Air University Review, November-December 1983

Levels of Strategy and American
Strategic Nuclear Policy

The debate over American nuclear strategy for deterrence is clearly in disarray. Deep divisions separate scholars, defense analysts, and policymakers about the nature of the nuclear threats that confront us, appropriate strategies to counteract those threats, and proper force configurations to support the deterrent purpose. Disagreement covers the intellectual spectrum of possible advocacy. At one extreme, harsh assessments of the Soviet threat have led analysts like Colin S. Gray to advocate a much more robust force structure and a plausible "theory of victory" in a nuclear conflict as the necessary ingredients for continued deterrence of Soviet nuclear aggression. ¹ Such suggestions appall other analysts and bring about ringing appeals for a return to more conventional deterrence conceptions grounded in assured destruction. ² As one British observer dourly concludes, " Form the surreal world of the analysts have emanated hypotheses about how to fight and survive a nuclear war that corrupts the Western concept of deterrence." As a result, "the outlook at the start of the 1980s is quite surprisingly grim. The risk of a holocaust is growing with every year that passes, and whether we shall avoid it is at least questionable."³

The strategic debate swirls between these extremes, manifesting itself most distinctly around two basic interrelated issues sufficiently well treated in the literature to need no more than passing mention here. The first issue is the evolving nature of the strategic threat posed by the Soviet Union. Focused on the continuing aggressive Soviet force modernization and expansion program that dates back to the latter 1960s and the discovery that the Soviets view deterrence differently than do Americans (for example, "war-winning" strategy, Soviet civil defense), disagreement is widespread about what all this means. Have the Soviets acheived some sort of nuclear superiority? and if they have, of what utility (if any) is it to them? Even more fundamentally, what are the true intentions of the Soviets for their thermonuclear arsenal? and how can we meaningfully relate Soviets’ capabilities to their intentions? Within these prickly and intractable questions, the stuff for worst-case analysis abounds.

The second issue is more evolutionary and technological: weapons arsenal capabilities have changed and expanded, at least theoretically, a great deal during the past decade, and uses of nuclear weapons formerly unthinkable (because unattainable) have become less so in some minds.⁴ The issue, resulting from the independently developed but mutually reinforcing emergence of multiple warhead delivery vehicles and increased warhead accuracy, has been the theoretical achievement of hard-target counterforce capability against a broadening but not comprehensive set of retaliatory targets⁵ and the consequent "window of vulnerability" and what (if anything) to do about it. These matters constitute the heart of the controversy over the MX missile system and ways to base it.⁶ In a more futuristic vein, the potential development of effective ballistic missile defenses (BMD) through projected advances in antiballistic missile (ABM) and "exotic" laser and particle beam (collectively directed energy transfer or DET) technologies⁷ offers the prospect for a parallel expansion in capabilities.

These technological innovations, particularly as they begin to enter the operational inventory, have raised considerable clamor in the strategic and especially the academic communities. The basis of misgivings is that these new systems possess characteristics suitable for missions either extraneous to or at odds with deterrence conceptualizations derived from assured destruction. Dissension reached a pinnacle in the late summer of 1980 when the Secretary of Defense Harold Brown announced so-called Presidential Directive ( P.D.)59 and has continued amid speculation the Reagan administration will supersede that guidance with planning for a protracted, winnable nuclear war.

Reaction has been strident and has helped to crystallize a debate about strategic policy as fundamental and profound as the debate about massive retaliation in the late 1950s that eventuated in assured destruction. Critics passionately condemned P.D. 59 as a sharp break from the traditional assured destruction deterrence base with dangerous implications that could make nuclear war more thinkable and hence more likely. Supporters, including many within the professional military itself, viewed the directive as at most an incremental change in operational policy, a position taken by Secretary Brown when he announced the document at the Naval War College commencement: "P.D. 59 is not a new strategic doctrine; it is not a radical departure from U.S. strategic policy over the past decade or so. It is, in fact, a refinement, a codification, of previous statements of our strategic policy."⁸

The irony, which gets at the heart of the entire debate over strategic policy, is that both sides are correct, from their perspectives. The difficulty arises because those perspectives are different, focusing on different aspects of nuclear strategy, so that a tendency and continuing danger exist that the parties will simply talk past one another rather than engage in a mutual dialogue from common reference points. If a constructive debate is to emerge, this problem needs to be addressed for what it is, a levels-of-strategy problem.

It is the central contention here that the nuclear strategy process, rather than operating in a seamless, deductively valid manner, operates at a minimum of three separate levels, each of which has strong implications for the others but which, in fact, operates in large measure independently of one another.

Levels of Strategy

"In simplest terms, strategy is a plan of action that organizes efforts to achieve objectives."⁹ This simple definition suggests that strategy has two basic components. The first is the plan of action: a response to some form of challenge posed by an adversary to our politically determined goals or objectives that changes as our perceptions of the challenge alters. In strategic nuclear terms, the political objective is to deter a thermonuclear aggression by the Soviet Union (or any future member of the "nuclear club"), and nuclear strategy at any time is the action plan that gives effect to that objective. Second, however, strategy is also "the process which connects the objective ends with the means to achieve that objective."¹⁰ When the dynamic nature of nuclear strategy is considered, one is confronted by the various levels at which strategy operates and the discontinuities that can exist between those levels.

Broadly speaking, nuclear strategy is made and carried out at three levels: declaratory strategy, development and deployment strategy, and employment strategy. (Desmond Ball refers to the two latter levels as "action" policy or strategy.)¹¹ Each level represents a complex of activities and missions that flow from the deterrent purpose, and each level tends to have its own reasonably distinct set of actors, dynamics, and operational constraints.

Declaratory strategy refers to the broad "set of public pronouncements made by the President, the Secretary of Defense or sometimes other senior administration officials regarding the requirements of deterrence, targeting policy, and strategic doctrine."¹² Based on subject perceptions and political judgments about the nature and intentions of our adversaries, what will dissuade them, and how our actions will be perceived by our adversaries, declaratory strategy is the most general and public statement about what deters. The nature of deterrence involves complex psychological and theoretical elements in addition to murky assessments about adversary intentions and philosophical statements about the propriety of different strategic options, so that strategy at this level tends to be abstract and theoretical. Often seemingly unrelated to day-to-day events in the political world, declaratory strategy attracts the attention of the theoretical intellectual community. It is on this level that almost all of the academic debate is confined.

Declaratory strategy is also the least continuous strategy level. Although declaratory strategy is influenced by the body of theoretical work arising from the academic community and defense "think tanks"—most of the conventional "wisdom" in the current debate is latter 1950s and early 1960s in vintage—it bears the imprimatur of the President or Secretary of Defense who adopts it as official policy, and public officials come and go more rapidly than, say, weapons scientists and engineers. The expert community can and does produce varying shades of opinion so that a new President can assemble an expert team to produce a new, or at least new-sounding, strategy generally conforming to whatever general predilections on defense he may hold. Moreover, defense and strategic affairs have an importance and glamour that appear to make administrations want to place their own distinctive marks on them. Hence, Eisenhower’s massive retaliation strategy gave way to Robert S. McNamara’s controlled response and assured destruction, which successively was replaced by Nixon’s strategic sufficiency, James R. Schlesinger’s limited nuclear options, and Harold Brown’s countervailing strategy. These name changes are often more cosmetic than substantive, but each generally changes declaratory strategy to some degree, reflecting changed perceptions of the threat and the balance of capabilities, among other things.

The other two levels of strategy are more implementary in nature, falling within the realm of military strategy or Desmond Ball’s action policy. Development and deployment strategy actually refers to two distinct operations that are related, since one cannot deploy a weapon system that one has failed to develop in the first place. (The obverse, however, is not true: one can decide not to deploy a successfully developed system.) Generically, development and deployment strategy refers to the process that begins with investigation of the weapons potential of some physical principle to the point that a finished weapon system or component enters the operational inventory. Collectively, the two processes have the purpose of force acquisition, but different actors and dynamics are prominent in each phase.

The development phase of development and deployment strategy refers to the process of scientific endeavor that begins with ideation of weapon systems possibility through the point that a successful weapon system prototype is produced. As such, it is roughly equivalent to the familiar research, development, testing, and evaluation (RDT&E) cycle. In turn, RDT&E can be divided into two subphases suggested by the different operations conducted: research and development, followed by testing and evaluation.¹³

Different actors predominate and make decisions that cumulatively constitute strategy within each subcycle. Research and development is the primary province of basic scientists (e.g., physicists and chemists) and engineers. Decisions about what to investigate and how to solve engineering problems are largely based on scientific criteria about physical properties of the universe. As W. K. H. Panofsky explains, scientific endeavor is relatively insensitive to strategic or policy direction because "‘pure scientists’ take pride in their ability and success in pursuing science for its own sake, unaffected by the potential application of end products of their achievements."¹⁴ Therefore, it is difficult to influence or control what will be discovered; if one already knew what scientists would find in their research, there would be no need for the inquiry. Moreover, the time line on scientific discovery is difficult to predict, much less control: scientific discoveries are made when they are made and cannot be ordered to meet a politically dictated strategic timetable. Efforts to influence the pace and direction of scientific endeavor are indirect, stimulating, or depressing specific research efforts through differential funding levels. As well, many weapon possibilities arise from scientific and engineering in nonmilitary research that may be related to military programs or be wholly unrelated. Often, these contributions are entirely serendipitous.

When basic research yields promising weapon possibilities the fruitfulness of which is a matter of developing practical applications, some decisional discretion occurs. Development is largely an engineering concern, seeking applications of basic ideas and designing prototype weapon systems incorporating the research findings and making engineering improvements on current designs. At this point, however, outcomes are not assured, making assessment difficult, so that decisions tend to be made primarily on the basis of likely technical feasibility rather than on some broader criteria of strategic desirability, and there is a natural tendency to pursue as many promising areas as budgetary constraints will allow. Those individuals responsible for making such decisions, mostly scientists and career officers, bring their own viewpoints and perspectives on the nature of the threat, desirability of certain weapons, and the like, which may or may not reflect the perceptions of political authorities up to and including the President. A classic case in point was President Carter’s purported "discovery" of U.S. neutron (enhanced-radiation) bomb research in a newspaper account of a congressional subcommittee hearing where an Energy Research and Development Administration (ERDA) official unintentionally made reference to the project during testimony.

During the testing and evaluation subcycle, when prototypes undergo operational tests, the results are noted and evaluated and necessary modifications are made; the critical point is in deciding whether deployment recommendations will be forthcoming. To some extent, the criteria for these recommendations are likely to be purely technical: Does the weapon system work at all or up to some usable standard, and is there a mission for it? Two decades of failure in the cruise missile program (largely because of guidance system deficiencies) comes immediately to mind as a major system whose deployment recommendation was delayed because of technically based deficiencies.

A bureaucratic dynamic in this process provides a bridge from development to deployment strategy. In the RDT&E process, weapon systems tend to develop constituencies within the industrial/defense bureaucracy that create internal pressures for positive deployment decisions. The most obvious advocates are those individuals with a direct interest in the system: the scientists and engineers who designed and developed the system; the agency or agencies that sponsored stages of development; and the service or services that would add the system to the operational inventory. Since no one wants the reputation for developing or sponsoring bad ideas, this basis for advocacy is natural and understandable, as is service interest in adding new (and presumably superior) components to the arsenal. Also, those defense industries that would be primary contractors or subcontractors for a system have a direct vested interest in positive procurement decisions.

Although those associated with developmental strategy remain active advocates in pushing for particular deployment decisions, they are not the central actors. Decisions about what weapon systems in what quantities enter the inventory and which cumulatively define deployment strategy are economic and hence political in character. The economics and accompanying politics are evident at a minimum of two levels: in the interservice allocation process of proposing and later dividing up the defense budget; and in the political decision process where defense allocations must compete with other budget priorities. Different actors with differing interests and motivations are involved in each phase of the economic process that supports deployment recommendations, with technologists interested in specific systems and theoreticians concerned with effects on the structure of deterrence in a support role offering expert advice in support of the various contenders. If it is true that policy is what receives funding, deployment strategy is at the heart of nuclear strategy writ large. The large points to be made are that the criteria used in making budgetary decisions are political and economic, they are made by politicians, and those decisions may or may not be swayed significantly by abstract notions about deterrence.

Determining what kind of defense budget will be proposed is largely an executive branch in-house affair. At one level, it is a competition between the services, where each presents its needs and where outcomes expressed as proportions of defense requests and allocations for each service (as well as trends in those percentages) take on both great substantive and symbolic value. At another level, the competition is between the Department of Defense and other agencies, where the chief arbiter and devil’s advocate (especially in the current administration) is often the Office of Management and Budget (OMB). The role of OMB Director David Stockman was particularly prominent within the early months of the Reagan administration.

Ultimately, of course, deployment is based on what Congress appropriates. Internal executive branch political processes result in budgetary tradeoffs and compromises where procurement patterns are altered on the bases both of strategic and nonstrategic requirements. More of the same is likely to occur in Congress when budget recommendations must compete with other national priorities for funding. Although both houses have members expert in defense issues on their Armed Services committees, the ultimate disposition of the defense budget, including those systems that can be procured and deployed, is done by the entire membership, many of whom may vote up or down a particular allocation on grounds entirely divorced from any notion of deployment strategy. The budgetary process is politics in its purest form, and since deployment strategy is the result of decisions about what to buy in what quantities, that level and hence overall nuclear strategy are guaranteed a political content.

Employment strategy, the third level, represents planning for the actual use of nuclear weapons in combat should deterrence fail. The most concrete manifestation is the single integrated operational plan (SIOP). The term SIOP is itself a bit misleading, because the SIOP is and always has been a complex series of different attack scenarios emphasizing varying levels of destruction and different kinds of target sets. Guidance regarding targeting priorities for the SIOP is provided by Presidential memoranda, such as President Nixon’s National Security Decision Memorandum (NSDM)-242 that sought to bring about limited nuclear options and President Carter’s aforementioned Presidential Directive 59. This guidance in turn is "spelled out in the Nuclear Weapons Employment Policy (NUWEP) issued by the Secretary of Defense."¹⁵

Within the parameters established by the NUWEP and the various Presidential memoranda, the detailed SIOP is crafted by the Joint Strategic Target Planning Staff (JSTPS), a body composed primarily of professional military officers. As a nuclear "battle plan," the SIOP serves two broad purposes. First, although its details are secret, its broad objectives are openly available through statements by public officials like former Secretary Brown’s announcement of P.D. 59 (he cited the priorities as "the things the Soviet leaders appear to value most—political and military control, military force both nuclear and conventional, and the industrial capacity to support a war,"¹⁶ a list essentially identical to the priorities listed by Ball in the current plan, SIOP-5D) ¹⁷ and unclassified congressional testimony. Making general contours public serves the deterrent purpose of informing our adversaries of the potential kinds of destruction they might have to endure in response to their nuclear aggression. Second, the plan provides the President with a carefully elaborated set of options for fighting a nuclear war at whatever level of intensity seems appropriate.

That the planning process for employment strategy should be "designed by military men, as a military operational plan"¹⁸ comes as no surprise, since it is the military’s role to plan for, and if necessary to fight, wars. Dominance of the operational element of employment strategy by the professional military does, however, enter yet another distinct set of actors with distinct orientations to the strategy process at this level. Professional officers rarely become involved, at least publicly, in discussions over declaratory policy, and, until recently, most theoreticians have demonstrated only passing interest in employment matters beyond a general preference for counter-value or counterforce targeting. The result is to facilitate a general lack of awareness by one group about what the other is doing and, when interaction does occur, to increase the prospect that dialogue will occur within separate frames of reference.

The fact that different actors operate at the various levels of strategy facilitates independent development at each level, but there is another vexing dynamic that virtually ensures some discontinuity. That problem is the time frame within which each level operates: all three levels have distinctive and independent time lines for their activities that make it virtually impossible to synchronize them at any given time.

Declaratory statements of strategy have the least sensitive constraints imposed by time: a President or Secretary of Defense can issue statements of declaratory strategy whenever he deems it appropriate. Certainly, there are constraints arising from the other levels and externally. A President cannot change strategies too often without appearing indecisive or foolish, and strategy must reflect judgements about what the public will support. Declaratory formulations also reflect the state of activity in the other levels of strategy in two distinct ways. First, declaratory strategy must reflect the current state of the art at other levels, or the declaration will lack credibility (for example, even if one has the perceived will to carry out a strategy, one must also have the hardware).

Second, declaratory strategy is used to provide guidance to and influence other levels of strategy. The motivation underlying assured destruction, as a means to influence the deployment portion of development and deployment strategy, illustrates the point. As Laurence Martin argues, "finite assured destruction was originally more a way of constraining procurement than an operational strategy clearly, thought through and actually intended for execution."¹⁹ In support of this contention, it must be remembered that there was active support within the military and elsewhere to deploy an intercontinental ballistic missile (ICBM) force of 2000-3000 missiles during the 1960s. Moreover, the emphasis on targeting noncombatants was never fully accepted by those responsible for the SIOP, for whom counterforce targeting was always more military and hence natural. As a result, in operational planning " 'assured destruction’ measures were no more than an insensitive—and quantitatively conservative—shorthand for the hideous reality of nearly any full-scale retaliation."²⁰

Whether the function of declaratory strategy is or should be to reflect reality at the other strategic levels or whether the function should be to provide policy guidance from which the other levels deductively flow is, of course, the central question, but the answer is prejudiced by the time line function. Of the three levels, declaratory strategy is least influenced by temporal constraints. Ball’s action levels, however, are much more sensitive to time constraints that are internal to their own processes rather than being the product of external assessments.

Development and deployment strategy is the most obvious case in point. Science proceeds at its own pace, and scientific and engineering discoveries cannot be finely calibrated to a precise timetable. The period from the time of conceiving the idea for a weapon to the time a usable system reaches inventory is generally measured in years. During that process, breakthroughs in development occur but cannot be predicted. The development stage of this strategy level is long and uncontrollable. Furthermore, deployment decisions are made over long periods of time. The arsenal components deployed today are products of research and development efforts initiated in the 1940s and l950s (the designs for systems in the air-breathing leg of the triad are 1940s vintage), and the predicted life span of strategic systems is measured in decades. Given these facts, current development and deployment decisions affect and must be measured against strategic needs in the l990s and beyond, just as decisions made two or more decades ago influence capability and hence strategy today.

Finally, the ultimate transition from strategy to war plan (employment strategy) has its own distinctive temporal dictates. The basic dynamic is that targeting as reflected in the SIOP will inevitably lag behind declaratory strategy and reflect capability resulting from developmental and deployment decisions. The reason follows from the way operational employment strategy is fashioned; the SIOP is constructed using the various methods described earlier as guidance, and it is a time-consuming technical task. According to Desmond Ball, the current plan, SIOP-5D, "includes some 40,000 potential target installations, as compared to some 25,000 in 1974 when NUWEP was promulgated and the development of SIOP-5 initiated."²¹ Since the warhead arsenal is less than one-quarter that size, a significant amount of time goes into setting target priorities. There is also the extremely technical, complex task of matching appropriate warheads from different sources to targets. In this matching process, one must allow for problems like MIRV footprinting limitations and consider cross-targeting requirements. All of this means that it can take years for a fully operational new SIOP to be developed. Thus, there will be a time lag while the new plan is being developed. To the extent the new guidance creates demands for change, a declaratory-employment strategy mismatch is inevitable. Because declaratory strategy and guidances change fairly often, this problem is dynamic and constant.

Too much of the literature and defense debate proceeds as if the problems associated with the interactions between levels of strategy do not exist. The academic debate, centering around assured destruction and its alternatives, rarely gets past the theoretical underpinnings of declaratory strategy, and when it does, its contribution is often a Greek chorus of appall and despair. At the same time, a great deal of the debate occurs as if it were divorced from any political context. At least implicitly, the debate over declaratory strategy assumes a fundamental rationality to the enterprise; once one has accepted certain principles about what deters, the rest is a mechanistic application of those principles. Viewed from the levels of strategy, however, the political elements are revealed as fundamental and critical. Declaratory strategy is made by the nation’s chief politician and his assistants and reflects a variety of political purposes (most prominent of which, of course, is preserving the national existence), and bottom-line deployment strategy is the culmination of the political process, appropriations.

The failure of so much analysis to view strategy in its political context is the most damning indictment of avoiding the levels of strategy problem. Decisions that cumulatively define nuclear strategy are made by politicians, and it is not surprising that those politicians regard strangely recommendations from theorists ignoring that basic reality. The scholarly debate emerges as a theological contest that can safely be relegated to the cloisters. The lack of communications between theorists and politicians frequently results in politically unacceptable strategy and strategically deficient policy.

Implications

In an analytically tidier world, the relationship between the various levels of strategy would be a simple deductive exercise where declaratory strategy was translated precisely into development and, deployment and employment strategies. As has been argued, such a view oversimplifies and distorts reality. In fact, there are discontinuities and even contradictions among the various levels arising at least partially from the two broad dynamics cited earlier: there are different individuals and institutions with different perspectives involved in strategic formulations at each level, and the internal dynamics of each level dictate a temporal sequence to strategic activity that virtually guarantees some discontinuity at any time. It is worthwhile briefly to view the current state of the nuclear debate in the levels of strategy framework.

The heart of the debate that has been going on since the early l970s has largely been over declaratory strategy (limited nuclear options versus assured destruction),²² with residual concern over development/deployment strategy (MX is or is not necessary given a mutual assured destruction [MAD] or limited nuclear options [LNOs] declaratory posture) and employment strategy (counterforce or countervalue targeting is or is not compatible with MAD or LNOs). Particularly when the debate is extended beyond declaratory strategy, there is at least the implicit assumption that development/deployment and employment strategies do or should flow deductively from current declaratory positions. Whether such a relationship ought to exist is a philosophical question that can be debated; such a formulation contradicts the way the process operates.

In one sense, the whole debate is, in Shakespeare’s phrase, "much ado about nothing." Certainly the debate about MAD and LNOs is overblown, in the sense that, at the operational level, MAD has always contained more finite targeting objectives (employment strategies) and the LNO position admits all-out countervalue exchange as the ultimate possibility, whether it is featured or not. A debate focusing on "pure" MAD or LNO positions hence distorts the policy debate, which occurs over shades of emphasis rather than at the extremes.

Understanding that declaratory strategy is neither MAD nor LNOs but rather the part of the mix emphasized serves two essential purposes. First, it moves the debate away from the extreme ends of the poles back toward the middle ground where real policy debates among those political and military actors who devise strategy occur. In the process, we create the possibility that academics and strategy makers can engage in dialogue instead of talking past one another. Second, understanding that changes in declaratory strategy are matters often of subtle reemphasis and repackaging creates a greater sense of continuity to the strategy process than does viewing the formulation of declaratory strategy questions in either-or terms. In the process, this recognition promotes an appreciation of the continuities rather than the discontinuities between the levels of strategy.

Linkage becomes apparent with both employment and development/deployment strategies. At the employment strategy level, recognizing that declaratory strategy in fact has always dictated a range of strategic options makes more natural a dual emphasis on countervalue and counterforce targeting, since limited options imply selectivity in targets attacked and these quite naturally contain counterforce objectives. Given the natural military professional inclination toward attacking combatants (counterforce objects) rather than noncombatants (counter-value objects), a counterforce-oriented SIOP (and guidance therefore as in P.D. 59) represents not so much a change in philosophical positions over what kinds of threats deter best as it does an improved linkage between declaratory employment, and development/deployment strategy. Developments in weapon systems capabilities are expanding the list of counterforce objectives that can be targeted. Furthermore, these advances in weaponry permit greater flexibility in one’s response to changes in adversary offensive and defensive capabilities. Such developments are a natural outgrowth of technological processes both in the U.S. and U.S.S.R. and reflect no more than the dynamic nature of weapon science.

Strategy that emphasizes a variety of options also suggests a development and deployment orientation investigating a wide variety of possible capabilities. This observation is clearly true within the development cycle of this level of strategy, but true discrimination occurs when deployment decisions are made. Within this cycle, political actors are most prominent, and deployment strategy is often effectively formulated on bases that are largely nonstrategic (for example, budgetary tradeoffs) rather than on the basis of clearly articulated deterrence grounds. Tradeoffs and compromise are the basic stuff of politics, and as long as the process does not produce strategically unacceptable outcomes (which it has not to date), it is natural and not pathological.

These dynamics, suggesting both sources of continuity and discontinuity, are complex and have some clear implications for theoreticians and practitioners alike. Two implications stand out for theoreticians (undoubtedly there are others). On the one hand, deterrence strategy as a complex interaction of the various levels of strategy clearly suggests that concentration on any one level is inadequate. The disservice such an emphasis provides is vividly demonstrated in shock and dismay over P.D. 59. If one had been looking at questions of MAD versus LNOs exclusively, the pronouncement appeared a dramatic and definitive statement of philosophy; viewed from the level of employment strategy as influenced by development and deployment strategic decision-making, P.D. 59 was little more than an incremental link in an ongoing process.

On the one hand, and relatedly, this mode of analysis suggests that theoreticians need to broaden their horizons to encompass all levels of strategy if they are interested in influencing policy decisions that affect the deterrent condition. Probably most critically, this implies the need to become involved in the critical deployment level where decisions are made that define arsenal characteristics, targeting possibilities, and limitations, and, hence, the capabilities that make different declaratory strategies credible or incredible. That involvement is likely to be the most effective in pointing out the relationship between the theoretical and the concrete. If nothing else, the contribution may be best in pointing out the long-range, nonobvious impact of discrete decisions on the structure of deterrence. If policy is indeed that which receives funding, the critical intervention point, where the greatest impact can be registered, is the political processes leading to funding decision for various patterns of force deployment.

For practitioners, the problem is not understanding the process, it is coordinating the levels better. At the operational levels of development/ deployment and employment strategy-making, there is too often only a shallow awareness of the theoretical implications of various decisions and a resultant surprise when objections are raised. At the same time, coordinating activity at the various levels more tightly can avoid logistical difficulties in selling strategies, as the P.D. 59-MX controversy illustrates. In logical fashion, the sequence of policy decision would have flowed from countervailance as declaratory strategy (emphasizing limited options) to P.D. 59 as employment strategy (to determine target coverage patterns necessary to carry out identified options) to developmental/deployment strategy to provide the necessary hardware for the employment strategy (the most obvious need arising from such assessment being additional warheads, which MX would provide). Steps two and three were reversed, and the result was contention. MX was criticized as providing excess, redundant counterforce capability, and P.D. 59 was condemned as revolutionary heresy. Closer coordination among the levels of strategy would not have removed controversy because the whole concept is controversial. Viewing the process as a sequential levels-of-strategy problem would, however, have reduced the confusion.

The bottom line is a plea to look more closely at all the levels that produce nuclear strategy, with the hopeful result that those making decisions at the various levels will be capable of meaningful dialogue that will produce better strategy at each level. Targeters need to realize that strategy is more than the SIOP, politicians need to understand that budget-driven decisions may have strategic implications far beyond the impact on the federal debt, and theoreticians need to realize that the constraints of the real world make deterrence strategy-making something other than an exercise in deductive logic. It all seems so obvious that it hardly bears emphasis, yet the evidence suggests that the obvious has not been so self-evident after all. Recognizing that there is a levels-of-strategy problem does not solve the disorder of nuclear strategy, but it at the least makes more sense out of why the disorder exists.

1. See particularly, "Nuclear Strategy: The Case for a Theory of Victory," International Security, Summer 1979, pp. 54-87; "Targeting Problems for Central War," Naval War College Review, January-February 1980, pp. 3-21; and (with Keith Payne), "Victory Is Possible," Foreign Policy, Summer 1980, pp. 14-27.

2. For a particularly eloquent statement, see Robert Jervis, "Deterrence Theory Revisited," World Politics, January 1979, pp. 289-324.

3. Nigel Calder, Nuclear Nightmares: An Investigation into Possible Wars (New York: Viking, 1980), pp. 13, 10.

4. My own assessment of these sources of contention, including documentation, can be found in "Strategic Uncertainty and Nuclear Deterrence," Naval War College Review, November-December 1981, pp. 27-41.

5. Some of my thoughts can be found in Nuclear Strategy in a Dynamic World: American Strategy in the 1980s (Tuscaloosa, Alabama: University of Alabama Press, 1981), pp. 96-99 and 205-16.

6. For an overview, see Donald M. Snow, "The MX-Basing Mode Muddle: Issues and Alternatives," Air University Review, July-August 1980, pp. 11-25.

7. For a somewhat speculative overview of the prospects, see Donald M. Snow, "Lasers, Charged Particle Beams, and the Strategic Future," Political Science Quarterly, Summer 1980, pp. 277-94; or

Barry J. Smernoff, "Strategic and Arms Controls Implications of Laser Weapons: A Preliminary Assessment," Air University Review, January-February 1978, pp. 38-50.

8. "Remarks Prepared for Delivery at the Convocation Ceremonies for the 97th Naval War College Class," Washington, D.C.: Office of the Assistant Secretary of Defense (Public Affairs), 20 August 1980, p. 6. Emphasis in original.

9. Dennis M. Drew, "Strategy Process and Principles: Back to the Basics," Air University Review, May-June 1980, p. 39.

10. Ibid. Emphasis in original.

11. Desmond Ball, "Dynamics of the US Strategic Nuclear Policy Framework," Defense and Foreign Affairs, April 1981, p. 5.

12. Ibid., p. 4.

13. One of the best discussions of this process remains Richard G. Head, "Technology and the Military Balance," Foreign Affairs, April 1978, pp. 544-63.

14. W. K. H. Panofsky, Arms Control and SALT II (Seattle: University of Washington Press, 1979), p. 3.

15. Ball, p. 4.

16. Brown, p. 7.

17. Ball, p. 6.

18. McGeorge Bundy, "Strategic Deterrence Thirty Years Later: What Has Changed?" in Christoph Bertram, editor, "The Future of Strategic Deterrence, Part I," Adelphi Papers No. 160 (London: International Institute for Strategic Studies, Autumn 1980), p. 8.

19. Laurence Martin, "The Determinants of Change: Deterrence and Technology," in Christoph Bertram, editor, "The Future of Strategic Deterrence, Part II," Adelphi Papers No. 161 (London: International Institute for Strategic Studies, Autumn 1980), p. 13.

20. Bundy, p. 8.

21. Ball, p. 6.

22. Donald M. Snow, "Current Nuclear Deterrence Thinking: An Overview and Review," International Studies Quarterly, September 1979, pp. 445-86.

Contributor

Donald M. Snow (B.A., M.A., University of Colorado; Ph.D., Indiana University) is Professor of Political Science and Director of International Studies at the University of Alabama, Tuscaloosa. He was a visiting professor at Air Command and Staff College in 1980. Dr. Snow is the author of The Nuclear Future: Toward a Strategy of Uncertainty (1983), Introduction to World Politics (1981), and Nuclear Strategy in a Dynamic World (1981) and is a frequent contributor or to the Review.

The conclusions and opinions expressed in this document are those of the author cultivated in the freedom of expression, academic environment of Air University. They do not reflect the official position of the U.S. Government, Department of Defense, the United States Air Force or the Air University.

Air & Space Power Home Page | Feedback? Email the Editor