—overall—is a virtue. In the government’s vast monopoly, it is essential. Naturally, the seamy side of competition being a perennial best-seller, the public record is replete with hair-raising anecdotes of sharp-elbowed investigators pursuing the same quarry to the benefit of criminals, enemies, and traitors. On a macro level, however, the throat-cutting is statistically insignificant. As a rule, competition impels agents to test their premises and press for better information; it results in the generation of more leads and the collection and refinement of more intelligence. In a world where the Supreme Court cannot decide a case without amicus briefs from innumerable interested observers, where Congress declines to pass legislation without the input of scores of experts, do we really want the President, in matters of national security, reduced to a single stream of intelligence-collection and analysis?
If turf-battling is not an enormous obstacle, does that mean there are no obstacles? Hardly. The real problems, though, are not bureaucratic but structural and philosophical. They have taken over 40 years to metastasize, and they would take a lot more than cosmetic surgery to reverse, even assuming the national will to do it.
As with much else in our national life, the bacillus now grown to plague America’s intelligence apparatus took root in the unrest of Vietnam and the upheaval of Watergate. The perception of national security became intertwined in those years with an increasingly unpopular war that ended badly. For a generation of activists soon to take up positions of influence in politics, academia, and the media, the antiwar movement inculcated a lasting aversion not only to the exercise of American military power but to the agencies tasked with assessing threats to our national security, not to mention the real-world grunt work of intelligence.
Watergate deepened the aversion. For one thing, the burglars included former intelligence officers. For another, President Richard Nixon enlisted the CIA to obstruct the FBI’s investigation of the break-in. For a third, his White House “enemies” operation featured spying against domestic political adversaries. Hot on the heels of these misdeeds, the CIA became enmeshed in other domestic spying scandals that were subjected to high-profile probes, first by a commission appointed by President Ford and, in 1976, by the celebrated Senate Select Committee chaired by Frank Church.
Perhaps the first consequence of this chain of events was a long-term decline in the authority of the executive branch of government. The decline stemmed from an illogic that often bedevils the aftermath of scandal: the tendency to confound the sins of a corrupt actor (in this case, Nixon) with a structural weakness in the system itself. In the mid-1970’s, the new operating premise was that, since robust presidential power was likely to be corrupted, it must therefore be scrutinized and shackled in every respect.
From this there followed a second consequence: a shift of national-security functions, prominently including intelligence-gathering, from the ambit of broad executive discretion to the area where executive action is regulated by Congress and the federal courts. Compared with the “intelligence failures” decried by journalists and politicians today, this shift engendered a continuing calamity.
In the constitutional license given to executive action, a gaping chasm exists between the realms of law enforcement and national security. In law enforcement, as former U.S. Attorney General William P. Barr explained in congressional testimony last October, government seeks to discipline an errant member of the body politic who has allegedly violated its rules. That member, who may be a citizen, an immigrant with lawful status, or even, in certain situations, an illegal alien, is vested with rights and protections under the U.S. Constitution. Courts are imposed as a bulwark against suspect executive action; presumptions exist in favor of privacy and innocence; and defendants and other subjects of investigation enjoy the assistance of counsel, whose basic job is to thwart government efforts to obtain information. The line drawn here is that it is preferable for the government to fail than for an innocent person to be wrongly convicted or otherwise deprived of his rights.
Not so the realm of national security, where government confronts a host of sovereign states and sub-national entities (particularly terrorist organizations) claiming the right to use force. Here the executive is not enforcing American law against a suspected criminal but exercising national-defense powers to protect against external threats. Foreign hostile operatives acting from without and within are not vested with rights under the American Constitution. The galvanizing national concern in this realm is to defeat the enemy, and as Barr puts it, “preserve the very foundation of all our civil liberties.” The line drawn here is that government cannot be permitted to fail.
For these reasons, prior to the post-Vietnam, post-Watergate revolution, executive-branch authority in matters of national security had been almost plenary. The constitutional checks held by Congress were largely trifles. The power to declare war was already nearly an anachronism—during the Civil War, the Supreme Court had ruled that, regardless of whether Congress acts, Article II of the Constitution actually obliges the President to respond with all necessary force to put down attacks against the United States. Even Congress’s power of the purse lacked much practical muscle, given the inherent political risk for a legislator who dared to withhold funds the President said were vital to national security.
In line with this, the executive branch had wide latitude to gather intelligence against potential threats. True, the CIA’s charter did not permit it to conduct domestic intelligence-gathering—that task being left to the FBI—but this affected only which arms of the executive branch could spy on our enemies in which venues. It did not, at least in theory, affect the substance of the information to be gathered.
But cataclysmic changes were ahead, and their harbinger was President Jimmy Carter’s acquiescence in the 1978 Foreign Intelligence Surveillance Act (FISA). Here, for the first time, Congress and the courts undertook to regulate the gathering of national intelligence, particularly by electronic eavesdropping, against agents of hostile foreign powers. In the Nixonian afterclap, it was adjudged that the executive could not be trusted unilaterally to wield this power, which might secretly be used against political opponents.
Of course, such wiretapping was already illegal, and the Nixon experience had amply demonstrated the political price to be paid for engaging in it. No matter. Henceforth, the executive branch would not be allowed to use whatever tactics it, as the branch with the most expertise and information, determined were necessary to protect the nation. Rather, it would be compelled to go to a federal FISA court newly created for the purpose, and, as with the procedure for criminal wiretaps, it would need to establish probable cause that the target was an agent of a foreign power. Electronic surveillance would be permitted only if the judges approved.
The impact on intelligence collection was serious. Previously, it would have been laughable to suggest that foreign enemy operatives had a right to conduct their perfidies in privacy—the Fourth Amendment prohibits only “unreasonable” searches, and there is nothing unreasonable about searching or recording people who threaten national security. (The federal courts have often recognized that the Constitution is not a suicide pact.) Now, such operatives became the beneficiaries of precisely such protection. Placing so severe a roadblock in the way of a crucial investigative technique necessarily meant both that the technique would be used less frequently (thereby reducing the quantity and quality of valuable intelligence) and that investigative resources would have to be diverted from intelligence-collection to the rigors of compliance with judicial procedures (which are cumbersome).
This was only the start of the debacle. Courts and the organized defense bar soon began to ply the FISA statute with hypothetical governmental abuses. What if, they worried, a national-security wiretap yielded evidence of an ordinary crime—not an unlikely event, given that terrorists tend to commit lots of ordinary crimes, including money laundering, identity fraud, etc. This was no problem under FISA as written: intelligence agents could simply pass the information to agents of the criminal law, who could then use the damning conversations in court. But what if such law-enforcement agents, for their part, were to try to use FISA as a pretext to investigate crimes for which they themselves lacked probable cause to secure a regular criminal wiretap?
In one sense, the suggestion was not out of line—wiretap conversations are devastating evidence, and defense lawyers routinely strain to have them suppressed. But the notion was logically absurd. If a criminal investigator was going to act corruptly, it would be far easier for him to fabricate evidence showing probable cause for a regular wiretap (by pretending, for example, to have an anonymous source who had bought illegal drugs from the target) than to trump up a national-security angle necessitating an additional set of internal approvals. Nor was there any indication that such chicanery was actually afoot. But reality is rarely an obstacle for those who see life as an ongoing law-school seminar. Gradually, courts rewrote FISA, grafting onto it a so-called “primary purpose” test requiring the government to establish not only probable cause that it was targeting operatives of a foreign power but also that its real reason for seeking surveillance was counterintelligence, not criminal prosecution.
As one would expect, this created among many prosecutors a grave apprehension about “the appearance of impropriety”—a hidebound concept governing lawyer ethics that is perfectly nonsensical in the life-and-death context of national security. Even as militant Islam began its terrorist war against the United States with the 1993 WTC bombing and the 1994-95 “Bojenka” plot to blow a dozen American airliners out of the sky over the Pacific, the Justice Department was worrying that agents and prosecutors might be perceived to be using intelligence-gathering authority to build criminal prosecutions. Often, the result was weeks or more of delay, during which identified terrorists who happened also to be committing quotidian crimes went unmonitored while the government dithered over whether to employ FISA or the criminal wiretap law. The insanity reached its apex in 1995 with the “primary purpose” guidelines drafted by the Clinton administration: henceforth, a firewall would be placed between criminal and national-security agents, generally barring them even from communicating with one another.
The damage from the firewall and the impediments to FISA has been incalculable. It took ten years to make the racketeering case against Sami al-Arian, the professor accused of helping run the murderous Palestinian Islamic Jihad from the campus of South Florida University, because the wealth of information collected by intelligence agents was withheld from their criminal counterparts. And that was a pittance compared with what happened in the waning weeks before the September 11 attacks. Zacarias Moussaoui, who had paid cash for pilot training (and was reported to authorities when his bizarre behavior—including intense interest in how cabin and cockpit doors worked—could no longer be ignored), was detained by the immigration service. Worried FBI intelligence agents were desperate to search his computer, but were turned down by supervisors who decided there was insufficient evidence to go to the FISA court. His al-Qaeda membership and numerous connections to the hijackers were not uncovered until after the attacks.
And the Moussaoui travesty itself pales in comparison to the story of Khalid al-Midhar and Nawaf al-Hazmi, excruciatingly recounted in Slate by Stewart Baker, general counsel of the National Security Agency during the early Clinton administration. The pair, who had trained to pilot planes, lived in California. In August 2001, an astute FBI intelligence agent was trying to find them, and asked the criminal division for help. But FBI headquarters stepped in and insisted that the firewall not be breached: criminal agents were to stay out of the intelligence effort. A few weeks later, al-Midhar and al-Hazmi plunged Flight 77 into the Pentagon, their manifold ties to Mohammed Atta and the other hijackers kept safely under wraps.
In attempting to “connect the dots” on how branches of our government erected barricades against efficient information-sharing, one cannot avoid addressing the most basic blunder of all. In the years after World War II, the designers of the CIA conceived of it as, in one sense, an analogue to the American military. Just as the armed forces are generally precluded by law from domestic policing (which is left to the FBI and other federal, state, and local agencies), so the CIA could not conduct its operations within U.S. territory.
The CIA, then, is confined to foreign intelligence and counterintelligence activities. When leads cross into U.S. territory, the FBI takes over—mainly through its foreign-counterintelligence division, which is separate from its law-enforcement side. This division of labor, and not simple rivalry, is the salient reason for the inter-agency warfare of the last half-century.
Turf aside, however, the structure is not analogous to the military doctrine of posse comitatus, which bars the armed forces from domestic policing. For if the United States were invaded by a foreign army, our military would respond; that would be a national-defense function, not policing. Similarly, hostile foreign operatives within the U.S.—plotting, recruiting, providing funding and material support to their principals—fit the mold of an invading foreign army far better than that of a criminal collaborator.
Yet U.S. law and tradition (strenuously supported by many of the same politicians who today bluster about the CIA’s lack of dot-connecting skills) rig intelligence as if it were Russian roulette: the agency whose raison d’être is to counter foreign threats to our national security is precluded from participating in investigations once they cross into our nation, while the agency that is expected to pick up the ball and run with it from there does so without the CIA’s depth of knowledge and expertise.
The ill-conception of this arrangement has become increasingly patent. With the info-tech revolution, al-Qaeda operatives seamlessly share information across borders with the click of a mouse, enabling them instantly to construct a complete picture of their prey. By contrast, the forces charged with keeping us safe from them are expected to complete awkward hand-offs as persons and information roam in and out of the country. The windfall beneficiary is, ironically, the terrorist operative who happens also to be an American citizen. Such an operative is not only protected by the full panoply of constitutional rights wherever in the world he travels but is radioactive to the CIA, which is no less fearful of the perception that it is spying on Americans than the Justice Department was about the appearance of misusing FISA.
It is bad enough that, prior to 9/11, terrorists could easily survive in the lacunae of our domestic intelligence apparatus. Worse, they positively thrived on the way it operated.
Throughout the eight years of the Clinton administration, as militant Islam’s jihad against America escalated, the federal courts became the linchpin of counterterror strategy. This began understandably enough. The 1993 WTC bombing was viewed as a domestic crime. Although, years later, investigators and journalists would link the bombing to al Qaeda, and al Qaeda in turn to prior terrorist acts against the U.S., at the time not much was known about Osama bin Laden, his network, and his national support systems in Afghanistan and Sudan. No one credibly could fault President Clinton for handling the matter as a court case or for not responding militarily. As the murder and mayhem grew, however, and as it became clearer that indictments were a pusillanimous response to suicide bombers geared to obliterate American embassies and naval destroyers, Clinton stayed the self-defeating course.
As Defense Secretary Donald Rumsfeld has observed, weakness is provocative. The fecklessness of meeting terrorist attacks with court proceedings—trials that take years to prepare and months to present, and that, even when successful, neutralize only an infinitesimal percentage of the actual terrorist population—emboldened bin Laden. But just as hurtful was the government’s promotion of terrorism trials in the first place. They were a useful vehicle if the strategic object was to orchestrate an appearance of justice being done. As a national-security strategy, they were suicidal, providing terrorists with a banquet of information they could never have dreamed of acquiring on their own.
Under discovery rules that apply to American criminal proceedings, the government is required to provide to accused persons any information in its possession that can be deemed “material to the preparation of the defense” or that is even arguably exculpatory. The more broadly indictments are drawn (and terrorism indictments tend to be among the broadest), the greater the trove of revelation. In addition, the government must disclose all prior statements made by witnesses it calls (and, often, witnesses it does not call).
This is a staggering quantum of information, certain to illuminate not only what the government knows about terrorist organizations but the intelligence agencies’ methods and sources for obtaining that information. When, moreover, there is any dispute about whether a sensitive piece of information needs to be disclosed, the decision ends up being made by a judge on the basis of what a fair trial dictates, rather than by the executive branch on the basis of what public safely demands.
It is true that this mountain of intelligence is routinely surrendered along with appropriate judicial warnings: defendants may use it only in preparing for trial, and may not disseminate it for other purposes. Unfortunately, people who commit mass murder tend not to be terribly concerned about violating court orders (or, for that matter, about being hauled into court at all).
In 1995, just before trying the blind sheik (Omar Abdel Rahman) and eleven others, I duly complied with discovery law by writing a letter to the defense counsel listing 200 names of people who might be alleged as unindicted co-conspirators—i.e., people who were on the government’s radar screen but whom there was insufficient evidence to charge. Six years later, my letter turned up as evidence in the trial of those who bombed our embassies in Africa. It seems that, within days of my having sent it, the letter had found its way to Sudan and was in the hands of bin Laden (who was on the list), having been fetched for him by an al-Qaeda operative who had gotten it from one of his associates.
Intelligence is dynamic. Over time, foreign terrorists and spies inevitably learn our tactics and adapt: consequently, we must refine and change those tactics. When we purposely tell them what we know—for what is blithely assumed to be the greater good of ensuring they get the same kind of fair trials as insider traders and tax cheats—we enable them not only to close the knowledge gap but to gain immense insight into our technological capacities, how our agencies think, and what our future moves are likely to be.
In considering the asserted “intelligence failures” of September 11 and beyond, it is worth bearing in mind this information bounty, which our government consciously decided to provide from 1993 through 2001 even as it was increasingly manifest that the enemy was growing more proficient, its attacks more deadly.
Although I have thus far been concentrating on the collection and analysis of intelligence here at home, a similar and complementary history can be constructed for what happened to our capabilities overseas. There, too, our intelligence apparatus was thoroughly compromised.
In particular, the collapse of the Soviet Union in the early 1990’s dovetailed with a severe economic recession that ultimately cost George H. W. Bush his presidency. For the CIA, this constellation of circumstances had two major, detrimental consequences.
First, desperate to cut spending wherever politically palatable, the federal government declared a “peace dividend.” This was a fantasy. Although the fall of Soviet tyranny was an enormous blessing, it also presaged a more challenging international environment, filled with threats diffuse, unconventional, and less predictable. Nevertheless, at the urging of many of the same elected officials now complaining about failure, including Senator John F. Kerry, intelligence spending was repeatedly slashed.
The second nightmare for the CIA was President Clinton. For the first President Bush, himself a former CIA director, intelligence had been a priority. For Clinton, it was a nettlesome chore—and one he largely avoided. Clinton had no time even for James Woolsey his own chosen director of Central Intelligence, declining to hold a single one-on-one meeting during Woolsey’s maddening two-year tenure. This freeze-out had the predictable effects: agency morale plummeted, officers abandoned ship, and Congress’s funding door slammed shut.
Human intelligence also fell into disrepair, having already fallen into disrepute. It is worth considering that almost all the terrorism prosecutions of the 1990’s took place after successful attacks. We managed to stop exactly two such attacks: the 1994 Bojenka plot against the airliners, and a 1993 conspiracy to bomb New York City landmarks. The former success was due to sheer luck (a fire, started by inept chemical mixing on the part of two terrorists, was detected by an alert Manila police officer), combined with a Pakistani informant who was induced to turn in the ringleader. The latter happened because an informant penetrated the blind sheik’s terror organization, recorded scores of conspiratorial conversations, and permitted agents to catch the plotters in flagrante delicto, stirring explosives. Sadly, that informant had actually infiltrated the group in 1991 but had been deactivated seven months before the 1993 WTC bombing (after which he was reinstated).
One cannot develop the necessary global network of intelligence informants without CIA case officers. As George Tenet, the current director, attested in a recent speech, by the time he took the helm in the fifth year of the Clinton administration the graduating class of case officers was at a historic nadir. As for the agency’s clandestine-services program, Tenet elaborated, that was in such a shambles that it will take until 2009 before it is functioning at an acceptable level.
Meanwhile, abjuring clandestine operatives, Clinton-era intelligence went hi-tech, making extensive use of satellite surveillance and other advances in remote eavesdropping. But with fewer agents to translate and analyze what was gathered, or to follow leads, the effort was ineffectual. Consider: the 1998 embassy bombings in Africa, carried out by an organization we had been focusing on for five years, took several months to plan; ditto the 2000 strike on the U.S.S. Cole (which would have happened eight months earlier, to the U.S.S. The Sullivans, had not the terrorists’ attack boat sunk from the heft of explosives). The attacks of September 11, 2001 were plotted on four continents for well over a year. We did not sniff out any of them.
As the CIA stumbled, the FBI was ascendant, opening a host of new legal-attache offices around the world. Generally speaking, this was a positive development: just as the terrorist threat was exploding, so too was the spread and sophistication of criminal syndicates, making it imperative for law-enforcement agencies to cooperate internationally. But timing is everything. The FBI was spreading its wings just as its most significant cases involved not ordinary crimes but national security.
Some of our best information is obtained from foreign intelligence services. Naturally, those services are much less forthcoming if they think that what they tell us will have to be revealed in court because of U.S. legal rules. Historically, that was not much of a problem when dealing with the CIA; it is, however, always a concern for a country weighing whether to share some sensitive or potentially embarrassing information with the FBI. The Saudis’ infamous obstruction of the FBI’s efforts to investigate the 1996 Khobar Towers bombing is an exquisite example.
In the Clinton years, no matter how many times we were attacked, all the world knew that our approach was to have the FBI build criminal cases. Indeed, Presidential Decision Directive (PDD) 39, issued in June 1995, announced that prosecuting terrorists and extraditing indicted terrorists held overseas were signature priorities of the administration. Nearly three years later, after several other attacks and public declarations of war by bin Laden, Clinton issued a press release that both trumpeted as a ringing success his strategy of having terrorists “apprehended, tried, and given severe prison sentences” and announced a new directive, PDD 62. This purported to “reinforce the mission of the many U.S. agencies charged with roles in defeating terrorism,” including by means of the “apprehension and prosecution of terrorists.” The embassies in Kenya and Tanzania were bombed less than three months later.
The mantra that “9/11 changed everything” is omnipresent. But is it true? It is certainly true in one crucial sense: our national anti-terrorism strategy is no longer to fight bombs and militias with indictments and press releases. The military has reemerged as the spearhead, with law enforcement in an important but subordinate role. The ramifications have already been positive: simply by responding with force to our enemies, we have not just eliminated thousands of terrorists but accumulated volumes of vital intelligence.
But much still needs to change, and the prognosis is not hopeful. For one thing, we speak of intelligence “failures” as if they were current lapses, to be laid at the feet of the poor saps left without a chair just as the music stopped. And we speak about “fixes” without coming to terms with the nature of the problem; until we do, any such fixes will at best be palliatives, and will more likely make things worse.
Take Iraq’s missing weapons of mass destruction. It may yet turn out that these will be found in Iraq itself, or that they were moved or hidden outside the country in the many months between when we first told Saddam Hussein we were coming and when at last we arrived to depose him. Still, for the moment the stubborn fact remains that the government said the WMD were there and they have not been located. Whose intelligence failure is that? Did our intelligence agencies “fail” in 2003, when, according to David Kay, even Saddam’s Republican Guard believed Iraq possessed the weapons? Or did they “fail” in the 1990’s when the government of the United States regarded the CIA, and spying, and human intelligence, and Iraq as one big pain that should just go away?
Hizballah killed well over 200 servicemen in the two Lebanon attacks of 1983. The blind sheik, and bin Laden after him, promised their adherents that a reprise or two of such “operations” would surely induce the Americans to cut and run from the Persian Gulf. Although we did not cut and run, we did stand by as Saddam Hussein put down a revolt we had incited with the materiel we let him keep. When Saddam tried to assassinate the first President Bush and when he expelled the UN inspectors, we lobbed a few missiles at useless targets—just as we did when bin Laden obliterated our embassies in Africa. In response to the Cole bombing, we did nothing.
Bin Laden struck us repeatedly in the eight years leading up to September 11. From the thousands in al Qaeda’s swelling international ranks, we plucked about 40 and indicted them, bathing them in all the rights of American defendants, and arming them with information from our intelligence files to prepare their defenses. One of these, Mohammed Daoud al-‘Owhali, had killed nearly 250 people by helping to drive a car bomb to the entrance of our embassy in Nairobi, and later confessed. Al-‘Owhali was a soldier in a war on America, probably among the most effective ever. He was held not as a prisoner of war but as a criminal defendant, questioned not by the CIA but by FBI agents, who actually tried to give him Miranda warnings. When he was given a civilian trial, a U.S. judge initially ordered his confession suppressed—which would nearly have guaranteed his acquittal—because he had not been advised of his right to have an American defense lawyer present: a right that, since he was in the custody of Kenya, he did not have. The judge later relented, but only after issuing an opinion holding that foreign terrorists who attack America overseas should be accorded the benefits of the constitutional system it is their mission to destroy.
Was September 11 the worst intelligence failure in our country’s history? Or was it, rather, a national failure, the failure of a country that allowed its sense of decency to overwhelm its instinct for survival and that effectively convinced its enemies that they could strike with impunity?
The problem with our intelligence apparatus, to repeat, is that we went on a national nap for over two decades. If an entity is systematically warped and mismanaged for 20 or 30 years—not by a single agency director or American President, but by a philosophy—it cannot be fixed overnight. You cannot wake up on Monday and say, “We need more informants,” and expect to have them embedded and reporting by the close of the business day. If those lobbying for quick fixes to the intelligence mess do not appear to understand this, might it be because they do not want anyone to start probing whose mess it actually is?
This is not to say that the U.S. intelligence apparatus needs fundamental restructuring. In my opinion, it does not. Instead, its primary needs are, first, time to reverse a quarter-century of sloth, and, second, adequate resources to build a new human-intelligence network. Beyond that, a few other things need to happen, but it is here especially that pessimism sets in.
Although there is no need to restructure the CIA and FBI, the division of labor between them must take account of new realities. Without losing the benefits of rivalry, it is imperative to eliminate the structural barriers that, assuming they ever made sense, make none now. In particular, in a national-security investigation, the overriding assumption must be that we are dealing not with potential criminals presumed innocent but with foreign enemies who must be brought to heel. This means that the CIA must be able to follow the trail of its intelligence into the U.S.
In short, I am proposing that the CIA be permitted to work in the United States against those who have been colorably associated with foreign powers, including terrorist groups. A number of safeguards can be put in place to assure Americans that we have not authorized Big Brother to run amok. In addition to requiring that the FBI be given notice and periodic updates, we could mandate that the CIA obtain authorization within 72 hours of the start of domestic surveillance.
My own preference is that this approval come from a responsible executive-branch official rather than from the courts. The FISA model, in my view, violates the principle of separation of powers, gets courts (which have no institutional expertise in, or ready access to, intelligence) into the business of micro-managing national security, discourages agents from pursuing investigations essential to public welfare, and confers upon enemy operatives benefits they should not have. Still, given that FISA is not going away, I would rather have a requirement to obtain FISA court authorization than a continuation of the outdated system in which, while al Qaeda can freely cruise from Peshawar to Peoria, the CIA gets turned away at the border.
Complementing this change, the FBI and the CIA should continue their increasingly effective cooperation outside the United States, with two caveats. The first is that the CIA (and the Defense Department) should be in the lead, the FBI in a secondary role except when the executive branch determines it is in our national interest to extradite to our criminal-justice system a terrorist held by a foreign sovereign. The second is that, the targets in this war being enemy combatants and not criminal suspects, they should not get Miranda warnings, American constitutional protections (except minimal due process, which our government must always accord), or lavish access to our sensitive files. Instead, they should be captured, held for however long active hostilities last, squeezed (humanely) for information, and, if they have violated the laws of war, given military tribunals.
Other commonsense steps to promote competent intelligence-collection were incorporated in the Patriot Act, enacted six weeks after the September 11 attacks. This act, however, has come under blistering assault; so vicious has the campaign been that sensible Democrats like Senator Feinstein and Senator Joseph Biden have been moved to join their voices to those of President Bush and Attorney General John Ashcroft in the act’s defense. But it may be too little, too late: there are now more than a half-dozen proposals making their way through Congress seeking rollbacks or repeal.
The Patriot Act’s intelligence improvements were vital, and nowhere more so than in the area of information-sharing. It dismantled the pernicious FISA firewall that prevented agents from pooling information. It authorized intelligence agents who were conducting FISA surveillance to “consult with federal law-enforcement officers to coordinate efforts to investigate or protect against” terrorism and other hostile acts. In addition, the act made it easier to obtain surveillance authorization, scotching the requirement that agents show that foreign counterintelligence was the “primary purpose” for their application in favor of the less burdensome certification that it was a “significant purpose.”
But it is these crucial improvements that have come under greatest fire. First, in 2002, the FISA court itself took umbrage at Congress’s demolition of the firewall and the (judicially invented) “primary purpose” test. Fortunately, the court’s attempt to reestablish the suicidal status quo ante was blocked. Next, however, an amalgam of libertarian Republicans and anti-Bush Democrats has promised to limit the term of the bill’s crucial provisions to December 31, 2005, when they are currently scheduled to “sunset” unless extended or made permanent by new legislation.
This bipartisan Senate cabal (led by Democrats Patrick Leahy, Richard Durbin, and Harry Reid and Republicans Larry Craig and John Sununu) wants not only to terminate the FISA sharing provisions but to end the sharing of grand-jury information; to restrict the information that intelligence agencies may obtain from communications-service providers (the same kind of information long available to criminal investigators probing health-care fraud and gambling); and effectively to destroy the valuable “sneak-and-peak” search warrant (another longstanding tool in ordinary criminal investigations) that allows agents, with court approval, to search a location for intelligence purposes but not to seize anything, thus keeping the targets unaware. No doubt, the next time something goes boom, these Senators and their myriad sympathizers will be among the first to wail about unconnected dots.
A political class that appreciated the stakes involved would not indulge in this sort of recklessness. It would not hasten to dub every episodic setback an intelligence failure without asking searchingly whether we have set our agencies up to fail. It would have the necessary perseverance, through the inevitable torrent of catcalling, to retrace a quarter-century of missteps. And it would construct its remedies on the basis of a correct diagnosis of the disease. Right now, when we need it most, this is not the political class we have.