Estimating Soviet Military Intentions and Capabilities
Raymond L. Garthoff
From the outset, the highest priority question for CIA analysis was the role of military power in Soviet ideological doctrine, state policy, and organizational plans. By mid-1946, there was full consensus in the American policy and intelligence communities that Stalin and other Soviet leaders operated on the basis of belief in historically destined conflict between the Soviet Union and the United States and other Western countries. This was seen as an inescapable challenge requiring a firm response to contain Soviet expansionist tendencies and to deter a Soviet military threat. In short, there was a Cold War.
The question remained as to whether the Cold War would eventually lead to a "hot" war; some believed war was inevitable, while others did not. American political leaders wisely were prepared not to prejudge the answer by taking the initiative. But there remained the estimative questions for intelligence: Did the Soviet leaders intend to use their military power to launch an attack on the United States and/or Western Europe? If so, under what situations, when, and how? If contingently, under what conditions? If not, what were the roles of military power in Soviet foreign policy?
The question of Soviet intentions and attendant objectives was the fundamental element of threat assessment. Soviet military forces and capabilities to carry out Soviet leaders' intentions necessarily constituted the second, but crucial element of that assessment. Soviet military strategy and doctrine for employing its forces and capabilities in support of their objectives and intentions was the third essential element of threat assessment. Finally, threat assessment needed to be made for both the near-term and the long-term. Judgments on Soviet near-term and ultimate intentions were constantly made, sometimes explicitly but more often assumed. Estimates of Soviet capabilities were the predominant focus of attention and received virtually all of the intelligence collection, analysis, and estimative effort. Discussion and conclusions on Soviet military strategy were sometimes addressed but most often implicitly and by assumption.
In all three respects, CIA and US national intelligence estimates and judgments were made not only in terms of the best, meaning "most likely," estimates or validated findings, but also (or instead) on the basis of "worst case" judgments of Soviet leadership intentions and state policy, military capabilities, and military strategy. The question of whether US policy should seek to deal with the most likely threat or the most dangerous and challenging threat was never posed directly. Basically, that was a judgment for US policymakers, not intelligence officers, to make. But intelligence professionals at times took on this responsibility by hedging, with a bias toward the most plausible worst-case threat. This was considered the most prudent course: an underestimate of the most threatening case could bring disaster and defeat, while an overestimate would (it was believed) only bring over insurance.
Implicit in this approach, though never explicit, was an assumption that we were dealing with an established objective, "given" Soviet intentions and capabilities, even if (as some would concede) we could only make our own subjective evaluation of intentions and hedged our estimates of capabilities. There was essentially and with rare exception no recognition that the underlying reality might be contingent and dynamic and at least in part reactive; that Soviet intentions, military programs to build forces and provide capabilities, and strategies for employment of these forces and capabilities could be significantly affected by US policies and actions (including our response to what we perceived to be the threat); and that, in fact, US intelligence assessments and their consequences in US policy and in military capabilities and strategy would influence the nature and extent of the very threat we were assessing. One reason for this blind spot was the tendency to see Soviet objectives, intentions, and capabilities as principally, if not exclusively, offensive.
In the late 1940s and early 1950s, especially after the outbreak of the Korean War in 1950, perceived Western military weakness led to intelligence evaluations of threatening Soviet military intentions. An internal CIA Intelligence Memorandum in August 1950, not coordinated with other intelligence agencies but circulated to them, stated that Soviet war readiness was such as to "suggest strongly that the Soviet leaders [believe that they] would be justified in assuming a substantial risk of war during the remainder of 1950." Moreover, there was no hesitation in concluding that "the USSR is vigorously and intensively preparing for the possibility of direct hostilities with the U.S." The first CIA-drafted National Intelligence Estimate (NIE) on Soviet Capabilities and Intentions (NIE-3, 15 November 1950) reported that "There is, and will continue to be, grave danger of war between the USSR and its satellites on the one hand and the US and its allies on the other." Further, the NIE warned that "the Soviet rulers may deliberately provoke such a war [‘general war with the Western Powers'] at the time when, in their opinion, the relative strength of the USSR is at its maximum. It is estimated that such a period will exist from now through 1954, with the peak of Soviet strength relative to the Western Powers being reached about 1952." 
From 1946 until 1952, a series of estimates dealt primarily with the question of a possible Soviet intention to initiate war. A parallel series of estimates was made of Soviet military capabilities for attack on the United States and the West based on "worst cases," usually postulated as one-and-a-half to two years in the future (a future that fortunately constantly receded, maintaining that same span), without addressing the likelihood of war. The separate estimates of intentions and capabilities raised dire possibilities of when the Soviet leaders could decide on war, but did not go so far as to estimate any time that such an attack would, in fact, occur.
NIE-3 had said that "at what point they [the Soviet leaders] will make a decision to launch a general war is not now determinable by Intelligence." But it promised future reports. And NIE-15 on 11 December 1950 even posited that "The overall situation is such that the possibility cannot be disregarded that the USSR has already made a decision for general war and is in the process of taking steps preliminary to its inception."  Other NIEs from 1946 through 1951 stressed the threat of Soviet initiation of war, and it was not until NIE-48, on 8 January 1952 that CIA and the other intelligence agencies reached a cautious and qualified judgment in a near-term estimate that "on balance we believe it unlikely that the Kremlin will deliberately initiate global war during 1952." This and several subsequent estimates introduced the somewhat reassuring formulation that the estimators believed "the Kremlin prefers to pursue its objectives through methods short of deliberate resort to war with the US and its allies." 
Only by the late 1950s and early 1960s would it become possible to acknowledge straightforwardly that the Soviet leaders had security concerns and defensive and deterrent objectives in pursuing their military programs, including bolstering their offensive forces for deterrence. (Again later, as we shall note, as Soviet strategic offensive forces grew in the late 1970s and early 1980s, there was a renewed tendency to impute offensive intentions to increasing Soviet strategic offensive capabilities.) What about the role of CIA's analysis in the late 1940s and the 1950s? From 1947 until late 1950 the Office of Reports and Estimates (ORE) and, from November 1950 until 1973, the Office of National Estimates (ONE) drafted the estimates on intentions and (with inputs from the military intelligence organizations) the estimates on military capabilities, with few significant dissents from what were generally consensus appraisals. 
CIA analysis in the formative years of the Cold War was weak and not very influential. This also was true of the other US intelligence agencies and their influence. Although CIA--and its predecessor, the Central Intelligence Group (CIG) --had a voice in this formative period, the most important judgments on Soviet intentions were based on evaluations by policymakers, not their intelligence advisers. ORE-1, the first CIG estimate, was prepared in July 1946 as a contribution to the internal White House Clifford-Elsey Report to President Harry S. Truman on Soviet objectives (as was a Joint Intelligence Committee paper submitted by the Joint Chiefs of Staff as JCS 1696). But both had far less influence than George Kennan's "long telegram."
The intelligence submissions helped to consolidate a consensus that largely had already formed among policymakers by mid-1946. Later, ORE 60-48 was used, but probably was not influential, in preparing NSC 20/4 in November 1948, launching a number of covert operational programs directed against the Soviet Union. The key Cold War charter, NSC 68 in April 1950, referred only once to a CIA evaluation, and that was on a question concerning military capabilities (an agreed-upon CIA military estimate, somewhat inflated as it turned out, of the number of atomic weapons the Soviet Union was expected to have, including an ominous 200 by mid-1954, setting that up as a new year of anticipated greatest danger). On Soviet intentions, NSC 68 ignored the moderate CIA assessment that even acquisition of 200 atomic bombs would not necessarily lead the Soviet leaders to attack. (This sensible judgment had been dissented to by all the military intelligence offices and the intelligence bureau of the Department of State and was abandoned by CIA under pressure only two months later in a revised consensus estimate). From 1954 through 1960 there were comprehensive annual estimates concerning Soviet internal and foreign political, economic, and scientific and industrial policies as well as military affairs, each dealing with the forthcoming five years. NIE 11-4-54, Main Trends in Soviet Capabilities and Policies, 1954-1959 (September 1954) was the first in this series. By this time, CIA had gradually developed political, economic, and scientific-technical research and expertise and was developing capacity in the military field. Nonetheless, the national estimates of Soviet military programs and capabilities continued through most of the 1950s to be based on contributions from the US military intelligence services, although modified by CIA drafters and the interagency process.
When CIA was established, there had been a general understanding that the Army, Navy, and newly established Air Force would exercise primary responsibility for military intelligence in their respective fields. Early Directors of Central Intelligence (DCIs), including Allen Dulles, and on into the 1950s, had accepted that division of responsibilities, although the DCI and CIA retained final authority for "national intelligence," and the boundary between that sphere and the narrower areas of responsibility of the military intelligence services was not clearly defined.
NIEs represent an authoritative distillation of intelligence information and analysis, and provide estimates that forecast behavior and future forces and capabilities. This is true for all areas of CIA intelligence analysis but especially for military analysis. Much political, economic, and scientific-technical intelligence is conveyed through other intelligence reports, especially political intelligence through daily intelligence publications. CIA military analysis is also made available in other publications but especially in NIEs.
There were a number of errors in estimated Soviet military force projections in the 1950s, but these did not involve disputes within the US Intelligence Community, nor did they become public or policy issues. There was an underestimation of the growth of military expenditures in the mid-1950s; a substantial underestimation of Soviet military manpower--mainly ground forces--in the early 1950s; an underestimation of the Soviet medium bomber force from 1954 to 1957; and an underestimation of the availability of uranium and the extent of U-235 production in the mid-1950s (although balanced by an overestimate of the nuclear weapons stockpile in the critical very early 1950s). There were substantial overestimates of the growth of the Soviet submarine force in 1955-57; a Soviet aircraft carrier force in 1954-56; all-weather Soviet fighter interceptors in 1955-56; and Soviet military manpower in the late 1950s. Although most of these errors were introduced by the military intelligence services (except for the CIA estimate on military expenditures and shared responsibility for the erroneous early nuclear weapons stockpile estimates), they were accepted by all the intelligence agencies, including CIA, and the errors therefore became errors of national intelligence made by the Intelligence Community. None of them, however, had important influence on national policy.
Two developments in the last half of the 1950s were more consequential. A serious overestimation of Soviet heavy bomber production and a corollary overestimation of refueling tankers from 1955 through 1957 led to a public outcry over an alleged "bomber gap" to Soviet advantage, imperiling US superiority and security. The error was a compound of insufficient information on the current situation in Soviet aviation development and mistaken assumptions as to Soviet military "requirements" for the future. CIA originally joined the other intelligence agencies in accepting Air Force projections, but later, detailed CIA analysis of Soviet aircraft production facilities--information obtained from U-2 reconnaissance flights--and other technical intelligence led to deflating the feared bomber gap by 1958. Instead of the 700-800 heavy bombers projected from 1955 to 1957, the Soviet Union never fielded more than 150 (plus 50 others configured as tankers). In parallel, the large Soviet medium bomber force (much larger than initially expected) was, in US intelligence estimates, expected to enhance its possible intercontinental capability by aerial refueling from a tanker force estimated in 1954 to reach 850 aircraft by 1959 (even though the Soviet Air Force at that point had no experience in aerial refueling); as late as 1956 the estimate was 350-400 tankers by 1960-61.  In fact, the Soviets did not acquire more than about 70.
The ‘Missile Gap'
By the time the "bomber gap" expired in 1958, concern over an even more ominous "missile gap" was brewing. The Intelligence Community had begun to get its first substantial intelligence on the Soviet missile development program in 1954 and 1955 from radar and other technical intelligence collection stations in Turkey, near the Soviet missile test range at Kapustin Yar. This supplemented information gleaned from returning German scientists and a few knowledgeable defectors and other sources. The first NIEs to estimate an initial operational capability (IOC) for a Soviet intercontinental ballistic missile (ICBM), issued from October 1954 to March 1957, were fairly close to the mark, all estimating an IOC in 1960-61.  The errors in the "missile gap" period began in November-December 1957 in the somewhat panicked reaction to Sputnik (the first Soviet earth satellite, orbited in October 1957) and the first Soviet ICBM test of August 1957. The estimates of that period included very early IOCs of 1958 or 1959,  although the previously estimated IOC was closer to the actual deployment program's nominal IOC in 1961 (actually with less than the arbitrary IOC definition of 10 operational launchers).
The NIEs of late 1957 through mid-1961 presented overestimates of a Soviet ICBM buildup much greater and more rapid than the US program, portending a serious threat to the United States and its strategic retaliatory forces and bases, with another time of greatest danger seen in the early 1960s. Again, Air Force Intelligence led the way and most egregiously overestimated current and future Soviet ICBM capabilities, but State Department intelligence and Joint Staff intelligence went nearly as far. CIA took a high middle position, with Army and Navy intelligence estimating far lower levels. The problem was not in analysis of available information (except for the Air Force position), but uncertainties as to Soviet missile test results, missile production capabilities, and intentions to press maximum effort. The feared "missile gap" was dispelled when various intelligence sources, above all, satellite photography during 1961, clearly showed that the Soviet leaders had decided to wait for a more suitable second generation ICBM before deploying more than literally a handful of the bulky SS-6s. Some evidence earlier available was now given more confident interpretation, in particular a long hiatus in Soviet ICBM testing in 1958-59, which by 1961 was recognized as evidence of unresolved problems requiring further attention (as CIA, Army, and Navy had believed), rather than as a sign the Soviets had no problems and were already rapidly deploying (as Air Force had contended).
The main causes of the vast exaggeration of current and projected Soviet ICBM deployment from 1957 to fall 1961 were the absence of adequate intelligence information and assumptions that tended toward worst-case estimates--considered more prudent than optimistic ones. The worst for current and near-term estimates were in the early years; NIE 11-10-57 in December 1957 estimated 100 ICBMs deployed by mid-1959 and up to 500 by mid-1961--when the actual figure in 1961 was four! The greatest overestimate for the mid-1960s was in NIE 11-8-60 in August 1960, estimating up to 700 by mid-1963, but even NIE 11-8-61 in June 1961 still estimated up to 400 by mid-1964. There was a spread, with Army and Navy intelligence consistently estimating lower (and more nearly correct) levels than the CIA estimates cited above, and the Air Force consistently estimating much higher figures. The Air Force in December 1960 estimated up to 950 ICBMs deployed by mid-1964 and 1,200 by mid-1965, and in June 1961 still estimated up to 1,150 by mid-1965 and 1,450 by mid-1966. In fact, Soviet ICBM deployment leveled off in 1963-65 at 209 operational launchers before beginning a buildup with third-generation ICBMs later in the 1960s. They did not reach the 1,450 estimated by the Air Force for 1966 until 1971. During the period of controversy over a "missile gap," an unrelated issue arose over a different segment of the estimates of Soviet military power. In January 1960, Soviet leader Nikita Khrushchev announced a planned reduction in Soviet military manpower by one-third, from what he declared to be an existing level of 3,624,000. The current NIE (11-4-58; 11-4-59 had been delayed and was still in preparation) had estimated the level to be 4,325,000 men, not including an estimated 400,000 militarized internal security and border guards that Khrushchev probably was not counting. Nonetheless, some in CIA believed Khrushchev's figure might be correct, and DCI Allen Dulles called for a Special National Intelligence Estimate (SNIE) to take a new and harder look at the evidence.
Although the subject was not intrinsically of high importance, Dulles's action marked a new departure by taking a more active CIA role in estimating areas of Soviet military power heretofore left to the military intelligence services. Improving estimates of manpower involved revising estimates of many secondary elements of the Soviet military establishment (for example, drastically reducing estimates of Soviet coastal artillery). Of course, each of the services provided a contribution focusing on the manpower of its corresponding Soviet counterpart, and when they did, the combined figures were higher than the previous NIE submissions, enlarging the discrepancy between the US military services' estimate and Khrushchev's statement. CIA took a very active part in pressing the military intelligence services to justify their claims in detail, and in many cases they could not. To be sure, it was natural that the Air Force had focused on numbers of aircraft and the Navy on ships, rather than on numbers of personnel. The military manpower estimates of both were found to be greatly overstated. The Army had naturally paid more attention to manning, but while they had a good handle on order-of-battle by units, estimating actual manning levels beyond units in East Germany was usually based on weak evidence. In all services, the basis for estimating actual manning of most support units was thin.
In retrospect (based on CIA analysis in 1956 and 1957) it had become clear that, while an estimate of about 175 Soviet Army line ground divisions was probably sound, the actual strength of the Soviet armed forces (mainly in the army) had risen from less than 4 million in 1948 to nearly 6 million in 1952-53, and had then gone down again to around 4 million by the late 1950s. US intelligence estimates during the early and middle 1950s had varied only between 4 and 4.5 million, and estimates of Soviet Army ground force combat strength had been steady at 2.5 million throughout the 1950s. The upshot was a new estimate, SNIE 11-6-60, prepared by CIA and concurred with by all the services, concluding that Khrushchev's 3.6 million men was "substantially correct." Moreover, the groundwork had been laid for more detailed intelligence collection and analysis by the Army and CIA leading to more refined and accurate estimates of the strength of Soviet divisions. Most of them were found to be not at the previously estimated average of 70 percent of strength, but overall averaging about half that level (with the closely observed divisions in East Germany being not typical, but among those maintained at a relatively high level of manning).
From the early 1960s on, CIA had both trained analysts and much more extensive intelligence information to permit sharing fully in national intelligence estimates of Soviet military forces and capabilities. NIEs continued, for sound reasons of not getting too far ahead of the data, to project estimated future forces only five years ahead. CIA did, however--with Secretary of Defense Robert McNamara's approval--also participate with the military services (and, after 1961, with the newly created Defense Intelligence Agency [DIA]) in preparing intelligence projections of Soviet military forces for long-term Department of Defense planning purposes in an internal Pentagon series called the NIPP (National Intelligence for Planning Purposes), building on NIEs but looking some ten years ahead, although of course revised and updated annually. CIA also took part with the military intelligence services in preparing the annual NATO military intelligence estimates (SG-152 in the 1950s, later MC-161 series).
Throughout the 1950s, the NIEs drafted in ONE and CIA's internal analyses of Soviet military developments thus refined and improved estimates of Soviet military capabilities. As earlier noted, in the late 1940s and early 1950s national intelligence judgments on Soviet intentions with respect to the use of military power tended at best to make guarded, short-term estimates that deliberate initiation of war was unlikely and at worst to repeat solemnly that the Soviet leaders (usually the impersonal "the Kremlin," and never the very real Stalin) might decide to go to war at any time. The frequent short estimates on Soviet capabilities to attack usually sidestepped the issue by saying what Soviet military forces were believed to be capable of doing if they were launched in an attack. It was, of course, appropriate and necessary to estimate what the Soviet Union could do in case of general war. War was possible, and the estimates did not say it was certain. Nonetheless, in historical retrospect it is clear that too much attention was devoted to general war as a near-term possibility. By the mid-1950s that danger had receded from the forefront of concern, and most estimates focused more usefully on Soviet courses of action short of initiating war. By the late 1950s, CIA carried further the proposition that although the Soviet leaders were adversaries with a hostile world view, they also were concerned with security and deterrence of war and not inclined to launch a war. NIE 11-4-58 in December 1958 stated that "the Soviet leaders will view large deterrent and other military capabilities as an essential support to their foreign policy and to the USSR's status as a leading world power." It further said it was probable that "in the Soviet view both sides are now militarily deterred from deliberately initiating an all-out nuclear war or from reacting to any crisis in a manner which would gravely risk such a war, unless vital national interests at home or abroad were considered to be in jeopardy." Nonetheless, it also still stated that the USSR was seeking in the long run "to achieve, if possible, a clear military superiority."  There was no dissent in the Intelligence Community to that estimate.
By February 1960, in NIE 11-4-59 and NIE 11-8-59 (both issued after delays in coordination), reaffirmation of Soviet pursuit of a deterrent capability and belief in mutual deterrence was no longer balanced by an equivocal reference to seeking superiority. To the contrary, NIE11-4-59 declared that "military power in the Soviet view should not be used recklessly to the hazard of the main power center of Communism. . . . The Soviet armed forces are intended in the first instance to deter attacks on the USSR and other communist states, and to insure survival of communist power should such an attack occur. . . . They are probably not intended for any consistent and far-reaching policy of outright military conquest." That broke the consensus. The chief of Air Force Intelligence registered a vigorous dissent. He did not accept the existence of "mutual deterrence" (nor Soviet acceptance of it). He also contended that "Soviet forces are as likely to be intended for a "consistent and far-reaching policy of outright military conquest as for any other purpose." 
NIE 11-4-59 deleted the previous judgment that the Soviets were seeking, "if possible," a clear military superiority; it also declared that "they probably do not count upon acquiring, by any particular date, an advantage so decisive as to permit them to launch general war with assurance of success and under conditions which would not gravely menace their regime and society." It went even further: "Moreover, even if the Soviets came to believe that they could win a general war, perhaps at a high but acceptable cost, it probably would not be their preferred course to initiate such a war. Instead, in view of their philosophy about war and politics, their preference would probably be to press home their advantage by political and psychological means. . . ." The NIE did flag the danger that the Soviets might in such circumstances "overplay their hand and take actions that would in fact precipitate general war," but it argued that "just as we exclude the deliberate initiation of general war from the policy alternatives the Soviets would adopt, we also exclude the witting assumption of serious risks of general war."  The Air Force again dissented, arguing that the Soviets were "vigorously attempting to acquire an advantage so decisive as to permit them to launch general war . . . ."  This image of Soviet intentions with respect to their growing military power in the NIEs of the late 1950s, highlighted by the lone Air Force dissent, was a marked change from the jittery concerns expressed by CIA and the other agencies a decade earlier.
NIE 11-8-59 on strategic capabilities posited the goal of the estimated Soviet ICBM buildup to be "as large as Soviet planners deem necessary to provide a substantial deterrent and preemptive attack capability," not decisive superiority. It did argue that the Soviets would "attempt to achieve a military advantage over the US" and would even consider it "desirable" if they could attain a decisive superiority, but also that they considered it necessary for such a buildup to avoid provoking a US preventive attack.
The Air Force dissented vigorously, arguing that the Soviet rulers would not be satisfied by a deterrent and contingent preemptive capability but were "endeavoring to attain at the earliest practicable date a military superiority over the United States which they would consider so decisive as to enable them either to force their will on the United States through the threat of destruction, or to launch such devastating attacks against the United States that, at the cost of acceptable damage to themselves, the United States as a world power would cease to exist."  This Air Force dissent reappeared in most NIEs on the subject for the next thirty years.
NIE 11-4-60 carried forward the new evaluation of Soviet intentions introduced in NIEs 11-4-59 and 11-8-59. Soviet military power overall, and long-range striking capability in particular, were said to be thought of and designed primarily for deterrence or response if deterrence had failed and not for "deliberate initiation of general war." Moreover, "while communist doctrine injects hostility and conflict into Soviet policy, it does not propel the USSR toward war as the primary instrument of policy. Indeed, the Soviets' fundamental belief in the inevitable movement of historical forces to their advantage leads them to prefer to avoid the risks of war." Soviet recognition of mutual deterrence was reaffirmed, as well as continuing efforts to gain military advantage, but even if they attained a "clear military advantage," their "preference would be to press it by non-military means, and to attempt to achieve their objectives without actually resorting to general war." Moreover, the judgment expressed in NIE 11-4-59 was reiterated, that the Soviets "probably do not count upon acquiring, at any foreseeable point in time, an advantage so decisive as to permit them to launch general war with assurance of success…" 
Air Force Intelligence dissented again, five times, to the evaluations cited above, asserting not only that Soviet strategic programs showed "a definite intent by the Soviet rulers to achieve a clear military superiority at the earliest practicable date," but also that "we are entering a very critical twenty-four month period in which the USSR may well sense it has the advantage. The Soviet leaders may press that advantage and offer the US the choice of war or of backing down on an issue heretofore considered vital to our national interests." 
By the end of a 24-month period there had indeed been a crisis, but one stemming from Soviet strategic military weakness, not strength, and ending with the Soviet Union, not the United States, backing down: the Cuban missile crisis. The actions of both sides in the crisis, however, reflected mutual deterrence. The long-term issue, whether the Soviet leaders were striving for a decisive military superiority and war-winning capacity, continued to simmer in Air Force footnotes of dissent but did not seriously arise again for a decade and a half, in the challenge of "Team B," to be discussed later. For all practical purposes, the issue of basic Soviet intentions, while still relevant to current events including evaluation of risks, was no longer featured in NIEs in the 1960s and the first half of the 1970s.
NIEs had reflected and followed alarmed political evaluations in the late 1940s and early 1950s, spurred by the coup in Czechoslovakia, the Berlin blockade, and the Korean war. They had not been based on any concrete intelligence information of Soviet intentions, and judgments based on inference from communist ideology cut both ways. Later changes in estimates in the 1950s were also based more on changes in the international situation, the post-Stalin changes within the Soviet Union, and observed Soviet behavior rather than on specific intelligence about Soviet intentions with respect to use of military power.
Meanwhile, there was a definite, if gradual, improvement in the intelligence foundation for estimates of Soviet military forces and capabilities throughout the 1950s. CIA's analysis advanced from building a database to evaluations based on more concrete information about weapon programs acquired from a wider array of sources, including intercepted communications (e.g., from tunnels into Soviet military communication lines in the Soviet occupation sectors of Vienna and Berlin), radar intelligence on Soviet missile testing (e.g., from stations in Iran, Turkey, and Alaska), and U-2 photography of the USSR and other communist states--succeeded by satellite photography from the beginning of the 1960s on. Espionage provided some useful information but very much less than technical means of collection. CIA moved into a lead in most of these fields, but information from all agencies was generally circulated among the members of the US Intelligence Community (and to some extent with close allies, notably the United Kingdom).
Doctrine and Strategy
With respect to the third element of threat assessment--doctrine and strategy for the employment of military power--CIA and the NIEs lagged behind the assessment of intentions and capabilities. Recognition of deterrence and preparedness for defense of the Soviet Union, and of the "counterdeterrence" of Western military power and other Soviet political uses of military power to bolster an active foreign policy, were reaffirmed in estimates. And it was assumed from the outset and seemingly confirmed by the steady enhancement of Soviet and Soviet Bloc military forces in Europe that Soviet strategy from 1946 on had, should general war occur, been postulated as an all-out conventional offensive against the West in Europe (and possibly also in the Mediterranean and Middle East). There is now some evidence that Soviet war plans in the late 1940s called for a defensive posture in Europe, but even if valid, that may simply reflect a decision that war in Europe would occur only if initiated by a Western attack. In any case, by the 1950s Soviet and later Warsaw Pact military strategy called for an all-out offensive to the West in case of war, either by preempting a budding Western attack or by a prompt counteroffensive as soon as possible after rebuffing such an attack. We still lack today adequate information on Soviet intelligence evaluations of Western military intentions, capabilities, and strategy during the Cold War, but such evidence as is available suggests considerable exaggerations of Western bellicosity and capabilities, including planning for initiation of war. Soviet intelligence estimates, like those of the United States and NATO, were always predicated on initiation of war by the other side.
In the 1950s, beyond correctly seeing a Soviet strategic concept calling for an offensive military campaign in Western Europe, and probably incorrectly believing there was also a Soviet proclivity to prepare to launch a war if conditions seemed propitious, the estimates reflected little appreciation of Soviet doctrine and strategy. As noted earlier, when it was necessary to go beyond the evidence to estimate future "requirements," there was a consistent tendency in the 1950s to underestimate numbers of Soviet medium range bombers and later missiles and to overestimate intercontinental bombers and missiles. From 1954, as Badger jet medium bombers began to enter the Soviet force, through early 1957, the NIEs estimated that the number of Soviet jet medium bombers would level off at about 700 aircraft by 1958; yet by 1960 the actual number was more than twice that. (The NIEs also continued long thereafter to underestimate the number that would be retained as the force gradually declined in size.) Similarly, the NIEs in 1959-61 greatly underestimated the number of Soviet medium and intermediate range ballistic missiles (MR/IRBMs) then being deployed. NIE 11-4-59 and 11-8-59 (both on 9 February 1960) and NIE 11-4-60 (1 December 1960) estimated the Soviet MR/IRBM "requirement" at 250 launchers, expected to be deployed by mid-1961, with that deployment then maintained. NIE 11-8‑61 (7 June 1961) estimated the force then at 225 and believed it would level off at between 275 and 300 in 1965-66. NIE 11-8/1-61 (21 September 1961) raised the expected level to 350-450, to be reached in 1962-63. By the time of NIE 11-8-62 (6 July 1962), it was apparent from continuing construction of new sites that the total would be much higher, and it was expected to level off at 550-650 by 1965. Finally, in NIE 11-4-63 (22 March 1963), with about 600 launchers completed and 75 others under construction, it was correctly estimated that the total would top off in mid-1964 at some 700-750 launchers (in fact, it reached 709). 
The problem with these estimates of Soviet medium bombers and medium range missiles was not primarily shortcomings in information, but error in estimating the Soviet view of their requirement for forces deployed to strike targets in Eurasia, especially forward-based US strike forces. Similarly, the erroneous assumption that the Soviet Union must be intending to build up a large aerial tanker force to enhance the weak intercontinental capabilities of the Bison and Badger aircraft was a case of projecting onto Soviet military planners the US view of priorities.
In fact, the whole series of failures to anticipate, and later to understand, the high Soviet priority given to what US estimators regarded as "peripheral" theater strike forces from the 1950s on was--in important measure--an example of "mirror-imaging," albeit one that critics of that process have not recognized. This mirror-imaging underlay such statements as the admission by the authors of NIE 11-8-63 of "difficulty understanding the Soviet rationale" for the priority placed on building up theater forces for peripheral areas. The NIE lamely suggested this represented "an apparent lag" in strategic thinking. Of course, to the Soviet military and political leaders those potential theaters of war, with NATO and US forward-based nuclear strike forces deployed on their doorstep, were not "peripheral."  There were later instances of failures, or divided views in the US Intelligence Community over whether certain Soviet military developments concerned peripheral theater missions or additions to intercontinental capabilities. The US propensity to give priority to the intercontinental balance was again a factor. In the early 1970s, the NIEs did not anticipate, but did observe and report, the deployment of some 310 variable-range strategic ballistic missiles--ICBMs in terms of inherent range capability but deployed with an orientation to targets on the Eurasian periphery of the USSR and replacing intermediate range missiles. The United States properly insisted that because these missiles had intercontinental range they must be counted in ICBM levels in the Strategic Arms Limitation Talks, and the Soviet side accepted the point without argument. But for other purposes it was relevant that their assigned mission be understood and made clear in NIEs. This was done, although there were divided views.
More significant, and more contentious, both in the US Intelligence Community and in strategic arms negotiations with the Soviet Union, was the controversy in the mid- to late- 1970s over the Soviet Backfire (Tu-22M) medium bomber. Air Force Intelligence, sometimes joined by other intelligence services (and the Department of Defense in an interagency policy dispute over the arms negotiations) argued that the Backfire bomber had some intercontinental capability that could be enhanced, while CIA and others contended that it was most suitable for and would be deployed for peripheral theater missions (as it eventually was). There were other considerations, but the US propensity to judge such matters in terms of its own focus on the intercontinental strategic balance still obscured recognition of the Soviet view of its requirements for Eurasian theaters of operations. The same was true of the reaction of the United States and its NATO allies to the large-scale Soviet deployment of SS-20 IRBMs facing Europe in the late 1970s and early 1980s to replace the SS-4 and SS-5 missile deployments of the early 1960s.
There also was a persistent tendency to underestimate Soviet perceived requirements to provide air defenses against potential air attacks from the United States, NATO, and other countries in the Eurasian theater. Soviet air defense against putative US strategic air strikes was well understood, and naturally figured in US operational evaluations. US national intelligence estimates described Soviet air defense forces in detail and projected future forces reasonably well, but they focused on these forces primarily in the context of US strategic air power. Consequently, while finding Soviet air defenses extensive, the intelligence estimators also found them consistently deficient in relation to US offensive air capabilities and were perplexed at the heavy Soviet investment in ineffective air defenses. From the Soviet standpoint, these defenses were being measured not only against the US Strategic Air Command (SAC), but also against the other air forces of the United States, NATO, and other countries deployed in Europe, the Mediterranean, the Middle East, and the Far East.
Later, the continued Soviet investment in air defenses ineffective against US strategic air forces and operational capabilities led some US estimators, especially in the Air Force and Department of Defense, to speculate (incorrectly) that such defenses as the SA-5 air defense missile system must have an antimissile capability, since its high-altitude range was not comprehensible given SAC's operational tactics.
Soviet development, and deployment near Moscow, of an antiballistic missile (ABM) system in the 1960s was reported. Although stated only as a potential development, widespread Soviet ABM deployment (up to 7,000 ABM launchers) was forecast in NIEs in the late 1960s, prior to Soviet agreement to the ABM Treaty in SALT in 1972, and to a protocol in 1974 banning all but a single ABM deployment area limited to 100 launchers (the existing Moscow system).
When the comprehensive 11-4 series of estimates on Soviet military policies and capabilities was discontinued in 1961, two new series were initiated, one on intercontinental attack capabilities (11-8), and one on strategic defense (11-3) (after 1974 combined as 11-3/8), in addition to several other new series on various military fields of activity. One was a new annual series (11-14) beginning in 1962 on Capabilities of Soviet Theater Forces (later titled Capabilities of Soviet General Purpose Forces, and from 1967 on, The Soviet and Eastern European General Purpose Forces). Naval theater forces were included, but an additional occasional series on other naval missions (11-15) was also instituted.
Soviet naval forces contributed to strategic offensive capabilities through submarine-launched ballistic missiles (SLBMs) and to strategic defense capabilities through antisubmarine forces, to counter US and NATO offensive missile submarines and aircraft carriers. These forces, were included in the strategic attack and defense estimates. The chief naval elements of strategic defensive and ocean-control forces were surface ships and submarine-launched cruise missiles (SLCMs).
Beginning in the late 1960s, the Soviet Navy acquired a new, third mission, in addition to contributing to strategic offensive and defensive forces and support of land theater campaigns: showing the Soviet flag in distant theaters. Although sometimes termed "distant force projection capabilities," the limited Soviet naval, marine, and air forces capable of maintaining a presence at sea and at a few available foreign ports and bases were always more important as a presence in support of foreign policy purposes than as a potential combat force.
Soviet military involvement in the "third world" was limited mainly to military advisory personnel in conjunction with arms supply, although there were air and naval reconnaissance activities and important intelligence-collection bases for intercepting regional communications established in Cuba and Vietnam in the 1960s (as well as naval auxiliary ships contributing to that mission). Soviet Army and Air Force combat units were rarely based abroad (briefly in Cuba in 1962 and in Egypt in 1970-72). Soviet naval patrolling was maintained in the Mediterranean Sea from 1967 on, and on a sporadic basis in other oceanic theaters. NIEs occasionally examined Soviet military presence and other limited capabilities for projecting military influence in the third world, but chiefly in terms of Soviet foreign policy, rather than military policy.
In assessing the central Soviet concern for the military balance between Soviet and Warsaw Pact forces vis-à-vis the US and NATO, more attention began to be devoted to analyzing Soviet military doctrine and strategic thinking as, in the 1960s, estimates of Soviet capabilities matured. With the demise of the old comprehensive NIE 11-4 series, although some aspects of Soviet doctrine were raised in the new military series noted above, most discussion of Soviet military strategy and doctrine was found in a new 11-4 series, variably titled but usually denoted as dealing with "Soviet military policy." The first of these, before the new numbered series was regularized, was SNIE 11-14-61, The Soviet Strategic Military Posture, 1961-1967 (21 November 1961). It proceeded from the conclusion that the Soviet leaders saw their military power as a deterrent and a support to their foreign policy, but "their political outlook, their military programs of recent years, and intelligence on their current intentions, all suggest that the Soviet leaders do not regard general war as desirable or a Western attack on them as probable."  The estimate stated that "such high-level discourse as we know about does not revolve around the question of alternative attack strategies and target systems which are at the center of US military attention." Instead, it pointed out, the two chief contending schools of thought were divided into primarily military leaders who believed that despite nuclear and other new weapons, "general war is likely to be protracted ground combat on a mass scale," and others who believed that "a general war is likely to be short, with victory decided primarily in the initial nuclear exchange." The current official Soviet military doctrine at that time was described as an amalgam of both views. This high-level debate among Soviet military leaders and theorists was said to have been deliberately encouraged by the regime in order to spark original and creative thought. "As a result, strategic doctrine is a lively and argumentative field of professional study in the USSR today." And so it was. NIEs do not as a rule indicate the sources of their information, and this was especially true in this instance because the source of extensive documentary material on which these statements were based was a still-active CIA spy in Moscow, Colonel Oleg Penkovsky. Although the drafters and chief coordinators of this Special NIE had the requisite Ironbark clearance for the documentary materials supplied by Penkovsky--principally Secret and newly initiated Top Secret editions of the Soviet General Staff's theoretical organ Military Thought--the SNIE did not refer to those materials directly and did not carry the restrictive codeword, thus permitting more widespread dissemination of the conclusions based on this material. Many of the judgments were similar to earlier ones; for example, that despite public statements that limited and local wars with the West would inevitably escalate into general war, such statements were "not necessarily to be taken as expressions of Soviet military policy." The Soviet leaders would shy away from such limited engagements, and if involved would avoid escalation, while acting "to minimize the chance of escalation to general nuclear war." Remarkably, concluding that "the Soviet leaders do not regard war as desirable or a Western attack on them as probable" did not in this instance provoke dissent.
Although SNIE 11-14-61 was a useful step toward more discussion (based on having more intelligence data) of Soviet military doctrine in dealing with capabilities, it was not immune from several then-current errors noted earlier, notably, expressed expectations that the Soviet medium-range missile force would soon level off at much lower levels than it did in fact, and that numbers of medium bombers would drop much more sharply than they did. More important, it foreshadowed an underestimation of the Soviet ICBM buildup of the latter 1960s when it expressed the estimate that the ICBM force would reach "several hundred" in the period 1964-67. In fact, by mid-1967 the Soviet Union had nearly 700 operational ICBM launchers, and the number was rising rapidly. The worst error was arguing that "several hundred" would seem to meet their deterrent and retaliatory requirements. The NIE did not say the Soviets would stop at several hundred, but it implied as much, and did not suggest the force would continue to grow rapidly.
A number of NIEs (usually the 11-4 series) in the 1960s and 1970s continued to examine Soviet military doctrine and strategy, but more was done in CIA's Directorate of Intelligence (DDI) (and, on operational doctrine, in the newly established DIA). The analyses in the DI were usually circulated within the Intelligence Community but not beyond it. SNIE 11-14-61, in underestimating Soviet ICBM deployment in the late 1960s in a passing statement, was not in line with the basic NIEs on Soviet strategic capabilities at the time. The "missile gap" gross overestimates of 1959 to mid-1961 were greatly deflated in September 1961 in NIE 11-8/1-61, but even that estimate and NIE 11-8-62 in July 1962 continued to project somewhat higher current ICBM forces and still higher projected forces than were in fact acquired and deployed. The next estimate, NIE 11-8-63 (18 October 1963) was on the mark with its current and short-run estimates (for 1963 through 1965), but it estimated only some 400-700 ICBMs deployed by 1969; the actual count was 858 in mid-1969. NIE 11-8-64 was much worse in its outyear estimates: it deferred the previous year's estimates of 400-700 to mid-1970;  the actual number was 1,371 operational ICBM launchers. Even the usually higher Air Force dissent had been only 600-900. NIE 11-8-65 estimated between 500 and 800 ICBMs by mid-1970, and between 500 and 1,000 by mid-1975. The actual force in 1975, limited by SALT, was 1,601 ICBM launchers. These estimated forces, clearly far below what the Soviet Union would deploy, were based on the explicit assumption that the Soviets wanted "a satisfactory deterrent," with no regard for other purposes of strategic military power.  NIE 11-8-66 was no improvement; again, while current and near term figures were accurate, future projections were far too low: 800-1,100 by mid-1971, and 800-1,200 by mid-1976. The actual figure by then was the SALT-constrained level of 1,601. The increased levels over previous estimates, although still far too low, were justified by assumptions that the Soviets would wish to have a deterrent plus some damage limitation and a buttress to Soviet foreign policy, while seeking also not to lead the United States to outmatch the increase in Soviet forces. NIE 11-8-67, while somewhat better, estimated only 1,000-1,300 by 1972 and 1,000-1,500 by 1977 (the low side if the Soviets chose simply to match the US level). NIE 11-8-68 still estimated 1,100-1,500 by 1978, and NIE 11-8-69 estimated at least 1,300; CIA declined to estimate a maximum, the military services set 1,800 as a maximum.  NIE 11-8-70 counted 1,371 currently (including 80 oriented against targets in the Eurasian periphery) and 1,565 (including 120 with peripheral orientation) in 1972. The estimate contained three alternative future forces in a spread--from 1,350 to 2,100--so broad as to be of little use, but at least the actual SALT constrained force of 1,601 fell within it. NIE 11-8-71 also carried alternative projections, but by then the SALT I Interim Agreement of 1972 was taking shape, and there was no further increase in new ICBM launcher construction, so the size of the missile force was fixed. CIA analysis supporting the NIEs was very important to the intelligence process, even though sometimes in error. It also helped establish a benchmark, or bounding parameters, even when other members of the Intelligence Community dissented. No agency had a perfect record, but CIA's record overall was good and probably the best in the Intelligence Community. Moreover, it was the least influenced by an institutional interest, especially in contrast with the military intelligence services.
CIA also played a special role in supporting policy during the most demanding Cold War crisis, the Cuban missile crisis of 1962. Although CIA had presided over a unanimous Intelligence Community NIE conclusion that the Soviet leaders probably would not deploy long-range missiles in Cuba--a serious error in a political estimate--CIA and SAC U-2 overflights did monitor the military buildup in Cuba and detected the medium-range missile deployment under way there before it could be completed. Throughout the subsequent crisis President John F. Kennedy and his National Security Council (NSC) "Ex Comm" relied on daily reports and briefings from CIA on all intelligence aspects of the situation, military as well as political developments. This support was very effective and was as complete and accurate as available evidence made possible. (There were errors of omission owing to gaps in information, but intelligence performance should not be judged on the basis of expectation of perfection.)
In other crises, such as the Soviet intervention in Czechoslovakia in August 1968, nonintervention in Poland in late 1980 and again in late 1981, and military intervention in Afghanistan in December 1979, CIA and the other intelligence services were able to monitor the buildup of Soviet forces but could not say whether they would be committed. Capabilities for intervention could be monitored, but the intentions of the Soviet leaders were not known. Intelligence monitoring of noncrisis developments in Soviet-Western relations also was stronger in observing actions than in divining intentions.
Although Soviet military power was mainly addressed to the central Cold War confrontation with the United States and its NATO and other allies, the sharp deterioration of relations with China from the late 1950s on compelled Soviet attention also to threat assessment and generation of military requirements vis-à-vis China. This concern was fortified as the United States developed relations with China throughout the 1970s. After clashes on the Sino-Soviet border in 1969 and through the early 1980s, the Soviet Union built up its forces facing China from 15 to more than 50 Army divisions, with a corresponding buildup in air force, air defense, and other forces and infrastructure. This expanded agenda of Soviet threat assessment required Soviet attention to potential requirements for defense in multiple theaters of war and complicated their evaluation of requirements vis-à-vis the West. US military intelligence agencies and CIA monitored and reported on this range of developments in NIEs and other intelligence reports. This was a tangential, but not minor, aspect of US intelligence assessment of Soviet military power, intentions, and capabilities.
CIA also played a central role in providing intelligence support in the field of arms control and disarmament. Scientific and technical experts from the Atomic Energy Commission and outside laboratories played the main role in support of the earliest series of negotiations on a nuclear test ban in the 1950s and early 1960s. CIA and other members of the Intelligence Community played a more active role in supporting later negotiations, especially in SALT from 1969 to 1979, as well as later arms negotiations of the 1980s and early 1990s.
Even in the 1950s, CIA analysts briefed members of disarmament talks and participated in the first, brief but major negotiation, the "Surprise Attack Conference" in 1958. CIA drafted the first substantial National Intelligence Estimate on the topic of arms control and disarmament, SNIE 11-8-58, The Soviet Attitude Toward Disarmament, in 1958. It did not predict early serious Soviet negotiation, principally on the grounds of relative Soviet military weakness. But after "their position is further strengthened, especially through the advent of a substantial ICBM capability," the estimate concluded, in the longer term "both foreign policy and security motivations may lead to growing Soviet interest in expanding the areas of serious disarmament negotiations." This proved to be a prescient estimate a decade before SALT. It was equivocal, indeed somewhat contradictory (owing to the coordination process) in its judgment on whether the Soviets would seek eventual military superiority. Nonetheless, the Air Force Intelligence chief (correctly) believed the overall tenor of the estimate "suggests a Soviet willingness to curtail or limit the development of their capabilities to a level of deterrence rather than to seek the early attainment of an overpowering military superiority." In his dissent, he wanted the estimate to say that Soviet acceptance of any limitation would only be entered into with the intention of "furthering their drive for world domination" and would "in no way lead them to lessen their efforts to achieve an overpowering nuclear delivery capability at the earliest possible time."  In the later SALT and subsequent negotiations, CIA provided support through coordination in Washington, provided representatives on the US delegations, and later chaired an interagency arms control intelligence staff. CIA also took the lead in helping develop policy positions on verification requirements, on the extent of reliance on "national technical means" of monitoring compliance, and on interagency compliance monitoring.
CIA's expanding work on analysis and estimates of Soviet military power in the 1960s led to the consolidation of several smaller units in an Office of Strategic Research (OSR) in the DI in 1967. DCI William Colby in 1973 fundamentally restructured the entire estimates organization. He abolished the ONE and replaced it with a group of National Intelligence Officers (NIOs) responsible for various geographic and functional areas (one was NIO for the USSR, another NIO for [Soviet] Strategic Programs, another for [Soviet] General Purpose Forces). The NIOs each had a single deputy but no staff, depending instead on others in the Intelligence Community to provide analytic support. After 1979, the NIOs were members of a loose umbrella organization called the National Intelligence Council (NIC), but there was no equivalent of the ONE Staff or the Board of National Estimates to provide serious collegial review of estimates. This change had many effects, some probably good but some less clearly so. For example, with separate, coequal NIOs for the USSR, Strategic Programs, General Purpose Forces, Science and Technology, and the like, there was no one with responsibility for the overall integrative coverage of Soviet military intentions, capabilities, and doctrine. The DIA played a larger role than before, particularly since the NIOs for General Purpose Forces (now called Conventional Military Issues) were usually generals, and in the 1980s the NIO for Strategic and Nuclear Programs was a former Defense Department specialist rather than a CIA officer. CIA analytic work continued to be important but in some respects it was less central, and this is evident in the NIEs. Also, within CIA in 1981 functional offices of the CIA's DI (political, economic, military affairs) were dissolved and replaced by geographic offices, one of which was the Office of Soviet Analysis (called SOVA).
One result of the mistaken low estimates of future levels of Soviet ICBMs that the Agency made from 1963 through 1968 was a heightened concern over the actual growing strategic forces from the late 1960s to the mid-1970s. For one thing, this led to a reconsideration of Soviet objectives; for another, it affected future estimates of Soviet capabilities against US forces, in particular the Minuteman ICBM force. Ultimately, it led to what may be termed "the second missile gap" alarm of the 1970s and early 1980s, a concern over qualitative improvements of the already large Soviet ICBM force that seemed to many to portend a "window of vulnerability" for the US deterrent and a "window of opportunity" for the USSR. The earlier described underestimation of the ultimate size of the Soviet ICBM deployment from 1963-68 was based in part on an erroneous assumption that the Soviet leaders would consider a strategic force perhaps even smaller than the one US had already fielded, or at least of comparable size, as sufficient for deterrence. NIE 11-8-64 had even stated, without dissent, that the estimators did not believe the Soviets would aim at matching the US in the overall number of strategic nuclear delivery vehicles (SNDVs). This was reiterated for the next four years. In 1966 the Air Force registered a mild dissent that they believed the Soviets did seek "at least numerical parity with the US" as a goal.
The reluctance of estimators in the mid to late 1960s to see parity with the United States as a Soviet objective is difficult to understand. They correctly saw that basic deterrent objectives could be met with a few hundred protected ICBMs. But the very fact that the United States, also primarily concerned with assuring deterrence, had already opted for more than 1,000 ICBMs and was rapidly deploying them (plus a very large and capable submarine-launched ballistic missile force) should have suggested both that the Soviets also would seek a redundant force to ensure a survivable deterrent and that they would want to emulate the United States if only to signify they had parity in strategic military power as they were striving for political parity.
Parity, of course, need not mean equal numbers of SNDVs, and could mean still less of each kind of strategic delivery system. In 1969, the strategic forces NIE finally concluded that "rough parity with the US" was the minimum Soviet objective and correctly noted that "in assessing the strategic balance the Soviets would go beyond numbers to consider operational differences in weapons systems and the interplay between offensive and defensive forces." NIEs in the early 1970s suggested that parity as the Soviets sought it "cannot be objectively measured; it is essentially a state of mind." From 1968 through 1972, as the Soviets reached and passed numerical parity with the United States in SNDVs operational and under construction (and with the US force level not increasing), the NIEs finally recognized that the Soviet leaders were seeking at least parity in numbers; the estimates also assessed them as believing that "marked superiority" was not attainable.  By 1973 the NIE stated that the Soviets had probably decided on no less than parity and would seek "a margin of [numerical] superiority if they can" (that is, if the US did not match a Soviet increase). In 1974, the CIA drafters of NIE 11-3/8-74 accepted that the major Soviet force-enhancement programs under way raised the question of whether Soviet leaders were seeking "some form of militarily or politically useful strategic superiority," but their answer was to doubt whether the leadership had "firmly settled on either acceptance of parity or a decision to seek clear-cut strategic superiority." The NIE reiterated the earlier estimates that the Soviets probably sought to achieve no less than equality with the United States, plus "some degree of strategic advantage" if they could. The three military services dissented, saying they believed the Soviets foresaw "a decisive shift of the strategic balance in their favor" yielding a superiority "umbrella" of extended deterrence permitting a freer pursuit of their objectives in the world without US "interference."  This divergence continued for the next several years, although by NIE 11-3/8-76 the CIA text suggested that the Soviet leaders "hope" that their strategic capabilities will "give them more latitude than they have had in the past for the vigorous pursuit of foreign policy objectives, and that these capabilities will discourage the US and others from using force or the threat of force to influence Soviet actions."
While this approximated the dissenting military services' position of 1974 (reaffirmed in 1975), by December 1976 all the military services went further in a new dissent that argued, inter alia, that "in combination with other military and non-military developments, the buildup of intercontinental nuclear capabilities is integral to a programmed Soviet effort to achieve the ultimate goal of a dominant position in the world," which they expected to move closer to in the next decade (the period of the estimate). The Air Force additionally declared that this estimate and its predecessors understated "the Soviet drive for strategic superiority" and asserted that the Soviets "have made great strides toward achieving general military superiority over all perceived constellations of enemies and for attaining a war-winning capability at all levels of conflict." The Air Force dissent further cast doubt on the past US conduct of a policy of détente, arms control, and trade, contending that the Soviet Union had "exploited" them to the serious disadvantage of the West. Finally, Air Force Intelligence went so far as to state that "at issue is whether present intelligence prescriptions [as expressed in the NIE] provide an adequate basis for averting global conflict in the decades ahead." This time there was a countering dissent by State Department Intelligence, arguing that the Soviet Union was undertaking a vigorous strategic program in order not to fall behind the United States in the qualitative competition, and expressing doubt that "the Soviet leaders anticipate any improvement in the USSR's strategic position vis-à-vis the US over the next 10 years which would substantially influence their behavior."  By the time of NIE 11-3/8-79 (March 1980), the CIA text (this one written in large part by DCI Stansfield Turner) had reverted to describing the Soviet position as one of "approximate nuclear parity" and mutual deterrence, and it concluded that the Soviets might feel they had a somewhat freer hand only in places where the US regarded its interests as less central, such as Afghanistan. There was no State Department dissent, but DIA and the military services all registered an overall dissent from the main volume of the estimate and further expressed the view that the Soviets "may now perceive that they have nuclear superiority," which would grow over the next few years, and might miscalculate US reactions because of uncertainty about US resolve. 
NIE 11-3/8-76 had also revived the issue of Soviet military doctrine on waging, surviving, and winning a general nuclear war. The estimate took the view that "the Soviets are striving to achieve war-fighting and war-survival capabilities that would leave the USSR in a better position than the US if war occurred." It did not say whether this was regarded as a realistic objective even in the long term. The State Department dissented, arguing that the Soviet leaders "do not entertain, as a practical objective in the foreseeable future, the achievement of what could reasonably be characterized as a ‘war-winning' or ‘war-survival' posture." The military intelligence services all joined in a dissent, expressing the belief that the Soviets did regard as "attainable" an objective of achieving the capability to wage and emerge from a nuclear war "with reserves sufficient to dominate the postwar period."  The same three-way divergence emerged again in NIE 11-3/8-80 three years later. Meanwhile, an unusual brief estimate on Soviet Strategic Objectives in 1977 had posed this and related issues, reaffirming the three diverging views without taking a position on the matter, but with an admittedly "more ominous interpretation" of Soviet aims. 
The ‘Team B' Exercise
The renewed debate over whether meaningful strategic superiority was a Soviet objective and over Soviet doctrine and the issue of winnability of nuclear war, had arisen in NIE 11-3/8-76 as well as later estimates, fuelled by an unusual exercise in preparation of an alternative "competitive estimate" by an outside panel, called "Team B." Owing to the growing unease over increasing Soviet strategic forces and capabilities in the mid-1970s, and criticism of détente and arms control from some quarters, several hard-line members of the President's Foreign Intelligence Advisory Board (PFIAB) began to criticize the NIEs (in particular NIEs 11-3/8-74 and 11-3/8-75) as too complacent both in respect to growing Soviet capabilities, as well as Soviet doctrine and intentions. This was when some of these members and others were organizing what became the Committee on the Present Danger and calling for a repudiation of détente and the SALT negotiation process. Their proposal was for an outside group with more skeptical and critical views of Soviet objectives to be cleared for the same data as the official Intelligence Community estimators and asked to critique the official estimate and draft an alternative one. At first rejected by DCI William Colby, it was agreed to in 1976 by DCI George Bush. The group did its work in the summer and fall of 1976 while NIE 11-3/8-76 was being written and coordinated by the usual interagency estimating process.
The Team B report on Soviet objectives dealt not only with Soviet intentions, objectives, and doctrine, but also with a number of aspects of strategic force development and capabilities. The report was highly critical of the conclusions of the official estimates, and in particular of the CIA drafters.  Moreover, the general thrust of Team B's critical conclusions was swiftly leaked to the press.
The Team B exercise was ill conceived and disappointing in its limited contribution to identifying ways to improve the estimating process, in part because its authors were more concerned with criticizing the message of the NIEs and pushing their own hard-line views. The process terminated without full-scale confrontation of diverging views. The direct impact on the estimating process and on the content of the NIE 11-3/8 estimates was small. As noted above, it did give a boost to the similarly minded Air Force Intelligence chiefs and to some extent to other military intelligence agencies on questions of Soviet objectives, and it made some CIA statements on that subject in the NIEs more cautious and equivocal. But it did not lead to significant changes. (While some of the Team B criticisms on a few specific aspects of the estimated force developments and capabilities seemed to be reflected in subsequent estimates, these were cases where additional evidence, not the arguments of Team B, led to the changes.)
In retrospect, and with the Team B report and records now largely declassified, it is possible to see that virtually all of Team B's criticisms of the NIE proved to be wrong. On several important specific points it wrongly criticized and "corrected" the official estimates, always in the direction of enlarging the impression of danger and threat. For example, the range of the Backfire medium bomber was considerably overestimated, and the number of Backfires the Soviet Union would acquire by 1984 was overestimated by more than 100 percent (estimating 500 when the real figure was 235). Team B overestimated the accuracy of the SS-18 and SS-19 ICBMs, feeding the unwarranted fears of a "window of vulnerability" for the US ICBM deterrent. Team B estimated that the Soviet Union would field a mobile ABM system, which it did not. It regarded as ominous, rather than reassuring, that no intelligence information had been acquired on Soviet development of a nonacoustic antisubmarine warfare capability, again raising concerns over a looming threat that did not arise.
Team B saw as a "serious concern" the possible upgrading of Soviet mobile intermediate range missiles (SS-20s) to ICBMs and criticized the draft NIE for estimating that the SS-16 mobile ICBM program would remain small. In the event, no SS-16s were deployed, and no SS-20s were upgraded to ICBMs. With respect to exotic technologies for ABM defense, Team B castigated the NIE for failing to draw more attention to the threat of Soviet development of charged particle-beam directed energy interceptors, stating that it would be "difficult to overestimate" the magnitude of the Soviet effort, yet by those very alarmist words it did so. The large-scale but ineffective Soviet civil defense efforts were also depicted as an important part of a Soviet design to be able to fight, and win, a nuclear war. Team B even suggested incredibly that the ABM Treaty helped the Soviet leaders "to pursue a goal of achieving assured survival of the USSR and assured destruction for its major adversary."
Team B also reported "an intense military buildup in nuclear as well as conventional forces" and criticized the NIEs for failing to describe adequately the scale of the Soviet military effort. While Team B was estimating a relentless, continuing buildup at a growing pace, it was later learned that, in fact, Soviet leaders had just cut back the rate of spending on their military effort and would not increase it for the next nine years. To be sure, the Soviet Union continued to spend a great deal on its large military programs, but it was not the limitless buildup in pursuit of a war-winning capability that Team B ascribed. Team B went even further. Its report argued at length that there was no constraining effect resulting from the requirements of the civilian economy. The NIEs were attacked for even suggesting that economic considerations might limit Soviet military growth, and Team B itself asserted that "Soviet strategic forces have yet to reflect any constraining effect of civil economy competition, and are unlikely to do so in the foreseeable future."
"Intentions" are, of course, even more difficult to estimate--and more liable to intentional or unintentional subjective distortion and error--than are evaluations of forces and capabilities. The main Team B report had focused on Soviet objectives and intentions and had presented an unjustified, ominous picture. Team B gratuitously and without foundation asserted that the Soviet arms buildup "certainly exceeds any requirement for mutual deterrence," which implicitly raised the question of why, then, the United States-- with such a deterrence objective--should have built up even larger strategic forces. Team B sought to make a case for a uniquely dangerous Soviet pursuit of world domination and readiness to countenance the use of its growing military power. It asserted, in a demonstrably untrue statement, that, "There is no evidence either in their theoretical writings or in their actions that Soviet leaders have embraced the US doctrine of mutual assured destruction or any of its corollaries. Neither nuclear stability, nor strategic sufficiency, nor ‘parity' play any noticeable role in Soviet military thinking." Moreover, it asserted that "while hoping to crush the ‘capitalist' realm by other than military means, the Soviet Union was nevertheless preparing for a Third World War as if it were unavoidable." Military power, it was alleged, was seen by Moscow as "an instrument by means of which, in the decisive moment in the struggle for world hegemony, the retaliatory power of the United States can be preventively neutralized, or, if necessary, actively broken." Finally, in the judgment of Team B, in a conclusion stated in italics for emphasis: "Within the ten year period of the National Estimate the Soviets may well expect to achieve a degree of military superiority which would permit a dramatically more aggressive pursuit of their hegemonic objectives, including direct military challenges to Western vital interests, in the belief that such superior military force can pressure the West to acquiesce or, if not, can be used to win a military contest at any level." 
There was room for debate as to Soviet objectives and predictive estimates in 1976, but it was evident at the time, and all the more so with the benefit of hindsight, that NIEs were not the place to argue for hard-line, or any other, policy. As it was, the NIEs in this period and later generally erred on the side of overstating the military threat.
Team B was so bold as to suggest a crucial military confrontation with the Soviet Union within the ten-year span of the estimate (a standard maximum span for such NIEs). Yet ten years later we saw not a Soviet Union threatening or launching global nuclear war for world domination, confident of its superiority, but an ever weakening Soviet Union with a radically revisionist leadership under Mikhail Gorbachev urging radical disarmament and launching a mix of concessionary negotiations and unilateral measures to decisively turn down the arms race and end the Cold War. It is clear in retrospect that what had been needed in 1976 was not a hard-line Team B, but a more imaginative and far-seeing "Team C."
As noted earlier, underlying the renewed debate over Soviet objectives in the 1970s and 1980s (e.g., parity versus superiority) --including the whole Team B episode--was the rise in Soviet capabilities as Soviet forces grew in size and quality. The estimates gave some attention to Soviet reactions to ongoing US developments, but not enough. For example, Team B and others in and out of the Intelligence Community asserted that the Soviet buildup in numbers of ICBMs and total SNDVs was "unprecedented" and represented a "relentless" buildup somehow inconsistent with détente and arms control. The Soviet buildup was rapid and large--1,050 SNDVs were added in the five years 1968 through 1972 (when the SALT I accord set a limit on ICBM and SLBM launchers). Yet this was hardly "unprecedented," nor did it upset the strategic balance, nor was it inconsistent either with détente or the arms control agreements or positions of both sides. In the preceding five-year period, 1963-67, the United States had added 1,060 SNDVs to its arsenal. Nor was it even very relevant. The reason the United States had unilaterally stopped building up numbers of SNDVs in 1967 was not because "the US faltered in the 1970s" or exercised "unilateral restraint" (as a NIC intelligence assessment asserted in 1983). It was because the US had shifted to a more effective way of enhancing its strategic superiority: increasing the number of multiple independently targeted reentry vehicles (MIRVs) on ICBMs and SLBMs. Accordingly, as the Soviets moved beyond parity in the number of ICBMs on operational launchers in 1969, and beyond parity in the overall number of SNDVs (including also SLBMs and intercontinental bombers) in 1973, they were falling farther and farther behind the United States in overall number of reentry vehicles (RVs, warheads, and bombs). By 1975, when the first Soviet MIRVed missiles became operational (fully five years after the United States had begun to MIRV its forces), the United States had greatly enlarged its superiority in numbers of warheads (RVs). Even after the Soviet Union added 1,000 RVs a year in the late 1970s, in 1980 the United States had a superiority of 9,200 operational RVs to the Soviet's 6,000. Even if the Soviet leaders had set a goal of strategic parity in terms of numbers of strategic warheads, they never achieved it. The Soviet Union came closest to equality only in 1990, on the verge of its collapse (with 10,700 Soviet to 12,700 US strategic intercontinental RVs).
Military planners, of course, had to consider many operational factors, including initiation of a war, in weighing the survivability and deliverability of warheads by both sides in a war. The technical complexities of theoretical advantage to one side or the other are immense. But they are also largely irrelevant. As the NIEs regularly stated, without dissent, the Soviet leaders never had capabilities that could give them confidence in surviving and winning a nuclear war. The debate was always over whether they believed such superiority was attainable and whether they were driven to seek to achieve it. The debates could never be resolved by intelligence information and even now cannot be fully resolved, although there has been no evidence that the Soviet leaders ever believed this to be the case, and in retrospect this seems clear.
Debating US Vulnerability
Although (or because) Soviet intentions could never be ascertained with assurance, the critical debates from the mid-1970s to the mid-1980s were over Soviet capabilities--or, to be more precise, overestimated Soviet capabilities--in the mid- to longer-term. From the late 1960s on, the key issue was whether and when the US ICBM force (1,000 Minuteman and 54 Titan missiles in silo launchers) would become vulnerable to a Soviet surprise attack--in particular, [they wondered] when the Soviet Union would have enough ICBMs with sufficient accuracy to threaten the Minuteman force.
From 1972 on, the level of Soviet missile forces--operational ICBM and SLBM launchers--was established by the SALT I Interim Agreement of five-year duration, and the NIEs from 1971 anticipated this. Given the steady level of Soviet heavy bombers at about 150 (with another 50 as tankers) from the late 1950s on, the buildup of SNDVs had for all practical (and estimative) purposes ended. But, as noted above, the United States had already set the new measure of forces some years earlier: the number of RVs, which were not limited by the arms control negotiations (and on which the United States had made clear it would accept no limitation).
The NIEs in the late 1960s and early 1970s overestimated how soon the Soviet Union would have MIRVs and join the RV race. In 1969 there was a dispute over whether a version of the SS-9 (Mod 4) was being developed as a MIRV. In an unprecedented interference with the intelligence estimating process, in June 1969 National Security Adviser Henry Kissinger interceded with DCI Richard Helms to "clarify" (in effect, to reconsider) a statement in the draft NIE (NIE 11-8-68) that the SS-9 variant was not a MIRV, a judgment contradicting a recent public statement by Secretary of Defense Melvin Laird. Helms asked the chairman of the Board of National Estimates to review the estimate on the SS-9, and the language--but not the basic finding--was revised. That ended that particular intervention. But a few months later, the issue of political influence arose again in a broader statement in the follow-on draft NIE (NIE 11-3/8-69). This time Laird interceded with Helms (through his aide William Baroody) to object to a paragraph stating that, even if the Soviets sought to achieve a strategic superiority sufficient to alter the strategic balance, they would not find it feasible to acquire a capability to attack the United States without receiving unacceptable retaliation. Helms did agree to delete that particular statement, although the sense of the Estimate was not changed. Even so, the State Department intelligence bureau inserted the withdrawn passage as a dissent.  By the next year or so it was accepted by all that the SS-9 Mod 4 did not carry a MIRV warhead. Each year thereafter, the NIEs estimated the Soviets would introduce MIRVed ICBMs soon and finally, in 1975, they did.
From 1970 through 1977 the NIEs tended to overestimate the numbers of Soviet RVs, but they projected a range of possible numbers and the actual figures were within the range estimates (although a "high within SALT" force included in the range in 1973-74 was twice as high as the actual level). Later, in every year from 1978 through 1985 even the lowest side of the estimated range exceeded the actual Soviet deployment of RVs. After the United States did not ratify the SALT II agreement signed in 1979, and the further limitations provided for by it were assumed not to be operative, the estimated range was much higher than the actual levels. Moreover, the estimates used illustrative "SALT-constrained," non-SALT-constrained, and SALT "breakout" alternative forces, so even when the SALT-constrained estimates were not excessively high, the other "equally valid" alternatives conveyed a pattern and impression of much greater buildup than in fact occurred.
Perhaps most critical for the key issue of Minuteman vulnerability were the estimates of missile accuracy. In the early 1970s, Soviet missile accuracies tended to be underestimated. Perhaps in part in overcompensation, the estimates from 1977 through 1980 in turn overestimated Soviet missile accuracies, although this did not become clear until 1983. Although these overestimates of the late 1970s were closer to Team B's alarmist projections, they were not influenced by the unproven assertions of Team B but by the fact that tests of the SS-18 Mod 4 and SS-19 Mod 3 in late 1977 and early 1978 yielded better results than expected.
NIE 11-8-73 had predicted a high threat to Minuteman by 1982, although it (and its successors) correctly noted that the overall US deterrent capability would not be threatened. By late 1974, NIE 11-3/8-74 put off the expected high threat to Minuteman to 1981-85. By 1978, because of mistaken newly estimated higher accuracies, it was estimated that the Minuteman force would be vulnerable by 1980-81. This opened the period of high concern over a "window of vulnerability" for the United States and a possible "window of opportunity" for the USSR.
By early 1984, when a delayed NIE 11-3/8-83 was issued, the reassessment of Soviet missile accuracy undercut the window of vulnerability alarm. So, too, did the 1983 report of the blue-ribbon Scowcroft Commission, even though it still had used the old, more alarmist NIE 11-3/8-82. The Scowcroft Commission made the same point that had been made a decade earlier in NIE 11-8-73, that even if the Minuteman force were severely threatened by the Soviet ICBM force (unless, of course, launched on warning or under attack), the United States had created a redundant triad of deterrent forces precisely to deal with such a contingency. NIE 11-3/8-84/85 gave the window its coup de grace by putting off the estimated time of Minuteman vulnerability by a full ten years, to the early 1990s. 
This was the most critical issue in the assessment of Soviet strategic capabilities in the 1970s and 1980s. There were many lesser issues, although some were significant in their negative impact on the arms control negotiations or in other ways. For example, from the mid-1970s into the 1980s, the Air Force (sometimes joined by other military services) revived the error of the 1950s by trying to create an intercontinental threat out of the Backfire medium bomber, obviously designed and later trained and deployed, as CIA consistently argued, for missions in the peripheral Eurasian potential theater of war. (This too was one of the red herrings boosted by Team B.)
The NIEs also projected a new Soviet heavy bomber to replace the 1950s vintage Bison and Bear bombers. In 1972, NIE 11-8-72 estimated a new heavy bomber would enter service in 1978. This date had to be put off time and again; a new heavy bomber finally appeared, in very small numbers, beginning only a decade later, in 1988. Estimates in 1975 projected a replacement of 94 percent of the old heavy bomber force by 1985; in fact, except for about a 15-percent replacement of newer-model Bears, there was no replacement until the trickle after 1988. 
Nor was that example exceptional, albeit extreme. From 1974 through 1986, every year's NIE 11-3/8 overestimated the rate of Soviet strategic force modernization. The initial deployment of new or modernized systems was overestimated in 10 out of 17 systems (and underestimated in only one). The rate of deployment of modernized systems was also generally overestimated. 
There were other alarms raised during the late 1970s and early 1980s to which intelligence overestimates contributed. Team B, for example, argued that the Soviet Union had a large and effective civil defense program as part of an overall design for survival and victory in a nuclear war. CIA had been studying the Soviet civil defense system since the 1950s, although it was not a priority target. Even before the Team B criticism, CIA had been reexamining the evidence and noted in NIE 11-3/8-76 that the Soviet civil defense program was "more extensive and better developed than we had previously understood." The estimate nonetheless concluded that "The Soviets probably do not have a highly optimistic view about the effectiveness of their present civil defense."  The Air Force, which had been pushing the issue, in a broader dissent included a hyperbolic assertion that "War survival and civil defense efforts to date have already placed the US in a position of serious strategic disadvantage by neutralizing much of its capability to destroy or damage effectively those elements of the Soviet leadership, command, military, and urban-industrial structure required for maintaining a credible deterrent balance." The State Department intelligence bureau went further than the estimate and directly challenged the Air Force view, stating that the Soviets simply saw civil defense as a "prudent hedge," and that Soviet "civil defense efforts will not materially increase Soviet willingness to risk a nuclear exchange and will not undermine the deterrent value of US strategic attack forces." Civil defense was a traditional Soviet program to provide damage limitation and had no necessary relationship to concepts of a war-winning strategy. Yet the other military intelligence services essentially supported the Team B and Air Force argument that the Soviet leaders saw their civil defense program "as an essential element in the achievement of the capability to wage intercontinental nuclear war, should one occur, and survive with resources sufficient to dominate the postwar period."
By NIE 11-3/8-79, CIA had carried out further studies showing that prompt Soviet casualties would be about 120 million (with 85 million dead) in case of little or no preparation for an attack, about 100 million (60 million fatalities) if urban shelters were fully occupied, and 40 million (with 15 million dead) if both sheltering and evacuation of cities had been fully implemented--and those were only immediate casualties. Moreover, those casualty figures would rise during the 1980s. The CIA text concluded a further discussion with the comment, "Our latest analyses . . . provide additional support to our previous judgment that present and projected Soviet civil defense programs would not embolden the Soviet leaders to take action during a crisis that would involve deliberately accepting a high risk of nuclear war." DIA and the military services expressed support for an "alternative view" that Soviet civil defense would contribute to Soviet resolve in a crisis, and thus affect "the strategic balance and perception of the balance in the USSR and elsewhere."
Although NIE 11-3/8-80 included a long discussion of Soviet doctrine on waging a nuclear war, there was only one paragraph on civil defense, increasing the estimated overall casualties to be expected, but omitting any discussion of the impact of civil defense on Soviet policy in crisis or war. And there was no dissent. The "civil defense gap" had essentially evaporated, with only passing reference to the subject in later NIEs. The NIEs did, however, describe the extensive efforts being made to provide protection for the Soviet leadership to ensure wartime command and control.
Another index of the strategic balance derived from the NIEs, defense spending, was publicly depicted in the late 1970s and early 1980s by both Carter and Reagan administration spokesmen as indicating an alarming shortfall in US defense expenditures. From 1976 to 1983, it was estimated that Soviet defense spending was growing at a rate of 4-5 percent per year. In short, there was a serious "spending gap." Then in 1982, analysts of the DI in CIA determined that in fact the rate of growth of Soviet defense spending since 1976 had been an average of only about 2 percent, and in weapons procurement the rate was almost flat, zero percent growth. While an encouraging finding from the standpoint of comparing performances between the Soviet Union and the West, it was a political embarrassment from the standpoint of the administration's campaign to boost support in the United States (and in NATO) for higher defense spending. In 1983, the new findings were made known without fanfare in testimony before the Joint Economic Committee of Congress. Administration spokesmen such as Secretary of Defense Caspar Weinberger simply stopped referring to the matter. The "spending gap," no longer useful, was quietly buried. 
During the period from 1980 (after the Soviet intervention in Afghanistan and then the change of US administrations) to the mid-1980s, US policy toward the Soviet Union was more consistently confrontational than it had been before. Given the controversies within the Intelligence Community and in broader policy and public forums, as exemplified in the position of the PFIAB and the Team B exercise, this raised a question with respect to the position of CIA and the estimative process, particularly in view of the fact that former PFIAB (and Committee on the Present Danger) member William Casey was the DCI after January 1981.
Although the Reagan campaign had emphasized the need to meet an ominous Soviet threat (including a military challenge that the previous three administrations were said not to have met adequately in a "decade of neglect"), and the new administration set out to build American military power, there was no quarrel with the Intelligence Community's NIEs. Director Casey did not involve himself directly in the Soviet military estimates as his predecessor, Stansfield Turner, had done. (Richard Pipes, now on the National Security Council staff, instituted an inquiry into what CIA had done with the Team B report four years earlier, but that was strictly a personal interest by the head of Team B and the new administration did not pursue the matter.)
The OSR in CIA's Directorate of Intelligence, in what proved to be a kind of valedictory, issued in April 1981 a comprehensive survey on The Development of Soviet Military Power: Trends Since 1965 and Prospects for the 1980s. As a comprehensive survey, it tended to present overall Soviet military power as a somber challenge. It was a more deeply researched and balanced study than the coordinated NIEs, although of course it was compatible with the current estimates.
One issue that had attended the national estimating process from the outset, and was raised directly in 1979 and 1980 by DCI Turner, was the question of including net assessments in NIEs. In NIE 11-3/8-79, Director Turner introduced the use of what he termed "quasi-dynamic indicators" of what forces could actually accomplish--short of full-scale war gaming, but beyond static indicators of force size and performance characteristics. This crossed a line laid down in the late 1940s, since dynamic evaluations by definition involved judgments not only of Soviet force performance but also of US and other Western force performance in their interaction, and the military services believed only the Pentagon should make such evaluations of our own forces. Moreover, the effect of Turner's approach was to reduce the impression of Soviet military capabilities and the confidence Soviet military leaders could have in those capabilities. The Director of DIA and all the military intelligence chiefs strongly objected. 
Questioning Net Assessments
In NIE 11-3/8-80, the controversy led to an unprecedented issuance of a split estimate, with two entire sets of key judgments, one representing the view of the DCI and the other that of the Director of DIA and the heads of all the military intelligence services. Elaborating on their dissent in the previous estimate, the DoD argued that the CIA estimate "produces misleading results with respect to trends in the military balance, sheds little light on the question of deterrence or escalation control, and comprises an unrealistic net assessment." They did not offer their own. Instead, they presented a traditional static description of Soviet forces and gross capabilities, and vigorously presented their objection of principle: "net assessments from a US perspective are not a proper function of intelligence . . . analysis based on a US perspective should be accomplished within the Department of Defense with intelligence as a full partner, and should not be included in a National Intelligence Estimate." The DCI directly rebutted this position and added that he did not "believe it in the national interest that DOD [the Department of Defense] should control all comparisons of the effectiveness of its forces with other forces." 
The divergence in terms of estimates of future Soviet behavior was clearer in tone than in substance. The DCI cited the historical record to conclude that "the relationship between the strategic balance and Soviet behavior in the international arena is uncertain," but agreed that "during the early-to-middle 1980s, when the Soviets' strategic capabilities relative to those of the United States would be greatest, we would expect them--as in the past--to probe and challenge the United States . . . " but that it was "highly unlikely" they would "go for broke" and "take military action against Western Europe or the United States." The heads of DIA and the military services argued "the Soviet leadership is now confident that the strategic military balance has shifted in the Kremlin's favor and that the aggressiveness of its foreign policy will continue to increase as the Soviet advantage grows." 
This estimate, it should be recalled, was made only months after the Soviet invasion of Afghanistan and in the midst of an election-year debate over Soviet strength and charges of US military weakness. In retrospect, it is not clear that Soviet strength was at its greatest in the early to mid-1980s, as both sides of the US intelligence estimative battle then conceded. What is clear is that the Soviet leadership did not believe the strategic balance had shifted in its favor or that its hands were freed to be more bold and aggressive in its initiatives. Indeed, in 1981, Soviet intelligence instituted an unparalleled alert against the possibility of a US and NATO surprise nuclear missile attack that persisted through much of the decade, with a peak alarm in late 1983. 
DCI Casey did not share Turner's interest in more actively controlling the military estimates, but in SNIE 11-4/2-81 he did reassert the right to make net assessments in NIEs, incurring the same dissent from DIA and the military intelligence services.  As in the early 1960s when the issue had previously been raised, the resolution agreed upon in the early 1980s was for CIA to work more closely with the net assessments office in the Department of Defense and not to include net assessments in NIEs.
Another issue with respect to the Soviet military assessments arose in the early 1980s, this time not as an issue between the DCI and the military services, but between members of Congress and analysts in CIA's Directorate of Intelligence on the one hand, and the DCI as well as the military services on the other. The issue was whether to continue to make overall static assessments of Soviet strategic capabilities and capabilities for theater operations in Europe relative only to US forces and capabilities. CIA analysts (and some in the military) saw the more relevant comparison as Warsaw Pact versus NATO forces and capabilities--the United States and the Soviet Union would not (even could not) fight alone in Europe. But those who wished to place emphasis on "the Soviet threat" did not want to diminish it by adding on the forces of the NATO allies.
Despite the importuning by analysts in CIA's Directorate of Intelligence, and independently by knowledgeable members of Congress asking for NATO-Warsaw Pact comparisons, DCI Casey and Deputy Director of Intelligence (DDI) Robert Gates (later Deputy Director of Central Intelligence [DDCI]) dragged their heels. NIEs on Soviet and Warsaw Pact capabilities had, of course, as earlier noted, been made ever since 1962. But in the overall assessments made to Congress and the public in the 1970s and early 1980s the US-USSR comparisons were used, for example, on levels of military spending or military manpower, as well as nuclear capabilities. Only late in the 1980s was this to some extent rectified.
Although net assessments and alliance capabilities assessments were eschewed, some interactive comparisons--of a new kind--did enter NIEs in the late 1970s and early 1980s. Assessments of Soviet perceptions of Western policies and capabilities were of course a legitimate and necessary consideration in intelligence assessments of Soviet policy. Sometimes, however, the estimates conveyed only thinly veiled judgments on US policy. Thus, for example, NIE 11-4-78 estimated that the "more assertive Soviet international behavior" observed in Angola, Ethiopia, and Afghanistan (even before the Soviet military intervention) in the latter half of the 1970s was "likely to persist as long as the USSR perceives that Western strength is declining" and its own strength is steadily increasing. That was a legitimate (if, in retrospect, questionable) assessment, although one senses the authors of the estimate themselves held the view that US and Western strength was declining.
By the early 1980s, direct statements by the estimators, no longer attributed to Soviet perceptions, entered the NIEs. NIE 11-16-83X in March 1983 referred to "the fact that the United States did not make a commitment to strategic force improvements comparable to that demonstrated by the magnitude and vigor of Soviet programs."  And a NIC memorandum in August 1983 stated flatly "the US faltered in the 1970s . . . while the Soviets continued modernization" of their strategic programs, and alleged "US unilateral restraint."  These were net assessments or assessments of US policy to which, however, the Pentagon raised no objection. Similar statements were being made by administration spokesmen from the President down, but intelligence assessments were not supposed to address US policy or performance.
The Reagan administration did introduce a new version of intelligence assessment to inform (and influence) public opinion. In 1981, and then from 1983 through 1990, the Department of Defense issued annually a lavishly illustrated publication titled Soviet Military Power. It presented an ominous picture of a massive Soviet military buildup, without adequate indication of countervailing American and NATO military forces or programs. Although based on national intelligence (plus uncoordinated Defense Department analysis), the data were selected and embellished to magnify the impression of a threat and to rally support for the US military buildup. DIA prepared Soviet Military Power, with informal consultation with CIA (and other) analysts, but the series was not coordinated nor was it an intelligence product; it was a Department of Defense public information publication.
By 1982, the strategic research analysts at CIA had discovered a disturbing trend in the annual NIEs estimating Soviet strategic capabilities: the aforementioned consistent overestimation of strategic forces modernization and enhancement of capabilities since the mid-1970s. The problem was simple, but any solution would be complex, and difficult to obtain agreement on. While projections of future programs for any individual weapons system or category of weapons might be reasonable, the aggregate force projections were more than the Soviets procured and would have required a massive increase in spending that was highly unlikely. Moreover, although the NIE 11-3/8 series now estimated "low" and "high" forces as general parameters, overestimation of the aggregate was consistently true even for the "low" force projections. The CIA analysts became aware of this problem in 1982, but they were unable to determine which of the individual programs should be scaled back or dropped. Finally, by 1986 a new methodology had been devised (involving thousands of Monte Carlo simulations for each system variable) to derive a forecast allowing for uncertainties. Although not determining which programs would be curtailed, this methodology would have permitted bringing the aggregate down to correspond to more probable overall force projections. A proposal by the Director of SOVA to insert a CIA DI footnote into NIE 11-3/8-86 explaining the matter was not accepted. The problem remained, and was soon compounded by uncertainties and divergent judgments on Soviet policy under Gorbachev. Meanwhile, the annual overestimates of future force enhancements continued.
Moscow's Economic Burden
There was also a growing concern in CIA's Directorate of Intelligence on the weight of the economic burden of defense even for existing programs, much less the forces projected for the future. But this was not reflected in the NIEs. In 1985, CIA analysts attempted, unsuccessfully, to get a caution on the Soviet defense burden on the economy into NIE 11-18-85. Also in 1985 an NIE on Soviet objectives in arms control, while correctly predicting that economic problems were not such as to prompt any significant early concessions in arms negotiations, also suggested that perhaps the Soviets were deliberately exaggerating their expressed concern over the economic impact of the arms competition in order to convince the West of their good faith in arms control. It did not concede a military burden on the economy to which Soviet leaders might respond. Moreover, it did not recognize, even as a possibility, that the new Soviet leadership might seriously be interested in ending the arms race, for political as well as economic reasons.
CIA's estimates of the economic burden of Soviet defense spending were controversial in the 1970s and 1980s, and the issue remains unresolved even today. The estimates that defense spending amounted to about 15 to 17 percent of GNP were almost certainly low, although the extent remains unclear. In the Soviet command economy, prices and costs were set arbitrarily; even Soviet leaders did not know the real extent of the burden. Accordingly, even the opening of Soviet records does not provide an answer. It does, however, seem clear that while Soviet leaders were not precluded from pursuing a wide range of military programs and were not compelled to seek negotiated arms limitations, they had strong incentives to seek to lower the defense burden, and this affected both their military programs and their foreign and arms control policies.
Meanwhile, the split in the Intelligence Community in the late 1970s following the Team B exercise and under DCI Turner contributed to a polarization of views in the NIEs on Soviet military doctrine and strategy. Basically, the issue was whether the Soviet leaders were seeking a decisive superiority and attainment of a war-fighting and war-winning capability, or were they pursuing deterrence bolstered by war-fighting capability while recognizing the unattainability of a war-winning strategy or capability. The issue could not be conclusively resolved by Soviet pronouncements or military programs, but the fact that the US military intelligence services supported the more dire alternative meant that the inclination to overestimate projections of future Soviet military capabilities, as well as of assertive current and future Soviet foreign policy actions, tended to dominate the NIEs. Again, the issue became even more salient with the advent of the Gorbachev administration in the Soviet Union.
From the beginning of Gorbachev's leadership in 1985 to his fall from power from August to December 1991, the highest priority tasks for US intelligence estimation were Gorbachev's political intentions, and his political capabilities, vis-à-vis the Soviet Union as well as toward the West. This was not immediately evident, but during the years 1985 to 1988 the fact that some intelligence analysts came to this conclusion was sufficient to make the issue of Gorbachev's political intentions at home and abroad of high, if controversial, importance. From 1988 to 1991, Gorbachev's intentions were more clear, and the issue increasingly became a question of whether Gorbachev could retain power and maintain his radical perestroyka of the Soviet Union and of Soviet policies and, if not, what would follow. Economic problems were recognized to be important in influencing Gorbachev's political agenda, and his failed efforts to find solutions to them (and indeed inadvertently to make them worse) were important.
Although military policy was an extremely important aspect of Gorbachev's reorientation of Soviet foreign policy, and eventually would be very important to his reformation of the Soviet system, it was not an early target for change. Military programs and procurement from 1985 through 1988 continued on the previously established basis of the twelfth five-year plan, which actually involved a modest increase in the rate of growth of military spending (from about 2 to about 3 percent). The NIEs were therefore on a sound foundation in estimating short-term military programs and capabilities. There were many signs that this would change, but the changes were not considered sufficiently clear or certain to modify the usual, somewhat overstated force estimates for later years. The errors of aggregate estimates of individually plausible projections discovered by CIA analysts in the early 1980s had not been corrected, but continued to be quietly adjusted downward each year as they proved too high, generally without modifying new future force projections.
The available declassified estimates of strategic forces in the latter 1980s through 1991 are spotty. They do make clear that as late as NIE 11-3/8-88 (the most recent substantially declassified strategic forces estimate), the projections for the future continued to be inflated. I would not fault the fact that the NIEs continued to estimate little-changed Soviet forces well into the 1990s, even though, in fact, by 1992 there was no Soviet Union (or Soviet armed forces). It was reasonable and prudent to estimate even as late as the summer of 1991 that a largely intact renewed Union, even if pared of some outlying republics, might remain. The kind of routine overestimation that I have in mind is illustrated by the familiar example of the Soviet strategic bomber force. NIE 11-3/8-88 in December 1988 was still estimating deployment of some 80 to 120 Blackjack (Tu-160) heavy bombers, and up to 150 heavy jet tankers, by the late 1990s. Although comparably precise estimates of the numbers of the newest ICBM variants being deployed (the SS-18 Mod 5 and SS-24 Mod 2) are not included in the summary sections of the estimate that have been declassified, these programs were described as involving "major changes in the force." Yet a CIA DI assessment a little over a year later stated that by 1988, the time of that estimate, the Soviets had already scaled back several intercontinental attack forces deployment programs, including the Blackjack, the SS-18 Mod 5, and the SS-24 Mod 2.
The key judgments section of NIE 11-3/8-91--the next after the 1988 NIE to be available in part in declassified form and the last of the series, issued in August 1991-- noted that the Soviets were reducing the estimated future number of Blackjack bombers, down from the 80-120 by the late 1990s previously estimated to some 40 by the year 2000. (There was no longer even mention of heavy tankers.) The estimate also noted that they had completed their SS-24 Mod 2 deployment, although the SS-18 Mod 5 continued to be deployed. The NIE included a general caveat, estimating that "barring a collapse of central authority or the economy," the Soviets would "retain and modernize powerful, survivable strategic forces throughout the next decade." While it may have been prudent to include an estimate of possible Soviet strategic forces and capabilities a decade hence by trend-line projections, would it not have been prudent in 1991 to estimate at least two alternative future force projections? The furthest the estimators went, noting a range of discordant views in Moscow, was to suggest that "the possibility remains, therefore, that a reformist regime [not the Gorbachev administration?] might challenge the need to maintain strategic nuclear forces comparable to those of the United States to ensure superpower status and might settle for a lower level of forces solely for deterrence." 
The greatest shortcoming of the NIEs on Soviet military power from 1987 to 1991 was the failure to give even cautious recognition to the radical changes already under way in Soviet outlook, doctrine, policy and strategy. NIE 11-18-87, evaluating Gorbachev's overall policies in November 1987, while belatedly acknowledging his reforms and including some astute observations on the internal scene, still stated that "On the national security front, Gorbachev adheres to traditional objectives: first and foremost enhancing the security of the homeland; expanding Soviet influence worldwide; and advancing Communism at the expense of capitalism around the globe." Gorbachev was said to seek to "ensure Moscow's ability to continue to fulfill military requirements," a formulation that begged the issue and obscured the key fact that Gorbachev had set new military objectives, doctrine, and requirements that would soon be evident. The NIE gave almost no hint of these developments under way and instead explicitly took as a "given" traditional Soviet "superpower ambitions, military power, and ideological predilections" that made it the adversary of the West "whether or not Gorbachev is successful." Instead, Gorbachev's "ultimate goal" in his new security policies was reduced to "slowing the overall momentum of US military programs." Gorbachev's new thinking on security and foreign and military policy was characterized as having "a substantial propaganda component. . . . Gorbachev wants to further the impression that his leadership is embarked on fundamental new directions." His main accomplishment in foreign policy was seen as "improving the Soviet image abroad," and he was said to have "better positioned the Soviet Union to put Western diplomacy on the defensive and to increase Soviet influence." Yet it was by then evident to many observers of the Soviet scene (including CIA analysts) that there was much more to Gorbachev's new thinking than improving an "image," furthering an "impression" of change, and maneuvering to put the West on the "defensive." The NIE even asserted that the most radical proponents of systemic reform under Gorbachev through the 1990s "would not end the competition, risk the relative gains they have made over the past 20 years, accept an inferior military position, or draw back from the pursuit of a global superpower status." Yet Gorbachev did all those things and more within the next three years.
Even a year later, NIE 11-3/8-88 in December 1988 flatly declared that, "to date. . .we have not detected changes under Gorbachev that clearly illustrate either new security concepts or new resource constraints are taking hold." Instead, "We believe it is prudent to adopt a wait-and-see attitude toward the prospects for longer term change in the Soviets' fundamental approach to war." Meanwhile, the NIE declared, "sufficiently compelling evidence is lacking to warrant a judgment in this Estimate that the Soviets already have begun to implement fundamental changes in their approach to warfare under Gorbachev." Moreover, "Thus far, we see no convincing evidence that the Soviets under Gorbachev are making basic changes in their approach to actually fighting nuclear war."
"Convincing" and "compelling" evidence was not long in coming. Less than a week after the issuance of NIE 11-3/8-88, Gorbachev announced in an address to the UN General Assembly in New York that the Soviet Union would unilaterally cut its forces by 500,000 men and--most importantly--withdraw six tank divisions and 5,000 tanks from Eastern Europe. The import of this reduction was even greater than the numbers; it meant unilaterally giving up the preponderant armored striking capability of the Warsaw Pact for any attack on the West, as subsequent estimates acknowledged.
Although none of the NIEs had even posed the possibility of a major Soviet unilateral reduction of its forces in Europe before Gorbachev's December 1988 announcement, CIA analysts were aware of reported studies begun in Moscow on such an action and were open to the possibility. A DI assessment in June 1988 had suggested the possibility of a unilateral Soviet cut, perhaps after testing the new administration in European arms negotiations in 1989, and had cautiously suggested that there was "a good chance that Gorbachev will, by the end of this decade, turn to unilateral defense cuts." 
It should not have required Gorbachev's announcement to alert the intelligence estimators to the changes under way in Soviet military doctrine and strategy and their implications not only for future Soviet capabilities but for already operative changed Soviet objectives and evaluations of the role of military power. In this respect, among others, the record shows that the analysts in CIA's Directorate of Intelligence were well ahead of the Intelligence Community as a whole, and their published analyses were well in advance of the NIEs in predicting (or at least catching up with) Soviet actions based on these doctrinal and policy changes.
The above-cited mid-1988 DI assessment of Soviet national security policy had reviewed in considerable detail the published debates by Soviet military and political figures and found that a close correlation of these statements with other (not declassified) sources led to the conclusion that "the Soviet leadership has come to believe that in an all-out nuclear war with the United States the USSR could not achieve victory in any meaningful way, even if it struck first." Moreover, calculations based on Soviet methodologies of damage assessment showed that US second-strike retaliatory capability had grown--not diminished--in the 1970s and had the potential to nearly double again by the end of the 1980s. "Under these circumstances, the prospects of achieving strategic nuclear superiority that could produce a meaningful victory in an all-out war may increasingly have seemed unrealistic."  This was a more forthright statement than the CIA-Intelligence Community position in the NIEs since the early 1980s and a direct repudiation of the Team B position taken by the military intelligence services since the late 1970s. Moreover, the CIA assessment argued that there was "persuasive evidence from both classified and open sources that the discourse goes beyond mere propaganda and involves fundamental issues that have potentially important ramifications for Soviet security policy and military forces over the longer term."  That was certainly true and had been for some time. It was reflected in a number of CIA DI analyses and assessments in 1987-89. But it was difficult to get meaningful discussions into the coordinated NIEs, and the result was usually a sterile standoff of "alternative views" or a brief statement in the text balanced by a dissent, neither going beyond general assertions. Only after the unilateral Soviet initiative of force reduction and retraction from Eastern Europe did the ice crack and begin to move. It also became easier to have discussion of such matters in NIC "Memoranda," less formal than NIEs but coordinated within the Intelligence Community by the appropriate NIO.
By spring 1990, a NIC Memorandum finally, and belatedly, acknowledged that "in 1986 and 1987 there was mounting evidence that the Soviets were reassessing their military doctrine,"  something only slowly acknowledged at that time in internal CIA analyses (and by nongovernment analysts on the basis of plentiful indications in open sources). It also noted that the reductions then well along in Eastern Europe were "consistent with a movement toward a defensive doctrine," and further that the Soviet proposal in the negotiations on Conventional Forces in Europe (CFE) "serves as one of the most convincing indicators to date of the defensive orientation of their military doctrine" as well as of "their intent to decrease the economic burden of the Soviet theater forces. . . ." Another NIC Memorandum in late 1989 had also noted that the unilateral reductions were not only "consistent with the announced Soviet shift toward a more defensive doctrine," but that "pursuant to the new doctrine's ‘War Prevention' tenet, the reductions will virtually eliminate the Soviets' already limited short warning attack capability." This was serious business and was related to the doctrinal revisions now belatedly acknowledged to have been under way since 1986.
The last relevant NIE statement preceding Gorbachev's unilateral initiative to reduce and withdraw forces from Eastern Europe had appeared only days before his announcement. It had considered the possibility of a unilateral action only to dismiss it beyond the possibility of some gestures for political effects in the West. It also stated with uncharacteristic certainty that in negotiations the Warsaw Pact states would not accept the NATO proposal that entailed massive cuts in tanks for minor NATO cuts, establishing parity between the members of the Pact and NATO. Yet that is precisely what the Soviet Union and other Warsaw Pact countries agreed to in the CFE Treaty within two years.
NIE 11-14-89, Trends and Developments in Warsaw Pact Theater Forces and Doctrine Through the 1990s, issued in February 1989, assessed the situation after Gorbachev's announced major reduction of Soviet forces in Eastern Europe; it would also be the last comprehensive estimate before the revolutionary changes in Eastern Europe nine months later, portending the disintegration of the Warsaw Pact. It did not foresee dramatic changes. It did, however, recognize that Gorbachev's policies were having a significant impact on Soviet general purpose forces (albeit with a DIA dissent arguing that further large reductions were unlikely and that "present evidence and future uncertainties make the element of continuity in Soviet military policy as important as the changes for US national security and defense planning"). 
The NIE stated that Gorbachev's military doctrinal revisions and other initiatives (such as the unilateral reduction) were "designed primarily to help the Soviet leadership reinvigorate the economy by shifting resources from defense to the civilian sector . . . ." Although previously NIEs had only reluctantly accepted an economic constraint on defense expenditure, this explanation was more readily accepted than one also attributing to Gorbachev's policy a new world-view and new political objectives, including the dismantling of the East-West military confrontation in Europe. The director of DIA did dissent from the judgment that the concept of "reasonable sufficiency" was a matter of avoiding the costs of an unabated arms race. But he did so only to argue that the concept was advanced primarily "to raise Western expectations regarding the prospects for substantial force reductions and to undermine support for NATO modernization" and that, if Gorbachev failed in these objectives by the mid-1990s, "the Soviets most likely would, despite the extremely heavy costs, revert to their traditional resource-intensive approach to develop the next generation of weapons and modernize their forces." 
NIE 11-14-89 nonetheless marked an important, if incomplete, change in evaluating Soviet military doctrine. CIA DI publications had, as earlier noted, been treating seriously debates and changes in doctrine both in open and classified Soviet sources. The NIEs had generally shied away from serious evaluation of changes taking place even in official doctrine. Although NIE 11-14 was not the most logical in the NIE series to deal with this subject, it was certainly affected, and it appeared at a time when some judgments seemed necessary. The result was an abandonment of the Team B arguments that the Soviets believed in attaining victory in a nuclear war. Indeed, even the record of past acceptance of this view in NIEs was brushed aside:
There is no indication that the Soviets have ever been sanguine about the consequences they would expect to suffer in a nuclear war. Moreover, evidence from the 1980s indicates the Soviets doubt they could prevail in any traditionally meaningful military-political sense because of the expected high levels of damage both sides would sustain from nuclear attacks. Since the early 1980s, Soviet leaders have explicitly renounced the possibility of achieving victory in a general nuclear conflict. We judge that the "no victory in nuclear war" position--publicly endorsed by Gorbachev and incorporated in the 1986 27th Party Congress program--is basic to the political dimension of Soviet military doctrine.
And it was further emphasized that the "military-technical" component of military doctrine was strictly subordinate to the political, and that military doctrine had "the status of state policy."
But what the NIE yielded to the political dimension of military doctrine with respect to Gorbachev's new thinking, it offset by saying that "we have not detected any changes in the military-technical dimension of Soviet military doctrine which clearly demonstrate that the Soviets have changed their nuclear war-fighting doctrine under Gorbachev." Specifically, "The Soviets will remain committed to the offensive as the preferred form of operations in wartime." 
DI analyses in CIA had discussed the presentation in May 1987 of a new Warsaw Pact military doctrine calling for a defensive doctrine at the military-technical level. Previously Soviet military doctrine at the political level was avowedly defensive but, once war had begun, Soviet military doctrine at the military-technical level called for assuming the offensive as soon as possible and pursuing it vigorously. The new doctrine called for a radically new approach, and it was hotly debated. While it was appropriate to weigh carefully its significance, the very fact of a vigorous debate in Moscow suggested that a real and significant change was involved. But despite other demonstrated changes under Gorbachev, the NIE stated flatly in an estimate covering the next decade that "the Soviets will remain committed to the offensive." Readers of the NIE would not know that, for example, the confidential Soviet General Staff journal Military Thought in January 1988 in its lead editorial on the new training year had emphasized that the new doctrine was meant as operational guidance, and that military exercises in 1988 and after gave much greater attention to defense, as did articles in the journal.  During the years 1989 through 1991, the Intelligence Community sought to catch up with the rapid rush of events. These events were predominantly political, although of course with important military implications and consequences. Soviet military spending, which had remained steady at about 3-percent annual rate of growth from 1985 through 1988, fell with a 15-percent decline in 1989, another 15 percent in 1990 and still more in 1991. As earlier noted, the estimates of strategic forces and capabilities for these years, except for one brief summary, have not been declassified. The declassified materials indicate only one brief NIE on Soviet theater forces after NIE 11-14-89, but there were also several NIC Memoranda in 1989 and 1990. None of these repeated an oddly anachronistic statement in NIE 11-14-89 that, although "we consider Pact initiation of hostilities without mobilization [the new situation after the unilateral reductions] to be extremely unlikely . . . we cannot, however, rule out the possibility that the Pact might initiate hostilities from a condition of partial mobilization if it perceives an opportunity to achieve decisive results against NATO, or a need to forestall NATO from achieving decisive results against the Pact." Several years into the Gorbachev administration, and with major changes under way in Eastern Europe, to draw explicit attention in 1989 to a possible Warsaw Pact decision to launch a war because of a perceived opportunity to defeat NATO was probably one of the most bizarre estimates in the history of the Cold War.
A constructively imaginative assessment of the military implications of the possible breakup of the USSR was prepared in CIA's DI in September 1991, as that possibility loomed ever larger. It represented the estimated thinking and views of the Soviet General Staff. The basic alternatives included one then being prepared by Gorbachev (losing only peripheral Baltic, Moldavian, and Caucasian members), one limited to Russia-Belarus-Ukraine and Slavic northern Kazakhstan, and one limited to Russia alone. The latter was, of course, soon the new reality. The estimators (as, no doubt, the Soviet General Staff they were seeking to emulate) assumed that even in that third alternative it would be possible to transfer many military forces if not fixed assets into Russia, so even the postulated "worst case" was not quite as bad from Moscow's standpoint as reality soon became. The analysis was thorough and professional. The one element not given due attention (in the redacted, declassified version, at least, no attention at all) was the dispersed location of tactical nuclear weapons, although by that time most had been returned from outlying republics (other than Ukraine and Belorussia/Belarus). 
The most appropriate concluding questions to pose are probably the most basic ones: What was the role of the CIA and its analyses in estimating Soviet military power, intentions, and capabilities? How good were the National Intelligence Estimates of Soviet military power--how accurate were the information and assessments, how complete, and how useful to policymakers? A very general overall judgment might be that the estimates often could have been better but also could have been much worse. Others making "estimates," in and out of the US government, were sometimes more on target; but many, many such "competing estimates" were terribly wrong. Within the US Intelligence Community, the institution of NIEs, and the reputation for their authority (if not always for their messages), played a positive role in keeping the various intelligence (and non-intelligence) elements in the US government within the parameters set by the NIEs.
Overall, the record of NIE assessments of Soviet intentions shows a tendency in the late 1940s and 1950s, and again in the last half of the 1970s and first half of the 1980s, to overstate a Soviet propensity to rely on military power and its offensive applications. The assessments of Soviet military forces and capabilities also displayed a persistent tendency to overestimation (with the notable exception of the systematic underestimate of future ICBM deployment in the years 1963-68). Thus the Soviet threat was generally overstated throughout the Cold War. It is, however, not easy to evaluate the impact on US policy and defense programs. Moreover, from the early 1960s on, the estimates of current Soviet forces were generally accurate; overestimates concerned projections of estimated future forces, and these errors were corrected with respect to current forces (even if projected future forces continued to be overstated). Least successful were estimates of Soviet doctrine, strategy, and policy objectives in the latter half of the 1970s and early 1980s, and especially in the very slow response to changes in the Gorbachev period from 1986 through 1991. CIA shared in these misestimates but was almost always closer to a correct understanding than the military members of the Intelligence Community, and CIA DI assessments were more correct and more informative than coordinated NIEs.
The DCI, after 1947 also the head of the CIA, never acquired control over the Intelligence Community. But he did have control over the estimating process and the content of NIEs. He could not preclude the expression of diverging and dissenting views, and on occasion was alone against an array of other members of the Intelligence Community. But to the extent he wished to do so, he could determine the official national estimate.
DCIs in most cases relied principally on the analysis of the CIA in reaching their positions, although that depended on more convincing evidence or arguments than those offered by other agencies. Occasionally, a DCI would disagree with CIA analysts and estimators (for example, DCI John McCone over Soviet intentions to deploy medium range missiles in Cuba in September 1962 or DCI William Casey over Soviet responsibility for international terrorism in 1981) or pursue a given line further than they would (as DCI Stansfield Turner in NIE 11-3/8-79 and 11-3/8-80). But by and large, while the NIEs were a coordinated and agreed-on product of the Intelligence Community, or at least of the DCI, from the 1950s through the 1970s, they were drafted in CIA and rested mainly on CIA analysis.
The role of CIA analysis in NIEs on aspects of Soviet military intentions and capabilities grew considerably in the 1950s and became dominant in the 1960s and 1970s owing to the development of greater expertise in CIA in areas such as economic analysis, military production analysis, scientific (including military, space, and nuclear) intelligence, and effective melding of all-source intelligence including overhead imagery, communi-cations intelligence, clandestine human-source intelligence, diplomatic and other reporting, and analysis of broadcast and published open sources. From the mid-1970s, with the institution of NIOs drawn from various quarters, the centrality of CIA military analysis was diluted.
How much were the Soviet military estimates used by policymakers (and their staffs)? It is fair to say that no one waited for an NIE 11-4 (or later an NIE 11-3/8) to see if war was coming or if the United States should or should not negotiate on arms. But the military estimates (and derivatives such as the National Intelligence Planning Projections) were used very seriously in the context of US defense policy, even though US military programs and forces were generally not tailored to match or counter Soviet military forces. It was for that reason that the military services' intelligence chiefs were so keenly engaged on matters affecting their own service programs.
CIA analysis was not always right, nor always accepted as the basis for judgments in NIEs. But it came to have a predominant role in them because it was more correct more often (apart from the fact that it was also ultimately the DCI's choice). On occasion, CIA analysis was not accepted by the DCI because of his own divergent views. CIA DI analyses in the Gorbachev period were often not reflected in NIEs (or disseminated otherwise to senior policymakers) by DDCI Gates because he held a different view. That was his right. But it was regrettable because the CIA analysis was far more correct than the view he held.
CIA analysis, and the estimates process of the Intelligence Community, grew greatly over the four-and-a-half decades of the Cold War. It was a long and tortuous passage and by no means always in a straight line of evolution and refinement. In fact, it was a twisted path. Nonetheless, there was not only improvement over time in deriving intelligence from a greatly expanding combination of rich sources of information, but also in building estimative methodologies and approaches. CIA analysis not only became the central element generating national intelligence, but also provided a stabilizing function in making intelligence more accurate and reliable, more timely, and more useful to policymakers.
The particular intelligence requirements of the Cold War have now passed into history. Lessons to be learned from a critical evaluation of the performance of CIA (and the other members of the Intelligence Community, and the Community itself) remain, nonetheless, important not only for history but also for considering how best to meet the intelligence requirements of a new era.
A panel moderated by Aaron Friedberg, Professor of Politics and International Affairs at Princeton University, discussed Raymond Garthoff's paper and provided views on CIA's military analysis of Soviet military intentions and capabilities during the Cold War. The panelists were Richard Betts, the Leo A. Shifrin Professor of War and Peace at Columbia University; Douglas MacEachin, former Deputy Director for Intelligence at CIA; and Steven Rosen, the Beton Michael Kaneb Professor of National Security and Military Affairs at Harvard University.
Commentator Richard Betts agreed with Garthoff that, given the limitations of the process, the record of US analysis of Soviet military affairs was generally good, especially the day-to-day analysis commonly referred to as current intelligence. The problem, according to Betts, was in the conceptualization of the terms intentions, capabilities, and politicization. Regarding intentions, Betts argued that it was necessary to distinguish between policy intentions and strategic intentions. Betts defined policy intentions as the USSR's motives and objectives with regard to its offensive and defensive weapons. He defined strategic intentions as how the Soviets would use force to win a war. Betts asserted that the Intelligence Community could not address Moscow's policy intentions issues because they were wrapped up in theology, ideology, and politics. He suggested that the Agency's analysts should have assumed that the inherent conflict of interest between East and West made war possible, even if neither side considered its own arms policies overly aggressive. Rather, they should have focused on discerning the Kremlin's strategic intent and assessing Soviet military capabilities. Betts opined that estimates of military capabilities are more useful because forces are more observable than intentions, although it is hard to quantify levels of training and the fighting ability of conventional forces.
Betts then discussed politicization, which he defined as the shaping of analytical conclusions to conform to policy predispositions. He defined three types of politicization. The first type is the deliberate corruption of analysis done by changing conclusions in order to conform to policymaker preferences. Betts concluded that this type of politicization of the Agency's analysis during the Cold War was rare. The second type is the slanting of analysis because one's biases and assumptions are based on a particular political point of view. Betts said this form of politicization usually was done unconsciously based on one's assumptions of how the world works. He maintained that we all have biases, based on our experiences, study, and ideology; anyone without some bias "hasn't learned anything about how the world works." Finally, Betts discussed the organized confirmation of biases in the analytical process, which he defined as making conscious and explicitly competitive the unconscious biases that organizations or individuals bring to a problem. According to Betts, this is what the Team B exercise should have been all about. It troubled him that the Team B exercise was terminated without a full-scale confrontation of diverging views.
Betts also introduced the idea that the most useful function of the NIE process was to force attention to the differences between what we know from direct observation, which is generally undisputed, and what we deduce from evidence, which may be questionable, as well as what we assume or guess on the basis of logic and our biases.
Finally, Betts said that the most difficult debates about nuclear strategy during the Cold War were the ones involving issues of nuclear deterrence; NATO's reliance on the threat of escalation and first use; measurements of parity and superiority; and the assumption of NATO inferiority in conventional forces. For Betts, these issues had major implications for US and Allied defense and domestic policies. In his view, they were never resolved satisfactorily.
Douglas MacEachin began his comments by supporting Garthoff's and Bett's contention that the unique political climate in which National Intelligence Estimates on Soviet military forces were produced had an enormous impact on the estimates. MacEachin argued that the biggest single trap in producing NIEs was their attempt to assess Soviet intentions, which were impossible to know. According to MacEachin, nothing is more subject to social-political atmospherics than one's perception of intentions. In his view, the United States needed to reexamine how it did NIEs in order to avoid the "intentions trap." He went on to say that Soviet policies during the Cold War also were driven by internal politics in the Soviet Union, which the Intelligence Community had no way of foreseeing. Attempts to do so, in MacEachin's view, would lead to "intelligence failures."
Overall, MacEachin concluded that CIA did provide good military intelligence to policymakers during the Cold War. However, he was skeptical that it had much overall impact or influence on US policy.
Steven Rosen took exception to Raymond Garthoff's harsh criticism of the Team B experiment. Rosen asserts that the Team B exercise was a positive step toward improving the overall analytical product. He maintain that because individuals are reluctant to give up their views of the world and their predisposed mindsets, it is beneficial to the analytical process to force people to confront divergent views or to force them to deal with ideas they don't like. While agreeing with Garthoff that Team B reached some wrong conclusions in certain of its projections and estimates, Rosen thought the exercise raised the level of the debate by forcing intelligence analysts to realize that the Soviet leadership had different views about nuclear war and nuclear weapons than we did. According to Rosen, the Soviet leadership believed that more was better and that nuclear superiority in offensive and defensive capabilities was "worth spending a lot of money."
Rosen agreed with Garthoff that part of the problem was the tendency by CIA analysts and others to assume that Soviet views on military affairs were the same as those that dominated American thinking at the time. Rosen cited, for example, the Intelligence Community's underestimation of the production of Soviet medium range bombers and missiles and its overestimation of its intercontinental range forces. Rosen argued that the erroneous estimates were caused by US analysts assuming that the Soviets were most interested in striking the continental United States, an assumption based entirely on Western thinking. In Rosen's view, the Soviets way of looking at the world differed markedly from that of US policymakers.
Rosen concluded by saying that it was best to let a new generation of scholars examine the legacy of Cold War intelligence because most of the experts currently engaged in the debate were personally involved in assessing Soviet military power during the Cold War. He suggested that the Soviet biological warfare programs and the growth of the Soviet naval forces needed further study. He also said that Moscow's preoccupation with China from the early 1960s on needed to be assessed more carefully because much of the Soviet military buildup was directed at China rather than at the United States. Rosen also questioned why communications intelligence and SIGINT were not discussed at the conference, given their central role in the internal US debate on Soviet military doctrine.
Raymond L. Garthoff
, a guest scholar at the Brookings Institution in Washington, DC, is a prolific author on Soviet affairs and former US Ambassador to Bulgaria.
 Intelligence Memorandum No. 323-SRC, Soviet Preparations for Major Hostilities in 1950 (25 August 1950), p. 1. The memorandum (pp. 1, 2) noted "only one field" in which the Soviet Union was "relatively unprepared" for a major war in 1950: "The USSR will not be capable of a large scale atomic bombardment in 1950; however, the Soviets are capable of employing against the continental US the 25 bombs estimated to be currently available." We now know there was no Soviet stockpile of ready weapons in 1950.
 NIE-3, Soviet Capabilities and Intentions (15 November 1950), pp. 1-2.
 NIE-15, Probable Soviet Moves to Exploit the Present Situation (11 December 1950), p. 1.
 NIE-48, Likelihood of the Deliberate Initiation of Full-Scale War by the USSR Against the US and its Western Allies Prior to the End of 1952 (8 January 1952), p. 1. Earlier, in 1951, the Director of Naval Intelligence had formally dissented from one guarded, equivocal estimate going so far as to say boldly that he believed it was "unlikely that the USSR will deliberately choose to precipitate or undergo the hazards of general war" in 1952. See NIE-25, Probable Soviet Courses of Action to Mid-1952 (2 August 1951), p. 5.
 Ibid. Italics added.
 A valuable supplement to the NIEs of the 1950s is a critical internal ONE retrospective analysis prepared by Abbott Smith, Deputy Chairman of the Board of National Estimates in those years: A Study of National Intelligence Estimates on the USSR, 1950-1957, undated . Although originally classified Top Secret, the document was declassified in 1992 but remains unpublished.
 The estimate used in NSC 68, although not cited by designation number, was ORE 91-49, Estimate of the Effects of the Soviet Possession of the Atomic Bomb Upon the Security of the United States and Upon the Probabilities of Direct Soviet Military Action (6 April 1950); the repudiation of the estimate discounting possible Soviet attack was ORE 32-50, The Effect of the Soviet Possession of Atomic Bombs on the Security of the United States (9 June 1950). (The agreed estimate on numbers of Soviet atomic bombs, cited in NSC 68, was given in ORE 91-49, p. 11 and ORE 32-50, p. 2. This estimate of the Soviet atomic stockpile was, incidentally, soon revised upward to 245 bombs by mid-1954, in NIE-3, Soviet Capabilities and Intentions [15 November 1950], p. 4.)
 Each was designated NIE 11-4 and the year issued; NIE 11-4-60 was the last in the series. Beginning with NIE 11-4-61 a new series with the 11-4 designator was devoted to aspects of Soviet military policy, as described later.
 These figures are all taken from the NIE 11-4 and 11-8 series from 1955 through 1958.
 See NIE 11-5-54, Soviet Capabilities and Probable Programs in the Guided Missile Field (5 October 1954); NIE 11-12-55, Soviet Guided Missile Capabilities and Probable Programs (20 December 1955); and NIE 11-5-57, Soviet Capabilities and Probable Programs in the Guided Missile Field (12 March 1957), p. 3.
 See NIE 11-5-58, Soviet Capabilities in Guided Missiles and Space Vehicles (19 August 1958), p. 3.
 See SNIE 11-10-57, The Soviet ICBM Program, and the NIE 11-4 and 11-8 series, 1958 through 1962. There is an excellent detailed review in the now declassified internal history of CIA in the Dulles years, written by a former Board of National Estimates member; see Wayne G. Jackson, Allen Welsh Dulles as Director of Central Intelligence, 26 February 1953 - 29 November 1961, CIA History Staff, HRP 91-2/1 Volume V, Intelligence Support to Policy, July 1973, pp. 39-138, at the National Archives and Records Administration, NN3-263-94-011. Here and throughout, I have followed the usual practice of referring to estimates of the numbers of ICBMs, or ICBMs deployed; strictly speaking, the estimates all referred to numbers of operationally deployed ICBM launchers. The numbers of ICBMs in the pipeline from production, or held in reserve, are not counted in the estimates.
 See SNIE 11-6-60, Strength of the Armed Forces of the USSR (3 May 1960); and Raymond L. Garthoff, "Estimating Soviet Military Force Levels: Some Light from the Past," International Security, Vol. 14, No. 4 (1990), pp. 93-116.
 See NIE 11-4-58, Main Trends in Soviet Capabilities and Policies, 1958-1963 (23 December 1958), p. 28, and see pp. 55-56.
 Ibid., p. 28.
 See NIE 11-4-59, Main Trends in Soviet Capabilities and Policies, 1959-1964 (9 February 1960), p. 26 (italics added) and see pp. 4, 25-26, and 39; and see also NIE 11-8-59, Soviet Capabilities for Strategic Attack Through Mid-1964 (9 February 1960), pp. 3-6.
 NIE 11-4-59, footnotes on pp. 26 and 27.
 Ibid., p. 39. Italics added.
 Ibid., footnote on p. 39.
 NIE 11-8-59, Soviet Capabilities for Strategic Attack Through Mid-1964 (9 February 1960), pp. 2 and 5.
 Ibid., footnote on p. 3.
 NIE 11-4-60, Main Trends in Soviet Capabilities and Policies, 1960-1965 (1 December 1960), pp. 3, 4, 22, 23, 25, and 36.
 Ibid., footnote on p. 3.
 See NIE 11-5-54 (7 June 1954); 11-7-54 (17 August 1954); 11-4-54 (14 September 1954), the only one to estimate larger numbers than 800; 11-3-55 (17 May 1955); 11-7-55 (23 June 1955); 11-56 (6 March 1956); 11-4-56 (2 August 1956); SNIE 11-6-57 (15 January 1957); 11-4-57 (12 November 1957); and 11-4-58 (23 December 1958). For the corrected figures, see 11-8-59 (9 February 1960); 11-4-59 (9 February 1960); and 11-4-60 (1 December 1960).
 See the NIE 11-4 and 11-8 series from 1959 through 1963.
 See NIE 11-8-63, Soviet Capabilities for Strategic Attack (18 October 1963), p. 17.
 NIE 11-8-70, Soviet Forces for Intercontinental Attack (24 November 1970), pp. 17-18, and NIE 11-8-71 Soviet Forces for Intercontinental Attack (21 October 1971), p. 15, registered split views on the first 120 such SS-11s. All agencies agreed that the missiles had both intercontinental and Eurasian theater capabilities but disagreed as to which were "primary." (CIA and DIA firmly believed they were primarily peripheral; others varied from regarding them as intercontinental to saying that their employment would be either.) By NIE 11-8-73 Soviet Forces for Intercontinental Attack (25 January 1974), pp. 3 and 10, the missiles were carried as ICBMs. The text, however, noted that these missiles (by this time, the final number of launchers was 310) had capabilities for either intercontinental or peripheral theater missions and were oriented for the latter. The total number of SS-11 missile launchers, including these 310, was 1,030.
 SNIE 11-14-61, The Soviet Strategic Military Posture, 1961-1967 (21 November 1961), p. 1.
 Ibid., p. 2.
 Ibid., p. 3.
 Ibid., p. 2.
 Ibid., p. 7.
 NIE 11-8-63, Soviet Capabilities for Strategic Attack (18 October 1963), pp. 13 and 32-33.
 NIE 11-8-64, Soviet Capabilities for Strategic Attack (8 October 1964), pp. 4 and 16-18.
 NIE 11-8-65, Soviet Capabilities for Strategic Attack (7 October 1965), pp. 11-13.
 NIE 11-8-66, Soviet Capabilities for Strategic Attack (20 October 1966), pp. 2 and 16-20.
 NIE 11-8-67, Soviet Capabilities for Strategic Attack (26 October 1967), pp. 3 and 12-13.
 NIE 11-8-68, Soviet Strategic Attack Forces (3 October 1968), pp. 2, 8, and 12-13; and NIE 11-8-69, Soviet Strategic Attack Forces (9 September 1969), pp. 4 and 9-11.
 NIE 11-8-70, Soviet Forces for Intercontinental Attack (24 November 1970), pp. 2 and 17-18.
 For the most complete and up-to-date analysis of the performance of US intelligence in the crisis, see Raymond L. Garthoff, "US Intelligence," in James D. Blight and David A. Welch, eds., Intelligence and the Cuban Missile Crisis ( Portland OR: Frank Cass, 1998), pp. 18-63; and the same essay in Intelligence and National Security, Vol. 13, No. 3 (1998), pp. 18-63. This essay cites and uses declassified CIA DI studies of 1962-64 dealing with the missile crisis.
 SNIE 11-6-58, The Soviet Attitude Toward Disarmament (24 June 1958), pp. 2 and 5.
 Ibid., footnotes of dissent on pp. 2 and 5.
 See NIE 11-8-64, Soviet Capabilities for Strategic Attack (8 October 1964), p. 2; and NIE 11-8-66, Soviet Capabilities for Strategic Attack (20 October 1966), footnote on p. 1.
 See NIE 11-8-69, Soviet Strategic Attack Forces (9 September 1969), p. 8.
 See NIE 11-8-70, Soviet Strategic Attack Forces (24 November 1970), p. 5; and NIE 11-8-71, Soviet Forces for Intercontinental Attack (21 October 1971) p. 6.
 NIE 11-8-69, Soviet Strategic Attack Forces (9 September 1969), p. 8; NIE 11-8-70, Soviet Forces for Intercontinental Attack (24 November 1970), pp. 5-6; NIE 11-8-71, Soviet Forces for Intercontinental Attack (21 October 1971), p. 6; and NIE 11-8-72, Soviet Forces for Intercontinental Attack (26 October 1972), p. 7.
 See NIE 11-8-73, Soviet Forces for Intercontinental Attack (25 January 1974), p. 4; and SNIE 11-4-73, Soviet Strategic Arms Programs and Détente: What Are They Up To? (10 September 1973), pp. 8 and 10.
 NIE 11-3/8-74, Soviet Forces for Intercontinental Conflict Through 1985 (14 November 1974), Vol. 1, pp. 10-11. Italics in original.
 NIE 11-3/8-76, Soviet Forces for Intercontinental Conflict Through the Mid-1980s (21 December 1976), pp. 2-6. The Air Force Intelligence diatribe was the swan song of Major General George J. Keegan, Jr., who soon retired and made a similar alarmist statement publicly.
 NIE 11-3/8-79, Soviet Capabilities for Strategic Nuclear Conflict Through the 1980s (17 March 1980), Vol. 1, pp. 3-4.
 NIE 11-3/8-76, Soviet Forces for Intercontinental Conflict Through the Mid-1980s (21 December 1976), pp. 3-6.
 NIE 11-3/8-80, Soviet Capabilities for Strategic Nuclear Conflict Through 1990 (10 December 1980), Vol. 1, pp. B3-B4.
 NIE 11-4-77, Soviet Strategic Objectives (12 January 1977), pp. i, and 1-4.
 See Intelligence Community Experiment in Competitive Analysis: Soviet Strategic Objectives, An Alternative View, Report of Team "B" (Washington DC: CIA, December 1976), p. 55. There were two other more narrowly focused technical Team B reports on missile accuracy and air defenses, which need not be considered here. For an excellent in-depth review of the Team B episode and the context of strategic alarms in the 1970s see Anne Cahn, Killing Détente: The Right Attacks the CIA (University Park, PA: Pennsylvania State University Press, 1998). Team B suggested that the NIEs downplayed the military threat because of "bureaucratic rivalry" between the CIA and the military services, with "the civilian analysts" controlling the NIE language and tempering the more pessimistic views of the military intelligence analysts. Report of Team "B," p. 4. To the extent that this was true, it was a credit to CIA.
 Report of Team "B," quotations from pp. 14 and 45-47.
 See Raymond L. Garthoff, Perspectives on the Strategic Balance (Washington DC: Brookings Institution, 1983), p. 8 for the figures above excepting those for 1990-91, which are given in Arms Control Today (December 1994), p. 29, and (November 1995), p. 30. The NIC quotation on US faltering and unilateral restraint is from NIC M83, The Strategic Weapons Spiral: Soviet Reactions (August 1983), p. iii.
 See Foreign and Military Intelligence, Final Report of the Senate Select Committee to Study Governmental Operations with Respect to Intelligence Activities, Ninety-fourth Congress, Second session (Washington DC: US Government Printing Office, 1976), Book 1, pp. 77-79; and John Ranelagh, The Agency: The Rise and Decline of the CIA, rev. ed. (New York, NY: Simon and Schuster Touchstone, 1987), pp. 494-499.
 Most of the relevant NIEs, especially the 11-3/8 series, for the 1970s and 1980s have been declassified at least in part. In addition, a particularly valuable internal CIA critical analysis of these estimates prepared in 1989 provides most of the information in this and succeeding paragraphs and in greater detail. Now declassified, this analysis by CIA's Directorate of Intelligence is titled Intelligence Forecasts of Soviet Intercontinental Attack Forces: An Evaluation of the Record, a Research Paper, (April 1989), including a lengthy supplement (cited hereafter as Intelligence Forecasts).
 The three paragraphs above draw heavily on Intelligence Forecasts, in particular the Supplement, pp. 123-127. See also Lt. Gen. Brent Scowcroft et al. Report of the President's Commission on Strategic Forces (Washington, DC, April 1983).
 Intelligence Forecasts, p. 10.
 See Ibid., p. iv.
 NIE 11-3/8-76, Soviet Forces for Intercontinental Conflict Through the Mid-1980s (21 December 1976), pp. 2 and 12.
 Ibid., pp. 5, 12, and 13. In NIE 11-4-77, Soviet Strategic Objectives (12 January 1977), p. 11, the State Department again dissented from a main-text statement that Soviet civil defense measures could "have a significant impact" on "perceptions of the likely outcome of a nuclear exchange."
 NIE 11-3/8-79, Soviet Capabilities for Strategic Nuclear Conflict Through the 1980s (17 March 1980), p. 23.
 See Soviet Defense Spending: Recent Trends and Future Prospects, an Intelligence Assessment, Directorate of Intelligence, CIA, July 1983; and see Raymond L. Garthoff, The Great Transition: American-Soviet Relations and the End of the Cold War (Washington DC: Brookings Institution, 1994), p. 41; and Garthoff, "The ‘Spending Gap,'" Bulletin of the Atomic Scientists, Vol. 40 (May 1984), pp. 5-6.
 The Development of Soviet Military Power: Trends Since 1965 and Prospects for the 1980s, an Intelligence Assessment, OSR, (April 1981), pp. 145.
 NIE 11/3/8-79, Soviet Capabilities for Strategic Nuclear Conflict Through the 1980s (17 March 1980), Vol. 1, pp. 14-16.
 NIE 11-3/8-80, Soviet Capabilities for Strategic Nuclear Conflict Through 1990 (16 December 1980), p. B-13.
 Ibid., pp. B17-B18.
 See Ben B. Fischer, A Cold War Conundrum: The 1983 Soviet War Scare (Center for the Study of Intelligence, CIA, 1997).
 SNIE 11-4/2-81, Soviet Potential To Respond to US Strategic Force Improvements, and Foreign Reactions (6 October 1981), see pp. 2 and 17.
 NIE 11-4-78, Soviet Goals and Expectations in the Global Power Arena (9 May 1978), p. 1.
 NIE 11-16-83X, The Soviet Approach to Arms Control: Implications for START and INF (8 March 1983), p. ii. Italics added.
 NIC M83, The Strategic Weapons Spiral: Soviet Reactions (August 1983), p. iii.
 Memorandum for Deputy Director for Intelligence [Richard Kerr], From Douglas J. MacEachin, Director for Soviet Analysis, Subject: NIE 11-3/8, Force Projections (22 April 1986), 7 pp. For the new methodology, see Projecting Soviet Military Forces and Weapons Procurement, A Technical Intelligence Report, Directorate of Intelligence, CIA, SOV 87-10006, (November 1987).
 See Kirsten Lundberg, CIA and the Fall of the Soviet Empire: The Politics of Getting It Right, Case Study C 16-94-1251.0 for the Intelligence and Policy Project, John F. Kennedy School of Government, (Cambridge MA: Harvard University Press, 1994), p. 11. NIE 11-8-85, Domestic Stresses on the Soviet System (November 1985).
 SNIE 11-16-85/L, Soviet Strategic and Political Objectives in Arms Control in 1985 (March 1985), pp. 4 and 8.
 NIE 11-3/8-88, Soviet Forces and Capabilities for Strategic Nuclear Conflict Through the Late 1990s (1 December 1988), p. 10.
 Ibid., p. 7.
 Gorbachev, Perestroyka, and Future Soviet Strategic Offensive Forces, an Intelligence Assessment, Directorate of Intelligence, CIA, SOV 90-10009JX, February 1990, p. iii.
 NIE 11-3/8-91, Soviet Forces and Capabilities for Strategic Nuclear Conflict Through the Year 2000 (8 August 1991), pp. 1, 3, 6, and 8. There was no 11-3/8-89, and 11-3/8-90 has not been declassified.
 NIE 11-18-87, Whither Gorbachev: Soviet Policy and Politics in the 1990s (November 1987), pp. 3, 7, 21-25, and 35. See also a contemporaneous memorandum by DDCI Robert M. Gates, Gorbachev's Gameplan: The Longer View (24 November 1987), p. 4, a cautionary, self-styled "reminder of the enduring [traditional] features of the regime and the still long competition and struggle ahead".
 NIE 11-3/8-88, Soviet Forces and Capabilities for Strategic Nuclear Conflict Through the Late 1990s (1 December 1988), quotations from pp. 6, 3, 7, and 6, respectively.
 Soviet National Security Policy: Responses to the Changing Military and Economic Environment, SOV 88-1004CX, an Intelligence Assessment, Directorate of Intelligence, CIA, (June 1988), pp. 22 and viii.
 Ibid., pp. 8-10.
 Ibid., pp. v and viii.
 For example, see Gorbachev and the Military: Managing National Security Policy, SOV 87-10061X, (October 1987); The Clash Between Civilian Specialists and the Soviet Military Over "Reasonable Sufficiency," FB 87-10021, (4 December 1987); Soviet Defense Policy: "Sufficiency" Doctrine Gains in Debate, FB 88-10020, (15 November 1988); The Changing Role of Civilian Advisers in Shaping Soviet National Security Policy, an Intelligence Assessment, SOV 89-10004X, (January 1989); The Nature of Soviet Military Doctrine, a Research Paper, SOV 89-10037X, (April 1989); and Gorbachev's Strategy for Managing the Defense Burden, an Intelligence Assessment, SOV 89-10036, (April 1989).
 NIC M90-10002, The Direction of Change in the Warsaw Pact (April 1990), p. 2.
 Ibid., pp. iii and 12.
 NIC M 89-10005, Soviet Theater Forces in 1991: The Impact of the Unilateral Withdrawals on Structure and Capabilities (November 1989), p. 2.
 SNIE 11-16-88 CX, Soviet Policy During the Next Phase of Arms Control in Europe (November 1988), pp. iii-iv.
 NIE 11-14-89, Trends and Developments in Warsaw Pact Theater Forces and Doctrine Through the 1990s (February 1989), pp. iii and ix.
 Ibid., pp. 1-3 (with the dissenting "alternative judgment" of DIA in the text on p. 3).
 Ibid., p. 6.
 Ibid., pp. 6 and v.
 By the summer of 1990, these references were all made public; see Raymond L. Garthoff, Deterrence and the Revolution in Soviet Military Doctrine (Washington DC: Brookings Institution, 1990), pp. 156-174.
 NIE 11-14-89, Trends and Developments in Warsaw Pact Theater Forces and Doctrine Through the 1990s (February 1989), p. vii. Italics added.
 The Implications of a Breakup of the USSR: Defense Assets at Risk, an Intelligence Assessment, Directorate of Intelligence, CIA, SOV91-10039X, (September 1991).