Politics & Policy

What About The Wall?

Commissioner Gorelick is not the right person to probe intelligence lapses.

For those of us who were in the trenches of the struggle against militant Islam beginning in the early 1990s, it is jarring to hear, of all people, Jamie Gorelick–now a member of the 9/11 Commission–hectoring government officials about their asserted failure to perceive how essential it is that the right pieces of intelligence get into the right hands. Equally bracing is to read the account of Gorelick’s star witness, former counterterrorism coordinator Richard Clarke–that hero of the last 15 minutes–who bemoans how, even though he “had asked to know if a sparrow fell from a tree” during the summer 2001, the FBI and CIA nonetheless failed to stitch together disconnected bits of information about al Qaeda operatives and flight schools.

No one in his right mind could say that intelligence breakdowns related to 9/11 are not worth exploring. At issue, though, is the proper explorer. One would have hoped the appearance of objectivity, never mind the reality, would be the 9/11 Commission’s guiding compass. Instead, the panel is beset by a gargantuan conflict of interest–and it’s starting to show.


Commissioner Gorelick, as deputy attorney general–the number two official in the Department of Justice–for three years beginning in 1994, was an architect of the government’s self-imposed procedural wall, intentionally erected to prevent intelligence agents from pooling information with their law-enforcement counterparts. That is not partisan carping. That is a matter of objective fact. That wall was not only a deliberate and unnecessary impediment to information sharing; it bred a culture of intelligence dysfunction. It told national-security agents in the field that there were other values, higher interests, that transcended connecting the dots and getting it right. It set them up to fail. To hear Gorelick lecture witnesses about intelligence lapses is breathtaking.

And Clarke, as counterterrorism coordinator, may have had fascinating ideas about getting at al Qaeda through the Uzbeks and the Northern Alliance, but he didn’t do a damn thing to highlight for the incoming Bush administration this loony construct that discouraged agents with potentially vital evidence from sharing it with their colleagues across the hall. Is it any wonder that the commission now finds itself fiddling over such minutia as whether “battle stations” should have meant regular meetings of the “principals” rather than the “deputies” during summer 2001 when, in reality, Rome had started burning years before.


To be sure, it would be heavy-handed to lay the entire culture of intelligence dysfunction at the feet of the Clinton administration. If we are to play the blame game, there is plenty to go around. The relevant history in fact goes back to the 1978 enactment of the Foreign Intelligence Surveillance Act (FISA) (50 U.S.C. §§ 1801-1862). FISA, a reaction to domestic-intelligence abuses during the Vietnam and Watergate eras, permits the federal courts to regulate and monitor the executive branch’s execution of two crucial investigative techniques–electronic surveillance and physical searches–in the context of national-security (or “intelligence”) investigations. This is in contrast to ordinary criminal investigations, where the use of those techniques is governed by other federal law–what is called “Title III” for wiretaps or bugs, and the criminal procedure rules that govern searches.

In ordinary criminal investigations, agents must present probable cause of a crime to get a judicial warrant to eavesdrop electronically or conduct a search. FISA, on the other hand, is not principally about criminal investigations. It is about national defense power. It focuses on foreign enemies, including international terrorists, who pose threats to the United States. To obtain FISA wiretap or search authority, the government need not show probable cause that a crime is being committed. Rather, FISA permits the government to “obtain foreign intelligence information” if “there is probable cause to believe that…the target of the electronic surveillance is a foreign power or an agent of a foreign power[.]“

The difficulty here is that the theoretical divide between what is a “criminal” matter and what is a national-security “intelligence” matter does not track reality. Espionage, for example, is both a dire national-security issue and a serious crime. Similarly, terrorists commit many crimes (immigration fraud, identity theft, money laundering, seditious conspiracy, possession of precursor explosives, and bombing, to name just a few) in the course of undermining national security. Thus, whether an agent’s investigative authority comes from FISA or the criminal law, he is in either event very likely to collect evidence that constitutes both national-security intelligence and proof of everyday crimes.

Far from being a problem, this should be a win-win situation for public safety. But it has been a problem because, in the quarter-century or so prior to 9/11, our nation regarded terrorism as primarily a law-enforcement problem. Here again, it is important to acknowledge that President Clinton did not invent this approach, though it may fairly be argued that he stayed wedded to it too long, to the exclusion of other, more effective counterterror tactics.

When you are in the national-security mindset, the survival of the country is at the forefront. In that stark place, it becomes easy to see and do what must be done to promote public safety–like making sure the right hand of intelligence knows what the left hand of law enforcement is up to. The mindset of law enforcement, however, is worlds apart, and in a very fundamental way–one of those things so patent it is usually missed.

Even the most serious criminals–mafia chieftains, serial murderers, drug gangs–do not pose any real threat to the survival of the United States. To put it in gamblers’ parlance, in the law-enforcement thicket, we are playing with the house money. We know most of the crooks will get convicted in a system that ensures due process, so the country is not at mortal risk if we presume them innocent and accord them presumptions in favor of the privacy of their persons, their effects, and their conversations. In the sphere of law enforcement, we have the luxury of interposing skeptical judges as a safeguard against prosecutorial overreach, of designing a system exceedingly deferential to civil liberties, and of being more exercised that a single innocent person might be railroaded than about the occasional miscreants who inevitably escape justice.


Prior to September 11, however, we imposed this law-enforcement mindset on the national-security framework–that is, we conducted ourselves as if we were playing with the house money when it was our living wage that was actually at stake. For present purposes, the relevant manifestation of this delusion involves hand wringing over whether the government should be allowed to use FISA-derived evidence in criminal prosecutions.

As written, FISA poses no obstacle to such use. This is significant not only because, as we’ve already seen, terrorists commit lots of ordinary crimes, but also because the ability of the government to threaten or carry out prosecution is among its most important tools for convincing terrorists and other foreign agents to cooperate and provide information.

FISA does, however, require that a national-security official in the executive branch–typically the FBI director–certify that “the purpose” of the surveillance is to obtain foreign-intelligence information (as opposed to a purpose to build a criminal prosecution). This is simply an explanation of why we are doing what we are doing. The requirement does not purport to restrict either the scope of the investigation or the permissible uses of any resulting evidence. Nor is it peculiar to FISA law. When, for example, criminal agents get a wiretap based on probable cause that the subjects are selling drugs, but then learn in the course of listening that the subjects are also committing murders, the murder conversations are routinely used in court to prosecute the murders. It doesn’t matter that the agents weren’t investigating murder in the first place; they had a legitimate authorization to be listening, the murder evidence was thus properly collected, and the public interest in bringing murderers to justice plainly outweighs whatever privacy interest drug dealers have in being able to plan other crimes.

It is, of course, a reality that in the law-enforcement realm, because it is hopelessly lawyerized and hyper-vigilant about civil liberties, we actually are forced to worry about such things as the privacy rights of drug dealers. Thus, we do find ourselves litigating ad nauseam such commonsense matters as whether, just because you properly heard something, you should be allowed to use it to protect the public. When you’re playing with the house money, you have the luxury of contemplating your navel and wondering why the sky is blue. Not so when national security is at stake. Unfortunately, however, Justice Department lawyers and federal judges, over two decades, came reflexively to treat FISA as if it were a creature of the law-enforcement realm they know so well, rather than a vital tool for the life-and-death of national security.

The shift began with the afore-described certification requirement. Soon after FISA took effect, the Justice Department began construing it not as a mere announcement of purpose but rather as something much more restrictive: as a substantive limitation on the use of FISA evidence in criminal cases. This, too, was no fault of the Clinton Justice Department; as the FISA Court of Review explained in a historic 2002 opinion (available here) this happened gradually during the 1980s.

As the Review Court opinion elaborated, over time the certification gaffe led to a “false dichotomy” in FISA derived information, sorting it into two supposedly distinct categories: mere intelligence, and information that was evidence of ordinary crime. With the government needlessly acting as if it may have done something wrong by acquiring evidence of crime outside the ordinary criminal means, it should come as no surprise that the federal courts, too, soon got into the act of concocting new “safeguards” not found in the text of FISA. Specifically, they created a “primary purpose” test under which FISA-derived evidence could not be used in criminal prosecutions unless the government demonstrated that the primary purpose of its investigation had been to collect intelligence; i.e., that FISA had not been used as a pretext to build a criminal case.

This judicially legislated hurdle was a serious error that the executive branch failed for years to challenge. As the Review Court held in 2002, FISA “as passed by Congress in 1978 clearly did not preclude or limit the government’s use or proposed use of foreign intelligence information, which included evidence of certain kinds of criminal activity, in a criminal prosecution.” (Emphasis in original).


Nevertheless, flashing back to when the courts were still warping FISA, two significant developments occurred: first, in January 1993, the Clinton administration came to power; and second, a month later, militant Islam announced its war against the United States by bombing the World Trade Center and rapidly following that atrocity up with two unsuccessful schemes–a Spring 1993 plot to bomb the United Nations, the Lincoln and Holland Tunnels, and the FBI’s Manhattan headquarters; and the 1994-95 Manila Air conspiracy to blow U.S. airliners out of the sky over the Pacific.

On this score, one can only shake his head at Dick Clarke’s celebrated conclusion that we had been fairly successful in the fight against terrorism prior to 9/11 since “only” a few dozen Americans had been killed in the preceding eight years. The 1993 WTC bombing occurred at high noon on a weekday–a time when the complex was generally populated by between 60,000 and 120,000 people. That the death toll was less than ten rather than in the thousands or tens of thousands–as it would surely have been had the bomb-laden van been parked just a few yards differently–was nothing short of a miracle. The New York City landmarks’ bombing conspiracy was thwarted only because the FBI convinced an informant (who had been cut out of the investigation months before the WTC bombing) to go back to work. The failure of Manila Air was sheer luck–incompetent bomb mixing caused a minor explosion that was detected by an alert Filipino police officer. (And I haven’t even gotten to the 1998 embassy bombings that killed over 250 people–mostly Kenyans and Tanzanians–or the 2000 Cole bombing that killed 17 sailors.)

In any event, by 1995 it should have been clear even to government lawyers that we no longer had the luxury of pretending that we were playing with the house money–of pretending that we could safely saddle national-security tools like FISA with conventional law-enforcement restrictions. The time, instead, was ripe for a searching analysis of whether FISA was constitutional, whether it still made sense, and, at the very least, whether self-inflicted wounds like limitations on FISA evidence and unsustainable creations like the “primary purpose” test should be ignored, discarded, or challenged.

But the Justice Department, with Deputy Attorney General Jamie Gorelick in the thick of important policy decisions, did not see it that way. Committed to the bitter end to the law-enforcement mindset, and overwrought at the mere possibility of violating the ill-conceived “primary purpose” test, DOJ made matters significantly worse. It imposed severe procedural barriers against competent intelligence gathering. As described by the FISA Court of Review in 2002:

[T]he 1995 Procedures limited contacts between the FBI and [DOJ’s] Criminal Division in cases where FISA surveillance or searches were being conducted by the FBI for foreign intelligence (FI) or foreign counterintelligence (FCI) purposes. . . . The procedures state that “the FBI and Criminal Division should ensure that advice intended to preserve the option of a criminal prosecution does not inadvertently result in either the fact or the appearance of the Criminal Division’s directing or controlling the FI or FCI investigation toward law enforcement objectives.” 1995 Procedures at 2, 6 (emphasis added). Although these procedures provided for significant information sharing and coordination between criminal and FI or FCI investigations, based at least in part on the “directing or controlling” language, they eventually came to be narrowly interpreted within the Department of Justice, and most particularly by [the Justice Department’s Office of Intelligence Policy Review (OIPR)], as requiring OIPR to act as a “wall” to prevent the FBI intelligence officials from communicating with the Criminal Division regarding ongoing FI or FCI investigations. . . . Thus, the focus became the nature of the underlying investigation, rather than the general purpose of the surveillance. Once prosecution of the target was being considered, the procedures, as interpreted by OIPR in light of the case law, prevented the Criminal Division from providing any meaningful advice to the FBI.

(Italics mine except where otherwise indicated.)

It is important here to remember that most FBI agents are not practicing lawyers. Yes, as the Review Court noted, the regulations left some daylight for cooperation between criminal and intelligence agents. In practice, however, things were much worse: Agents basically understood that criminal and intelligence investigators were not allowed to compare notes, period. That is, the problem is not simply the wide range of perfectly appropriate communications that the regulations expressly prohibited, but is magnified by the ethos the regulations instilled: While justifying communications required a series of mental gymnastics, the agents knew that they could not get in trouble if they just avoided communications entirely.


In this light, Clarke’s sibilant hissing about the failure of agencies to connect the dots or pass information upward in the chain of command rings especially hollow (See Against All Enemies, pp. 236-37). Under the regulations that were, as he well knows, in effect throughout his tenure as counterterrorism coordinator, it was generally impermissible for an intelligence agent to communicate with a criminal agent across the room, much less to press a comprehensive picture upward to ultimate decision-makers. Did Clarke ever think of dismantling the wall? Evidently not. Here’s what he writes on the subject: “The Justice Department normally reviewed FBI requests for national-security wiretaps with a skeptical eye. Justice correctly wanted to insure there were no abuses, lest Congress restrict their ability to do any electronic surveillance under the Foreign Intelligence Surveillance Act” (Id. at 212) (italics mine).

I don’t know whether Clarke has any legal training. I don’t know what he supposes an “abuse” would be. I don’t know whether he actually comprehended that there were abundant separation-of-powers reasons to question whether FISA, even as permissibly written, is constitutional, much less whether Congress could or would attempt to impose further restrictions at a time of terrorist peril. But it is manifest that he knew about the wall, knew it was self-imposed, knew it could be raised or lowered based on Justice’s perception of the threat environment at any given time, and therefore knew it was possible to argue for the wall’s eradication. He doesn’t claim to have done that. Instead, it appears he was a supporter of the notion that, in matters of national security, the FBI–which he now says was not aggressive enough–needed to be reined in by DOJ.

Clarke blames incompetence and apathy for the failure of our agencies to appreciate the significance of pre-9/11 information that our country had been infiltrated by two identified al Qaeda terrorists, Khalid al-Mihdhar and Nawaz al-Hazmi, who would eventually guide Flight 77 into the Pentagon. A different take is offered by Stewart Baker, general counsel of the National Security Agency in the early years of the Clinton administration. In a courageous and forthright account published by Slate in December 2003, and aptly entitled “Wall Nuts,” Baker surmises that “on Sept. 11, 2001, that wall probably cost us 3,000 American lives.”

Baker recounts that an FBI intelligence agent who was trying to find the two terrorists during summer 2001, asked for help from the law-enforcement side of the house and was turned down flat by headquarters. The agent’s responsive pre-9/11 e-mail, quoted by Baker, is chilling: “[S]ome day someone will die–and wall or not–the public will not understand why we were not more effective and throwing every resource we had at certain ‘problems.’ Let’s hope the [lawyers who gave the advice] will stand behind their decisions then, especially since the biggest threat to us now, UBL [Usama Bin Laden], is getting the most ‘protection.’”

Baker’s conclusion? “We couldn’t find al-Mihdhar and al-Hazmi in August 2001 because we had imposed too many rules designed to protect against privacy abuses that were mainly theoretical. We missed our best chance to save the lives of 3,000 Americans because we spent more effort and imagination guarding against these theoretical privacy abuses than against terrorism.”

It would be the height of unfairness to blame Clinton, Gorelick, Clarke, Baker, or any other single official, agency, or presidential administration for the systemic weaknesses that were in place when the 9/11 attacks occurred. Responsibility for the attacks lies squarely on the shoulders of the menace that is militant Islam. To the extent our intelligence system was compromised, the breadth of that problem was much broader than the FISA wall, predated the Clinton administration, and was not adequately addressed at any time before the hijacking attacks.

But all that said, the FISA wall was a rudimentary blunder, and it is fair to ask whether someone who had a hand in constructing it ought to be playing such a central role in examining why our intelligence failed. The 9/11 Commission should be a constructive, bipartisan effort to find a way forward, not an exercise in the gotcha politics of an election year. The commission’s fact-finding, moreover, should be dispassionate; it should not be freighted with agendas that create incentives–wittingly or not–to maximize or minimize some contributory factors at the expense of others.

In finding the way forward, for example, there is little that the commission could do that would be of greater service to the nation than to urge the essential preservation of Patriot Act information-sharing reforms. The Patriot Act dismantled the FISA wall. It exhorted intelligence and criminal agents to pool information, even going a step better by removing the complementary barriers that prevented criminal agents from sharing grand-jury information with their intelligence counterparts. But, precisely because of the same hyperventilating about theoretical civil-liberties abuses that inspired the wall in the first place, those vital reforms would have been left on Patriot’s cutting-room floor had the act’s sponsors not agreed to sunset provisions. This effectively means the FISA wall will be back on January 1, 2006, if a new law is not enacted to make information sharing permanent–or at least to extend it beyond the end of next year.

This would be a catastrophe. Compared to the FISA wall, issues like the meaning of “battle stations” and Clarke’s proposals for attacking al Qaeda (which even he concedes would not have prevented 9/11) are trifles. The commission needs to recognize this and address it, without fear or favor.

Andrew C. McCarthy is a former chief assistant U.S. attorney who led the 1995 terrorism prosecution against Sheik Omar Abdel Rahman and eleven others. While he consults part-time with the Defense Department, the opinions expressed herein are strictly his own.


The Latest