Created: 8/1/1974

OCR scan of the original document, errors are possible









Inn poper wai prepared under'fice of Scientific Intelligence, Central Intelligencenquiries regardingpaper coo be addresied or to the Author.

Thii report has teen prepared under contract with the Central Intelligence Agencyhe views expressed ire tho* of the contractor and not necessarily those of tbe CIA.



The study seeks to accomplish five objectives of which the third is the central one:

To indicate what is methodologically permissibleecisionmaking process in the Soviet defense sector.

To present an overview of current interpretations thatbear on an important aspect cf this process: the roles and relationships of middle-level and low-levelparticipants. These participants includedefense sector watchdogs of the Soviet leadership,elements in the armed services and theministries, and designers, research scientists, and military representatives.

To provide critical assessments of these current The interpretations are criticised either for overscepping the bounds of what is methodologicallyor for failing to fully use the information they contain within these bounds, To conform to thestandards which the study presents at the outset, the study analyses are themselves presented only as.

To present the implications that the hypotheses hold for future decisionmaking analyses of particular Soviet weapon system developments.

To illustrate hew these implications, in fact, applyarticular weapon development case study and to demonstrate their utility in addressing such key decisionmaking questions as how new weapon systems are initiated in the USSR.

The study maintainsautious building-block approach is the appropriate way to analyzeecisionmaking and enable this analysis to eventually contribute to an understanding of largeras the dynamics of the arms race between the two superpowers. This building-block approach emphasizes tbe need to gradually accumulate enough knowledge to permit valid generalizations

about Soviwtecisionmaking. Such knowledge would accrueonsiderable number and variety of case studies of specific Soviet weapon system developments. The approach also emphasizes che need to be aware of conceptual biases that may result in the misuse of such information cnecisionmaking that presently exists and may be forthcoming in the future.

The study finds chat current interpretations may have fallen short in exploiting existing. so doing they haveassumptions that may prove misleading, particularly for analyses of future Soviet weapon system developments. Basic criticisms of these assumptions are given in the form of five hypotheses:

Current interpretations underestimate the management demands of modern complex weapen systems; thus, they tend tothe ability of the top Soviet leadership to supervise directly and competently ongoing weapon system developments.

Current interpretations have tended to abstract the defense sector from the larger Soviet political context. As aalitical control factor in Sovietecisionmaking, which may place importanton the watchdogs of the defense sector and affect design competition, has been sutraerged.

Current interpretationsicture cf designand the existence of an urge toward state-of-the-art technological advances in che Soviet defense seccor that is basically contradictory. To remove this contradiction for purposes of validly determining the initiator ofconservative or "advencurous" weaponariety of factors affecting the specific relations among designers, research scientists, and Service personnelexplicit attention.

Current views about the pervasive impactconstant shares" principle in the Soviet defense sector are The emphasisairly steady and "equitable" apportionment of resources minimizes the significant breaches of this principle that have occurred on both the Service and defense-industrial side of che equation. In so doing, such features of the defense sector as theof the defense-industrial ministries and the strained relations between miniscers and deputy ministers, which may be particularly important in the future, have been slighted.

5. Current interpretations ofecisionmaking tend coiew of possible interest-group relationships in the Soviet defense sector chat rests on incorrectabout the nature of the Soviet defense-industrial ministries. This viewne-on-one relationshipiven Service and defense-industrial ministry. And it ccmes close to treating the defense-industrialas monolithic entities, equating their interests in particular weapon systems with the interests of theServices. Consequently, efforts to determineactivity in the Soviet defense sector are left with little basis fcr ptcperly appreciating such keyas the differences in intensity of interestarticular weapon system, possible sources of opposition, the relative power of "interested" parties, etc.

The implications these hypotheses hold for analyzing particular weapon system developments ac'd up to two needs. Cne is to take into account certain organizational features hitherto slighted or obscured. The other need is to pose certain questions about these developments that seem to be absent frcm the considerations of current analyses or to have been given lew priority. According to the hypotheses, one might expect to find, particularly in the development of future weapon systems, such organ!zaticnal features as:

Mdtual assistanceumber of defense-industrial

A larger coordinating and management role for inhouseto take up the slack left by the defense sector

Possible efforts to play the various wacchdegs against each other on the part of those who are "supervised" by these watchdogs

A particularly act've role for military representatives and deputy ministsrs

Increased reliance on "outside" elements such as research scientists attached to the Academy of Sciences.

The questions that the hypotheses suggest should be posedrelateetermination of the initiation ofconservative or "adventurous" weapon system programsossible interest-group activity in the initiation of new weapen programs. With regardhe hypotheses suggest that questions

should raised about the relationship of and design shop personnel as well as about the relationship of design shop personnel and the Service customer. The questions are designed to elicit information concerning ccmm'jnications between the involved parties and their relative power and weapon interests. The questions that should be raised with respectre designed to bring cut differences in the intensity cf interestiven program at the defense-industrial ministry level and at the design bureau and research institute level of the participating defense-industrial ministries. These differences are crucial in determining whether Service and defense-industrial ministry interests are roughly comparableiven weapon system. They are crucial in determining whether subordinate elementsefense-industrial ministry might expect to find actual opposition to the program on the part of other subordinate elements or at the ministry-level. And, if this is so, the differences are crucial in determining whether the subordinate elements mightarticular need to engage inactivity with Service elements and try to override such opposition.

The organizational features highlighted by che hypotheses and the questions they elicit are "tested"ase study (not included in some copies for classification reasons) of the development of an "exotic" weapon system that broadly represents future Soviet weapon system development programs. The study emphasizesingle case cannot test all the hypotheses that have been generated or even thoroughly testhese hypotheses. Nevertheless, it finds thatdata gapshypotheses are useful and generally valid. They can help to determineew weapon system is initiated, whether interest-group activityart in such initiation, and hew effectively the development programodern weapon system is managed and its elements coordinated. In keeping with thepremises of the overall st-jdy, it is acknowledgedirm verification of the hypotheses requires similar case studieside variety of weapon areas.




I. necisionmaking in rhe Scviec Def-nseIntroduction

Larger Context of the Study: Scviet

an Approach

the Study

II. Composite Picture cf che Current Understanding of the Cverall SovietProcess

A. Organizations Involved in theProcess 8. Steps in the Decisionmaking Process C. Salient Characteristics cf the Scviet Approach toecisionmaking

III. Hypotheses Concerning Possible Heeded

Refinements in the Composite Picture of the Scvietecisionmaking Process

verstatedTop Political Leaders

ossible influancePolitical Control Factors

cr.tradiccingDesign Conservatism andTechnological Advances

ervasiveness and"Constant Shares" Principle Overstated

omplexity andPossible Interest-Group

IV. Summary and Conclusions: The Hypotheses and Their Implications for Analyses of Particular Soviet Weapon System Developments


Appendix--Case Studyarticular Weapons System

(available enly in certain copies of the paper because cf classification reasons)

. Apparent Basic Organizational EntitieseD Decisionmaking Process

FigureBasic Organizational Entities


Figure II- parent Basic Organizational Entities oftcisicnmakin; Process

Figure III-l. General Evolution of the DefenseMinistries


"Wc just do not have an adequate explanatory model for the Soviet-American arms race." (Ref. 1)


Itherished hope that an important by-product of the SALT experience will be an improved understanding by both parties of each other's strategic perspectives and domestic constraints. It ishoped on. side that the Soviet Union willesson ln strategic thinking that will at least help to put it on the same wavelength as the United States in calculating the rewards and penalties of future Soviet weapon deployments.* Implicit in thisunrealistic hope, is the admission that in the past the USSR's development of its strategic arsenal was not fully canpreh-nsible. standards.

The foremost casualty of such an admission is the long-held theory that the arms competition between the two superpowers can be satisfactorily explainedimple action-reaction phenomenon, whereby each side determines its weapon deployments an the basis of what is needed CO counter the actual or expected deployments of the other. In its simplest form, this theory focuses solely onstimuli to the development and deployment of weapon systems

An interpretation of the Soviet agreement to the ABM treaty, which holds chat che Soviets had already learned much up to and during SALT I, is discussed in Ref. 2.

and depends for its credibilityorrelation of the timing and characteristics of the. and Soviet "companion"

Because of its preoccupation with the international dynamic, and its assumption that the United States and the USSR can be regarded as mutually interchangeablehe action-reaction theory attachos little importance to an explicit examination of the similarities and differences between the weapon systems decisionmaking processes of the two countries. This particular deficiency has lad in recent years to the development of alternative theories. In the words of one student of these theories:

The newer proposition holds that the arms race behavior of the state-actors is determined not so much by theof threat, as by "the games that bureaucrats play." The range of models for the elucidation cf this proposition is formidable indeed. At one extreme,devise an action-reaction model wherein the principal actors are. Air Force, Wavyomewhat astrategic budget ceiling, and wich che Soviet Unicn performing an essentialegitimization function

While the conclusions offered by extreme proponents of chis new "bureaucratic" approach may prove to be as wide of the mark as the explanations of the arms race yielded by the acricn-reaction theory, this approachhole has at least highlighted the area within which the first questions about Soviet. strategic arms behavior should be posed. In other words, determining whether and to what extent international stimuli dictate che weapons policies of either the United States or the USSR requires an understanding of theof their respective weapon system decisionmaking processes.

Since issues of fundamental importance to the security of the United States are involved in these larger determinations, the effort to understand the decisionmaking processes in the Soviet defense sector represents something moreearch for knowledge for its own sake. lue to the enormity of the problem, it is useful to

keep In nine! chat on. side cf cht strategic equation, an understanding of. weapons acquisition process is still far frcoi complete. Cn the Soviet side, the lack of adequate data cn the workings of the defense sector makes che hopereakthrough in understanding the arms race even more rtmoce. The task ahead for those analyzing the Soviet defense sector may be likened tc an arch-eological dig. At the present stage of thesw scattered pet-sherds have been unearthed. y yield seme important insights, but surely not enoi'gn to permit an adequate understanding of theof the site's inhabitants. Indeed, there seems to be nofor the careful and patient digging needed to accumulateartifacts to cake such an understanding possible. The most that this study can reasonably aspire to Is toit more light on those artifacts that have already been frund and to indicate where future excavations may prove fruitful.


1. The Problem ot Coverage

If lt is prudent tc fight shy of any hard and fist general theory of the arms race between the two superpowers until both of their overall weapon systems decisionmaking processes are better understood, so too prudenceautiouslock approach within the onfines cf these processes cr. either side. Accordingly, this study docs not seek to provideeneral interpretation of the Soviet weapon systems decisicrmuking process, oromprehensive explanationecisionmakinghole within that process. Rather, it focusesingle facet cf Sovietecisionmaking, the roles and relationships of middle-level and low-level participants in. These are the people most intimately concerned withn the USSR. The middle-level ranks include leadership elements in the services and the defense-industrial ministries and the organizations and personnel that are charged with monitoring and coordinating weapon system

development programs. The ranks of the low-level participantsweapon systems 'esigners, resesrcn scientists, and service.

Two caveats concerning the coverage of this study must be noted. The first is that the line between middle-level and high-levelcannot be too sharply drawn. Clearly, no account of the monitoring and coordinating activities in Sovietan afford to avoid explicit consideration of the role of Dimitri Ustinov, for example, who, given his expertise and positionandidateof the Pclitburo, also is involved in high-level decisions onnd defense policy generally.

The second caveatore fundamental methodological problem. tudy of this sort can be approached in several different ways, each of which has particular merits and shortcomings. One way of examining the roles and relationships of middle-level and lew-level participants in the Sovietecisionmaking process would be to elucidate those roles and relationships as they arein the developmentingle weapons system. The merit of this ease approach is that itood chance of being asrigorous and valid as the peculiar nature of the Soviet defense sector allows any analysis to be. ase is carefully chosen, the analyst may be able to support his argumentssubstantial" amount of empirical evidence. The deficiency of the single-case approach is that its narrow focus may produce only very limited insights. Since the representative nature of the case cannot be taken for granted, there is no way of knowing whether and to what extent the insights revealed by the case may apply beyond it.

Another approach would be toeneral treatment of the subject matter. The roles and relationships of middle-level and low-level participants in Sovietecisionmaking could be analyzed simply by extrapolating from broad-ranging analyses of the overall Sovietecisionmaking process. The merit of this approach is that it keeps Che subject matter of the study in

perspective by relating it to the la-ger decisionmaking context and at the same time opens the way for an exploration of case may not reveal. The shortcoming of this approach is that the conclusions it producesar short of the methodological rigor and validity that direct empiricalmakes pcssicle. Since this general approach is at second remove from the evidence on which the analyses cf the overallprocess are based, it must assume that tha analyses have been validly derived from the evidence. Even if this assumption should beethodological dilemma nay be encountered. If the analyst seeks to make told extrapolations from the analyses, to convey as uny new insights as possible, he may go far beyond what the evidence underlying those analyses will sustain. If, on the other hand, he is cautious, the whole exercise sight be sterile. At best, his conclusions would be heavily prejudged by the analyses cf the overall decisionmaking process that he used as his point of At worst, his conclusions would be superfluous in terms of shedding new light on the workings of the decisionmaking process.

Given the difficulty of securing adequate data about the Soviet environment--and particularly about tha setting in which the decisions on Soviet weapons are taken--the goal of methodological purity oust inevitably be compromised if efforts to understand Sovietprocesses are not to comealt. It is important,for the analyst to be aware of when and where compromises are made and to what extent he is making them.

In thisethodologically ideal approach is clearly Such an approach would calleries of case studies of the roles and relationships of the middle-level and low-levelln decisionmaking on particular weapon system developments. It would also callirect analysis of the evidence bearing on many aspects of the overall decisionmaking proves* so that thereached in the case studies could be properly related to the larger decisionmaking context. Failing this, the study should

adopt an approach thatiddle groundhediscussed above. The coverage of the study should bebroad to shed new light cn the overall decisionmakingit should be sufficiently detailed to facilitate

tr. empirical verification of the study's conclusions.

Determining the appropriate coveragetudy of this sort is only one part of the methodological problem that has to be confronted. The other part has to do with the concepts that the study utilizes to illuminate the subject matter on which it focuses.

2. The Problem of Concepts

The task ofruitful and valid approachtudy of this sort involves an appreciationariety of potential biases to which almost any examination of Soviet weapons decisionmaking may fall prone. This particular susceptibility to biases of analyses of Soviet weapon decisionmaking problems is no doubt basically due to the murky nature of the subject. Because the data are both scant and spotty, it becomes tempting to apply uncritically concepts and notions derived from other settings and to generalize froa United and possibly unique examples. Although no approach can reasonably claim to be bias free, it is incumbent on the analyst to attempt to take into account, and make explicit allowance for, those biases that pose the most obvious challenges to the credibility of his For ease of discussion, the bias dangers that this study will be particularly wary of have been grouped into three categories: those that say stem. ind other Western (weapons)theories; those that may issue from notions of decisionmaking in the Soviet civilian sector; and those that may inhere inof particular Soviet weapon decisionmaking experiences.

a. Bias Problems Associated. Weapon Decisiorrraking Theories. Cf the various theories of bureaucratic decisionmaking that have come to the surface in recent years as putative successors to the action-reaction approach to analyzing. relations,

none has been more sophisticated or better articulated than those advanced by Graham Allison. In brief, Allison puts forth twoparadigms as devices to explain Soviet. strategic"organizational process model" and the "bureaucratic politicshe considers either separately or in tandem as inherently more fruitful and accurate than what he refers to as theolicy

Allison regards the rational policy paradigm as the implicit model used by. strategic analysts in the past. This model accepts individual governments, oer se, as the relevant policymaking actors and operates according to the dictum that "to explain an occurrence in foreign policy simply means to shew hew the jovernment could have rationally chosen that action* (Ref.. SO). Bythe organizational process model sees the foreign policies of states as being largely che product of the specific interests and established behavior patterns of the various organizations "on top of which government leaders sit" (Ref.. The bureaucratic politics model in turn focuses attention on the personalities who head up those organizations and views foreign policies as being heavily determined by the individual political skills of thesein pressing their own and their organizations' interests (Ref.

Allison's theoretical contribution, particularly his distinction between organizational interests and routines, on the one hand, and the personal interests and political skills of the individuals uhc head organizations, on the other, has potential utility forSovietecisionmaking, as will be discussed later. For the present, however, what is significant is that Allison's mcdelsumber of aethodological snares, he first place, the dataor the bureaucratic politics model seemformidable even on. side, co preclude easy transfer to the Soviet defense sector, about which data are in short supply.

Perhaps even more important: is the cautionary message that lies in an amplification of Allison's criticisms of the rational policy paradigm. The weakness of the rational policy paradigm is not only that it is questionable whether states invariably cr mainly act as monolithic policymaking entities guidedrational" calculus of what they seek to achieveis other states, but also that this modelniversal standard of rational behavior. Cne should at least be mindful that what is irrational in the light. experience is not necessarily irrational in the eyes of Soviet policymakers. Indeed, perhaps the very search. analysts for alternatives to the rational policy paradigm has been stimulated by the fact that. perspective certain of the Soviet actions in the past have appeared "irrational."'

Emphasizing this quite obvious point about national differences adds up to something quite differentere reinforcement of Allison's criticism of the rational policy paradigm. Because the dangers inherent in assuming that Soviet policy is. standards of rationality are supposedly avoided by using methods other than this paradigm, the broader relevance of this danger may be missed. Thatredible approach toecisionmaking must be wary of resting its conclusions on anacceptance of the applicability to the Soviet side of. organizational behavior or the political behavior of thein the United States who head up the organizations. Allison himself has given perhaps the most vivid example of what an incautious acceptance of his alternative models may lead to by statingo be sure) that, according to the organizational process model, the Soviet Union would be unlikely to come to an ABM agreement with the United States (Ref.

This estimate was largely based cn an apparently mistaken evaluation of the PVOs (Soviet National Air Defense forces) ABM capability and/ or the PVOs political clout.

Sy way of briefly illustrating che necessity for avoiding the pitfalls of hastily reading Sovietecisionmaking.ew additional examples follow. Michael Annacost, in his study. The Politics of Weapons Innovation, has detailed thebetween. Air Force. Arty in the development Of an IRBM. In order to apply Armacost's insights about checharacteristics and technological implications of service rivalry to the Soviet side, it would be necessaryare minimum to make appropriate allowances for the fact that the primaryfactor of this rivalry was that it was above allission role Any generalization about Soviet Service rivalry on the basis of Armacost's analysis would therefore obviously first have to take into account whether such ccbipetition wasto what extent--for the Sovietrograms being examined.

A similar caution is warranted in trying to illumine the behavior of Soviet "defense contractors" with insights gained from analyses. organizational behavior in. For example, one recent theory advanced as an explanation for the selection of weapon systems and weapon systems contractors in the United States holds that:

f one of the eight (aerospace) production lines is opening up, it willew major contractilitary Ordinarily, the new contract will be structurally similarollow-on contract. esignseripheral factor of the award

However accurate or inaccurate such an interpretation. weapons procurement incentives may be, clearly any transfer of this notion to Soviet soil would have to take intoost of organizational peculiarities in the Soviet defense-industrial environment, most importantly, the special funoing patterns and incentive structures builtigh-priority sectorlanned economy.

Finally,iew of the weapons svstem selection process that conflicts somewhat with the Interpretation given above, the authors

of another study-trong incentive by industrial contractors for "state-of-the-art" advances in weapons technology on thegrounds: "Contractor operating executives realize thatquality can seldom hurt the production follow-on potential of their product and will often help it Again, whether this sort of incentive makes much sense in the Soviet environment can of course only be judged on the basis of an understanding of the special economic dictates imposed by the Soviet planning system on the high-priority defense-industrial ninistriee.

None of the above words of caution about the ucilityderived models in analyzingecisionmaking is intended to suggest that theecisionmaking process is so totally unique aa to defy illumination by insights gained from analyses of. experience in this area. Clearly,. analyst couldeaningful description of the Soviet process, much less compare it with. process, if that was the case. Thenature of the purely technical factors impinging onecisionmaking would alone make this conclusion unwarranted. To be sure, just as it would be improperethodological standpoint to force Sovietecisionmaking. mold, so too it would be improper to. experiences out of hand because the Soviet weapons decisionmaking environment has certain special characteristics.

The principal utility of insights derived. weapons would seem to be in pointing to areasnd theshould be explored on the Soviet side but which are not readily suggested by the Soviet data themselves. This utility would be vitiated, however, if the concepts or theories taken from. weapons decisionmaking experience wereore prominent role as cither substituting for the data they are supposed to Illuminate or being the sole determinants of which data are to be examined. It would be no more credible to assume, for example, citing an earlier illustration, that the Soviet defense

industries shy awiy froa "state-of-the-art" advances In weapons technology, mainly or only because they have no concern for follow-on potential, than that theyontrary urge primarily because of suchcr the saae basic reason. defense contractors.

b. Bias Problems that Stem from Interpretations of che Soviet "Civilian" Cecisicrr^king Div iron-rent. If it is incumbent on the analyst to be *ary of reading Sovietecisionmaking. terms, it is also necessary to be cautious about depending cn information and concepts gleaned from Soviet decisionmaking practices in the civilian sector to provide the keys to the defense sector. There are basically two types of pitfalls to be avoided in this relying on evidence of the existence of certain organizational practices and behavior in the civilian sectorubstitute for direct evidence on the defense side; and being insufficientlyto the special conditions of theituation within the Soviet environment in utilizing civilian decisionmaking concepts to evaluate such evidence as does exist on the defense side.

The first of these pitfalls isroduct of the scant and spotty data about Soviet defenseunclassified data. onsequence of the data problem, it becomes tempting to assume that organizational practices discernible in the civilian sphere also apply in the defense sector. In his analysis of planning and management innovations in the Soviet civilian economy, forRobert W. Campbell attempts to demonstrate that Soviet space and military efforts havepillover effect on managementin the civilian economy. In the process, Campbell focuses on several areas of management innovation which, although notare at least being bandied about in the Soviet civilian economicsystems concept, new methods of quality control (the Saratov syscem,nd reliability assurance andmethods (Sovie; versions of Program Evaluation and ReviewPERT, and Critical Path Method, CPM) (Ref.assim).

Because of the evident applicability of aucb management and planning techniques to weaponsleastampbell deduces their widespread and successful use in tbe Soviet defense sector <Ref.

Whether or not Campbell happens to be correct is, of cccrse, entirely beside the pointhat matters is that if he is correct it is only by chance. ystematic and directof Soviet weapons decisionmaking practices, which is not evident in his study, his conclusions about the use of the managementin the defense sector are hardly warranted on the basis of deductions drawn from an observation of the emergence of thesein the Soviet civilian economic sector.

Another illustration of this methodological snare is thatby David Granick's extrapolation of Soviet defense andractices from informa;icn he has gleaned about Soviet behavior in the metal-fabricating industry. Taking into account what he has observed about the differences.ractices and Sovietpractices in the civilian metal-fabricating context, Granick contrasts. bias toward an experimental approach to development problems inith what he posits asof both Soviet civilian anda theoretic approach. As he puts it:

Cue would expect that research and development scientists in the Soviet Union would be considerably more likely than those in America to prefer the firstheoretic) approach, and that their first efforts at the solution of any problem would be theoretic. Such preference would be dictated both by the Russian cultural bias and by thefactor proportions existing in the two countries

atter of fact, the Soviets have given some hints of the use of such management and planning techniques in ttej defense sector. For one such hint, see. Skugarev and L. V. Xuden, Critical Path Planning in the Nayy. Joint Publications Research ServiceNo. bull. ase for the widespread use and success of CPM, let alone other similar techniques, requires much greater substantiation than the partial andof such Soviet spokesmen.


As vith Campbell's study, there may be scne truth in what Granick says aboutpproaches, Nevertheless, whatever credibility may attach to his conclusions, concerning the preferenceheoretic approach in the metal-fabricating industry, can hardly be extended automatically to similar conclusions about Soviet. Hot only dees his analysis pay scant attention to the special priority status and customer relations that affect the variousof the Soviet defense industrieshole, but it takes no account whatsoever of what mayarticularly importantof the preference for theoretic or experimental approaches in Soviet defensepeculiar role of Soviet designers and their relations with research institute personnel.

If the methodological deficiencies illustrated above can be said to stem primarily from the effort to find surrogates for direct evidence about the workings of Soviet, the second type of pitfall derives from the ordering and evaluation of directwhen it is secured. The principal potential offender in this case is the interest-group approach to Soviet politics. As ato the long-held totalitarian mooel of Soviet policymaking, interest-group analyses have clearly played, and should continue toseful and necessary role. They nave illuminated both the conflicts aixj bargaining that occur at the top levels of Soviet decisionmaking and, more particularly, the incentives and leverage that groups further down the decisionmaking ladder may have inthe decisions reached at the top.

At present, the interest-group approach, even as applied to civilian policy matters in the USSR, is confrontedasic task. In the face of objections chat it isargeorced transplant from the soil of Wescern pluralistic democracies, this approach must demonstrate its general methodological validity byufficient array of rigorous concepts that areto the Soviet setting.* in meeting thisumber

For one analysis that effectively points out the manifoldinvolved in transplanting Interest-group analyses tosee Ref.

of important analytical distinctions have been made that sake sense in the Soviet political environment." Thesecntrastmere interest-group existence and actual interest-groupa differentiation between issue-oriented and organization-orientedistinction between the Impact ofexerted influence and individualiscrimination between successful and unsuccessful timing of interest-group activicy relative to the involvement of the party leadership in the issues considered, and so cn.

However much progress the establishment of these conceptual distinctions may represent In analyzing some areas of Sovietthe interest-group approach stillong way to go before it can be relied uponona fide model for determining the nature and extent of Soviet interest-group activity generally. The methodological snare inherent In this approach issues directly from its partial successesew way of looking at Sovietlife. It become? tempting to assume that the conceptualand criteria that have been convincingly applied in some areas of Soviet policymaking are also appropriate and adequate in examiningthose involving the defense sector.

Beyond the slippery problem of even identifying interestin the defense sector on the basis of certain presumedand attitudes* there is also the problem of determining appropriate indicators of interest-group activity. An example of how interest-group analyses applied successfully in other areas of Soviet policymaking may prove misleading or inadequate in the defense sector Is provided by Stewart's evaluation of the repeal

These distinctions and others emerge from such analyses as those given in

Although overstated, Cdom's analysis of the tenuous assumptions cn which the basic distinction between party and military interests restsseful critique in this regard.

of production education in the Soviet civilian sectcr during Khrushchev's tenure. Cne of Stewart's key conclusions is that "individuals and groupings may publicly solicit support forpolicy proposals frca other individuals, institutions, or. To be sure, for anyone who has kept tabs on the different treatment accor-led military and foreign policy issues by Pravda and ?ed Star free time tc time, the publicity criterion can hardly be regarded as totally inappropriate ininterest-group activity in tha Soviet defense sector. It is certainly doubtful, however, given the Soviets1 unusually tight security cn weapon systems information, that public debate can give as accurate oricture of che relevant interest-group activity in this domain as it can inreas as educational

This is not to say that the workings cf the Soviet defensedefenseno reflection of che larger Soviet political, economic, and scientific environment. After all, the Soviet leaders in the civilian sector whom interest groups ray wish to influence are also the ultimate decisionmakers thatgroups in che defense sector have to try to influence aa well. Indeed, even if the arena of maximum interest-grcup impact and activity in the defense sector should prove to beevel below that of the top leadership, those who are "influenced" at that level are likely to be conditioned in their responses by concerns for what the traffic will bear at the top. Similarly, however special the situation of the defense Industries may be in terms of their priority status in funding, manpower, and the organizationalthes' oerait, that situation can hardly be regarded as having no characteristics in common with the civilian industries which

There is probably no more vivid indicator of the Soviet attitude on this score than that provided by an after-hours encounter between Colonel-General Cgarkov. delegate during SALT I. At this ercounter, the former admonished. for revealing information about Soviet weapons to the Soviet civilian delegates (Ref. IS).

operate with them undor the sane planning MCMnua. The point is, nevertheless, chat considerable caree taken both In tilling in rhe data gaps ir. th* Soviet defense sector Dy extrapolations froa the civilian side and in ordering and weighting data about th* workings cf the defense sector by utilizing ^models* that have been fleshed cut with concepts suggested by civilian policymaking,

In attempting toorkable balance between acknowledging and overdrawing the distinctions between the Soviet civilian and defense sectors* the errors associated with extrapolating frcm the Soviet civilian scene are probably easiest to avoid, There is an aspect of tills"negative"never* thelesSp requires as much conscious attention on the part ot the analyst as th* application of Sovietodels to. If one must exercise caution in keeping the focus of theapproach appropriate to the subject, on* nust ilso be wary of assuming that* against th* general background of Sovietuccess, inefficient or wasteful practices in th* civilian sector automatically denote the cppcsit* in defence.

c. Bias Problems Associated with Analyses of Parcicular Soviet Weapon Systems Decisions. Assuming that th* analyst ofecisionmaking exercises caution in applying concepts and theories taken fran. weapons procurement environment and the Soviet civilian policymaking setting, there is still another source of potential bias to which he must be attentive. If anything, this bias danger--the temptation to generalize from what may beeasonable picture of decisionmaking practices in the case ofweapon systems--may be the most insidious of all, preciselyit is embodied in interpretations that emphasize, and address themselves pointedly to, the special charact*ristics of the Soviet defense sector.

The uneven nature of thethe unclassifiedon the Soviet defense sector isjgt contributor to

this te .station. Taking comfort in che weight cf tha availableit is easy totance rha. implicitly assumes that the conclusions reached on decisicnraking practices in areas about whichata base is comparatively rich are generally valid for Soviet weapcns andecisionmakinghole. The weapons system area chat seems most likely to support this tendency to generalize is the area of Soviet military aircraft development. This is an area that has yielded perhaps the richest source of information on Soviet weapons decisionmaking, even cn an unclassified level.

The analysis by Alexanders indicative of the generalise. l- acknowledging his deliberaten aircraft, Alexander nevertheless feels justified in using it toicture of Soviet weapons declsicn-aakinghole.* Such features as endemic designendency to shy away from pushing "state-of-the-art" technological advances in the interest of achieving simple, reliable, and time-sensitive designs, and, above all, the preeminent role of the designer emerge as characteristicseneral Soviet approach to.

Obviously to suggest that sweeping conclusions of this sortreally stand on the evidence provided by one area of weapondecisionmaking is not to imply that many of these conclusions may not turn out to be substantially accurate. Rather, it is tothat, unless and until comparable information is available on the decisionmaking processeside range cf other Soviet weapon systems, the above view or any otherf Soviet weapons

This is clearly implied in the title of the study and is exolicitly set out in the study's conclusions

Even without substantialerious deficiency inmodel seems apparent that would disqualify ic aj "the"decisionmaking model in this regard--nc substantial account is given of che customer's role and impact.

In oealing with other aspects of aircraft decisionmaking,Alexander's model may be on safer ground. For oneof his views (as similarly expressed in an earlier study,, see


decisionmaking based principally cn the development of aircraft should be recognised as such.

Just as lt is .'mportanc to be wary of hasty generalizations free the evidence of aircraft decisionmaking practices, so too, it Is necessary to te cautious about generalising from practicesin other weapons areas. If,road methodologicalthe special character of the Soviet weapons decisionmaking environment must be kept in mind. weapons procurement and Soviet civiliano too it is necessary to be attentive to whatever conditions or circumstances may take the process associatedarticular weapons system special.

One must alsc be careful of letting "aircraft decisionmaking models" determine the selection and evaluation of data on theaffecting the development of other weapon systems. To assume, for example, that the heavy handhief designer must be present where no heavy hand can really be confirmed by the evidence, or to slight the evidence attestingetermined effort co push the state-of-the-art technologically would be to confirm the nodel at the expense of new insight. By the same token, It must be acknowledged that, if treated judiciously, "aircraft decisionmaking models" canseful service. Because they highlight the significance of certain features of the peculiar environment ofthey are likely to be of real use In opening up fruitful lines of Inquiry that would otherwise be hidden. If, for example, therolehief designer is nowhere in evidence, then the key question is at least raised as to which mechanisms are employed to perform his functions.

3, Methodological Requirements for this Study

In sum. the methodological problems discussed above suggest that the following general requirements should be met by this study: First, although the focus of the study is on che roles andof middle-level and low-level participants in the Sovietecisionmaking process, this aspect of the process is sufficiently

important to the overall process to merit broad coverage. Accordingly, the study should directly relate the roles and relationships of the middle-level and low-level participants to the larger decisionmaking context and shed as many fresh insights on the overall decisionmaking process as is possible.

Second, since broad coverage of the subject matter cf the study makes it necessary to rely heavily on data that are not directlyfrom independent empirical investigation, the conclusions of the study must be treated with appropriate caution. Indeed, these conclusions should be sufficiently detailed and specific to facilitate their subsequent empirical verification in case studies of particular weapon system developments.

Third, since weapon systems andecisionmaking processes can be expected to have certain corner, elements, especially technical oies, from country to country, the study should utilise, where possible, insights derived. weapons decisionmaking practices. However, because of inherent bias problems, it should lot rely on those insights to substitute for direct evidence on the Soviet side or to serve as primary criteria for selecting and evaluating direct evidence.

Fourth, since weapon systems andecisionmaking processes are likely to reflect certain characteristics of the overall national policymaking environment, the study should use, whereconcepts and information provided by Soviet civilianactivities. Again however, because of inherent bias problems, it should avoid filling gaps In the data on Sovietwith this information or applying those concepts, particularly those associated with the interest-group approach, without due regard for the special nature of the Soviet defense setting.

Finally, sinceecisionmaking practices observed in conjunction with some types of Soviet weapon systems may be applicable to other weapon systems as well, it should pay particular heed to the characteristics highlighted by those observations. Nevertheless,



because different characteristics may be suggested by the evidence on other systems, che study should be guided principally by the evidence using it to test and hopefully expand those observations rather than automatically accepting them as generally valid.


The purpose ofas been to give the reader asackground as possible before he acquaints himself with hew the Sovietecisionmaking process is presently understood to work, and before heritical look, through tha author's eyes, at this current understanding with respect to the roles andof the middle-level and low-level participants.

Section II represents an overview of the Soviet defensedecisionmaking process. Itomposite picture that condenses and amalgamates the findings of the best available, recent general studies of the process. The overview serves two purposes. The first and more general purpose is simply to establish the largercontext ln which the middle-level and low-level Sovietarticipants function. The second purpose is more pointed and critical. It is to acquaint the reader with the perspectives of the roles and relationships ofarticipants that are contained in current analyses of therocess. The perspectives provide the basis for the critical appraisals that constifite che cere of the study.

The critical appraisals are formatted in Section III as several hypotheses that bear on the roles and relationships of middle-level and lew-level decisionmaking participants. The hypotheses point cut problem areas in the interpretations of the overall process contained in the composite picture. They are generated by manifestand apparent gaps in the composite picture. The studies from which the composite is drawn do not reflect the implications ofand readily available evidence of the workings of SovietR&D, which in some instances these studies themselves contain.

The treatment of concepts and notions drawn. weapons


decisionmaking and Soviet civilian policymaking practices areand, in seme cases, inappropriate. Also, the studiesa hasty generalization of decisionmaking practices cbserved in particular weapon areas. Note that the term "hypothesis" is used deliberately. The analysis does not assume the level of validityirect detailed investigationide range of empirical data would make possible.

Aary of principal points contained in each hypothesis is given in Section IV. To enable subsequent verification of the hypotheses empirically, the specific implications cf each hypothesis for use in analyzing particular weapon system developments are also presented. In some cases, specific characteristics that should be noted in these developments are pointed out. In other cases, appropriate questions to be posed about these developments are enumerated.

A case study illustrating the application of these hypotheses to the analysisarticular weapons system is appended." The case is examined in terms of three key questions about Sovietecisionmaking: How are new weapon system programs initiated? How are the elementseapons system program coordinated and Does interest-group activity occur in the developmenteapons system? References chat are cited in the report, except for the Appendix, are listed after Section IV.

Since the information on which the case study is based is classifiedigher level than that of the body of the study, the Appendix is available only in those copies of the study similarly classified.


This section amalgamates and condenses analyses contained in several recent studies of the Sovietecisionmaking process. Unless otherwise indicated, the view given on particular aspects of this process should not be created as being specifically

As mentioned in Section I, the composite picture derived from the findings of these studies represents an overview whose purpose is twofold. First, it establishes the larger decisionmakingin which the middle-level and low-level Sovietarticipants operate. Second, it provides the perspectives on the roles and relationships of the participants contained in current analyses of the overall RSD process.

An overview of the current understanding of how the Sovietecisionmaking process works can hardly hope to do justice to all the interpretive nuances embodied in this Just asiew must inevitably simplify theof the process, so too it will understandably minimize the areas of disagreement among the various studies from which it is drawn. The composite picture of the overallprocess presented here is, therefore, not intended to convey the impression that all its points are uniformly endorsed. By the same token, areas of particularly vivid disagreement among current analyses of the process are noted wherever possible.


The basic organizational entities involved in th" process byew weapon system is developed fall roughly into three "customer,"and "coordinacor/oonitor/ultimate

1. Customer

This category comprises the Ministry of Defense and itselements. In one recent study, the basic function of the Ministry cf Defense inecisionmakingis summed up as follows:

The Ministry of Defense, the major consumer of the products oi advanced military technology, is at the sameentral participant in tha process that generates such production. The ministry and its subordinate agencies dominate in the framing of the technological choices of technologies tobepursued and weapons to be deployed...

The elements of the Ministry of Defense most directly involved in this basic function are located both at the ministerial/general staff level and at the Service level. The authority at the former level amountsinistry-wide responsibility for new weapons policy and coordination. Apparently, over the years, its actual locus Ms varied between elements at the ministry level and those at the general staff level. The same variation in locus has apparently existed for this top level authority over the Servics. For example, between thend, the author? .or new weapons policy and coordination was located at theevel in the Special Directorate of New Weapons under General M. I. Nedelln. However,3 to the end of, the responsibility appeared to devolve to the Scientific Technical Committee (XTX) of the General Staff under General A. V. Cerasieov. And this top level function, with respectithin tie Ministry of Defense, probably shifted again to the ministryen Gerasimov's formerGeneral N. N. Alekseyev, was appointed Deputy Minister ci Defense 0 Hmmmmmmmmmmmf


Each of Che five Services has under it che equivalent (orain Weapons Directorate, charged with performing the basic custorcer function for that Service. This function includes feasibility research relevant to new weapon sysceros, the formal generation of new weapon requirements, and the monitoring of the development (and production) and testing of new 'weapon systems. The pertinent directorates for the five Services are: Soviet Air Forces/ Aviation Engineering Services Directorate, Soviet Rocket Forces/Main Directorate for Rockec Armamentoviec Ground Forces/Main Rocket and Artillery Directorater Defense Forces (PVO)/Fourth Mainhoviet Naval Forces/ Directorate of Rockets and Artillery and Main Directorate forand Armaments.

The Soviet Navy, for example, has two such directorates with responsibility for naval RSD programs: the Main Directorate for Shipbuilding and Armaments and the Directorate of Rockets and Artillery

Subordinate to each directorate, and most directly chargedthe funccions of the Service with respect tocientific Technical Sommittee (NTK). It is incommittee thac che formal technical requirementew weapon system is formulated. The requiremenc ison the relevant RSD and producing elements of thetechnical committee also establishes and has direct chargeof military representatives. The teams are sent toweapons RSD and production carried out by the producer forumber of research institutes (Nils) and testtechnical committee jurisdiction, enable each Serviceew weapon system is feasible and to make sure itacceptable once it is developed. However, the extent ofcapability of the Services is by no means clear. Itbelieved char this capability is rather limited

The followingf che precise responsibilities invested in che Aviation Engineering Services Directorate or Che Soviec Air Forces illustrates che kinds of resources che Services have available to them in discharging their functions wiCh respect to new:

research generally applicable to requirements definition and generation

equirementsand development/ production monitoring

The Leningrad Military Engineering Academy ieeni A. Hozhayskiy, theAir Engineering Academy imeni Professor N. Ye. Zhukovskiy

The Aviation Technical Committee (itself)

Scientific Testing Institute of the Soviet Air Forces, che Scientific Testing Inscitute for Aviation Instruments, the Scientific ResearchInstitute for Aviation Medicine.

that to augment these in-house resources, thearticular Service nay also be able to drew on scientlscs and engineers from oucside organisationsossibly from ocher Services,omponents of tho producers,o serve in an advisory capacity on the directorate's technical committee

2. Producer

If the basic customer rolt in the Sovietprocess is essentially filled by the aforementioned elements of the Ministry of Defensej.the producer role directly encompasses the activities of eight other ministries and their subordinate elemenrs, and it indirectly involves at least six additionalas wellhe Academy of Sciences The eight ministries most directly involved and cheir general areas of weapon specialization are:


Defense Indus cry (HOP)

General Machine(HCH)

Aviation Industry (MAP)


Radio Industry (MRP)

Electronics Industry (HEP)

Shipbuilding Industry


Machine Building (MM)

Armed vehicles, artillery, rockets, small arms, and aircraft inftHM

Ballistic missiles, space launch systems, upper stages, and nonrecoverable spacecraft

Aircraft, aerodynamicdefensive missiles, recoverable spacecraft

Nuclear weapons end nuclear propulsion plants

C omm unica tion/na vi gaclonal/ guidance eauipmont,

Electronics components

Naval vessels, underwater weapons, fire control systems

Ammunition, explosives, fuses and projectiles, and solid propellants

Of these eight ministries, four may be regarded as system developers (MCP, HCH, MAP, and MSP) and three (MSM, HEP, and NM) as basically providers of subsystems and components. The Ministry of the Radio Industrypecial category. To some extent,roducer of radars, it qualifiesystem developer as wellubsystem and component producer.

Tne six other ministries involved indirectly inrovide subcontracted support to elements of the eight defense-industrial ministries. These are: the Ministry of the Automobile Industry; the Ministry of Construction, Road, and Municipal Machine Building; tha Ministry of Electro-Technical Industry; the Ministry of Electro-Technicalhe Ministry of the Chemical Industry; the Ministry of Instrument Building, Automation, and Control Systems; and the Ministry of Heavy Power and Transport Machine Building

The extenc and nature of che involvement of che Academy of Sciences tn Soviecre relatively obscure. While it is generally acknowledged that various research institutes of che academy have participaced inver tno years, some hold that this involve-none has been confined mostly to subconcracc work onnd has teen, in the main, quite minimal.

Producer elements most intimately involved in Cherocess are the research institutes (Nils) and design bureaus (KSs, CXBs, SKBs) subordinate to each defense-industrial ministry. The scope of such resources is base illustrated by che face chac che Ministry of the Defense Industry has had at times as many asesearch Institutes,esign bureaus, and upwardsplants in its domain

The function of che defense-industrial ministries with respect co weapons research has been summarized as follows:

The defense-industrial ministries plan, initiate, and conduct che applied research programs needed to support military systemesearch planning is based largelyinistry'sof future system requirements andtechnology. Research programs of tne defense-industrial ministries usually are initiated lnof future systems requirements. Only rarely is applied research initiated in response co explicit system requirements pfJ^BHiVViBnl

The same study cogently describes the relacionship between tbe research institutes and design bureaus, and their basic functions in:

Scientific Research Institutes play anrole in support of design bureaus and specific weapons development programs bynecessary applied research activities, preparing design handbooks and specifications, conducting cround environmental testing and prototype testing activities, and evaluating design suitability during the various phases of the weapons development cycle. actual development ot specific weapons systems,is che sole responsibility of tne chief designer who has been awarded the development program after competition 'with one or sore


Note, in this connection, thatefense-industrial ministry may be blessedealth of design bureaus (and researchs Illustrated in the case of the Ministry of Defense Industry, relatively few design bureaus tendarticular weapons field. When they do, they are usually headed up by elderly designers of long-standing andcompetence, such as Yakovlev, Sukhoy, and Anconov ln tha aircraft field, and Chelomey, Yang el, and Xorolev in Che area of ballistic missiles. The role of luminaries is not necessarilyto design bureaus char serve as prime contractors, responsible for turningomplete weapon system. For example, in the aircraft field, considerable status is also accorded designers who provide important subsystems for new aircraft, notably engine designers.

s funded in three different ways: out of the budgets of either research institutes or design bureaus of che defense-industrial ministries (apparently, the nose usualut of Che State budgec, or out of che Ministry of Defense budget which sometimes directly fundsHBmmmmml ttention given co che applied research activities of the defense-industrial research institutes is exemplified by theytarn which more Chanercenc of theutlays in che Ministry of the Aviation Industry went Co these institutes

3. Coord irator/Moni tor/Tic irate Decisionmaker

The activities in this third category oi organizational entities involved in the Soviet defensedecisionmaking process range free monitorinc, and coordinating ongoing programs to making che final decisions to go aheadoweven on occasion,providing the impetus to initiate the development of a'new system. Several organizational entities, either formally constituted and identified in the Soviet national policymaking hierarchy, orore shadowy or informal nature, are involved in these activities (Fig. Cn the formal side, they include che Council of Ministers, the Scate Planning Commission, the Politburo, and the Defense-Industries Section of the Central Committee Secretariat. Less formal, or less visible, participants are the Defense Council, the Politburo Defense Subgroup, che Military.Industrial Commissionnd che personage of D. F. Ustinov.

hole, formal and visible entities seem toa less prominent role char, that oi thei lkHBM ouncil of Ministers' role inecisionmaking seems to be quite minimal, although in the policymaking hierarchy it is the "formal-superior of one of the key decisionmaking bodies: the Military-Industrial Ccnmision (VPX). Similarly, although the State Planning Commission has the ultimate task of integratinglans into the overall economic plan, its impact on theecisionmaking process appears insubstantial. Furthermore, even the Politburohole, in its forael activity of giving the final stamp of approval on new weapon projects sesms toreat extent, by the decisions of its defense The Defense-Industries Section of the Central Committee Secretariat, headed by I. D; Serbin, is probably significant intaff support function for two "shadow" organizations, the Politburo Defense Subgroup and the Defense Council, Also, it apparently carriesonitoring function with respectrograms, however, the extent of the section's

capability in providing both these functions is rather murky. For example, there is no firm indication of the approximate size of the section's staff. What does seem clear, is that Ustinov has more power and influence than the section's head, I. D. Serbin; also, the section has evinced no capability In providing top leadership with systems-analysis type evaluations of the cost-effectiveness cf new weapons

The Politburo Defense Subgroup tops the hierarchy of informal or shadow organizational entities that apparently carry the major clout inecisionmaking. Its probable members include Brezhnev, Kosygin, and Podgomy. This subgroup seems capable of hammering out the essential top-level decisions onssues before the Politburo formally deliberateshole. The Defense Councilodern-day version of an organization that has gone through several incarnations since it was originally made prominent by Stalin. It probably comprises the members of the Politburo Defense Subgroup, tht Minister of Defense, and various ad hoc members such as the Chief cf Staff, and Ustinov, ard others. The Defense Council represents the highest-level entity in which the respective views of the top politicians and the military can be airedariety of military issues, including. It standsey avenue of military access to the top leadership. However, its significance in this respect may be somewhat attenuated by purely personal contacts between military leaders and the top Soviet politicians, and by Marshal Grechko's current statusull member of the Politiburo.

probably, the two most active participants in the coordinating/ monitoring/ultimate decisionmaker category are the Military-Industrial Commission (VPX) and D. F. Ustinov. The VPK, never openlyto exist by tne Soviets,upr'.ministerici body whose functions in coordinating and monitoringmong the defense industries and between the defense-industrial ministries and the customer (Ministry of Defense) are apparently carried out

on several levels. On one level, Che ability of the VPK toles function may devolve quite informally, simply on Cherelacionship between L. V. Smirnov, its head, and D. F. Ustinov. Smirncvrotege of Ustinov's and is apparently subordinate co him. More formally, two discrete levels of activity exist in theollective organ and an administrative staff.

The collective organ has been identifiedecentonsisting of the ministers of the eight defense-industrialand the Minister of Defense wich I. D. Serbin also included

However, Serbin's "membership" is somowhac

debatable* Because modern weapon systems maynpucs from several different defense-industrial ministries, the needorum in which top representatives from all theand the customer ministryt seems apparent. it is not an that clear as to how effective such meetings are, or how frequently the meetings occur. Two considerations seea to bear on this issue. The first Is that the ministers of the defense industries obviously wear two hats--a3 members of the VPK and as the principal bearers of responsibility for the fortunes of their particular ministries. onsequence, it is entirely possible that conflicts of interests hamper the effecciveness of che collective organ ol Bi^Hm The secondis chac clearly all the ministries are noc directly involved in every weapon system. This may mean chat only ministers whose ministries are involved in the developmentarticular weapon system attend the meetings in which that system is It ^ay also mean that such meetings occur in addition co the formal meetings of all the defense-industrial ministers. In either case, much of Che function of the VPK collective organ might be performed by restricced meecings, in which Che conflicc-of-interest problem is largely subdued.

The Vpx Alsotaff whose activities possibly help to override whatever stopgaps occur in the workings cf the coasis-sion's collective organ. This staff has been characterized as

The administrative staff is directly subordinate tohairman (I. V. Smirnov) and consistsirst deputy chairman, several deputy chairmen, administrative personnel, and technical specialists. The" staff is structured functionally with each deputy chaiman apparentlyajor Industrial area

The staff, accordingly, se*.ms toapability to effectively monitor and coordinaterograms.

D. F. Ustinov is usually regarded as the lynchpin of political control over theecisionmaking process in the USSR. He has this role because of his standingandidate member of the Politburotatus that L. V. Smirnov, I. D. Serbin or any of the defense-industrial ministers do notis long involvement and expertise in the weapons field (in contrast to the other top politicalnd his having "tutored" many of thm defense-industrial managers now subordinate to him in key positions.The role and function accruing to him have been described thus:

The role of the Party keyman, D. F. Ustinov,crisisnterjecting himself into the administratier, and coordination of thisat all levels,road range of programs, and in the greatest detail imaginable, rMkes him the de facto supra-itanager of all systemiVflVffaVmmmmmmmmm


To briefly illustrate the current understanding of the roles and relationships of the various entities in the overallprocess, it is useful to summarize what is thought to be the sequential steps by which the needew weapon system might typically be transformed into an item ready for serial production.


For the sake of convenience, these steps are broken down into two basic categories. The first includes tne part of theprocess chat ranges from identifying the needew weapon system to the Politburo decision (communicated to the VPK and the defense-Industrial ministries and their components) to proceed with system development. The second category encompasses the steps in the actual developmentm weapon system up to the decision to proceed wich serial production. Typically, steps in the first category heavily involve the Ministry of Defense and its subordinate elements. Also, other entities participate, either formally or informally. These are the VPK, Ustinov, relevant defense-industrial ministries, and members of the top political leadership particularly concerned with defense matters.

1. First Stage

weapon system need is identified.

method for meeting need isfeasibility is determined.

service sounds out MOD (Ministry ofStaff.

Top level military authorities consult Ustinov and Breihnev.

Weapons Requirement Draft (TTT) is prepared.

Military service command reviews,

General Staff and MOD review.

Smirnov and Ustinov review,

Appropriate defense industry ministries review.

VPK (Military-Industrial Commission) reviews.

Ustinov consults Gosplan (State Planning Commission) on resources.


The entity that takes the first step is probably the most difficult to pin down since, theoretically at least, the percepcion of the needew weapon system can emanate from any one of several sources. For example, it may be generated by officers in

the field (and then transmitted to the Servicet may stem instead from the Service leadership itself or from the ministry or General Staff levels in the top hierarchy of the Ministry of Defense. Alternately, this perception might come directly from one of the top political leaders. Atifferent level, it may arise from the technological possibilities comprehendedesigner in the defense-industrial sector. Presumably, all these possible sources, or even other sources, are reflected in the variety of weapon systems deployed by the Soviets over the years. However, assigning precise responsibility for the inspiration of any one system is an open question.

As discussed earlier, steps,re the basicof the appropriate Service weapons directorate, particularly its technical committee which can draw on the resources of che Service research institutes. re essentially informal steps. They are designed to determine the sore of reception the ideaew system can expect to meet at the higher levels of the Ministry of Defense and the opposition, or the kind of support top politicians might give the ideaequirement for it is formally generated by the Service. Basically, the remaining steps in this firstaim at securing formal acceptance of the new system requirement by appropriate elements of the military, political, and defense-industrial decisionmaking hierarchies.

2. Second Stag?

The second category of steps commences with the endorsement given the proposed new system by the Politburo Defense Subgroup after successful review of the system by the Defense Council. Thisoccurs after stepn the first category, in which Brezhnev would probably endorse the new system in rhe Defense Council review meeting with top military leadership. Council acceptability, under present circumstances, would basically determine the position taken

by che Politburo Defense Subgroup in presenting the proposed system for formal consideration by the Politburo. However, che subgroup would presumably enlist some staff aid from Serbin's section and possibly seek some advice from specially convened ad hoc groups of weapon experts before it solidifies its position. As with thethe acceptability by the Defense Council would probably favor acceptance by the Politburo, followed by its forma; approval.

Once Politburo approval is given, the locus of decisionmaking activity shifts co the Military-Industrial Commission (VPK). Again, after appropriate staff work (by the VPK),

rime system developer ministry/executor is assigned overall responsibility, based on product line charter, for fulfilling theof Defense) requirements. During this procedure, the VPK staff is probably in working contact with Gosplan (State Planningofficials. When major support is required from other defense-industrialegotiated protocol is signed at the ministry level. This protocol spells out interministerial tasking with expected completionnce signed, the protocolarty, Council of Ministers, and VPK decision and is binding on all participating ministries and facilities

The next basic step is for the "primary'1 defense-industrialto establish an Expert Commission, which exists for the duration of the entire development cycle. The commissionof representatives of the customer, research institutes, design bureaus of the "primarynd appropriate shops of other ministries expected to provide subsystems and components for the new system. Under the aegis of the Expert Commission,

equestreliminary designreproject study is formulated and, in tiieindustry, is levied or several design bureaus for competition

The designer or competing designers prepare the preproject study, utilizing design handbooks written by the relevai


institutes for the system in question. Performance criteria in che handbooks constrain the designers. But, presumably, the handb. jks contain information on the "latest" advances in che state of the art, which benefits the -iesigners. Military representatives of cheService, from che Ministry of Defense side, are called upon to monitor the preparation of the preproject study and, in che case of the "winning" designer, che subsequent development of the new system.

When the preproject study or studies are completed, the Expert Commissionecision co proceedock-up of the new system by the winning designer. After the mock-up by the winning designer isock-up Commission is established to evaluate it for the customer:

For military aircraft, the chairmen of these commissions are probably representatives of the Scientific Testing Institute of the Soviet Airhis commission thoroughlythe mock-up andetailed The report is forwarded through the Air Force to the responsible ministry's Expert Based on this report, the Expertdesignates one design for continued effort chrough the detailed design and experi-phase

The detailed design supposedly accounts for changes recommended by the Mock-Up Commission, and when it is finished the Expertconvenes for another review. Based on this review, thegives the go-ahead for prototype construction. Thisis usually performedreproduction plant attached to the design shop.

In the Aviationumber ofanywhererc constituted for flight test

After cests at che appropriate defense-industrial ministry'sa state commission is convened co supervise customer testing. The scate commission comprises members from the relevant Service and the primary producer ministry.

Ac che end ofocument containing detailed technical specifications on che new system is prepared. The document serves, subsequently,rimary quality-control device for serialor the system. When it is signed,rocess affecting che new weapon is essentially complete. The stage is then set for the high-level decisions that determine the production and ultimate deployment of che system. It is emphasized that throughout the actual development of the new system, it is closely monitored not only by military representatives, but also by Ustinov, the Military-Industrial Commission, and the Defense-Industries Section of the Central Committee Secretariat.


In rounding out the composite picture of the overall Sovietecisionmaking process, feecures eepbasited by currenc analyses of the process bear highlighting.

1. Supervision by Top Political Leadership

Effective supervision of new weeoon systems development is apparently exercised by the top political leadership. This iswell exemplified in che supervisory capacity of D. F. Ustinov. In thisuotation cited earlier bears repeating:

The role of che Parry keyman, D. F. Ustinov,crisis manager" interjecting himself into the administration and coordination of this system at all levels,road range of programs, and in the greatest detailmakes him the de facto supra-managar all System eleoent;

His effectiveness in ehis role is characterized, by the same scudy, as follows:

Ic appears that cheystem, if it is to operate with any effectiveness,the direct involvement andof the highest level of the Party;

be sure, Ustinov is hardly regarded asolitary agent in exercising this supervisory function. Two other individuals also seem to have particularly important roles to play: L. V. Smirnov (Chairman of the Military Commission) and I.erein (head of the Defense-Industries Section of the Central Committee But evidently, they too are generally considered to be basically subordinate and responsive to Ustinov. Smimov'sis Implicit in the characterization of the Hilitary-Industrial Commission as:

resided over by L. V. Smirnov buttoCentralComml ctee Secretary D. F. Ustinov

Similarly, Serbin's subordination to Ustinov is implied in the view


he Defense Industriess headed by I. D. Serbinlso serves Ustinovupport staff on defense-industrial matters

Ustinov's effectiveness and control are not merely the products of the particular subordinations described above, nor of the weapons expertise he is implicitly credited with by dint of training and long experience in dealing with defense-industrial matters. They areonsequence of -that is described as cronyism. He is able totop managers of the defense industry, because many of them either began their careers and developed under his tutelage when he ran the defense industry, or they shared early industrial work experience and perspective with him. For example, S. A. Afanasyev (Minister of General. V. Bakhlrev (Minister of Machine. M. Novikov (Foreign Economic Commission),

K. H. Rudnev (Minister of Instrument Building, Automation and Control. C. Serbln, L. V. Smirnov, G. A. Tyulin (Deputy Minister of General. R. Udarov (Deputy Minister of Generalnd S. A. Zverev (Minister of the Defensemong key defense industry managers, were at one time proteges of Ustinov or worked under him. These kinds of long-time contacts ensure that Ustinov knows the personal quirks, strengths, andof tne defense-industry managers. Consequently, much of his

interaction among top managers is based on cronyism

2. Apparent Lack cf Systems Analysis Type of Evaluation

The sort of effective control exercised by the top political leadership, via Ustinov, over the developmentew weapon system should not obscure another important feature of the Sovietecisionmaking process. As noted earlier, the top leadership's overall capability to determine the needew system and to monitor its subsequent development apparently does notystems-analysis type of evaluation capacity. As one study has concluded:

The apparent lack of an independent analytical capability at the top level for systematic and substantive correlation of military requirement with technological possibilities and economic costsutative weakness of the system that could leave the top political leadershipto decisionmaking inputs furnished by a

industrial bureaucracy below

The same study also concludes, however, that

he inability of the highest echelons of the Soviet policy pyramid to do in-house optimal planning has not inhibited and may indeed have abetted the determined expansion of Sovietpower over the last half-dozen years. Thus, though the bias of the system may have tilted decisionsro-defense direction, the results seem *uch chat the Soviet leaders, withew exceptions, can hardly view them as Indeed, through the leadership's eyes,

th* system's record may seem sufficiently ia-ressive co block major changes within

3. Conservative Design Philosophy

Another important feature of Che Sovietprocess worth noting is the one chat is revealed in the very design characteristics or Soviet weapon systems. For example, in Soviec aircraft inasically conservative design philosophy is said to predominate Several elements apparently converge co produce this conservative bias. To meet requirements of easy operation and repair for th* aircraft and to develop these aircraft quickly, there are incentives to keep designs simple and to

se subsystems and components already ln existence

which means, in effect, to

se as much off-the-shelf hardware as possible

The basic perceptions of th* designers themselves seem to be at the cencer of Chis conservative bias.

Because the designers mayighertechnological gamblesre notand because th* military hasthe conservatism of Sovietth* military probably are moreseek subscancja^_technological advances ry

Note ehat che weapon syscems area where chis conservatism seems most vivid is also the area where che role of designer luminaries appears most pronounced. Insofar as "such luminaries in Che aircraft field resisc bold innovations, th* very conservatism of their designs seems to testify to Chair power to fend off radical departures proposed by che relevant Services.

Presumably, one of the penalties designers My wish to avoid by sticking wich simple, conservative designs is che loss of future contracts that might come with Che failure of an avant-garde system. Again, this may be given particular weight in the area of aircraft development, since it is here also that design competition seems to be heavily featured. It has been noted that the competition that seems so endemic in aircraft development may not be as pronounced in the cesa'of"other weapon systems. For example,

While the sequential phases of aircraftmay in part apply to Soviet ballistic missile development progra-as, there is some evidence that che latter have rot settled into the routines and procedures followed in aircraft development. It appears, for example, chat th* Soviets tried and chen discardedompetitive prototype approach Co ballistic missile development that they have ton found suicable in cheffla,

Besides ch* conservacive biases on che part of eminent designers, note also that the latter are noc the only participants in theecisionmaking process who may be set in their ways. Indeed, an often noted characteristic of the major actors in the defense-industrial sector is the advanced age and long job tenure of the top managers-.not only in the case of such supermanagers as L. V. Smirnov and D. F. Ustinov, but with respect to the defense-industrial ministershole. Here Likely than not, the age factor in this segment of the defense RSD decisionmaking network complements and reinforces th* conservatism observed on th* part of top dssigners.

4. Pufhing the State of the Art Techrologicaily

By way ofomewhat more balanced picture of the impact of dosign conservatism on the Soviecs' overall approach to, noce chat some studiesovioc concern wieh pushing the scat* of che arc technologically:

Major decisions on basictrategy and funding-in the Soviet Union appears to be directed toward advancing the general state of

the art in certain military technology areas rather than being undertaken in support of specific weapon system developments. This somewhat open-ended approach to mucliendency tonded growth expenditures ^BBHammmmamV

Given the stated "separation" of urges to push the state of the art from the development of specific weapon programs, the presumed locus of the principal beneficiaries cf these urges is the researchof the defense-industrial ministries. Presumably, thesewould themselvesriving force behind the impetus in state-of-the-art advances as well as the beneficiariesolicy with this bias, probably, therefore, it is the defense-industrial research Institutes in particular that are being referred to in the following statement:

In some instances, the stimulus to explore new paths in weapons technology may derive in part from knowledge of the adversary's technological potential and activities, but the principal incentives are seen to be internally seatedesire to push the state of the art as far as it will go

S. "Constant Shares" Principle

Another significant phenomenon that apparently characterizes the Sovietnvironment is the "constant shares" principle. Its existence and impact have been describad as follows:

There is some indication that once established, an institutional claimiven share of resources tends to perpetuate itself,ind of sanction that makes it bureaucratically difficult to shift resources rapidly in new directions, and especially to cut back the accustomed level of allocation. This tendency may apply both to the military claimhole on resources, and to the internal pattern of allocation within the military establishment itself. Although there have been gradual shifts over time in distribution of resources among various elements of Che military, tne apparent persistence of relatively stable shares of resources, especially with regard to procurement expenditures,


suggests that the Soviet planning system and bureaucratic structure together nayuilt-in inertia resistant to the reordering of priorities frommajor changes in the internal or external environment |Hmm

While this description is given in terms of its application to the Ministry of Defense and its components, it can be taken to apply as well to the defense-industrial ministries. For example,eneral claim is implicit in the following statement:

When the interests of various powerfulcollide, some fairly simple solution to bureaucratic revalaries may be imposed to organizational peace within the system

Note that in the presumed broad applicability of the constant shares principle to the defense-industrial ministries, the elements ofrinciple are implicit in assertions of the continuity andof top personnel, witnessed in the defense-industrial sector. The notion that there is minimal "rocking of the boat" exists in this sector. otion, which is in the interests of preserving established managerial fierdoms and avoiding internecine feuds, is clearly implicit in statements that the defense-industrial ministers haveas powerful men with long tenures, have essentiallybasic self-sufficiency for their ministries.

nimize their dependence upon other sectors of industry and to minimize their career risks, ministers characteristically have sought self-sufficiency in all erdeavors essential to meeting casks VMmmmmmmmmmmffammV

S. Potential for Interest Group Activity

A final feature of the Soviet defense RSD decisionmakingworth highlighting is the inherent potential for interest group activity. As one study pointedly asserts,

It is probable that special interest ties have developed between the various military services

and those organizations in the defense-industry sector which develop and produce weapons. An alliance probably exists, for example, between the various working elements of the Ministry of General Machine-building, which probably is responsible for the design and production of strategicmissiles, and the militaryof the Strategic Rocket Forces

The study then proceeds to indicate other likelylation-ships between Services and defense-industrial ministries, based on their shared participation in particular types of weapon systems. These alliances include:

Ground Forces and the Ministry of the Defense Industry in the weapon areas of infantry material, armor, tanks, and rockets and artillery

Ground Forces and the Ministry of General Machinebuilding, also in the area of rockets and artillery

e Navy and the Ministry of Shipbuilding in the weapon areas of surface ships and submarines

Navy and the Ministries of the Aviation Industry and Radio Industries in the areas of long-range air, ASMs

National Air Defense (PVO) and the Ministries of the Aviation Production and Radio Industries in the weapon areas of fighters, AAMs, and SAMs

National Air Dei?nse and the Radio Industry in the area of ABMs

Air Force and the Ministries of the Aviation Industry and Radio Industries in the weapon areas of fighters, AAhs, bombers, and ASMS

Air Force and the Ministry of the Aviation Industry in the area of transports

Strategic Rocket Forces and the Ministries of General Machinebuilding and the Radio Xrdustry in the weapon area? of ICBMsHIHLbV

The obvious common interest of two such powerful and well-entrenched sets of bureaucracies as the Services and the defense-industrial ministries is to promote particular kinds of weapons systems. This could result in strong internal pressures being

exerted on top level decisionmakers tor the development andof new weapon systems. And such systems may not necessarily be warrantedrational" appraisal of the international However, this does not mean that incessant internecineoccurs between particular combinations of Service/defense/ industry "allies." For example, the constant shares principle is presumed to act as an important damper or. such rivalry

indeed, the invocation of the constant shares principleackdrop of virulent pressure to develop new weapon systems byange of powerful alliances would seem to have theeffect of producing steady, across-the-board increases in the Soviet arsenal. Also, co some extent, the rivalry between these alliances is probably held in check by the mediatory activities of the General Staff and Ministry-level elements of the Ministry of Defense in the competition among che Services. Presumably, D. F, Ustinov also mediates, especially in defense-industry competition in Chese Service/defense-industrial ministry alliances "J|

Finally, it is important to note that there appear to be fairly broad parameters within which interest group activity in defense RSD can operate. On the one hand, no matter che particular weapon interests that may separate one set of Service and defense-Industry ministry allies from another, they can an be expected co unite in the general cause of promoting military spending versus civilian spending. On the other hand, alliances on behalfarticular weapon system may cut across Service lines. For example, the long-range Air Force mayreater common interest with naval aviation than with Air Force tactical aviation. The latter, In turn, may find its greacesc common interesc wiCh the National Air Defense forces and elements of the Ground Forces


A detailed, direct examination of all the significant elements of the Sovietecisionmaking process is not feasible within the confines of this study. Consequently, the findings of several recent analyses haveigaaated toiew of the larger decisionmaking context to which the principalof the study relate. In developing this view, certain problem areas have been identified, particularly those that bear on the roles and relationships of middle-level and low-level participantsresearch scientists. Service representatives, defense-industrial ministers, defense sector watchdogs, et al). These problem areas are expressed in the form of hypotheses. Hopefully, they provide the means of more fully understanding the overall process ofecisionmaking.

Note that these hypotheses should be regarded as only hypotheses. Given the confines of the study, they are not the products of ainvestigationide range of direct empirical data on the workings of Soviet. They are rather broad Inferences. These inferences are based on what seem to be inconsistencies and anomalies in the composite picture of the Sovietprocess. They are also based on gaps in this picture that might be filled by taking into appropriate account certain insights drawn. weapon decisionmaking and Soviet civilianpractices, and the implications of available evidence of the workings of Soviet.


The extent ci direct competent supervision exercised by th* top political leadership over the development of new weapon systems is probably overstated in most recent studies of Soviet RSD.

The technical expertise and combined work capacity of th*designated as the principal political watchdogs ofre the chief considerations underlying this hypothesis. Th* composit* picture gives th* clear impr-sslon that cn both points th*olitical leadership has ample resources at its disposal to keep effective tabs on the simultaneous developmentroad array of new weapon systems. Indeed, it might be said that this impression contrasts starkly with the leadership's inadequacy on another score: it apparently lacks tha means to perform systematicanalyses in deciding whicn new weapon systems are worthy ofproduction, and deployment. While there is no reason to assumeweakness" in one area cf weapons RED decisionmakingleadsweakness" in another, there still appears to be ample grounds on which to question whether the top leadership is, in fact, as fully in command as it is made co seem.

1. Managementof Modern Weapon Systems

The impression of effective and pervasive political supervision

of en-going weapon system development programs is festered in a

number of ways. In tho most generalt is fostered by such

statements as:

Over half the Politburo members have technical backgrounds and are equipped by experience co handle problems affecting the

More pointedly, this impression is fostered by statements citedCMC emphasize che -xtenc and depth of D. F. Ustinov's personal involvemenCyriad of decisions affecting che on-going dovelopmenc of new weapon syscems. Whatever the background and incentives of either Ustinov or even the abovementioned Politburo


members to actually perform the kinds of tasks they are assumed to do, one can cast doubt on their respective abilities to keep close cabs on weapon system developments. In the case of Ustinov in particular, the effect of the customary characterizations of his role is to endow him with almost superhuman competence and

Of.course, this kind of characterization is fully in keeping with certain traits cf Soviet policymaking in an earlier period. Stalin, for example, presumably involved himself on occasion even in relatively minor decisions on the development of various weapon systems. Whether the "Great Genius of Mankind" possessed sufficient technical expertise to make effective decisions on those occasions is not possible to say. What is possible to say is that the nature and complexity of the technologies employed in modern weapon systems would seem to make it quite difficult for Ustinov to be personally as effective as he is assumed to be, in making similar decisionsroad range of development questions affecting such systems.

The kinds of demands for expertise and the workload burden imposed on those charged with supervising the development of modern weapon systems are particularly well presentedecent analysis of the Polaris program in the United States.

But the PBM (fleet ballistic missile) was too important, too big for the program's Director, Admiral Raborn, to rely simply on the good judgment of his Technical Director. Although giving assurance to others that the Polaris would be developed successfully and on time, he needed assurance of his own that this would be so. Thus, Admiral Raborn had to have independent sources of information through which he could check on Admiral Smith and the progress of the FBM program. The Chief Scientist and Engineering Consultant, men whose functions were never clearly defined or even distinguishable, but whosewas broad and whose knowledge of thewas extensive, reported to Admiral Raborn on program developments and opportunities. The weekly management meeting, which Admiral Raborn


naver missed, provided another check on che Technical Director as it offered anreview of the program throughreports of the Program Evaluation Branch, the tecnnical branches, the field offices, and then addition, Admiral Raborn was the beneficiary cf the conflict between the Director of the Plans and Programs Division and the Director of the Technical Division over budgets since it kept him informed of che alternatives for resource alloca- tions. He also kne* personally the presidents of all major contractors and travelled constantly to their plants asking quescions and making Finally, with the assistance of the Havel Ordnance Laboratory, hepecial advisory committee of technical naval personnel to review independently the program's technical decisions and test results

This quoted description is noc incended to suggest thacsupervision of Soviet weapon syscems need. practices, nor that Ustinov need be Che essential counterpart of Admiral Raborn in the development of each new Soviet weapon system or of anysystem. AdmiCtedly, the Polaris weapon system is highly complex. But the magnitude of Che management problems involved in its development does impose some kind of realistic perspective on Ustinov's position as the designated "supramanager"road array of modern Soviet weapon systems, some of which may be ascomplex as Polaris.

What may prove particularly misleading in assessing Ustinov's positionupramanager is that, quite aparC from che historical precedent of Scalin's personal involvement in weapon decisionmaking details, past circumscances in che Soviec defense RSD decisionmaking environmenthole, or circumscances related co certain kinds of weapon syscems, such as aircraft, campc generalization to the present and future. For example, in certain weapon system areashe overall marag' *nt problems may have been and perhaps still are considerably more simple than those exemplified by the Polaris system. If, as is asserted in the composite picture of the Soviec defense RSD decisionmaking process, chief designers,

particularly prestigious ones, are entrusted with much of theauthority inew aircraft, the burdenupra manager like Ustinov would be presumablyreat deal. che need for his intervention would appear to be mitigated by the previously stated reliance on off-the-shelf hardware for new systems and by the essential "self-sufficiency" of the defense--industrial ministry primary developer. Presumably, these factors would obviate many of the coordination problems associated with an extensive reliance on other defense-industrial industries.

It is not entirely clear as to what extent the reliance onand trusted designers, the ability to utlize off-the-shelf subsystem and component elements, and the sufficiency of in-house resourcesiven defense-industrial ministry characterize the broad spectrum of present or recent Soviet weapon system On technical grounds alone, It appears quite plausible to expect that the significance of such factors in facilitating Ustinov's supervisory activities would vary greatly according Co the kinds of weapon systems involved. In this regard, reliance on thedecisionmaking model" nay tond to obscure certain These are situations in which the role and competence of weapon designers are less pronounced than those of the aircraft designer luminaries, the ability to use off-the-shelf hardware items is not at the level past aircraft designs may have indicated, and the self-sufficiency of the "primary producer" ministry is inferior to that attributed to the Ministry of the Aviation Industry.

2. Nature and Limits of "Informal^ Decisionmaking Procedures

A relatedhich may be of prime importance inche comparative effectivenessupramanager like Ustinov, is the personal relationship factor. It seems reasonable to grant that informal ties between various significant participants in the Soviet defense RSD decisionmaking process have counted for much in making the processby facilitating the formaldescribed in the process or circueventing them when they did

noe work. In Ustinov's case particularly, it has been noted that one important byproduct of his career has been the accumulation personal tiesost or key defense sector personnel. on how beholden these people have been to Ustinov and could be relied upon to perform effectively and with demonstrableto him, his supervisory burdens have been doubtlessly eased.

While the general importance of the informal element in Soviet decisionmaking in the defense sector has been well and broadly recognized, the variable nature of this element dies not appear td have been given sufficient emphasis. Just as. practices described in the Polaris case help to focus or. the magnitude of management problems which the very technical complexity of modern weapon systems seems likely to impose on either Soviet. decisionmaking in general, so too. experiencen examining particularly che informal element in Soviet decisionmaking. The basic distinction drawn by Graham Allison ln his discussion of decisionmaking models, between the organizational process model and the bureaucratic politics model, is particularly germane (Ref.. In essence, the organizational model bases itsabout decisionmaking influence and likely policy outputs on an assumptionhe forcal organizacional entities of the decisionmaking process pretty auah tell the whole story. Once Che proper organizational entities involvediven policy issue are identified, their standard operating procedures are accounted for, and their links to other encicies are described, the model userseasonable right to expect the predicCionarticular policy out corse.

By contrast, the bureaucratic politics model is regarded (and rightly so) by Allison as being much less capable of entertaining predictive outcomes. While acknowledging the basic significance jf most formal elements of the decisionmaking process, this modelemphasizes the personality factorey variable. It recognizes that theiven organization on relevant policy

issues is likely co very over tiie and froa issue to issue, even if ic is involved with Che sane organizations in these issues. Although the organisations Tay remain che same, different personnel will occupy different key positions in these organizations, have different personal contacts Chan Cheir predecessors, and employ different political bargaining skills on different issues. These distinctions make organizational relationships inherently dynamic and the policy outcomes of these relationships much more difficult co predicc Chan the organizational process odel win Allow.

In the present case, Che particular significance of thesesuggests the necesslcy of avoiding the tacit assumption that che benefit of informal ties to Ustinov is either ubiquitous throughout the defense sector or invariant in any particular area of chis rector. To be sure, the organizational stability of che entities in the defense sector--and in che USSR as ache long tenure of many key defense sector personnel should be given their proper due in conrributing to Ustinov's informal effectiveness over time. Nevertheless, it would seem unlikely that the influence of che personal relationship factor would be uniform between weapon system Jcvelopments. While it may seem obvious when explicitly staced, one would certainly expect the significance of informal ties to vary considerably in facilitating Ustinov's functionupra-manager of Soviet weapon system developments. This variance would depend on the particular ministerial leaders, designers ecnvolved ln the developments.

3. Relationships and Support Capabilities of the Defense Sector Watchdogs

Finally, in assessing the ability of the top politicalto competently and directly supervise on-going weaponprograms, it is crucial to consider the assistance that Ustinov is likely rendered by key management figures, such as Smirnov and Serbin, and by staff support elemenCs. Clearly, to appreciate chat modern weapon systems impose considerable management demands because

of their technical complexity does not mean that Ustinov can only perform effectively if he personally possesses detailed technical knowledgeroad array of weapon areas. Indeed, lt is because he cannot possibly possess such knowledge, notwithstanding his long experience and undoubted diligence, that such assistance assumes particular importance. However, on this score, lt is also doubtful whether the composite pictureully accurate iupression. One problem here is that there is room to question whether the key management figures of Smirnov and Serbia are amenable to the supine subordinate rolea they are assumed toelative to Ustinov. In fact, they may clearly acknowledge Ustinov's superiority andseek to assist rather than obstruct his efforts. Nevertheless, as discussed in some detail below, in the light of Soviet politicalthe formal positions of both these men do not rule out, by any means, their capacity to take, or at least aspire to take,action.

Logically, the extent to which Smirnov and Serbin are personally capable of independent action would seem to affect the extent of assistance their staffs may give Ustinov. Note, in this connection, that the staff elements of the Military-Industrial Commission which Smirnov chairs do not appear to be as likely to gear their efforts to personally support Ustinov as do the staff elements subordinate to Serbin. This is due simply to the different nature of the two organizations. The representation of the presumably powerful ministers of all the defense industries gives the Military-Industrial Commissionapacityustification for an independent status greater than that of Sorbin's organization.

As indicated in the composite picture, some studies flatly assert that it is Serbin's staff upon which Ustinov principally relies for staff support. On the one hand, this makes perfectly good sense, given Ustinov's obvious need in this regard and given the realities of Soviet political protocol. arty Secretary "withouto to-speak, Ustinov's needizeable

personal staff to fulfill his (unctions would, presumably, putlass with only one other--General Secretary Brezhnev. On the other hand as noted earlier, it is less than certain that Serbinesignated Secretariat head of section would be entirelyin placing his staff resources at Ustinov's disposal. At the very least, then, the effectiveness of an arrangement in which Ustinov oust rely heavily on Serbin's staff, although possiblya limited, personal staff at his disposal, seems inferiorituation in which Ustinov could rely solely on his own personal staff.

Note, clso, that wherever Ustinov's staff resources lie the great size oftaff that management demands of modern weapon systems would likely dictate, is not inleast not in the composite picture. However, it may well be that on this score, as on others, today's process reflects the decisionmaking problems of an earlier day which did not make such possible deficiencies seem so onerous.


The Soviet defense PSD decisionmakingv be shaced bv Important political control factors that the composite picture does not take into account.

The first hypothesis considers the issue of effectiveness and extensiveness of direct top leadership supervision of Soviet weapon system development efforts. This oneeparate but related aspect of the political leadership's approach to weapons RSD decisionmaking: the question of political motivation. Have certain practices and certain organizational arrangements come into being in the defense sector because of the Soviet politicalconcern to keep the power of various key defense RSDin check? On the surface, it is tempting to point to theview of Ustinov's position asefinitive answer to

this question. However, the accuracy of this view eay be in doubt, which raises skepticlsci about how completely Ustinov, as thesupramanager of, is in fact relied upon to ensure political control in the defense sector.

1. Existenceeparate Political Control Factor

The first issue to be confronted is whether the composite picture gives sufficient emphasis to the political control factor, even in itsf Ustinov. While asserting that Ustinov enables the top political leadership to keep effective tabs onweapon system developments, the impression that comes across is that Ustinov's supervisory function is performed overwhelmingly for functional^ reasons that are pointedly military, economic, and scientific. In other words, he becomes involved in on-gcing weapon system developments to make sure that schedules are net, coordination problems are solved, parochial interests are overridden, etc. This secures needed weapon systems of high quality and reliability for the Soviet arsenal on time and without wasting economic and scientific resources.

Admittedly, it is difficult topure* politicalfrom such military, economic, and scientificnctionaln goals, since, to the extent that these goals are met, basic domestic and Internationa- political ends are obviously served as well. it seems hardly out of keeping with Soviet political practice and ideology manifested over the years to suggest that whilo these ends are important, the means are alsohe stark and relevant question is whether the Soviets have been, or would be, amenable to organizational arrangements in tho defense sector that secureigger, better, and more cost-effective weapons arsenal, if, at the same time, tho arrangementserious weakening of the Party's claim to omniscient guidance of Soviet society and control of the defense sector. If che obvious answer is accepted, then Ustinov becomes important for another reason besides that of his presumed effectivenessrisis manager and

expediter that enables che Soviet political leadership toigh level of military prowess without recourse to otherarrangements. He also becomes important, because heserves the political goal of asserting and validating the Party's necessary guidance of this important aspect of Soviet life at the same time.

In viewing Ustinov's positiontrictly political light, certain anomalous features of the organisational arrang. entsSoviet defenseare inevitably underscored. One is the apparent incorapatibilicy between the power, influence/and essential autonomy of the defense-industrial ministers and top designers, in particular, and the Party's concern to inhibit the emergence ofpower centers. It may well be, as is generally eaintained, that the prerogatives of these defense sector personnelong way toward explaining the success of Soviet weapons developments over the years. However, it is unlikely that these prerogatives are also in keeping with the political goal of Percy control and guidance. Vf;ether such prerogatives are by default or by design is che crucial question. It would bc incrediblo to assume that che Influence and authority of Che defense-industrial ministers and che top designers, which have been widely touted by Western analysts, would somehow have escaped the attention of the Soviet political leaders. But it would be no less incredible to assume that, having appreciated the infringement on Party authority inherent in this situation, the political leadership would not at least have sought some middle ground between heavy-handed control that would hamper theof designers and defense-industrial ministersaissez-faire attitude in which no real check3 were placed on the letters' power.

2. Political Control and "Functional' Effecciveness: alance?

Clearly, che very ability of Ustinov, Smirnov, and Serbin co prod, cajole, threaten,epresents this middle groundertain extent. Nevertheless, it raises seme questions. Has a

balancebeen struck between effectiveness and control' Are certain less apparent attempts at control in evidence? If thishas not been terribly effective, are there other inherent obstaclesoncern for the successful development of new weapon systems that prevent it?

Regarding the first question, which relates to considerations of the first hypothesis, political control of the defense sector (to the extent thai: considerable autonomy on the part of keyis checked) would seem to be deficient. s Is notase of the oft-cited evidence of the authority of top designers. It also seems to be endemic, for example, in the very relationship of Ustinov and his erstwhile proteges and cronies. As noted in the first hypothesis, it is doubtful whether gratitudeong acquaintanceship keep Serbin and Smirnov comfortably subordinate to Ustinov. imilar caveat would seem to apply to defense-industry leaders and tho like. There is no reason to suppose that once such individuals are in positions of formal authority they would not chafe under Ustinov's restraints and on occasion seek to resist them. While some individuals may be more acquiescent than others, the status of top designers and defense-industrial ministry leaders gives them an ability to exhibit what may be called the "Earl Warren Syndrome." As President Eisenhower discovered and later bemoaned, an appointeeigh-level position cannot be counted on to be unaffected,iscomfiting way to his patron, by his new position.

But if the positions of Ustinov, Smirnov, and Serbin aretorue balance between Party political control and weapon development effectiveness, it may well be asked whether other control mechanisms are utilised to curb the autonomy of top designers and defense-industrial ministry leaders. With respect to the latter, one such evident mechanism is the Military-Industrial Commission. Besides the personal authority ofwhen backed up bywould seem to be an inherent curb on the autonomy of individual defense-industrial Ministry leaders

inherent in their joint membership in the collective organ of the commission. Since the defense-industry ministers identifynot only with the parochial interests of their respective ministries, but also with the "larger" common interest of the def enae-industrial sectorhole, presumably some check, onautonomy occurs. For example, if the membersj Politburo canommon front despite their attachment to individual areas of interest, it would seem toeasonable analogy that theof the Military-Industrial Commission can do the same. from the standpoint of the Soviet political leadershiphole,ommon front might wellixed blessing. The forum'presented by tha commission would facilitateowerful common front to push the general interests of the defense sector as compared with the economyhole.

The other side t- the coin, which would seem to partiallythis danger, is that commission members do, after all, wear two hats. Except for possible threats to tha preferential treatment of the defense sector, or for beckoning opportunities that wouldlosing of ranks, members of the commission might, for th* most part, be expected to press for th* interests of their respective ministries. Such interests help to keep the defense-industry ministers divided. And in this respect they would appear to keep in check the power of these ministers to Infringe on the authority of the top political leadershiphole who "represent" bothand defense constituencies. By the same *cken, if one comnisslon is hampered in exercising effective curbs on individual ministers because of the parochial intercuts of its members, they would be left considerable leeway to deal with "pure" defense matters.

3. Design Competition and Political Control

The control mechanism represented by the Military-Industrial Commission seems fairly evident, even if the purely politicalbehind the top leadership's relationship to the defense sector is not explicitly acknowledged. But certain, less apparent,


attempts at cootr*ol oay ba brought to bear as well. There may well be another mechanism that operates to facilitate political control whose significance is obscured by the tendency to view it in military, economic, and scientific "functional" terms. This mechanism is competition, particularly design competition, which is frequently citedrominent feature of the Soviet approach toecisionmaking. Those who utilize the decisionmaking practicesin the aircraft areaacit model ofive such competition special emphasis. Howeverr as noted in the composite picture of the overall decisionmaking process, other analysts also point out that in the missile field, for example, tha extent of this competition may be somewhat less than that observed in the case of aircraft developments. Whileontrast, by itself, can hardly suggest any definitive answers to the question of what sort of attitudes the Soviets bring to the issue of competition inoes call toumber of provocative anomalies.

Competition probably varies in intensity from one area of weapon development to another. But what is particularly interesting, is that tho weapon area in which competition seems most pronounced is also the area in which design conservatism is especially featured and the role of designer luminaries is given unusual prominence. On thethe most tempting explanation for the competition observed in the Soviet defense sector is that it serves an evident function in achieving the Soviets' military, economic, and scientific goals. Presumably,.it induces individual designers to turn out the kind of designs that will permit the rapid deployment of reliable systems of high technical quality and with minimum waste in economic resources. In particular, competition would seem to be of special utility in dealing with the problem of technical uncertainty. This is assumed to occur in two ways. Tho first is represented by the inducements in the direction of simplicity and reliability presented todesigners by their need to compete. The second is by the

provision of backui systems, mado possible by the multiple approachesommon design problem occasioned by Che competition.

In fact, if competition in Soviet defense RED is mainly or solely prompted by such "functional" concerns, the intense competition observed in the aircraft weapons area would seem to be somehwat anomalous. Why should such competition be apparently greater than that in. the area of ballistic missiles, where the inherent problems of technical uncertainty, for example, would appear to loom largercertainly as largein aircraft development? Is* the touted conservatism of Soviec aircraft designroduct of the competition that occurs in this area, or is it, for example, derived from the long-standing perceptions that the designers bring to chis competition? If the conservatism of Soviet aircraft design is significantly the produce of various faccors only peripherally related to competition, it follows that competition is not as crucial to the reduction of technical uncertainty as might be assumed. Finally, it might be asked whether the prominent role of designer luminaries observed in the aircraft area i3 not also out of keeping with the presumed "functional" utility of competition. Why should competition be particularly Intenseeapons area that seems to have specially benefitted froa the long-standing competenceoterie of eminent designers?

It Is stressed that such questions do not add up to thethat functional concerns do not or should not provide the Soviec political leaders with significant reasons to utilize compecicion in. However, they do suggest that other "political"may come into play as well. One possible set of suchmay have to do with the political leadership's concern with essentially placating top weaponensure that each shares the action. Of course, this kind of political consideration would attest more to the influence of these designers than to theability to demonstrate its political control. For this very reason, though, it is doubtful that the political aspect of

competition is entirely confined to the concern of the political leadership to avoid roiling the waters. Indeed, the credibility ofolitical concern would not be established in the action sharing that joint participationiven design competition may

apparently small In brief, the fact as evidence of the

political clout of top designers. But the apparent fact thatlosers are not severely penalised can be construed as evidence.

This suggests that competition may alsooliticalaspect. Even though the penalties for the losersesign competition may not be great, it must be granted that the designers' desire to win may be considerable. After an, the bonuses and prostige that accrue to competition winners aro respectable For che political leadership, tho payoff in terms of political control isiven designer,op one, cannotcount on receiving weapons development "contracts" because of his prestige and past accomplishments. To the extent thatonsideration helps to keep himo important military, economic, and scientific ends of Soviet defenseare served. At the same time, this consideration would also seem toarticular designer more responsive than he would otherwise be to the authority exercised by the Party's watchdogs in the defense sector. Clearly, if the political leadership was willing to impose severer penalties on the losers of design compeitions, thiswould be even greater. Apparently, Che leadership is unwilling to do so, which assumedly testifies to che importance of Chegoals of Soviet defensethat might be negatively affected by such penalties. And in this respect, the functional goals of Sovietay help to give topodicum of political leeway.

4. Importance of Political Control Over elm Defense Sector Waccndogs

The compromises that tha political leadership seems constrained to accept in the operations of the Military-Industrial Commission and in the possible use of competition for political control purposes serveackground for an even more fundamental compromise tna"toccur in the defense sector. The principle of "divide andhich is ieplicit in the foregoing comments on the political utility of design competition and the parochial interests of defense-industry ministers, may also be evidenced in the very arrangements involving the watchdogs of the defense sector. In other wcrds, just as the top leadership utilises various mechanisms to keep the power and influence of designers and ministry leaders circumscribed, so too it may utilize other means to keep the watchdogs of the defense sector appropriately responsive to Party authority. In the light of the common picture oi Ustinov,tatement is admittedly not easy to accept. It is particularly hard to accept if Ustinov's position is assumed to be adequately defined in terms of the military, economic, and scientific goals of the Soviet political leadership, and that he helps to secure these goals with consummate effectiveness. However, if the basic goal of maintaining political control over the workings of the defense sector is explicitly taken Into account as well, it seems reasonable to entertain the possibility of Ustinov's having to operate under certain significant constraints.

The relevant issue is not the well acknowledged politicalthat Ustinov still lacks full Politburo membership. I: is whether the organizational sotting in which Ustinov operates as tha top "administrator" of the defense sector reflects certain built-in constraints on his authority, whose removal would facilitate his carrying out the political and "functional" responsibilitiesto him. This raises the question of whether such constraints aro in keeping with general Soviet political practice. ursory glance at the Soviet political scone suggests that the basicof divide and rule has been employed to keep the politically


powerful in check. At the test obvious level, the parallel lines of authority embodied in the initial establishment of separate Party and Governmental hierarchies reflect this principle. The deliberatethat Stalin fostered among his principal lieutenants reflects it. And the division of labor between Kosygin and Brezhnev upon the ouster of Khrushchev also reflects it.

rfh,ie"I: can hardly be claimed ia the light of Stalin's, Khrushchev's, and Brezhnev's ascendancies that this principle has been uniformly and successfully applied, it seems highly questionable that :ts applicability toefense sector would be ruled out. If the Soviet leadership is appreciative of the need to invest Ustinov with sufficient power to dial with the powerful Marshals and ministers of the Soviet defense establishment, as viewed by most, it is hardly likely tha; tha leadership has remained! unappreciarive of therisks that such power may carry.

In fact, the very arrangements utilized for tho top leadership's supervision and control of the defense sector workings reflect an appreciation of the neea to irhibic Ustinov from accruing more power than i* essential to his political and other tasks. In some sense, the triumvirate of Ustinov, Smirnov, and Serbin as the principal watchdogs of the defense sector is an aberration from "standard" Soviet political arrangements. Indeed, from the standpoint of formal organizational requirements, Ustinov is the odd man out. In terms of the parallel governmental and party hierarchies, the roles are filled by Smirnov and Serbin, respectively. The fact that Ustinov has an added role over and above those of Smirnov and Serbinthat it is more important than either the formal Party or governmentto the extraordinary importance the Soviec leadership attaches to defense matters and to itsof the needupervisor with Special authority to ride herd on the powerful interests in this sector.

At the same time, as noted in che firsc hypothesis, Che very fact that Ustinov is constrained to coexist with Serbin and Smirnov

restricts his capabilities. While presumably benefitting from their aid, Ustinov -aust also invariably contend with the implications of each of these cronies having positions that give thea each statusertain independence fron his. It may well be that this is nota burden that top leadership is willing to bear, but that it is in fact desired by it. Since the positions of Soimov and Serbin permit thea"some independence from Ustinov, the top leadership is not so readily dependent on Ustinov's evaluations ofnd other defense matters, even though on most occasions it presumably would defer to his judgments. The existence of these somewhat muddy lines of authority miynt well be acre preferable to the top political leadershipituation in which Ustinov was truly the czar of the defense sector.

A situation in which Ustinov operatedingle large staff at his personal disposal, and without other watchdogs withformal status to deal with, mightore effective way for political leadership to keep the Marshals and defense-industrial ministers in line and attain the functional goals ofhan present arrangements permit. A; the same time, it would make Ustinov (or hilotentially formidable political power towith. Therefore, an acceptance of some slack in the political control exercised over the ministers and top designers in the defense sector, and in the capability to urge the sector to achieve its military, economic, and scientific goals mayrice the political leadership is willing to keep paying to avoid the emergence ofolitical contender.


As presented in the cocpcsitepicture, the emphases on bcth Soviet desiTr. conservatismesponsiveness to state-of-the-art advances in technology are basically contradictory.


In its treatment of the Soviet approach to weapons technology, the composite picture obscures several aspects of the workinginoviet defense HO environment. These aspects will help determine the source of new weapon system development initiatives. Individual studies from which the composite picture is drawn differ somewhat in the weights they assign to design conservatism and to the impulse toward state-of-the-art technological advance. But none seems to acknowledge that these designated characteristics of the Soviet philosophy of -cCoons development are difficult to reconcile.

1. Contradictory Characteristics Eephasl^.ed in Current Interpretations

These characteristics seem particularly contradictory in terms of the assessments given to the functions of various relevantentities in the composite picture. For example, it is generally held that of the organizations from which weapon system initiatives might emanate, the Services are probably the major source ofin general and of radically new weapon concepts in particular. Also, it is held chat while designers may alsoole asof new systems, they are likely tomaller roll than the Services and be more conservatively inclined than the Services in the systems they seek to develop.

Other widely held interpreta-iors also "jeer on the issue. For example, designers, parcicuiarly eminent ones, are. conceded to have considerable personal influence in the development of particular weapon systems. Also, the Soviets are assumed cotrategy that aimsetermined advance in certain broad areas of military technology. This strategy is essentially separated from specific weapon system development efforts. Accordingly, research institutes appear to be the logical direct beneficiaries of this strategy and might be expected, therefore, to be advocates of state-of-the-art advances. Finally, with lospect to the locus of research institute activities, it is generally held that the bulk of military research occurs in tho defense-industrial institutes. The research capability of thenly adequate for performingscudies on weapon systems and assisting in tests.


Consideredhole, these views seen tcecided influenceetermination of iniciadve sources for new weapon systems and on the overall question of the likely technological "conservatism" or "adventurism" of systems. They contribute in rwo ways. They provide flat answers. But equally important, the views embody contradictions. And these contradictions can help pinpointtful areas tor farther inquiry. These contradictions are best revealed through juxtaposing some of the views.

The Services are assumed tcrime source of new and "technologically adventurous" weapon system ideas. But, at the same time, they are acknowledged as havinginimal research cape-city from which such ideas would most likely emanate. On the other hand, the defense-industrial ministries are conceded to have most of the Soviet Unicn's ailitary research facilities, which, presumably, in turnarticular interest inilitary research strategy of state-ci-the-art advances in broad arear af Kihnclogy^ Yet the weapon system initiatives the defense-induS"rial ministries exert are held to be both fewer and morehan those emanating from the Services. Note, in particular .rat thr Ministry of the Aviation Industryefense-industrialh possesses some of the most prominent andesearch entities (TSACI ands also widely touted as exemplifying Sovietesign conservatism. The urge "to push the state of the art as far as it will go") is one explanation of the weapon system choices that the Scviers make. But can this urge be easily squared with the notion that the eminent designers of Soviet systems are mostly conservative in their weapons philosophy with considerable influence in determining the characteristics of tha systems they develop?

At the very least, these contradictions suggest that the broad views which have revealed them should be more modestly stated; also, certain conditioning factors that affect these views should be given explicit attention. The contradictions drew particular attention to

two sees of relationships that would seen to bt crucial in two aspects: determining the principal initiatorew weapon system, and determining whether the system, is likely to be"conservative* ox "adventurous." The first set involves the defense-industrial design bureaus and research institutes. The second set involves the defense-industrial design bureaus and the Services.

2. Design Bureau and Research Institute Relations

In the case of the first set ofuaaber of important considerations seem necessary before the impact of an urge to advance the state of the art technologically can be judged. communication, and power are the key considerations. Asconstrued, designers and research institute personnel have almost diametrically opposing incentives with respect to weapons technology. To meet such touted criteria as timeliness, reliability, and operational simplicity in serving the needs of the military customer, designersasic incentive to shy away frcoadventurous designs. To the designer, such designs are likely to carry greater risks of failure than those of conservative ones. Also, they may make it more difficult for him to meet For example, if off-the-shelf components and subsystems are harder to use in adventurous systems, coordination problems withand subsystem developers (particularly with those in other defense-industrial ministries) are more likely to arise.

By contrast, research institute personnel seem somewhat insulated from these risks. Their funding is apparently separated from the "contract" relations between producer and (ultimate) customer

chat affects the situation of the designers.* Moreover, since their funding is determined by an overall strategy of state-of-the-art advance in military technology pursued by the Soviet leadership, they are likely toositive interest in demonstrating theirto the strategy. Thus, researchers working in particular areas would especially sees prone to demonstrate that the next crucial breakthroughs were likely to occur in their areas and that,they deserved special funding consideration.

Given these op-osing incentives, even the most effectivebetween research institutes and design shops would appear inadequate in making sure that state-of-the-art advances werein the designs for new weapons. As It is, it is impossible to assume that this communication is, or has been, all that effective. Considerable .rJoraal contacts presumably occur between designers and research scientists. But the impression conveyed by the composite picture is Chat design handbooksery significant "formal" role in communicating new research findings to the designers. Apparently, the primary function of tne design handbooks is to ensure that the designerariety of technical standards in producing his design. It may be questioned, then, whether these handbooks can adequately fulfill their otherapprising the designer of newthe saae time. And, it certainly seems questionableeans of communication, which the designer Tay basicallyas infringing on his authority, would be the test way to inform him of new research findings and get him to appreciate them, against the backdrop of his basic disincentives to pursue adventurous designs.

Mote that the funding for specific weapon programs may come from the State budget for tba defense industries and not from che Ministry of Defense budget. This consideration is held to enhance Che likelihood of military receptivity to weapon programs that the designer may initiate. Prom the designers' standpoint, whether the funds for specificcome from che State budget or froa the Ministry oi Defense to the defense industries would not seem toaffect their interest ln winning specific "contracts" and avoiding risks in doing so.

tcMnisa of coercion, design handbooks raise da final key issue chat bears on the relationship between designers and researchrelative power. Since the designers are constrained to conform to the standards set by the design handbooks, research scientists (who, by the way, are represented In Expertasay beignificant measure of power'over the designers. 3ut it is doubtful that this extendsapability of forcing designers to turn out technologically adventurous designs. Indeed, because of the prominence accorded designers, their relative power would seem, in general, to be clearly greater than that of research scientists.

A judgment on the impact or an urge to advance the state of the art technologically is conditioned, then, by: the opposing incentives of defense-industrial research scientists and designers, theeffectiveness of the communications between them, and theof power on the side of the designers. And thesemake it difficult to generalize that an impulse for state-of-the-art advances in weapons technology, emanating from the defense-industrial ministries, gets reflected in the weapon designs produced by these ministries.

The difficulty ineneral case, however, does not rule out the possibility ofarticular case. For exdople, the incentive issue may vary considerably, depending particularly on the designer Involved. ounger designer may be more amenable toadventurous designs than might an older one. esigner would be confrontedompetition situation might also Influence his attitudes about incentives. In terms of thediscussion in the secondesigner's success or failure in previousthe penalties he suffered because ofalsoole. If his penalty was light and if his design was conservative, his tendency to stick withdesigns might well be reinforced. On the other hand, if he had been heavily penalized "cn an earlier occasion for losing with a

conservative;hadumber oi previous competitions--the designer eightochnolcgicaiiy advencurous fasign as the only way ro recoup his fortunes. Finally,eluctance to incur the coordination, problems of utilizing 'new" subsystems provides designers with an impetus co use conservatism, it is important to consider weapon system areas in which, by the very nature of the-system involved, off-the-shelf use of component hardware amy rot be possible in any event.

Similarly, in the communication issue, it may well be that in some weapon system areas the informal relations between designers and research scientists drastically reduce the significance of design handbookseans of coomunieating new research findings. Also, the comparacive powsr of defense-industrial research scientists and designers is likely toreat deal. o means is it proven that designer luminaries have dominated all areas of weapon Ncr is it proven that they will dominate in the future. On the whole, research scientists do not seem to be as prominent as designers in Soviet. But in some cases, they may be as prominent or even more prominent than the designers.

Other factors also deserve consideration in determining weapon system initiatives and the likely conservatism or adventurism of Soviet weapon systems. One is the tie between component and subsystem developers and research institute personnel. This relevanton the defense-industrial side should be considered whendesigner-research scientist relations. Even if the chf ef designer of an overall syscem is conservatively inclined and opes for che use of off-the-shelf hardware, the hardware aay be quite sophisticated because of Che advencurism of its developer in an earlier period. It may not be as simple and reliable as the designer may want it to be. But the hardware may be crucial enough to his design CO constrain him Co accept it, thus causing him to be inadvertently less conservative in his design than he would otherwise be. The point, therefore, is thateneral case for th* impact of state-of-the-art impulses

in technology (embracing from the defense-industrial ministries) on weapon designs cannot be convincingly made, based on the contradictory views contained in the composite picture, such an impact cannot be ruled out in particular Instances. And to evaluate this impact requires an explicit look at the various key aspects of thebetween the relevant research scientists and designers outlined above.

J. Designers and Service Customer Relations

Likewise, this requirement is evident in the second sac of relationships mentioned earlier between designers and their Service customers. Indeed, if the case is Co be made that certain Soviec weapon systems are che product of an urge co advance the stace-of-the-arc in weapons technology (emanating froa che defense-industrialt is necessary toonstrate not cnly that Soviet weapon designers respond Co chis urge but chat, having responded To it, they can Chen "sell- their idea tc an appropriate military customer. InformationJPj^HemmmmmmmmmBmVmfjmlmVmYalhe central role thac should be assigned co che militaryof Che various Services.

mfirmatlon of the general view that the Services are more prone to accept technologically adventurous syscems than the designers themselves. The military de-%jr*;r, on

1-Mid to he high. Citing the extensive avionicsWestern aircraft, military officials urgedotdemand may indicate

In the

ous aircraitj

acquisition environment, the phenomenonreferred to as "gold-plating" and need not'be synonymous with technological advenrurism. Therefore, for the Soviets, such gold-placing should probably be viewed as IncUcac-ing anosc^hy in Che broad! the term. The push coward the heavy use of avionics in Cho example given above does not necessarily mean that sCate-of-the-art advances in avionics are required. However. it does Man chat technologically adventur-

1 (pay be needed to accoTtnodatc these avionics.

iomparecively "adventurous" accidie fry the military, it alsoreflect deficiencies in the research support directlythem, also attesting to the Generally held view on thisexample,aecanQ' 'or 'vionics did not:

account for such obviously important conditioning factors as the weight iifferentj.' between Soviet and Western equipment IHBm-'"

If the Services are generally receptive to idventurousit does noc necessarily follow that theandesigners in this direction. Soviet weapon designs canas conservative. Also, designers can beconservative in terms of their weapon philosophies. Bothunderscore the general impression ofand authority in weapon oesign. If,trie adventurous de-nan is made by tho military

are not necessarily well thought out, che ability 1to

resisc them would seem to be particularly strengthened. Suchmay be the product of poor communication betweeninstitutes ard the military authorities that seek toweapons design requirements on the designers. Onhand it may stem frcm the comparatively small researchof che Ministry of Defense. It is impossible to say which onethe cause. To the execneresltc chese

demands cn convincing technical grounds, the effect would be 'he sane.

These considerations may cast some doubt on the general proposi-cionstate-of-the-art" impulse from the military side may propel Soviet weapon system developments. Note too that several of the mitigating factors discussed in the relationship betweenand defense-industrial research scientists may apply as well. For the reasons mentioned above, it cannot be assumed that all

a glaring deficiency ir. design hand-

_pjoksgenerallydecisively undermine their ucilityeans of communicating "state-of-the-art" advances to designers. They often lag several years behind the level of As sucn, to che extent chat they are hce^-xi by designers, they would reinforce rather than weaken the tendency to design conservatism.


designers ore, or will be, prone to conservatism in weapon design. Hot can it be assumed that they have, or will have, sufficient power to resist demands to be more adventurous.

However, as in the previous discussion, explicit attention must be givenariety of relevant factorsetermination can be made in the casearticular weapon system. With respect-etermination of the desire and ability of the military to impose an adventurous design requirementivenumber of considerations need looking into besides those bearing on theand power of the given designer. Theseeterrinarion oi whether the Service involved has any particular incentive to push for an "adventurous" weapon program, what sort of researchit has to help make its requirement technically sound and convincing, what sort of previous working relationship it has had with the designer, etc. For example, the first consideration would require an estimate of the recent fortunes of the Serviceccapared to other Services). The second consideration is the possibility that the research backup for the Service may not beto the Service itself but may be drawn from other sources as well. For example, is there evidence of substantial liaison between the Service and relevant research scientists on the defense-industrial side, or with scientists attached to the Academy of Sciences? If the Services and the defense-industrial scientists each have discrete incentives to advocate state-of-the-art advances, their combined impactonservative designer could te considerable.

4. Special Role of Military Representative3

The thirdrelationship of designers and their Servicethe role of military representatives to the

fore.* Information

For purposes of discussion, military representatives are treated in the singular. It is recognized that standard Soviet practice is apparently to assign teams of such representatives to monitor weapon programs.

illuminates this role. Militaryare clearly crucial in establishing satisfactorybetween designers and the Services: they providedirect and1

year tour cf duty is common for military representatives fmmmmmmmV An affiliation of this length would seem toumber ofimplications in the relationshipesigner and his Service customers.

On the one hand, it suggests that the customer should be able to accrue considerable leverage. ive-year tour of duty shouldilitary representative to become thoroughly familiar with the operations of the design shop to which he is attached. This would give hisore accurate appraisal of the strengths and weaknesses of the designer whose work he is monitoring than that affordedhorter tour. He would be able to inform them of instances in which the designer was deliberately avoiding adventurous technological paths to minimize the risks of personal failure, etc. Prom che designer's standpoint, it would seem reasonable to suppose that the representative's long tour of duty would induce him COto the wishes of his Service customer. Facedengthy affiliationiven Service representative, the designer's lack of cooperation could well lead Co prolonged bickering, tenseetc. that he would prefer co avoid.

On Che other hand, che representacive's tour of duty could also carry negative implications for Service influence over designers. If the designer possesses considerable stature, he may respond to the prospect ot five years of troublesome relationsealousrepresentative by trying to get the representative replaced by someone more amenable to him. engthy affiliation between the designerilitary representative would seem to holdfor Che latter to acquire divided loyalties. tmnmmmmmmmV


instances in which military representatives subsequently acquired

important management jobs in the Ministry ^BBHI. IC is impossible

to say how widespread chis practice is in che defense-industrial sector.0 However, where the possibilities for sinecures in defense industry design bureausignificant incentive would also appear to exist for the military representatives to avoidindividuals, such as designer luminaries, who might be important in helping them secure these appointments.**

A systematic appraisal of the role of military representatives seems crucial in determining che ability of Che military Services to secure che cooperation of designers in Service-initiated weapon system developments. This seems particularly true in technologically adventurous developments. But military representatives also seem important in assessing the initiatives exercised Ly che designers

Note that in Che two cited examples the Individuals were identified as having been assigned to research inscicuces (not design bureaus) as military representatives. Moreover, one of them subsequently became chief engineer of the inscicuCe to which he had been earlier assigned. This practice would seem significant in terms of the potential forontinuing liaison becween che Services and research institutes. inecureefense-industrial research institute could suggest something quite differentinecureesign shop. Againsc the backdropresumed common interest in scate-of-the-art advances in technology by the Service and che research institute involved,inecure would tend copressures on the relevant designers for che development ofadventurous syscems.

In this period, the Ministry of Che Aviation Industry was called the Scace Commiccee for Aviation Technolcr,.


This example ia illuminating on several counts. First, iis accurate, it demonstrates convincingly thee designer-initiate, development programs do take place; the Services or other Ministry of Defense elements are noc the scle sources of program initiatives. Second, it suggest* quite strongly that, at the very lea*c, Che acquiescence of the military representacive assignediven designer cay te required to make the initiative possible andif the designer should lack the backingowerful pacron elsewhere in cheetwork. Finally, it indicates that che military representative mustppreciation of what the traffic will bear within the precincts of his Service, in terms of che Service receptivity to programs initiatedesigner.

The very possibility ofeminent designers-seeking to initiate weapon programs does not seem at all surprising,

^ This

of initiative would be supported by che very status of such designers. It would presumabLy earnumber of useful contacts in powerful political circles and the like. Moreover, since the Soviets keep design teams pretty much intact between development programs carried

out for the military customeresigner has the possibility of at least using this talent for his own ends, which would not be afforded him if the teams were disbanded. he same

oken, note that while che Ceams may be kept pretty muchinfringement on the manpower resourcesesigner maytime to time. mWKKkWkaaaaaKaaaMMamX nQted two instances inprograms in other

design bureaus |

But he also noted that this wa-sual practice.

These considerations ciuld either contradict or reinforce each other inoviet designer's ability coeapons development program. If the maintenance of design teams gives theesource capability to be used for his own initiatives, lt would also seem to deprive-him of an incentive to use this capability.

(This would help to undercut the notionoviet version of what is touted by some as the "follcw-on imperative*. weaponspractices, Similarly, if the loss of manpower resources would make it more difficultesigner to exercise initiative, the prospect ofoss would seem to give him an incentive to try torogram As with the relationships discussed earlier, che mix of incentive and capability can be expected to vary from case to case. The coabination oost likely to result in aeffort toew program would be one in which he-had considerable manpower resources at his disposal (betweenperformed for the customer) but, based on past experience, had some reason to fear encroachoenc on these resourcesew program requiremenc was not forthcoming soon.

The importance of Che military representative inesigner to exercise initiative is underscored in two fundamental ways. Ic appears highly unlikelyesigner could undertake toew program on his own without the military representative being aware of his efforts. Whacever che reasons for assigningrepresentatives to design bureausive-year period, its effect gives the miliCaryapability for Surveillance that shorter assignment practices would not readily permit.

The military representative is also importantesigner's initiative because of che entry he provides to the cuscomer. It might be that che designer can personally count on the influence of high-level patrons to support his initiative. But if he can't, the military representative would seem to be the crucial link Co those who, afcer all, have to "buy" che designer's program if hisis not to be stillborn. The earlier noted circumscances that bear on the working relationship between military representatives and designers appear to be of utmost significance here. esigner whose relationsilitary representative were basicallywould probably be severely hampered in exercising weapon development initiatives. But if these relations wereespeciallyey military representative had reason to expect a


sinecure in the relevant defensedesigner's initiative would likely te greatly facilitated.

However, another obstacle must te overcome if the designer's efforts are to te effectively supported by the military The program initiative must show promise of winningacceptance in the Service. It is inconceivable, in the Soviet setting, ttat the support rendeie.i| y the militarywould have been forthcoming simply because of the latter's personal appreciationtfamV ideas. The military representative would probably have been very wary oi giving emtemaTAihis supportfirst sounding out key Service officials as to their reception ofideas.

This leads to the final point about weapon developmentand the contrast between designer ccn serve ti sir. and the urgestate-of-the-art advances in weapons technology. It might be asserted that if, as generally assumed, the Services are the strongest advocates of technologically adventurous programs, they are likely to be most receptive to designer initiatives of this sort. But it can also be asserted that if the designers have ample roason forrisks in the programs they develop for requirements levied by the customer, they would be even more fearful of failure in thethey initiate themselves.

As with the other relationships discussed above, the weight of these considerations can be expected to vary from case to case. It depends on the Service, the designer, and the militaryinvolved. In generalizing on past and perhaps present Soviet practices, it seems prudent to shy away from the assumption that designer-initiated weapon programs have been technologically For the likely designers who have attempted to exorciie initiative! are those who have probably avoided failure by past conservatism, who have gained the ear of powerful patrons by their past successes, and who have the most to offer militaryin earning theirperhaps


Th*onoss erdcf the "constant shares" principle fa*n the echoes! to oiiture are probably overstated.

The notionconstant shares" principle -as operated in the past to damp dour, bureaucratic feuding among the Services and, by extension among the iefense-industriai ministries asdoes net seoc unreasonable. er, co keep the principle'sin perspective and to fix on its possible furje implications, it seems prudent toumber or considerations that would appear to contradict the impression of its pervasiveness andl-pacc.

1. Service Breach of "Constant Shares" Principle

On the Service side, the case is probably the strongest that the constant shares principle has been widely and effectively appLied. Nevertheless, granted that the Soviets havetrong incentive toelative balance cf apportionments among the Services, one significant departure fromractice must be given its due. As the opposition voiced by all the Services--except the PVO--attested, the formation of the Strategic Rocket Forces (SRF)9lear and violent breach of the constant shares

Although the formation of the SRF mayuiteof the principle on the Service side, otherless apparent but more comrnon--have occured on theside. Regardless of the longevity of various-dniscers, the fact remains that there have teenreorganisations of the defense industrial ministries over(Fig. recent past, the most

significant of these reorganisations have been as follows:

a. he State Committee for Electronics Technology (now the Ministry"of the Electronics Industry) was forced basically out of elements separated from the State_

Committee for Radio electronics (new tho Ministry of che radio Industry)

b. he Ministry of General Machine Building was formed basically out of elements separated from the Scace Committee for Defense Technology (row theof che Defensenda the Ministry of Hachine Building was also formed basically ouc of elements separated from the Ministry of the Defense Industry.

In terns of evaluating che impression of che role of che con-scant shares principle on tha defense industries side, conveyed by-the composite picture, that such reorganizarions have occurred and been noted is noteworthy. These reorganizations have not been as attention-getting as che formation of che Strategic Rocket Forces, but they have been chronicled.* Moreover (iopliciCly ache violacion of che constant shares principle that theserepresent, has also been taken into account. For example, onees appropriate attention Co the creation of entirelyacilities; =ut it asserts CMC:

As the new elements of che system were created out of che resources of preexiscing ministries, the resources required co perform che new or oxpanded functions were always transferred to them. For chis reason, as RSD facilicies have been transferred, for example, from MOP (che Ministry of the Defense Industry) co NCM (che Ministry of General Machine Building) or to MM (the Ministry of Machinehere has notignificant increase in theof all typeserforming institutionsthe sysCem

In discussing the fate of che Ministry of the Defense Industry in theanother study also notes thac:

It appears chac many, If not most, of che scientific research institutes and OKBs (design bureaus) active earlier ln weaponere disbanded, ard their personnel assigned to eitheracilities orrganizacions

As noted above, the Ministry the Defense Industry hasrime candidate of these reorganizations in recent years. These reorganizations and earlier ones affecting the ministry arein detail

In light of these statements, che role of che constant shares principle presented in the composite picture does not emphasize apparent violations of the principle enough. The significance of these infringements is suggesced by che very stress placed on che .ncentives of def ense- Indus trial ministries to seek self-sufficiency for their ministries chac che composite piccure features. It seems difficult co reconcile these incentives with an acceptance of the minor import of Che past violations of che constant shares principle in che defense-industries sector. Indeed, the various reorganiza--tions in the sector would seem to carry implications of potentially great political significance. For if industry ministers have soughc self-sufficiency, Che pise infringements on chelr ministries could hardly have been suffered lightly by them.

2. Pastations of the Defense-Industrial Ministries and the "Constant Shares" Principle

A detailed exposition of political and other ramifications of-che past reorganizations of che defense-industrial ministriesa cask chat is far beyond che capabilities of the present study. The most that can be ventured at present is to hint at che possible relevancy of an examination of these reorganizacions co certain key issues of Soviec defense RSD and defense policy For example, in analyzing the motives behind the formation of the SRFt has somecimes been maintained Chat

he SRF was the logical organizational resultuccession of technological advances which culminaceo,eparate and discinguishableof weapon systems andhe technological thesis holds chat attainment ofapability would result ineparate command organizationmftaal

A consideration of the reorganization affecting che missileof che Soviec defense industries would tend co dispute this tmesis.

The relevant reorganization in the defense-industrial sector occurredhen the Ministry of General Machine Building (whose basics ballistic oissiles) was formed primarily out of the resources of the Ministry of the Defense Industry. It appears reasonable tc suppose Chat an organizational reflection of pressing technological needsiven weapon system areain the institutional framework where these needs are first required to ben the defense industries. Vet, it is curious that the Ministry of General Machine Building was established only some six years after the formation of the SRF. Moreover, it was established only after seme of the most significant Soviet missile programs ofere well under way.

The timelag evidenced here uggests that there is particular reeson to question the notion that the formation of the SRF was an evident response to technological necessity. But it also hints at possible political considerations that were recognized. The formation of the SRF met with strident opposition from all cheServices, save one, the PVO. Consequently, it may have beenimprudent co risk adding to this opposition ty directly infringing on the domain of an important defense-industrial ministry at the same time.

It is possible that political considerations of this sort entered into the pictur ineorganization in the defense-industrial seccor simultaneous with the major reorganizacion in the Services. But it also seems possible chat political concerns subsequentlyart in determining che timing cf theof the Miniscrv of General Machine Building. The Ministry of the Defensetore the brunt of thisuccession of particularly eminent and presumably powerful ministers over the years. In its early incarnations during World War II and the postwar period, the Ministry of the Defense Industry was headed by. He was succeeded briefly by A. V. Domrachev in In. N. Rudnev (present head of the Ministry of Instrument Building and Automated


and Controlf Ustinov, assumed thethe Ministry ard retailed this position' hecrairmar. of then-it tee for the Coordination ofResearch. Rud. ev's successor was Scirnov,nd present chairman of tne Military-crial(VPX). Smirr-cv held th- jest of Mir.isoer of" when. Sverev, who continues to headH*aBBBBBJ

A detailed assessmentost of factors affect ins thepower of the Ministers cf ch* >fense Industryous stages in the Ministry's evolution would seem necessary coa firm case for the impact of political considerations on the Tin. is try's fate. ursory look at ch* successive incumbents of the top office in the Ministry suggests that an encroachtient on its territory eipht haveiberat*ly avoided, when particularly eminent and powerful administrators occupied chat office.* for exaaple, while che Scraterio Rocket Forces were being forted, Rud-ev was its occupant. arcicular political factor oioht have been at work during che tenure ofs helping tothe reorganisation affecting the Ministry untilo rev hat less eminent--and rorecook over. Zverev has suffered che indignities of an encroachment cn his domainq separace occasions5 when che Ministry of General "achihe Building was formed and8 when the Ministry of Machine suildira was established). This may well rescify to his comparative weakness with respect to his predecessors.

A possibly significant caveat is that during the last years stinov's tenure as head of theinistry of General Machine Building was also estabUshed. This entity was scrappedowever (see Fig. Such questions as to whether Ustinov was comparatively weak politically at the time, cr whether the constellation of political forces was such chat even powerful opposition cn his part could be overriden, neea addressing before evaluating the significance of the caveat.

Ofhis is not eo suggest that important tecnnologitai reasons did netole in the reorgeni oaticn decisionsthe Ministry cf the Defansa Industry. Indeed, these reasons nay have been ultimately decisive in contrast to the decisionsto the establishment of the SRF. But it is to suggest that when the dee lag between the establishcenc of the SRF and that of the Ministry of General Machine Building is considered, and when thestatus cf the occupants of the top office in the Ministry of the Defense Industry is duly recognized, the role of political considerations in affecting these reorganization decisions should not be ruled out.

These considerations serve as an example of the regimo'sto risk antagonizing both Ministry of Defense elements and Defense Industry cicoents at che same tine. But in addition, they may be taken to indicate generally the political leadership's appreciation of, and wariness toward, the power of both of these defense establisheentecielly their combined power. To tm sure, cheshares principle reflects such an appreciation by copas well. The difference, however, is that che constant shares principle conveys che impression that the leadership can pretty euch avoid direct challengests authority by keeping che principals concenc with generous and "equitable" resource allocations. There may be occasions though, as suggested by che considerations noted above, in which sore of ch* principals will require co be affronced. When Chis happens and when che constant shares principle is Itself violaeed, che topct basically by elnimizing che possible challenges co ICS authority through the avoidance of making too many powerful enemies at the same time. If the leadership is accencive to such possibilities, one significant implication is that further organizational chances affecting defense RSD might not be -as improbable as the composite picture would lead no* to expect.

3. Mutual Assistancehe Defense-Industrial Ministries

A related aspect of rhe top leadership's perspective on the defense sector, which rhe composite picture seems to submerge, is the issue ox compensation for those whose terrirory has beenupon. If, as rhe above considerations indicate, some-violations of the constant shares principle may be una void* tie at times, and if powerful metiers of che defense establishment can be expected to be deeply antagonized by these violations, then ir seems logical to suppose chat seme cenpensarcry mechanisms might ruse into play. To antagonize someone like Zverev might noc be as significant asa Sairrcv, for example. Yet, if the defense-industrial sector is to function smoothly, it seems improbableverev would continue as head of an important defense-industrial ministry without some compensation for the infringements in his domain.

A consideration of this sort seems likely to reinforce aapparently fostered by che technical complexity of modem weaponiminution cr che self-sufficiency cf particular defense industriesonsequent increased isnrribution by several ministries iniven weapon system.

It would be important for the leaders of miniscries toeasonable assurance o: access to che resources of other ministries, when needed. esult, losing seme cf one's own resources might te easier to tear. Presumably, che authority of the top leadership would be particularly effective in making such an assurance stick. If mutual assistance of this kind occurs, it would seem to Take reorganizations in the defense-industrial sector more palatable politically. At the same tiee, however, it might well reduce che cechnological imperatives for such reorganizations.

Again, it is important to emphasize that this is something quite different from what the-constant shares principlein terms of minimizing bureaucratic in-fightinj, the basic effects mighc seem to be the same. In the sense that the conscanC shares principle sanctions che abiliCyiven minisrry to preserve its

share cf the resource pie, in the inceresc or maintaining -ninisteriai self-sufficiency,figntin; nay te presuxede held in check because ci this self-sufficiency. By contrast, anof che operation cf mutual assistance among che derense-induscrial miniscries suggests that bureaucratic infighting may be kept in bounds, because there is an alternative co ministerial self-sufficiency. This mutual assistance may, cf course, be resisted, and the defense-inciustrUl ministries may attempt to preserve their self-sufficiency in the face or the demands imposed by technically-complex modern weapon systems. If this resistance occurs,in the defenseby technologicalight te more likely. Future efforts to preserve the constant shares principle may .veil precipitate, therefore, the verychat represent the grossest violations of this principle.

finally, an appreciationhe breaches in the operation of the constant shares pri-iciple might also shed some light on the relative status of the defense-industrial leaders and on their The mutual assistance roted above cay serve to mitigacs antagonistic relations between defense-industrial ministers. But it seems likely char infighting wculi still occur. The relations between ministers who have been direccly affected by past organizational conges would seem tootential for strain at cheeast. Zverev's relations with S. A. Adanasyev of the Ministry oi General Machine Building and with V. V. Bakhirev of the Ministry of Machine Building might, in part, reflect the resentment engendered by che escablishmenc of these miniscries ouc of che resources of Zverev's miniscry. imilar resentment would also seem to have teen possible in the relations between A. I. Shokin and V. D. Mlmykov (recently deceased). Shokin's ministry (Ministry of the Electronicsas formed1 basi :ally out of the resources Separated from Kalmykov's ministry (Ministry cf the Radio Industry). Obviously, the extent to which such resentment has in fact occurred and has been sustained over thewhether it has really mattered in che working relationships cf the individuals involved--depends on a

number ol ocher considerations. These are che compensation afforded che injured parties, the co Ha bora Tien required of these individuals in particular -eapen system developments, and che posslbilicies for ccapetiticn between thee in plumping for different weapon system developments.

The relacionsnip between Shokin and Xainvkov sujgasts the possible significance or other factors affecting the roles andof defense-industrial Tin is cry leadership elements chat the conscant shares principle Sty submerge. The constanc shares principle aces co suscain an icage of similarity of aetlcudes and power of the various derense-industrial ministers. This image is particularly fostered by the notion chae the ministersind oi managerial gerontocracy, which has accumulated management skill, technical expertise, and rcliticai acumen by long years of service in th* defense sector. On the whole, che picture of this gerontocracy mayairly accurate ore. However, ie is also importanc to Veep in mind possiblyistinctions ter-een its members.

A minister, such as P. V. Deoenryev, who has headed up the Ministry of the Aviation Industryay, for exaraple, have cultivated more relationships with important political, military, and managerial personalities than someone Like Bakhirev, who hasefense-industrial minister (Machine Building) for only six years.onsequence, Dementyev .might be more adepc than Bakhirev in utilizing informal relations to promote che interests of his And he eigne be more adept in circumventing obscacles chac the formal defense RSD decisionmaking process has created and which affect the efficienc performance of weapon development casks by his ministry. Obviously, this does noc mean Chat Bakhirev would have no supporcers of consequence. Indeed, the very establishment of the Ministry of Machine 3utidingerceived need on the part of the top leadership. Bakhirev might be expected to have certain powerful backers in cop leadership circles--in addition eo certain

"natural" allies in rhe Military. Nevertheless, Dementyev's looser tenuremister rayifference in che overall ability of che cwc men tcrhe system for cheir ana ends.

In considering such distinctions, note chat although some ministers have had car.ures longer than those of ethers, all of the defense-industrial mirdsoers have had long experience in rhe sector in one capacity or another. Also, note Chatelative polirical clout of individual ministers is likely co depend significantly on che weapon sysrtr. areas for which they have responsibility, "or example, S. A. Afanasyev (Minister ofchineas considerable political weight, despite his relatively short tenure as miniscer (ei;hcimply because ballistic eissJle deveicp-aencs fall within his administrative purview.

Most of the defense-industrial ministers are in theirs and older. Therefore, it jculd be manifestly risky to seggest that the differences in their tenure as ministers would have an important bearin; cn th* degree of conservatism Chat may shape their perspectives or. weapon system developments. Ir. examining Soviet policy making, generally, underre urn scarce* even the use of actuarial table* toe attitudes of the individuals involved should be approached with great caution. However, this is not tc say that all distinctions among the defense-industrial ainis-cers Chat partly relate ro the age factor are unlikely to affect Che atticudes these ministers may bring to weapon system The relationship between Shokin and Xalmykov, noted before, suggests the relevanceistinguishing factor chat may prove to be especially significant in the case of future Soviet weapon Shokin presumably accumulated appropriate administrative know-how and technical qualificationseputy minister under Xalmykov. Accordingly, he was doubtless regarded by rhe topas particularly well-quaufled to head up the Ministry of tht Electronics Industry, wren ie was formed1 out of resources separated from the Ministry of che Radio Industry. Recently,

P. S. Pieshakov, another deputy minister c" che Ministry of the Radios appointed minister upon the death or xaimykov.

*. Importancescars of the trial Ministries

Given che reiati/ely advanced, years of :ne ministers in che other defense-industrial ministries, the importance of the depupy minister's"post deserves hiihiighcinj. First, deputy ministers in individual ministries might .veU be reoirded as che likeliestto succeed the incumbent ministers. Obviously, che choice of outsiders is always possible. 3ut the deputy ministers would seem co possess an inside crack cn ministerial posts, becausee experience and know-how they are able co accumuiace in che weapon sysren areas in which their respeccive ministries are involved. The post of Minister represencs an obviously important lure. the deputy miniscers mi-hc be expected to he soce of tha cost zealous and ccopecicive individuals in che Soviecnviron-oent. The oeal and ccopet.civenesseputy minister may be direcced not only co advancing the incerests of his own ainiscry, but also advancing himself over other depucy miniscers within his ministry. Insofar as che idencificacion of specific deputy ministers in the Soviec defense-industrial sector has been possible, it does not seem unusual for each ministry co have several deputy ministers. Before Kaloykov's, the Ministry of Che Radio Industry, for example, had one first deputy minister and ten deputy ministers

Given cheentlves presented co chem, andetting in which any one of several likely candidates mighc te chosen for the job,ould noc be-surprising if deputy ministers sought early on to cultivate relationships with powerful political,and managerial cypes to give chem an edge over their rivals. For example, it seems possible Chat such cuicivacior helped ?leshakov secure the pose of Minister cf che Radio Industry upon Kalmykov's demise. If formal status is an indicator, Pleshakov was not

necessarily rha likeliest candidate for the job. Cm pose of first deputy minster is held by C.pslatky, while Pleshakov was cne of ten deputy minsters. Hc-jever, Revhlkav had secured come recenc prominence as "The" deierise-irlustrial representative at SALT. Whether Pleshakov's presence it SAL?solely due tc unique technicalhe -nay have possessed is difficult to say. The support of powerful backers xiy ilso have neiped hir. sec-re the SALT assignment. Alternatively, he nay have gainod these backers by his performance at SALT.

If deputy eirdscers in the defense-industrial ministries are affected fry che sorts of considerations describeeumber of implications follow. Although the) supporthe incumbent minsters would seem ccatural goal for them to seek, ic appears quite likely that chars would be some strain in their relationship. eputy minister is out to make his mark he might be expected to bo frustrated by the very prospect of having toongbefore room ar rhe tcpavailable. Sue he may also chafe at the minister's relucranceepart from "safe" weapon development programs. The deputy minister's presumed receptivity co "bolder" weapon system ideas would not be merely rhe product of his relative youthis the minister, butperhapswould be fed by Jn incentive to gain quick recognition. eputy minister's responsibilities are narrower than these cf the minister. Consequently, his willingness ro support "bold" programs might also cone frcei his greater familiarity with the MOKfc of the research scientists and deiigrers with whom he comes in cenract. eputy minister finally achieves the post of minister he eight also prove more amenable to "bold* programs than his peers. Again, this is not merely because of his relative youth, but because sach programs might be necessary in promoting the interests of his ministry againstministers whose longer tenure has given them more influence than he possessesewcomer.

A sicuacion or inherent strain between ainlscers and deputywould seem to offer rcssitliicios roc only for deputyso play -interest-group" policies" at ch* lower levels of che defense RSD decisionmaking hierarchy, but also for taking deputyespecially useful to the top ieadersnip. Becauseesire to earn tne sacking of powerful patrons, deputy ministers would seen to be particularly well-placed to ease che task of th*n keeping the powerful defense-induicrni ministers in line. They .may perform -his role quite informally as an idjuccj to ch* staff efforts of Ustinov, Serbin, and ch* Military-Industrial Commission (VPK). However, cn occasion, they -ay also beecond "formal'1 hat: ormal positior with the Vp.<.


Despite appropriate caveats on the Service side of the *cuaricn. che composl c* oiccrc -ay crscurc tr* ccocl**iry and intensity ci possible tne refers* sector cvcn che broas one-on-cne reiatior'hicsen the Servicesefense-Ir.-usTTjal Ministries.

The composicemphasizes che evident common interestsiven Serviceefense-industrial ninisery inparticular weapon systems. In so doing, it calls acc*ncion co the existence of powerful alliances in tho Soviec defense seccor chat mayonsiderable impacc in determining th* sis* and shape of the Soviet arsenal. The combined -weight of the Strategic Rocket Forces and the Ministry of General Machine Building may count for much in urging the USSR to develop and deploy strategic offensive missiles that might not otherwise ssea warrantedrational"of the international environment. However, th* significance of other possible interest groupings should net be overlooked, despite th* view that the locus cf possible interest-group relationships in promoting particular weapon_ syscems exists ma inly or only at the Service and defense-industrial ministry level.


tha comcosice picture has partially acknowledged chis linica-cion. Ic notes interests on the Service side chat cue across Servicehe cc-rmon incerescseen che aviacicn component of the Soviecand che Scviet (long-range) Air Force. It notes furcherumber o: defense-industrial ministries vayocnon interestiven Service inarcicular weapon system. vercr.el*ss, che ccmposic* picture does not bring out thef individual defense-industrial ministries being less thann their weapon system interests.

1. Special Characteristicsoviet Defense-Iriusrrial Ministry

The composite piocure fosters che impression of abetween parcicoiar Services and And oing, ic comes close to portraying ch*in che Soviet defense sectoray

chatdifference between Soviet.

of che Aviation Industry, for example, is

far core Chan che equivevltnt. aerospace firm, While it may te quite true chat Soviet defense-industrial ministries differ free their civilian counterparts byingle customer who possesses iwnense paver (the Ministry oft does not mean Chat the relationship of the defense-industrial ministries co che customeropsided one in power terms. Unlike. Depart-eenc of Defense, which can dealumber of independent firms for Che development and production of parcicular weapon syscems, the Soviec Hiniscry of Defense has no such opcion in dealing withse-industrial ministries.

If the Ministry of Defense wanes milicary ai-crafc developed ard produced, ic has co depend on che Ministry of the Aviation Industry eo develop and produce them. If It wants radars, che Ministry of che Radio Industry is che only source. If it wants submarines, theof Shipbuilding is Che only source, and so on. Moreover, if che common view of the imporcance of Service incerescs in Soviececisionmaking"is accepted, and if Service differences and rivalries are considered, the notionmonolithic" Miniscry


or Defenseiven defense-industrial ministry is sere-what irjporcpriate co ceyir. with. Similarly,. Depart:f Defense does roe aceonclicnic organization in confronting the firms ic coals with to develop and produce weapon systems.

The ixportsnce si hijhlishcin; what oay seem to ie an crvious difference between rhe Sevier. weapons procurementis tSat, for ore, it rails sttar.ticn to the fact chat ar.Soviet armed Service is not necessarily in the driver's seat when dealingiveninistry, oviet Serviceefense-industrial ministry is analogous. situation in which rhe Service dealsumber of "individual- firms which are part of aningle identity (rhe conglomerate speaks fcr the firms' common interest vrhile permitting tht firms CO 'compete-with each other).

The difference between the Soviet. weapon procure-err environments underscores rho necessary dependenceiven Soviet Serviceefense-industrial ministry to fulfill its particular weapon neeis. Buc ic also underscores rha flexibility the ministry may have in responding co these needs. efense-in-ustrial ministry does noc necessarily dependingle Service customer. Fortht Ministry of the Aviation Industry develops aircraft fcr ch* che Soviet National Air Defense Forceshe Air Force, and tht Navy. Similarly, tht Ministry of Clit Radio Industry may ceor producing,iven time, radars for the PVO and guidance and control subsystems for the Strategic RocVet Forces. Likewise, the Ministry of General Machine Building may be developing orreals for the Strategic Rocket Forces and SL2Ms for the Soviet Navy.

Clearly, different degrees of dependency must be taken into account. The fact thatan secure ballistic missiles only from the Ministry of General Machine Building would hardly lead che Ministry to view ic*ontracts"ith the Soviet Navy)

as giving it ouch leeway co slight Che needs of sucn an important customer. The dependency of the defenso-industrial ministries cn particular Service customers night beary, net enly from ministry to ministry, out also over timeiven ministry. The balance oi the workload of the Ministry of the Radio Industry between, for example, radars for the PVO and guidance and control systems fcr che SRF eight shift in one direction ir. one period and in the other direction in another period. Therefore, che peine is not chat che defense-industrial ministries can afford co slight certain Service customers because theyc solely depend on chem, but that the importanceustomerinistry may be neither absolute nor static. As such, che willingness of thet3 support tho interestservice client may also vary in incensicy, at the ministry level.

Variations of this sort wouldco be particularly imporro-ic co consider in situations where che cutccmersiven ministry might be directly cc-pecing toew weapons system. For example, if the Ministry cf had to cheesew alr-crefr for tne Soviec Stavyew aircraft for che PVO, which party might be more likely to win the support of leadership elements of che Ministry of the Aviation Industry to press its case? Fretun-ably, one factor among others to be sure, would be the minister's evaluation of che sice of Che expecced development and production "contract" in each case.

2. Differences between Ministry-Level and Subordinace-Eleoenc Weapon Incerescs

An appreciacicn of che limits of che one-on-one view of Service/ defense-industrial ministry relationso consider relative degrees of "interest" at the ministry level. It also calls attention Co important differences between ministry level interests and chose of subordinate elements within the ministries. It is here, ac che level of these subordinate elements, that the relacionship between customer and producer in the Soviet setting may

come closest tche key reHticr-shi? in. weapons prer.virorT.enr between Service suitsstrs sr.ifirms. Ir bears repeating chat rhe flexibility ar tht mini scry level, ino rheneeds of garriouU Services, prcr-tiy creates within fairly Mrscaers. , still, this flexibility .jculd ceem to be inherentlythat availablearrirulirureaus, for example,erense-irdusrrial ministry. The fortunesarticular design bureau ray be tied moreo rhe forcr-lc* eras is the ministryhola.

The difference between ministry level and desi;r. bureau la*.el dependency or. Service customersutter cfcr.-cerning possiblerelationships. siven Service mayarticular iesijr. shop cor si can CI -or* receptive co its weapon reeds than leadership elements of the iritis try zzthe design sliop is subordinate. esigner tight expect his Service customer to be mere acer.ble ro certain ofater, systea ideasculd his rvr. ministry, lh face, a* rheree the Ar.tcnov iasign bureau indicates (seeesigner may neet with outright op-ositicn by leadership slemer.r* cf his ministry, while finding supporters for his idea* .vithin rheof his Service customers | ^

This suggests an important difference. andScviet weapgnnvircr.mer.ts. The defense-industrial ministry is an organizational entity that has no equivalent inU.S. setting. On the one hand, because itarticular ares of weapon systems, che defense-industrial ministry mightreaterto promote such systems than that of an. weapon* contractor. Cn the ocher hand, also because of Irs "monopolistic" positionertain weapons system area, the defense-industrial ministry might have better bargaining leverage in dealing with rhe Service customer than. contracting firms. The defense-industrial ministry could represent an important obstacle that

individual Service cuscoiers In che United Scaces do ncc have co contend with. The defense-industrial ministry's resiscance :oweaponvinced by che cuscomer service or by weapon syscem initiativesn one of ics design bureaus eigne

somecices prove decisive in blocking weapon programs. However, che very pocenciai for opposirion of chis sore ac chelevel, may well serve Co stimulate interest-group activity at che lower levels.

Another implication of che differences between oiniscerial level incerescs and che interests of subordinate elementsefense-industrial ministry is chat there seeos to be considerable room for conflic- and competition among the subordinate elements. Thein this instance is not th* design competition involved ln the developmentarticular weapons system. Itivalry chat might be experienced between design bureaus,iven defense-industrial ministry, which have cies co different Services. Presumably,ivalry would be latent most of the time. But it coula come into the open when tfc* Ministry of Defense and che cop leadership might be faced with the need co choose becween two weapon syscems chat are each siacedifferent design bureauarticular defense-industrial ministry.

For example, che choice might beew bomber program for che Soviet Air Force (long-range) and an lntercepcor program for th* Soviec Macionai Air Defense Forces (PVO). Clearly,lgn shops as ch* Qyushin OKB and Che Mikoyan OXB, respectively, would have conflicting interests. To b* sure, decisionmaking situations in which the choice between programs of chis sort amountsimple either/or proposition would appear to be relatively infrequent. Buc even if che go-ahead were given on both programs, one might suffer ac the hands of che ocher u> ceres of prograa size. Trade-offof chis kind may not be all that infrequent. When they occur,

che conflicting iactrsltl of che affected surordinate elementsiven defense-industrial ministry tight well emerge. Rivalry between the Services involved would be paralleled by rivalry between che design bureaus in the defense-Industrial oinistry.

Such rivalry could stimulate intensive and extensive inceresc-group activity. To influence the program decisions, the Services and the design bureaus, with common interests in the developmentarticular weapon system, woulde and obvious incentive to ally in order to pressure theirsuperiors effectively. The Service would seek to influence ministry-level and General Staff elements in the Ministry ot Defense. The design bureau would seek Co influence leadership elements /ithin its defense-industrial Active supporc might be sought from research institute and production elements as well. Furthermore, special efforts might be made to win over key political types outside the ministry, sinceituation of this sort the backing cf th* defense-industrialleaderiven design shop could no more be Mken for granted than ministry-level (Ministry jf Defense) support for the Service.

Note chat the operation of the constant shares principlein hypothesis likely to heavily influence th*and virulence of the interest-group activity described above. Taken in its extreme form, tho Soviet leadership's rigid adherence to the constant shares principl* wouldong way towardany incentives for interest-group activity in the defense sector. If the Services, th* defense-industrial ministries, and their subordinate elements were truly assured of getting the go-ahead for weapon system programs chat affected them, they would have little incentive to seek to influence the decisions on these matters. However, this does noc mean chat if the Soviet leadership adheres co Che consMnC shares principle, powerful interests are absent fron the defense sector. It may very well be that ato antagonise such interests and stimulate them inco action

will provide an ultimate rationale, foronstant shares principle.

As indicates in hypotnesis So. however, it is doubtful that the constant shares principle has operated as pervasively and with as much impact as the composite picture suggests. This deficiency highlights certain aspects cf the interest-group issue. The compes-ite picture emprascces the practice of mutual assistance among defense-industrial ministries as beingecessary surrogate for the self-sufficiency of defense-industrial ministries. Such practices would seem to reinforce the importance of makingbetween ministry-level interests and the. interests ofelements within ministries. efense-industrial ministerertain reluctance tc share his resources with other ministries, surely the reluctance mustot greater on the part of thedesign shops that are negatively affected. Indeed, such transfers might te expected to take place more frequentlyiven ministry.

prospect of transfers of this kind would seem to provide designerstrong incentive to seek allies among traditional Service customers to get new "contracts." By the sameesigner mayreater incentive on occasion than the Minister of his defense-industrial ministry may have to engage in programs where mutual assistance is required. If instead of facing the loss of his own resources in this situation, che designer gains entry to the resources of others, his interest in development programs that call for inputs from other elements in his ministry or other ministries may well bedespite the coordination problems he may meet in running such programs.

3. Demur/ Ministers andepresentatives asv Inarticipants

The role of depuCy ainiseers should also be considered in dis-tinguisning ministry-level incerescs from chose of subordinate As noted earlier, che interests of deputy ministers may noc coincide with chose cf che Itiriseers of che def miniscries-tecause of che relaciv* youth, the career ambitions, and the narrower range of responsibilities of che former. Consequently, che incerescs of iepucv ministers in parcicular weapon systemmay ie tore incense than those cf che ministry leaders. The depucy ministers may nave closer contacts with parcicular design shops. They mayore amenable to bold weapon system programs and programs chat call for mutual assistance among ministries. They mi;hc also te expected to have actively cultivated important and powerful patrons outside their Hiniscries to gain the inside track for suceeding the incus-.bent miniscers. Aseputy ministert be one of the likeliest candidatesesigner to seek, as ar illy in pushing or defending his weapon interest, if he cannot co-nt on che automatic support of Ms ministerarticular we-oons system program.

In terms of the connection becveen Service interests and derense-industrial ministryifferenciacion becveen upper-level and lower-level fecuses acceneicn cn cheof military represencacives. As discussed in hypothesis No.ilitary representatives, by dint of cheir five-yearand hopesossible sinecureefense-industrial ministry, can be expected to Identify fairly closely with the parcicular interests of the design bureaus they are assigned co. Just as particular designers might be affected most ionediately and heavily by che faceiven weapons system program, so coo che careers of milicary representatives might be affected. The fact chac military representatives are the closest linkesign shop and che Service cuscomer makes cheir parricipaeion in interest

group activity extremely likely ar. the design bureau level. If, asesigner may not be guaranteed support at thelevel, the hacking cf the military representative may prove crucial.

Hone of this is to suggest that the supportilitary representative, the backingeputy minister, or even theof an important political leader will automatically translate weapon system interests at the lower levels of the Sovietecisionmaking hierarchyuccessful influence in weapon-system decisions. The surest identification of any interest-group existence does not, by any means, establish che influence of thac group. The nature oi the relevant decision must be considered; che forces, both pro and con, oust be identified and weighed; and che key question as to 'whether che decision would have been taken even without the interest group whose existence one has established must be answered. Obviously, the issue of interest-croup activity in che determinationeapon system program cannot be considered at all,cmmcnility of interests is first establishedariety of possible participants. Even common interests do not suffice, if communication among those with such interests is found lacking. This discussion has cried to indicate certain differences in che level 3nd extent of possible weapon system interests in the Soviet defense-industrial ministries. Such differences may help in determining where and when common interests may emerge in the case of particular weapon programs.



The extent ef direct competent supervision exercised bv the ccp political leadership ever the development cf new weapon systems la probably overs-ated in mcstncles cf Soviet 3SD.

The iepressicn of direct competent supervision, across che board, by the top leadership ofpon system developments is This conclusion is based cn what isout therealities of modern weapon systems, the questionable ubiquity of certain Soviet weapons development and decisionmaking practices (both formal andnd the apparent deficiencies inelement arrangements. Of course, this dees not mean that such supervision was not, or is not, forthcoming in particular priority weapon systems developments. Nor does it mean that ^his supervision was not widespread in the past. What it does mean is that thepractices of the post--at least with respect to theby top leadership--do noc automatically indicate present across-the-board management effectiveness. Nor do they indicate ;hat they will prevail in the future.

These caveats seen to hold the following implications forspecific Soviet weapon system developments:

1. In analyzing future weapon system developments, the caveats should te given special emphasis. Thisonsequence of two basic considerations. Cn* is that, on th* whole, the technical complexity of futureis likely to increase. reater number of elements of different defense-industrial ministries will probably need to contribute, and the use of off-the-shelf subsystems and componentsbe less feasible, thereby increasing coordination

prcblecs. Ache system! wculd probablyiceabl* strain on cht relevant expertise that individuals iik*v, Serein, candirectly frSQ cheir sucpcrt scarfs. eccrxl ccnsider-acicn involvesesigners and def anse-ir.duscrial ministry leaders. A* thev replace members cf checercr.rccracy chat have enabled Jstincv co ase ercnyisaeans of marageeanc, che infcrcal eie* withscinov facilitates hi* supervision cf ch* defansa sector, will likely diminish In importance. Obviously, this will not preclude new lnfcrrai Sue khachercciacicns will replace them cnas dsri.mcv has develcoed in thretof managerial responsibility in ch* defense sector seems doubtful. imilar phencaencn seems likely cc occur in che case of any of che for Ustinov's pesicien, once he leaves che scene.

Since the overall management demands imposed uscn cop leadership will probably increase, evidence of direct involvement by cop leaders In particular future weapon system develeprents rayore reliable indicator of the importanceiven system than ic has been. The operacing assumption in analyses cf past andSoviec weapon system developments is crac ch* top leadership's across-the-beard decailed involvement in supervising on-gcing weapon syscen deveiepments can be taken as norma 1. logically, Chen, such lr.vclvem.ent can not by icself really serveeasure of whac che Soviets regard as priority systems, Ey contrast, fucure weapon system manacamenc prcblecs in general may inducediscrimination and, hence, prompt directin fewer system developments. Presumably, che importance cf the system wouldey criterion in determining this involvement.

Greater coordinationd perhaps new forms of coordination--by people actually developing the systemsbe needed. Thea* are augured by the generalIn coordination problemsiminution of the significance of informal ties. The general increase would stem from decreased use of off-the-shelf hardware and the need to secure subsystem inputs from elements of several defense-Indus try ministries. Because ofeffcrc, coordination problems might tend to be handled more by in-house elements andore formal way. This could rang* frcra increased activity andcn the parr of defense-induitrial ministry leadership elemer.eseputy ministers)ore active role for Ministry of Defense elementsrepresentatives).


4. The Academy cf Sciences cay be used -Cre, and che staff support rescurcts cf Serbin, the Military-Industrial Ccmnissicn, and perhaps "Jstincv personally may be The strains cn the expertise or the topsupport staffs that future systems may wellseen, tc infer these possibilities. Th* needed increase in the staff support resources of Serbin, the Military-Industrial Commission, and perhaps Ustinov perscnally, woulti be sizeable. Due to certainproceed" barriers in the latter case inthe Academy of Sciences ray also be tapped mere extensively--or acre evidently--thar. in the past, as another cc=plem*ntary source of expertise.


The Soviet defsnsedecisionmakinge shaped by important political ccncrol factors that the ccrocsice picture cces not take into account.

When the political control elements cf the Soviet leadership's relationship to the defense sector are singled cut fcr explicitomewhat different view of the workings cf Soviet?SD than that presented by the composite picture emerges. The composite picture festers two related impressiens. Cne is that the Soviet cop political leadership keeps tabs on the activities of the defense industrial sector with great effectiveness. The second is that the leadership's motivations in this regard are confined mostly to making sure that the sector serves the Soviet Union's military, economic, and scientific purposes. iscrete political control purpose is also acknowledged to exist, and if it is viewed in the broader context of the long-standing political practices evinced in Soviet civilian policymaking, important constraints and ccmprcmises, that the ccmposice picture submerges, are revealed.

On one level, these constraints and ccmprcmises directly affect defense-industrial ministers and designers. The apparent virulence of the former's parochial interests inhibits che utility of the forum offered by the collective organ of the Military-Industrial Commission for these ministers toowerful cenmon defense interest. At

che Sana time, by caison of chis inhibition, lncivicual ministers wculd appear cc gain ism leeway in dealing with purely defense oaccers that affec: them. Similarly, design ccmpecicicn wculd appear co offer che means fcr keeping individual eminent designers responsive co higher political authority. Also, design ccnpecicion would help secureVs military, eccrccic, and sciancific gcals. Sue, presumably, becauseoncern tc see Chat these gcals are notTaacership wculd be reluctant tc severely penalise che losers cf chis ccnpecicion. Such reluctance would seen eo mitigate the ccncrcl exerciud over eminent designers.

Cn anotherhese constraints and compromises directly Affect the capabilities cf che political watchdogs cf the defense seccor. In so doing, they indirectly work co enhance the actcnemy of the above caned participants. The basic tradeoff here involves two needs. Cn the cne hand, thereeed eo invest theseparticularly Uscincv--wlth sufficienc auchcricy eo deal with Marshals, ministers, ano designer luminaries. Cn th* ether hand, chereeed to keep UscinCv himseli in his proper place. The organioaticnal arrangements bearing cn che cop level political supervision cf che defense-incuscrial aecccr would appear co be somewhat less effective than ocher pcrslble arrangements. They would b* less effective ln securing tight policical ccr.crol over middle-level and low-levelin this sectcr and in securing che military, economic, and scientific goals of the Soviet state. But they also ace Co Jnhibie the poeencial for political challenge by the watchdogs themselves. Because of the basic nature of these concerns, it seems likely that the political leadership would seek to strike the same balance for any of Ustinov's possible successors. However, depending cn who is chosen, che relaclve weights-of these concerns maysomewhat.

The first hypothesis suggests that future weapon systemneeds might well give the Soviets an incentive to augment or alcer present supervisory practices in. The present

hypothesis suggestsiscrete concern by cop leadership for political control would definitely limit the kinds cf management alternatives considered.

If the significance of political control elements in cheof Soviet defense SSD is explicitly acknowledged, the following implications for analyzing particular Soviet weapon systemseen appropriate:

present hypothesis gives added weight tohighlighted in the first hypcthesis. Afor political control is apt to make itdrastic changes in top leadership'sin defense SSD will occur. Therefore,that may help to compensate for managementin the case of technically ccmplex modernwithout roiling political waters, are likelymore significant and widespread. This wouldapply particularly to the use of individual talents

in the Academy cf Sciencesource of outside By contrast, Ustinov's ability toarge personal staff would seem less likely than even the first hypothesis allows.

The practice of solving coordination problems in house, suggested by the first hypothesis, will Drobably be underscored. This is because political control concerns make it inadvisable to effect chances that would enhance the supervisory capabilities of the Party's defense watchdogs. arger role for deputy minister types, military representatives and, in general, acre intimate liaison between the Services and Cheir cliencs, areby this implication.

Despite the difficulties involved in altering the basic organizational arrangements pertaining to the Party's watchdogs, some changes may occur. ugurs for increased leeway and autonomy by middle-level and lew-levelarticipants. Consequently, efforts

by the top political leadership to seek effectivecontrol ever these participants should not be ruled cut. It would_ seem likely that the leadership would strive to maintain the current division of labor among the Party's wacchdegs. For example, if Ustinov's personal staff is increased, ic mighc well beby increases in Che scaff elemencs subordinate to Smirnov and Serbin. Inicuacion, che systems-analyse type evaluation capacity, which the cophas appeared to lack up to new, may find anniche. Of course, because of its political-

importance, this capacity cculd be sec up independent of any of che present wacchdcgs. By ch*eken, though, ic could be ir-ccrpc rated in th* staff elements of one cf tha watchdogs, providing th* other two vere duly con pensated by staff increases.

Competition,ossible political centre! aechanisn, may be acre lipcrtinc than it has been. The difficulty of caking che sorts of organizational changes that,would be potentially volatile would rake the competition mechanise more important. However, its actual us* vculd not te dees-rained byeneralilcne. In political terms, much would depend on whether the defense SSB environmentc apparently has featured up tonumber of eminent designer luminaries with ccnsiderable political clcut. Similarly, ic may be that ccmpetiticn might provecounterproductive in certain weapon areas. Moreover, it may occur at che subsystem level as well as at the system level, given che likely cechnicel cceplex-ity of future weapon systems and the possible diminished use of cff-ehe-shelf hardware for these syscems. Ic seems improbable that the developers of these subsystems would on Che whole attain the eminent scacus enjoyed by famous syscems designers of the past. Consequently, che chances are chac the ceo petition in such cases vculd fcr the most part not be promoted by political concerns. Competition might be politically palatable. 3uc it should not be assumed chac the competition observed in particular future weapon system developments ismotivated,esigner luminary participates in the development, and che technical utility of such competition is clearly found wanting.

A cercain independence in che activities of topwatchdogs in monitoring particular weapon syscem developments should he expected. This is inferred frcm the political considerations chat affect che working arrangements of the top leadership's watchdogs in the defense sector. The first hypothesis suggests that watchdog Involvement on the whole may indicate che prior-ley statusiven weapon system oevelopmenc more truly Chan in the past. But where chis participation may be observed, tight coordination among theseshould not be automatically assumed Co exist,the status of che system. This does not necessarily imply bitter infighting among ch* watchdogs or their. However, it does mean that th* people whose activities are monitored by the watchdogs may seek to enhance their own autonomy by playing the variousagainst each ether. In summary, the potential fvr political leeway by defense-industrial ministers.

c ti should not be discounted, even in weapon davelcpeenc situations in which leadership stts fit to encouragt close supervision.


As presented in tht composite picture, rhe emphases cn both Soviet design* co stace-of-the-arcs re basically ccnc.-ac

The iBpre*sicn in ch* composite pictureomfortablein Sovietetween the features cf designand responsiveness to stac*-o'-cbe-art advance* in technology is not easily sustained. In an important sense, these features appear to conCradict each other. Of course, they do noc contradict each other when cn* considers char research scientists in tht defe-se-industriel ministries and the Services should be particularlyto what ray beechnologically adventurous design philosophy and that designers leanasically conservative design philosophy. Nor do they contradict each ether In thechac tha latter may give ground to the former in somepermitting the urge toward state-of-the-art advances to be reflected in actual Soviet weapon system designs. But they doin maintaining that Soviet designers are both powerful and conservative andesponse to-the urge to advance the state-of-the-art technologically may constitute an important driving forcethe weapon systems the Soviets actually develop and deploy. If Soviet weapon designers are powerful and conservative, they tight well be expected to try to frustrate the efforts in the direction ofadventurism undertaken by defense-industrial researchand tha Services. And they might well shy away frees aadventurous course in their own program Initiatives.

To be sure, in particular instances, Soviet designer* may depart from the practices generally ascribed to them in succumbing either to the individual or combined influence of the scientists and che Services, or in exerclsing'cheir own initiatives. The Services and

the defdustrial research scientists each have apparentlyco plump fcr state-cf-the-arc advances in technology. Each cculd presumably exercise independenr influence cn designers. And designers ehemselvesotential for development initia-civesechnologically advencurcus direction. Because of chese possibilities, che emergenceechnolcgically advencurous wsapen sysrem dees noc auccmaticaliy sake ch* idencificacicn of che system's iniclaccr easy. In each case, ic seems necessary co cxpllclclya variety of issues that affect che relationships between che research sciencists and designers and between che designer and the Service. These issues include the Incentives, ccmmunicaeicns, and power of the participants concerned and, especially in the case of the designer-customer relationship, che parcicular role of therepresentative.

The basic chrusc of this discussion has been to idenclfy and spell cut various condicicning factors. And these should bewhen utilizing design conservatism and the urge co state-of -che-art advances in technology as indicators of initiative origin in weapon syseve programs. Accordingly, che significance of the present hypothesis fcr analyzing individual weapon programs can beby briefly recapitulating chese factorseries ofthat should be answered when crying co determine ch* iniCiatcrarcicular weapons program:

Is ch* sysccm in quescion conservative, or technologically adventurous (even exotic)?

Isrcroinenc designer involved in ch* prograa? What is ch* designer's past record in terms of ch* kinds of systems he has develcped? (Conservative va.Successes vs. failures?) (Service customer?) exhibited?).

3*. Are prominent research scientists involved ln the prograa? What are their past interests? What is their status rel-aciv* to che designer involved? What sort ofhave chey had with che designer ln che past? With subsystem developers? What is their relationship with the designer's customer?

Which Servic is involved? What is its standingto o'..iec services? Does it have anyreason co he interested in che develcpcenc of an exotic or conservaclve system?

Whac is the role of che military representative? Does he-have an apparent long tenure with che relevant design bureau? If so, what sere cf working relacicnshi? has he had with che designer in che past? What sore of relations does che military represencacive have vith che research-instituces irvclved? Is there evidence or past jobfor oilicary representacives in the relevr.cministry? The design bureau? Th*institutes?


The pervasiveness and impact cf the "conscanc shares'* principle featured in che cccpcsice picrore are probably overstated.

The impression of the pervasiveness and impact cf the constant shares principle cenveyed by the composite picture appears to be somewhat wide of che mark. This opinion is based on the evidence of pasc reorganizations in che military establishment and che defense-industrial ministries, che impecus toward "mutual assistance" aceng the defense-induscrial ministries, and che differences among defense-induscrial ministers. This is not to say chat, in broad cutlirf, the conscant shares principle isoal that many adherentsCO on various levels of the Sovistecisionmaking hierarchy. Noi is co claim ehae so far as che principle is applied ic is not useful in helping co keep bureaucratic rivalries in check. However, pasc reorganizations in the defense sector and theimperatives of technically complex modern weapon systems suggest tha* it has been blatancly breached on occasion. And chat since the jelf-sufficiency of che defense industrial miniscries is considered an essential attribute cf the conscant shares principle--barring the way to mutual assistance among cheseSoviets might be compelled to blatantly breach the principle in the future, if they Co apply ic rigorously. Moreover, the difference- among che defense-industrial ministers suggest that if the relative influence

of Individual ministries matters in the allotment of the resource pie, the portions that the ministries receive nay be more variable andthan the constant shares principlein the future.

These caveats suggest the following implications for analyzing particular weapon system development programs:

The participation cf elements from severalministries would not seem unusual, especially in the case of future Scviec weapon system dsvelopsentechnically ccmpiex weapon systems are likely to make it increasingly difficult for any one defense-industrialto "hoard" sufficient resources to minimize che need for contributions*from ether ministries. The topencouragement of mutual assistance would beby an appreciation that if such assistance did net take place, further reorganizationseministries might be necessary. The leadership might well expect these reorganizations to be mere disruptive and engender deeper antagonisms, than might result if defense-industrial ministers were encouraged to "share" their resources.

A specfll burden would be placed cn the party watchdogs in the defense sector. In order for ccmpiex weapento be developed efficiently, the politicalshould appreciate tho need fcr mutual assistance among defense-induscrial miniscries. The need fcr active involvement by the watchdogs in weapen system development programs would be increaseo not only by che technical complexity of che programs as suggested in hypochesis So. 1. Above and beyend the proliferation of relationships among defense-industrial ministry elements chat che very complexity of che system might stimulate, management problems nay be increased by che sheer reluctance of various miniscries to cocperace. For example,arcicular ministry may see che concribucion of one of his research inscitutes or design bureaus to

a program performed under the auspices of anotheras detracting from the successful completion of his "own" programs.Presumably, he would try to resisc making chis concribucion. inicuacion, che acclve incervencion of the tcp leadership's watchdogs would seem necessary to ensure that the necessarywas initiated and maintained.

suggested in hypothesis Ho.uture weaponin particular may bear witness co cheinability to actively monitor the ongoing

development cf all weaocn systems. rying cc cotermine whether cheparticipation cf so? laacership's wacchsccs in *weapons programc was ortcioitatec by the pricricy status cf ch* program orager.anc problems of tho sere described above, it wculd seem necessary cc gec seme iLw cn the past relationships cf tho ministries ir.vcived, cheir currenc weapon orcgrans, etc. It would be important to address such cjssti-ns as to whether the reasons for antagonism between thea arOse out of case relationships, and whether they were each working cnhat competed fcr the same. If nc particular basis fcr acsucing ch*of fcccdracging cr outright resistare* in ch*cf the ministries involved were fcunc, che active participation cf the party's wacchccgt in theweapon program might then be assumed with -ereas indicating che very priority of che system.

The role of deputy ministersiven weapons systemrogram be apparent and should be given special attencicn. Ir. central, such activism would seem to be prompted by the ambitions cf these ministers, as discussed above. Their role as cccroinaccrs and monitors aighe be particularly expected, if tho active involve-sent of che cop leadership's watchdogs is minimal. In such situatiens, the incentives cf deputy ministersn th* suopcrc of powerful backers in advancing their careers might be counted on by the cop leadership co make che deputy ministers acceptable surrccates for the leadership's watchdogs. Also, the activeof deputy ministers in weapon system programs that are both technically complex (andumber cf ministries) and technologically adventurous wculd be expected. Secaus* of their career ambitions and relative youth compared with ch* ministers involved, they might be more given to support mutual assistance with other ministries than would theirand they would be less wary of bold programs as such. Indeed, the potential of the position cf deputy ministers wculd sees considerable enough fcr research scientists and service elements to seek the deputyout as allies, in pushing for the initiation cfadventurous weapon system programs.

. S

appropriate civeacs cn che Service side of the ec.uacicn, the cc-posicev obscure che cccolexicy and intensity cf possible interesc-oroup relationships in che defense sector bycn che breadn-cne relationships between che Services and Defense-Industrial ministries.

The composite piccure focuses on che existence cf importanc cca-mon incerescs becwesn che defense-industrial ministries and the Soviec armed services in parcicular weapen syscem areas. As pocenciallyas this cne-cn-one relationship may be in determining certain weapon system decisions, ic may obscure significant and ccmpiexgroup relationships chac occur below che defense-industrial ministry level. An appreciation of the possible common interests at the lower levels is important in two differenc ways. Consider the cases in which supporteapons syscem program may be forthcoming ac the defense-industrial ministry level, therebyne-on-one service-defense-industrial ministry relationship tc operate. Here, an appreciation of the lower level interests affected by the program can enable one to roughly estimate the intensity cf the icin-iscry level support; it can indicate the kinds of internal pressures that the ministry's leaders are subjected co. In cases where supporteapons system program may be lacking at the ministry level, an appreciation of the lower level incerescs can indicace whecher che resiscance of che ministry can be possibly overridden.

Perhaps che mosc vivid recent example of the need to go beyond the one-on-one relationships ftaturtd by the composite piccure is

| success in carrying out the fmmWM program. By the implicit logic of che nocion of one-on-one service/defense-induscrial miniscry relacionships, che program could appearuice nacural produce of the common interests of the Soviet Air Force (long-range) and che Miniscry of che Aviation Industry.

irs; itep In appreciating che relevant differences in weapon system interests that ray influence the decisions affecting particular weeper, syscen programs, ic is important co not* the special qualities of the Soviet clefndustrial ninistries. They are not the equivalent of weapon system contracting firms in the United States. They are analogous. industrial conglomerates in that they "control" the development end production cf particular areas of weapon systems. Also chey may have several Services as theircusccmers. Because cf their situation, che defense-industrial ministries say net have as incense an interestiven Service in the developmentarticular weapons system.

Those in the defense-industrial sector who seem likely to have the most intense interests in particular weapon systems are located below the ministry level. It is hereervice interesteapon system, and certainly the interestservice'selements, may be most closely approximated cn the defense-induscry side. Designers, military representatives, and deputyin particular would seem to have both the greatest incentives and the greatest opportunity co seek ccmocnality in either promoting or opposing particular weapon system programs. Their careers are lil.ely to be sore directly tied to the fate cf particular programs than those of defense-induserlel alnlscry leaders; and chey would seem co have considerable opportunity for direct and sustainedwlch each other.

Elements of competition may come inco play at the ministry level in situations where the sizeeapons system program to beby one ministry might be effectedoncurrent program carried out by another ministry. But an additional element ofcan occur below the ministry level to stimulate interest group activity. Since che defense-industrial ministries nay have

acre Chan one Serviceotential exists fcr opposing Service-designer alliances co ariseingle defense-indus-crlal miniscry. Obviously, the successiven interesc grouping at che lower level of che defense RSD decision-caking hierarchy, in seeking co affeccystem decisions, will depend cn acre than just che intensity cf its ccrmcn interestarcicular program. Those with less incense interests out mere pcwer to influenceisay have co be wen over as well. However, it may beelatively high level of interesteapon program is necessary, to actually prompt the formation of interest groups in che first place. If chis is so, auch of che real inceresc-crcup accivicy that may be found in che Soviec defense seccor--es opposed to individual efforcs co affecc weapon decisions--weald likely occur below che miniscerial level of che defense-industrial miniscries.

This discussion suggescsetermination of che exiscence of inceresc-group activityarticular weapons syscea development shculd include che following consideraeiens:

1. The importance of the program in question to che Service and the defense-industrial oinistry (or miniscries)should be ascertained. Cn che Service sice, chis requires scce estimate of che Service's currenc standing relacive to ocher Services: whecher che system augurs to fill an important gap in che Service's weaponswhecher the system, seems likely to aeeeinposed by. weapen programs, etc. On che defense-industrial ministry side, ic would be particularlyco get scae appreciacicn of che position the system occupies in che currenc and prospective mix of weapen prcgrans carried our for che miniscry's various Service clients. Foris chearge program thac has been given priority scacus by the Miniscry of Defense and che top political leadership? Does it cunpete for resources needed for important weapen system programs underfor ocher Service cusccraers? Does theof the system require inputs from other defense-industrial miniscries? Is the system technologically adventurous or conservative? Answers co chese kinds of quescions are likely to heavily Influence che eccitude of th* leadership in ths defense-industrial miniscry (or miniscries) involved in the program.

Below the .ministerial level, che following kinds z2 questions shcuic Vine is che chief designer fcr che Wat is his status? What sere cfinfluence has he exhibited Lipast? Hew io> porcanc is che prccrar: cc his future standing? What nas. been his past relationship with ch* minlsttrfs? cf his ministry? Dees he have any discernible ties ce ether political and military types whese influence could be important fcr che program ir. question? Is che system technologically adventurous or conservative? What i3 the'3 likely attitude co technologicallyand conservative systems? Cces the'systen require inputs from etheruscrial ministries?hat is the designer'3 attitude tc such programs? Has thehad goodong-standing relations with therepresentative assigned tc nenitcr the program: Is there any evidence c: che military represencativs's status within his Service? Soes he have any'discernible ties to important superior officers whose influence eight bear on ch* program in question? Is the releeputy minister in evidence? If so, are there any indications of his scacus within che miniscry, his relacicns with the minister, with ether keythtypes? These kinds of questions neec answering in order to determine the intensity cf interest of lew-level participants in che developmentarticular weapons system program. They also need answering in crier co determine whether che program could engendercn the part cf other elements in theministry. In combination with the assessment cf the situation ac the ministry level, these dacermir.seions might then serveasis fcr judging the likelihood of actual inceresc-grcup activity cn behalf cf the program.


Hoist,. and Sovietand Armsccasionalf Policy Study, University of

RAND Corporation, Worldwide Soviet MilitaryPolicy, Thomas W.antaApril

Colin Cray, -The Arms Race is Aboutoreian Policy., Winter

Graham Allison, "Conceptual Models and the Cuban Missilen Bureaucratic Politics: Readings in American

Foreign Policy, a. Haioenn and Ar *

Michael Armacost, The Policies of Weapons Innovation: The Thor-Jupiter Controversy,, Hew

James R. xurth, "Aerospace Production Lines and American Defenseesting the Theory of che Military-industrial Ccmpiex, Steven Rosen,. iei, Lexington,

Hereon J. Peck and Frederic M. Scherer, The Weapons Acquisition Process: An Economic Analy*Ts,

Robert w. Campbell, "Management Spillover from Soviec Space and Militaryoviet Studies, v. XXIII, No ased on larger study, Space Spillovers In the Soviet Economy, Final Technical Report of Research under NASA Grantloomington, Indiana,

David Granick, Soviec KeCal-Fabricacinc and Economic Development,,V.


ch^artz and William R. Keech, "Group InfluencePolicy Process in ehe SoviecmericanReview,eptember

12. Philip D. Steware, "Soviet Ineeresc Groups and che Poiicvorlp_Poa^, XXII, No., Oct-

H. Cordon Skilling and Franklynnterest

Groups in Soviet Politics. Princeton University Press".

Princeton, Mew


E. Cdcm, "The Partyroblems, Sept ember-October


Yorkn: The Storv of SALT,ew

16. The RAXD Corporation, Weapons Acquisition In the Soviet -Union. United States,rthur J.anta Monica, California,

t?. The RANDn SovietR Santa Monica, California,


The RAND Corporation, Controversy inhip CaseOOl-PR,




22. The RAND Corooraticn, The Interaction Process and Its In-fluence on Major Soviet Arms Decisions, T. W. Wolfe anc

onica, California,SECRET).

Original document.

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