Origins of CIA's Analysis of the Soviet Union
Donald P. Steury
In the forefront of President Harry Truman's mind as he signed the order establishing the Central Intelligence Group (CIG) in 1946 was concern that the permanent, peacetime intelligence organization he was creating not serve as the cornerstone of an American "Gestapo," Nazi Germany's pervasive and oppressive secret police organization. At the same time, Truman felt the need to respond to widespread concern that the lack of an overarching national intelligence organization in the United States had been responsible for the strategic surprise at Pearl Harbor just five years before. And his own experience in the year he had been President showed him the policymaker's need for regular, timely intelligence reporting. Truman's way out of this dilemma was to choose a minimalist solution. The CIG was created as "a sort of holding company whose main functions would be the coordination of departmental intelligence."1
The Director of the CIG was given a permanent intelligence staff of just 29 (17 of whom were on loan from other departments).2 He was dependent on the Departments of State, War, and Navy both for staff and for funding. As created, CIG had two functions: the planning and coordination of all federal intelligence projects and the making of high-level estimates of foreign situations for the President and senior government officials. Although meant to be "the very last word in accuracy and timeliness," the estimates were not supposed to result from independent research but to be the product of the correlation and evaluation of analyses produced by the "departmental" intelligence organizations.3
As Sherman Kent presciently observed in 1946, "The CIG... [is] in for difficulties. If it has a soul to call its own, this and its heavy responsibilities are about all it can claim exclusive ownership to."4 The original concept of CIG may have been "reasonable and derived from real informational needs, [but] institutional resistance made implementation [of this concept] virtually impossible." The military services and the Department of State jealously guarded their preexisting control of information and their role as policy advisors to the President. Under such circumstances, CIG's original mission was "an exercise in futility."5 The military resented having to provide military data to a civilian agency and felt that "civilians could not understand, let alone analyze military intelligence data."6 Although the War and Navy Departments eventually assigned officers to CIG, they never granted CIG access to US military data. No less hostile to CIG's intelligence-producing authority was the Department of State, which almost immediately challenged CIG on the issue of access to the President. When Truman asked CIG for a daily intelligence summary, Secretary of State James F. Byrnes insisted on his department's prerogative to provide the President with daily policy analysis. The result: Truman received daily summaries from both CIG and the Department of State.7
It is true that in Spring 1946, the National Intelligence Authority--the Secretaries of State, War, and Navy and the President's personal representative, who were to supervise the CIG's Director of Central Intelligence (DCI)--authorized the CIG to conduct independent research and analysis "not being presently performed" by other departments.8 By October of that year, DCI Lt. Gen. Hoyt S. Vandenberg had created the Office of Reports and Estimates (ORE), with a staff of more than 300 intelligence professionals and clericals-- sufficient to make CIG an independent intelligence producer. At the same time, CIG received its own clandestine collection capability, the Office of Special Operations (OSO), created by "returning" the War Department's Strategic Services Unit (SSU).9 With the SSU, CIG received a number of trained personnel, originally from the World War II Office of Strategic Services, with experience in both analysis and operational matters. Despite this, ORE depended on the Department of State for raw intelligence and relied primarily on unclassified sources of material--in part because OSO's intelligence product was highly compartmented and not accessible to ORE, and in part because ORE still lacked ready access to military intelligence. Moreover, the President's interest drove ORE to concentrate on producing a daily summary of international developments. Even as CIG discovered that it was easier to collect and analyze its own data than coordinate the work of obstructionist departments, it also learned that Truman liked and expected to receive its Daily Intelligence Summary. The pressure of events and the priority of responding to the President thus focused ORE's efforts on current reporting rather than on long-range forecasting.10
The result was that CIG "drifted from its original purpose of producing coordinated national estimates to becoming primarily a current intelligence producer."11 CIG produced just four estimates on the Soviet Union in 1946, two of which were analyses of Soviet Bloc propaganda broadcasts.12 Another, Soviet Capabilities for the Development and Production of Certain Types of Weapons and Equipment, was just two pages long and contained the first of a series of wrong-headed projections concerning the development of Soviet atomic weapons capabilities.13 The very first estimate--ORE 1: Soviet Foreign and Military Policy--represented exactly the kind of "high-level estimate of foreign situations" CIG was created to produce, but it stands out as virtually unique among the crop of estimates ORE turned out.
The passage of the National Security Act and the creation in 1947 of the Central Intelligence Agency (CIA) did not greatly alter the situation. Under the new DCI, RAdm. Roscoe Hillenkoetter, "The Agency experienced un-directed evolution in the area of intelligence, never fulfilling its coordination function, but developing as an intelligence producer."14 The list of intelligence estimates published in 1947 is rather longer than that for 1946 and contains a number of high-level estimates, but topics of immediate concern predominate. Of the 15 estimates declassified to date, six are in the nature of "situation reports" describing developments in various countries. Four others discuss the ongoing implementation of Soviet regional policies or likely Soviet reactions to US actions under consideration. Perhaps the most comprehensive estimate is ORE 14, Future Soviet Participation in Long-Range International Air Transport.15
None of these documents represent judgments outside the purview of either CIG or CIA, and all contain information of importance to the formulation of US foreign policy. But the predominance of such a current, situational focus suggests a preoccupation with "answering the mail," to the detriment of the longer range, more comprehensive intelligence assessments which the nation's central intelligence organization might have been expected to produce. Nowhere does one see the kind of comprehensive, formative intelligence documents produced by CIA's Office of National Estimates (ONE) in the 1950s and 1960s or by the National Intelligence Council (NIC) beginning in the 1970s. Moreover, the predominantly political tone of many of the estimates (especially the "situation reports") suggests a duplication of intelligence functions better performed by the Department of State.
In reviewing the intelligence process for the National Security Council, the 1949 Dulles-Jackson-Correa report somewhat wistfully concluded, "The principle of the authoritative National Intelligence Estimate (NIE) does not yet have established acceptance in the government. Each department still depends more or less on its own intelligence estimates and establishes its plans and policies accordingly."16 The report blamed ORE for not asserting itself enough in the estimates process and for failing to fulfill its mission as a coordinating intelligence body. That indictment was largely in line with the facts, but it failed to allow for what CIA veteran Ludwell Montague called "the recalcitrance and incompetence of the departmental intelligence agencies."17 Lack of cooperation on the part of the departmental agencies only isolated ORE further, contributing to its general failure to function as a producer of coordinated, high-level estimates.
Nonetheless, the problem was deeper still, interwoven into the fabric of the newly created intelligence organization. In fact, neither CIG nor the early CIA was capable of meeting America's postwar intelligence requirements. CIG had been created to prevent the kind of strategic surprise that had brought the United States into World War II. But by 1946, although avoiding another Pearl Harbor was a paramount requirement of the postwar American Intelligence Community, strategic warning was only part of a wide spectrum of intelligence requirements. The experience of more than four years of total war was formative for postwar American strategic culture. By the end of World War II, a nation's war-making capacity was seen as but the expression of its total potential economic and military power defined in the broadest possible terms.
The shadow of Soviet military power settled across the European continent as the Soviet Union first infiltrated and then ruthlessly imposed dictatorial communist regimes upon the peoples of Eastern Europe, all the while moving to confront the Western Allies in Germany, Greece, Iran, and, eventually, Korea. An accurate appraisal of the full military and economic potential of the Soviet Union came to be viewed as an essential component of the role of the US Intelligence Community in assessing the burgeoning Soviet "threat"-- one fully as important as achieving an accurate forecast of Soviet intentions.
The dimensions of the postwar intelligence problem were mapped out by Sherman Kent briefly in a 1946 Yale Review article 18 and then comprehensively in a 1949 treatise, Strategic Intelligence for American World Policy. Kent focused the lens of American intelligence on what he called the strategic stature of a potential enemy. By this he meant a nation's ability to influence an international situation in which the United States had a "grand strategic interest."19 Kent saw a nation's strategic stature as having three components, the most important of which was its war potential--the fully mobilized potential of its society, economy, and military to wage war. In any given situation, a detailed understanding of a nation's strategic stature was important, not only as a measure of what courses of action were possible (e.g., its capabilities--military or otherwise--for action), but as one indicator of its likely courses of action. For example, Soviet deployment of massive ground and tactical air forces capable of deep-strategic (e.g., offensive) operations to Eastern Europe was an indicator, not only of the political and strategic importance Moscow attached to the region, but of likely Soviet strategy in the event of war. A nation's war potential would only be fully employed in wartime, but Kent warned that in the developing Cold War the gap between a peacetime posture and one fully mobilized for war was narrowing rapidly.20
Kent's voice was the most authoritative to speak on intelligence matters, but it certainly was not the only one. America's strategic culture blossomed intellectually in the postwar period, and questions of strategic intelligence received considerable attention. Kent himself was strongly influenced by what probably was the first postwar book to be published on the subject, George S. Pettee's Future of American Secret Intelligence. An intelligence veteran of the wartime Foreign Economics Administration, Pettee concentrated almost entirely on the industrial and socioeconomic elements of national power.21 To cope with the complexities and dangers of a world arena shaped by industrialization, Pettee called for a postwar intelligence organization with far more emphasis on research and analysis than ever in the past. His influence on Kent is to be found in the latter's conceptualization of national war potential and even in the basic organizational schema applied to Strategic Intelligence.
Pettee's little book has all but disappeared from the American political consciousness, but his contribution was part of a large body of work influenced by the experience of total war. In the postwar world, foreign intelligence analysis meant building a comprehensive picture of state and society--one that demanded a significant, ongoing research effort. This was particularly true as American intelligence confronted the Soviet Union--a nation combining a large, offensively minded, standing military; rich natural resources; and a vast, if not necessarily modern, industrial base with a powerful and effective security apparatus that kept secret virtually every aspect of Soviet industrial, military, and technological development. It would have been impossible to achieve a strategically significant body of knowledge about the Soviet Union without the development of sophisticated tools of intelligence collection and analysis. Whatever their talents, the CIG's 29 intelligence officers simply lacked the time or resources to perform analytical work on such a level. In effect, they lacked an institutional basis for the intellectual authority they were expected to wield over the national intelligence process. ORE was perhaps better placed, but it was confounded by the institutional difficulties of inserting itself into a preexisting and bureaucratically entrenched national security establishment.
To some extent, ORE was also attempting to impose itself on a process that already was under way. The War Department's Joint Intelligence Committee (JIC) had been created in 1941 to counter Maj. Gen. William J. "Wild Bill" Donovan's appointment as Coordinator of Information, the first step in creating the Office of Strategic Services. In 1945, the JIC had begun applying its experience analyzing the Nazi military-industrial base to estimates of Soviet war potential and to projections of likely postwar Soviet behavior. JIC analyses encompassed the political, economic, and ideological dimensions of Soviet power as well as the more traditional military aspects of weapons development and war planning. 22
By 1946, the War Department's estimative process had acquired considerable momentum. Thus, when the DCI on 29 April issued a directive (CIG 8) calling for "production of the highest possible quality of intelligence on the USSR in the shortest possible time," it was the JIC, rather than ORE, that became the focal point of the analytical effort. Although the directive expressed the intention that "CIG would take over formal sponsorship of the project at the earliest possible moment," in practice CIG was virtually excluded from the process.23 Intelligence actually was to be produced by a Working Committee comprising representatives of the Department of State, G-2 (Army Intelligence), the Office of Naval Intelligence, and Air Force Intelligence. CIG was not represented. The Working Committee was to be responsible for creating a digest of factual strategic intelligence on the Soviet Union, to be compiled as a Strategic Intelligence Digest. Based on the digest, Strategic Intelligence Estimates were to be prepared by member agencies as needed to meet their own requirements. CIG was to act in a "supervisory capacity"24 and function as adjudicator between departments. The military, however, resented having to defer to CIG to process and distill raw intelligence data.25 CIG's impact on the process thus was minimal.26 ORE's own response to the CIG 8 directive was ORE 1, Soviet Foreign and Military Policy, a landmark estimate, but one that failed initially to achieve the prominence or impact of JCS 1696, a much larger document, produced by the War Department's JIC, and published under the aegis of the Joint Chiefs of Staff.
Ludwell Montague, principal author of ORE 1, was critical of the methodology employed in JCS 1696, and the document's balance and alarmist conclusions about Soviet intentions have since been questioned.27 Nonetheless, JCS 1696 responded to prevailing concerns in the Washington national security establishment, and it was to this document that the White House staff turned in drafting its own appraisal for the President.28 Truman's reaction was dramatic. "This is so hot," he concluded, "it could have an exceedingly unfortunate impact on our efforts to try to develop some relationship with the Soviet Union."29 While ORE 1 was more concise and perhaps more balanced in its analysis of Soviet intentions, it did not offer much that was not in JCS 1696. In short, the new kid on the block still had a lot to prove.
The opportunity to do so was provided by the Soviet Union. On 21 November 1947, the Soviet Military Governor in Germany, Marshal Sokolovskiy, opened a meeting of the Allied Control Council with a violent outburst attacking the Western Allies. In December, Soviet Foreign Minister Vyacheslav Molotov disrupted a Quadripartite Foreign Minister's Conference in London and, in January, Soviet guards began regularly harassing trains transiting East German territory en route to the Allied garrison in West Berlin. Over the winter, Gen. Lucius D. Clay, the American Military Governor in Germany based in Berlin, began noticing an increased Soviet security presence in meetings with his military counterparts. Meanwhile, local intelligence officers reported recurrent consultations between the Soviet-controlled zone of Germany and Moscow. On 20 January 1948, Sokolovskiy rejected Clay's plans for currency reform inside occupied Germany.30
Finally, on 5 March 1948, Clay felt compelled to voice his growing unease in a cable to Washington:
For many months, based on logical analysis, I have felt and held that war was unlikely for at least 10 years. Within the last few weeks, I have felt a subtle change in Soviet attitude, which I cannot define but which now gives me a feeling that it may come with dramatic suddenness. I cannot support this change in my own thinking with any data or outward evidence in relationships other than to describe it as a feeling of new tenseness in every Soviet individual with whom we have official relations. I am unable to submit any official report in the absence of supporting data but my feeling is real. You may advise the chief of staff of this for whatever it may be worth if you feel advisable.31
Unbeknownst to Clay, that same day the American Commandant in Berlin, Col. Frank Howley, decided to express similar misgivings in another cable to Washington:
After weeks of calm, last 2 Kommandatura [the quadripartite governing council in Berlin] meetings, 26 February and 2 March, showed such increased Soviet violence in attacks that it is believed that General Kotikov, Senior Soviet Member is acting under new instructions. Attacks are thoroughly prepared, unprovoked, and often unrelated to any incidents of the meeting.
...The apparent pattern, with reference to Soviet intentions in Berlin, which may be temporary or permanent, includes the following elements:
--Effort to build case that quadripartite government is unable to operate in Berlin...
--Complete opposition to agreement of any kind in quadripartite meetings.32
According to Secretary of Defense James V. Forrestal's biographer, Walter Millis, Clay's telegram "fell with the force of a blockbuster bomb."33 Howley's cable, dispatched independently later that day, seemed only to magnify the crisis. Although Clay later denied that he intended the cable as a war warning,34 it was interpreted as such inside the Pentagon. That same day, Gen. Stephen Chamberlin, Chief of Army Intelligence (G-2), hand-carried the "war warning" to Gen. Omar N. Bradley, Chairman of the Joint Chiefs of Staff. Secretary of the Army Kenneth C. Royall also saw it that afternoon; his first response was to ask how long it would take to get a number of atomic bombs to the Mediterranean, ready for use should the Soviets initiate a military action. Meanwhile, G-2 formed a task force under Col. Riley F. Ennis to begin a crash estimate of Soviet intentions.35
Incredibly, although the cable was received with the utmost alarm inside the Pentagon, the Army was in no hurry to inform anyone outside the Department of Defense. Not until three days later, on 8 March, did Secretary of Defense Forrestal brief a closed-door session of the Senate Armed Services Committee on Clay's telegram. On 11 March, Gen. Chamberlin, the G-2, called DCI Hillenkoetter to request a meeting of the inter-departmental Intelligence Advisory Committee the next day.36 Not until that meeting did representatives of the Office of Naval Intelligence, Air Force Intelligence, or the Department of State see Clay's cable, although by that time they had gleaned through the rumor mill some idea of what was happening.37 On reading the cable, Director of Naval Intelligence Thomas Inglis noted that "this was the very function for which CIA had been established," and proposed that Hillenkoetter appoint a CIA representative to chair an ad hoc committee to study the situation and prepare an estimate of Soviet intentions.38 With a stroke, Admiral Inglis transformed what up to that moment had been "an Army matter" into a national intelligence problem.39
As Washington mobilized to deal with what it still perceived to be a crisis, intelligence officers stationed in Europe were polled for any supporting data. With this, CIA's Berlin Operations Base first heard of Clay's cable.40 Surprised by the apparent extremity of the situation, Dana Durand and Peter Sichel of CIA visited Clay's intelligence chief in the Office of the US Military Governor in Germany. All agreed that further Soviet measures short of war were likely, but that war itself was unlikely--an opinion that prevailed generally throughout Allied intelligence establishments in Europe.41 This consensus in the field took the edge off Clay's "war warning" and reduced the sense of immediacy prevailing in Washington.
The next day, Saturday, 13 March, the ad hoc committee met for the first time under the chairmanship of CIA's DeForrest Van Slyck, an analyst from ORE's Global Survey Group. Hillenkoetter left Van Slyck to run the meeting, but bustled in and out with trays of coffee and sandwiches.42
CIA's Van Slyck and G-2's Col. Ennis later were identified as "the principal protagonists" in the meeting--those with the most timely information and staff on the ground in Europe. The Army leadership had seized on Clay's cable as a means of justifying increases to its budget, which was about to come up before Congress. While Ennis was thrashing out an intelligence response with Van Slyck, his colleagues in G-2 were drafting an "Estimate of the World Situation" that called for augmenting the regular Army and recommended bringing "our machinery for general mobilization to an alert status." This Army document went on to warn that "The risk of war is greater now...than was the case six months ago...[and that]...war will become increasingly probable.... The Soviet Armed Forces...overshadow the whole of Europe and most of Asia.... The United States has no forces in being which could prevent the Soviet [sic] overrunning most of Eurasia.... Present forces...are incapable of offering more than a weak and unorganized delaying action in any of the likely theaters."43
In the ad hoc committee, G-2--supported by Air Force Intelligence--was equally dire in its conclusions but was restrained by the need to achieve a consensus supported by evidence. (The Army, in its "Estimate of the World Situation," admitted that it lacked conclusive evidence of an immediate Soviet intention to initiate hostilities.) Ennis led off by demanding that the estimate include a recommendation for universal military training. Van Slyck angrily refused, saying he was "running an intelligence estimates committee, not an appropriations committee." Considerable difficulty was experienced in reaching agreement on the language to be used. Although none of the intelligence organizations argued that war was likely or imminent, the Army G-2 and Air Force Intelligence refused to agree to a direct statement that war was unlikely. Nonetheless, by the close of business on Sunday, a unanimous agreement had been reached on a statement that war was improbable for at least the next 60 days. Van Slyck drafted a response to be given to the Intelligence Advisory Committee (IAC) the next day.44
We now know that the Soviet premier, Joseph Stalin, after consulting East German leaders, had decided to initiate actions designed to push the Western Allies out of Berlin over the course of 1948.45 The results were nothing like what he expected. Indeed, had Stalin deliberately set out to increase US military spending, he could not have chosen a more propitious time. The Pentagon was on the verge of requesting a supplementary budget authorization for fiscal 1948-1949. At stake for the Army was a general expansion and universal military training. The newly created Air Force hoped for expansion to 70 combat groups. The Navy was looking for continued funding for its postwar aircraft carrier force, based on the first of a new generation of supercarriers.46 On Thursday, 11 March, the day before the first IAC meeting, Gen. Bradley and the service chiefs left Washington for a meeting on the proposed budget at Key West, Florida. They returned, having decided to make the supplemental budget request, to face an ad hoc committee estimate that the Soviet Union was not ready for war. The service chiefs, supported by Air Force Intelligence and the Army G-2, rejected Van Slyck's draft. Only Admiral Inglis, the Director of Naval Intelligence, stood fast behind the estimate his department had helped write. 47
On 15 March (Monday) Van Slyck presented the ad hoc committee's conclusions to the IAC, but it would not agree to the estimate. Hillenkoetter, however, had been to see the President and returned with a demand for answers--definitive, yes or no answers; that morning; with no elaboration--to three questions:
1. Will the Soviets deliberately provoke war in the next 30 days?
2. In the next 60 days?
3. In 1948?48
After some debate, the IAC drafted the following answers, consolidating (1) and (2) and deferring (3):
I. An examination of all pertinent available information has produced no reliable evidence that the USSR intends to resort to military action within the next 60 days.
II. It is not believed that the USSR will resort to military action within the next 60 days.49
Theodore Babbitt, CIA's chief of current intelligence, hand-carried the answers to the White House in the form of a CIA estimate while discussion continued in the IAC meeting. At this point, the G-2 raised again the issue of universal military training, but further delay was avoided by agreeing to deal with the issue in a separate document.50 On 16 March, a fuller statement--allowing for the possibility that "some miscalculation or incident" might result in war--was issued as Intelligence Memorandum 21.51
A series of escalating Soviet provocations, culminating in the blockade of Berlin and the Allied airlift, kept the ad hoc committee alive until the end of the year. It produced a series of estimates, beginning with ORE 22-48, Possibility of Direct Soviet Military Action During 1948 (2 April 1948), followed by two supplementary updates.52 The gist of all three estimates was that the Soviet Union was unlikely to deliberately initiate war in the foreseeable future, despite its military preponderance in Europe. A further supplement, ORE 58-48, expanded this argument with a careful evaluation of the risks, advantages, and disadvantages conquest of Western Europe would bring to the Soviet Union. This estimate concluded that the potential risks of such an action were so great that the Soviet leaders "would be unlikely to undertake this operation...unless they anticipated an attack or became involved in military action through accident or miscalculation."53
Although the process was attended by considerable difficulty, in the end the ad hoc committee served its purpose. The analyses it prepared began with a short-term projection of Soviet intentions but rapidly evolved into an effort to place Soviet actions into the much broader context of the strengths and weaknesses of their overall strategic posture. The result was a much more balanced estimate that gave due weight to the restraints operating on Soviet military power, while acknowledging the undoubted military superiority the Soviets enjoyed in Europe. The Cassandra-like tone of the Army's "Estimate of the World Situation" shows what could have been expected had the departmental intelligence agencies been allowed to function without the benefit of a "national" consensus.
As the Berlin crisis deepened, the ad hoc committee Estimates proved to have both immediate and long-term relevance for both policymakers in Washington, and those stationed in Europe. Thus, when Marshal Sokolovskiy's deputy notified his Western counterparts on 30 March that, effective the following midnight, all Allied traffic through the Soviet zone would be forced to submit to inspection, both Washington and the Office of the US Military Governor in Germany were reasonably sure they faced a political challenge, not an effort to provoke a war. By the same token, when all ground access to Berlin was severed three months later, Truman could be reasonably certain that the city could be supplied by airlift without deliberate interference from Soviet air defenses. That confidence was no doubt shaky at first, but by winter the confrontation over Berlin was clearly a struggle of endurance that, barring accident, was not expected to escalate into war. Thus it was that, when the Dulles-Jackson-Correa Survey Group reported to the National Security Council, they singled out the actions of the ad hoc committee as "the most significant exception to a rather general failure...in national estimates.... This case illustrated that, when properly used, the existing interdepartmental arrangements can, under the leadership of the Central Intelligence Agency, provide the President and top policy-makers with an authoritative intelligence estimate."54
The key to the success of the ad hoc committee was CIA's ability to exert intellectual authority over a process that closely involved the departmental agencies. That Van Slyck was able to do so was less due to CIA's position inside the Washington national security hierarchy than to the circumstances under which he assumed chairmanship of the committee. The Berlin crisis was so obviously a national intelligence problem that it transcended the bureaucratic lines that had divided the Intelligence Community for the previous two years. It was at once a crisis that was developing daily--even hourly--and an enduring confrontation with profound implications for national security policy and the survival of the Western alliance. Understanding Soviet intentions meant anticipating their actions on a daily basis, while comprehending their behavior in the context of a long-term national strategy. The alternative to a "national" approach to the intelligence problems presented by Berlin was, quite simply, paralysis. Even so, as the Dulles-Jackson-Correa survey noted, the event was "largely fortuitous," and quite dependent on the statesmanlike judgment of the Director of Naval Intelligence, Admiral Inglis; CIA just happened to be there.55
The fuller implications of the estimates prepared over 1948 did not become apparent until 1950. In moving to implement the recommendations of the Dulles-Jackson-Correa report, Hillenkoetter's successor, Lt. Gen. Walter Bedell Smith, adopted the workings of the ad hoc committee as an example to be replicated in the organization of CIA.56 The 1950-51 "Smith reforms" dissolved ORE and replaced it with a Board of National Estimates, headed by William L. Langer and then Sherman Kent, both of whom brought to bear considerable authority from their experience in the Office of Strategic Services. The Board of National Estimates was given analytical support by an Office of National Estimates, with the intention that it would rely exclusively on the departmental agencies for research support, although, perhaps inevitably, it became a major research organization in its own right. In January 1952, both were made part of a newly established Directorate of Intelligence (DI).57 CIA's own particular contribution derived from analyses performed by the Office of Scientific Intelligence (OSI), which produced intelligence on weapons research and long-term scientific developments, and the Office of Research and Reports (ORR), responsible for analysis of the Soviet economy. This latter responsibility was assumed from the State Department in exchange for acknowledging State's primacy in political analysis.
In theory, CIA assumed responsibility for the Soviet economy as one of those national topics for research and analysis "not being presently performed" by other departments. In fact, both ORR and OSI gave CIA considerable analytical depth, completely independent of the national intelligence process. Officers in ORR quickly demonstrated that they interpreted their mandate for economic intelligence in the broadest possible terms. As the founding head of ORR, MIT's Max Millikan, noted somewhat wryly, "The distinction between economic and military or political, or scientific intelligence is wholly arbitrary."58 To Millikan, the degree to which a country was able to mobilize its economy for military purposes was a profound indicator of likely intentions. The first function of economic intelligence was "to estimate the magnitude of possible present or future military or other threats to ourselves and our allies," a task that included estimating "the character and location of possible present or future military or other threats...[and]...the intentions of the USSR or any other potential enemy." "A potential enemy can undertake successfully only those military operations which its economy is capable of sustaining," he wrote in 1951. In the short run, operations might be determined by the manpower available and stocks of weapons and materiel, but in the long run, "the military potential for anything but the briefest campaign...[depends]...upon the total economic resources available to a nation, including those necessary to support the civilian economy as well as those necessary to produce and operate the instruments of war."59
With a mandate this broad, ORR was able to build a comprehensive picture of Soviet war potential that provided a constant, reliable check upon analyses prepared in the departmental agencies. In effect, ORR institutionalized and provided a sustainable intellectual base for the authority wielded by DeForrest Van Slyck in 1948.
A panel moderated by Frederick P. Hitz, Lecturer in Public and International Affairs at Princeton University and former Inspector General at CIA discussed Donald Steury's paper and provided views on CIA's early analytic efforts on the Soviet Union. The panelists were R.M. Huffstutler, Chief Operating Officer of the Aegis Corporation and former Executive Director and Director of the Office of Soviet Analysis at CIA; Dr. Melvyn Leffler, Edward Stettinius Professor of History at the University of Virginia; and former Dean of the College of Arts and Sciences at the University of Virginia; and Jack F. Matlock, former US Ambassador to Moscow and currently the George Kennan Professor at the Institute for Advanced Study at Princeton University.
Commentator Rae Huffstutler discussed the development of CIA's analytical capabilities in the late 1950s through the early 1980s. Huffstutler agreed with Steury that neither CIG nor CIA were capable, in the early Cold War period, of preventing another Pearl Harbor. He contrasted CIA's analysis in the 1950s with the far more detailed and sophisticated estimates of the 1970s. Huffstutler described four major developments that shaped CIA's analysis over that 20 year period: (1) the revolution in technical collection and analysis; (2) the change in Department of Defense force-planning guidelines in the early 1960s under Secretary of Defense Robert McNamara; (3) the emerging preeminence of the national estimates process; and (4) the growth of a cadre of analysts at CIA and elsewhere in the Intelligence Community.
According to Huffstutler, the development of an extraordinary national technical collection and processing capability allowed the CIA and other agencies to address key questions regarding Soviet strategic capabilities. Secretary of Defense McNamara's demand that the threat portion of force-planning documents be based on National Intelligence Estimates (NIEs), changed strategic military assessments from departmental edicts to debates chaired by CIA's Board of National Estimates. It also changed the scope and the depth of the strategic estimates, Huffstutler argued. The unconstrained "military requirements approach" to estimating prevalent during the 1950s, Huffstutler also claimed, gave way to a multidimensional and more integrated approach during the 1960s. Thus, CIA analysts were drawn into an area that was once the province of the military services. Huffstutler further argued that just as the Board of National Estimates was drawn into the strategic assessments process by McNamara's directive, it also emerged gradually as the arbiter of assessments on Soviet political behavior in the fields of foreign policy, domestic Soviet policy, and arms control.
Finally, Huffstutler briefly commented on the analytical skills of the CIA. Although in the 1950s, CIA had a diverse set of analytical experts in a wide range of fields, they were relatively small in number. Some 200 analysts worked Soviet issues in the 1960s according to Huffstutler. By the early 1980s, there were over 1,600 analysts in CIA's Directorate of Intelligence alone; half of them working on Soviet issues. In addition, the Agency sought expertise from industry and from American universities. Moreover, the turnover of CIA analysts was also extremely low, running less than 3 percent during the 1980s.
Looking at the overall performance of the CIA during the early Cold War, Melvyn Leffler argued convincingly that, despite shortcomings, the Agency built a comprehensive picture of the Soviet state and of the communist system and of the threats they posed to American society. Although the Agency did not predict the timing of the Soviet atomic bomb, the North Korean attack, or China's intervention in Korea, Leffler said that the Agency's analysis was far more nuanced, far shrewder than popular or even scholarly accounts currently suggest. According to Leffler, CIA defined security in terms of correlations of power, and power was defined in terms of control of energy resources, industrial capacity, and raw materials production. Thus, the rivalry between the United States and the Soviet Union was--for the CIA--not primarily military, but was a competition over aggregations of power potential.
According to Leffler, CIA saw three basic problems for American security: (1) to keep the still widely dispersed power resources of Europe and Asia from being drawn together into a single Soviet power structure with a uniformly communist social organization; (2) to persuade the people and political authorities of states in the intermediate regions that their political aspirations and security interests could be satisfactorily identified with the United States; and (3) to maintain the social fabric and structure of the United States. Leffler asserted that CIA's analysis sought to establish a larger strategic framework for assessing Soviet intentions, Soviet threats, and more importantly, American security interests.
Agreeing with Steury's analysis, Leffler said that the intelligence assessments produced during the 1948 Berlin crisis were basically sound. The Soviet Union wanted to avoid war, because it was weaker than the United States and knew it could not win a protracted war. CIA analysts concluded that the Soviets would, nevertheless, try to exploit economic hardship and revolutionary nationalism. Hence, revolutionary nationalism in the Third World had to be dealt with effectively. According to Leffler, CIA grasped the virulence of revolutionary materialism, grasped its indigenous roots, and worried about the Kremlin's ability to harness it for its own ends. The CIA also perceived the roots of the Sino-Soviet split before most other observers, and it understood that the key to Soviet power was a strong economic base.
All of this, Leffler claimed, did not translate into clear policy choices. CIA's nuanced assessments created problems of their own for US policymakers. Leffler argued that since "reality is always messier, grayer, and blurrier, than we would like," we need to focus attention not only on intelligence assessments, but, on how policymakers have used intelligence.
The final commentator, Jack Matlock focused his remarks on the question of how policymakers have used intelligence. Speaking from his own experiences, he stated that usually one could discount the assessments from the military intelligence services because "they never make an assessment that would, in any way, undercut their service's requests for funds." Matlock went on to say that in making political decisions you wanted an unbiased analysis, so you looked to CIA or to the State Department, depending on the subject. The biggest advantage Ambassador Matlock saw in creating CIA was in moving the thrust of intelligence analysis out of the Defense Department into a more neutral body.
Referring to former Senator Patrick Moynihan's criticisms of the Agency's analysis, Matlock argued that the purpose of intelligence is not necessarily prediction. For Matlock, the only solid basis for prediction is what happened in the past, and this can lead to shaky assumptions because inevitably there comes a time when people react differently than in the past. Matlock, in general, agreed with Leffler's assessment that CIA's analysis of the Soviet Union, particularly from the late 1960s through the 1980s, was much more accurate than most critics are willing to admit.
Donald P. Steury, visiting professor at the University of Southern California in 2001, is a senior historian on the CIA History Staff at the Center for the Study of Intelligence.
1. Sherman Kent, "Prospects for the National Intelligence Service," Yale Review (1946): p. 123.
2. William M. Leary (ed.), The Central Intelligence Agency: History and Documents (Tuscaloosa, AL: University of Alabama Press, 1984), p. 25.
3. For example: The intelligence organizations of the Department of State, the War Department, and the Navy Department. Kent, pp. 126-27.
4. Kent, "Prospects," p. 127.
5. Leary, CIA, p. 24.
6. Ibid., p. 25.
7. Ibid., p. 25.
8. Ibid., pp. 25-26.
9. Ibid., p. 26.
10. Ibid., pp. 26-27.
11. Ibid., p. 24.
12. Center for the Study of Intelligence, Declassified National Intelligence Estimates on the Soviet Union and International Communism (Washington, DC: CIA, 1997).
13. 20 ORE 3/1: Soviet Capabilities for the Development and Production of Certain Types of Weapons and Equipment (31 October 1946), p. 1.
14. Leary, CIA, p. 23.
15. Center for the Study of Intelligence, Declassified National Intelligence Estimates.
16. Leary, CIA, p. 28.
17. Ludwell Lee Montague, General Walter Bedell Smith as Director of Central Intelligence: October 1950-February 1953 (University Park, PA: Pennsylvania State University Press, 1992), p. 43.
18. Kent, "Prospects," p. 123.
19. Sherman Kent, Strategic Intelligence for American World Policy (Princeton, NJ: Princeton University Press, 1965), pp. 40-41.
20. The three components were non-military instrumentalities (the ability to influence events with political and economic means); force in being (the capabilities of its standing military); and war potential (the fully mobilized potential of its society, economy, and military to wage war). Kent, Strategic Intelligence, pp. 40-65, passim.
21. George S. Pettee, The Future of American Secret Intelligence (Washington, DC: The Infantry Journal Press, 1946), see esp. Chapters 2 and 3, pp. 25-46.
22. Larry A. Valero, "The American Joint Intelligence Committee and Estimates of the Soviet Union, 1945-1947," Studies in Intelligence (Summer 2000), pp. 5-9.
23. Foreign Relations of the United States (hereinafter, FRUS), 1945-1950: Emergence of the Intelligence Establishment (Washington, DC: US Government Printing Office, 1996), p. 345.
24. FRUS, 1945-1950, p. 346.
25. Leary, CIA, p. 25.
26. Leary, CIA, p. 25.
27. Valero, "Joint Intelligence Committee Estimates," p. 11.
28. Melvyn P. Leffler, A Preponderance of Power: National Security, the Truman Administration, and the Cold War (Stanford: Stanford University Press, 1992), 130-38; see also Ibid., p. 11.
29. Valero, "Joint Intelligence Committee Estimates," p. 11.
30. William R. Harris, "March Crisis 1948 Act I," Studies in Intelligence (Fall 1966), pp. 3-4.
31. Ibid., p. 7.
34. Lucius D. Clay, Decision in Germany (Garden City, NY: Doubleday and Co., 1950), p. 354.
35. Harris, "March Crisis," p. 8.
36. Ibid., pp. 9, 11.
37. Ibid., p. 10.
38. Ibid., p. 13.
39. Ibid., p. 8.
40. Murphy, et al., p. 55.
41. Harris, "March Crisis," p. 15.
42. Ibid., p. 16.
43. Ibid., pp. 16-17.
44. Ibid., pp. 16-17.
45. Vladislav Zubok and Constantine Pleshakov, Inside the Kremlin's Cold War: From Stalin to Khrushchev (Cambridge: Harvard University Press, 1996), pp. 48-49.
46. Kenneth W. Condit, The History of the Joint Chiefs of Staff: The Joint Chiefs and National Policy, Vol. 2, 1947-1949 (Wilmington, Michael Glazier, Inc., 1979), pp. 191-212, passim.
47. Harris, "March Crisis," p. 19.
48. Ibid., p. 20.
49. Ibid., p. 21.
50. Ibid., pp. 16, 21.
51. Ibid., p. 21.
52. ORE 22-48, Addendum: Possibility of Direct Soviet Military Action During 1948-49 (16 September 1948); ORE 46-49 The Possibility of Direct Soviet Military Action During 1949 (3 May 1949). For a full text of all these documents, see Donald P. Steury (ed.), On the Front Lines of the Cold War: Documents on the Intelligence War in Berlin, 1946-1961 (Washington, DC: CIA, 1999), pp. 136-57.
53. ORE 58-48, The Strategic Value to the USSR of the Conquest of Western Europe and the Near East (to Cairo) Prior to 1950 (30 July 1948), p. 3.
54. Harris, "March Crisis," p. 13.
55. Arthur B. Darling, The Central Intelligence Agency: An Instrument of Government, to 1950 (University Park, PA: Pennsylvania State University Press, 1990), p. 338.
56. DeForrest Van Slyck saw things rather differently. As he later reported to DCI Hillenkoetter, "It was 'virtually impossible'...to get a `completely objective' estimate from the 'Service departments;' they were `unable to free themselves from the influences of departmental policy and budget interests.' It was their tendency `too readily to translate capabilities into intentions' without giving due weight to political, economic, and psychological considerations of wide range. These points [he wrote] were equally `applicable to the State Department.'" D. Van Slyck to R.H. Hillenkoetter, 23 December 1948; quoted in Darling, Instrument of Government, pp. 339-340.
57. Leary, CIA, p. 31.
58. Max Millikan, "The Nature and Methods of Economic Intelligence," Studies in Intelligence (Spring 1956), p. 4. Although published in Studies in 1956, this article actually reprints a memorandum written by Millikan during his tenure as D/ORR.
59. Ibid., p. 2.