National Security Intelligence

by Loch K. Johnson

  Key issues of intelligence accountability

  Who should be informed?

  Several prominent issues have dominated the ongoing debate over how legislative oversight should be conducted with respect to the intelligence agencies. The first issue is who should be kept informed on Capitol Hill. Before 1974, the answer was just a few lawmakers on the small intelligence subcommittees of the Armed Services and Appropriations Committees – if indeed the DCI deigned to tell them anything, or if the lawmakers were (unlike Senator John Stennis) willing to listen. Today the “witting circle” on the Hill has widened to include members of SSCI and HPSCI, as well as (on some topics) selected members who serve on the appropriations, armed services, foreign affairs, and judiciary committees – with the top four leaders of Congress (two from each party) also informed on some emergency occasions. Intelligence officials, though, are not beyond playing games with these reporting rules, sometimes attempting to whisper only in the ear of a committee chair – DCI Turner's ploy with Eddie Boland early in their relationship, until Boland began to wonder how much he could really peer into the activities of the intelligence agencies while locked in a bear hug by the DCI.

  On other occasions, intelligence managers may seek to limit their briefings on a finding, or some other key intelligence decision, to a “Gang of Eight” (the top four congressional leaders and the top four SSCI and HPSCI leaders from both parties); or even a “Gang of Four,” by limiting the witting circle to only the leaders of the two Intelligence Committees (or sometimes the phrase is used to refer to the four top congressional leaders).74 The Gang of Four concept is strictly an invention of the executive branch, used from time to time by presidents of both parties since 1980, as during the attempts by the second Bush Administration to inform only a few lawmakers about the NSA metadata program and the CIA's adoption of torture. Nowhere is the Gang of Four provision allowed by statute; and the Gang of Eight is a group, it should be emphasized, that was supposed to be used only in times of emergency. Yet, in 2002, the second Bush Administration briefed – all too briefly, according to recipients – just the Gang of Eight for its questionable use of warrantless wiretaps, an operation revealed by the New York Times in 2005.

  In another bit of trickery, intelligence managers often do their best to make sure that no staff are present for briefings on covert action to the Gang of Eight or the Gang of Four. It is a clever strategy, relying on the fact that members of Congress usually don't have enough time or expertise to question an operation deeply, while experienced staff aides might raise serious objections for the members of Congress to consider. As Kathleen Clark notes, the executive branch will inform the Gang of Eight sans staff support, then claim that it had the support of Congress for an operation – as if eight lawmakers represented all of SSCI and HPSCI, let alone the full Senate and House.75 Sometimes the staff on SSCI and HPSCI are partisan, inexperienced in the procedures of oversight, or have limited knowledge about intelligence. Good staff are vital, because the members are often too busy to address the details of intelligence initiatives advanced by the CIA and the other secret agencies. “The staff has to take the lead,” concludes a former CIA officer who has observed Agency–Hill relations for many years. “You ultimately must choose the right people and pay them well to stay.”76

  Clark has concluded that “the Obama Administration has continued the Bush administration practice of resisting robust intelligence disclosure to Congress.”77 The Obama Administration, for example, sometimes refused to report to Congress on covert actions beyond the Gang of Eight. It bears repeating that the intent of Congress in the 1980 and the 1991 Intelligence Oversight Acts was to allow a Gang of Eight just in emergencies, after which (“in a few days” – interpreted to mean forty-eight hours) the privileged Eight were expected to ensure that the full complement of SSCI and HPSCI members were properly briefed. Here is the standard to which the Congress should return, insisting at the same time that if there is a genuine need for a Gang of Eight briefing, then a few senior and knowledgeable staffers will be included in the session and the entire SSCI and HPSCI membership will be informed that the Gang of Eight provision has been temporarily evoked.

  Sometimes the favorite number of lawmakers to brief on Capitol Hill has been zero – a “Gang of None” – as was clearly the case with Operation CHAOS, COINTELPRO, SHAMROCK, MINARET, and the CIA assassination plots. The longest-serving DCI, Allen Dulles (1953–61), once said that he felt obliged to tell the truth only to one person: the President – if he asked.78 On another occasion Dulles widened the circle by one. “I'll fudge the truth to the oversight committee,” he said, “but I'll tell the chairman the truth – that is, if he wants to know.”79

  In bold contrast, the intent of the 1980 and 1991 Intelligence Oversight Acts was to have all the members of SSCI and HPSCI briefed (along with a few staff in attendance), not just some subset. The Committees, though, must insist on this approach or intelligence managers will sidestep the full-briefing rule.80 Certainly the two oversight committees have demonstrated their trustworthiness over the years, with virtually no major leaks from them – far fewer than the flood from the executive branch. Further, during its long history Congress has never had any “plumbers in the basement” – unlike the executive branch during the Nixon Administration with its Watergate conspirators.81

  A well-known controversy regarding reporting to Congress by a DCI occurred during the tenure of Richard Helms. In 1973, while appearing before a Foreign Relations subcommittee in the Senate, he was asked whether or not the Agency had been involved in secret efforts to overthrow the Allende regime in Chile. He said no, but that was not a truthful answer. When lawmakers complained about this misleading response, the Justice Department tried the DCI for perjury. In plea bargaining, Helms agreed to a misdemeanor charge and was fined $2,000, evading a two-year prison sentence. He continued to believe that he had acted appropriately before the subcommittee, because (in his view) the CIA had not been involved so much in a coup d’état as in an effort to prevent Allende's election in the first place; moreover, Helms maintained, he was simply honoring the secrecy strictures placed on him by President Nixon. Yet the end result was that members of Congress were misled (although some knew the truth from other briefings). Rather than lie to lawmakers, Helms could have said that, as is always the case with allegations about covert action (whether real or imagined), he was unable to discuss such matters in public, but that he would be happy to meet with members in a closed session to set the record straight on Agency operations in Latin America or anywhere else.

  When should reporting to Congress occur?

  Another central issue has been the question of when intelligence managers should inform lawmakers about their activities. Before 1974, the answer was: whenever – and if – the managers felt like it. With respect to the Hughes–Ryan Act, the answer for covert actions evolved into an understanding that the briefing would take place “in a timely fashion,” defined during the floor colloquy that preceded the vote on this law to mean within twenty-four hours. The 1980 Intelligence Oversight Act – short in length, but long in reach – changed the formula powerfully to mean “prior notice” (except in an emergency, which allowed reporting – recall – to just the Gang of Eight for a day or two, after which full briefings to SSCI and HPSCI were expected). Ante facto reporting had won out over ex post facto reporting and, in the process, intelligence oversight acquired much greater strength. Lawmakers were now in a position to object to an operation before it was already under way. For example, HPSCI chairman Peter Hoekstra (R, Michigan) once protested to President George W. Bush directly about a specific covert action and the President reportedly modified the operation based on this criticism.82

  The 1991 Intelligence Oversight Act retained this reporting requirement, but clarified during floor colloquy before the vote on the proposal that the emergency delay provision (which permitted reporting only to the Gang of Eight) could be used by the executive branch for a full forty-ei
ght hours – but no longer. At that point, the Gang of Eight was expected to report to the rest of the SSCI and HPSCI membership.83 During confirmation hearings, the (successful) DCI nominee Robert M. Gates testified that he would strongly consider resigning from office if a delay in reporting to Congress went beyond just a few days at the most.84

  What information should be reported?

  The question of what kind of information should be reported is obviously important, too. Again, before 1974, this decision was left to the discretion of the DCI. The Hughes–Ryan Act then required reporting on important covert actions. The far-reaching Intelligence Oversight Act of 1980 made it clear that Congress wished to be kept informed of “all” important intelligence activities, which remains the gold standard – although one the executive branch has periodically ignored.

  Which agencies are expected to report?

  Further debate has surrounded the topic of which of the sixteen intelligence agencies must report to lawmakers (or seventeen, if one includes the ODNI as a agency, rather than an administrative office). Before 1974, the answer was: the reporting, however infrequent, would be carried out by the CIA, since the DCI was housed in its headquarters building at Langley. After 1974 (Hughes–Ryan), the answer was modified to include the White House along with the CIA – at least when it came to covert actions, which henceforth required a presidential finding. Then the 1980 Oversight Act stressed that all government agencies and other “entities” would be expected to keep the oversight panels on Capitol Hill informed, if they were engaged in intelligence activities. The broad language included the NSC staff, a matter that became controversial during the Reagan Administration when that staff chose to ignore this provision as it launched its super-secret Iran–contra covert actions. The staff (including the NSC's lawyers) argued that the Council was not an official member of the Intelligence Community and, therefore, was immune to the provisions of the 1980 Intelligence Oversight Act – an interpretation never intended back in 1980 and rejected in 1987 by a majority of the lawmakers involved in the Inouye–Hamilton inquiry into the scandal.

  How long should lawmakers serve?

  The question of committee term limits has been of significance, too, in understanding the lack of effective oversight on Capitol Hill. At the time of its creation in 1977, HPSCI limited its members to four terms (eight years), although the chairman and ranking minority member were allowed to stay on for ten years. The theory behind these restrictions is that high turnover among members would keep them from becoming co-opted by the intelligence agencies – a phenomenon all too common on Capitol Hill in other policy domains. Driven by the same theory, SSCI had similar term limits, until 2005, when its members decided that greater longevity among members – and therefore greater experience and knowledge – trumped the earlier concentration on countering co-optation. In recent years, HPSCI also jettisoned term limits. Now intelligence overseers in both chambers of Congress have a longer period of time in which to become genuine experts on the arcane aspects of espionage – or to be co-opted.

  Who should enact legislation?

  The normal passage in Congress of authorization bills, followed by appropriations bills to pay for the policies, has sometimes fallen apart in the intelligence domain. Often the Appropriations Committees strike their own deals with intelligence managers, despite what SSCI and HPSCI have recommended. For example, the Intelligence Committees have been scorching in their criticism of some recent satellite programs, yet the Appropriations Committees – lobbied effectively by intelligence managers in an end run – have funded them anyway.85 More disquieting still is the fact that SSCI and HPSCI, caught up in partisan wrangling, proved unable to pass an authorization bill for several years during the early 2000s because of strong disagreements over policy between Republican and Democratic lawmakers on the panels. This polarization raised serious questions about the ability of SSCI and HPSCI to perform as responsible intelligence overseers. Given this state of affairs on the Intelligence Committees, the Senate leadership assigned the Government Affairs Committee (rather than SSCI) to reform the Intelligence Community after the 9/11 Commission concluded in 2004 that intelligence oversight had become dysfunctional on the Hill – “the low point in the history of the [Intelligence] Committees,” according to a seasoned observer.86 In more recent years, both Committees have been able to pass annual authorization bills, but considerable partisan strife continues within their confines, as dramatically displayed during the wrestling over SSCI's Torture Report. The most obvious remedy to this dilemma is for the congressional leadership to select SSCI and HPSCI chairs and ranking minority members who can get along with one another in a spirit of bipartisanship – no mean feat in this unfortunate and ongoing time of partisan warfare in Washington, DC and across the country.

  Who is in charge?

  Finally, the lines of authority for intelligence on Capitol Hill have been a jumble. Each of the following committees in both houses claim intelligence jurisdiction: Appropriations (money), Foreign Relations and Foreign Affairs (international affairs), Judiciary (FBI and the FISA Court), Armed Services (tactical military intelligence), Homeland Security (domestic security), and, of course, SSCI and HPSCI (which presently have sole authority only over the CIA and the ODNI). This many committees is a sure prescription for confusion, as well as a major surcharge on the time and energy of intelligence managers, be they the DNI or program managers at the agency level. In the old saw: when everyone is in charge, no one is in charge. Intelligence programs can be overlooked by lawmakers because they think one of the other committees is responsible. Zegart reports, for example, that the FBI counterterrorism reforms in the late 1990s fell between the stools of SSCI and HPSCI, on the one hand, and the Judiciary Committees, on the other hand, as each counted on the other to examine these new proposals.87

  The ideal solution would be for SSCI and HPSCI to have more extensive jurisdiction over most aspects of intelligence, aided by the staff of the Government Accountability Office (GAO, an arm of Congress), as well as an outside oversight panel (a proposal discussed in Chapter 6). In conjunction with SSCI and HPSCI, the Judiciary Committees could still review FBI and FISA matters, given their close relationship to domestic legal considerations. On the whole, though, the other committees presently involved in intelligence reviews have more than enough work on their plates already when it comes to government oversight, without the added burden (which they often ignore anyway) of responsibilities for the secret agencies.88 Moreover, several lawmakers on SSCI and HPSCI also serve on the other national security panels in Congress; they can report back to Armed Services and the other committees on intelligence developments as they see fit.

  The placement of more intelligence activities within SSCI and HPSCI would require strong congressional leadership, however, and, some would say, divine intervention as well. Nevertheless, in 1975–77 powerful forces opposed the creation of SSCI and HPSCI in the first place and these forces were defeated by a feeling in the nation and on Capitol Hill that intelligence oversight had to undergo fundamental change. Reform – so badly needed – can happen again, if the will is there.

  The roles played by lawmakers as intelligence supervisors

  These years of experimentation in the United States from 1974 until now about how to maintain a proper balance between efficient spy agencies, on the one hand, and the sanctity of civil liberties, on the other hand, have been turbulent, see-sawing between intensive oversight at times (in reaction to major alarms) but, more often, a lackadaisical approach to standing guard against the abuses of secret power. In dealing with their intelligence oversight duties since 1974, members of Congress have adopted one of four major roles: that of the ostrich, the cheerleader, the lemon-sucker, or the guardian, with fluctuations by some lawmakers between the different roles according to the circumstances and personalities of the time.

  One can imagine other roles that intelligence overseers could assume. During the Aspin–Brown Commission inquiry, for exampl
e, a member of that panel, Senator James Exon (D, Nebraska), rarely bothered even to attend meetings and hearings, let alone read draft reports, interview witnesses, travel to inspect U.S. intelligence programs overseas, or engage in other tasks carried out by his colleagues on the Commission. Exon showed up for the first session; once more midway through the inquiry when a celebrity witness came before the Commission in secret session (former Secretary of State Henry Kissinger); and, finally, when the panel's report was finished and the commissioners were invited to brief President Clinton and Vice President Gore on their findings in the Oval. In terms of role types, one could invent the “prairie dog” just for Exon, for he poked his head out of his underground hole every now and then to see what was going on, only to disappear quickly if it looked like work was on the horizon.