Created: 9/19/1972

OCR scan of the original document, errors are possible



Controlled OHvem Limited Dithibuiion




Soviet Military Research and Development




The following intelligence) organizations participated in ths preparation of the estimoln:

The Central Intelligence Agency and lhe inielligence orgoniialiorrt of lhe Depori-ntenli of Slate and Defame, ond the NSA.


lhe Deputy Director of Control Inielligence

Ihe Direcior of Intelligence ond Research. Deportment of Stale lhe Direcior, Do fern* Intelligence Agency Ihe Director, Nationol Security Agency

The Director, Division ol International Securitytomic Energy CommUtion Abstaining;

The Atclttont Director, Federol Bureau ol InoeMigolion, ond tho Special Aitiitont

to the Secretary of the Treaiory. the uibjecl being oulude of their jurisdiction.










A Research and Development Facilities 5

cientific and Engineering 6

C. Outlays for Military Research and Development and Space Programs 7



Military Compared to thc CiviUan Sector in the USSR11

Compared to US Military Research and12



A. Orgaoiiational Aspects12

echnical Considerations13

C. Imph'catioQS forrturc15






top crcnn



This Estimate addresses the potentia] of Soviet military research and development. It first appraises the general magnitude and rate of growth of resources available for thisthemen, and money, and how efficiently these are used. It then assesses how effectively Soviet military research and development meets military requirements. It does not attempt to predict specific Soviet technological advances. This aspect of the problem isin part in the series of NIEs on tbe various components of the Soviet military forces.




USSR has long accorded high priority to) on military weapon systems and relatedtechnologies, including space programs. It has madein thc resources devoted tond hasrelatively satisfactory level of efficiency with which theused. Comparable results have not been achievedcivilian pursuits, but thc Soviet leadership now appears to begreater emphasis and attention.

resources, we have made estimates of whatare spending each year on theirrograms.recognize that such estimates cannot be compared, exceptwith estimates for similar expenditures in the US becauseproblems including different currencies, price structures,priorities, and strategic goals. Paragraphs IS tof the7resent our approach to the estimates, whichcomplex and independent methodologies, and the results thatThe results could understate or overstate the truea wide margin. Nonetheless, the two independently-derivedare broadly consistent; they indicate that duringin Soviet expenditures forlus space hasin support of the space effort. In this same periodrates of increaseacilities and manpowerrates of growth are now less than thatxpenditureswhole.1

is virtually impossible to measure the effectiveness ofR&D. Although the Soviets have demonstrated thcsolve advanced technical problems, we do not know whetherproducts reflect fully the original requirements fornot- We believe that the Soviets have established theirhich seems to emphasize the expeditiousof systems that willob simply and reliably.

1 Foi the view* ot Vice Adm. Vinccni P. de PoU. USN, Uie Director. Defense Agency: Mai. Ceo. Phillip B. Davidson, the Asm tint Chief of Staff foe Intelligence, Department ot" thc Aimyj Bear Adin. Eail F. Kectanui, die Director of Naval Intelligence. Department of die Navy; Assistant Chief of Staff. Intelligence. USAF, on e? Uma Ira of Soviet erpenditores for military PAD, see their footnote* tof Uie teat, and tonnex II,

expeditious approach is followedast RctDwhich tends toward conservatism. New ideas andsubjectariety of planning constraints and must benumerous levels and agencies. And the Soviets oftenredundancy of effort, judging that thc hedge against failuretlic greater expense involved.

foresee little change in the way the Soviets go aboutout their. The success that they have enjoyedwork against any major changes in procedures, at leastnear future. The various systems wc expect them to introducefuture will, for the most part, continue to representpresent systems through subsystems upgrading or theestablished developmental trends. In general, the Sovietsfavor this approach as contrasted with the search for radicallyuntried



I. The Soviel leadership views science and technology as fundamental to lhe growth of military power. Tbey also see it as tbe keytrong, modern economy andource of international prestige and influence.research andR&D)igh national priority. The highest priority Is given to military programs and some space programs, and to developments

' ;Ii and development aswhole Includes bote and applied research in science and engineering, and hi the design and development of prototypes andh-artcriginal (nvratiitalion for the advancement of icientific knowledge;taich Is directed toward discovery ol new iciewtjfic knowtedt* with spenafic product oejeetirw:ctively ooetcernod with problcon* cocnueUf cd in translating reaearch fiodeigsuki emeial acafn-ttfic knowledgr inlo specific prciducti The concept of MD as used In thb Kstimate also Include! lostlne and evaluation, and tliua equates to the US concept ol luseaich. development, teslinB. and evaluation (It, D.

which contribute to an image ofparity with, or superiority to, the US. Lesser, but growing, attention is paidelated to civilian pursuits.

Tbe Soviets are well aware of dieof basic research in oSe sciences and support it generouily. In many fields,arc free to pursue their work with few constraints.onsequence, great strides have been made in key scientific areas. Tlie USSR bu also worked oo building up industrial technology to supportoals lor the military and In space. But the Soviets are still weak in converting thc results of their basic research Into practical applications, especially in. In the past few years. Soviet leaders have attached increasing importance to closing ihis"

The Soviets have decreased, but by no means have eliminated, their dependence on foreign technology in such key areas as ia-

strumentation and computers, which constitute supporting technology important to botl)power and sustained economic growth. In these areas the Soviets continue to depend heavily on the acquisition of advancedequipment and technology. Development in these field*riority goal, but progress has been slow, specialty in non-military applications,ubstantial commitment of resources.

hc Soviets have established three basic policy goals in civilian HAD which reflect concern with the low level of efficiency in this area. These are: to apply nowmore rapidly to tho civilian economy; to improve the research environment with better equipmentore effectiveinforrnation system; and toetter nii'.ins of exercising control over the widely dispersedffort withoutinitiative. These goals, the subjecti of

various official government decrees, have been

realized onlymall scale.

hc extent and direction ofill be influenced by thc SALT agreements. Thc Soviets will certainlyto pursue an activeunder the agreements. Their willingness to limit the numbers of strategic weapons, while permitting qualitative improvement!,confidence that their HAD programs can meet competition with Ihe US. In some instances, resources for this purpose could bo shifted from additional deployment of present systems. For example, tbe Soviets piobably will continue to develop more and better reconnaissance"nationalensure verificaHoti of USwith thc SALT agreement It itto assume lhatn such improvements is already underway.


hc Importanceo tbe Soviet state Is indicated by the rapid growth ofmen, andhave boon poured into these activities,Intond space programs. The exact meaning of this effort is hard to define ln more than general terms. It isdifficult to relate this effort and itsof success to similar US efforts. It is not clear Just what activities are covered byfor what the Soviets refer to as "science" (see Annexoreover, neither tha pattern of engineering employment, nor the level of efficiency of tbe opera [ion is well understood. It is even more difficult to discuss these matters with rogard ton particular. In this section wc evaluate theInputs to Soviet. We later address tbe question of how effectively these resources are applied.

A. Research and Development Facilities

he number and size of facilities' fordevelopment, and testing of Soviet weapon systems and space hardware have grown rapidly in the postwar years. Numerous scientific research and design institutes,production plants, and major testing facilities wereexample, tho huge complexes at Tyuratam, Saiy Shagan. and Kapustin Yar. If the Sovietseed for additional factories toromising line of technological investigation, they undertook their construction with little hesitation.of facilities, however, has not always resulted in development of operationalThus, current and future construction of new facilities indicatesotential, and not necessarily production and deployment programs.


The resource base (or HAD in support of Ihe Soviet aircraft, missile, aod spacein terms of the area of roof cover of identifiedrapidly in the,ate ofoear, largely In support of new design programs. Since thot has continued to growear. From what little we know from the spotty data available, we believe that growth inacilities has been slower than that (or the aerospace industries.

B. Scientific ond Engineering1

he supply of highly trained andscientific and engineering manpower has grown sharply over tbe past two decades. During this time substantial shifts haveIn the occupational composition of this technical elile, reflecting shifts In demand since the, most notably for skills associated with military and. The average annual growth in numbers ofand natural scientists, which reached II toear during the. hasear since then. Overall, the Soviets0 hadcientific workers in science and cn-

' The Soviets do not pnhtUb data on the number of weaken actually engaged in either civilian orRAD andave not asvuialiad any such data from ctaaMfied sources. Tbe Soviets do publah.ever, many seta of iratiitra oowhich bear directly onverall BAD effort For example they regularly publish data on the number of Soviet worker! holding decrees in engineering andacientifie woikeis" and a* cm-ployed In "science and science lervfcw" Not all of thisngaged in worl in HAD. Many of these worken hold administrative, maintenance, or service positions and do not engage directlyRAD work See Anne* A, pageor additional details on this and other aspects of Soviet scientific and engineering manpower.

gmecting andillion engineers,enrollmenli in science andindicate that this manpower pool will continue to grow atear over the next five years toillion scientific workersillionhe. The advantage to the Soviets of this large body of trained manpower,in engineering, bar been neutralized Somewhat by the assignment of trainedto non-technical jobs.

e estimate theork force in tlse USSR has grown from aboutillion0 to closeillion0 (about one-fifth of these are sclcntisU and engineersollege and university levelhe rate of growth liasearrowth over theears could continue at about this rate, but more likely will decline slightly.

IL Tbe Soviets attract the top scientific and engineering graduates to positions within therganitaHoM whichcarry more prestige lhanrganizations. Workers in therganizations are provided superiorand olher facilities, better opportunities for advancement, aod housing preferences.

The quality of Soviet training ofand engineers Is as good as it is in tbc West in many scientific aad technological fields- One of the reasons may be earlyof students to the scientific anddisciplinesntroduced in tbe fourth grade and one-third of the secondary curriculum is devoted to science and matbe-rrntics.

oviet scientific and technical educnlion has shortcomings. Until very recently, Ihe aim was to develop special-

ized skills enabling the individual to perform efficiently only in specific areas, and tbe Soviet system has successfully fulfilled this aim. If has been less successful when tho need arises for adaptability. The narrow specialties in which many Soviet scientists and engineers are trained produce limitations on their ability to integrate contrirxitioiu from disciplines other than

o add to their pool of scientific and engineering manpower, tbe Soviets havepart-time advanced training, but notenalty in thc overall quality of training. Moreover, many part-time students arc from older age groups and consequently do not Iiave aseriod to use their acquired skills. In contrast to full-timegraduates of part-time programs arc not required to take assigned jobs. Anof tbe quality of theirhe fad that those who manage,art-time basis, to receive engineering degrees and who seek engineering positions are usually hired onlyast resort. Most part-time graduates remain in jobs requiring only technician skills.

C Outloys for Military Research and Development and Space Programs"

lie conventional way to nmrcsent the resources available fornd spacethe research institutes, test facilities, manpower, and otheris by means of tho total expenditures involved. There are, however, no data that permit this to be done for the USSReasonably straightforward fashion- Soviet financial data

'The derivation of tlte estimates of these cipeodi-tuica.iicussion of lhe inherent problem*aie presented in Anno B,

from official Soviet publications, and thebut incomplete irdormation on Soviet facilities and programs observed in satellite photography, must bo supplementedarge amount of Indirect data, subsidiarywtrapolations, and assumptions toan estimate. There is no way of confi-dently telling how much error is introduced by each step in the process or whether, and to what orient, thc errors offset one another or cumulate.

ven if an accurate estimate offor Sovietlus space programs could be derived, expressing It in terms permitting useful comparison withUS expenditures Is fraught with further problems. Thc US and tbe USSR havecurrertcies. economic priorities, price structures, institutional approaches, strategic goals, military tactics, and technicalmentionew areas ofAnd even if RAD expenditure estimates in the US and the USSR were expressedommon currency, tbe comparison could still only bo used in the most general fashionross measure of the relative effort. Moreover, an equal Input of money does not imply an equal military achievement or capability. Used with other information, however, estimates of expenditures arc helpful as onealbeit assessing thefforts and priorities of the US and the USSR.

e have madeon analysis of Soviet financialSoviet expenditures for militaryrogram) plus space programs in rubles. This analysis has been the principal basis lor our pastThe ruble estimates have boon con-verled to dollars by meansuble/dollar

(alio This ralio is based on available data on US and Soviet prices for complex, high lechnology equipment andough approximation ot* tbepurchasing power ol ihc ruble and the dollar in the area of luVD. The validity of tlie dollar estimates derived from Soviet financial dala depend* upon three factors: (a) the extent to which the coverage of "sciencu" expenditures conforms to Western definitions; (b) the accuracy of our distribution of the Soviet expendituresmilitary and civilian programs; and (c) the extent to which the ratio used to convert the rubles to dollars represents the average purchasing power of theesources. The rate of growth of expenditures in) isby the deflators used to convertunits into constant units, and therefore by any inaccuracies in Ihe deflators.uncertainty exists for each aspect of these calculations and it must be recognized that Ihe estimate could overstate or understate the Soviet nffortubstantial margin.

s another approach, we have estimated directly what theserograms might have cost In dollars if they liad been carried out In thoumber of studies which attempt to cost particular aspects of thcffort direcdy in dollars have been done In tho Intelligence Community, but no single study has presentedestimates of the cost in thc US of all Sovietlus space piuaysats. Far purpose* of this Estimate the work done in tbc Intelligence Community was collected and combined in order to develop for the first time this approach for estimating theof fhe Sovietlus space effort We call this method the direct-costing

method. The validity of thc direct-dollar costs depends upon the precision with which the Soviet programs have been identified, defined, and projected both for past and future years, and upon (lie accuracy and relevance of the direct-costing relationship! which have been applied. Again much uncertainty isnd-this approach could also overstate or understateubstantial margin the true magnitude of the Soviet programs.

f ilitany RAD Plus Space Eipendilures. The Tabic below compares thc estimates for Soviet expenditures onlus space derived from analysis of Soviet financial data with those developed by combining direct-cost estimates of Individual programs


0 Dollars)











. .


figures la this Table tie prrsenied as <DUtle catinuWa rounded to one decimal place for rearer* of na ol presentation, not because we haw uw eordldaac- Id then Imphed byestimates. Beeaoie of the manyoferror a* thesee hava no bast) foranp. The rradVl aOrMM* at directed So the per-taaea*tneehoSoL>|rfs and anurnpUoosthne fif-orea ara baaad.


hc est una teafrom Sovietdata arc roughlyimilarand grow at about the same rate as those based on the direct-costingexcept (or more recent years. There is no basis for considering ooe more accurate than the other, nor do the two series representlikely range. Nevertheless, the two scries are reasonably consistent despite the uncertainties ol the task and the completely differenton which they are based. Thisencourages live presumption that the general magnitudes which result are in accord with reality.'

ilitary RAD Ejpendituns Alone. These two approaches have been used to derive estimatesxpenditures directed at purely military goals. The rust subtracts the costs of Soviet civilian space programsnmanned lunar exploration andprobes) and military space programs already operationalatelliteand communications) from the estimates based on Soviet financial data; these space oosts hove been estimated by direct costing. The second calculates costs ofrograms directly in dollars.

Ada. Vans* P. de Pots. USN, the Defense Intelligence Agency, Maj. Gen. Phillip B. Davidaoo, tba Asaiatant Chief of Staff for Intelligence, Dapartnactf of tha Army,m. Earl F. Bactanus, the Director of Naval IntcOigeiicc. Department of tba Navy; and Maf Ceo. Georgeeegao.he Assistant Chief of Staff. Intelligence, USAF, do not believe Ihat the general conftttsncy of remits obtained from tba too metboootociei should encourage Use presumptioned above. They believe that neither iriediodology producei very credible re-sulti. but tbey haw considerably moce confidence in the direct-cost inn approach.urtherf tbii view are their footnote toB.)




(Law Space')



The figures in this Table are presented aa tingle estimates rounded to one decimal place for leasoia of eaie of presentation, not becausehave them them implied by such estimates. Because ot the many aoureea of poarlble eiror in thesewe have no basis fe* itaange. Tha reader's aftcntsna hi dmctad la ibeparagraph* olescribing the methodologies and asmmptlocii on which these figures are baaed.

'Does not include National Acroaanuca! and Space Admlnlilratioo-rype spacenor operaliooal Department of De-fsnae tjpraUo

f necessity, both series again appear roughly consistent, except in more recent years. The data imply that after tbe, Sovietspendituresvery little if at all. and that most of the growth in Soviet ItfrD came from the spaco programs. This conclusion is bestIn light of the fact that Sovietforere already high at the begmning of. The relatively slow rate of growth duringercent areflects

TOP cccncT

high levels ol acfivily throughout. Al lhe beginning ofote ballistic missiles, aerodynamic missiles, and military aircraft were under devcaooment dun at the end. Although aircraft arc now more advanced and more costly to develop, most of tho missiles now in test are modii(cations of rail-tins' ones, rather lhan new systems, and consequently are less costly to develop.

contrast. Soviet space programsibeir infancy at the beginning of (lieIhcheipacas launches and estimatedspace programs grew about nine-fold.that Soviel space programs0 cost the Soviets almostrograms.

believe tlie resultingpredominant growth element Inhas been the spaceaof actual dewJopmeests inand not the result of attributingto space programs and low costsones. In both tlie US and thesame plants and test facilitiesand space development Analogous.identical, methods have been usedcosts of tbe space and missileThey both, in turn, useto those developed to costThus we discern no reason lorupward bias in the estimates ofof Soviet space programs whentorrjgrams.

xpenditures for MiUtaiy. Tlie Ninth Five-Yearalls for expenditures for "science"orlus space) ofillion rubles, or an increase ofercent over tbe previous five-year period. This rcpicseiils an average an-

nual Increase for the period ofn current rubles. It Is not unusual for actual expenditures to deviate from thefigures, and it may well be that reported foreriod will fall below tbe planned levels. In any event,inflation will probably make real growth less thanercent. If past rates of inflation continue, the projected rate ofercent in current rubles would equate to aboutear in constant rubles.

lone may also Increase byear. There are,various considerations that could lead us to osiimate slower or faster rates of growth. Supporting an estimatelower rate of growth is the apparent predominancen recent growth and the fact that the expansion of facilities that carry out Sovietas slowed toear. On (he other hand, in the SALTSoviet leaders may wish to pushrograms more rapidly than in the pasL An appropriate upper limit may be established by thc following set of assumptions: If, (u)ill grow at lhe same rate as tho overall planned rate; (b) civilian space expenditures will not grow; and (c) thegrowth will be for; then an arithmetic calculation indicates Ihatould growateour. Theselowear,ilghe viewedeasonable assumed range for future possibilities. Under these assumptions, Soviet cx[ictid it tires foray grow fo the amounts as indicated below:

3 S

Soviet Financialt-llCoitina 8 3



stimates of the efficiency with which tbe resource inputs into Sovietnd space programs are utilizer" have aon our evaluation of these programs. Meaningful measures of efficiencyctivities, however, are difficult to develop even fn Ihe US where information is more complete. It Is especially hazardous tothc efficiencyn differentwhich have different priorities, price structures, and practices. But some insights can be obtained by comparingeneral way Sovietithn tho USSR and withn the US, and relating these comparisons toaspects of the Sovietstablishment.

Military Compared to thein Iho USSR

ear split level in efficiency.s more efficient than civilian for several reasons. Il Isin that die customer is able lo present his requirements tostablishment and to participate in the development,and production process through military-technical committees-lsohas had first claim on resources,tho brightest scientists and engineers and the highest quality material andIts overriding priority has reduced the supply problems that plague. For caample. the military sector absorbs most of the available computer capacity: numerous Soviet scientists wonting on civilian projects have complained about the handicap ofaccess to computers- Workerslso receive higher wages because of extensive bonus supplements.

the Iwo basic characleristics thatmilitary fromand higherformer isimportant. For example, when highhas been assigned to selectedsuch as computers, tlie increasedof funds and materials has notgood results quickly.and bureaucratic barriers toand communication, prevalent inoften have frustrated the effort.andiffer significantlydegree of interaction between tbeand iheir the Ministry of Defense In theprocess Is very great, while inresearchers, designers, andusually physically andThere Is no evidence thathave been made reoeiitly in theand planning ofust find their general approach tosatisfactory.

The Soviet leadership has beento makeore efficient and responsive to national needs, especially in areas which will contribute to economic growth. The Soviets in recent years have grantedanagers greaterhave forced them to becomeaccountable, have adopted moraetaryschemes, and have tried to bridge the gap between research and production by establishing research organisations at aof key industrialirom.

Prospects for makingore efficient arc not outstanding; It Is unlikely that there will be much "spinoff" from thc classified military programs, or that resources will be shifted fiom. where they

probably can be utilined much morethan in. Nevertheless, pressures from the leadership forin efficiency in production in general may loud to the allocationreater share of resources, to, hut probably only if the total resources allocatedontinue lo grow.

C. Soviet Compared to US Military Research and Development

t is difficult to compare (be relative efficiency of Soviet and US. Part of this difficultyesult of tliemix of resources used in the two countries. In tbe USSR manpower Is relatively plentiful and capital equipment relatively scarceto the US. The Soviets therefore use more lower priced manpower and less higher priced equipment to do tbc same fobs. Onedministrator has saidorker produces twice as much as his Soviet counterpart. Our understanding is that this difference initsdue partly to tbe much greater array of equipment at the disposal of the US worker. But these judgmentsn general do not tell us much aboutn particular. Wc have no data thata direct measurement of the relativeof Soviet and USn using resource inputs."

uble/dollar ratio for the output of Soviet military RAD and spare programs wouldpecific relationship between tbe eflirtrnciea of USSoviet programs il compareduble/dollar ratio for inputs to these pioxrams. Ruble/dollar ratios fully tr ft tentative of the relationship between the ruble and the dollar prices for the outputs of. and inputs tn. Soviet program wuuld belder to attempt to measure icltDve efficiency this way. If more information heroines available io the future, suchmeasure uiulJ be undertaken.


hile l! is possible to make somestatements about Use efficiencyn relation to Its civiliani| Is difficult, if not impossible, to(he effectiveness of Sovietbe ability to solve the technical problems posed toany absolute or objective standard. In order to do so we would have to assume that end products reflect fully (he requirements levied onrocess and that our estimates of tbe performance of those end products are essentially correct. Neither of these assumptions can bo demonstrated with confidence. What It is feasible to do, however, is to set forth what we know about howctivities are managed and controlled within the USSR andraw some inferences from this base.

A. Organizationalctivities todayair degree of latitude for cross fertilization and cooperation between their variousThis is unlike tbe situation thatunder Stalin, when the best scientific minds were often forced to carry out their rcsearch under conditions ofn emnd close scrutiny by state securitythat restricted thc free exchange of ideas and concepts that is desirable for rapidand technical advances. Nevertheless,must still work within tho confinesast and cumbersome bureaucracy which tends toward conservatism, and they still do not enjoy the freedom of exchange thatthe US scientific community

'See Annes C, pageurther discussion of thc organon ol Soviet military

n lhc US, lhc development indof advanced weapon systems arcmainly by private industry under coo-tract to fioverrunent agencies In the Soviet economy, however, there is no privateirutitutions havingunction are operated directly by one or another agency of the govcsTimentontrolled by the Pohtburo which receives advice on policy planning involvingfrom tbc staffs of the Centralthe Council of Ministers, and tho Defense Coundl.ecisionroject has been made, the project is then turned over to th* Mihtarv-Ir>doxtrta](VPK) of Ihe Coundl of Ministers for implernentation ond supervision. Prom thai point on, the VPK acts as the controlling peakyramid of organizations, each having some clement of the project to implement or on which to provide advice and guidance.

hc bureaucratic relationship! overseen by tho VPK were already working before its establishment. Sovietn any one type of weaponas fighterlongigh degree nf ongoing reoperation belween research and design, production, and military consumer. Representatives of design bureaus, production plants, and service users constitute ad hoc teams to followrogram fromlo deployment. But this degree of multiple involvement also has its drawbacks. Any RAD project is subject to influenceariety of organizations, each of which has its own ideas about what needs doing and how il should be done Moreover, suchrequires standardization, and any new ideas or concepts must be justified through numerous echelons and agencies and are subjectariety of technical,nnd planning constraints and biases.

B. Technical Considerations *

oviet basic research In fieldslo military developments probably is about equal in quality and scope to that of the West. Theoretical work it generallyFor example, Soviet theoretical work io aerodynamics js among Ihe best In tbe world. Puil her, thc Soviets have looked for theoretical solutions which would permit simple hardware designs using provenThis approach seldom loads toperformances, but it does speedand ease manufacturing, operational, and maintenance problems. And the Soviets have, where possible, taken advantage ofin other countries to further their knowledge or lo by-pass certain steps in their own developmental processes. In developing their missilery, they iriitiaDy relied heavily on captured German equipment andas well as on the talents of manyscientists and engineers whom they took prisoner during World War II. In the early days of nuclear weaponry, they clandestinely obtained much data on US and other Western lechnirjuea. But the Soviets have alsoown approaches and high technical competenoB. and their presentn this field Is,istinctly native produd. In some fields they continue to rely heavilyubstantial effort for theof foreign technology.

n developing their own approaches to. the Soviets have generallyredisposition to keep theof production and use in mind, totlie essentiab, and tu devdop that system which will most simply and reliably domrcd fob. Their carry missile guid-

Armet D. pageucuuxxi ofical aehkvrnwnli and coasldaatiimi In ipocific> al areas.

ance took advantage o( their wii ihmlfj|Hll rocket-engine technology and avoided Uio need for lightweight digital Qicnputcn which were thenrimitive state of development. Theiras designed topecific fob of interception and to be easy tomaintain, and fly. Once having estabesign approach, tbcy hare thento make many incremental changes, ratherew quantum Jumps. In this thoy prefer to use standardized, proven, off-the-shelf components rather than make mafor changes in subsystems with each new weapon system. They effectively use conventionaland proven tocfaniques to give military products the desired service life, and tothe time needed to develop them.

Their predisposition to put outunsophisticated models early in the research effort, and then to upgrade them through subsequent model changes, permits the Soviets to field an operational model early, and to make iinprovementsesult ol con-tinued field experience and improvements in technology. Early surface-to-air missiles (SAMs) have been steadily improved bymodifications, the mission flexibility of theas improved through Ihe held ing of overodifications: and the early Soviet nuclear submarines apparently metspecifications only yean after Ibeir initial operations. Tho succession of Incremental changes permits tlie most recently deployed weapons to be relatively up to date, but It also sometimes gives theumber of older models which do not represent theof technology.

In several instances where they have departed from (his philosophy, they have run into trouble. In the ease of thepace booster thoy encountered numerous problems which resultedery poor record ofduring the first Ihree years of llight test

ing. Their largest space Ixioster,ehicle, has yet to be successfully flown. Their solid-propel lant intercontinental ballistic missilehcis having problems. These, and other troublesomemay have tempered their willingness to take bold strides beyond proven concepts or techniques. We know of no deployed weapon system which fuDy used what we estimated was the state-of-the-art In their technology at tho time of Initial design.

ot to say that they have never been successful In exploring new nativeThey have, for example, shownin their approach to the devrJoprncnt of antiballrstic missile and antisateUire systems. Tbe areas where thc Soviets are likely to be venturesome are those where they believe the US. or some other Westernaking advances that could slgnificantiy affect the Soviet strategic position. There arc. forindications that the Soviets have begun HAD on laser beam weapons. Such work may have been undertaken in the belief that the US badimilar program. Under such circurnstances, we believe that Sovietact swiftly toince they cannot afford to do cthcrwbe.

ne characteristic of the Sovietoach tohich usually resultseliable productelatively short time is that of competition between two or more design teams. This aspect has been mostin fhe development of aircraft andIt appears thatade loew weapon system, the general specifications are handed to several designers wiih the ur>derstanding that final acceptance will depend on thc evaluation of the product. Within the general guidelines laid down, each design team then examines the technology necessary to meet tbo requirement and pro-

top scene r-

to develop an experimental prototype. It probably is at thii point (hat the individual designs are era mined. Two or more arc then chosen for further work, often with instnic-tionsentral design bureau towhatever features wereesign that was being dropped

he process generally continues until tlie systems have reached the flight test stage and then the decision is made as to which to put into series pioduction. In some in-stances one design clearly is favored and the otber iMOJocts are abandoned. More often, the refected system undergoes limited deployment Regarding this competitive approach In the development of aircraft, five prototypes ofnterceptors were flown inMigs and two Sukhoy types. One of the Migs was selected for use in Frontal Aviation as was one of thc Sukhoy types. Another of the Sukhoy aircraft was selected for use in air defense. In thc case of missiles, thendere developed concurrently as were tliend thcn each cue the former was widely deployed while the other was deployed only in small numbers, or not at all. Thcndere probably also in competitionne anotherevelopmental program aimedmall ICBM suitable for widespread deployment. One system, theollowed tbe off-the-shelf approach while thc other, thenvolved tlie new solid-propellant technology. Thebviously won but, even so, dieas deployed in limited numbers.

the competitive approachreliable, useful systems, it mustcostly, especially when even theis rewardedodicum ofThe Soviets apparently feel,the rewards gained by thb approachtho espenso Involved.

C Implications for the Future

some majorstablishment and itswe consider unlikely, we foreseein tlie way tlie Soviets go aboutout their. The successhave enjoyed inuccessionweapon systems will probablyany major changes in theleast in the near future Thec oJtpect them to(he future (discussed in detail in tbeNIEs on the Soviet militaryfor the most part, representon present systems throughor continuation of establishedtrends, rather lhan throughof radically new concepts.approach in weapon systemwill almost certainly continue,reliance on proven techniquesBold innovative programswill only be undertaken when thegroat enough to justify tho risk*even then the chances of success willas great as those taken under theircompetitive approach.

top crxnn



I. Tho Soviets do not publish data on thc number o( workers actually engaged in cither civilian or miliiary research and developmentnd we have not acquired any such data from classified sources. The Soviets do publish, however, many sets of statistics on manpower which bear directly on their overall RAD effort. For example, they provide data annually on tbe number of Soviets graduating in engineering and on persons classified as "scientificnd on the work force"science and sciencee believe this information to bo reasonably reliable and therefore use Itasis for making our estimates on the level and rate of growth of this manpower. This Annex presents these estimates. We also provide some US-Soviet comparisons in these fields and. finally, we briefly examine thr* Soviet educational system which supplies this manpower.

B. Scientific Manpower

Thc Soviets annually publish extensive data on the total number of what they call "scientific workers" in the USSR. They break down this information by field ofby geographical location, and even byorigin. In providing these data theuse the term "scientific workers" which is much different In concept than thc category known in the US ashe Sovietsthis term as including all persons holding graduate degrees from universities in science and alt persons engaged in research regardless of their educational background. Their defini-

ncompasses such fields as research on law and art which are not classified asin the US.

s used in the context of this paper draws manpower primarily from the fields of natural science aod what the Soviets callsciences" (dbtinct from engineers, as such) and we therefore included thesein discussing Soviet "scientificn tlse decade ofhis segment trebled (sea,hile the total civilian labor force expanded byonger period (about tbe pastears) the rate of growth of this segment rose markedly during thend, followedecline in tbc rate of growth in tlie latter part of that decade:

Thou si ii >i> of Workers

A vuuee Annual Cnowm or Snjeno CATXConns or SCuueririC Wonxxjis

IWningc nut


. 7J)

complete Soviet definition for "sesenBtsp iiWo" reads as follows: (a) acadecoictara who are fuB or cc*rcsr>oadau aaeanbers of an academy of science; (b) all peraaiu who have an academic derree of doctor oc candidate of science, or an academic tide of professor, lecturer, senior or junior

rewirch c: Jr: i.v

or character of work- (c) persona conducting ecien-tifie research work and scientific pedagogical work Id higher cdut-alioiu! Institutions, regardless of degree or academic tide, and (d) oevor Specialists who do not have an advanced degree or academic title but are doing reheard work in industiial enterprises, nnd design organiiattom


and Mathematics

Vatarinary S.

and Pharmaceutical S<


autistic* up0 vera obtained Irom official Soriel publications. Standard statistical method* sere used to establish an average rate ot powlb uphia average rale was then projected

only* these engineers classined as "scientificersona wiih graduate degrees in engineering, and engineers irrespective olworking in scientific research organitatiom. Thoy comprise aboulercent of ell personi classified by tbe USSR ea engineers {seeI,.

pattern of growth reflects the step-up in the Soviet strategic nuclear missile and space effort resulting from decisions probably made in the. Tho number of workers in thc fields of physics, mathesriatics. andessential to militarygrown about eight-fold0 {during this period the total civilian labor forceby onlyhis total lias grown only by someercent,during this recent period the share of those with advanced degrees has increased.

C. Engineering Manpower

ngineers are actually engaged In tlie direct performance of basic and appliedand in tbc development and control of manufacturing technology. They design and build weapon and system prototypes, and maintain facilities and instrumentation, Tlic term "engineer" as used by the Soviets, and therefore as used in this report, denotes a

person who hasiplomaiigher educationalWhen compared roughly with US data, however, official Soviet figures which report the employment of such persons overstate substantially the number of persons actually employed as engineers. Wo base thison an official past Soviethich indicated that as many as one-half of the persons withdegrees worked in managerial, administrative, or other essentially non -engineering occupations (For example, Soviet Party leader Leonid Brezhnevegree in metallurgical engineering and is therefore counted as an onguiecr inand administrative institutions.}certain Soviet categories of engineeringartography, architocture. andarc not counted as such In the US. Nevertheless, we brieve that the published

"The eentiHi ofeennu which hat Included these dala.


dataeasonably reliableol trends and areas of emphasis in allocating engendering manpower. These data show that the total number of employedfor certain years1 was as follows:




We have projected these figures6 by major engineering specialty on the basis of current university enrollments.I presents this projection for selected yean.

he patterns of growth of Sovietis similar to the growth of selected categories of "scientifice, raped growth0 with the greatest increase inriod, followed by ain the:

AftOwrn or Sovar

Peiceniaj'e Bate of Engineer!

latest year for which official Soviet data are available.


7 .


Geology ProapoeUoa

MiaeraJa KiplonUoo

Power Eneineennr.


Machine Building and Instrument Con-

Electronics, Electrical Equipment Con-

and Automation

Radio and Communication*



Other Specialncluding Consumer Producla, Geodeey. Meteorology, etc.



" ** D^DCtl" l> published annually. Ercept for the year.


This figure include* both machine building and electronic, opeclallies.

wor'kW inyear is caloeorised aa "acienUfic

in scientific reaearch and teaching aeganlsatiaaa.

The most rapid growth was in thc specialties on whl. Ij military HAD is heavily dependent.

D. Em ployn Research and Development

umbers of sctantiits and engineers,do not equate to rrmploymont. Many of thc scientific workers are employed as teachers, and are therefore not employed in RAD. MosI of the engineers are employed in production, construction, transportation, and otherositions. Moreover, most employeeso not have degrees io science or engineering.

n approximation of total full-time equivalent Soviet employmentthe social sciences and humanities) can be derived from reported Sovietin "science and sciencehis employment is defined by the Soviets asoi administrators, professional staff, and non-profesiional staff in scientific research cstablirlirncnts; in design, experimental, and testing organizations; and in surveying,and hydro meteorological work. By subtracting reported numbers of survey, and hydromcteorological service workers and adding estimates of numbers of scientific workers employedt higher educational and Industrial enterprises, weat an approximation of numbersersonnel in universities, enterprises, and at research, design, experimental, and test or-Sanitations. These numbers are set forth in,

mployments estimated to have grown from aboutillion00 (some one-fifth of these are scientists and enginccriollege orlevelhe rate of growth hasear since the mid-

, compared to an averageercent for scientific workers in scienceercent for engineers. If. as Isthe number of scientific workers and engineers growsateear through the, employmentill probably grow no faster, and, most likely, will grow more slowly,

E. Some US-Soviet Comparisons

Because of conceptual differences in the statistical reporting of the two countries. It is difficult to make valid comparisons involving US and Soviet science and engineeringThe narrower concepts of "scientist" and "engineer" as used in the US generally will result in an overstatement of Soviet total numbers In each of these areas when direct comparisons arc attempted. Keeping thisInomparison of the trend and sire of scientific and enguscering manpower nevertheless does provide an insight intostrengths and relative priorities.

We can estimate with some confidence tbe total number of personnel holding college degrees and working the equivalent of full timen USSR. We compare thb total with comparable US data in thelow. As can be seen, growth in the USSR has boonore rapid pace than in the US. But wc








I860 IM1 2 3 IMS 6 7 - 9 0

3S8 67







in Research IiUUUiU-Denagnerim*nUl I'lanU, and Teal al i

Employed In Universities

in Induitrial EuUr-


Wiih College Education


Employment in "i

leporled)employment la geologicninauttaies,plants, and leal eataMiahnnu.

Number, ef aoeuUia reaearth inatiUalee (estimate) io gro.0) are .abUacUd teem.Wsdenle engagedl research .aMiluU* (climated to growOO

|workers"nivemUea (repotted) leu aoeial- at unlvomiiea deported total aociall research inautuiee) giro. total numbers ol natural kmiiU.u, miliiary aoientiita, and engineere at

' ToUl employmentt Indiulria]aeumed u> b. ihree times ihe number oforker..

rig eeieolific -oekerr. atiiutea (reported)ni.emtio,

- TalaluraJ Ud miliury ae.entUu aad engine, employed by HAD enUrprieee, urdetraitiea, aad iaduetrial aslarpruea. plus gradual* .tudenU workiagl research imUI.Ua

dvanced degnmia

eUbluhroenU (reported) represent, ibe abate lot all such aoraluU aad

mindfulmaller share of the totallircctly engaged in research than in thc US. Wc also believe thatersonnel are leas efficient than are similarrofessionalarger number of support workers, while his US equivalent has more and better equipment. In view of these differences, we believe that the two series of figures cannot be used to make evaluative iudgmcnts regarding theof thc two countries.

here were considerably more scientists in the US than there were scientific

workers in tlieespite the superior Soviet pace of growth of natural scientific

workers (see purngraph. the gap between tho two countries has not narrowed grcady.0 the number of US natural scientists in the military-orientedphysics, math, andthat

"In ihe US. rng'neen are aB person* mttmaBw aw gagedngUManag workevel requiring knowl-edge of engineering, physical, life, or mathematical sciences equivalent to lhat acquired'-iufynar college course In one et there fields.are penoos wiih such training but euiienllr employed In poiiUoiu thai do not lerpjlre It.

top scene*

tho Soviet Union by lomo four timra. Tills margin has since been cut in half probably because of changing US national interestsreducedupport lorand space.

F. Training of Soviet Scientists and Engineers

n the Sovietigh premium is placed upon technical and specialized, rather than liberal arts education. Science andarc considered to be the foundation of national strength in modem times andthey receive unique emphasis at all levels of schooling. Secondary schoolingthe base for early (and mandatory)to the sciences trom which selected individuals arc chosen for professionalMoreover, the quality of professional training in scientific, engineering, and applied fields probably is ccmparable substantively to that offered in the West. One reason is the early exposure to scientific and technicalPhysics is introduced hi fourth grade and one-third of the secondary curriculum bto science and mathematics.

he USSR hasniversities, where most scientists arc educated, aodechnical institutes which provide much of the specialized engineering manpower. There isarge part-lime educational training program that provides higher educationadvanced degrees) in science and technology. The rate of increase in theIn higher educational institutions has slowed somewhat in recent years, but anIncrease in enrollment0 Is forecast by Soviet leaders.remier Kusygin predicted an increase fromillion toillion studentsctual enrollment for the academic

illion Indicating that the Soviets were well on thc way toward that goal. The official target0tiulioo.

on the training ofindustrys reflected in theand absolute increase inenrollment in engineeringchooltotal5 million studentsnrolled in higher educationin engineering fields of study.illionercent) ofenrollmentillion were inOver the same period,the natural science fields trebled.

scientific and technical educationUSSR has several shortcomings. Untilthc aim was to developskills, enabling the individualefficiently in specific areas. Totlse Soviet system hasuccess.the need is for adaptability ofthen the success of their system IsThe narrow specialties ioscientists and engineers are trainedsome limitations on their ability tocontributions from disciplinestheir own.

lie quality of Soviet academic degrees since thes comparable to that in the West. Fust-degree (diploma) holders from university and other prestige institutions arc somewhat ahead of holders of USdegrees. In some casos they are comparableS masters degree. The Soviet "kandidat" degree it roughly equivalent to the US doctor of philosophy. Tbe highest Soviet degree,as no US equivalent.

nlatest year on which data arehigher schools gradu-


tudents in (Masin Ihe USe engmcering. or nearly fivo times as many engineers as US colleges. The amount was up from three tunes as manyn the natural sdences, tne US hasonsistent and wide lead gradu-

ating0 well over twice as manyas theomparedverall,0 the USSB heldwo-to-one edge over the US in natural science nnd cngmecring graduates combined.





here ue serious conceptual andproblems in trying to measure resources used for research and development) within any large and complex nationaland special problems for the analyst who tries to collect and interpret Soviet data. Even in the US. there is no clear line of demarcationnd the preparatory stages of series production, nor between innovativethat results in the development of new products or processes, and production design directed at providing rnodified products or processes. Conceptually,annot always be clearly differentiated from.unding is defined in the US as outlays by the Department of Defense (DOD)f.iliii, development, testing, and evaluationilitary spaceexpenditures by the Atomic Energy Commission (AEC) for research directly related to militaryThis definition is buieaucraticaliybut sotnewhat deficient in conceptualbecause itin die private sector which may Iiave important implications for military and space capabilities.

problems are compounded inSoviet. The data areand Soviet accounting practices andcannot be matched with those usedUS. It is especially difficult tomilitary anddetermine |ust where the Soviets choose

to draw the line between development and production. The conventional way totbe resources availablend spacethe research institutes, test facilities, manpower, and otherby means of the total eapenditures involved. There are, however, no data that permit this to bo done for the USSReasonably straightforward fashion.

ven if an accurate estimate offor Sovietlus spacecould be derived, expressing it in terms permitting useful comparison with similar US expenditures is fraught with further problems. The US and the USSR have differenteconomic priorities, price structures, institutional approaches, strategic goals,tactics, andew areas of difference. And evenxpenditure estirnates In the US and the USSR were expressedommonthe comparison could itill ony be used in Ihc most general fashionross measure of the relative effort. Moreover, an equal input of money does not Imply an equalachievement or capability. Used with otltcr information, however, estirnates ofare helpful as oneassessing thefforts and priorities of the US and the USSR.

I wo basic approaches to estimating the cost of Sovietave beenhis NIE. Each is useful in its own way. but each has its own limitations. One deals with an analysis of published Soviet

primarilylive olher involves the direct costing of observed and estimated Soviet programs and facilities. The first starts with thc derivation of Sovietlus space expenditures (expressed in rubles) from published Soviet financial data. The second applies US costs to known, estimated, and projected Soviet militaiy BAD programs. This Annex discusses the complex problems encountered in these two approaches.


S. Analysis of the Soviet budget and other financial data has provided some insight into the total Soviet BAD effort, and that portion which is related to defense and spacebut it does not supply the detailtoigure for military BAD alone. The Soviets tlsemselves say almost nothing specific about their military PAD or even about thc total amounts spent for it. They do publish considerable generalincluding expenditureBADhole, and these data have been used toapproximations of thc portion devoted tond space. So lar, however, Soviet financial data has yielded noabout resources devoted to particularor missions.

outct Expenditure Data. The USSRtittle information regarding thc concepts and methodology employed in thc collection and presentation of their expenditureWe believe that the Soviets report their financial supportn terms offorlthough there arcinstitutional and proceduraltbe evidence indicates that thc Soviet concept of "science" is compatible withdefinitions. Tlie Soviets Itaveexpenditure data to tho United Nations (UNESCO) fnr inclusion io studies.

These data have licen consistent with data published within thc USSR under the rubrichey have used such Western terms as "basicappliednd "development" when discussing their science effort. They have shown themselves to be familiar with thc expenditure figuresn the US published by the National Science Foundation, and have discussed their "science" expendituresimilar context.

he Soviets provide three importantfigures for"science" as they call it. Hie most Inclusive of these figures is described as "Expenditures for Science from thc State Budget and Othere believe tint this figure includes aD important outlays for science in (he USSR, and we use it as the control totalur overall budget analysis. It Is referred to hereafter in this Annex simply as total funds. The Soviets issue two other official expenditure figures which they identify as components of total funds. One figure is the allocation labeled "Science" in the Social-Cultural category of theState Budget. We refer to thisas science budge* funds. Tbe third figure reported in Soviet statisticsecord of expenditures for capital investment for science which, under Soviet accountingIs grouped with other capital Investments rather than included under "Science" in the budget" When the science budget funds and capital investment funds are subtracted from total funds the result Is another component, which we label other funds. The source and applications of thc other funds are onlyexplained in the Soviet literature.he distribution of total funds for science among these three components was relatively stable. Science budget funds represented about

" Although Iheie sic some funds for capitalincluded in the science budget funds, these are in addition to those in capital investment funds.

top GrxnoT

percent, capital investmentnd other funds aboutercent of total fundi Thiswitheconomic data and subsidiary judgments on Soviet scientificthe framework (or further analysis leading toof expenditures forhis space programs.

Derivation of Military Research and Development Plus Spoco Expenditures

Science Budget Funds. This account is composed of two subaccounts, tho AlbTJnion budget and the budgets of the UnionThe AHUnion portion typicallyfor aboutercent of the science budget funds. According to the Soviets, it supports workheoretical nature" and "other works of national importance" The budgets of the Union Republics support thelr subordinatc academies of science and other scientific organizations that are concernedwith BAD Of local interest. We believe that Republic budgets are not presently being applied to military or space activity.

The Soviets do notreakdown of thoir science budget funds among. military RDT&E, and their space pro-grain.owever, they didetailed resourceexample, wages and salaries, instruments, books,of expenditures foreriod under the AILUnion and the Union Republicsbudgets. In the case of the Union"science" budgets are devoted to localexpendilures aboutI the announced totals,

Tlie sum of the detailed costs did not, however, equal the announced totals in all thc Ail-Union budgets. It leftarge annual residual that grew steadily over7 period Analysis of the official

ciitenditure data and pertinent economic bt-eraturc suggests that this residua! is the rnasor source of funds for those activities considered sensitive by thcmilitary RDT&E, nuclear, and the space program. Soviel publications, for example, havethe All-Union budget allocation for "science"ource of funds tor "work of nationalnd they associateexpenditures with ballistic missileand tbe space program.

Thc most recent budget handbooks, published2ave not repeated the detailed breakdown (or science budgetpossiblyesult ofby Western scholars In unclassifiedon the possible military rignificance of tbe residuals- Although the means ofdetails for the Ail-Union account are no longer available, the trend of thc7 can be used (or estimating the breakdown of the budget expenditures for science in

he unexplained residual amounted toercent of science budget funds from all the All-Union budget7 it had increased three and one-half times and accounted foreicent of the All-Union budget. Also,7 almosterceni of the annual Increase in science budget funds from all the All-Union budget was being allocated lo this residual. Although there were no signslowdown in the, we Judge that the dramatic growth of thc residual could not have continued much longer. We have

For Ihc views of Vice Adm. Vincent P. dethe iHrector, Defense Intelligence Agency;Phillip B. Davidson. Ihe AssistantfIntelligence. Department of the Amy; Rearf. Iwctaaus. the Director of Naval Navs Otm Btsmm

Kees-eo,he Assistanti Staff.ISAK. see their lootnote lo paragraph 6S,

accordingly projected die share of lhe All-Union budget aocounied for by the residua) to grow atdeclining rateaximum ofercent4 and to remain at that levelc believe that almost all, if not alb of this large portion of the All-Union component of science budget funds is applied to Sovietnd space programs and that little or none of the remainingis so applied.

apital investment. The Soviets give no Indication of the amount of capital investment funds used to construct or equip

* Evenercent of the leisac* budget funds trom the All-Union budget were allocated tond space programs, the estimated expenditures for these pros-rains presented0 in TaMe below, would be increased by less tlianercent

or space facilities. Presumably, however, capi-tal investment supports alictivities financed by the operating expenditures in-cludcd in science budget funds and ether funds. Therefore, capital investment has been apportioned betweennd space programs on the one hand, andrograms on the other, to conform with the civilian-military split estimated for operating expenditures covered by these two funds.

lAer Funds. There are few reliable data about the source or application of these funds. Other financial accounts that thecall "enterprise funds" and "university research funds" are the source of about one-half of other funds, and these are the only sources of such funds that can be identified. Industrial enterprises and other commercial



(Billion Current Ruble*)

Budget Funds

Military Plus


Military Plus


Miliiary Pluj




Military Plus



lus Space (Con-



Kxeept (or bottom two line*.

figures in this Tabic are presented as single values rounded to one decimal place for reasons of ease of presentation, not because wo have Uie confidence in them implied by such values. Because of the many source* of possible error in these figures, wn have no good basin lorange. The reader's attention in directed to lhe pertinent paragraphs of this Annex describing the methodologies and assumptions on which these figures are hosed.


organizations normallymallor their revenues to. The enterprise mayontract directlyesearch institute, or it may pay thc funds to ils ministry and Die ministry may then support research which Ii iisefcJ to tho whole industry.

IS. The amount deducted from enterprise revenue varies considerably among different ministries or production organizations. It Is difficult to estimate tho amount that may be deducted from revenues of enterprisesin military production and used to finance military. On thc one hand, if the practice In the defense-related industrial sector were thc same as in all other Soviet industry, then as much asercent of these enterprise funds might be used for military purposes. On Ihe otherood argument can be made that because of the strong, centralized administration and support for. the practice of generating funds fort the enterprise level is more limited. What evidence we havethat deductions for military industries arc, in fuel. less than for dvilian.

s for tlie rest of the other fundswc have little specific Soviet data oo sources or applications. The Soviet Ministry of Defenseomplex organizationalto support, coordinate, andrograms. All tha military icrvloes have institutes which conduct research tonew or improved weapons systemand to establish system requirements Thc services operate teat facilities to conduct suitability tests at tho end of development progiams as part of tho process of accepting new system? for scries production andMilitary representatives hold important positions on industrial commissions formed to evaluate design proposals. They areassigned to the administrative api>ara-

tus of industrial ministries and to research and design institutes involved in theof weapon systems. The personnel of the Ministry of Defense working on these andatters within or outside thestry are probably financed from tho Defense Hud got rather than from science budget funds. It seems .likely that these costs are thenin total funds, where they would show up In other funds. It is also possible thatoperational military space programs such as satellite reconnaissance, which carrybe funded from (he Defense Budget rather than from science budget funds and their costsin the other funds portion of total funds.

aximum, one could reason thatercent of tbe other funds coming from enterprise funds and alt, or virtually all, of the rest of olher funds support activity related tor space. If this wore correct, about DO percent of the total of Other funds should be counted as expenditures fornd space On the other hand, it is also reasonable to argue that rhe portion offunds applied tor spacehat no more thanercent of the balance of other funds should be similarly assigned. On tliis basis aboutercent of the total of other funds would be forlus space. The data doore precise or confident distribution of the total of other funds. We have chosen the approximateof this range and Includedhird of other funds in our estimate forlus space. Admittedly, this ii andistribution, but even If the proper share wore as high asercent or as low asercent, the totalstimate for any particular year9 would vary only by aboutercent in cither duectJoik.

9owever, the pattern ol science funding changes, and judgments about the application of olher funds have more effect on tlve trend and magnitude of Ihelus space estimates.

Funding Changes

Soviet spendingnd space bad reached high levels by the, and most Western analysts expected that both the growth rate and absolute Increasesould decline substantially.lo these expectations, however, theannounced increases in total fundsillionercent)illion (II percent)illionercent)

At the same time the pattern of funding changed significantly. Science budget funds, which had accounted fee someercent of total spending, increasedairly moderate rate ofercent per year. Actual expenditures oo capitalwhich had represented only abouter coot of totalnot available, but plan figures and statements by the Soviets indicate that no unusual increases were planned for this component. Other funds, in contrast, grew dramatically byerceni9n another significant departure from the stable pattern of earliernd wage costs failed to keep pace with tbe growth of total funds. Estirnated total employment inby onlyercent9ercentage costs, which normally account for aboul half of total funds, took less thanercent of0 increases.

Several alternative explanations, not all mutually exclusive, havo been offered tofor the puzrluig growth in other funds9a) the increase is due to a

broader definitionnvolving lhelo the "science" account of certainwliich previously had been carried in some other category; (b) existing science accounts may have been inadomtately covered or incompletely reported by thc Soviets in the past, and the new figures reflect aof the deficiency; and (c) certain aspects of Sovietexample, civilian industrialbe uscreaslrsg very rapidly.

None of these explanations can be either confirmed or completely eliminated.Soviet data often show someand unexplained changes. In tillsIiowever. it seems unlikely thatelocation of existing accounts took place. In the past there have been redefinitions of thc science categories, but these wereby changes in the historical data aa well. For example,hen costs for museums, libraries, and science exhibits were transferred from the science account to an education account, this adjustment wasin figures for earlier years. Abo,9hen new sources of funds were added lo total funds, figures for prior years were adjusted.

It Is possible that the Soviet Central Statistical Administration is measuring more accurately and completely areasthat it has slighted in the past For example. Soviet literature Indicates that most oferformed by industrialwas not previously included in thefigures. Therefore, in9 thcStatistical Administration undertook an investigation to quantify the cost ofbeing carried out by industrial plantsi need from production funds, rather than science budget funds. Tho addition of these previously unreported costs couldfor some of the increase9 and


Became lhe activity wai performed by industrial production personnel, we would notomparable increase in the growth ol identifier manpower and wage costs, and this did nut occur.

t is difficult to explain the largein olher fund* as representing only the rapid growth of certain current programs or the initiation of major new programs. TbeCOuld represent an expansion ofr of space programs, but we wouldthe development of major new systems to be centrally funded by the State Budget and included in science budgethe growth of science budget funds, in fact, does seem to allow for ample funding of all large-scaleprograms which have beenor can be reasonably postulated.wc wouldeal Increase in thaffort to be accompaniedomparable growth in employment and wage COStX Militaryhich we think may be financed by otherof the Ministry of Defense for adminiitrative and liaison personnel, military institutes,almost certainly could not have been expanded so rapidly. If this reasoning is correct, there remains the difficult task of identifying therograms which are receiving such generous financial support.

ne possibility occurs tosharp growth in other funds may reflect Sovietto stimulate the introduction of new products, and new manufacturing technology. That is. one can hypothesize that Sovietare shilling and that relatively more emphasis is being placed on this area. Tlie need to improve and speed up theof new technologyonstant theme in tho Soviet press. Tbe Soviet literature speaks about funds to implement new technology and for lhe development of product.on. These funds appear lo be in (ended in particular lo

offset tbe costs associated with Iheof new products and to subsidize thecosts incurred by enterprises during the early stages of new production.of the uses of these funds all center on the conceptually gray area between prototype development and series production. In the Western view, oxpendituies of this nature would be stretching thc conceptual limits. The Soviets, however, may wellmoney spent on these problems as properly counted in support offfort.

2S. This explanation would abo account for the failure of scientific employment and wages to grow as rapidly as expenditures. To the extent that wages or bonuses are covered by these fundi, most of tbc payments would be lo workers in industry rather thanrganizations.

oviet efforts to improveroad scale could benefit both defense and civilian objectives. Becauseindustry is so much more extensive and has greater problems in introducing newIbe rapid increase in other funds, if accounted for by this explanation, is probably largely for Civilian production.

hese explanations are only reasoned guesses, and none is fully90 we haveange to the estimates of the portion of other funds going lolus space. On the low side the military RAD plus space share of other funds is assumed to continue to growate ofercent For the higher estimate we assume ihat thehi* space share continues, to accounthird of other funds. This range introduces an uncertainty ofillion rubles in the figures

top srG-rt&f-


of Ihe Analysis of Sovfel Financial Oafa

nder lhe assumptions described so far, the calculations show (hat thelus space portions oi science budget funds, capital invesimenl, and other funds together account forercent of total funds. Analysts in other research organizations have estimated the military PAD plus space shareunds to be as low asercent or as high asercent This mvcrgence ofpoints up the inherent uncertainties and reveals that widely different conclusions have been deduced from the same body of data.

hows how the methodology, described up to this point, Is appliedample

Adjustment for Price Changes. In the Soviet Union, Just as In the US. some of the growth in science expenditures Is due to priceeneral revision of tho prices of corarrtodlties was initiated in the Soviet Unionnd throughouthe Soviet Governmenteries of upward wage adjustments. Soviet data on manpower, average wages, and prices for materials and construction have been used to calculaterice index. This index has been applied


Science and Military Research and Development Plus Space Expenditures8

Budget fur-as (announced)

Sourco of funds

Capital Investment (announced)

illion rubles

In billion rubles



Oilier fund.".


Estimated militaiy' space share

illion rubles

to thc current ruble estimate in order toa constant ruble estimate that betterthe growth of theffort in constant prices.

.resents estirnates of total Sovietlus spacefor the periodn8 rubles. It also shows category in current rubles.

Advantogos and Disadvantages of This Approach

eriving estimates of Sovietforlus space programs from Soviet financial data enables us to startocumentary base; one can have tome confidence that the figures for total fuiuls represent expenditures fornd space in the Soviet Union. The method permitswith other Soviet data, such as total budget appropriations, and irdormahon on education and employment of scientists and engineers. The share of these sciencelhat is allocated tolus space, however, roust be derived oo the basisarge number of inferences subsidiary estimates, and subjective judgments. Thoabout the meaning of the recent increases in published statistics onfor "science" only points up the degree to which, with little evidence to go on, our estimatesirect function of ourAnother disadvantage of financial data analysis is that thc results, expressed in rubles, cannot be compared directly with dollarforlus space in the US;omparison requires the developmentuble/dollar exchange rateinally, the methodology does not supply the detail necessary to makebreakdowns of the aggregated data and. in particular, toigure forlone.

espite tlte problems connected with thc method based oo Soviet financial data, however, it doesasis for aderivation of the cost of thend space effort as reflected in Soviet statistics. Although the resulting figures are open to adjustment in the future as we learn more about Soviet accounting methods, they doseful picture of the aggregateffort and that prwtion related to various defense and space programs, especially whoa viewed in conjunction with estimates based on direct costing.


nother method of estimating the cost of Sovietnd Soviet spaceis to estimate what Soviet programs ivould cost in dollars If they were carried out in theumber of studies which attempt to cost particular aspects offfort directly in dollars have been done within thc Intelligence Community, but no single research study has presentedestimates of what it would cost tn the US to undertake all Sovietlus space programs. For purposes of this Estimate, thc work done in theCommunity was collected andIn order to develop for the first time this aHemarjv* approach for estimating the magnitude of tlie Sovietlus space effort We call this method the direct-costing method.

his section describes the techniques used to derive costs of Sovietlus space programs0 dollars. Thc results are presented in Table BIT.ecause different kinds and amounts of data areon different aspects of Soviet militarylus space activity, the methodology is not uniform. For thc major Soviet militaryas ballistic missiles, defensive missiles, and militaryderive tho costs of system development, testing, and



01 3 * 5 7 8 9 0





DOD-Type Space RAD





DOD-Type Space Operations


6.9 0 9 1 1 *

figures in ihis Table are presented aa single values rounded lo one decimal place for reaaona of ease of presentation, noi because we have the confidence in them Implied by auch values. Because of the many aoureea of possible error in these figures, we have DO Rood bast* forange. The reader's attention ia directed lo tbe pertinent paragraphs of ihis Annex describing the methodologies and assumptions on which these figures are based.

Cost estimates have been either calculated directly0 dollars or converted0 dollars by applying the US gross national product (GNP) deflator for federal government puroliases of goods and services.

Includes expenditures for research, development, design, fabrication of flight lest vehicles, testing, and evaluation for lhe weapon systems involved. Does not include coals of development of nuclear weapons, nor general electronics research. Includes military airerafi only; excludes civil transports, such aa the SST.

Includes expendituresE on military apaceanalogous to those funded as "astronautics"n the DOD in thc US.

Includes DOD-type expenditures for military sciencen other equipment not included In the two lines above, and program-wide management and support. Calculated alercent of the total of the lines above.

1 Includes expenditures forelivitios such as ore funded by the ABC in the US.

Expenditures! for apace programs such as those funded under general support in the DOD in the US.

for programs such as those funded by National Aeronautical and Space Administration (NASA) in the US.

1 Totals may noi equal tbe sum of lie parts because of independent rounding.

from US costich we callince many space programs can

"CERs (cost estimating relationships) describe the relationships between several key characteristtcseapon or space system (such as take off weight, speed, and thrust In tlie case of aircraft) and the oast of developing tlie system. In someFR is developed for each functional cost elementdevelopment, prototype production, flightn otherER is developed to embrace all costsingle equation. In cither case the CERingle equation relating characteristicsost) is used for each broad category of weapon or spaceirctaft, unmanned spacecraft, space boosters, ballistic missiles,ERs are notfor each lype or model.

also be costed in this fashion, CERs (or cost models) are the basis for someercent Of the costs derived by the direct-costing method. Simple US analogy is the basts for anotherercent of the costs; US analogy is used for other space programs, AEC-typc programs,n ordnance, naval ships, andStudy of overhead photography of Soviet facilitiesombination of various other methods are the basis for anotherercent each. Finally, someercent ol thc costs are derived simply byactor, based on US analogy, to thc sum of the above


More specific explanations of tho above methods are set forth In the followingwith rumples.

ost Fjtimottnr, Rela-

tionships. We Illustrate tho use of CEllsescription of their employment in costing BAD for Soviet aircraft Of tbe elementsdevelopment, testing, and evaluationhe CERs for aircraft are the basts for estimating the costs of "D,he estimate of the cost of research {theogether with capital investment in R. D., is based on an analysis

The CERs for aircraft were derivedS Air Force contractor on the basistudy of the costs of developing, testing, and evaluating US aircraft in relation to such parameters as lhe weight and speed of tho aircraft. Separate CERs were developed for various cost elements: initial designand prcproduetion tooling for thedevelopment of the engine, production of the test aircraft, flight tests, andsupport."

Tbe CERs were utilized to calculate tlie cost of each model of aircraft developed by the Soviets in the past (whether It reached test and deployment orhose currently

under test, and those protected in the future."

The inputs into tbe CERs for this calculation were estimates of the weight and speed of Ihc

"The cost estimates did not Inchade the cost of tooling fee series production or the cost of sawtcs produced aircraft. Thc devdoprneail cost of thepackage in Ihe aircraft wai not included, but the manufacturing eotll of tho avionics package in. test aircraft wai

Protections of fuiuee system were hosed generally en lhe Tugh^nMidle-be SovietNIEs. Proyectiont of space systems were based

on. "Tlie Soviet SpaceECRET.

Soviet aircraft; the power, type, and number of engines; the complexity and quality of the aircraft; and the number of test vehicles. The results of the calculations were separate costs for each of the cost dements The costs were expressed initially5 dollar prices, which were then adjusted3 dollar prices using factors developed by the US Air Force. The costs for each aircraft model were summed and phased over time, usiog certain observed bench marks (date of first test. Initialcapability, etc) as well as US experience to nidge the rapidity of system development Since the costs olAE" were estimated by aircraft model, there was an easy basis For distinguishing between tho costs ofaircraft and the costs of civilian aircraft

wo problems remained. The first was to estimate the cost of research (thef RDTftE) and of capital Investment

in RDTotE-L

he second was to apportion the costs of research and of capital investment between civilian and military aircraft This was done by assuming that the afjpegate cost ofn aircraft should be divided between military and civilian purposes in each year in thc same proportion as the costs ofE" alone. When the total costn military aircraft for each year had been calculated in this fashionS dollars, the annual figures were adjusted0 dollars by multiplying them by the price index lor federal government purchases of goods andomponent of the GNP.

ERs were used In similar ways ton some space programs and on ballistic and defensive missiles. The inputs to the CERs for space launch vehicles, for Instance,estimates of the pounds of thrust for each slagc, Ihc number of developmental tests, anil the typo of instrumentation, among other


In developing inputs to CICIIs for missiles, certain questions arose willi regard to coverage. These were resolved as follows: for ballistic missilesF

{used as inputs to the estimate

For defensive missiles, tlte costs were defined lo include the costs of the development of associated radars and their construction and programming at test ranges, but not the cost of construction of anradar, even if it were the firsthe Dog House).


^]US Air Force studies of the cost of constructing and equipping similar facilities in thc US8 furnished appropriate factors of costs[2

such costs wasactor ofercent, based on US analogy, for the costs of roads, utilities, etc. The costs of the facilities going into ope rati oh each year was summed and used as tho estimate of the total mvesiment In each year.

he annual operating costs of research institutes supporting the development ofand missiles were develoriedT"

Factors developedAir Force studies of Ihe costs offacilities in thc US8 wereorder to arrive

at an estimate of the costs of operating these institutes. It was judged that the cost of design bureaus would be covered by costs derived from the cost models but not the costs of research bureaus, which carry on more general irsearcb-

S Analogy. Costs of Sovietsimilar to those funded by the US AEC are used as an eaatnple of the utilization of simple US analogues.n nuclear weapons in thc USSR was assumed to cost one-third as much as that of the US, on the grounds that the Soviet test rate is about ooe-third that of the US. The cost of weapon tests was derived by multiplying the number of Soviet tests each year by an average costS test An amount for development ofreactors, again based on US analogy, was added to these costs to obtain the total costs.

Various Other Metliods. The costing of the Salyut program furnishes an example of the useombination of other costing methods. The program uses two major types ofSoyuz ferry spacecraft and the Salyut station itself, on the one hand, and the space launch vehicles to place them in orbit on the other. Soyuz costs were scaled up from the cost of thc Vcstok/Voskbcd(which had been estimatedSon thc basis of estimates of increased sire and complexity. The Salyut station was cos ted as an analogroposed US space station. Thepace launch vehicle for the Soyuz spacecraftomposite of theooster and the Venik upper stage. The purchase and launch costs of this launch vehicle, and of theaunch vehicle for the Salyut sution, were derivedost curve showing thc cost in the US of placing payloads of various weights in orbit

n estimate was made of tbe number of Salyut stations that would be launched, by the time tho program is complete. Annumber of Soyuz ferry vehicles was also estimated The numbers of launches were multiplied by the cost of each booster/spacc-craft combination and the products were added to get the hardware cost of the total program. Non hardware costs based on US analogy, were added to the hardware costs to get total costs. The total costs were then phased over time, based on US experience, and keyed to the first flight These costs were then adjusted0 dollars by application of the price deflator for federal government purchases of goods and services.

djustments to Cover Additionaln anyffort there are alwayscosts that cannot be estimated on the basis of any of the above methodologies. In some cases, the activity cannot be isolated for study; in others thc activity is so small or so widely scattered that the effort to isolate it is not justified. To get at these costs,based on US experience, were made to the total of the programs already estimated by the approaches described above. Thesewere used for activities corn parable lo those funded under BAD in the US DOD on the one liatid, and for space programs funded by NASA on thc other.

n the US DOD there areccounts called military science, oilier equipment, and program-wideand support. Military science covenresearch in physics, chemistry, and other disciplines, such as would be carried oo in the USSR by thu Academy of Sciences. Other equipment is catchalln general electronics, command and control, DOD e* pendilures on nuclear.ost of other programs. Program-wideand support includes the housekeep-

ing costs ofacilities and test ranges, and the coils of general management and support

In the US. the costs of this natureinortion oi the Defense Budget have equaled about OO percent of tboures for weapon systems and space programs. In estimating Sovietunder theseimilar percentage is used.

In the case of space, the methodsso far do not cover the costs ofsuch as advanced research, andand support Adjustments to cover such additional costs were made, based largely on US analogy. These adjustments amounted to aboutercent of the total of costs esti-maled by otber metliods.

Because direct costing produces costs for many discrete programs or groups ofprograms, these costs can be aggregatedarge variety of ways. For purposes of comparison with US expenditures. It isfor us to aggregate themannerto (hat followed in the US, even though the Soviets do not do so. The results of the direct-costing method are presented Inl. pagenianner.

As can be seen inI. (he major growth in expenditures forlus space programs duringas in apace programs, military and civilian, and0 was largely attributable to tbe estimated drop in expenditures for civilian (NASA-type) space programs.

Advantages and Disadvantages of ihe Direct-Costing Method

lie direct-costing method allows us to specify pieciseb' what Is being costcd and therefore to gel some appiedation of thoimplications of particular Soviet it&D

programs. It also permits direct comparison ol" Soviet programs wiih US programs. But considerable uncertainty is also associated with this methodology. CERs may call for more information than intelligence canin these cases either engineeringhave lo furnish some of the inputs to thc equations or simpler CERs havee found. The manner of phasing the costs over the development cycle isatter of Judgment, as much of the development has to take place before wc have any evidence ol thc program. Moreover, we do not know for sure howctivity that never progresses to test is being undertaken.costs for such early work arerelativelyimilar problem is that annual cost estimates for recent years are particularly dependent upon projections of future programs. For example, Soviet costs0 and thereafter can vary hy hundreds of millions of dollars, depending upon what judgments are made about tbe pace andofehicle development and the programs associated withlunar landing, very large space station, planetary probes, etc. Analysis of facilities can lead to overestimating costs when programs are cut back and budgets reduced.

53. Despite the many problems connected with the direct-costing method, enoughaspects of the Sovietnd space effort have been examined toa basisystematic quantification of tlie Sovlol effort based on direct observationrograms. Although the resulting figuies must, of course, be considered asat this time, they do provide apicture of the Soviet programs, especially when viewed in conjunction with tbobased on Soviet financial data which wo translate to dollars by meansuble/ dollar ratio


he conversion of ruble expenditureto Hollar equivalent or dollar covts to lubles. requires thc use of an appropriate exchange rate,ubWdollar ratio, which reflects the comparative purchasing power of the ruble and the dollar in thc area.uble/dollar ratiohould quantify precisely the financial outlays needed in Ihc US and fn (he USSR toroduct or outputatisfactory method forutputhole lias never been devised.

hen output cannot be measured tbe standard approach Is to measureollar valuation of theutput would be achieved by calculating the dollar cosl of all the manpower, materials, equipment, and facilitiesuble/dollar ratio based on inputs would be appropriateepresentation of that for output if both countriesimilar mix of capital and labor inctivity; If the relative prices of manpower, materials, equipment, and facilities were approximately the same in both countries; and II the productivity of tbe inputs woe roughly comparable.

or several reasons, however, including distortions introduced by thc administered price structure of tbe USSR, and theand bureaucratic differences that exist, the structure of Soviet inputs and their prices differs grossly from that of the US. In tbc

"The "official" ruble/dollar rate of exchange (one ruble equals IIH) Isealistic Indicator of the purchasing pen of tfae ruble aa relataon lo the dellaa. Th* ruble isreely convertible currency in Inter-national markets, ill use Ir limited to Internal tians-aetinaa and accounting In the USSR and In lotra-CEMA trade.there iti value is eaiablbhed hbI-laterally by gnvcnuneal decree, not by the market aiechiniam,

-rOf CnCRCi-

US. for example, scientific instruments and computers arc of higher quality and are used more intensively than in tbe USSR. UScosts are much higher rclativo to other costs than is the case in Ihe Soviet Union. Accordingly, the US uses relatively lessand more capital equipment In its RAD programs than docs the Soviet Union. In these circumstance* an attempt to measure one country's activity using the input and price structure of the other is not very meaningful.

o improve the frame of reference for establishing an RAD exchange ratio, thoicfore. we have considered Soviet and US prices for the most complex products which arecomparable in tbc two countries. There arc, of course, still problems of limited data in this approach. We have little recenton Soviet priceside range of high technology equipment In tbe key area of electronics, however, substantial research has been done. This workasis for making reasonably informed judgments aboutruble/dollar prices for productwhich include communicationscientific instruments, testing andequipment, radars, avionics, amiOverall, thc estimated ruble/doUar ratio for electronics equipment is aboutuble is equal toith rubles adjustedfl prices, and dollars adjusted0n thc military procurement cate-gories the ratio for electronic equipment is SInd for instruments, testing, and measuring equipment ItIFor all militarythe estimated weighted roWe/doUar ratioland theratio1

his procedure permits us to approach our objective ofonversion ratio that represents the dollar cost of tbe Soviet RAD offort as il il were being conducted

in the US with US resources and supporting technology. It is likely that an RAD con-venion ratio based on ruble/dollar ratios for electronics and other advanced machinery and equipment more accurately reflectsefficiency of inputs in RAD than one based directly on the prices of inputs. The limited sample on electronics and other com-plex equipment which has militarysuggestsounded ratio ofeasonable appro rin for converting Soviet rublefor RAD to US dollars.

here is still another way touble/dollar ratio. The method of calculating Soviet expenditures for military RAD plus space based on Soviet financial data described earlier in this paper givesuble estimate of these costs. The direct-costing methodsubsequently provides dollar estimates of costs for whatattbc same activities Thus, if each of the two

series accurately measured what it purports lo measure, one would secure weighted ruble/ dollar ratios for each of the years covered by

simply dividing the ruble estimate by tlie

dollar estimate.

n practice, matters are not so simple. For lessons already given, it is not possible lo be confident that the two scries cover tho same things, or that they accurately measurethey do in fact of the two series is revealing and(see1 i.. For one thing, it appcais lo rule out overall ruble/dollar ratios as low as that derived by applying US prices to Soviet Inputs, or as high as the officialrate of one ruble. For another, it suggestsuble/dollar ratio dose1s about right. This is thc approximate ratio yieldedomparison of Soviet nnd US prices formachinery and equipment










. .

figures io tail Table are praamled a* single values rounded to oa* iiriaaal place for reasons of ease olotwe hav. Ueidenre in then implied by sack values. Because of tha assay sources of possible error in tfaeat tigurra. we have no good basis foraage- Use reader'a attention is directed to the pertinentof tills Annex describing the meuiodologica and assumptions on which thcec figuics aie baaed.



side from ils rough agreement wiih the ratios derivedomparison of tltese two independent estimates, there arc tt priori reasons for placing some confidence in lite ratio of one ruble. This ratio Is based on the cost of outputs, rather than of Inputs, and therefore incorporatei anfor relative efficiency. Moreover, inthe cost of advanced machinery and equipment, one is dealingombination ol inputs, processes, and products which can be assumed to represent the purchasing pr-wet parity between the ruble and the dollar in advanced areas of technology such

and space. Thus, this paper1s the ratio to convert estimates of the ruble costs of Soviet RAD into dollars

Ot We recognizeeparate ruble/ dollar ratio for each year, ratheringle ratio for the entire period, would better reflect changes in the mix ofnd space activities. But we have no good basts for such adjustments and. in any case, the potentials for error abcady in the data are much greater in magnitude than likely year-by-yearhe ratio. Consequently, we haveingle ratio throughout the period.

V provides thc dollar values lor Soviet expenditures oolus




The figures in Ihis Table are presented aa aingle rabies rounded to one deoiosal place for reason" o( rase of presentation, not because we hove Use eon-fidence ia them implied by eeeh values. Because ol Ihosources of possible error inuiea.ase no rood basis for statins;range. Tha readee'a allention is directed to lha pertinentei Ibis Annex oeaerflwg tbe and uiL-pUom oo which these figure* are baaed.

'onversion ratio ofo the ruble.


(lint icsult dam the application of this ratio to tlie estimate based oo Soviet fmarscial data


hows tbe estimates for Soviet expenditures on military BAD plus spacefrom analysis of Soviet financial data and the applicationuble/dollar ratio of



0 Dalian

Soviel Financial



The figures in this Tabla areounded to oca dacimal place lot raaaooaof presentation, not becamehava Uiethen implieda Bnaior of

Uieucea ol poaaibla error inigures, we have ut, good buttange. The reader'* niicnlioii ia dueelcd to lhe perUnanl para-Ctsphi of ibis Anneaethodologiesumptions on vhtcb these fig am are

FromV. .

From Table IMI,3S.

ith those developed by combining direct-cost estimates of individual programs.

be estimates derived from Sovietdata are roughlyimilarand grow at about the same rateote based on the direct-costing methodology, crcepl for recent yean. There is no basis for considering one more accurate than the other, nor do the two serieskely range The validity of the dollar estimates derived fiom Soviet financial data depends upon three factors; (a) the extent to which the coverage of science expenditures conforms to Western definitions of RAD, (b) thc precision of the distribution of the Soviet expenditure*military and civilian programs; and (c) the accuracy of the ratio used to convert tho rubles to dollars. Thc rate of growth of the ruble scries is affected by the deflators used to convert currency units into constant units, and therefore by any IriaaAiracies in theConsiderable uncertainty exists for each aspect of these calculations and It must be recognized that tho estimate couldor understate the Soviet effort by amargin. The validity of the direct dollar costs depends upon tlie precision with which the Soviet programs have beendefined, and projected both for past and future years, and upon the accuracy andof the direct-costing relationships which have been applied. Again much uncertainty is present, and this appioach could alsour understate,ubstantial marginue magnitude of the Soviet programs. Never the less, thc two series are reasonably consistent despite the urxertunbes of the task and the completely different approaches on which diey are based. This consistency en-


presumption (lie general

magnitudes which result are in accord with reality"

hc only year in which results are at considerable varianceere theprobably results from tl* sharp andupturn in reported Soviet expenditures when considered in relation to the difficulties in identifying and txtttihg Sovietlus space programs in the rnost recent years. Much current Soviet Red) activity has not yet produced an identifiable prototype or test vehicle, and tbis tends to introducebias (or tlte most recent years in tlie estimates based on direct costing. Tins phe-

- Vice Adin. Vincent P. deSN. the Defense total ligence Agency, Mai. Can. PhuTrpavidson, the AauataM CW at Staff for Intefligrnee. Depaiunent ol theBear Adm. Bart P. rtecianus, thu Director of Navalattartment ol the Navy; aad Uaj. On Ceoex*orean.ha Assistant Quel ol Stall. IrMrsugence, USA P. do not believe dial the general comtttency of remits obtained from tha two rnatbodoiagka chouU encourage the ptrauaapOOn staled above. They believe thai neither methodology produces very crediblrbul thay have considerably mora confidence in the dmct col ting approach. Tbay betsev* that the lad olMailedated in paragraphf Utii Annoi) maker the remits obtained through thc firuiidal data method highly questionable. They believe that the retiaaner far more raorstf yean cV-rivrd through lhe financial data method depend loo heavily on erbopolation of thc data. Farther, thai' note thai la tbe letuhs et the analyiU of Soviet financial dalan apjuient absolute decline in ernenditurej for military RAD Irons IOCS6 which ihey find Inecoriitrmt with th* dirott nomi results and with their apptreiirior. ot the general magnitude of the Soviet effort. While believing thai lhe direct costing approach provides an aggregated total estimate that ta ala&ost oarUuofy lew. they have more, though not bigh, confidrnor in thai method. In view of thoir doubts concerning theol the financial data inethadoAogv. 'hey would dracoutace the presamption that thearallel results cantaken ai supportive of either methodology.

nomcnon may account for Ihe rapid drop in estimated space expciiditurcsime is likely to provide new dala to supportwith both methodologies and as thc estimates for recent years are refined andthey may move more closely together, particularly those


be (wo scries of figures presented so far cover expenditures for bothlus space. The next step is to remove from tbetn thc estimated cost of civiliannmanned lunar exploration and planetary probes) and the estimates cost of operational space programs onas been completedatellite reoonnaitsanoc and communications) in order to get figures forirected at purely military goals. In thc case of (he estimates of military PAD plus space derived from direct costing, the procedure is simple andto drop for each year the costs calculated for tho aforementioned space activities. Thisbe done, however, within the series based on Soviet financial data, because the necessary detail is not available. Consequently, It tt necessary to subtract from tlic scries derived from Soviet financial data thc cost of civilian and operational space programs derived from the direct-costing method.esult each series is reduced by the same amount in each year, and thc absolute difference between the tworepresentinglorw--remains (he same as that for the two series when they representedlus space (seeI,.

f necessity, both series again appear roughly consistent, except in more recent yean The data imply that after the. Soviet military RAD expendituresvery little if at all. and (hat must of



0 Dollar*

Financial Data*


ures in this Table are presented aa single values rounded to one decimal place lor reason* ol ease ol piescatation. not because w* havehem implied by such valuta. Became oi the many sources of possible error in llioae figures, we havo no good bsaia forange. Tbe reader's attention is directed to tbe pertinantof this Annei describing the nirlboeWosnea and assumptions on whieh these figures are based.

for military RAD and apace from, page ft, less DOD-typt apaceand NASA-type apacespendituiea IromI.

of expenditures for weapon systems RAD. DOD-typ*. otber DOD-typ* RADC-typeiornI, page 3S.

the growth In Sovietame from die space programs. Thii conclusion Is beatin light of lhe lact that Sovietfor military RM) were already high at thc beginning of. Tlie relatively slow rate of giowtli duringjGa--avcragmgercent areflectshigh levels of activity throughoutt the bcginiiiiig ofore ballistic missiles, aerodynamic missiles, and

military aircraft weir under develcrpmcnt than at the end. Although aircraft arc now moreand more costly to develop, most of the missiles now In test are modifications of existing ones, rather than new systems, and are coruequendy less costly to develop.

n contrast. Soviet space programs, at the beginning of. were in their in-fancy. During the. the numbers of space bunches and estimatedfor space programs grew about nine-fold. Wc estimate that Soviet spaceof nil types0 cost Ihe Soviets almost aa much as military RAD programs.

e believe the resulting estimate that the predominant growth element inat been the space effortrueol acrual developments in the USSR, and not (he result of attributing high costs to space programs and low costs to military ones. In both th* US and the USSR tbe same plants and test facilities support missile and space development. Analogous, If not identical, methods have been used to estimate costs of the space and missile programs. They both, tn turn, use methodologies similar to thoseto cost aircraft airframes. Thus, we discern no reason for systematic upward bias in Ihe estimates ol the costs of Soviet space programs when compared torograms.

stimates for bothnd foi space based on direct costing may well be low because of tlienoted earlier for direct costing tothe results for the most recent years. To the degree thai this is so. thestimate based on Soviet data would be high, because it is derived by subtracting tbefor space based on direct costing.


Tho Ninth Five-Yearalls for expenditures for sciencelus space) ofillion rubles, or an increase ofercent over the previous five-year period. This represents an average annualfor the period of aboutercent in current rubles. It is not unusual for actual expenditures to deviate from thc planning figures, and it may well be that reportedforeriod will fall below Uie planned levels. In any event,inflation will probably make real growth less thanercent If past rates of inflation continue, thc projected rate ofercent in current rubles would equate to aboutear in constant rubles.

lone may also increase byear. There arc, however, various considerations that could lead us to estimate slower or faster rates of estimatelower rate of growth is the apparent predominance ofn recent growth and tho fact that theof facilities that carry out Soviet military

as stowed. Studies show that during the early andacilities to support space programsn aircraft andgrew at roughly the same rate asfor these programs. It may betherefore, that the erpansion of facilities declined toear at the end of. On the other hand, in the SALT context. Soviet leaders may wish torograms rnore rapidly tban in the past. An appropriate upper limit may be established by tlie following set ofIf, (a)ill grow at the same rate as the overall planned rate; (b) civilian space expenditures will not grow; and (c) the remaining growth will be for. then an arithmetic calculation indicates that, could grow atear. These rates oflow ofear,ighe viewedeasonable assumed range for future possibilities. Under these assumptions, Soviet expenditures loray grow to the amounts indicated below:

D 9 S

Soviet Financial1 8 3



Thii Annexrieriptictof Iht overall arrangement of (A* principal InRituTiftu omd orgawUertoas totalled at SamtHond determentiUon/ at petit off emphailied. Thr discussion ii general, and thettfoie only Oie keuait covered

In any detail.

op Party-Government LcoeL BAD if controlled at the apex ol state power by tbe Politburo of the Communist Party and by tbe Council ot Ministers All basic decisions on the scale, direction, and organization of BAD are made, or subject to approval, at that level of authority as il by directorsiantSince there is no private sector in the Soviet economy, all institutions having an BAD function are operated directly by one oragency of the government.

dentification of key personnel andand their roles in tho SovietBAD sector is derived almost exclusively from classified materials^

jVVhilc the available data permit us toroad assessment of the management of Soviet military BAD, major gaps remain in our understanding of how decisions actually are made at all levels from the Politburo down to the design bureau. Specifically, little is known about thc impact of interest groups and key figures on decisions at tlie various levels. In addition, there is only fragmentary information on the Defense Council, Ihc Military-Industrial Commission, tho role of the Central Committee Secretariat.

and the Ministries for Genera! Macliine Build, ing and Medium-Machine Building in military BAD policy making and management^

^hasase fo* theof assumptions and hypotheses about management and policy making in this'scctor.

man Politburo establishes broad policy on BAD. It is served in this function by various staff organizations controlled through its Secretariat. Within the governmentalthe Council of Ministers, and Itscommittees, implement the national policies established by the Party. It is in the key governmental committees, the StateCommitteeleared to by its acronymnd the State Committee (or Science and Technologyhat many important decisions are made and co-oidination of BAD at tbe national level takes place. The USSR Academy ol Sciencesadvice to these bodies. Including counsel oo the exploitation of resources and theaspects of economic planning.

hc State Planning Committee conducts planning ofup toears into the future) and males provisions (or thc input of thc materials needed. Its principal science planning Is encompassed in

top GEcnrr

Five Yearlie plan sets forth thc main scientific and technological problems and the requirements fur their solution.

The State Committee for Science and Technology has tbe overall resporafoility (or coordinating non-military scientific andplanning and planning for basic andresearch. Its function is to ensure thcutilization of the results of this HAD. The GKNT establishesand appliedpriorities on the basis of an annual list ofkey problems of nationalcompiled from the recommendations of country-wide councils on scientific piob-letns. It also allocates science funds for these projects.

Thc Academy of Sciences advises the government on matters relating to science and technology, including the exploitation ofand thc various aspects of economic planning. It is the most prestigious Soviet scientific institution, and has primaryfor basic and theoretical research; it also conducts some applied research.esearch institutes employing the best known Soviet scientists, and many of these Institutes do work in support of military needs.

"Two of theic tub of the cuiicot) cooccn science ted tees-scaogy. On of these onphxizo tbe speeding op of scientific progrc" ind the carrying outnified technological policy. Industry is to be mechanised and automated more 'itemivrl* with modem clectieruc ond computer technology, quality la to be faired and st.odardUatlon broadened, and iclenlCfic and technical Informso be made mora readily available. TW other of these banc lastsIkedevcfapn*ro. of fundamental and applied scientific research" and application of ils results in lhe national economy more ispadty. Alto envtuoned are various incentive schemes thai would make it advantagec-u. lor organisations and individuals lo tryring aboul the Hflcienl and prompt uie ol RAD results by industry snd the mill' Ury.

Party Politburo receives adviceplanning involving mi'iioy FADfrom the staffs of the Centrallive Council of Ministers, and theCouncil. One of tbeandidate member olis D. P. Ustinov who hasthan three decades managingprograms. He probably is the keyfigure in matters concerningof military RAD

Mih'Ury-lndtistrialof the Council of MinistersRAD programs and plans, andof tbe various ministries involvedwork. It is headednder lhe control and guidanceand is made up of representativeseight ministries engaged in militaryIts primary (unction is toRAD, the production ofand plans involving the work ofone ministry of the defense-relatedThis function includes designationresearch institutes, design bureaus,lo carry out RAD andarticular program. Itfinancial, materiel, andto support specific programs.

he development and productionubmarine-laurachcd ballistic missile illustrates bow the VPK discharges its coordinationAt least five of the defense-relatedprobably would be involved in thbTbe Ministry for General Machine

"The principal branches of the Soviet economy arerghl mnusoscs oompnle the so-called "defense industrialefense Industry (conventionaleneral Machine Mil ( missile,ehicles).Bu.WWand solidedMai Machine Building (nuclear weapons andviation Industry. Electronics Industry. Radio Industry, and Shipbuilding Ii-Suitry-

Building would be responsible for developing (he missile itself, Insuring thc compatibility of thc missile with the submarine would be Ihe job of the Ministry of ShipbuildingElectronic subsystems would beby thc Ministry of Radio Industry which would get lhe necessary components from tho Ministry of Electronics Industry. The warhead of thc missile would come from the Ministry of Medium Machine Building. Tho VPK would supervise thc coordioation of tbe many components of these ministries Iromof the project to completion of

inisjOTtof Level Thc headquarters of thc Ministry of Defensehc defense-related production ministries, andat various levels of each of the five armed forces components have military-technical commiltccs thatdvice and guidance to thc VPK. These committeesand coordinate various research plans and examine new projects Tbey also advise their respective commanders in chief on weapons development and procurement. At tlie MOD level, support for thb activity Is drawn from pertinent sectors of the scientific community such as academics of science and higher educational institutions

mfiiufc ond Bureau Level Mostwork on military products tsby scientific research institutes orbureaus which are part of the ministries theythe defense-relatedThese organizations arc headedhief designer and usually have bothand production facilities. Tbe design bureaus generally depend on the basicdone by others, but many of the larger ones conduct their own. "Central design bureaus" coordinate and supervise the work of the many design bureaus that work on thc same typo of product, such as fighter aircraft.

IS Early in the developmentarticular piece of military hardware, the Sovietswillpecial commission within one of the defense-related ministries to make recommendations on each phase ofincluding preliminarynockup, detailed design, constructionrototype and, finally, testing. Tbenormally will include representatives Irom olher industrial ministries, scientificinstitutes, the military users, and the chief design bureaus, tlsereby assuringand coordination between production, and user. Final tests are conducled by tin- special commission to ensure that performance meets the military'sOperational tests are then conductedilitary test crganizatiou to determine If series production of the hardware ought to follow"

principal Soviet ttuhtary test ranges foe the various kinds of armaments are,largest ol the Soviel testintercontinental bal* lUtlc omuls. (lCBMi).and aU large spaoa boosters are fl.gfct tested

3 Me-ml has been

mad in recent years f*

aupport, and otheri this facility. Tlie ft'enolia rangedevoted to naval missiles and is currentlywith ike devs^opaaeat of the newest andthese. Kspurrmie oldest mtmle leitil the primary teal center formmiWi of lvs> than intercontinental range.has been used lor tbe lasting of certainmlsslleiiamrmloy "bate of the Plight Test Initilut* responsibleacceptance tests on all aircraft Thefilingsed lor both surface-toand su'tn-surface aerodynamic vehicleslaud and sea targets. Tha Emba test centerulr of tbe deveiopanent of laeOes! SAMmust heavily instrumentedused for testing aiitibsllittic andSAM svatenw

tinal are Ibruclearii-ir;facilities.



ultimate ineuiure ol IheSoviet military research andis its ability to solve the complexproblems posed by advancedthis Annex we evaluate then major technologicalin the military context. Nomade to review specific products andthis purpose, Ihe reader isthe Estimates, on tlie variousthc Soviet military forces.

A. Materials

Soviet metallurgicalsufficiendy well developed to rneetmilitary requirements. RAD in thissupported adequately themanufacture of such weapon systemshigh performance aircraft, andThe Soviets effectively usematerials and proven techniques toproducts the dnsired service life,decrease the time needed to developmaterials for airframes, they appear toadvantage of high strength alloysweight. Soviet titanium technologydeveloped Higher strcngtli steelsare being developed, particularly forbulls. Materials technology foris weak in such areas as highprotective coalings, but RAD invacuum techniques and highprocesses are believed to bethat in the West.

s iu (lie past, progress by the Soviets in tho important areas of metallurgy will con-

tinue to benefit from their knowledge ofin foreign technology. The Soviets have repeatedly demonstrated an ability to combine effectively their own RAD effort with knowledge of foreign accomplishments to solve materials engineering problems.

n polymers, such as plastics, fibers, and rubbers, has improved markedly In both quality and quantity over thc pastears. Thc Soviets are doing workevel comparable with that in (he US on (he development of stronger, fiber-reinforced, polymers some of which are stable at high temperatures. Nevertheless, there hasotable deficiency in the application ofprocesses discovered in the laboratory. In the area of synthetic rubber, for example. Soviet capabilities lag many WesternThereap between competence in research and industrialresearch on synthetic rubber is of high quality, but in developing the fadUties for producing this material in quantity, the Soviets arc seriously deficient.

B. Electronics and Sensor Subsystems

5 Materials, Components, and Devices. The quality of Soviet fundamental research Lnmaterials is approximately equal to that of thc US. ln areas directly ln support ofneeds there appears toood coupling between basic and applied research. The Soviets have available all conventional types of electronic materials and they havesome with special performance such as

shortcomings have retarded the development and use of advanced computer programming techniques.

n spite of these problems, thc Soviets probably are able to satisfy their requirement lor computers in high priority militaryFor the present, thoy probably will continue to follow their general practice of employing computer designs andtechniques that are just able to meet thc minimum required performance of computers in military projects.ew instances the Soviets appear to have deliberately avoided using computers by adopting design concepts which do not requirehe advent of the newer Soviet computers will inevitably provide the additional experience needed to advance the development of programming techniques and to expand (line sharing systems.

so^istic Missile Guiiance.

solar cells and radiation detectors wiih high rcmperatuic resistance* -charac* eristics ofvalue (ot some military and space uics.

he quality of research on electronic component] and devices is generallywith that ol* thc Free World; actual practical developments in these areas, how-ever, lag tlsbsc otree World. Worst in discrete devices such as microwave transistors and dtodes is adequate to provide near term Improvements in Soviet electronic warfnrc, avionics, and phased-array radar equipment. Studies continue in opto-escctronics using lasers, and logic applications have beenwhich may greatly increase the capacity of future information systems. Corssiderable research is devoted to components used in computers.

he mayor shot teeming In this area of technology is poorly developed manufacturing lechniques. The quality and reliability of finished Soviet components therefore israther low by Western standards. Quality tn this area depends upon cleanliness, the use of high precision tiutrurnenU operated by competent technicians, the establishment of adequate testing and quality-control rtaadioda. and the use of high purity gases aqd other materials. Soviet industry has met theseonly marginally.

omputers.n computer theory and design isar with that of tlie West But In the rnjinufacturc of computers and related equipment, the USSR is presently developing prototypes comparable to US models first producedhc translation of basic research into high quality endas lagged thc West due largely to irtadequate manufacturing procedures and an inadequate supply ol parts. The Soviets are also far behind thc West in the development of storage: memories anil input-output devices. These



hile there ate many uncertainties in assessing the ptecisinn of Soviet baUisticguidance systems, we are confident that the Soviets are not yet approaching the limits of accuracy attainable by improvingno change in the basicsystem design. The quality of guidancecclciothefactor in achieving very high aornsraeies. The sum of the evidence pointsag in the USSR's capability to produce very precise gyroscopes and accelcromcters. This is due in partreoccupation with theoretical, rather than practical. Investigations.problems have also hindered Iheof high quality inertia! components- For example, the USSR's continued dependence on ball bearings (rather than gas bearings or other suspension techniques) and tho use of aluminum (rather than beryllium) In these critical assemblies have been primarilyboth for shortened operationaland for significant reductions in quality.

hc SovieUigh prioritynffortange of new approaches to improving conventional gyroscopes and ac-ceieromcters. They probably are also worling toward developingottona]ensors, such as tlie laser and cryogenicIn developing new gyroscopes and hc-ccleromcters. the USSB probably willon the complex engineering and man-

" Dei pile the difference in drs-re. both tha US and the USSB use hiertial autdancc foi (heir baDirMe rois. lilef. This technique insures that the miMllr* are immune lo laminiiu: becauia lhe eiiidanc* tyUora ii contained in the miiiili' No radio link with tlie (round it reoulred. Tha prune ctvncxwO areeeSeromemi (which meamre accekrattontveo direction) and Jyroieopw (whkh meirurr deviationeference direction).

ufacturing problems, since much of thogroundwork in these areas already has been laid.

the Soviets elect to develop stillguidance, we would expect themto more sophisticated techniques,stdlar-ioertial, or tentiinal,

J stellar-inertia I

particularly at-

tractive for missiles Uurichcd from mobile platforms, such as submarines, in order to reduce errors resulting from uncertainty in thc precise position of the launch platform. In any case, the quality of gyroscopes and ao celerometen will continue to be important factors in determining guidance system

Radar. Soviet radaras generally followed traditional radar design principles, but major advances have been made in thc development of radars incorporating phased-array technology. Present research appears to be concentrated on improving existing radar designs and using proven concepts. The Soviets probably now lead the West in the development of large radar antennas. Tbcy still lag. however, in such areas as signaland solid-state radar develop men ts.

Because of rhe different designused by the US snd USSR radars for Iheir strategic defense, direct and specific comparison is difficult. Although the level of sophistication of Soviet radar toclinologylo be somewhat less than that of US designs, tbe Soviets have deployed somesuperior to similar operational US equip-rrsent. This in capabilities Isatter of kind than degree; timeumber of examples of US radars withcapabilities which the Soviets have not demonstrated and probably cannot match. Still, the Soviets haveegree

ol independence, especially In ihcir large phascd-array radar development whicha high level of confidence in their own engineering decision-makinga-pabihty.

The further development of radarfor ballistic missile defense, over-the-horizon detection, space surveillance, and low-altitude air defense is being given highConcentrated and competent research is underway in such areas as signal processing, data processing, and electronic steering appb-cable to such systems, which probably will lead to significant advances within the next five years.

Communications. The Soviets, over tbe years, havearge scale, highrogram io all areas ofWhile this program has resultedigh level of technology in some of these areas, Ibeir overall research efforts havein systems inferior to those developed in tbe Western world. Strong points of Soviet efforts include the theoretical aspects ofamong which are information, modulation, and coding theories. They have excelled in propagation andspects associated with high frequency (HF) long distance radio communications.

With few exceptions, among them laser, tbe Soviets have come up short when compared to tbe West in almost all aspects of high capacity multichannelR&D. This is especially true concerning high trcquencies (VHF and UHF) which are used to take advantage of thn greater band-widtlis. Illustrative of tbe Soviet lags their troposplseric scatter systems which have failed to attaio the level of technology of the West,in terms of the development of high capacity systems.

Future Soviet efforts probably will be concentrated in systems dcvelcpmcnt asto phersomenological or. Such areas as filtering technology and HF wideband components and devices will probably be stressed. Tlie overall prospects for future progress in these areas appear good. One possible limiting factor during the next several years, however, may be the it.ability of the Soviets to provide high-quality. HF. electronic componentsarge scale.

Lasers. Tbe Soviet Union occupies an advanced position in. with itsprogram second only to that of the US. The Soviets have demonstrated competence in all areas of laser research which are of importance to military applications. They are actively engaged in Use developmentislti-kdowatt gas dynamic laser, and although they have admitted to achieving powers ofew watts, tlie techniques they are employing are thc same as those used hy the US to surpassilowatt power level. Similarly, tbe Soviets are pushing theof ekctrically-excited and chemical lasers, other devices with the potential for large power outputs. All of these lasers are being studied In tbe US as possible weapons.

Another area of Soviet expertise Is thc Mghpowu pulsed laser field, especially nec-dymitsm-glass lasers capable of producing short pubes of light at extremely high-power levels. Tho Soviets have used these huge,devices in studies related to plasma generation and controlled thermonuclear(CTR) and probably have extended their Interests in laser/plasma interactions to include those aspects related to nuclearresearch.

general direction of Soviet laserit expected to remain the sameon high-power lasers (pubed and


laser/CTII studies and, very(ho application of lasers to weapon systems. They are faced with thc same difficulties as face the US, however, in perfecting lasers for weapons purposes. These include both theof tbe required high energies and the transmission of these energies through tlicto (he target, as well as Ihe problems associated with pointing and tracking. It is highly unlikely that io the neat five years the Soviets can overcome the existing obstacles and succeed in developing laser kill weapons. Other military laser devices such as target illuminators, range finders, and similarrequiring low power also may be under development

C. Propulsion

issile and Space Propulsion. At the end of World War II, the Soviets hadtheissile with its throttle-able rocket engine. Theyhis engine fn their early missiles and refined the thrust throttling featureoint well ahead of comparable US technology. This technique obvfalcs the need for onboard digitalby varying the missiles thrust so that it canreprogrammed trajectory to its target. It also obviates the need for vernier {control) engines In some Soviet missile de signs and permits other simplificarionj as welt The Soviets also led the US in developing the materials and techniques Deeded for very high chamber pressures to increase theof rocket engine combustion.

onetheless, tlie USSR has lagged (he US in many important areas of missile and space propulsion technology. Notablearc the development of liquidengines (an essential ingredient in the US Apollo program) and the development of largo sol id-pro pelocket motors (long

used by (he US for both missile and space applications).

iquid Propellants. Soviet work on liquid-propcUant systems has resulted in the production of reliable rocket engineside variety of missile and space launchRoth thc US awl (lie USSR progressed from non-storable liquid propellantsiquids (hat could be storedissile to Improve reaction times and to simplify handling. While the US then turned to solid propellants for its ballistic missiles, thc USSR continued with liquid-propellant systemi, which still serve as tbc backbone of Use Soviet ballistic inissile force.

he USSR's more recent ballisticuse nitrogen-base oxidizers withfuel, which provide the highest energy of all conventional slorablc propellants. If tho Soviets elected to improve theof their intercontinental ballistic missile (ICBM) propellants. we would raped (hern lo use additives with their present propellant combinations, rather than turning to propel-lanls such aswould be more efficient, but would greatly increase toxicity and handling problems. Fluorine or other high-energy propellants, however, probably will be used eventually in the space program.

olid Propellantargefor manufacturing solid propellants. (he Soviets have flight tested only two solid-propellantCBM and theedium-range ballistic missile (which is made up of the upper two stages of thenly thos known to be operationally deployed, and thus far only In limited numbers. Solid propellant ICBMs have some unique advantages, such as ease of han-dhng and simplicity of construction. But they require stringent environrnental controls and do not lend themselves to the guidancelong used by lhc USSR.

The Sovicl production capacily appears to be more than adequate to meet thc USSR's present solidhe SS-M, and an array of defensive missiles and tactical rockets. Additional solid-propellant facilities are under construction We believe, therefore, that the SovieU will press forward with solid-propel. There are Indications thai they arelarge prototype soltd-propellant rocket motors for ballistic missiles and probably for space boosters, and that effort appears to be movingteady pace.

rftgh-h'twrgy Propellants. We believe that the USSR has been working oo theof Idgh-ersergy propeUanU for the space program since theul they have not yet been used in any flight tests.

The constructionnd static test facilities suggesU that the Soviets have been conducting research aimed at exploiting liquid hydrogen and liquid fluorine as high-energy propcllaiiU. It is difficult to predict thestatus of these programs, but thoprobably ore about ready to flightiquid-hydrogenpace systemiquid -fluorine engine could reach flightstage within two years.

Marine ond rVotxif Propulsion Tbc iteam propulsion planU of die latest Soviet naval ships and submarines are of higher power per unit of wolght and volume thanUS systems. The major Soviet naval combaianl ships pack more horsepower, ond are capable of higher sustained speeds than US ships The SovieU were thc first to develop and operate ships utilixirig all-gas turbine pro pulsion planuower rating not yet achieved by the West. Certain Sovietare the fastest in the world Soviet literal me indicates thai there wdl bo con

tinucd development and expansion of the uses of steam planu incoqxwating higher steam temperatures and pressures, and superheating.

There is no specific informationon current developments In Soviet naval and rnarine nuclear propulsion. We can,make some gerveralirations based on thc past perforrnance of units using nuclearsystems. Thc SovieU have chosenpropulsion planU of high power to get maximum speeds for their nuclear submarines, while paying tho price of nolser operation and more difficuli maintenance. Nuclear fuel cores must be replaced more frequently than is the case with US reactors. The SovieU are aware of current Western research de-velop-mcnU in this field such as the concept of the consolidated nuclear steam generator, whichore efficient and compact reactor system, and on thc natural circulation reactor, which requires no pumps and therefore is quieter, and thoy may now ben these areas.

Aircraft Propulsion. Turbine engines for their aircraft arc well designed, but thehave had problems because of thethey follow In producinguse of low quality materiab have resulted in engines that need frequent overhaul and which have relatively short lifetimes. This approach also produces larger and heavier engines that are less efficient than Western engines because of their lower operating temperatures. Fuel consumption rates for most of the knownturbine engines are higher than for those of the West, especially when the afterburner is in use. Thc increasing use of advanced metal alloys has produced modestin the performance of Soviet engines, mod notably in operating temperatures and fuel consumption, bul the Soviets are still far behind Western standards for the time that elapses between engine overhauls.


System) Design

eneral. It is particularly difficult to evaluate lhe effectiveness of Soviet design of entire weapon systems, as such evaluation presupposes our knowledge of the design goals and performance standards. We do not have ihis information. In the absence of such criteria, we often tend to evaluatesystems designs on the basis of their effectiveness relative to similar US systems. If Soviet strategic military objectives (and hence weapons development objectives) were the same as those of the US, this wouldalid exercise. But Soviet objectives, design approaches, and production and performance standards are different from those of lhe US. Consequently, weapon systems Ihat mayless effective than comparable USoften lag US developments for reasons of less pressing force requirements, not for reasons offfectiveness.other systems Ihat appear to lead the US often do so merelyeqube-ment was identified earlier in the USSR and they have little relation to technical capa-biJiries.

tierc arc, however, commonin all Soviet weapon systems design that are Indicative of the effectiveness of their general approach. Their weapon systemsuniformly to be relatively simple inrugged in corasriuction. and reliable in operation. They seem lo use conservativepreferring many small technicaltested at each step, rather than greater steps. When they have departed from these principles, thoy have been less successful.

hc effectiveness of their BAD asin systems design thus consists not so muchew large bold steps, as it docs in persistently pushing ahead with ngrowing effort on many fiontt. This is not

to say thc Soviet approaches lack Imagination, rather that they limit risk taking, ami relyonsiderable momentum of effort to carry them forward.

eillttic Missises. The Soviets have tended to be conservative in dcslgnfng their ICBMs and other surface-to-surface missiles. Ihey have chosen relatively simple and rugged systems, which, after flight testing, have proven to be reliable^

JTbe Soviets tend to stick with proven design concepts for their ICBMs and notechnological weakness has been evi-dent. For tbc most part, present designs evolved graduallyesult of Improvements In earlier missiles. During the flight testing of specific missiles, which usually involves aboutirings and lasts about two years, the Soviets occasionally experience serious difficulties but these have been overcome in most instances. Some missile programs, however, never reached the flight test stage. We know of atozenthai were abandoned, some after construction of facilities for their launch indicating selectivity and competition Inork In this area.



to Soviet ICBMs overdecade have included alienation ofand shape of ie-entry vehiclesof tankage to permil greaterpayload, the use of multiple RVthe introduction of penetration aids.required to flight test thesehas varied coruidetably. On theabout two years has beeneach one. and the number of frightranged from about

Space Launch VehlcUi. Six distinct space boosters,ide range ofhave been developed by tlie Soviets. Through thehose boosters were reliable adaptations of proven ballistic missile designs, usually with additional upper stages, thereby simpUtying development. The Soviets have bad dismal results with their RAD on boosters developed solely for use in space.

Thc first booster developed exclusively for the Soviet spacehas been plagued with problemi since itsabout four yean ago. Its most recent tests have been successful, but there wereailures out of the firstttempts at launch. The only otherecificaliy designed for space missions Is Iheehicle which failed oataslrophically both times test flights were attempted. Wc do not know the reason for this poor record.testing o* design concepts, poor quality control during rnanufaoture. and less than vigorous prelaunch checkout procedures, may all be contributing fact on.

Aircraft. Tbe Sovietsery strong capability in both theoretical and applieddesign that has grown steadily inpersonnel, and capabilities. Tltey have demonstrated their capabilities in these fields by effectively designing all types ofvehicles. Since thehe So

victs have developed at leastistinct lighter types, ofre not in operationurrently are in test Eight bomber prototypes havehe same period, of which two were deployed and one is in flightarge number of helicopters, trans posts, and utility aircraft prototypes have appeared, of whichere produced and de-ployed In addition, the Soviets have devd-oped, tested, and deployed several variants of each basic design subsequent to its initial appearance.

Soviet theoretical work in aerodynamics is among tho best In tlie world. Whereas thc earliest research was concentrated on high speed phenomena, current pirjgrams aroways to improve aerodynamicat all speeds in association withvehicles. The strength of the Soviets in this general area grows outong-term and continuous effort which evaluates and applies research results only after extensive experimental work. Tbc Soviets also haveillingness to invest in environmental test facilities well in advance of actual needs. Research is characterized for tbe most part by low cost, simple solutions to specific design requirements. Thus, for example, large weights arc used to eliminate flutter in aircraft wings rather than lighter designs requiring complex and expensive construction; protruding(flow fences} which affect tbe flow of air over the wings are employed on swept wings.

Basic theoretical and experimentalhas been broadened in recent years. Presently, emphasis is being placed on thc technology of vertical, short takeoff andaircraft. Another aspecthich could see application in thc near future is the study of supercritical airfoils which reduce Ihe drag on aircraft.

Thc Soviets have mounted what they term "an all-out national effort" on laminar low techniques which increase the range of high performance aircraft. Western nations abandoned this technique after initial tests concluding that weight and powerpenalties would offset any gains. The persistent effort by tbe Soviets indicates that theyeasonable prospect lor successful development.

The most significant Soviet advances have been made in the area of appliedThc Soviets tend to concentrate on single purpose designs which facilitateand maintenance while accepting some performance penalties.esult,designs In the past were often crude, with protruding structual components. They have made great strides in overcoming these and other undesirable aspects of aircraftas demonstrated by the Foxbat aircraft.

Naval Systems. The effectiveness of thc Soviet approach to designing naval systems is difficult to evaluate, primarily becauseis based on Soviet, not Western needs and strategy. And since we have notthe Soviet decision-making process, we can only make informed guesses as to bow they perceive these needs-ursory examination of Soviet naval developmentreveals that thc quality of thc systems it has produced has been very unevento Western efforts. Some programsto be comparable, otlier surpasstechnology, and still others seem to lag Western developmentsubstantial amount.

In some cases, the Soviets havepursued lines of development which the West has abandoned or dismissed asIn others they have given up or notprograms which tbe US regards as

essential. It is tempting to conclude from all of this that Soviets erratic and poorly organized and that it is incapable of solving basic problems in such areas as afloat logistics or sonobuoy manufacturing.ore realistic judgment is that Soviet navalare merely systematic means of ptoducing thc hardware to carry out Soviet navalshaped by the Soviets' perception of their needs and the threat facing them.

The effort whichariety of anlisbip cruise missiles is perhaps thc best illustration of this thesis. Although the West abandoned its early efforts in the submarine-launched cruise missiie field, the Soviets have developed, adapted, and refined thc antiship missiie cencept to tbe' point where cruise missile-equipped subsurface, surface, and air units arc now thc backbone of the Sovietpurpose fleet. The magnitude ofirect result of Soviet realization that the major seaborne threat facing them inas Western carrier-based aviation, and that, as dramatically Illustrated by the Cuban missileounterstrategy is still required today and in tlie foreseeable future to limit Western options for intervention in distant areas. Thus the cruise missile effort is eminently justifiable in strategic terms. It is abo more economicalarge fleet of aircraft carriers, which did not fit with tbe traditionally defensive strategy of the Soviet Navy.

If cruise missiles represent an example of naval development patterned on peculiarly Soviet needs, there are other programs which were either abandoned, or never undertaken because there was no perceived need totliem. The Soviets have, as yet, no attack carriers or nuclear-powered surface warships, at least partly because there has been no doctrinal need for them.on-


power, largely sel(-sufficient In stra. tegic materials, with no overseas allies or long sea lines to safeguard, the Soviets couldon building an "anti-navyhey therefore steered clear of expensive power projection forces with great endurance, emphasising submarines and high-speed, and heavily armed surface warships.

ut not all of the differences can beaway in strategic terms. There are many requirements which are common to both East and West Some of these hare been adequately covered by Soviet development programs. Examples are the excellent naval guns and underwater ordnance which the Soviets have produced. Nevertheless, there is one key area in which the apparent lack of progress remainswarfare (ASW) detection. Despite frequent discussion of the magnitude of thc submarine-launched ballistic mrssile threat to the USSB and praise of Soviet antisubmarine forces, their hardware development, particularly in the areas of sonobuoys and moored acoustic detection systems, seems poor by Western standards. Thereumber of possible explanations for this apparent lack ofof which probably contribute to Iheir poor showing. It is possible that the Soviets have not yet decided on tlie final course of their ASW strategy^

Or thc

answer may simply be poorack of skilled technicalWhatever the reason, it is clear that Soviet naval RfltD is not always immediately responsive to stated Soviet needs, even those of thc greatest strategic importance.

Another apparent anomaly is theappearance of bulky, unsophisticated components in some of their newer systems. Whether deliberate or dictated by reason of economy, maintainability, or simply by poor management techniques which left large stockpiles of older components to becertain Soviet electronic and cruise-missile systems in particular have relatively simple components adequate to do the (ob, but without undue sophistication. Bather thanoor quality of effort, this might very well indicate the ability to plan andsystems designedission with minimum cost, while avoiding problems of maintainability, cost, and reliabilityupon more sophisticated systems.

Defensive Missile Systems. Soviethas produced some impressivein defensive missile systtans,of the priority given by the Soviets to problems of territorial defense, this has been an area in which thc Soviets have beenactive and willing to allocate large portions of available resources. In general, the Soviets have attached great importance to thc creationefensive missile force able to cope with an immediate air orthreat, rather than to thc developrnent of highly sophisticated systems able to cope with such threats anticipated for rhe future. The result has been the continued existence of defenses which appear aoacruotustic.are their defense systemaround Moscow in theor defense against tlie "thousand plane raids" that were typical of World War II, and the present Moscow antiballistlc missile (ABM) defenses which cannot cope with theaids now carried by US ICBMs. Thc Soviets therefore have had to devoteeffort to thc continual modification and improvement of older systems so as to main-


their viability in the (ace of morethreats. In large measure this is the resulteculiarly Soviet set of imperatives rather than the result of any limitations in Soviet technology.

When there haseed for high technology, the Soviets have generally met the requirement without reliance upontechnology from (he West. Indeed, the approach taken by the Soviets Id the design of air defense systems is an original one which spurned even the contributions made byscientists taken Io Ihe Soviet Union after the war. Most important in this regard was the early Soviet development of track-while-scan systems able to cope with many targets simultaneously.

Thc Soviets have tended to take simpler, less complex approaches to solving many of their defensive missile prablems. Rather than buildingle set of equipment the ability to cope with many types of threats, they have proliferated thc number of deployed sets of equipment and have filled Identified gaps in

coverage through the addition of new systems which become integratedomplete air defense network.igh level of technological sophistication has lioen shown in certain cases. For rwamplc, the propulsion most likely employed by their defense missile appears toybrid ramjet scJid-propdlant system tavestigated but not yet built by the US. And in their ABM system, (he operational flesibdity of tbe Calashwhich is achievedather elaborate fuel management scheme, has provided the Sovietsong-range interceptor not matched by anything yet flown in the US.

here are, to be sure, notable gaps In high technology in Soviet defensive missile systems. These include tbe lackigh acceleration ABM interceptor, and testing of look-down airborne intercept radars. These appear to he highly desirable developments which the Soviets should logically pursue In order Io meet requirements lliey must feel. Their absence though may bo the result of design choices not entirely apparent to us.



document was disseminated by rho Central Intelligence Agency, thiifor the information and uie of lhe recipient and of persons under his jurisdiction onto know basis. Additional essential drtiemlnotion nvay bo aulhoriicdithin iheir reipecKve departments:

o. Director of Intelligence and Research, for the Oopoitmoni of Stat*

Defense Intelligence Agency, for tho Office of the Secretory of

Defense and iho organization of tho Joint Chiefs ot Stall

Chief af S'ofl for Intell-gence, Deportment of ihe Army, for the

Deportment of the Army d- Assistant Chief of Novo! Operationsor ihc Department of the Novy

e. Assistant Chief of Staff, Intelligence, USAF, far the Department of the Air Force

of Intelligence, AEC, lor lhe Atomic Energy Commission

Director, FBI, for the Federal Bureau of Investigation

of NSA. far Iho National Security Agency

Director of Control Reference, CIA, for any other Department or Agency

This document nay be retained, or destroyed by burning In accordance with applicable security regulations, or returned to rho Central Intoll-gonce Agency by arrangement with the Office of Central Reference, ClA.

When this document is disseminated overseas, lhe overseas recipients mar retain iteriod not tn exeats of one year. At Ihe end of this period, the document should either be destroyed, returned to the forwarding agency, orshould be requeued of the forwarding agency to retain il in accordance with2

thle of this document when used separately from the test should beFOR OFflCIAl USE ONIY


While Hawse National Security Council Deportment of State Department of Defense Atomic Energy Commission Federal Bureau of Investigation


Original document.

Comment about this article or add new information about this topic: