Fifteenth DCI, Robert Michael Gates

Robert Gates: Preemptive Reform

New world out there. Adjust or die.[1]

The White House announced Robert M. Gates as Webster’s successor in May 1991. Gates, a professional CIA officer and the first analyst to be selected for director, possessed a keen sense of how the White House used intelligence from his tours on the NSC staff. While serving as the agency’s DDI in the 1980s, he had earned a reputation as an activist manager willing to make changes, and he had served as deputy to DCIs Casey and Webster. In the face of congressional momentum favoring intelligence reform, President Bush believed he could trust Gates to deal with the pressures growing for change without sacrificing executive branch equities or needed intelligence capabilities. Bush also had confidence that Gates’s unsuccessful nomination to be DCI in 1987 would not bar Senate approval.

The administration worked with key senators to ensure Gates’s confirmation. Gates was well aware that dealing with Congress might be the most important, and difficult, aspect of his job. Years earlier, like his predecessor William Colby, he had recognized that intelligence professionals needed to answer to Congress as well as to the executive branch, a fact of life some of his colleagues did not want to accept. By 1991, no DCI would have had any other option. In his confirmation hearings, Gates indicated he would be careful to involve Congress in undertaking, in the post-Cold War period, “a not-to-be-missed opportunity to reassess the role, mission, priorities and structure of American intelligence.” He said he would recommend to the president a fundamental review of US intelligence needs to the year 2005 and promised to bring Congress into the process of formulating proposals to respond to those newly defined needs: “The two Intelligence Committees should have the opportunity to participate even before these proposals come before the Congress.”[2]

Gates’s confirmation hearings turned into an unexpectedly protracted affair that delayed his taking office until 6 November 1991. The delay was caused not by his role in Iran-Contra, but rather by complaints from former analysts at CIA that he had “politicized” intelligence while serving as DDI under Casey. Effective rebuttal testimony on his part and by supporters saved the day, and in the end the Senate approved his appointment in a 64 to 33 vote.

Ensuring White House Support

Gates, who was deputy national security adviser when he was appointed DCI, came to the job with his White House working relationships well established. In addition to enjoying the confidence of President George Bush and his national security adviser Brent Scowcroft, Gates had worked successfully with both Secretary of State James Baker and Secretary of Defense Richard Cheney and their principal subordinates during the event-filled first two-plus years of the Bush administration. He had won praise for running the NSC “Deputies Committee” during a period when that forum gained unusual prominence in policymaking and dramatic international developments formed a crucible that forged strong bonds of comity among Bush’s principal national security officials. Bob Woodward’s The Commanders, a book about Bush administration national security policymaking prior to and during the 1991 Persian Gulf War, portrays Gates as an accepted partner of senior administration officials.

Before he left the NSC Staff, Gates put in place a policy foundation to support his reform agenda for intelligence. Starting as early as June 1991, he had drafted a national security policy review document that tasked policymakers with laying out for the Intelligence Community, in a comprehensive and prioritized way, the key countries and issues deserving attention. This NSC action, National Security Review (NSR) 29, was formalized on 15 November 1991, soon after Gates became DCI. The target date for its completion was March 1992, at which time a fresh definition of what intelligence was expected to do in the post-Cold War era was to be delivered to the DCI. A major purpose of NSR-29 was to show that what he would do as DCI was not parochial bureaucratic self-justification but rather linked formally to requirements defined for him by policy officials. “This effort,” he told an audience of CIA employees in December 1991, “will allow us to correlate resources and requirements we can or cannot meet at different budget levels—in essence, to let the customer decide what to do without at different budget levels. This is a monumental, and historic, undertaking.” A statement issued by CIA’s public affairs staff on 5 February 1992 called NSR-29 “the most far-reaching directive to assess future intelligence needs and priorities since 1947.” Gates monitored the progress of the project closely, at one point pressing his successor on the NSC staff to complete the policy review on schedule in order “to stay ahead of Congressional action.”

Gates also told his CIA colleagues in early December that he would not await the final outcome of the NSC exercise before embarking on an accelerated effort “to move boldly toward a very different shape for the Intelligence Community.” His main theme was integration: “I believe the hitherto loose aggregation of the Intelligence Community must become a much more tightly integrated, coordinated and managed entity than in the past. Protection of turf and old thinking must give way to the demands for greater efficiency, more cooperation, less redundancy and duplication, and better use of fewer resources.” He asserted that the president and other senior administration officials were on board with this approach, and he noted change was already under way for intelligence within DOD. “If we don’t make changes,” he threatened, “they will be imposed upon us.”

Change

Gates foreshadowed his approach to leading change in the Intelligence Community in testimony he gave in 1987, when he had first been nominated to be DCI. Then, he cited Casey as a model leader who had demonstrated how a DCI lacking formal authority over intelligence agencies other than CIA could achieve coordination and cooperation throughout the community. Without mentioning DCI Turner, he contrasted this approach with that of “trying to force people to recognize his authority,” which he believed had led to “serious disharmony” in the community. “That would not be my approach,” he said in 1987.[3] Accordingly, in 1991, Gates showed no interest in redefining or strengthening his authorities along “DNI” or any other lines of reform. Instead, he sought to use such authorities as he had to lead a community through consensus, not subordination, showing interest only in gaining the ability to reprogram funds among NFIP accounts and limiting even his most ambitious reforms to staffing and organizational changes that did not intrude on existing bureaucratic equities.

Armed with White House support, Gates was convinced the best course in dealing with top-level intelligence budget and program issues was to work closely with Secretary of Defense Cheney in reaching mutually agreed decisions. Gates was well aware of differences in culture and viewpoint between DOD and CIA officials, and of the differing perspectives among various community leaders. He fostered comity among the community leaders, commenting on occasion with satisfaction about the spirit of cooperation that he felt existed. He continued biweekly meetings with community leaders and held two offsite conferences where more strategic planning took place. This community leadership style was indeed similar to Casey’s. While not hesitating to take decisions on community matters, Gates often previewed his decisions, solicited comments, and avoided confrontational initiatives. Also, as shown in discussions and decisions regarding imagery, Gates demonstrated he was not tied to parochial CIA positions in all cases.

With respect to the community, Gates recorded some private thoughts in notes around the time he took office. He was happy to note the CIA/NSA cooperation achieved under Webster, judging that the working relationship was “much improved on hard issues but rough spots here and there.” He questioned whether the “flailing” NRO needed more direction as it contemplated changes being urged upon it. He saw DIA as “feeling [its] oats” and “moving to expand in every direction,” including pushing some “crazy ideas on HUMINT.” He wondered about the usefulness of the IC Staff and mused regarding whom he might choose as a community deputy. Overall, he viewed the community as a whole as working well, but noted “tensions rising as budget pinches.”

Gates felt that he could handle both his leadership of CIA and his community role effectively. He did not aspire to a Dulles or Casey-like role in overseeing clandestine operations personally, and the changes he wanted within CIA were in behavior and practices, not structure or mission. He solicited ideas from CIA officers with respect to possible changes, and he received some quite thoughtful ones (including some on his community role) to which he responded in detail. In his first days as DCI, Gates attended an offsite meeting of DO leaders to establish a good working relationship with them.

Task Force Whirlwind

Immediately after taking office, Gates established task forces to take up issues of organization and process involving both CIA and the Intelligence Community. He deliberately created a kind of “blitz” atmosphere by issuing numerous personal tasking memorandums (quite a few in fact were dated the same day), often with short deadlines. One list of the task forces shows that he commissioned a “first round” of 14 in his initial months as DCI (almost all of them tasked in November 1991), and then a “second round” of an additional 10 from March 1992 onward. This initiative aimed at establishing an image of dynamic personal leadership, a kind of “hundred days” approach often used by chief executive officers to show that they intend to provide top-down guidance at the outset of their tenure and take a direct hand in shaping events.[4] Gates made it clear that the reports were to be submitted to him, not to a board or committee, and—recalling how his abrupt introduction of change in CIA’s Directorate of Intelligence had provoked negative reactions—he encouraged wide participation and dialogue in the exercise. He circulated some of the reports widely, soliciting employee comments and insisting that each comment come directly to him for review, not through the chain of command. In a number of cases, he penned reactions and appreciation notes to those who had submitted remarks he had found especially useful or thoughtful.

Gates handpicked almost all the people conducting these studies, and he often had a solution in mind for the issues his boards discussed. Indeed, the selection of topics itself already identified the areas he felt needed to be addressed. Gates had always been a student of how the intelligence profession worked in Washington, avidly learning lessons from his assignments both within and outside the Intelligence Community. He had been a staff assistant to senior policy and intelligence officers since the late 1970s, and he had himself occupied senior policy and intelligence positions since the early 1980s. He therefore felt ready to try out ideas that had been in his mind for years.[5] His self-confidence was bolstered by the support he enjoyed from the president and senior administration figures.

Although some of the task forces addressed issues solely or mainly having to do with CIA, those efforts also helped the DCI in his community role. They showed that he was ready to change his home agency as well as push for change elsewhere. Moreover, some of the CIA task forces dealt with issues important for the agency’s role within the community (most importantly, improving its support for military operations), or for basic DCI responsibilities (e.g., avoiding “politicization” of intelligence products). In the case of a study aimed at developing electronic dissemination of finished intelligence, he asked CIA to serve as a pacesetter for the kind of change he wanted to occur community-wide.[6]

Still more task forces addressed community-wide issues. Two advocated improved staffing support for the DCI in carrying out certain fundamental responsibilities (community management and coordinating foreign intelligence relationships), and three suggested new arrangements for the top-level management of collection disciplines (imagery, human sources, and unclassified or “open sources” of information). Each of these was important enough to Gates’s community role to warrant more detailed description in subsequent sections of this chapter. Two others, not further discussed, addressed how the production of NIEs could be enhanced and how to improve warning intelligence. Finally, on a close-hold basis, Gates commissioned parallel studies of how the NFIP might be pared if deep cuts were applied to intelligence spending. This quick-look project was little known outside the groups working on it (it was not included on the standard lists of task forces) because it was a private task to enable Gates to contemplate an unlikely contingency without upsetting community leaders by producing speculation that Gates either anticipated or would willingly accept drastic cuts.[7]

Foreign Intelligence Relationships

Gates moved quickly to buttress his authority in determining community policy regarding foreign intelligence relationships. DCIs for decades had been in charge of ensuring the coordination of US intelligence activities abroad, including the liaison arrangements of all US intelligence agencies. They set policies determining what could be exchanged or discussed and with whom, and CIA chiefs of station in countries around the world acted as the DCI’s agents in ensuring that DCI policies were followed. International military-to-military relationships and signals intelligence organization relationships, however, had been conducted by DOD agencies for many years with little close DCI attention.

The challenge that had arisen while Webster was DCI was that the purposes of much US intelligence liaison and information sharing activity shifted with the end of the Cold War. If, for instance, Poland was no longer ruled by a communist regime nor was a member of an antagonistic military alliance, why should the United States share intelligence about it with its NATO allies? A senior CIA DO officer wrote a memorandum to Gates early in his tenure conveying concern that uncoordinated actions by some US intelligence agencies could be unwise and, if left unchecked, could erode the DCI’s authority in this area of community activity. Gates used this memorandum and tasked DDCI Kerr to suggest what should be done.

In December 1991, Gates accepted Kerr’s recommendation and appointed Ambassador Hugh Montgomery, an OSS veteran who had served abroad with CIA and had also headed the State Department’s Bureau of Intelligence and Research, to be his special assistant for foreign intelligence relationships. The DCI now had a subordinate with a small staff to whom he could look for support in dealing with a range of new issues: protecting US intelligence equities within NATO as the alliance adjusted to the loss of its major adversary, ensuring that enhanced FBI authorities with respect to counterintelligence overseas did not lead to uncoordinated actions, examining how best to provide US intelligence to international organizations involved in peacekeeping or arms control, and devising policies to govern commercial sales abroad of technologies developed originally for US intelligence use (e.g., satellite imaging systems). Gates informed cabinet officials and the president’s national security adviser of his action, noting that he had already gained the approval of the president and the secretaries of state and defense.[8]

Gates’s initiative caused some heartburn in DOD, partly because he used the word “management,” apparently leading some to fear more intrusive DCI action than his authorities warranted. He discussed the issue with Secretary Cheney, and Gates and Cheney issued a jointly signed message to the field explaining the initiative to chiefs of station, military attaches, and signals intelligence officers. The episode showed Gates’s ability to work with the secretary of defense on strengthening an area of DCI authority, in effect using the secretary’s authority to enhance his own.

Congressional Initiatives and White House Concerns

As 1992 opened, Gates signed out the DCI’s annual report to Congress for 1991. He had signed such a report in 1987, when he was acting DCI, informing Congress about the events of 1986, when Casey had been DCI. This time, the events covered were mainly those that had occurred under DCI Webster, such as the progress of the DCI centers, but he included mention of his blitz of NIEs and task forces, underscoring that he and the community were accommodating, even welcoming, the challenges of the post-Cold War world and the need for change. He claimed that the community had in fact been dynamic for some years, not fixed in static fashion only on the Soviet target, implicitly arguing that intelligence reform within the executive branch was a viable alternative to legislated changes. He stressed the need for US intelligence to reduce policymakers’ uncertainty in a changing world, which he argued supported an “overriding need to preserve flexibility in intelligence systems and structures.”

The Bush administration feared that Congress would saddle the executive branch with unwanted and possibly unhelpful new structures and processes and supported Gates’s efforts to preempt legislation through his own reforms. National security adviser Brent Scowcroft wrote to the chairmen of the SSCI and the HPSCI in January 1992 to explain the administration’s position. “Charter legislation,” he wrote, “is not necessary or helpful” and may create “an unfortunate distraction from more important issues.” He conceded that structural changes were necessary but asserted that “it is markedly easier not to have to do [them] by legislation.” No doubt this letter was intended to bolster Gates’s bargaining position in dealing with Congress on legislative initiatives, but it also made clear that an attentive White House would be closely following the new DCI’s efforts.

One of the concerns the White House had about intelligence legislation was that Congress might negotiate separately with different departments and agencies. At one point, Scowcroft voiced his concerns in this regard, and Gates replied that there should be a meeting of legislative aides from around the executive branch to deal with this possibility. Secretary of Defense Cheney feared that mischievous amendments could be added to bills, making it hard for the administration to safeguard even negotiated legislative initiatives worked out cooperatively between the legislative and executive branches, and like Scowcroft he wanted to hold firm against any new intelligence reform laws. Gates stayed attuned to their concerns as he proceeded with the task of negotiating acceptable legislative outcomes. He was more willing than Scowcroft or Cheney to accept a bill that did not contain unacceptable restructuring ideas, but his established relationship of trust with both of them enabled frank discussion of their differences and agreement on legislative strategies that kept the executive branch unified and up to date on the latest developments involving Congress.

In February and March 1992, the HPSCI and the SSCI opened hearings on proposed legislation aimed at “reorganization” of the Intelligence Community. Former DCI Colby, former DDCI Inman, and former NSA Director Odom testified at the initial HPSCI session, presenting varying viewpoints and cautioning that executive branch actions might make legislation unnecessary. At the initial SSCI session, former DCI Schlesinger cautioned senators to “move cautiously and permit the system to evolve—rather than provide a pre-set and untested blueprint.” At the same session, Odom offered personal testimony about how imperfectly he understood the entire intelligence enterprise despite a long career allowing him different perspectives on it, thus cautioning the senators not to believe that their grasp of intelligence allowed them easily to dictate improved structures.

At one SSCI session, former DDCI Frank Carlucci and two generals argued against the need for a new law, noting the excellent working relationship between the new DCI and the secretary of defense. Witnesses did not support the DNI idea, but they devoted some attention to the best formula for providing leadership for DOD’s intelligence empire.[9] The Democratic chairmen of the SSCI and the HPSCI, Senator David Boren (D-OK) and Representative David McCurdy (D-OK), had to deal with Republican colleagues skeptical of the need for legislation, and both stressed that they hoped executive branch actions as well as legislative ones could be taken cooperatively and in coordinated fashion, and even that legislation might not be needed at all.[10]

In fact, Senator Boren, who had bonded with Gates as the two men worked together through the lengthy confirmation process, actively assisted Gates in his community leadership role. Boren played a key role in gaining for Gates a temporary congressional authorization to reprogram funds within the NFIP. Limited to one year and by the requirement that the affected agencies consent to the reprogramming, the step at least allowed some more room for the DCI to operate below the “top-line” budget level in managing community resources. Also, on occasion they conspired to have Boren advocate radical changes so that Gates could present a change he wanted to undertake as a more “reasonable” position that deserved support.

National Security Directive 67

Concerned about the potential for unwanted legislative change but encouraged by the congressional testimony pointing out precisely that danger, Gates finalized his plans for change and began implementing them in the spring of 1992. On 30 March 1992, President Bush signed National Security Directive (NSD) 67, the end result of the NSR-29 process begun in November 1991. It asserted that policy requirements for intelligence support had “changed markedly” and would continue to do so. Attached to the document was a summary statement of the new list of post-Cold War requirements, divided into four levels of priority (“critical 1, 2, and 3,” and “valuable”). They were described as covering the period from 1992 to 2005, and the president approved them as the basis to be used by the DCI and the community for both resource management and production.

In this document, the president also approved the DCI’s recommendations for resource allocation in the coming years and his recommendations “for comprehensive restructuring of the Intelligence Community.” Specified under the latter rubric was the replacement of the IC Staff with a new Community Management Staff (CMS), measures to strengthen the NIC and community management of resources and requirements, improved coordination and management of the four major collection disciplines, the reorganization of the NRO along functional (i.e,. INT, as opposed to agency) lines, and initiatives to improve intelligence support to the needs of the military. “These measures, together,” the president declared in NSD-67, “represent the most dramatic reconfiguration of the Intelligence Community in decades,” and “must be implemented without delay.”

Gates was pleased with the result of the NSR-29/NSD-67 exercise, telling his senior staff at CIA that he never thought when he wrote NSR-29 that it would turn out so well and on time. The characterization of the changes approved as “comprehensive restructuring” and “dramatic reconfiguration” of the Intelligence Community was hyperbole, but it may have helped Gates in preempting additional congressional changes and in garnering him support within the community. All could see that the president was behind his program of change, and the generally worded endorsement seemed to apply to future changes Gates had in train, such as reform of the imagery collection discipline. The DCI’s relationship with the White House was more collegial than had been the case in 1971 when President Nixon attempted to force a larger community role on a reluctant Richard Helms, and DCI Gates was more enthused about his community leadership role than Helms had been 20 years before.

Gates delivered a major “Statement on Change in CIA and the Intelligence Community” on 1 April 1992 in both open and closed congressional sessions. He described the process leading to the presidential directive signed just two days earlier, the task forces he had commissioned, and the historical context for the initiatives he was taking. Just as he had carefully crafted a detailed statement to salvage his confirmation months before, he pulled together a comprehensive argument to prevent unwise reform or radical budget cuts from gaining momentum. The thrust of his remarks, drawn from NSD-67, pressed the case that the world presented American intelligence with a broader and more diverse set of requirements than ever before. The community’s budget would now depend on whether Congress agreed that the new problems were as demanding in terms of resources as the Cold War had been. Gates also signalled his interest in community affairs by selecting VAdm William Studeman, NSA’s innovative director, to be his deputy.

Community Management Staff

Gates was aware of the Childs report done a year earlier, and he received IC Staff comments on that report soon after he became DCI. He also had available to him, of course, the IG report on the IC Staff. Childs, who served as an adviser to Gates and Kerr in Gates’s first months as DCI, suggested abolishing the IC Staff and setting up a much smaller community staff. In his conception, there were two principal functions the DCI needed to have staffed: collection management and resources management. Childs envisioned the DCI’s collection management responsibility being staffed within the NIC by a new deputy and small staff who would develop broad collection strategies. The existing DCI collection committees could be given over to the functional collection discipline managers. The key function of resource management would be the focus of a new DCI community management staff that might have a core of only 30 to 40 officers. It would concentrate on strategic planning and annual program guidance, identifying cross-program issues, resource trade-offs, and alternative solutions to problems for consideration by the DCI and other community leaders. Childs emphasized to Gates that it was “essential that you use the staff as the principal mechanism in carrying out your Community management responsibilities.” Even the existing IC Staff would have been more effective, Childs argued, “had it enjoyed stronger leadership and support from the DCI/DDCI.”

Gates named Childs and Cheney’s intelligence aide, Richard L. Haver, to head a new task force to look at the issue of how he should be staffed for his community role. Gates’s mantra regarding management was simplicity (for the most part, the DCIDs written about the new entities set up during his tenure were concise, spare descriptions). He wanted a lean, new staff aimed principally at assisting his personal leadership. He had always operated with minimal staff support even when he occupied senior positions, and he envisaged Intelligence Community changes that simplified both structure and process. The IC Staff, which employed more than 250 people, seemed to him too much a collection of routinized activities reflective of community consensus than an instrument of change. He wanted a new structure designed from scratch.

Gates already had concluded that the new staff would be much smaller. He believed that the major collection disciplines, the INTs, were best handled if a single individual were responsible for each, and he had determined that the DCI committees responsible for each INT’s requirements and policy should be transferred to an INT program manager. These committees had for many years been part of the DCI’s IC Staff in order to assure DCI control, but Gates believed he would be satisfied with a simpler, top-level requirements system. Indeed, he seemed to believe that would leave him freer to coordinate, or manage, the community at the top level. It would also permit a smaller, more focused DCI community staff since the DCI collection committees made up more than half the size of the existing IC Staff. With “the creation of three vertically integrated collection disciplines with clear responsibilities,” Gates told his CIA audience in his December speech, “I believe a significant restructuring of the Intelligence Community staff toward purely a DCI management and budget role would then be possible. This staff could then provide major assistance in pursuing a far-reaching Intelligence Community division of labor to reduce costs and redundancy and free resources to address new requirements.”

With no need to staff INT-specific collection management at the DCI level, Childs and Haver were free to concentrate on resources management as the primary function of a new DCI community staff. They worked for about three weeks with 10 senior officers from various elements of the Intelligence Community and then outlined the results of their efforts in a memorandum to Gates. They stressed the importance of the connection between the staff and the DCI: “you have to use it—and be seen to use it—as the principal element you rely on for day-to-day staffing and exercise of your Community responsibilities.” To that end, they recommended it be co-located with him at Langley, reversing the Bush/Turner initiative of moving it downtown 15 years earlier. To dispel perceptions of excessive CIA influence over it, they emphasized that the staff had to be “truly representative of the Community,” with a carefully considered “balance and level of individual agency representation.”

Childs and Haver specified for Gates four areas where they felt he needed staff support in his community role: policy, planning and evaluation, requirements management, and program and budget development. Policy logically included foreign intelligence relationships, but they recognized that Gates might wish to keep his recently appointed special assistant for this topic organizationally separate (as indeed he did). Planning and evaluation needed to address strategic goals, systems, military support, and what they called community architecture and infrastructure (affecting information handling, training, or other community-wide programs). They envisaged requirements being managed on three levels: with customers (the NIC would do this), across the collection disciplines (which they assumed would be vertical organizations), and within each collection discipline organization. As for program and budget matters, they urged close cooperation with DOD on NFIP development, and they suggested that existing DCI authorities “could be exercised more aggressively.” They endorsed new statutory authority for the DCI to reprogram funds within the NFIP, and they noted that if the DCI were given full budget authority, then he would need a community comptroller with a sizable staff.

Structurally, Childs and Haver drew an analogy between OSD and the DCI’s office, suggesting the latter be expanded to include the new community management staff, the NIC, and any other elements Gates wanted reporting directly to him. They recommended that the DCI’s chief community officer be designated the Executive Director for the Intelligence Community and that the NFIC be designated an executive committee that would be a principal source of advice and “the top-level means for you to manage the Community.” They suggested some expansion of the NIC (including the appointment of two deputies, one for estimates and one for evaluation), and the retention of the NFIB to help in the preparation of estimates.

Gates adopted the team’s main recommendations, tweaking them a bit to suit his taste and making use of comments he got from officers with whom he had shared the team’s report. He shaped his community staffing into two parts, the new Community Management Staff (CMS) and the NIC. In the most basic terms, CMS handled resource matters, and the NIC handled substantive matters. Several arrangements were made to interleave the functions of senior officers of the two staffs so that substantive tasks were connected to resource management. Gates chose Haver, a protégé of Studeman’s as well as Cheney’s assistant, to head CMS (Fritz Ermarth, who had headed the main evaluation unit in the IC Staff in the mid-1970s, already headed the NIC).

CMS was indeed the lean management staff Gates preferred, numbering fewer than 100 people. Its head was the new Executive Director for Intelligence Community Affairs, or EXDIR/ICA (an analogous position to the CIA’s executive director, a position that Gates resurrected in 1992), assisted by three deputy directors, responsible for resource management, planning, and requirements and evaluation. The requirements deputy headed a new entity created by Gates, the National Intelligence Collection Board (NICB), which consisted of representatives of the collection disciplines and of the main all-source producer organizations, as well as of the main DCI and OSD staffs. This board provided the DCI with a tool by which he could orchestrate coordination among the INT stovepipes and between producers and collectors. Finally, CMS provided staff support to the new community Executive Committee (a community body analogous to CIA’s executive committee), a group of the topmost leaders of the community’s agencies that replaced the NFIC as the top-level board dealing with resource management issues.

Gates also enhanced the community role of the NIC. First, he made the chairman of the NIC a member of the new community executive committee to emphasize the role of consumers’ substantive intelligence information needs in the consideration of resource matters. Second, the NIC chairman was given two deputies, one for estimates and one for evaluation. The one for estimates assisted him in the continuing traditional task of formulating NIEs, and the one for evaluation, who had a small staff, strengthened his ability to play a role in guiding collection as well as production activities within the community. Third, the vice chairman of the NIC for evaluation was made chair of a National Intelligence Production Board (NIPB) comprising the chief production leaders in the community. This new body replaced the old Intelligence Producers Council chaired by CIA’s associate deputy director for intelligence, thus giving greater community coloration and higher visibility to the DCI’s leadership of analytic work within the community.[11] Fourth, the NIC chairman was made the person to whom the venerable DCI production committees (e.g., the Joint Atomic Energy Intelligence Committee) now reported instead of the head of the IC Staff (another step enabling CMS to be smaller than its predecessor). Finally, Gates announced that he planned to move the NIC downtown in order to stress its community nature and independence, to encourage its external contacts, and to enhance its availability to the policy community. He also gave the NIC the task of providing the executive secretarial support needed by the NFIB, which was retained as the topmost community board dealing with estimates and overall intelligence policy issues (previously the secretariat support to both NFIC and NFIB had been unified).

Gates sorted out lesser issues in short order. He rejected suggestions that he have dedicated community legislative, legal, or public affairs staffs in favor of using the staffs already at CIA for both his community and his CIA needs (thereby keeping his staff organization in these areas simple). He also rejected the task force’s recommendation that he use the term “assistant DCI” for his CMS and NIC heads (one comment he received on this notion warned him the title sounded too much like a position that should require Senate confirmation). He agreed that CMS could have a few additional officers staffing important DCI community responsibilities that did not seem to fit anywhere else. One was a foreign language coordinator, and another was a new open source coordinator, the seed of a fifth INT. Also placed within CMS were an R&D council and the DCI’s community-wide counterintelligence function (CIA’s Counterintelligence Center had never taken on community functions). Regarding the production of national intelligence estimates, Gates dictated greater emphasis on alternative views and use of outside experts in an attempt to combat potential tendencies toward insularity and single-outcome analysis.

Gates wanted more officers from various intelligence agencies to join his community staff and other staffs or centers that had community roles, and he changed CIA’s security policy to allow non-CIA officers to be stationed at CIA headquarters for up to two years without having to undergo the full lifestyle polygraph examination previously required of them. He acknowledged to CIA’s security chief that this change treated detailees differently from permanent CIA staff employees, but he justified it on the grounds that their access to CIA secrets was constrained by their limited periods of assignment and by internal security compartmentation practices. This action showed both Gates’s knowledge of the nuts and bolts of organizational change and his seriousness about fostering contact and integration across the community.

Gates and the INTs

Gates’s mantra for change for the community was integration, and integrating the intelligence collection disciplines other than SIGINT constituted a major target for advocates of intelligence reform in the 1980s and early 1990s. For many, NSA was the model because one manager, NSA’s director, was responsible for the overall management of all US SIGINT activities, national and tactical, and thus could achieve efficiencies and be held accountable for the entire INT. Gates was attracted to the idea of integrating the other collection disciplines into vertical organizations, or at least the semblance thereof, as one step toward better community management. He was also keenly aware, however, of the concomitant need to have devices that laterally cut across these vertical organizations—pejoratively called “stovepipes” by many—so that all sources of collection could work together with greater efficiency and less duplication and be responsive to DCI and external customer guidance.

When Gates requested comments on the Childs study of a new community management structure, he elicited an interesting memorandum from Jimmie Hill, the deputy director of the NRO, who articulated concerns held by many intelligence professionals. If “functional managers” or strong “stovepipe” agencies were stressed for the different intelligence disciplines, Hill wrote, “it will be absolutely vital to create some unit and a structured process that prevents or at least compensates for the natural tendencies of a ‘stovepipe’ organization.” What were these tendencies? In Hill’s view, such organizations will try “to be all they can be” in addressing every intelligence problem, they will husband their best information “to seek advantage in the competition for resources and favor,” and “most dangerous and insidious, they will deliberately attempt to be the sole sources of ‘truth’ with respect to the doability, knowability, and interpretation of data within their functional discipline.” He also charged that they would be drawn to concentrate on “hot current topics,” seeking to gain favor with customers by trying to be the first to report information on such topics and thereby compete for favorable attention and resources.

Hill cited CIA, NSA, and DIA as having been guilty of such tendencies over the past 30 years when they thought they had unique control of data. He drew attention to imagery—the collection discipline in which Gates was contemplating the greatest change—as the “single area” where the tendencies he warned about were absent. Why? “With no single IMINT authority,” Hill asserted, “it has been virtually impossible for any single agency, authority, functional manager to withhold important information for ‘competitive advantage’ or to insist, without serious scrutiny, that something is undoable or unknowable from IMINT.” What Hill seemed to be arguing in favor of was transparency and perhaps also distributed rather than centralized functions where possible. Thus, his arguments stand, in effect, as the “anti-model” to a complex, vertically structured organization.[12]

Hill realized Gates was headed toward single functional managers for each INT, including imagery, and he was really arguing for compensating measures to counteract the tendencies about which he was warning. He acknowledged that there were problems with imagery that needed attention, but he offered no path to fixing them other than recommending a strong DCI staff capability to guide change. He was emphatic that the DCI should not concede leadership of the INTs to DOD: “If all DOD elements of the NFIP are consolidated under an OSD manager and staff, the game is essentially over for the DCI.” Hill’s answer was for Gates to have “a very strong and reasonably large community staff (probably around 400),” with its most senior element being equal in rank to the heads of collection agencies such as NSA. This recommendation was similar to others made over the years by those who would strengthen the DCI’s community role, but it did not fit Gates’s personal preferences or operating style.

Whereas Turner had wanted a “vice president” to manage on his behalf all intelligence collection, Gates wanted four or five senior subordinates in charge of collection operations, each responsible for managing one of the major INTs. This was a more modest but more realistic goal than Turner’s, and it made use of existing organizations and people in ways that did not arouse bureaucratic hackles. In one respect, however, it was a more ambitious and potentially more effective effort. Whereas Turner’s collection deputy never attained anything more than top-level, general direction over tasking, Gates’s collection chieftans were program managers with full managerial authority and control over all their agencies’ collection operations.

Imagery

Early on, Gates had marked imagery for special attention. For this INT, Gates had told his CIA colleagues in December 1991, “we need to find a mechanism, as with signals intelligence, where a single organization can manage the tasking of both national and tactical reconnaissance assets to satisfy specific requirements in the fastest, least costly way.” He praised DIA for its role as executive agent in charge of imagery tasking during the recent Gulf War, thus indicating his willingness to give DOD a strong role in leading a new imagery “stovepipe” organization. It should be noted, however, that the process he praised was one in which the DCI retained formal top-level control of the INT even during wartime; it had been the DCI who in 1990 had granted executive agency status to DIA to manage the national imagery assets in the run-up to the Gulf War.[13]

The Gulf War had left in its wake a sense of dissatisfaction within the military services about imagery support to military operations. Postmortem examinations showed that many of the problems lay within DOD systems for disseminating imagery products, but there was nonetheless a natural tendency in DOD to believe that more DOD control over national imagery assets would be helpful in addressing the perceived shortfalls. From Gates’s perspective, the main equity was continuing DCI authority over the tasking of national imagery capabilities for strategic policy support. As for the numerous other problems needing fixes, why not let an imagery organization and DOD take the lead in solving them?

Gates commissioned a task force to examine the issue of the appropriate organization of national imagery. The experience of the Gulf War loomed large because of difficulties in acquiring synoptic coverage of broad areas and in moving images around to the users who needed them to plan and conduct military operations. But there were other issues as well, and the commission did an excellent job of surveying rapidly the main issues involved and recommending the establishment of a national imagery agency that would include COMIREX (the DCI’s Committee on Imagery Requirements and Exploitation), NPIC, and the Defense Mapping Agency (DMA). COMIREX had been the longstanding DCI collection committee for national-level imagery, and it was the unique control point for collection operations of the small number of major imaging satellites. National-level analysts at CIA and DIA, many of whom worked in NPIC, viewed it as particularly successful in making the process of tasking and feedback transparent and successful. They understood the capabilities of the satellites well, and they had a good sense of when the images they wanted could be obtained. Military users often felt less confident in their understanding of how the system worked and hence less well served by this DCI mechanism.

The report recommended placing the new agency in DOD because DOD was both the primary user of imagery products and the location of the major challenges. But the new agency was to have, like NSA, “strong ties to the DCI,” who would administer a new national imagery program within the NFIP, govern collection requirements in peacetime, and protect imagery sources and methods. The report anticipated further adjustments of collection priorities and increased foreign interest in imagery products, and it foresaw the advantages of exploiting new technologies and developing seamless interoperability of tactical and national systems. The report also dealt realistically with the management aspects of change, urging that the distributed nature of the imagery exploitation process “be protected” and that a mix of professionals from different parent agencies be involved, at least initially, as was done in the NRO.

Gates worked closely with Secretary of Defense Cheney and Chairman of the JCS Colin Powell in discussing the report’s recommendations and organizing the new agency. Powell, however, refused to place DMA, whose capabilities were critical to military operations, under an organization over which the DCI would exercise certain authorities for tasking. For his part, Gates retained NPIC within CIA. So, in the end, owing to concerns about both DOD and DCI equities, Cheney and Gates created a “Central Imagery Office” (CIO) within DOD on a more limited basis than the national imagery agency envisaged by Gates and his task force. It provided central management, planning, and tasking functions, but left exploitation of imagery and production of maps where they were.

In fact, a lot of the angst in DOD about the idea of a large new intelligence agency came from the military “operators,” those responsible for warfighting. Tactical combat veterans had a visceral mistrust of the NSA model cited by the blue ribbon panel because of their experience with signals intelligence. Imagery was at the heart of the targeting for the new precision-guided weapons coming on line, and they did not want the development of imagery assets, the integration of imagery or imagery-derived information, or other aspects of the new “smart” weapons dependent on some unwieldy and distant organization filled with civilians. They welcomed improvement in combat support from national imagery systems, but their concern over how tactical systems might be affected and controlled was too great to overcome.

Gates explained the decision essentially as being forced by caution, using the “first do no harm” adage. He stated that officers in all agencies—including CIA, the military services and commands, and DMA—were worried that proceeding quickly to form a large, new agency would endanger activities that were being performed reasonably well. Also, the connection between the national and tactical imagery assets had not yet been worked out, and this would be mainly a DOD task. The result seemed disappointing to those who had worked hard on a more advanced and integrated approach, but it did establish a DOD-based imagery organization that anticipated the course of future change.

By April 1992, as they were reaching final agreement on their joint decision to set up CIO, Gates told Cheney he was concerned about the transfer of some CIA authorities and resources to the new office and needed formal assurances that would set out “a proper legislative history which explained the circumstances and intent of the establishment of this organization.” Cheney obliged and provided Gates with a memorandum assuring Gates that his responsibilities as DCI would be fully taken into account in the DOD charter of the new organization and explicitly recognizing the NSA model: “the intention is that the CIO’s relationships be similar to those of the National Security Agency (NSA).” This was in fact done in May 1992 in a DOD directive establishing CIO, and Gates issued his own DCID describing CIO as “a joint Intelligence Community-Department of Defense activity within the Department of Defense.” Cheney, upon Gates’s recommendation, named longtime IC Staff deputy director William Lackman to head the new office, and both DOD and CIA careerists staffed it.

Gates noted in his letters of appreciation to the task force members that the decision reached did not adopt in full the organizational recommendation they had made “primarily due to concerns at Defense.” He told them that they had nonetheless persuaded Cheney and Powell that “they have a problem which they are now prepared to address with structural changes,” and that an ultimate solution nearer their vision was still possible. “It is just that the process of integration,” Gates wrote, “will take considerably longer than you all contemplated or recommended.”

Even so, what Gates had done with imagery marked a significant modification of the DCI’s community role. National imagery collection and tasking had been especially associated with the DCI and with CIA since the beginning of the U-2 program in the 1950s. Gates’s willingness to place it in a DOD agency, therefore, departed from the policies of previous DCIs. This step established a path along which Gates’s successors would go even farther in the 1990s, causing some CIA veterans to conclude that too much had been “given away” in terms of decreasing CIA’s influence over both collection tasking and analysis in the imagery arena.

NRO Goes Public

Another reform adopted after a task force commissioned by Gates had done its work quickly in early 1992 was the reorganization of the NRO into collection discipline sub-organizations (imagery and signals), replacing the tripartite structure of Air Force, CIA, and Navy programs (referred to as programs A, B, and C, respectively) that had existed for 30 years. Gates seemed determined to encourage intelligence agency heads to act to integrate better within their organizations as well as to cooperate more between themselves. This move coincided with the NRO’s co-location of its program headquarters in a single new headquarters in northern Virginia, accomplished in 1992. Finally, in September 1992, Gates prevailed on a skeptical Cheney to allow the declassification of the name and existence of the organization and of the names and titles of its top three officials.

The SSCI had been threatening to include in its intelligence legislation a statutory basis for the NRO and declassification of the “fact of” its existence. Marty Faga, director of the NRO, recommended acknowledging the NRO publicly, pointing out that the president would be hard pressed to claim “grave damage” to the nation’s security to forestall mention of the organization’s name in a law. The move did in fact obviate the establishment of a legislative charter for the NRO, thus achieving an administration objective of keeping the charters for DOD’s major intelligence agencies a matter of executive branch determination. Deputy Defense Secretary Donald Atwood brought the NRO out of the closet in a decision announced to the public on 18 September 1992, the 45th anniversary of the founding of the US Air Force and the CIA, the two organizations most critical to the NRO.

The Other INTs

Gates also moved to strengthen the top-level structure for human source intelligence, or HUMINT, giving it a single chief with community-wide responsibility and constructing a requirements and tasking entity that knit together the main organizations involved in this kind of collection. Gates designated CIA’s DDO as the National HUMINT Manager (formerly he had nominally been the CIA representative to the DCI’s HUMINT committee) and charged him with coordinating human source collection more rationally and efficiently. It fell to this new INT commander to deal with DIA’s initiative to form a Defense HUMINT Service within DOD.

To manage human source intelligence requirements better, Gates set up a National HUMINT Requirements Tasking Center (NHRTC) to be manned by officers from CIA, INR, and DIA. Each of these three organizations contributed a senior officer, and Burton Gerber, the first chief of the center, was a respected senior DO officer. The tasking levied on human source collectors was to be coordinated by the center, and the tasking of non-NFIP assets was to be worked in consultation with the departments and agencies involved. DIA was to handle further subdividing of tasking assigned within DOD, and the center’s tasking was to take into account information that could be obtained from open sources or other INTs.

As for measurement and signature intelligence (MASINT), a category formally recognized as a collection discipline under a DCI committee only in the 1980s, Gates wanted it to have an organizational home stronger than just a DCI committee, so the Director of DIA agreed to house this community function within his agency. As with the other INTs, the requirements committee for MASINT was subordinated to the modest new program office established within DIA. Finally, Gates created an embryonic open source collection discipline organization and placed it within CMS. He selected a senior DIA officer to head the “Open Source Coordination Office” and asked him to begin the task of planning strategically for how the entire community might best take advantage of what most thought was already a daunting amount of openly available information on many subjects. Open source intelligence (OSINT) thus became a program entity and began to be referred to as a fifth INT in the 1990s.

Gates also commissioned a task force on the subject of “Classification and Control for the 1990s and Beyond.” He was returning to an issue where Turner had attempted a major reform, rationalization of the maze of security compartments and control systems that probably few people knew in its entirety. As with other issues, costs were a driving concern, and the hope was that a simpler security system could be managed more cheaply and efficiently. Also, new and growing uses of intelligence in the post-Cold War world in connection with law enforcement, environmental and economic issues, coalition warfare, and UN peacekeeping placed new pressures on the old security practices. No practical result came of this effort under Gates, but it anticipated further efforts by his successors.

Gates also commissioned two panels to review particularly sensitive collection programs that he knew were being targeted for cuts by some in Congress. For one, he chose Dick Kerr, who produced a careful report that showed how the program could be altered to remain productive in the post-Cold War era. In the other case, he chose R. James Woolsey to head a study of the National Reconnaissance Program. Woolsey produced a thoughtful review that pointed the way to a prudent continuation of a valuable program in the changed world of the 1990s and beyond.

Requirements

A consistent theme of DCI efforts to guide change in the late 1980s and early 1990s was attention to defining and re-defining the intelligence information “needs” of the major consumers of intelligence. Gates hoped that NSD-67 would be helpful in avoiding charges that intelligence simply picked and chose what it wanted to do rather than doing what was directed and needed. He realized also that simply having an updated list of needs from the president was not enough; the DCI had to do something with it. In his confirmation hearings, Gates drew attention to his actions as DDCI to improve the requirements and evaluation function on the IC Staff to show that his new initiatives reflected a pattern of interest that went back several years.[14]

After setting up CMS, the new community collection and production boards, and the new imagery and HUMINT offices, Gates in July 1992 set up a task force under his CMS deputy for requirements and evaluation to devise a new, simpler community requirements process. The task force recommended establishing community “issue managers” to set forth the key questions to be addressed across the community and to develop “national intelligence strategies” detailing collection and production actions to be undertaken by the various intelligence agencies to meet customers’ needs for information. Gates approved the new process in one of his last acts as DCI in January 1993 and directed development of an implementation plan to turn what was essentially an idea with only a weak, general consensus among the agency heads into a workable process. This work he commissioned began the process of institutionalizing a role for issue managers as a tool of so-called matrix management, whereby coordination across programs is facilitated by staff officers appointed by the DCI whose responsibilities were defined by substantive topics rather than type of activity.

Gates and Cheney

Gates worked to keep his relationship with Secretary Cheney a positive and cooperative enterprise. By requesting Cheney’s aide, Richard Haver, to be the head of CMS, Gates signaled his desire to work closely with Cheney and other DOD intelligence leaders as he carried out his community role. Gates knew that future DCIs and secretaries of defense would be unlikely to have as close a relationship as he enjoyed with Cheney, but he felt the risks to the DCI’s community role inherent in his steps to integrate his office’s efforts with the Pentagon’s were worthwhile if they facilitated a more unified national intelligence community. Gates usually was the one pushing for changes in his sphere of authority against a skeptical Cheney. This was true even for areas such as imagery, where institutionally Cheney might have been expected to be the demandeur. In February 1992, for example, Gates urged Cheney to meet with his task force on imagery management and hear their story of what needed to be changed and why. Cheney agreed but said he “needed to be convinced something was broken and about the proper solution.”

Gates also had to contend with Cheney’s assistant secretary for intelligence, Duane Andrews, who had his own ideas for organizing DOD intelligence. Gates wanted him as a senior member of his community team, and in January 1992 he wrote him a short note urging him to join the biweekly luncheons hosted on a rotating basis by the various intelligence leaders. “I think it would be beneficial to all concerned,” Gates wrote, “if you could attend the Intelligence Community luncheons. Many Community issues get discussed and resolved in that more informal setting.” Gates valued these sessions as a vehicle for exercising DCI community leadership (recall his experience in chairing Deputies Committee meetings at the NSC), and he wished for as much comity and participation as he could attain (he felt he had good relationships with other key leaders such as the heads of NSA and DIA).

Gates consulted closely with Cheney, and also with Powell. He shared with them in draft the changes he intended to make within CIA to make that agency more responsive to military support needs. He asked them to provide an Army general to become the new associate deputy director of operations for military affairs at CIA. (Powell did not accede to Gates’s request for a “snake-eater,” but he did promise that the officer would have ready access to the chairman’s office.) The main intelligence support the US military always wants from CIA is more and better information from DO human sources, and Gates’s placement of the new position in CIA’s operations directorate was attractive to DOD. The DO’s support to the Gulf War had been of concern within DOD in the fall of 1990 as the US forces in the theater built up to invasion strength. But as military operations neared and then were conducted, a better liaison arrangement had been constructed that had been gratifying to the military, and Gates, wanting to build on that success, named the CIA officer who had gained Pentagon confidence in that success as the deputy to the newly appointed general at CIA.

The organization of DOD intelligence was in flux during Gates’s tenure as DCI. Lt. Gen. James Clapper, USAF, took over DIA around the time Gates became DCI, and he enjoyed considerable running room to make improvements in his agency and in military intelligence generally. He was given responsibility for managing the General Defense Intelligence Program (GDIP), a main constituent part of the NFIP, a duty that had been taken from his predecessor and given to the ASD/C3I’s office. Clapper moved to consolidate HUMINT within DOD, strengthening DIA’s hold on it by designing a “Defense HUMINT Service” and promising to build up its clandestine capabilities. The latter intent was of some concern to CIA’s DDO, who wondered just how DIA’s efforts meshed with his own orders to improve CIA’s support to the military. Clapper also was seized with the notion of building up his personal role as “Director of Military Intelligence,” which he saw as somewhat analogous to the DCI’s role as community leader, only in his case over the defense or military intelligence community. This issue would build into a disagreement between Clapper and Gates’s successor, R. James Woolsey, but it did not become a problem between Clapper and Gates.

Community Management Review

As summer faded into fall in 1992, Gates worked to limit the damage he saw coming in Congress on intelligence funding. As Cheney had feared, a Senate floor amendment proposed to cut the DCI’s budget further than recommended by the SSCI, and Gates appealed for relief directly to its sponsor, Senator Dale Bumpers (D-AK), as well as lobbying against it with others. In the wake of the amendment’s defeat and final resolution of the funding issue for 1992, Gates went through agonizing discussions on formulating guidance for the next budget cycle (FY 1994). He decided he would direct NFIP program managers to design programs that took into account the trimming recommended by his various task forces but not to shoot for any particular dollar figure that either he or Congress might impose. In a memorandum conveying this guidance, he told them to set their future resource levels to meet requirements, not to fit “an artificially contrived number.” Commenting that this would probably result in a smaller budget reduction than had been discussed up to that point, he worried aloud that his guidance “might create the temptation to avoid necessary measures better to integrate the Community functionally.”

To counter that tendency, Gates ordered CMS to “develop a detailed program addressing specific recommended divisions of labor, integrated activities, functional realignments and other steps to forge much closer coordination, cooperation, and integration in the Community in FY-93 and beyond.” Rather than creating another flurry of individual efforts responding to his personal tasking (although he did appoint a group to address training issues community-wide), Gates took a broad approach in his last assault on improving community management. In October 1992, Gates asked the most senior officers with community-wide responsibilities to undertake a “community management review” with the objective of achieving the “functional integration of the intelligence community.” The leaders were Adm. Studeman (DDCI), Duane Andrews (ASD/C3I), Haver (EXDIR/ICA), and Rae Huffstuttler (EXDIR/CIA). CMS provided central staff support to the exercise. The principal community functions considered were collection, production, infrastructure, and management.

With a presidential election looming, the review exercise had an almost desperate quality to it. It was energized by fears that congressional funding decisions would involve drastic cuts and that innovative measures were needed to prevent these cuts from eviscerating intelligence capabilities. It also involved sessions where “out of the box” thinking was encouraged to see if at least some radical ideas could be seriously considered. For example, regarding collection, an orthodox-thinking group suggested a senior collection council to assure cross-INT coordination whereas a group of self-styled “iconoclasts” suggested a director of collection, perhaps under a DNI-like cabinet officer for intelligence, was needed to assure integration and accountability.

The final report, which was not classified, was longer on conceptual descriptions of problems and solutions than on specific recommendations. It did suggest, however, establishing a DCI Security Commission to examine security policy in conjunction with the departments of defense and energy, an idea acted on by Gates’s successor. The report also encouraged co-location of analysts and collectors, urged completion of the work under way on requirements, suggested a program for integrating community R&D efforts, and dealt with other management issues. A final recommendation advocated establishing an overall “community architect” to integrate planning and development of communications, information handling, and other electronic connectivity across the community.

End Game

The clock had run out for Gates by the time the report was finished, and no new structures or processes emerged from this last flurry of staff work. The main impact of the conceptual thrashing was to pass on ideas to Gates’s successor, who had the benefit of the continuing service of Studeman and Haver, the main formulators of the chief ideas. As his task force grappled with potentially radical changes and the final budget decisions loomed, Gates kept in close touch with Pentagon leaders, gaining promises of help on the budget front from Cheney and Atwood to the extent they could do so within the DOD budget.

In December 1992, before he knew R. James Woolsey would be the nominee of the incoming Clinton administration as DCI, Gates expressed pessimism about the future of US intelligence in a conversation with Deputy Defense Secretary Atwood and Assistant Defense Secretary Andrews. He believed that avoiding a disaster on intelligence funding required raising the profile of intelligence and fighting off unwise reform plans. They discussed the DNI proposal, complete with options for adding the FBI’s counterintelligence responsibility and maybe even all DOD intelligence to the new intelligence chief’s portfolio, and Gates recalled that Turner’s similar plans 14 years earlier “thankfully” had not come to pass.

Also in December, Gates met with Senator Dennis DeConcini (D-AZ), who was slated to replace the retiring Senator Boren as chairman of the SSCI and who had urged a sizable additional cut in the NFIP. DeConcini, who had voted against Gates’s confirmation, had been hard on Gates during 1992, giving him credit for his task force efforts but calling him a “company man” and charging him with not following through on his reforms.[15] Gates gained DeConcini’s promise to withdraw his budget cut request by pointing out that the appropriation committee had already subtracted a large sum from what the SSCI had authorized and urged him to get to know the community and its business better in order to bolster his effectiveness as SSCI chairman and not simply set budget targets that might damage needed capabilities.

After the election of Bill Clinton in November 1992, the president-elect’s transition staff formed an intelligence transition team headed by HPSCI staff chief Jack Keliher and SSCI staff chief George Tenet. Gates supplied the transition team with materials about intelligence processes but did not take them into his full confidence regarding planning for the future. This wary approach softened once R. James Woolsey was named as Gates’s successor. Gates was pleased with the new administration’s choice, and welcomed Woolsey as a teammate in the process of planning the next steps in intelligence reform. He ordered that the community management review study be shared with Woolsey, and Woolsey’s cooperation with Studeman and Gates in Gates’s final days eclipsed the transition team’s formal report. In a move emphasizing continuity, DDCI Studeman was told he could forget about his pro forma resignation letter. Gates told his senior staff at CIA that he thought Woolsey had done an excellent job heading the task force that had examined the NRO program and that Woolsey would try to build on the change already under way rather than alter course to fit a brand new agenda.

In Retrospect

Gates’s tenure as DCI bears comparison with Colby’s. Both were ambitious CIA career professionals chosen as DCIs by the White House in the hope they could lead the nation’s intelligence efforts competently during times of change. Both tried to work cooperatively with activist congressional overseers by preempting potentially unwieldy congressional ideas with their own initiatives. Both were true believers in the need for some reforms in intelligence and willing to take a leading role in attempting them, although both worked for presidents who were not notably reform-minded. Both willingly, even enthusiastically, used the term “management” prominently in connection with fulfilling their community responsibilities. Both attempted to use “issue managers” to assist their leadership over community programs. Both had a quiet determination and a strong sense of self-confidence in their leadership roles.

With respect to their community roles, Colby’s efforts were more modest and had little lasting effect with the notable exception of his alteration of the estimative process in setting up the NIOs. Gates’s efforts were more ambitious, and while he left office before he could do much to follow through on his initial moves, more of his changes took root and survived or anticipated future change. Colby eventually lost the president’s confidence and was fired. Gates kept the president’s confidence in his ability to handle change, but he lost out to the vagaries of elective politics when Bush failed to win a second term. Making no effort to stay on in the next administration despite the president’s reminder that the DCI position was traditionally non-political, Gates left office in January 1993 having not yet reached his 50th birthday.

Footnotes:

[1]Handwritten notation that Gates wrote for his own use in preparation for attending a conference of leaders of the CIA’s DO held 7–8 November 1991, immediately after Gates became DCI. Presumably, he wished to deliver this message to the DO, but it applied just as easily to the entire Intelligence Community.

[2]Nomination of Robert M. Gates, Hearings before the Select Committee on Intelligence of the United States Senate, 102d Congress, 1st session, vol. I, 443.

[3]Nomination of Robert M. Gates, Hearings before the Select Committee on Intelligence of the United States Senate, 100th Congress, 1st session, 156.

[4]He also directed (in consultation with Robert Straus, US ambassador in Moscow) the rapid preparation of a series of NIEs on the USSR, which had disintegrated and was about to declare its formal end. Although the end of the Soviet Union meant the end of a particular focus of US intelligence and policy, Gates recognized that the fall-out of the event needed close observation and continual reassessment.

[5]In 1974, working on an assignment at the NSC staff, he had written a memorandum suggesting ways of improving analysis and intelligence policy support on Soviet affairs that reached then DCI William Colby, who passed it on to CIA’s new political research office (where it was less than enthusiastically received). In 1978, again on assignment at the NSC staff, this time as an assistant to national security adviser Zbigniew Brzezinski, Gates participated actively in a group led by Brzezinski’s deputy, David Aaron, that addressed presidential concerns about the state of political intelligence. During 1980–81, working on the top floor of CIA at Langley, he wrote numerous thoughtful memorandums first to DCI Turner, then to DCI Casey and DDCI Inman, critical of the way analysis was being done in the NIC and CIA’s DI, helping to pave the way to his being appointed DDI in early 1982. Immediately after becoming DDI, he blitzed CIA’s analytic directorate with ideas and programs to back them up. The pattern of personal leadership he showed as DCI was thus well formed long before 1991.

[6]Other CIA task forces addressed increasing the number of case officers in the field, reducing the volume and increasing the quality of analysis, improving internal communications, and enhancing CIA “openness.”

[7]Gates tasked separately the DDI at CIA (actually, the DI’s Collection Requirements and Evaluation Staff, or CRES) and the director of the IC Staff (actually, the deputy director, Bill Lackman) to lay out what the DCI might do if faced with deep cuts. The staffers working on the projects at the two places became aware of each other’s work, and, without coordinating, compared notes on their results. The results of the competing exercises were quite similar, essentially recommending that the DCI continue to support the most important, truly “national” intelligence capabilities. This account is based on the recollection of the author of this study, who was head of the DI staff involved at the time of the exercise. The episode is an example of how DCIs have sometimes used a CIA staff to provide a check on their IC Staff with respect to a community matter. “Insider” DCIs often turn to CIA officers to watch out for their community leadership role, fearing that non-CIA officers will adopt consensus views that concede too much to other agencies’ equities rather than giving primacy to protecting the DCI’s authorities. “Outsider” DCIs sometimes choose to avoid CIA officers’ advice regarding community issues for fear of being led toward parochial CIA views rather than broader, community interests.

[8]In his memorandum to them, Gates noted that some departments and agencies had been lax in not coordinating all intelligence activities with CIA’s chiefs of station—the DCI’s man on the scene abroad—but he also acknowledged that some station chiefs had not fulfilled their duty to represent the interests of non-CIA US intelligence organizations abroad, declaring: “Both sides of the equation need correction.”

[9]One issue worth noting was differing advice regarding whether the chief OSD officer handling intelligence should be separate or linked with C3. Adm. Inman told HPSCI that he thought a separate intelligence assistant secretary would be useful in DOD. Gen. Paul Gorman, USA, and Gen. Al Gray, USMC, however, opposed this strongly, urging that the integration of command and intelligence was important. Don Latham, a former assistant secretary of defense for C3I, also opposed splitting out the intelligence functions from C3 within DOD.

[10]A regional connection of a sort existed. Not only were both committee chairmen from Oklahoma, Gates was himself from Kansas, home state of the Republican minority leader in the Senate, Senator Robert Dole. In 2002, Gates became president of Texas A&M University, and his neighbor as president of the University of Oklahoma was David Boren.

[11]The NICB and NIPB were cross-linked with both CMS and the NIC. The new vice chairman of the NIC for evaluation was made the vice chairman of the NICB, thus linking the NIC with the requirements functions administered by CMS, and the CMS requirements deputy was made a vice chairman of the NIPB, thus linking the collection requirements business of CMS with the consumer support and analytic production functions of the NIC.

[12]Former Assistant Secretary of Defense Donald Latham testified against establishing a new agency for imagery intelligence, predicting it would create a more bureaucratic, stovepipe separation between IMINT and SIGINT when what was more desirable was a more integrated architecture encompassing both.

[13]Basic plans and directives over the years since World War II have envisaged the transfer of control of major intelligence activities from the DCI to the secretary of defense during wartime. The post-1945 wars fought by the United States never were formally declared, however, so those plans have essentially become a dead letter. CIA and the top federal intelligence official probably would continue to function independently in any future war rather than be subordinated to the JCS as was the case with OSS and Donovan in World War II.

[14]Nomination of Robert M. Gates, Hearings before the Select Committee on Intelligence of the United States Senate, 102d Congress, 1st session, vol. I, 476 and 482–83.

[15]Washington Post, 7 December 1992: A1.