Uncertainty is necessarily the lot of the planner, since the deals with the future. Uncertainty can never be completely removed. However, it can be compensated for, and to do so is a continuing responsibility of those who plan military forces. Primarily this can be done by insuring, in so far as we can, that future weapons and forces will be adaptable to the right range of defense needs or, as defense planners often put it, by insuring flexibility.

Modern weapons need not of themselves inexorably beget more sophisticated, expensive and destructive devices. It is true that pressures exist which, if unrestrained, would needlessly proliferate and duplicate weapon designs. There is a tendency to do what potential opponents are doing, or thought to be doing, or predicted to be able to do. But restraint and direction are exercised. In acquiring the capability to defend the United States, direction is given and decisions are made at the highest level, by the President with the advice of the Secretaries of Defense and State. Assumptions, concepts, economic analyses, divergent estimates of potential danger to our national security-all periodically are translated into an explicit defense program for the United States. At least twice a year the Department of Defense updates this program and projects forces five years or more into the future. Specific decisions flow from a continuing interaction between perceptions of what is technologically and economically feasible, and defense policy-the broad guidelines for achieving national objectives by the direct or potential use of armed force. In turn, specific program decisions and their subsequent results condition both future weapons development and future defense policy.

To say that there is uncertainty in tomorrow or virtue in flexibility is hardly novel. Reconciling the two, however, is a job calling for much diverse information and refined analysis. To evaluate the process, we must examine further the uncertainty which surrounds weapon choices and consider how best to minimize it and live with it. Let us first consider some uses and limitations of intelligence estimates and then examine the interaction of strategy and technology. In the light of these two exercises, the timing of decisions leading to full-scale weapon development takes on special importance, because such decisions rough out critical boundaries of the capabilities of our future forces and the sensitivity of that capability to the actions of others. Finally, within these boundaries, we can attempt to indicate some general factors which enhance force flexibility and some examples of the sort of flexibility which may be demanded of weapons currently under development.


There are, of course, numerous broad areas where uncertainty complicates choices about the composition and overall magnitude of our forces. What, for example, are the prospects in fifteen or twenty years for widespread nuclear proliferation, for shifts in alliances, for a largely disarmed world, for increased or decreased Soviet or Communist Chinese militancy, for external aggression and subversion directed against developing nations, for a world realignment leading to conflict between the developed and the developing? Here the face of possible long-range political and strategic futures varies from perpetual peace to Armageddon. The temptation is strong to satisfy one's own policy predilections by immediately trying to leap over the ravine of short-range and mid-range uncertainty in order to program developments and deployments which meet (or encourage) one or two of these long-range futures. But this ravine-the next five years-is far better lighted than the distant future. Within it the most pressing aspects of external military challenges appear. Here uncertainties about specific opposing deployments narrow; estimates of opposing capabilities in ballistic missile defenses, bombers, bomber defenses, submarines and advanced tactical aircraft become more precise. The pattern they present becomes extremely important in stimulating shifts in our own defense policy and development priorities.

By and large, our possible responses to five-year forecasts have to do with force procurement and deployment. At ten years, the forecasts are much less reliable, and our responses tend to become related more to questions of the development of particular weapons and systems. However, whether the forecast is for five or (less reliably) for ten years, intelligence gleaned across this wide range of subjects plays a critical role in determining what weapons and forces to proceed with part way, all the way or not at all, and in the timing of specific decisions.

Intelligence estimates provide no panacea, particularly if one insists on viewing policy-making as beginning with the "threat," a protean word whose meaning includes intent as well as capability, and embraces not only the number of ballistic missiles an opponent may have but the number of his targets that we want to be able to destroy. Because our planning is projected forward at least ten and sometimes fifteen years, intelligence estimates are made of what forces the other side will have far in the future. This may be a necessary exercise, but we sometimes tend to depend too much on the results. We may know with reasonable precision how many missiles the Soviets will have in their active force a year or two from now, because they must already have begun work on their deployment. We may have a fair idea how many they will have four or even five years from now, though a less precise idea of what kind of missiles and what their payloads will be. Yet in fact we know very little about how many or what kind of missiles the Russians will have ten years from now, because they probably have not reached a decision themselves. And even if they have, they might change their minds. Consequently, estimates beyond the immediate future enter, at best, a realm of educated extrapolation. They can readily pass beyond intelligence information, through telepathy-what others plan but have not done anything about-to clairvoyance-what they have not yet decided.

How many missiles the Soviets will have at the beginning of 1977 does not, of course, depend solely on their technical and productive capabilities-on what they can do. That is one kind of intelligence estimate often given- generally by people who want to "play it safe" by gearing our plans to the worst imaginable (even if not feasible) contingency. The development of Soviet forces will be influenced by the technical possibilities, of course, but even more strongly: (a) by Soviet military and foreign-policy goals; (b) by the Soviets' own fiscal requirements and allocations; and (c) by what the United States does, and the signals that we transmit, accurate or inaccurate, intentional or unintentional, about what we plan to do. When we try to look ahead ten to fifteen years, what we get are not intelligence estimates at all, but planning estimates. They should be treated as such. The "estimate" becomes more an examination of the question: "If countries A, B and C plan well (or poorly), what will they do in response to U. S. course of action X or Y?" This is the kind of question we ask ourselves frequently about military and political actions projected over the next year or two. We must learn to ask it (and try to give an answer) about developments and deployments other countries may make over a period of the next ten years. Our objective in military planning should be to provide ourselves with the capacity to take initiatives of our own, or to make timely responses to the force deployments of others at later times, as their actions inevitably become much clearer to us.

We are rightly concerned about the time that elapses between a decision to acquire a certain capability and the actual existence of that capability. This lead time may be as long as ten years. It can, of course, be compressed to some extent by giving a development program high priority. The urgency of a particular program often is based on intelligence information about an opponent's technical capabilities (or an imputation to him of our own present or future technical capabilities) and-always less certain-his intentions. But lead time cuts both ways. We often forget that our opponents also require substantial time between decision and the achievement of a particular military capability. We sometimes are tempted to act as though the Soviets have a capability in being as soon as our calculations indicate that a particular piece of hardware is technically feasible. In fact, to develop, produce, deploy and train is no mean task for any nation when large and complex new weapons systems are involved.

A good example of the two-edged nature of lead time and intelligence estimates is the intercontinental ballistic missile, which the Soviets were working on in the early 1950s. Knowledge that they were developing an I.C.B.M. was a major factor in the initiation of our own program in 1954- 55. But the Soviets did not have a substantial I.C.B.M. capability until 1962, an elapsed time of about ten years. Starting two or three years later, we managed to finish about two years earlier than they-in addition to developing during the same period the solid-propellent Minuteman and an entirely separate ballistic-missile system, the submarine-launched Polaris. This is not to say that we can always make up a late start that well. Complacent contemplation of past achievements would be extremely dangerous. However, this example does show that the interval between our discovery that the other side is thinking about something (or our conjecture that they are thinking about something because we just thought of it) and the time they actually have a useable weapon generally will be long enough for us to take appropriate counteraction if our decisions are made expeditiously. Our counteraction need not always take the form of a parallel response. If we have been skillful in our evaluation of intelligence and wise in our choices of options, we may offset a particular challenge by making a quantitative rather than qualitative change in weaponry, by adaptation or minor improvements in existing weapons, or by producing a new weapon for which early development has already been done. On occasion we may also get good results by making a doctrinal change-that is, a change in how existing weapons are used.


Whether strategy or technology comes first is (perhaps fruitlessly) debatable, though flow charts frequently are drawn by operational or technological planners depicting one or the other as the fount of most really significant activity. There appears to be no foolproof cyclical pattern encompassing the two, nor any yin-yang replacement of one by the other. If, as some suggest, a pendulum swings between strategy and technology, its period has not been convincingly measured. At a given time when we ask what technologies should be developed, strategic aims must provide the direction. Yet, if we examine how our strategy got to be what it is, clearly technology had a big part in getting us there. Broad strategic direction must be given if there is to be coherence in and among development programs and the military forces which result. But the effect of technology on strategy also should be continuous. Technology provides opportunities which may or may not be recognized immediately; strategy, while usually dominant in terms of authority, must be informed by knowledge of these opportunities if it is to be fully effective.

At the onset of the cold war in the late 1940s, technical factors-our possession of the atomic bomb and aircraft which could deliver it-dominated our military strategy. The American monopoly or preponderance of nuclear power and its relatively low cost, combined with our abrupt demobilization of conventional forces, set in motion a strategy heavily committed to deterrence based on the Strategic Air Command's capability for nuclear retaliation. This strategy was, of course, a response to the evidence of Communist aggressiveness that accumulated in the years immediately following World War II, and it lasted through the 1950s. After the Korean War, the Air Force consistently received just under half of the total defense budget, and the lion's share of its money went to strategic forces.

However, as the Soviet Union strengthened its own nuclear capability, the mutual devastation to be expected from a general nuclear exchange increased the likelihood of less dangerous forms of aggression. By 1961, strategy per se probably regained ascendancy over technology. Strategy would remain partially captive to nuclear technology; but in limited confrontations the end of extreme U.S. nuclear predominance had put a premium on uses of military force more believable than threatening the extinction of powers which, as they died, could cause the United States grave damage. Consequently, our ability to apply carefully measured, limited force increased in importance, though our ability to assure the destruction of any aggressor who might strike us retained first priority, as it does today. Development emphasis shifted to modernization of conventional forces, while long-range strategic nuclear capabilities were concurrently improved.

Since technology and strategy are so closely related and since so many fields of technology are growing increasingly complex and expensive, it is often assumed that lead time grows correspondingly and that there is unbearable danger in not having every technological bet covered. This assumption is too simplistic, and it would be impossible to pursue every promising idea to the point of producing hardware. Instead, we must seek innovation by providing adequate resources for basic research and exploratory development. We must hold open our option to adopt weapon systems which look promising, by developing those components which require long lead times. But we have no business buying insurance against every conceivable action by potential enemies, without regard for how likely or profitable that action may be.

Here the purposeful influence of strategy can be valuable, for it lends central direction and consistency to the commitment of technological resources. It determines whether a project should be pushed hard, placed on the back burner, or taken off the stove altogether. During the 1950s, for example, the Jupiter intermediate-range ballistic missile enjoyed a high priority as a feasible interim system until missiles of intercontinental range could be developed; hence, this radically new weapon took only three years to develop. On the other hand, in small-arms development, which during that period was given a low priority, the M-14 rifle was scheduled for eleven years in research and development and was to be followed by gradual replacement of the old M-1 rifle over a period of seven years.

An important part of the interaction between strategy and technology is a nation's total technological and productive capacity. In spite of current political evolution away from bi-polarity, there remain two supertechnical powers in the world: the United States and the Soviet Union. There is a good deal of merit in making clear to third powers that in the face of the capabilities of the two supertechnical powers the entry stakes into this technological competition are higher than the return in military capability and the political power it buys. This is a most important message to communicate, for the acquisition of weapons of mass destruction by additional powers, including some nations whose limited resources could be better used for their economic development, is likely to be far more destabilizing than acquisition of new weapons by the U.S. and U.S.S.R.

We have security today because deterrence has worked, and to the degree that deterrence remains assured we are more secure than in the past. But in another sense we are less secure, because the physical consequences of the "failure" of deterrence are worse. Technology has caused this. But since the Soviets detonated an atomic device in 1949, we have demonstrated that, with two principal nuclear powers involved, deterrence of general nuclear war can remain reasonably stable despite successive qualitative advances in weaponry, and despite asymmetry of purpose between the two great powers since World War II.


Considering the limitations of long-range intelligence and the complex interaction between strategy and technology, there is much to recommend a pragmatic approach to the development of major new weapons. The decision to proceed with full-scale development of a major new weapon system is an important event; almost without exception it indicates a large commitment of resources, and it defines, limits or expands the conditions under which future combat forces will be able to fight effectively. Normally the decision is made by the President or at the highest level in the Defense Department with Presidential concurrence.

The first important indication of full-scale development comes when the Secretary of Defense approves "contract definition" for a new weapon. It follows a period of "concept formulation," the thorough analysis of military need and concept of operation, studies of technical feasibility to determine "best" characteristics, cost and scheduling. The objective is to check earlier assumptions and to get a good contract, one which is unambiguous and is based on a fixed price or incentives which will properly motivate the contractor. The conditional decision to proceed with full- scale development can then be reviewed on the basis of firm costs, schedules and capabilities, and of any changes in conditions since the contract definition began.

Because the commitment of resources is large, decisions to undertake full- scale development or deployment must be made deliberately. Deliberation does not imply unwarranted delay, however.

The decision to go ahead at least a small part of the way may be made without irrevocable commitment to complete the project. Research and early development generally require 30 to 40 percent of the total time from conception to operational capability, and account for perhaps 5 to 20 percent of the development cost. The number of options that one can afford to take out in this way, while not unlimited, is none the less large. In sufficiently important cases we may have to go through full-scale development while reserving decision on deployment until the indications of need become apparent. An indication might be an opponent's deployment of a similar system, or more likely, of a related one which ours may be designed to neutralize. If we start after the other side has begun to deploy a new system, we may not have time for both development and production. Nike-X is an example of a system where development was pushed without a decision having been made as to deployment. However, it should be noted that a deployment decision on Nike-X depends on the nature and weight of the offensive forces we may expect to be used against us rather than on whether another country displays an anti-ballistic missile system. Unless the defense is much less expensive than the offense (informed analyses show it to be more expensive), the correct counter to an opponent's A.B.M. deployment is an improved or increased offensive force.

Here again the directing and restraining influences of strategy deserve emphasis. If anything, strategy and the criterion of utility in support of national objectives merit increasing influence on development decisions. The development community argues too often for going ahead "because you can do it." This is a fine reason for mountain climbing, but not for multi- billion dollar development programs. In building big systems, developers lacking explicit guidance tend to do the "best" job they can. A hazy operational concept encourages excessive performance requirements at the cost of utility, reliability, simplicity and early availability. Conversely, but less frequently, the same effect can result from too detailed an operational concept or one which through ignorance makes excessive demands on available technology.

The health of our technological base directly supports our ability to move ahead rapidly with full-scale development. Contrary to the belief of some critics, there is no paucity of basic research and exploratory development money. The Department of Defense spends about $ 1.5 billion a year for these categories alone and $800 million more for advanced development short of full systems development.1 Though none of this is tied in detail to a specific military strategy, very little of it is largesse to be dispensed annually for the general advance of technology.

Development of technology represents one level of effort; development of big systems is an entirely different and far more costly effort. Costs rise dramatically as we leave research and exploratory and advanced development and move into engineering development and procurement in quantity of equipment for the field. The jump is likely to be from a level of tens or hundreds of millions for advanced development to a billion or even ten billion dollars for deployment of a major weapon system. Hopefully, by the time we have completed 5 or 10 percent-the development part-of the total cost of a large system, we can be confident of a project's high operational utility. But this magnitude of cost is a compelling reason for the Secretary of Defense to insist-as he does-on a convincing demonstration of operational utility within an overall national defense strategy before he will make decisions tantamount to full-scale development and procurement. This is precisely the problem faced by advocates of large anti-missile defenses or a follow-on bomber to replace the B-52 Gs and Hs.[i]

The latter, for example, has required a massive study effort to determine its characteristics. Less than two years ago, few would have predicted that the 6-52 would be delivering conventional bombs on the scale we have seen in Viet Nam. Yet we must attempt to predict whether this role will exist for the manned bomber 15 years from now. Bombing guerrilla supply areas with non-nuclear weapons requires a large payload, but high speed is not essential as long as enemy air defenses are absent or relatively primitive. On the other hand, in a nuclear role supplementing our strategic missiles, a bomber may need high speed to penetrate advanced defense systems, but a very large payload may be less important. Is a satisfactory compromise feasible? Do we need two distinct types of manned bombers? Should we develop an optimum bomber for only one role? If so, which role? The questions whether to go ahead with any new bomber, if so what kind, and how far to commit it to development thus bring into focus a wide range of complex technological questions which must be squared with an even broader range of strategic and political issues.

Sometimes the divergence between the needs of strategy on the one hand and the use of technology to advance a poorly thoughtout concept of operation on the other leads to cancellation of a development project before commitment to, or before completion of, full-scale development. Invariably the result is traumatic for individuals who have invested years of talent and energy in the project. Nevertheless, in varying degrees a cancelled project leaves a heritage of improved technology. The XB-70 program has significantly advanced our knowledge of structures, engines and aerodynamics of large aircraft in supersonic flight. This knowledge cost us a billion and a half dollars. It is doubtful that the information was worth that price. But full-scale development and acquisition of the B-70 truly would have been a waste, and of a different order of magnitude. It could easily have led to a $10 to $15 billion program, inadequately related to strategy. The same is true of many other programs.


If continued military strength is still necessary to achieve a more peaceful world, the single most important characteristic of future forces will probably be flexibility. The United States must be able to deter or combat aggression in such diverse forms as subversive insurgency or general nuclear war. The variety of contingencies which our military systems must be prepared to meet is very great, and the most efficient composition of tactical forces employed in future limited conflict can never be fully known long before the fact.

The development of tactical forces inevitably has called for choice and compromise. Our technical capabilities and the variety of the possible small-war situations we may face present us with far more options than our resources will allow us to take up. For example, to design one or more types of aircraft solely for each specific tactical air function-air superiority, interdiction, reconnaissance, close support-not only would reduce the flexibility of our tactical air force but would be prohibitively expensive. Also, we must make some compromise in meeting the vast numbers and kinds of contingencies that may occur in widely differing environments: from the arctic to the tropics, from Europe to Latin America. This is why our tactical aircraft have been designed or modified to provide a dual capability, nuclear and non-nuclear, and to perform outstandingly in one tactical function and creditably in one or more others-thus providing the total force with great flexibility at all levels of conflict. But against a capable enemy, all people and all units cannot do all jobs equally well-if they are to do anything well enough. To meet unique missions and fight in unique environments requires a high degree of specialization, and this in turn requires an ability to train or retrain skilled personnel rapidly; a modern, dynamic and rapidly expandable training capability is absolutely essential to force flexibility.

In the emphasis and publicity given to items of hardware, it is sometimes forgotten that the first necessity is for a highly qualified professional force-in-being capable of quick reaction and rapid doctrinal change. Adaptability and willing institutional support, as opposed to mere obedience, are important elements of a total capability. These are not achieved overnight, although the policy decision to effect change may have to come that quickly. The important requirement is that military doctrine be alive and responsive to shifts in national policy-whatever the individual frustration which may be caused by shifts in the prominence of particular types of weapons and military units.

Second, flexibility is improved by the existence of competent, well- equipped and readily available reserve forces. Whether or not they are activated in a given crisis, their availability provides greater leeway in the employment of active forces. As in the 1961 Berlin crisis, they also serve to reinforce the known U. S. strategy, lending credence to a position which it is most important for the opponent to comprehend fully.

Third, flexibility is enhanced by a broad research and industrial base. We cannot hope to equip our forces completely for every conceivable contingency. However, in particular combat situations, such as Viet Nam, the cost of achieving success can be reduced by a capacity to make swift innovations tailored to the immediate circumstances. In this respect the United States has a relative advantage because of our unequaled economic and technical strength.

These three factors, among others, greatly affect our ability to maintain flexibility in our armed forces. But current choices of specific weapons still under development will also set future boundaries within which policy initiatives can be taken. Acquisition of these weapons will generate new political problems and ease others. It may be instructive to consider a few of these developments and how they both derive from and influence defense policy.

Dramatic increases in range, payload and versatility for both our fighter and airlift forces come quickly to mind-specifically the versatility of the variable-wing F-III fighter and the huge C-5 cargo aircraft. Both will soon be in our inventory. Ten C-5s could have handled the entire Berlin Airlift, which required more than 140 C-54s (the equivalent of a civilian DC-4). The range of these two aircraft, without refueling, will give us much wider choices as to how many troops and supplies need to be based overseas. For example, the "basic mission" of the C-5 calls for delivering a 100,000- pound payload a distance of 5,500 nautical miles. For shorter ranges of 2,500 miles, the payload increases to 265,000 pounds.

During Operation Big Lift in 1963, we airlifted 15,000 troops, a combat division, from Texas to Europe in 63 hours. This took 204 aircraft. Forty- two C-5s could do the same job in 13 hours. But while Big Lift was a great operational success, even with older aircraft, a number of Europeans expressed apprehension about American "neo-isolationism." While this concern was ill-founded, it is apparent that the broad political implications of our mushrooming airlift and fighter capability have to be continually reëxamined. The structure of our overseas bases will certainly be influenced by the increased capabilities of our troop and cargo carriers and fighters. It will also depend on many strategic factors, such as our estimate of how rapidly conflict might build up in specific locations, In some places, such as Europe, political and military considerations may require us both to have forces immediately available and to have others with which we can reinforce rapidly.

Still another area where current development is likely to affect future policy is our response to subversive insurgency. The Defense Department has given high priority to the development of weapons designed for the unique environment of Southeast Asia. Moreover, we are cutting lead times sharply. Thus far we have been quite successful in translating known technology into more effective non-nuclear ordnance and improved capability for aerial reconnaissance, finding targets, night and all-weather operations, and bombing accuracy. But our performance in these areas still leaves much to be desired; enormous changes for the better can be expected in the future. We are also seeking imaginative and effective ways to remove or transform incipient sources of insurgency wherever they may occur. For example, one of the most important tasks of Air Force Special Air Warfare Forces is to help the people in developing areas to better their economic and social environment. We would prefer to teach indigenous military forces how to use aircraft to fertilize crops rather than to drop bombs, to improve rather than interdict transportation. The problems of these and comparable American units, such as the Army's Special Forces, are as much political as military. Undoubtedly, their experience and success in Viet Nam will have a great bearing on their role in future programs.

Our greatest continuing concern must inevitably be in that area where we stand to lose the most, quickest: general nuclear war. Here we can afford no gaping holes either in our capacity to retaliate or in the technology to ensure that capability. Our most important current effort is the modernization of our strategic offensive forces through acquisition of Minuteman III and Poseidon missiles. This constitutes a forecast of our intention to deter general war by making it unprofitable for any potential enemy. It is important that enemies, allies and neutrals alike be spared misinterpretation. Even beginning development of a particular weapon system can have such an effect, since foreign perceptions are influenced before the weapon is translated into combat capability. By maintaining strategic deterrence, we gain considerable freedom at lesser levels to employ technology in support of strategies that will make limited aggression increasingly unprofitable.

Until greater congruity of purpose exists among nations, military forces will continue to be a foundation of policy. We must be wise enough to manage the interaction between weapons development and defense policy so that our weapons are always responsive to policy and our policy is based on a full consideration of the options made available by technology. This will remain a supreme challenge to policy-makers, who from a position of final responsibility must look at a partially obscure tomorrow and a distant future that seems opaque. They will find, almost invariably, that it is easier to produce military hardware than it is to know what policy to follow.

[i] This is aside from perhaps another $2 billion in support of the equivalent categories of broadly applicable research and technology from other government agencies-the National Aeronautics and Space Administration, Atomic Energy Commission and others.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now