CIA HISTORICAL REVIEW PROGRAM RELEASE AS SANITIZED
I
INTELLIGENCE ASPECTS Of THEONTENTS
Mature of the
The Fstiimates and the Chanpos in
Estimates of Soviet ICBH
Estimates of Soviet ICBM
Estimates of Soviet Force Effectiveness
Tho evidence and Its interpretation,
The Hlstortcai
Test Range
Production and Deployment Evidence
Influence of Soviet Statements
Public RefJec*1ons
Clarification and Correction of Estates,
Cross-conf iniing Evidence
Timing Considerations
TOJL-SKRfT"
t IWlmlbralt: II.
1NTFL1 KjLNCE ASPECTS OF THE "HI SSIP"
NATURE OfISSUE
The "miss! lo gap" crisisrisis In American confidence about the future adequacy of US power to deter Soviet aggressiveness. The issue was whother the defense programs of the Eisenhower administration in thes would extend America's superiority in strategic strength Into thes, or whether Instead the administration's policies were allowing the USSR to shift the strategic balance markedly In Its own favor.
The "gap" or "lag" was often viewed In very simpleis, the comparison between the number of nuclear-armed Intercontinental
ballistic missiles OCEMs) the Soviets were expectod to have operational In the ensuing few years and the number the US planned to have operational in Those yoars. More sophisticated interpretationsexample, ones which edded shorter range nlullM Into the ralance, and ones which calculated ho- many US bases the Soviets might krzzk outurprise ICO* attack. In these as in the simpler interpret ion, however, the nub of the Issue was the expected relativef-the two sides In IC&Vs, the degree to which this would affect the overall strategic balance, and the effect on the American deterrent.
US Intelligence was the authority responsible for estimating tno Soviet side of tho stratoglc equation. eries of National Intelligence
TQR-SKRET
estimates (Hits) issuedheeles Intelligence Ooord (USIBI set forth tho judgments which contriSuted to tho "missllu gap" crisis. These were issued as highly-classified documents within tho Executive Branch and formed tho basis for highly-classlMed brioflngs to Congressional Committees. As 1ho controversy unfolded, much was said publicly by Administration officials, legislators, and ne-smen about the nature of these estimates. Their terminology, methodology, and accuracy were all debated at length. So was the effect on the "missile gap" of adjustments made In the estimates from time to tine.
Although the NIEs contributed to the "missile gap" crisis, US Intelligence was not wholly responsible for It or for the public reaction to It. These aspects resultodomparison with US programs. omparison, Allen Dulles pointod out repeatedly, was not within the purview of the Intelligence community. "We are not In the buslnoss of passing upon whether there ishingmissilee said He was not in the business of revealing his estimates In public, either. But Intelligence was caught In the middle Inmajor policy issueepublican AcminisT-atlonemocratic Congress In the years between Sputnik and the US Presidential elections. Tho public ventilation of this Issue and of the Wis was not the doing of Intel1igence.
JOMEOgf
TOtSKtff
ICS
On Ihe other hand, there Is no denying that the Key lac* about the NIEs of the period was that US Intelligence was grosslySoviet ICBM prospects tor thes. Anddid not learn the truth and correct Its gross error
THE ESTIMATES ANO THE CHANGES IN THEM
During the years7he USIB completed sixteen separate estlmatos dealing with Soviet 1CEM development and deployment. This was an unusually large number ot estimates to be made on one topic. The reasons included the difficulty of the problem, the uncertainties Involved, and the critical nature of the subject.
The NIEs were ultimately the responsibility of the Director of Central Intelligencehe position hold by Allen Dulles throughout the period. The estimates wore not exclusively the work of the XI or the CIA, however. As chairman of USIB, the DCI was joined by the other top Intelligence officers of the government, both civilian and military, in considering and passing final judgment on draft NIEs.' The drafts were prepared by CIA's Board ot National Estimateshen under Sherman Kent, on the basis of contributions Invited from all agencies having competence in the subject mattor. BNE coordinated these drafts
* During most of the period under consideration, the membership of USIB comprised the DCI and the intelligence officers of the Department of State, the Army, Navy, and Air Force, the Joint Staff of the JCS, the National Security Agency, the AEC and the FBI. The FBI member reguiorl abstained from NIEs on Soviet militory programs. Letohe director of the newly-croated Defense Intel itgence Agoncy was added.
-J-
with representatives of the USIU member-agencies be'oro submitting them to USIB for review and approval. For estimates of missile development and performance, BNE usuallyrecoordineted contribution from the Guided Missile and Astronautics Intelligence Committee of USIBhose members were technical missile specialists from each agency.
Numerous consultants from outside the intelligence community were also called on by the various agencies as they analyzed the data and prepared contributions to estimates. , CIA alone used approximatelyonsultants from outside the Intelligence community, largely from Industry, the universities, end governmont, to help In analyzing various aspects of tho missile problem. In addition,8 onwards, the OCIroup of senior and respected outside experts under Mr. L. A. Hyland of Hughes Aircraft, to review the evidence periodically and to offer their Judgment as to the reasonableness of the estimates being made on the status of the Soviet ICBM program.
All of thisonsiderable number ofhundreds In the analytical and estimating business, not considering those concerned with information collection and processing. And the process of making estimates in this field was so time-consuming that the BNE and agency representatives working on these NIEs were sometimes in
session (or weekstretch. One NIE, originally scheduled (or completion in the latter partroved so troublesome that it was not finally passed by USIB until The tirao taken, however. Is perhaps as good an illustration as any of the determination of ONE and the XI to hear all relevant evidence and argument so as to arrive at the best estimate possible. Throughout the entire period, there was never any serious suggestion from any official or agency or consultant of the Executive Branch that any information or opinion failed torompt. Impartial hearing.
The estimates were heavily caveated in an effort to warn users that the evidence was fragmentary and inconclusive. Ranges ofof various sorts were. "probable" dates vs.ates, and spreads of high and low numbers. While these practises were not new, they were trcre frequently applied to the ICBM estimates than was usual in NIEs, and the estimates were thus quiteated.
Another element of complication was the presenceajor dissent registered by at least one member of USIB in every estimate in the series from the beginning0 onwards. Often several members registered separate dissents disagreeing with the USIB majority and also with each other. The most extreme case was the NIE issuedn which the DCI stood alone In his estimate of The
MU/WJ
probable Soviet ICBM deployment program and two other alternative programs, with their adherents among members ot USIB, woro written Into the text of the estimate.
Estimates of Soviet ICBM Deployment
The main elements In the NIEs contributing to the "missile gap" crisis were estimates of "when" and "how many." These questions were the sole ones in the stapler, numbarsHutchIng version of the "gap" controversy. Questions of "how good" were vital to the moreversion which sought to measure the Soviet threat against cities and SAC bases and the other targets ot the deterrent forces of that time. Theso ICBM performance aspects of the NIEs will be discussed In later paragraphs.
The many complications In the ICBM estimates, the frequency of revisions, end the attempts to refine the terminology, make It difficult even In retrospect to compare successive estimates In the serlos. Even now there is the possibility of confusion which was voiced by critics of the Elsenhower Administration of the time. But in order to deal witharge number of successive estimates, we must try to strip away the nuances and get at the Impression thoy created about SovietcapabiIities.
TOP^KftfT
Id
On this stripped-down basis, tho NIEs olall Into several groupings by time:
--Inefore the first Soviet ICBM shots and the launching of Sputnik, USIB estimated that The USSR would probably have an initial operational capability (IOC)ow ICBMs0 In very guarded terms It was estimated to be "feaslblo end reasonable" for the Soviets tostockpile"perational ICBMs by the endy the endy the endssuming that IX was at the earliest part of the spread date.
n Its first post-Sputnik estimate on ICBMs, USIB odvanced the probable IX date toot Indicated that the Soviets could have an "operational capability"CBMs In the periodond withnor atear later.
Way to8 various adjustments were made which pushed the probable dates beck. In December, the probable IX date was givenho probable time ;erlod by which the Soviets could have an "operational capability"CBMs wasond The possibility was left open that the Soviets night haveut this possibility was rated unlikely.
TOE-SKRff
l
9 andhe estimated IOC date was changed This date was held tlwoattor. The ICEW programhole was describedcrash" program. The basis for measuring the threat was changed to "ICBMs onnd the probable number of operational Soviet launchers was then estimated successively as follows:
Made
0 CIA or USIB Majority (Spread of dissents)
0 CIA or USIB Majority (Spread of dissents)
1 CIA or USIB Majority (Spread of dissents)
1 CIA or USIB Majority
(Scread of dissents)
0
35
30
For
150
a
In
For
)
:
tIe more)
For
)
)
the table above, It shouldted trateneral rule Air Force produced dissenting estimates higher than tnose of the CIA or USIB majority, whereas Armv was usually on the low side. Army was sometimes joined by The Joint Staff and the State Department sometimes estim-'ed on the high side. The foregoing summary shows that theof?8 were the most alarming from the pointiew of "how many"
-tt-
TOP.SeCRET
and "how scon." Tho worst case presented by these estimates was that the Soviets might be capable otCBMs readymploy as early asr the beginning of The general trend was to back off from the Initial alarming post-Sputnik estimates. The transition in measurlng-stlcks from "operational ICBMs" to "ICBMs on launcher" made it very difficult to compare the estimates made before andut the estimators' intention was to make relatively gradual adjustments when changing the numbers and dates In each of the NIEs through No very marked change In the impression of the threat occurred until US Intelligence discovered that the Soviet deployment program bore no resemblance to the existing estimates. That discovery resulted In the sharp downward revision of the estimatesI.
TOPJiKftfT
For purposes of comparison, the actual growth of the Soviet force of operational ICBM launchers as seen from the vantage point8 Is
pi yict erformance
The NIEs also estimated Iho oftoctiveness of lovlet ICQMs, as they lit lor all other weapon systoms. The general trend in this aspect of the US assessment was to Improve tho estimated performanceof the Soviet system as time passed, evidence accumulated, and more was learned about Soviot and US experience lr. testing ICBMs. The ostimete of Soviet operational 1CW accuracy was first put at. CEP at ICC and was Improved gradually until Int was estimated.edlo-lnertlel version at IX. Warhead weight, estimated atounds Inas raisedndnd finally, from9 onwards,ounds or more.
Estimates of Soviet ICBM reliability were irore complicated. They wore more heavily dependent on tho deployment, scheme to be used by the Soviets anc on training methods and standards, which US experience showed could vary greatly. In general, as the estimators learned sore about the
complexities of big missiles, and especially problems involved In building an operational force, they became more csjtious with respect to tho reliability figures they presented. 'educed rare factors and incorporated ranges and spreads. The reliability of Soviet ICBMs was givenlatercent 9roken down Into in-commIssion rate, on-launchor reliability, anc in-flight reliability, each of them givenpread figure dopending cn conditions. ombined spreadercent was estimated1 for that year.
TOkSKItff
Estimates of SovieI force Effectiveness
The estimates ot missile performance, such as accuracy and reliability, -ere normally contributed by the technical specialists of tho GMAIC committee. Though sometimes hotly debated by BNE and the USIB representatives, these estimates were almost always occepted with little or no alteration after they left the technical subcommittee. Once accepted, however, theyBNE and the representatives In their Judgment about the number of ICBMs the USSR would find It worthwhile to have, which was also called Soviet "requirements." As indicated above, the OCI considered that comparative or net estimates Involving judgments about the effectiveness of US forces were beyond the purview of intelligence. But Soviet requirements for ICBMs to be employed against US targets could be calculated without regard for US force effectiveness. Because the ICBM was unstoppable. Its effects against fixed targets like the bases of US deterrent forces could be calculated solelyunction of its performance and tho physical character of the targets themselves.
So0 andn an effort to make more realistic forecasts of Soviet ICBM programming, US intelligence made calculations of the threat ICBMs would pose to US fixed targets of various kinds. The calculations showed that surprisingly small numbers of ICBMs having the performance characteristics credited to them would be sufficient theoretically to destroy the major cities of the L'S and the SAC bases and other soft targets which then housed the grest bulk of our retaliatory
iur icwtf'
llltU/'Ai
systems. It also appeared that the timo period during which the US base structure would be highly vulnerable would be quite brief, because of the planned advent of hardened US facilities.
The numerical forecasts made by CIA and the USIB majority were not slavishly geared to theso "requirements" calculations, which were only one of meny inputs to the estimates. Indeed, the estimators concluded thatumber of reasons such theoretical calculations should not be thought to represent. In tho Soviet mind, an opportunity to build forces which could permit them to attack the US without receiving unacceptable damage in retaliation. This was spelled outclearly inf On the other hand. In that NIE as well as the accompanying, the estimators made It equally clear that they thought the Soviet force would probably soonajor factor In the strategic and political competition between the USSR and the US. Their key Judgments made on this question In the two NIEs wore that "long-range missiles will enable the USSR to overcome its inferiority to the US in nuclear strategic attacknd further, that "the USSR will1 have its most favorable opportunity toecided military, political, and psychological advantage over the US by the rapid deployment of operational ICBMs." ent on to show how the probable growth of the Soviet force would soon put In jeopardy the principal US cities and would,onstitute "on extremely
TQE-SKRFf
i
dangerous threat" to soft US deterrent force bases.
Thus although those NIEs quite pointedly stated that the USSR would not expect to build forces which would nullify 1ho US deterrent, thoy did spell out the probable Soviet force effectiveness for the1 time periodanner which was directly relevant to the "missile gap" Issue in its moro sophisticated form. Tho conclusions drawn were tho logical consoquonco of examining what wero considered to be reasonable and probable Soviet deployment programs.
THE EVIDENCE AND ITS INTERPRETATION,I960
A fundamental fact about the ostlmates ofs that the estimators were having to employ Inadequate evidence and had an Inadequate basts for its interpretation. As to the evidence Itself, the Sovlot security system was the principal reason for its Inadequacy. Tho Soviets made special efforts to guard their missile developments from Western observation. In the early years, they kept German missile scientists in ignorance of tho Soviet program while they were boing milked of their experience, so that when returnod to Germany they would have no Information. Khrushchev once admitted that even the names of Soviet missile engineers wero classified. Those extraordinary Soviet measures did not prevent US intelligence from monitoring test range activities by radar and other technical means. Out Intelligence wasong time quite unable to penetrate Into Soviet strategic planning and technical design or In
0
finding ways to search tho vast territory of the USSR for production and deployment locations. S Intelligence was in fact quite myopic--It could see missile test activity but could not see tho other aspects of the missile program which were going on around the testing.
Asase for Interpretation, Intelligence suffered from the newness of missiles and from the lack of US experience with them. Americans7 wore unable to answer such questions as what materials of what weights were required to protect an ICBM nosecone on re-entry. Differing answers were forthcoming from US designers when asked whether fixed or mobile deployment was most desirable. And sharp conflict was generated by questions dealing with the utility of missiles for various purposes, because the US was Itself debating their strategic role.
The Historical Context
In these circumstances, the historical context had an unusually great Influence on the judgments made by the estimators. Tho historical context always has an Important bearing on the making of estiamtes, even In the military and technical fields, because.in the final analysis the NIEs are about human decisions, their causes and tholr consequences. But when direct evidence on the actual path of events and decisions Is sparse, the estimator must rely rore heavily on the general appearance
of the surrounding terrain. Several elements of the historical context
0
had considerable influence on the estimators as they addrossod themselves toUM problem in the's:
--Sputnik, though accurately forecast by Intelligence, hadbeen jolting in its Implications about Soviet missile capabilities and technical prowess generally.
US had committed Itself vory heavily to deterrence by threat of massive retaliation against tho territory of the USSR Itself, and It was logical to believe that the Soviets wished urgently to acquire counteracting capabilities of their own.
--The Soviets had an acknowledged capability toarge Intercontinental bomber force, and earlier In the decade had bogun what was believed to haverogram to build one, but this program had been curtailed presumably In favor of mlsslles.
--The Soviets were known to hove developed snort and medium range ballistic missiles, and they were believed to have deployed them extonslvoly enough so that they could now focus on ICOMs.
eneral sense of Intensified Soviot competition with the US in many fields, fosterod by the ebullient Khrushchev and highlighted by expressions of alarm from US observers outside the Intelligence community. Including critics of the Administration In Congress and elsewhere.
TOP-SECRET
All of these factors Influenced the estimators to give credence to the elements of evidenco suggesting an early and urgent Soviet ICBM buildup. Conversely, in these circumstances the estimators were very slow to conclude that an absence of information meant an absence of activity. Thus the contextonsiderable bearing on the Initial post-Sputnik overestimate and made each downward adjustment, howeveratter ot lengthy argument and soul-searching.
The historical context also pushed the estimators into what in retrospect seems like overprecislon and over-refinement In light of the meager evidence available at the time. The middle ands were years of sharply climbing weapon system costs and of rapidlyead-times. US defense planners sought to overcome these problems to some degree by putting much pressure on Intelligence to come up with finite predictions against which they could plan. They were especially Insistent In The missile field, where everyone was learning and the strategic Implications wore critical and the Soviets seemed toeg up on the US. The estimators responded as best they could, and In the circumstances they tended to resolve their uncertainties In the direction of prudence."
Tost Range Evidence
b
(luring the entire period under consideration. From the first ICBM test inhis evidence was good enough to tell promptly when an ICBM had been fired, and usually whether or not It had been generally successful. Coverage of the test range also provided an opportunity to accumulate, though very gradually, evidence about the performance and configuration of the system. Soma hints about Its likely deploymentleast that it would be heavily dependent on the Soviet raitderivedhotography of the rangehead beginning in the same month as the first ICBM firing.
From the data on the first firings from Tyuratam In the fall7 (two ICBMs and two earth satellites, allombined with knowledge of the firm foundation the Soviets had acquired throughand extensively testing shorter range missiles at Kapustln Yar, It was concluded that the USSR was making maximum use of proven components and had already largely solved most major ICBM component problems. The7 estimate rested in part on this conclusion, and also on the belief that the Soviets would seek toubstantial ICBM capability as rapidly as possible, wouldapid program ofests In the very near future, andery high priority if not "crash" program to deploy the system.
The adjustments made in the estimates in89 were based very largely on test-range evidence. Throughout that period the
JOMKRET"
I
test range was virtually the only source of what Allen Dulles called "hard" evidence onIs, evidence known to be true as opposed to such things as Soviot statements. Although the Soviet program suffered some test-flight failures, reliability was generally good. The pace of testing was much slower than originally expected, however. The estimators concluded that the ICBM program wascrash" program, but was "orderly" Instead. Tho estimated IX date was pushed back, eventually This was the time when the Soviets had In fact accomplished aboutuccessful ICBM shots,umber of space launchings.
During this period, tho Intelligence analysts had to do some wrestling with the problem of negative evidence from the testthe apparent absence of successful test firings, especially in tho letter parteries of Inquiries was made in order to bo sure that there was no significant chance of either undetected shots from Tyuratam or the oxlstence of other ICBM tost ranges. At one point. Senator Stuart Symington, the Senate's loading expert in air and missile matters, challenged tho credibility of the DCI's testimony In briefings to the Senate. He did notove that the number of test-firings was as small as Mr. Dulles was reporting, inoetter to the President alleging that there was intelligence Information to indicate many more test firings ond other evidences of Soviet accomplIshmant.
0
This too resultedengthy investigation, which finally showed the Senator's allegations to be unfounded.
The test-range data, while clearlyowerful and reliable Soviet IC8T4 and space booster, also pointed to potential logistic and operational problems because of its very large size and non-storable propellants. By the endhe NIEs acknowledged that the gross weight of the vehicle mightounds or more. Insteadounds as originally estimated. The estimators mentioned the possibility that the Soviets would experience deployment problems with the first-generation system, but they failed to draw any corresponding conclusions about the size and pace of deployment.
9he NIEs correctlymaller, follow-on ICBM having less bulk and greater flexibility than the first-generation system, but again without drawing any particular implications about force programming. The Soviets actually beganaunch facilityecond-generation systemnd It was first photographed while still under constructionission In US monitoring systems collected evidence of the test-firing program from its Inception In0 or As late asowever, the estimators could not tell whether the new facility wasew ICBMimplified deployment method for the original system. Technical analysis of the radar
r
and oiher monitoring"Kit clorlly tv* Issue
by1 It had become clear that not just on'j but two dHterent follow-on systems wore under tost at Ty.ratem, end it wjs also dear that the programming of the Soviet force buildup had been very significantly affected.
Production and Deployment Evidence
Ashe estimators concluded on the basis of test-range data that the Soviet system was ready for operational use, but they were in the awkward position of having also to say that "the hard evidence at hand does not establish whether or not series production of ICBMs has actually begun, nor does It confirm the existence of operational launching facilities."" The estimates made on Soviet ICBM deployment at that time did not, as is usual In "ordor-of-battie" assessments, beginirm base of Information and understanding about the current situation, from which future projections ere then made. The es'lmators' Interpretations began Instead with the judgment that The ability of US Intelligence tormation on ICBM production and deployment was inadequate. rc" evidence could not be relied on as the basis tor anould the absence of such evidence be taken as reflecting an abserr* of activity. This
* aragraphs,
TC. IIUW6U
judgment was strongly supported by the various consultants the XI end USIB called in to appraise the evidence and hear the arguments. Inor example, the Hyland Panel reported to the XI that It found "the quantity and quality of the total evidence available to be dangerously unsatisfactory. In particular is the situation true regarding evidence on production facilities, training and deployment.'
, therefore, the estimators had to try to bring to bear whatever elements of Indirect evidence and US experience they and their advisors thought relevant to quantity production and deployment of missiles. The rate of test-firingsactor in the estimate because It was thoughVto Indicate the pace of the overall program. So were such things as US training schedules and the estimated production capacities of various airframe and automotive plants in the USSR. In general, this type of evidence was used mainly to attempt tofeel" for what wouldeasonable size and reasonable schedule for the Sovietprogram, rather than to put any firm constraints on it. Adjustments in the numerical estimates prior1 were therefore based on general estimative Judgment, even though sometimes they were assumed to be based on hard facts.
One result of the need to seek out indirect evidence of this type was gradually to give CIA production and deploymentetter
tr *
Wi
understanding ot what was involvedig missile program. Probably the most significant aspect of the improved understanding was the realization. In9 andhat the meaningful measure of anrogram was the number of operational launchers rather than operational missiles. There were two main reasons for this: one was that the threat to the US was best measured by salvo capability and not total missile Inventory; the other was that the pacing elements inissile forco were the building of launchers and ground support equipment and the training of launch crews rather than the manufacture of missiles themselves. This caused the change in the unit of measure used In the estimates inut more important It focussed the analysts more on launcher deployment and less on missile production.
Shortly thereafter, InIA setuided Missile Task Force of both scientific and production specialists, to focus greater analytical effort on the pace and magnitude of Soviet missile production and deployment. This group was set up by Robert Anory,hen Deputy Director for Intelligence, and was charged by him to "track the fox to its lair." This was one of several such special groups In thewore sot up for essentially the same purpose in USAF Intelligence hoadquarters and In the Intelligence section of tho Strategic Air Command. The conviction in each agency was that there must be some evidence of deployment, especially since the estimated
i','. i
IOC data had passeduildup of launching silos had presumably been under way for some time without detection. In SAC's case there was the further conviction that suspected deployment sites must be identified as candidates for targeting.
Lists of suspect locations were drawn up by these groups, collection guides were issued, and bits end pieces of evidence were weighed and argued before the GMAIC committee of USI6. Moreocations In the USSR had been considered as candidates for ballistic missileby Of thehich were written up in some detail by GMAIC at that time, only four were evaluated asCBM deployment locations on the basis of the very fragmentary data then available. AM the othors were considered doubtful, negative, or undetermined. None wore evaluated as probable or confirmed deployment sites. As it turned out, the0 list actually included the only two ICBM deployment complexes the USSR had at thata first-generation complex, and Verkhnaya Salda, second-generation complex under construction. (The list also included Polyarnyy Ural, where it was later determined that the USSR probablyirst-generation complex and then cancelled It some time) But there was no way of sorting these out for sure from the other names on the list, which turned out tootpourri of MRBM and IRBM sites, other missile-related facilities, and many non-missile facilities. In any case, such lists had little bearing on the numerical estimates made because the
TOP-^OT l
estimators did not bo' ieve they could rely on the ability ot Intelligence to acquire the necessary Information.
Another contributor to the effort to accumulate ovidonco onwasverflight program. Twenty-eight missions were accomplished over various parts of the USSR and East Europe between Many Important Installations Including the missile rongehoads at Kapustln Yor and Tyuratam were covered during tho four-yooro of this program. In the ontlro program, however, no photography was acquired of any ballistic missile deployment complex which was sufficiently far enough along In construction to be recognizable. (Later It turned out thatad photographed the Verkhnaya Salda complex, but so early thatingle roll spur was visible.) Again, these negative results wore not regardedasis for altering the estimates. Pertly this was because tho area coverage ofas quite limited. Moreover, until the first ICBM complex was seen and recognized, there was the nagging concern that deployment activity might not be distinguishable from othor Worse yet It might be invisible because It was concealod or mobile.
Tho first photographic reconnaissance satellite mission was conducted Inew months aftor tholight. Nothing was learned about missile deployment from It. The second successful satellite mission, inhotographed 'ho ICBM complex at Plesetsk, but low sun angle, small scale, andcover prevented its identification at that tine. It was not until the next two satellite
TOFU&KKET
ll
missions. In tho middlehat the photographic coverage revealed for the first time the extent of Soviet deployment of ICBMs and other bal11stlc miss!le systems.
Influence of Sovlot Statements
From the political and psychological point of view the ICBMoviet weapon from the very first test-firingnd so were the Sputniks and Luniks. The Soviet leaders had longense of strategic inferiority to the US, and they were clearly determined to take the political high ground with this new weapon and to use it_to advantage In International affairs. They had done so with medium range missile capabilities when they rattled their rockets at the time of Suez and Hungary. With the ICBM and space launching programs theyteady barrage of statements, boasts, end "disclosures" designed to convey the impression that Irreslstable new capabilities were now in the Soviet arsenal and further, that the "correlation of forces" between East and West was irrevocably altered.
In general, there were three types of Soviet statements about missiles which the intelligence analysts had to evaluate: boasts about milestones, hints about plans, and statements about policy. Khrushchevrinciple source of all three types, but other Soviet officials also made such statements In the relatively free exchange of visits and correspondence which existed at the time.
JQE-SKRET
TOPJfetKTT
Mti/W
As to the boasts of milestones, Khrushchev, Oefense Minister Hallnovsky, and otherstoady How of commentary, designed to show rapid progress toward operational ICBM capabilities. Khrushchev claimed publicly In8 that series production of IC8Ms had begun. In December of that year, another Soviet official told Dr. George Klstiakowsky of Harvard (soon to become Science Advisor to the President) that numbers of operational ICBM bases already existed. Inhrushchev told Vice President Nixon that Soviet launching facilities were cheap because they were mobile. In November of that year, he claimed one plant wasockets per year, and In0 he said dispersed and camouflaged launching bases existed. Each such statement was analyzed In detail andart of the estimators* stock of information. Such statements obviously did not prevent the estimators from gradually moving away from their very alarmist initial post-Sputnik estimates, though the8 boasts did cause USIB tolight "hedge" Into one estimate, to the effect thatonsiderable operational ICBMannot be ruled out as Impossible"ery high risk program.* In general, however, Intelligence evaluated these early Soviet claims as exaggeration from an emerging capability to an existingIs, dealing In futures.
* See,nd
TQR-SKRET
On the other hand,eomod reasonable that some suchere valid hints of real Sovlot plans. eor plant remark was not consldoroo yrounds toromewhat lower figure which Intelligence had arrived ateasonable number toauncher deployment program. But some other statements he made privately, foremark to Ambassador Harriman about the Soviet budget for missiles over the coming several years, were taken into account in estimating that Soviet force goals would probably call for several hundred operational launchers.
Finally, policy statements the Soviet leaders made to tho Supremo Soviet were usually taken quite seriously. By far the most significant of these were the announcements by Khrushchev and Mallnovsky Inhich stated that rocket forces had become the most Important branch of the Soviet armed forces andne-third reduction In military iraipower, to be accomplished In the nextonths or two years. Ineparate arm of service called Strategic Rocket Forces -as established. US analysts interpreted these measures as signs of Soviet confidence that attainments In nuclear and missile weapons would permit them to meet their military requirements with tower men. Policy statements of this sort also soomed clearly tooviet conviction that the ICBM wouldevolutionary effect on the USSR's strategic Situation. The US Intelligence estimators shared this conviction. It
TQt-SKftFT
I',', iiinvwi
influenced their sense of the likely tempo of the Soviet program and affected their evaluation of the sparse evidence o' other types they had To work with.
PUBLIC REFLECTIONS OF ESTIMATES
The public repurcussions came principally in the form of statements and counterstatements by defenders and critics of the Administration's defense programs. Many such statements were charges that the post-Sputnik levels of American defense effort were inadequate. Many Included recommendations for Increasing the tempo of US strategic missile programs or for taking actions to reduce the vulnerability of US strategic air power to surprise attack
The degree to which the statements were based specifically on Intelligence estimates varied considerably and often could not be determined. Some were very generalized judgments like thatockefeller Panel report Inhich said "It appears that the United States is rapidly losing its lead ove- -he USSR In the military race." Others were straight leaks ofanson Baldwin column of9 whichuoted the number of Soviet ICBM shots to that date, as well as tfc= *hen-current intelligence estimate on the times by which the USSR couldoperational capabilities with various numbers ot ICBMs. Suffice it tothat despite the highly-
classified nature of the estimates and of the briefings given byairly accurate picture of the findings on this subject was revealed to the public by the various spokesmen of the Administration, Its critics in the Congress, and tho columnists.
Tho expressions "missile gap" or "missile lag" were of course not used by Intelligence officers, who made no such comparisons. The first public use of the expression was probably by Joseph Alsop,olumn which appeared In The expression "missile gap" quickly came to have wide public currency and to obscure the comparisons of overall US and Soviet military strengths with which Administration spokesmen sought to counter their critics. Inefense Secretary McElroy was drawn Into admitting publicly that in thes the USSR could have three times as many ICBMs as the US planned to have. In0 his successor. Secretary Gates, who had justow ogtlmate of probable numbers of Soviet operational launchers, sought to show publicly that tho new estimate narrowed the differences. Me Immedistoly became embroiled In controversy over whether tho comparison was valid and whether Intelligence was justified In changingcapability" estimateprobability" estimate. In late January, Mr. Ouiles attempted to sot the record straight bypeech which described the methodology at some length and stressed that incorporating "the elements of programming and future
TO^SKRET
iAH/'U
intentions" Into estimatesormal oice procedure. Iful the next day Senator Symington chargede intelligence bookseen juggled so the budget books may be balanced."
Also inalculations of the vulnerability of US retaliatory forces to missile attack were made puolic by Gen. Power of the Strategic Air Command. The general announcedpeech ThatCBMs and en equal number of shorter range missiles, the Soviets under conditions of no warning "could virtually wipe out our entire nuclear strike capability." ebruary, Joseph Alsopolumn entitled "The Oulleshich quoted the DCI as having told the Senate Space Committee that the Soviets wouldCBMs operational by To be sure, Alsop said, by that time SAC would have more weapons to be hit and thus the Soviet requirement would have climbed, but who was to say that Mr. Dulles' estimates were not too low?
Thus in the spring0 the public press carried numerical Information from credible sources whichh the simple and sophisticated versions of the "missile gap" Idea. Even the downing ofn May Day was considered toearing on the question because its overflight activities were said tothat the intelligence estimates had been based on solid evidence. Tha figures and sources
IDWKfflT
on the subject. Citing these and other respected authorities on defense natters, the Democrats were able to sustain the issue in the Presidential campaign.
CLARIFICATION AND CORRECTION OF1
In the spring and summerhree elements of evidence finallyajor clarification of the status of the Soviet ICBM program. From these elements the estimatorsasis for sharply reducing their previous estimates of the ICBM threat and for making estimates on this critical subject much more confidently than had been possible before. This did not quite happon all at once, Out almost so. It began to be reflected In an NIE in1 and was the main featureupplementary follow-up NIE in September.
Cross-confirming Evidence
The first new element of evidence was the initiation of Soviet testi of second-generation ICBMs. Tho first successful Soviet shots occurred
1
follow-on vehicles were being tested, it was not immediately recognized that these were new types of ICBMs; this wasossibility in an NIE comploted innd was reported as a
certainty in the NIE of ead-time, this
TQR-SECRET
res wmu/ui
made it clear that the Soviets had docidedecond-generation program at least as earlyB. It tended to contirm that the simplified launch facilities built at Tyuratam9ew system ratherolution to the logistical and operaflonal problems of the old one. This In turn strongly suggested thatof the first-generation system would be quite limited.
Second, two very successful satellite reconnaissance missions. In June androught the first positive identifications of ICBM deployment In the field. The first-generation complex at Plesetsk was photographed and several second-generation complexes were found undor construction. More Important from the estimative point of view, these Installations were found to be large, readily-recognizable on the basis of test-range signatures, end dependent for their logistic support on the Soviet rail network. asis was thus established for searching for complexes by regular satellite photography, for identifying,and timing the launch facilities as thev were constructed, and for ruling out the existence ot such facilities at locations where good-quality coverage produced negative results. In terms of geographic coverage, these initial satellite operations provided information on about half the total railroad route mlloage In the USSR. Coverage which was believed to be good enough to tell whether an ICBM complex existod or not was achieved over about half of the portion of the USSR within
TOE-SKKFT
rco i
which deployment was most likely. Only five confirmed or possible ICBM complexes were found in all of this photography;as the first-generation complex at Plesetsk and the other four were second-generation complexes under construction.
emorandum to the DCI after the June photography had been scanned, the EWE assessed It as "the most important breakthrough into the Soviet long range ballistic missile program since the acquisition of radar coverage of test firings and TALENT) coverage of test range installations seme years ago." The menx> went on to say that the results gave promise "that frequent coverage by this means can reduce the problem of current and near future Soviet strength in long range ballistic missiles to one of order-of-battle rather than, as heretofore, one of estimating on the basis of inference from Indirect evidence.'
The third element of clarification was the acquisition from Col, Penkovskiy in1 of vital Inside information on the status of the Soviet ICBM program. While Penkovskiy had begun providing Information inhis was his first contribution on this particular subject. From senior Soviet officers connected with the missile program, Penkovskiy had learned that Khrushchev was deliberately magnifying his missile successes to Impress the West and that the USSR then had ateryi ted capabiy to stri ke at the US wlth missiles. Mions of rubles
itjpatuccf
I
were being spent on The missile program, however, and now types ol missiles and fuels were being worked on. While Its Soviets did nor have the hundreds of ICBMs they pretended to have, they would have them some day
These three typos of evidence were mutually cross-confirming. Theyound basis tor key new judgments about the Soviet program In the NIE of The estimators concluded, first, that In8 the Soviets had decided to deploymall force of first-generation ICBMs while pressing toward second-generation systems, and second, that the effect of this decision had been tolow plateau" of deployed strength In ICBMs which would probably not increase markedly In the months Immediately ahead.*' The "low plateau" was
estimatedperational launchers asI; the correct number later proved to be four.
From that time forward additional information, much of it from satellite photography, permitted further retineme-ts and adjustments in the estimates. Since then, the current progress of Soviet ICBMhas been followed and recorded on anattie" basis, and it has been possible to build an accurate andate picture of current and near-term strength in ooerational launchers. Thereumber of loose ends, of course. The chief intelliaence officer of the
ee NIEI,nd 3.
iur>tcKti
USAF continued toarger Soviet capability than the XI and the majority of USIB. onsiderable time SAC held to the old lists of suspect deployment locations and was unwilling to use the photography end the test-range signaturesasis for negating the presence of launch facilities. But it was the1 NIE which withdrew the intelligence basismissile gap" and established for the first time an unlnflated view of the Soviet ICBM threat.
Timing Considerations
From the foregoing review It is clear that thereumber of factors contributing to the gross overestimates, and that these factors Included very general influences on the estimators as well as specific elements of evidence. Certainly tho limitations of the collection capabilities of US Intelligence at the time hid certain developments which. If known and recognized earlier, could have broughtlarification and correction of estimates sooner.
1 the discussion in the estimates of follow-on or second-generation ICBMs was very cursory. Tho estimators tended to speak of the ICBM. Their analyses of production and doploymont considerations were basedingle, smooth, uninterrupted buildupingle system. Partly this was because the testing record made the first-generation system look so good and because their information did not
show clearlyollow-on system was irmincnf. Partly it was because the limited US missile experience provided the estimators with an inadequate background to appreciate the operational difficulties the Soviets might encounter with their first-generation system. Partly also it was because the historical context of thes had conditioned the estimators to expect an urgent Soviet ICBM buildup.
The facts as now available show that Intelligence might theoretically have known as early8 that the Soviets were beginning to design second-generation ICBM systems. Given the Soviet security system, however. It Is not realistic to expect Intelligence to penetrate into design bureaus at such early stages, or to penetrate the strategic and operational decision-making which authorizes the bureaus to proceed. By the springhotograph was taken of new test-range facilities under construction, and perhaps at thatetter understanding of the missile deployment problem would have told the analysts that the ground support equipment to be used at these facilities vould be so different from that of the first-generation system that it must be for something new and different-
Even if Intelligence had appreciatedhat there would be discontinuity in the Soviet program, however, It seems unlikely in retrospect that this knowledge alone would drastically have altered the estimates. Foilow-on weapon systems can be desicned, tested, and
!
Introduced in both the US and USSR without sharply breaking the (lowuildup when the degree of urgency and commitment of resources is sufficient. In retrospect, therefore, it appears that the Soviets simply did not authorize an early ICBM buildup of the magnitude we had expected, but that instead they had decidedore gradual buildup, together with their MRBM, IRBM, and other military programs as well as the impact of their space activities, would be consistent with their strategic, political, and economic objectives as they then saw them. Their difficulties with the first-generation program, end the cutback of It suggested by the possible cancellationomplexerely made the buildup even slower.
If It is accepted that Intelligence cannot expect to know about Soviet decision-making and design bureau activities in the preliminary stages, then virtually the only way of knowingeapons program Is through monitoring the actual progress of the later testing and deployment phases as they occur. In the case of the Soviet ICBM program the concurrency between testing and deployment activities, and the very large regions where deployment might occur, made satellite photography of likely deployment areas tho real key to understanding the program. The studies of factors affecting production and employment which were used to bolster the meager evidencenade it possible to Interpret tho "hard" photographic evidence better when it was finally
negative conclusions, the estimators had no basis for correcting their Inaccurate picture of the Soviet program. This was finally achieved1 through the employment of satellite collection systems which had been under development ins and were first employed shortly before the Presidential elections.
Once the deciding elements of evidence were in hand. Intelligence moved promptly. The satellite mission which BNE describedbreak through" was concluded onhe BNE memorandum to the DCI was completed onew estimate was formally initiated at JCS requestuly, the USIB representatives beganNE draft in August, and USIB approved the revised estimate oneptember.
Original document.
Comment about this article, ask questions, or add new information about this topic: