THE EARLY COLD WAR YEARS Woodrow J. Kuhns Editor e inact Center for the Study of Intelligence | Central Intelligence Agency , 1997 in ais , 4 These documents have been approved for release through the Historical Review Program of the Central Intelligence Agency. This publication is prepared for the use of US Government officials, and the format, coverage, and content are designed to meet their specific requirements. US Government officials may obtain additional copies of this document directly or through liaison channels from the Central Intelligence Agency. Requesters outside the US Government may obtain subscriptions to publications similar to this one by addressing inquiries to: Document Expedition (DOCEX) Project Exchange and Gift Division Library of Congress Washington, DC 20540 or: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Requesters outside the US Government not interested in subscription service may purchase specific publications either in paper copy or microform from: Photoduplication Service Library of Congress Washington, DC 20540 or: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 (To expedite service, call the NTIS Order Desk (703-487-4650) Comments and queries or this paper may be directed to the DOCEX Project at the above address or by phone (202-707-9527) or fax (202-707-0380), or to the NTIS Office of Customer Services at the above address or by phone (703-487-4660). This study will be available on the Internet after 20 March 1997 at www.odci.gov/csi. Assessing the Soviet Threat: The Early Cold War Years Foreword The documents in this volume were produced by the analytical arm of the Central Intelligence Agency (CIA) and its predecessor, the Central Intelligence Group (CIG), between the latter’s founding in 1946 and the end of 1950. During this formative period of the Cold War, President Harry S. Truman struggled to understand the menacing behavior of the Soviet Union and his erstwhile ally, Joseph Stalin. The analysts of CIG/CIA contributed to this process by providing the President with daily, weekly, and monthly summaries and interpretations of the most significant world events. They also provided ad hoc papers that analyzed specific issues of interest to the administration. Because more than 450 National Intelligence Estimates dealing with the Soviet Union and international Communism have been declassified since 1993, this volume features the current intelligence that went to tue President in the Daily and Weekly Summaries. Although some of this material has been available to scholars at the Harry S. Truman Library or has been previously released through the Freedom of Information Act, much of it is being made public for the first time. Taken as a whole, this vol- ume provides the first comprehensive survey of CIA’s early analysis of the Soviet threat. President Truman’s directive establishing CIG on 22 January 1946 created the first civilian, centralized, nondepartmental intelligence agency in Amer- ican history. His purpose was to end the separate cabinet departments’ monopoly over intelligence information, a longstanding phenomenon that he believed had contributed to Japan’s ability to launch the surprise attack against Pearl Harbor. As he stated in his memoirs, “In those days the military did not know everything the State Department knew, and the diplo- mats did not have access to all the Army and Navy knew.” Truman also was irked because reports came across his desk “on the same subject at different times from the various departments, and these reports often conflicted.” He intended that CIA, when it replaced CIG in September 1947, also would address these concerns. This volume focuses on the difficult yet important task of intelligence anal- ysis. Although less glamorous to observers than either espionage or covert action, it is the process of analysis that provides the key end product to the policymaker: “finished” intelligence that can help the US Government craft effective foreign and security policies. During World War II, American aca- demics and experts in the Office of Strategic Services had virtually invented the discipline of intelligence analysis—one of America’s few unique contri- butions to the craft of intelligence. Although it was not a direct descendent of the Research and Analysis branch of OSS, CIA’s Office of Reports and Fstimates built upon this legacy in difficult circumstances. The analysis reaching policymakers in these first years of the Cold War touched on momentous events and trends. Whether the Cold War was the result of a clash of irreconcilable national interests or of a spiraling series of misperceptions, an examination of the current intelligence provided to President Truman during this period—sometimes right, sometimes misleaG- ing—opens a fascinating window on what the President was told as he made his decisions. Equally interesting is the portrait of the analysts, their problems, and the impact on their work of the bureaucratic process, as presented by the editor of this volume, staff historian Woodrow J. Kuhns. Dr. Kuhns makes clear that the lot of the analysts was a difficult one in these early years. Many had been dumped on CIG by other departments that no longer required their services. They were subjected to frequent reshuffling and other forms of bureaucratic turmoil, and they operated under severe time pressure and sometimes with little information at their disposal. CIA’s first analysts are not to be erivied. We have ended this study in 1950 because by then the lines on both sides of the Cold War had been firmly drawn. US leaders had reached their con- clusions about Soviet intentions; had formed their opinions about Mao Zedong, Ho Chi Minh, and other revolutionaries; and had formulated their policy of containment in NSC 68. In addition, a new Director of Central Intelligence, Walter Bedell Smith, implemented a sweeping reorganization of the Agency’s analytical arm in late 1950, breaking the Office of Reports and Estimates into three smaller but more clearly focused offices. The CIA thus entered a new phase of the Cold War with revitalized analytical capabilities in a new Directorate of Intelligence that embodied President Truman’s intention to ensure that the US Government was provided with nondepartmental intelligence based on all available sources. Michael Warner Acting Cnief, CIA History Staff Assessing the Soviet Threat: The Early Coid War Years Preface During World War II, the United States made one of its few original contri- butions to the craft of intelligence: the invention of multisource, nondepart- mental analysis. The Research and Analysis (R&A) Branch of the Office of Strategic Services (OSS) assembled a talented cadre of analysts and experts to comb through publications and intelligence reports for clues to the capa- bilities and intentions of the Axis powers. R&A’s contributions to the war effort impressed even the harshest critics of the soon-to-be dismantled OSS. President Truman paid implicit tribute to R&A in late 1945 when he directed that it be transplanted bodily into the State Department at a time when most of OSS was being demobilized. The transplant failed, however, and the independent analytical capability patiently constructed during the war had all but vanished when Truman moved to reorganize the nation’s peacetime intelligence establishment at the beginning of 1946. “Current” Intelligence Versus “National” Intelligence The Central Reports Staff, home to the analysts in the Central Intelligence Group (CIG), was born under a cloud of confusion in January 1946.' Spe- cifically, no consensus existed on what its mission was to be, although the President’s concerns in creating CIG were clear enough. In the uncertain aftermath of the war, he wanted to be sure that all relevant information available to the US Governinent on any given issue of national security would be correlated and evaluated centrally so that the country would never again have to suffer a devastating surprise attack as it had at Pearl Harbor.” How this was to be accomplished, however, was less clear. The President himself wanted a daily summary that would relieve him of the chere of reading the mounds of cables, reports, and other papers that constantly cascaded onto his desk. Some of these were important, but many were ' The name of the Central Reports Staff was changed in July 1946 to the Office of Research and Evaluations, and again in October 1946 to the Office of Reports and Estimates (ORE), by which name it was known until it was abolished in November 1950. CIA veterans typi- cally use “ORE” as the shorthand name for the analytical office for the whole period 1946- 50. > Truman wrote in his memoirs that he had “often thought that if there had been something like co-ordination of information in the government it would have been more difficult, if not impossible, for the Japanese to succeed in the sneak attack at Pearl Harbor.” Harry S, Tru- man, Memoirs, vol. 2, Years of Trial and Hope (Garden City, NY: Doubleday, 1956), p. 56. duplicative and even contradiciory.* In the jargon of intelligence analysis, Truman wanted CIG to produce a “current intelligence” daily publication that would contain all information of immediate interest to him.* Truman’s aides and advisers, however, either did not understand this or dis- agreed with him, for the presidential directive of 22 January 1946 authoriz- ing the creation of CIG did net mention current intelligence. The directive ordered CIG to “accomplish the correlation and evaluation of intelligence relating to the national security, and the appropriate dissemination within the government of the resulting strategic and national policy intelligence.” Moreover, at the first meeting of the Nationa! Intelligence Authority (NIA) on 5 February, Secretary of State Byrnes objected to the President’s idea of a current intelligence summary from CIG, claiming that it was his responsi- bility as Secretary of State to furnish the President with information o7 foreign affairs.° Byrnes apparently then went to Truman and asked him to reconsider. Admiral Sidney Souers, the first Director of Central Intelligence (DCI), told a CIA historian that Byrnes’ argument ran along the line that such information was not intelligence within the jurisdiction of the Central Intelligence Group and the Directer [of Central Intelligence]. President Truman conceued that it might not be generally considered intelligence, but it was information which he needed and therefore it was intelligence to him. The result was agreement that the daily summaries should be ‘factual statements.’ The Department of State prepared its own digest, and so the President had two summaries on his desk.’ * See Arthur B. Darling, The Central Intelligence Agency: An Instrument of Government to 1950 (University Park: The Pennsylvania State University Press, 1990), p. 81. * Current intelligence was defined in National Security Council Directive No. 3, “Coordina- tion of Intelligence Production,” 13 January 1948, as “that spot information or intelligence of all types and forms of immediate interest and value to operating or policy staffs, which is used by them usually without the delays incident to complete evaluation or interpretation.” See United States Department of Siate, Foreign Relations of the United States 1945-1950, Emergence of the intelligence Establishment (Washington, DC: US Government Printing Office, 1996), p. 1,110. Hereafter cited as Emergence of the Intelligence Establishment. * “Presidential Directive on Coordination of Foreign Intelligence Activities,” United States Department of State, Emergence of the Intelligence Establishment, pp. 178, 179. Also reproduced in Michael Warner, ed., The CIA Under Harry Truman (Washington, DC: Central Intelligence Agency, 1994), pp. 29-32. ° “Minutes of the First Meeting of the National Intelligence Authority,” Emergence of the Intelligence Establishment, p. 328. The National Intelligence Authority was composed of the Secretaries of State, War, and Navy and a representative of the President, Fit. Adm. William Leahy. ’ Darling, The Central Intelligence Agency, pp. 81, 82. This uneasy compromise was reflected in the NIA directives that outlined C!G’s duties. Directive No. |, issued on 8 February 1946, ordered CIG to “furnish strategic and national policy intelligence to the President and the State, War, and Navy Departments. . . .”* National Intelligence Authority Directive No. 2, issued the same day, ordered the DCI to give “first priority” to the “production of daily summaries containing factual statements of the significant developments in the field of intelligence and operations related to the national security and to foreign events for the use of the President... .””’ In practice, this approach proved unworkable. Without any commentary to place a report in context, or to make a judgment on its likely veracity, the early Daily Summaries probably did little but confuse the President. An alarming report one day on Soviet troop movements in Eastern Europe, for example, would be contradicted the next day by a report from another source. Everyone invo' ved eventualiy realized the folly of this situatio., and analytical commentaries began to appear in the Daily Sumniaries in December 1946—episodicaily at first, and then regularly during 1947. The Weekly Summary, first published in June 1946 on the initiative of the Central Reports Staff itself, was also supposed to avoid interpretative commentary, but its format made such a stricture difficult to enforce. From its inception, the Weekly Summary proved to be more analytical than iis Daily Summary counterpart * National Intelligence Authority Directive No. |, “Policies and Procedures Governing the Central Intelligence Group.” 8 February 1946, Emergence of the Intelligence Establishment, pp. 329-331. After CIA was established, National Security Council Intelligence Directive No. |, “Duties and Responsibilities,” issued on !2 December 1947, again ordered the DCI to produce national intelligence, which the Directive stated should be “officially concurred in by the Intelligence Agencies or shall carry an agreed statement of substantial dissent.” National Security Council Intelligence Directive No. 3, 13 January 1948, gave CIA the authority to produce current intelliyence: “The CIA and the several agencies shall produce and disseminate such current intelligence as may be necessary to meet their own internal requirements or external responsibilities.” See Emergence cf the Intelligence Establishment, pp. 1,119-1,122; 1,109-1,112. * National Intelligence Authority Directive No. 2, “Organization and Functions of the Cen- tral Intelligence Group,” 8 February 1946, Emergence of the Intelligence Establishment, pp. 331-333. Interestingly, Sovers, who drafted both NIA Directive | and Directive 2, continued to believe that CIG’s principal responsibility was the production of strategic and national policy intelligence. In a memorandum to the NIA on 7 June 1946, Souers wrote that the “primary function of C.1.G. in the production of intelligence . . . will be the prepara- tion and dissemination of definitive estimates of the capabilities and intentions of foreign countries as they affect the national security of the United States.” “Memorandum From the Director of Central Intelligence to the National Intelligence Authority,” 7 June 1946, Emergence of the Intelligence Establishment, p. 361. The Confusion Surrounding “National” Intelligen: e Similar disarray surrounded CIG’s responsibilities in the production of “strategic and national policy intelligence.” The members of the Intelli- gence Community simply could not agree on the policies and procedures that governed the production of this type of intelligence. Most of those involved seemed to believe that national intelligence should be coordinated among all the members of the Intelligence Community, that it should be based on all available information, that it should try to esti.nate the inten- tions and capabilities of other countries toward the United States, and that it should be of value to the highest policymaking bodies. The Gvvil was in the details. High-ranking members of the intelligence and policy communities debated, without coming to a consensus, most aspects of the estimate production process, including who should write them, how other agencies should participate in the process if at all, and how dissents should be handled. Some of this reflected genuine disagreement over the best way to organize and run the Intelligence Community, but it also involved concerns about bureaucratic power and prerogatives, especially those of the Director of Central Intelligence. the newcomer to the Intelli- gence Community. Even the definition of “strategic and national intelli- gence” had implications for the authority of the DCI and thus was carefully argued over by others in the Community." DCI Vandenberg eventually got the NIA to agree to a definition in February 1947, but it was so general that it did little to solve the problems that abounded at the working leve!.'' Ray Cline, a participant in the process of producing the early estimates, wrote in his memoirs that It cannot honestly be said that it [ORE] coordinated either intelligence activities or intelligence judgments; these were guarded closely by Army, Navy, Air Force, State, and the FBI. When attempts were made to prepare agreed national estimates on the basis of inielligence available to all, the coordination process was interminable, dissents were the rule rather than '’ Emergence of the Intelligence Establishment, p. 367. '' The NIA agreed that “strategic and national policy intelligence is that composite intelligence, interdepartmental in character, which is required by the President and other high officers and staffs to assist them in determining policies with respect to national plan- ning and security. . . . It is in that political-economic-military area of concern to more than one agency, must be objective, and must transcend the exclusive competence of any one department.” “Minutes of the 9th Meeting of the National Intelligence Authority,” 12 February 1947, Emergence of the Intelligence Establishment, p. 492. After the establish- ment of CIA, National Security Council Directive No. 3, 13 January 1948, similarly defined national intelligence as “integrated departmental intelligence that covers the broad aspects of national policy and national security, is of concern to more than one Department . . . and transcends the exclusive competence of a single department. . . .” See Emereenc e of the Intelligence Establishment, p. 1,111. the exception, and every policymaking official took his own agency's intelligence appreciations aiong to ihe White House to argue his case. The prewar chaos was largeiy recreated with only a little more lip service to central coordination." In practice, much of the intelligence produced by ORE was not coordinated with the other agencies; nor was it based on all information available to the US Government. The Daily and ‘eekly Summaries were not coordinated products, and, like the other publications produced by ORE, they did not contain information derived from communications intelligence.'* The Review of the World Situation, which was distributed each month at meet- ings of the National Security Council, became a unilateral publication of ORE after the first two issues. * The office's ad hoc publications, such as the Special Evaluations and Intelligence Memorandums, were rarely coordi- nated with the other agencies. By contrast, the “ORE” series of Special Estimates were coordinated, but critics nonetheless condemned many of them for containing trivial subjects that fell outside the realm of “strategic and national policy intelligence.”' Whatever CIG’s written orders, in practice the President's interest in the Daily Summaries, coupled with the limited resources of the Central Reports Staff, meant that the production of current intelligence came to dominate the Staff and its culture. National estimative intelligence was reduced to also-ran status. An internal CIG memo stated frankly that “ORE Special Estimates are produced on specific subjects as the occasion arises and within the limits of ORE capabilities after current intelligence requirements '* Ray S. Cline, Secrets, Spies, and Scholars: Blueprint of the Essential CIA (Washington, DC: Acropolis Books, 1976), pp. 91, 92. Cline rose to become Deputy Director for Intelli- gence (DDI) between 1962 and 1966. Another veteran of the period, R. Jack Smith, who edited the Daily Summary, made the same point in his memoirs, The Unknown CIA (Washington: Pergamon-Brassey’s, 1989), p. 42: “We were not fulfilling our primary task of combining Pentagon, State Department, and CIA judgments into national intelligence 2stimates. . . . To say it succinctly, CIA lacked cloui. The military and diplomatic people ignored our statutory authority in these matters, and the CIA leadership lacked the power to compel compliance.” Smith also served as DDI, from 1966 to 1971. '* Smith, The Unknown CIA, pp. 34, 35. ORE began receiving signals intelligence in 1946 and was able to use it as a check against the articles it included in the Summaries. Security concerns prevented its broader use. Signals intelligence was sent to the White House by the Army Security Agency (from 1949 on, the Armed Forces Security Agency) during this period. CIA did not begin including communications intelligence in the successor to the Daily until 1951. '* The delays involved in interagency coordination made it difficult to mect the publication deadline while still including the most recent e. ents in its contents. George S. Jackson, Office of Reports and Estimates, 1946-1951, Miscellaneous Studies, HS MS-3, vol. 3 (Washington, DC: Central Inteiligence Agency, 1954), pp. 279-287. National Archives and Records Administration, Record Group 263, History Staff Source Collection, NN3-263-95-003. '* See the discussion of the Dulles-Jackson-Correa Report on page 10. are met.” It went on to note, “Many significant developments worthy of ORE Special Estimates have not been covered . . . because of priority production of current intelligence, insufficient personnel, or inadequate information.”' ®T his remainec ue even after ihe Central Reports Staff evolved into the Office of Reports and Estimates (ORE) in CIA."” If the analysts in CIG, and then CIA, had only to balance the competing demands of current and national intelligence, their performance might have benefited. As it happened, however, NIA Directive No. 5 soon gave the analysts the additional responsibility of performing “such research and analysis activities” as might “be more efficiently or effectively accom- plished centrally.”' * In practice, this meant that the analysts became respon- sible for performing basic research as well as wide-ranging political and economic analysis. To accommodate this enhanced mission, functional analysis branches for economics, science, transportation, and map intelli- gence were established alongside the existing regional branches." A high-ranking ORE officer of the period, Ludwell Montague, wrote that this was a deliberate, but covert, attempt to transform ORE (or CRS, a staff designed expressly for the production of coordinated national intelligence) into an omnicompetent . . . central research agency. This attempi failed, leaving ORE neither the one thing nor the other. Since then, much ORE pro- duction has proceeded, not from any clear concept of mission, but from the mere existence of a nondescript contrivance for the production of nonde- script intelligence. All our efforts to secure a clear definition of our mission have been in vain.” ‘© Memo from Chief, Projects Division to Assistant Director, R&E, “Proposed Concept for Future CIG Production of Staff Intelligence,” 1 July 1947. CIA History Staff Job 67-00059A, Box 2, Confidential. Nevertheless, during it: existence ORE did produce over 125 estimates, 97 of which were declassified in 1993 anc 1994 and deposited in the National Archives. '’ This point is made repeatedly throughout George S. Jackson, Office of Reports and Estimates, 1946-1951. Jackson himself served in the office du. ing the period of this study. '* National Intelligence Authority Directive No. 5, “Functions of the Director of Central Intelligence,” 8 July 1946, Emergence of the Intelligence Establishment, p. 392. '* The Scientific Intelligence Branch of ORE was established in January 1947 and shortly thereafter incorporated the Nuclear Energy Group, which had been in charge of atomic energy intelligence in the Manhattan Project, within its ranks. At the end of 1948, the branch was separated from ORE and elevated to office status, becoming the Office of Scientific Intelligence. *” Montague to Babbitt, “Comient on the Dulles-Jackson Report,” !1 February 1949. National Archives and Records Administration, Record Group 263, History Staff Source Collection, HS/HC 450, NN3-263-94-010, Box 14. Montague’s reference to a “deliberate but covert” attempt to increase the responsibility of ORE refers to the efforts of DCI General Hoyt Vandenberg to boost himself, and CIG as a whole, into a dominant position in the Inteiligence Community. Opposition from the other departments largely scuttled his attempts in this direction. See Emergence of the Intelligence Establishment, p. 366.