Fifth DCI, Allen Welsh Dulles
Allen Dulles: Reluctant Manager
My authority for coordination is a recommending one and not a mandate.
When Dwight Eisenhower became president in 1953, there was only the beginning of an intelligence community in the sense of a truly cooperative enterprise. Compared with 1946, however, one could judge what had been achieved as a creditable accomplishment. Smith’s success in creating a positive atmosphere for, and the mechanisms of, collaboration among the intelligence chiefs had fulfilled the main recommendations of the 1949 Dulles-Jackson-Correa Report. The creation of NSA in the waning months of the Truman administration had carried out the plan outlined in the 1952 Brownell Report. Eisenhower accepted the CIA and the rest of the Intelligence Community as he inherited them and at the outset expressed no desire to alter either organizations or authorities.
The new president was knowledgeable about and favorably inclined toward intelligence as an integral element of national security policymaking. He made it clear that more and better information on Soviet military capabilities and intentions was of highest importance in an era of potential devastating surprise attack. He also made it clear that he would look to the DCI to provide the kind of strategic intelligence that he felt he and the NSC needed to formulate and carry out policy at the national level.
Dulles as DCI
President Eisenhower selected prominent attorney and wartime OSS veteran Allen Welsh Dulles to be his DCI. Dulles was a person of stellar reputation in the intelligence field as well as the brother of the new secretary of state. DDCI at the time of his appointment, he had been since 1950 a senior official working under DCI Smith in guiding the rapid growth of CIA’s clandestine service and its activities overseas.
Dulles felt no need to ask the president for revised direction or authorities. The practices of community leadership set by Smith seemed adequate to him. Dulles’s personal style fit naturally with the committee-oriented kind of collaboration that had developed as a main mechanism of intelligence community coordination. He emphasized a positive, collegial interchange of views and, like Smith, worked to expand cooperative efforts where he could.
President Eisenhower, however, was anxious to give him new tasks to improve the nation’s intelligence posture. Eisenhower actively employed eminent US citizens, especially scientists and engineers, as leaders and members of presidential panels that sought to improve the US ability to counter the Soviet threat. He had MIT President James R. Killian, Jr., form a Technological Capabilities Panel to examine the possibility of harnessing growing US technological capabilities to garner key intelligence information about the denied areas in the USSR that human source intelligence had failed to penetrate. The panel recommended new strategic reconnaissance programs such as the U-2 aircraft and urged that the CIA take the lead on them. The president acted on the panel’s advice and promptly assigned the task to the DCI and CIA.
Dulles, who favored espionage and covert actions, did not welcome the assignment. But, under presidential pressure, he embraced the challenge and turned to CIA to lead the way in developing a secret strategic reconnaissance aircraft. The program was, after all, aimed at conducting secret intelligence operations overseas, and therefore within the scope of CIA’s charter. Under Dulles, CIA’s Richard Bissell and his US Air Force counterparts managed an exemplary joint program.
This presidential initiative had profound implications for the DCI and his community role. The U-2 airplane (and later the CORONA satellite), along with the associated camera development and photo exploitation activities, marked the beginning of a new dimension of community teamwork between the CIA and the Defense Department, especially the Air Force, on strategic reconnaissance programs.
Dulles regarded his own and CIA’s leadership of a technical intelligence collection program such as the U-2 enterprise as not posing problems of perception or conflict with other parts of the Intelligence Community. CIA, after all, was growing in other ways as well, with respect to both clandestine operations and intelligence analysis. He would have been content with the conclusion offered by his official CIA biographer: “CIA became more acceptable as a coordinator as the component parts of the Agency grew in stature….” CIA officials saw these areas of growth, which did not duplicate programs elsewhere, as appropriate for CIA.
Thus, for example, the CIA photo exploitation shop, started in the early 1950s, invited other organizations to join, and two days before Eisenhower left office in January 1961 it was rechristened the “National Photographic Interpretation Center,” or NPIC, a name this joint CIA-military endeavor wore with pride until the late 1990s. It is worth noting, however, as an indicator of the kind of leadership that Dulles preferred, that while Dulles welcomed the participation of non-CIA organizations in NPIC, he did not choose to confront the Air Force’s decision to set up a separate Strategic Air Command photo-intelligence organization. He was for centralization and coordination of national intelligence activities, but not at the expense of fighting divisive bureaucratic battles or attempting to monopolize or otherwise insist on sole control of entire fields of intelligence activity.
During Dulles’s tenure, the joint management of strategic reconnaissance programs by CIA and DOD was relatively smooth and involved only a few senior executives. This involvement of the DCI, however, in technical program management would change forever how overseers and others would come to view his community role. Inevitably, the programs grew in complexity and size, coming to pit increasing numbers of senior officials in antagonistic bureaucratic battles. One historian commented: “as the community became larger and as technical systems required larger budgetary allocations, the institutional obstacles to coordination increased.”
Pressure for Greater Coordination Grows
Dulles’s collegial leadership style was congenial to other intelligence chiefs, but not satisfactory in the eyes of the president or of the prominent citizens he employed as consultants in advising how best to strengthen the national security efforts of the government. In 1955, an outside panel chaired by Gen. Mark Clark, USA (ret.), examined intelligence as part of a larger review of the federal government conducted by a second Hoover Commission. It urged more DCI attention to the coordination of intelligence and less to the internal management of CIA, which the panel felt could be delegated. Dulles thus found himself on the receiving end of criticism similar to what he had helped to author in 1948. And now, looming behind the recommendation, was a president personally interested in his response.
Eisenhower himself applied pressure on Dulles to put more personal time and effort into his community role. In December 1956, the president’s new Board of Consultants on Foreign Intelligence Activities, headed initially by Killian, recommended that the DCI be encouraged to exercise a more comprehensive and positive coordinating responsibility and stated that he could do this within his existing authority. At the same time, the board felt the NSC should revise the NSC intelligence directives to the DCI, and it pressed Dulles to do more to integrate the community and reduce duplication of effort and to appoint a deputy to handle the CIA while devoting more time himself to community matters.
Dulles responded in part by reassuring his community colleagues that they had been cooperating well and should do even better. In addition, he turned around the deputy proposal and suggested appointing a senior deputy, but having that person take up community duties. He thus freed himself as DCI to continue to pursue CIA-centered activities connected with human agent operations and national intelligence production. His nominee for the new community job was a retired US Army general of excellent reputation, Lucian Truscott, Jr., to whom he gave the task of leading the way in revising the NSCIDs. He also sent the president a memorandum for him to sign urging all concerned to work harder on coordination.
President Eisenhower chose to accept this half-a-loaf approach, and in August 1957 he issued a memorandum to the NSC and the DCI concurring in his consultants’ recommendation and in Dulles’s appointment of Truscott as a “Deputy Director for Coordination.” The memorandum starts with language commending “strong centralized direction of the intelligence effort of the United States” through the NSC and the DCI. It then shifts, however, to a rhetorical assertion that “the exercise of a more comprehensive and positive coordinating responsibility by the Director of Central Intelligence can be of the utmost value to the entire intelligence community and strengthen the national intelligence effort” and urges all members of the community to render the DCI their “fullest possible cooperation.” Eisenhower’s exhortation was a notable presidential endorsement of DCI community leadership, but hardly a marshal’s baton.
This important interaction between the president and the DCI was not simply one of exchanging memos from afar regarding the recommendations of others. Eisenhower personally urged Dulles to be more forceful in leading the community. By some accounts, he had been somewhat taken aback at how much intelligence activities had come to cost, and his twin concerns that those resources be managed efficiently and take advantage of modern technology drove his desire to see more active leadership and management by the DCI. Dulles mounted a last-ditch defense. He told the president that he simply was not the kind of administrator or manager the president apparently wanted and that perhaps he needed another DCI. Faced with this response, the president backed off: “I’m not going to be able to change Allan [sic]…. I’d rather have Allan [sic] as my chief intelligence officer with his limitations than anyone else I know.” In effect, Eisenhower took an approach similar to that adopted by Dulles regarding community affairs: urge and persuade, but don’t force confrontations in which china is broken.
New Board and New Directive
President Eisenhower’s interest in forcing intelligence integration remained strong, however, and this episode was not quite the end of the story. His consultants on intelligence, frustrated over what they had learned about the lack of clear-cut assignments of responsibility and the difficulty of establishing efficient centralized processes, sent Eisenhower another report urging the creation of a single United States Intelligence Board (USIB), unifying the Interagency Advisory Committee (IAC) and the United States Communications Intelligence Board (USCIB). The president told Dulles to adopt the suggestion. The boards met to consider the proposal and rejected it. Dulles duly responded to the president’s consultants explaining the disadvantages of the scheme. They in turn responded that they had been misunderstood and that the reasons why communications intelligence had previously been kept separate should not prevent the formation of a unified top-level intelligence policymaking board.
In the meantime, the president decided to act. In March 1958, at a joint meeting of the IAC, the USCIB, and the NSC with the president in the chair, discussion about adopting a new set of NSCIDs brought out a point about the “separateness” of some kinds of intelligence (in this case, the JCS was asserting its right to handle a topic separately from the community as a whole). The president elicited an explanation of the value of “separateness” and then rejected it, dressing down the hapless DOD participant who had voiced it and explaining that the objective of integration and unity of effort was paramount. He then announced that there would be a unified USIB despite all objections and gave the community six months to set up the new process and revise the NSCIDs (with language in them referring to the IAC) that were being approved that very day.
The actions taken in the ensuing six months to implement the president’s decision broke new ground with respect to codifying a strengthened DCI role in leading the Intelligence Community. A new NSCID 1 issued in April 1958 (and reissued in September to account for the creation of USIB) gave the DCI an explicit formal mandate to “coordinate the foreign intelligence effort of the United States, in accordance with the principles established by statue and pertinent National Security Council directives,” a stronger formulation than was in the 1947 National Security Act (which had not endorsed the similar mandate given to Vandenberg by the NIA in 1946 and surrendered by Hillenkoetter in 1947). It also authorized the DCI to declare that a consensus existed in the community and forward a view to the NSC in the absence of an insistence by a community member that a dissent also be sent. It included rhetorical language praising integration as a goal, and the Director of the Bureau of the Budget (BOB), in concurring on the September 1958 version of NSCID 1, stated his continuing concern that the DCI’s leadership of community coordination was just barely well enough supported in the final draft to warrant its support.
The issuance of these stronger statements of DCI authority may have owed something to the president’s impatience with the persistent lack of support for integration among members of the IAC. Work undertaken in 1957 to revise NSCID 1 had resulted in wrangling reminiscent of the late 1940s debate over “collective” versus “individual” responsibility and a potential backsliding from precedents set in the early 1950s. This episode showed how important outside intervention to bolster DCI authority continued to be. In a sense, too, the new codification of stronger DCI leadership took aim as much at Dulles as at his colleagues heading other agencies. His non-confrontational style sometimes permitted wrangling to drag on and frustrated those who wished for crisper decisions and a stronger hand at the helm. In April 1958, on the day after the new NSCID 1 went into effect, Dulles, still reluctant to force the presidentially ordered merger, gathered the IAC and USCIB together and had the president’s assistant for national security affairs (Robert Cutler) and the chairman of his board of consultants (Gen. John E. Hull, USA, Ret) explain the president’s strong personal interest in unifying their two groups into one board. This extraordinary use of presidential assistants testifies to the dependence of the DCI’s authority on the attitude of the president and the DCI’s relationship with him.
A particularly notable aspect of the changes was a strengthening of the DCI’s authority with respect to signals intelligence. Separate NSCIDs on communications intelligence (COMINT) and electronic intelligence (ELINT) were joined in a new NSCID that treated them both, and NSA was recognized to have unified direction and control over both disciplines even though they would be handled in many ways separately at lower levels. The new structure erased the “separateness” doctrine that had exempted COMINT from other intelligence activities recognized under the 1947 Act as being led by the DCI. (Under the NSCID previously governing it, USCIB had not been considered a part of the Intelligence Community.) Also abolished, since the new unified USIB would be reporting directly to the NSC, was the special committee of the NSC to which USCIB had reported. For the first time, the DCI’s function as the president’s national intelligence coordinator was extended to policies governing both COMINT and ELINT.
The USIB System
The new USIB had two members in addition to those belonging to the old IAC. One was a “special operations” assistant to the secretary of defense, and the other was the director of NSA. The addition of these positions marked the beginning of a new phase of DCI chairmanship of the community. The step integrated more closely into the community structure the largest US intelligence agency and recognized the role of the secretary of defense as a kind of co-leader with the DCI of the community.
USIB, like its predecessors, presided over a growing complex of committees that coordinated intelligence activities in a variety of areas. Some were new topics that arose as issues in the 1950s. For example, attention to automated data processing extended earlier efforts to make standardized systems of collating, storing, and making available to users intelligence data of all kinds. Another committee took up discussions on the gathering and sharing of information about foreign states’ development of rockets and missiles. These committees rationalized tasks, decided relative priorities, and in other ways coordinated the activities of the community.
Eisenhower sought to keep the costs of defense, intelligence, and other national security activities as reasonable as possible, and that goal motivated much of his interest in integration, efficiency, and centralization of responsibility in the intelligence field. Killian and others in 1957 recommended a centralized gathering of accountings of the costs of doing intelligence. In response, the DCI formed a cost estimates committee in 1959, in part to preempt actions threatened by BOB, but he gained only limited insight into the activities of other agencies from the available budget data.
Another area of DCI leadership was to promulgate the basic objectives and requirements of national intelligence as guidance to the community. His staff established and regularly updated so-called priority national intelligence objectives (PNIOs). Also, various DCI-sponsored groups handled the specific requirements for the new overhead collection systems. The Ad Hoc Requirements Committee for the U-2 program took its target recommendations to the president himself for approval. When CORONA finally came on line in 1960, this committee merged with a satellite-oriented group to form a single Committee on Overhead Reconnaissance (COMOR).
USIB established a Critical Collection Problems Committee (CCPC) in 1958 to work on ways to attack difficult collection challenges in new ways, often with the help of technology. USIB also issued annual reports to the NSC to provide accountability on the DCI’s behalf. This committee work was inevitably process-bound, but it fulfilled the vital function of sharing information laterally among community elements and connecting individual tasks performed at operational levels with larger national needs.
One Last Try
The Eisenhower-Dulles working relationship concluded with yet another study of national intelligence. This time, a BOB initiative devolved into a DCI-commissioned “Joint Study Group” led by senior CIA officer Lyman Kirkpatrick that produced a report in December 1960 with 43 recommendations, all related to the objective of improving the coordination of US foreign intelligence activities. A major thrust was improving the management of intelligence in DOD, including the establishment of a DOD focal point for intelligence in the Office of the Secretary of Defense (OSD): “Great strides toward a more closely integrated community would result from improved intelligence coordination within the DOD.” Asserting that USIB had “slighted its managerial responsibilities,” the study group urged the DCI to enhance USIB’s role in planning for the long term and in making program and budget data among intelligence agencies more comparable to foster better review and coordination. Although continuing to view USIB as the DCI’s principal instrument of community coordination (by now CIA was no longer generally viewed as fulfilling this function), the group also urged that the DCI organize a coordination staff that would be “a full-time group of intelligence professionals owing primary allegiance to the intelligence community rather than to any one member agency.”
The ideas of this group, along with some final recommendations by the president’s consultants, provided grist for two productive top-level meetings that Eisenhower chaired in his last weeks as president. These meetings led in January 1961 to the establishment of NPIC at CIA and later that year to the formation of the Defense Intelligence Agency (DIA) and the transfer of DOD intelligence leadership from the JCS to OSD. Eisenhower knew well the bureaucratic tendencies undercutting his insistence on integration and ordered that the military services give NPIC their requirements to fulfill and that no other center should be permitted to duplicate NPIC’s work. For his part, Dulles took on the responsibility for NPIC despite some hesitation from his administrative chief, Lawrence K. (“Red”) White.
The Kirkpatrick study group had judged that the DCI “now has ample authority to carry out his assigned role as coordinator of the foreign intelligence effort of the United States,” but the president’s consultants on intelligence pressed Dulles until the end of the administration to do more in his community role. In responding to their almost exasperated plea that he tell them something they could recommend to allow him to exercise stronger leadership, Dulles declined to ask for new authority or organizational change and in effect suggested that the changes accomplished in 1958 had meant little in practical terms to the collegial style of leading community affairs that suited him so well. He recognized that his methods of negotiating agreements took longer than others often wished and involved frustrations, but he argued that they knit bonds between people and organizations and brought better results in the long run: “Once achieved by persuasion rather than by fiat, the coordination is likely to develop more effectively than under orders which might be subject to evasion or delay in execution.”
The president’s consultants provided a final counterpoint to Dulles’s refusal to do more. The DCI, they told Eisenhower in their final communication to him, “should divest himself of CIA so he could focus on “the even more important duty of coordinating, integrating, and directing all US foreign intelligence activities.” Within a year, a prominent American who took that charge with utmost seriousness and wanted to do something about it would be appointed DCI.
Dulles memorandum to the executive secretary of the NSC, 24 December 1960, cited in Wayne G. Jackson, Allen Welsh Dulles as Director of Central Intelligence: 26 February 1953—29 November 1961, Volume II, Coordination of Intelligence, 113. This five-volume official CIA biography of Dulles as DCI has been declassified and is available in the national archives but—unlike the Darling and Montague works covering the first four DCIs—has not yet been published in book form. Volume II is devoted entirely to the issue of “coordination” under Dulles.
Jackson, Allen Welsh Dulles, Vol. II, 11.
For a good history of this story, see Philip Taubman, Secret Empire: Eisenhower, the CIA, and the Hidden Story of America’s Space Espionage.
Karalekas, History, 72. Karalekas prepared her history of CIA as part of Senator Frank Church’s (D-ID) select committee’s work in the mid-1970s, by which time many observers had come to believe that the DCI should be doing more in the way of community-wide leadership and management.
Clark Task Force Report, 70–71, cited in IC21: Intelligence Community in the 21st Century, Staff Study of the Permanent Select Committee on Intelligence, House of Representatives, 104th Congress, CRS-9.
The full text of this memorandum is in Warner, ed., Central Intelligence, 49–50.
 Communication from Gordon Gray to Wayne G. Jackson, 21 September 1970, cited in Jackson, Allen Welsh Dulles, Volume IV, 84. See also Stephen Ambrose, Ike’s Spies: Eisenhower and the Espionage Establishment, 242–44, and Karalekas, History, 72–75. Karalekas cites Dulles’ decisions not to push CIA’s strategic military analysis or the establishment of a committee on foreign guided missile developments harder as missed opportunities for institutionalizing greater DCI leadership over community endeavors, concluding that “Dulles’ indifference to this area of responsibility [community coordination] allowed the perpetuation of a fragmented government-wide intelligence effort.”
In an interview in 2000, Gen. Andrew Goodpaster, USA (ret.), who had been a presidential aide quite likely to have been present at the meeting in 1958, did not recall the specific episode but affirmed the president’s strong feelings against “separateness” and for integration with respect to intelligence. Interview of Goodpaster, 14 March 2000.
Declassified versions of the full texts of three issuances of NSCID 1 under Eisenhower (April 1958, September 1958, and January 1961) are in Warner, ed., Central Intelligence, 51–66.
USIB was defined by presidential decision, not DCI mandate. The DCI could and did deal over the heads of these officials, but their cooperation and effort was vital to progress and success in intelligence efforts. Therefore DCIs expended considerable time and effort to build cooperative and fruitful working relationships with them, both via USIB and bilaterally.
USCIB had taken ELINT as well as COMINT under its cognizance in 1955. In 1959, NSA formally announced the use of the term “signals intelligence” (SIGINT) as encompassing both COMINT and ELINT. The challenging problems associated with managing SIGINT have been a recurring spur to efforts to improve the integration and coordination of intelligence. They played heavily into the president’s, his consultants’, and the DCI’s thinking at this time. When DCI R. James Woolsey bestowed a special medal on famous British “wizard war” veteran R. V. Jones in 1994, Jones commented that he questioned even today the wisdom of joining COMINT and ELINT in the same organization. Some intelligence controversies are never resolved!
DOD per se had never been, and is not, as of 2005, considered a constituent member of the Intelligence Community. Rather, the individual intelligence agencies chartered under DOD, such as NSA, are community members.
White, worried over the potential impact on CIA’s budget of future NPIC costs, suggested to Dulles that “maybe NPIC is something you ought to let the Defense Department have.” Dulles put his glasses on his forehead and replied: “Red, you don’t think after I’ve taken all these pictures I’m going to let somebody else develop them?”
Dulles memorandum to the executive secretary of the NSC, cited in Jackson, Allen Welsh Dulles, Vol. II, 113. Jackson observed that “his modus operandi meant that substantive issues on which there was disagreement with other government agencies did not get settled promptly or decisively [the example cited by Jackson—and later, in her history based in part on Jackson, by Karalekas—involved the formation of a committee to coordinate intelligence on foreign guided missiles, which took two years to accomplish].” Jackson, Allen Welsh Dulles, Vol. I, 53.