National Security Intelligence

by Loch K. Johnson

  Unhappy that the Senate investigators now had inadvertently been provided with the Panetta Report, some people in the Agency, including Brennan, tried to turn the tables on their Hill tormentors by claiming that the SSCI staff had stolen the document by way of illegal hacking. The tradecraft of the CIA abroad often includes the use of deception and manipulation; these dark arts were now turned against “Sissystan,” as Agency officers had nicknamed SSCI. The Senate Intelligence Committee had become, essentially, just another foreign country targeted for covert action. Even Senator Chambliss, one of the Agency's top cheerleaders in Congress, blanched at these accusations from the D/CIA. “John [Brennan] didn't handle that right,” Chambliss managed to say in mild rebuke.46 Another SSCI member observed at the time: “It's WWIII between the CIA and Senate.”47

  Feinstein wrote to Brennan, reprimanding the Agency for its search into SSCI's computers, which she said might well have violated the separation of powers doctrine that lay at the heart of the Constitution (not to mention the CIA's founding law in 1947, which clearly prohibited Agency operations inside the United States). The Director took weeks to respond. A Senate colleague of Feinstein's told the SSCI Chairwoman that Brennan's intentions on behalf of the CIA were “to intimidate, deflect, and thwart legitimate oversight.”48 Subsequently, Brennan would concede in testimony before the full Committee that a “trust deficit” had opened up between SSCI and the Agency, and he apologized to Feinstein and Chambliss for the hacking transgression.

  The CIA engaged in other forms of harassment against the Committee's investigators, too. Its Acting Chief Counsel went after the SSCI staff with a vengeance, perhaps distressed by the fact that his name had come up more than 1,600 times in the SSCI Torture Report (each of which he managed to have deleted during the White House declassification wars). He referred the names of the SSCI staffers who wrote the Torture Report to the Justice Department for investigation.

  In the light of these activities, the Agency's understanding of and allegiance to the existing laws and rules on accountability became a subject of deep concern on Capitol Hill and in the media. Senate Majority Leader Harry Reed (D, Nevada) warned Feinstein: “Look, you can't stand by anymore! The CIA is leaking stuff. They're making your staff out to be the bad guys!”49 The Agency's efforts to spy on SSCI staffers was an unprecedented violation of the oversight relationship – an operation carried out with impunity, despite the strong, even angry complaint delivered by Senator Feinstein in a press conference.

  Brennan appointed a panel to look into the charges. Three of the five members of this group were CIA officers, and the other two were outside individuals known to be friendly toward the Agency. After a brief “investigation,” the CIA absolved itself of any blame and announced, with its practiced skills of agitprop, that the whole episode had been just a “misunderstanding.”50 In sharp contrast, the CIA's own Inspector General had reported earlier that the Agency's hacking into SSCI computers was improper, as was the effort by the CIA Acting General Counsel to file a crimes report with the Department of Justice against Committee's staffers. The Agency's IG, David Buckley, soon found himself discredited and ostracized inside the CIA. He resigned, as his own organization – the brotherhood – closed ranks against him.

  Despite this opposition from Langley, in 2015 the findings of the SSCI Torture Report significantly influenced the Senate vote of seventy-eight to twenty-one in favor of banning America's use of any interrogation techniques in the future that were not specifically authorized in the list of less draconian methods printed in the U.S. Army Field Manual.

  A Mixed Record

  A former SSCI Chair, Jay Rockefeller (D, West Virginia), summed up the CIA's response to the interrogation inquiry as an “active subversion of meaningful congressional oversight.”51 To some degree, however, SSCI and HPSCI bear blame for allowing the CIA torture and the NSA metadata programs to move forward for so long before they took a closer look at them. Accepting at face value Hayden's “tummy slapping” remarks (as happened, too, with National Security Adviser Robert McFarlane's assurances in the 1980s about the non-existence of the Enterprise during the Iran–contra episode) proved to be a painful mistake. Overseers must learn to be more skeptical and tenacious – using hearings to dig for the truth, with witnesses sworn under oath, as well as resorting to subpoena powers to obtain access to key documents and witnesses – until they are reasonably sure that all the facts are on the table.

  Further, as several CIA directors noted in an op-ed, once they were finally informed by the executive branch about the interrogation activities, SSCI and HPSCI “missed a chance to help shape the program – they couldn't reach a consensus. The executive branch was left to proceed alone, merely keeping the committees informed.”52 Journalist David Ignatius concurs. He accepted the SSCI report as “immensely valuable,” but tagged the oversight committee for never properly addressing “Congress's own failure to oversee these activities more effectively.” Ignatius wondered: “Did the members of Congress push back hard, as we now realize they should have? Did they demand more information and set stricter limits? Did they question details about the interrogation techniques that were being used?” His correct response: “They did not.”53 Legal scholar Michael Glennon asks: “Where was SSCI while all this was going on?”54 He might have added: And where was HPSCI? Neither Committee visited any of the CIA's controversial black sites overseas, for example.

  At the same time, SSCI earns high marks in persevering with its inquiry into the interrogation program, despite great adversity and insufficient cooperation from the executive branch (for instance, the White House, at the CIA's instigation, blocked 9,000 pages of documents from going to SSCI during its investigation, on grounds of executive privilege). The Torture Report is no doubt flawed. It would have been especially good to have more bipartisan participation, at the staff and member levels. The Republicans could have made this happen, but deserted the effort as an early means of trying to slow down and discredit the inquiry. The GOP members would have had plenty of opportunities along the way to insert their dissenting views into the report, setting the record straight when they thought the Democrats were off-mark. In this sense, the Republican strategy did the CIA a great disservice. Only courageous Olympia Snowe (R, New Hampshire) displayed a bipartisan spirit on SSCI and a willingness to search for the truth with her colleagues across the aisle.

  Critics have charged the Democratic staff who worked on the report with allowing their ethical objections toward torture to trump a dispassionate weighing of the program's effectiveness. The staffers deny this, arguing that their purpose was to lay out the facts.55 At any rate, the facts did indicate that the CIA's resort to torture was not very effective, just as the value of the NSA's controversial sigint programs were routinely exaggerated by intelligence officials – and even by Senator Feinstein.56

  Senator Feinstein, the accountability hero throughout this doleful episode, captured the Torture Report's normative sentiments: “We're not Nazi Germany. We don't torture people. We don't waterboard them 183 times until they nearly stop breathing. We don't put them in coffins and attach them to walls for, like, 100 hours.”57 Hayden dismissed her views as “emotional,” but she was not alone in her outrage. Senator McCain for one, himself a victim of torture at the hands of ruthless North Vietnamese interrogators during the war in Vietnam, said that “this question is not about our enemies; it's about us.”58

  A shock theory of intelligence accountability

  As this look at the different phases of intelligence accountability suggests, the oversight performance of SSCI and HPSCI has fluctuated widely during the Cold War and since. Several observers have commented on these ups and downs. Writing about oversight in the years before the Church Committee investigation, for example, Ransom noted that intelligence accountability had been “sporadic, spotty, and essentially uncritical.”59 Even after the introduction of the New Oversight in 1975, students of the subject have discerned little serious attention to th
is responsibility in recent years.60 The chief cause of the inattentiveness derives from the nature of Congress: lawmakers seek re-election as their primary objective and they usually conclude that passing bills and raising campaign funds is a better use of their time than the often tedious review of executive branch programs. This is especially true for intelligence review. The examination of secret operations must take place in closed committee sanctuaries, outside of public view. Absent public awareness, the chances for credit-claiming – vital to a lawmaker's re-election prospects – become difficult.61

  An examination of intelligence accountability in the United States since the Year of Intelligence in 1975 indicates a cyclical pattern of stimulus and response. A major intelligence scandal or failure – a shock – transforms the perfunctory performance of oversight into a burst of intense program scrutiny. This burst is followed by a period of reasonably attentive oversight activities that yields remedial legislation or other reforms designed to curb inappropriate intelligence operations in the future. Then comes the third phase of the shock cycle: a return to a middling practice of oversight. Political scientists McCubbins and Schwartz offer the useful metaphor of “police patrolling” and “firefighting” to highlight these differences in commitment by lawmakers to oversight responsibilities.62 Patrolling consists of steadily checking up on the executive bureaucracy: the shining of a flashlight into darkened windows, jiggling the lock on the door, walking the streets with a keen eye. In contrast, firefighting requires an emergency reaction to a calamity after it occurs. Lawmakers qua firefighters jump on the fire truck when the alarm sounds and try to put out the conflagration. A prominent member of Congress has recently used the policing analogy. “There has been no cop on the beat,” said Representative Henry A. Waxman (D, California), chair of the House Oversight and Government Reform Committee, who accused Republicans of abandoning their oversight responsibilities. “And when there is no cop on the beat, criminals are more willing to engage in crimes.”63

  Sometimes the high-intensity police patrolling that follows firefighting can last for months and, if the original shock was strong enough to produce extended media attention, even years. Once the firestorm (an intelligence failure or scandal) has subsided and reforms are in place, however, lawmakers return to a state of relative inattentiveness to intelligence activities – low-intensity, or perhaps even non-existent, police patrolling. This pattern is depicted in Figure 5.3.

  Figure 5.3 The cycle of intelligence shock and reaction by congressional overseers, 1975–2006

  Source: Loch K. Johnson, “A Shock Theory of Congressional Accountability for Intelligence,” in Loch K. Johnson, ed., Handbook of Intelligence Studies (New York: Routledge, 2007), pp. 343–360; figure at p. 344.

  To reach the level of a shock (or “fire alarm”), an allegation of intelligence failure or impropriety has to have sustained media coverage, with at least a few front-page stories. In 1974, for example, the New York Times had an unusually high run of reporting on the CIA from June through December: some 200 articles. In December alone, nine stories on the Agency made the front page – unprecedented at the time. Here was a steady drumbeat of chiefly negative reports, setting the stage for a strong public – and, therefore, congressional and presidential – reaction to the most explosive of these news items: Operation CHAOS, the CIA domestic spy scandal. On the eve of the next major intelligence scandal, the Iran–contra affair the Times carried eleven front-page stories in both October and November 1986 about possible intelligence abuses related to a covert war in Nicaragua. The number jumped to eighteen front-page stories in December, setting the stage for the Joint Committee investigation into the scandal in 1987.

  Congress has greater authority than the media to investigate intelligence operations – from the power of the purse to the power of the subpoena and, ultimately, impeachment. The media, though, seem to have more will-power to conduct oversight, driven in part by a profit motive to sell newspapers by exposing government scandals and failures. Intense media coverage may not be enough in itself to bring out the firefighters on Capitol Hill. The warrantless wiretaps and the CIA torture practices during the second Bush Administration, for instance, garnered considerable media attention but no major congressional inquiry. The nation continued to feel under the threat of possible terrorist attacks and, in this climate of fear, investigations into these shocking revelations were never pursued. Such considerations as the personalities of congressional overseers – especially the attitudes of the SSCI and HPSCI chairs – and the existence of divided government can play a role, too, in determining whether an intelligence shock leads to a major fire-fighting episode.64

  The intelligence shocks

  Since Congress began to take intelligence oversight seriously, near the end of 1974, lawmakers have devoted about six years of their time to intensive investigating (firefighting), stimulated by five major intelligence controversies or shocks (alarms), presented in the text below and in Table 5.1. The rest of the time – the vast majority – has been spent in police patrolling. Sometimes this patrolling has been vigorous (the Senate Torture Report is a conspicuous example, at least for the Democrats on SSCI); but, for the most part, it has been carried out in a perfunctory manner.

  Table 5.1 Type of stimuli and oversight responses by lawmakers, 1974–2016

  Major Alarm No. 1: The Domestic Spy Scandal (1974)

  The government of the United States responded to news allegations of CIA domestic spying (Operation CHAOS) with the Church, Pike, and Rockefeller inquiries. The findings of these panels led to the creation of SSCI and HPSCI, the FISA warrant requirements of 1978, and the Intelligence Oversight Act of 1980 with its dramatic requirement of prior notice to Congress for all important intelligence operations.65

  Major Alarm No. 2: The Iran–Contra Scandal (1986)

  The Inouye–Hamilton Committee, which examined the Iran–contra allegations, revealed unlawful intelligence operations carried out by the NSC staff and a few CIA officials.66 Its findings led to enactment of the CIA Inspector General Act of 1989, creating an IG with responsibilities to keep Congress informed of Agency improprieties; and the Intelligence Oversight Act of 1991, which clarified covert action definitions and tightened its approval procedures.

  Major Alarm No. 3: The Ames Counterintelligence Failure (1994)

  In response to the discovery of Aldrich Ames's treachery, the Congress created a presidential–congressional panel of inquiry, known as the Aspin–Brown Commission and led by former Secretaries of Defense Les Aspin and Harold Brown. A U.S. intelligence failure on the Horn of Africa, leading to the deaths of several American soldiers, contributed to the sense in Washington that a major investigation into intelligence practices was warranted. The Commission published a report calling for reforms across the board of intelligence activities, but with special emphasis on the need for a stronger DCI and for revealing to the public the annual aggregate intelligence budget figure.67

  Major Alarm No. 4: The 9/11 Attacks (2001)

  The failure of the intelligence agencies to warn the nation about the catastrophic terrorist attacks against the American homeland in 2001 led Congress to form a joint committee of inquiry (the Graham–Goss Committee) and, subsequently, to urge the creation of an even more comprehensive presidential investigative panel (the Kean Commission) to probe further into the disaster.68 Moreover, HPSCI released to the public a critical evaluation of the CIA's humint activities around the world, stressing the lack of good assets in key locations.69

  Major Alarm No. 5: The Absence of WMD in Iraq (2003)

  In light of an incorrect intelligence judgment about the likely presence of WMD in Iraq, expressed in a National Intelligence Estimate of October 2002, Congress supported the creation of a presidential commission on intelligence (the Silberman–Robb Commission) to investigate this analytic failure.70 Moreover, SSCI undertook a parallel but separate probe into the faulty WMD estimate. The Committee focused on the CIA's errors, but elected not to examin
e questions about the poor use of warning intelligence by the Clinton and second Bush White Houses.71

  Major alarms – No. 6, followed by No. 7 – might normally have occurred in the wake of the NSA sigint collection controversies (warrantless wiretaps, metadata) and the CIA torture program, respectively; but the impulses toward intelligence inquiries (firefighting) by lawmakers and the White House were dampened down by the ongoing potential danger of future terrorist attacks. A sense prevailed among many (although not all) key officials in Washington that security considerations would have to supersede matters of liberty and privacy – at least for a while.

  As shown in Table 5.1, Congress embraced eighteen significant intelligence reform initiatives during the time span from 1974 to 2016. Eleven of these stemmed from major investigations (firefighting), and seven occurred outside the context of a response to a major fire alarm (police patrolling). Of these seven, the first made it easier for the Intelligence Community to shelter its secrets during legal proceedings in a judicial court, by permitting in camera (private) review by judges of any classified documents involved in a law suit involving the Intelligence Community. The second, the Intelligence Identities Act of 1982, was the result of a conclusion reached by members of Congress that a law was needed to provide stiff penalties against anyone who revealed, without proper authorization, the name of a U.S. intelligence officer or asset.72 The third consisted of placing limits on Freedom of Information Act (FOIA) requests to the secret agencies in some instances, especially related to CIA covert actions. The fourth and fifth initiatives dealt, respectively, with ensuring that the Intelligence Community reported to Congress any use of U.S. journalists in intelligence operations, and with improving whistle-blowing procedures.73 The final two police-patrolling initiatives (six and seven) contributed to the passage of laws that supported some key features of the NSA metadata program (the FISA Amendments Act of 2008), while, subsequently, also providing new safeguards for privacy from NSA intrusions (the USA Freedom Act of 2015). While these police-patrolling influences were important, they fell short of the more sweeping changes wrought by the five fire-alarm events: such reforms as the Hughes-Ryan Act, FISA, and the Intelligence Oversight Acts of 1980 and 1991. These major reforms – the outcome of specific shocks – highlight a number of fundamental issues related to the effective conduct of intelligence accountability. These issues are explored next.