153. Draft Memorandum From Secretary of Defense McNamara to President Johnson0


  • Recommended FY 1965-FY 1969 Army and Marine Corps General Purpose Forces (U)

I have recently completed my review of the Army and Marine Corps General Purpose Forces for FY 1965-FY 1969. The recommended program will form the basis for the preparation of the FY 1965 Budget. This memorandum summarizes the main factors I have taken into consideration in determining United States requirements for these forces.

I believe we should adopt, for planning purposes, the Army and Marine Corps force structure summarized in the table on page two. Where they differ from my recommendations, the forces proposed by the Army are shown beneath mine in parentheses. In particular, I recommend we:

Maintain the current Army structure of 16 Divisions, 7 Brigades and 4 Armored Cavalry Regiments with support elements.
Maintain the present active force of three Marine Division-Wing teams.
Continue the program to evaluate the Air Assault Division begun with your approval last year.
Provide $2,081 million in FY 1965 for procurement of Army weapons, equipment, and materiel.
Provide about $227.3 million for procurement of Marine Corps weapons and equipment, exclusive of aircraft.

In addition, I recommend disapproval of Army proposals to expand the Air Mobility tests; to add additional maneuver battalions to existing divisions; to add additional combat and support units to STRAF; and to add two Hawk battalions. The Deputy Secretary also recommends inclusion of $45.6 million in the FY 1965 Budget for initiation of procurement [Page 566] of Shillelagh.1 Appendix 1, page 36,2 presents a more detailed basis for these recommendations.

[Here follows a tabular breakdown of all requested general purpose forces.]

The Marine Division-Wing teams I am recommending are those recommended by the Joint Chiefs of Staff and proposed by the Secretary of the Navy. The Army divisions and brigades I am recommending are concurred in by the Joint Chiefs of Staff, except for the Chief of Staff, U.S. Army, who with the Secretary of the Army recommends the larger forces shown in parentheses in the table on page two. In the case of the Air Assault Division, the deployment decision can be deferred until next year. Since the test program I have approved is still in its early stages, a decision to program this division at this time would be premature.

The estimated total obligational authority required to procure and operate these forces is shown in the following table.

Total Obligational Authority by Fiload scal Yeara (Billions of Dollars) FY 62 FY 63 FY 64 FY 65 FY 66 FY 67 FY 68 FY 69 FY 65-FY 69
Prev. App’vd. 6,328 6,195 6,765 6,411 6,090 5,860 5,744 ——
SecDef Recom. 6,353 5,614 5,569 5,494 5,509 5,292 27,478
Army Prop.(18 Div.) 6,765 7,370 7,584 7,356 7,169 6,876 36,355
Marine Corps:
Prev.App’vd. 1,348 1,436 1,017 931 839 688 738 ——
SecDef Recom. 954 1,014 876 809 776 409 3,884
USMC Proposed 1,003 1,250 993 863 911 409 4,426

Because the issues and risks associated with decisions about the major combat units of the Army and Marine Corps are so important, I will discuss in Section II of this paper the reasons underlying my force recommendations.

[Page 567]

II. General Basis for Force Level Recommendations3

This analysis of our ground force requirements is based on consideration of these six areas:

A basic statement of strategy.
A comparison of NATO and Warsaw Pact ground force strength.
The special situation in NATO.
U.S. force requirements for various conflicts around the world.
Tactical nuclear posture and strategy.
Conclusions on contingencies, force structures, and deployment.

A. A Basic Statement of Strategy

The Joint Chiefs of Staff state the objectives on which General Purpose Forces should be based as follows:

“U.S. General Purpose Forces should be sufficient in quantity, quality, and mobility to meet the early reinforcement requirements of NATO and, while maintaining current or similar overseas deployments, meet the estimated requirements of any one of the most likely contingency plans of the commanders of the unified and specified commands. In combination with available Allies, these forces should have the following capabilities without a major deployment of reserves other than the Marine 4th Division Wing team:

  • “a. Resist, without using nuclear weapons, major non-nuclear assault by Soviet Bloc forces against areas where vital U.S. interests are involved. As a minimum, U.S. and Allied forces should be able to hold long enough to convince the communists of the risks involved in their present course of action.
  • “b. Defeat, without using nuclear weapons, non-nuclear aggression by Soviet Bloc forces at levels less than major assault in areas where such aggression would directly threaten U.S. vital interests.
  • “c. Execute successfully without counting on the use of nuclear weapons any one of the contingency plans of the unified and specified commanders outside the NATO area, [3-1/2 lines of source text not declassified].
  • “d. These general purpose forces must provide an adequate force structure to incorporate, protect, and employ tactical nuclear weapon systems.
  • “e. Accomplish their prescribed general war missions in the event of an all-out [less than 1 line of source text not declassified] Soviet attack.
  • “f. Maintain control of required land and sea bases, sea areas, and essential air and ground lines of communications with non-nuclear weapons, if possible, but be prepared to use nuclear weapons, if necessary to accomplish this objective.
  • “g. Conduct counterinsurgency actions.”4

B. Comparative Ground Force Strength

The key issue in designing a non-nuclear posture is the size and strength of the enemy. For many years Soviet and ChiCom forces have been greatly overestimated in western military thinking. This has led to an overestimation of the forces required to stop them and, consequently, to an air of hopelessness about our prospects of winning in a major non-nuclear war. As the discussion below shows, however, our prospects for successful non-nuclear defense along the Sino-Soviet periphery are better than is commonly supposed. The problems we face are related more to readiness, deployment capability and tactical mal-deployment than they are to matters of manpower and equipment resources.

NIE 11-4-635 estimates that the Soviets get about 145 divisions and cadre divisions out of their theatre ground force of 1,950,000 men, and that about 80 of these units, having 80 per cent of their manpower, can be called combat ready.6 This number of divisions can be maintained within the estimated total manpower only if the active Soviet division “slice” is very small. If all 145 divisions were manned equally, the slice would be 13,500; if only the 80 operational divisions were manned, their slice would be 25,000. Obviously, reduced strength and cadre divisions use some men, so the actual slice must be less than 25,000. In addition, the Soviet Army surely has manpower requirements not directly related to its divisions, and must fill them at the expense of the slice.

Further, despite the NIE’s estimate of 145 divisions and 1.95 million men in the Soviet theatre ground forces, there have been indications that the Soviets may be maintaining smaller numbers. Recently, a joint CIA/DIA group concluded that the number of divisions and cadre divisions may be smaller than had been estimated. The group’s view is that the total is between 115-135 such units, and that the number of combat-ready divisions is in the range of 58-75, in contrast to the NIE’s 80. So far as manpower goes, the group’s view is that a high degree of confidence [Page 569] can be placed on a range of 1.8 to 2.1 million men in the total Soviet ground force.7 Computations on this basis indicate that the slice would be 15,000 if available manpower were spread evenly over 125 divisions. On the more likely premise that support forces are furnished in proportion to the division’s combat-readiness category, the slice for combat-ready units is about 23,000.8 Since the value of 1.95 million men represents total ground force strength in the CIA/DIA group’s view, as opposed to the NIE’s estimate that it reflects theatre ground force strength only, further subtractions would have to be made for activities not directly connected to division support. These could produce a slice as small or smaller than 20,500.

In contrast, if one were to divide the U.S. Army’s total manpower by its combat-ready divisions, the resulting slice would be about 60,000 men. Many of these men, however, are also in activities not directly related to support of the divisions (e.g., 25,000 in ARADCOM, 10,800 in South Vietnam, 30,000 in Intelligence and Security, 48,000 in Command and General Support, etc.). Allowing for this, the U.S. Army division slice, worldwide, is on the order of 34,600. Although division slices are not a precise measure of effectiveness, the differential between this figure and the Soviet slice of less than 23,000, indicates that a Soviet division’s support is slender or that it is to be generated on mobilization.

The Chinese Communist Ground Force Order of Battle

The Chinese Communist Army is organized into 34 Field Armies of 132 divisions, of which 117 are at operational strength. Unlike their Soviet counterparts, these are primarily infantry forces. Their support consists of separate armored, artillery, anti-aircraft, and armored cavalry units, plus engineer, service and support units. The non-divisional combat support provided is roughly equal, in numbers of men, to Soviet levels [Page 570] but administrative and logistic support forces comprise only about 25 per cent of the total forces. The latter is low even by Soviet standards, while both combat and logistic support are deficient by U.S. standards.

The Chinese Communist Army is still essentially a World War II model and even by these standards is lacking in firepower and mobility. In many cases it has not reached its own authorized levels of equipment. The main strength of the Chinese Communist Army is in manpower, an advantage which does not stand up well against a well-equipped, modern Army. Its main weakness is logistical, compounded by its having been designed and built to operate with large scale Soviet logistical and technical support. Without Soviet assistance, China may be able to supply itself with small arms and ammunition, but it could not produce the needed quantities of armor, artillery, vehicles and other heavy equipment.

Comparative Combat Effectiveness

The development of military force structures, especially in an alliance, is not an exact science. Quite apart from the variables introduced on one’s own side, there is the difficult matter of arriving at a precise estimate of the enemy’s potential. A substantial margin of uncertainty seems inescapable, and while it is obviously unsafe to underestimate the enemy, an overstatement of his capability is equally dangerous. It can lead to a strategy of desperation and to pricing necessary and valuable military capabilities out of the market. Thus, although it would be simpler to design our military posture to deal with a single objective and a single estimate of enemy capabilities, we must confront the full range of uncertainty, maintaining both a hedge against unfavorable possibilities and the ability to take advantage of favorable circumstances.

The uncertainties stem, in large part, from the intangibles inherent in ground combat. Many things other than sheer numbers and the amount and quality of equipment influence the combat effectiveness of ground forces. Tactical surprise, morale, training, the ability to keep large forces healthy in the field, generalship, terrain, the air situation, defense versus offense, and sheer chance can and will affect the outcome of any war, campaign, or individual battle. But, before any differences between the two sides in these factors can be taken into account, it is necessary to establish some basis for comparing the measurable aspects of opposing ground forces on the assumption that the non-measurable factors are equal. One obvious measure is strength in men. Another can be derived from the relative firepower which has been measured by an Index of Comparative Firepower (ICF). This technique was originally developed by the Army as a means of resolving arguments arising during maneuvers. It has been increasingly applied in studies and in wargaming of specific tactical battle situations, although it was not designed for application on a strategic scale over an entire campaign. The Secretary [Page 571] of the Army has stressed this limitation, with respect to both the ICF and a parallel index which takes into account some items beyond raw firepower, the Index of Combat Effectiveness (ICE).9

I accept these reservations, but feel that, so long as it is clearly understood that these indices do not take into account the intangibles mentioned above and are not an infallible means for predicting the outcome of an entire campaign, they are useful in determining relative pre-hostilities combat potential—certainly the best tools now available. Some such tool is clearly needed. In addition to their obvious task of fighting effectively if war comes our military forces must, in peacetime, deter aggression and provide a power base for the conduct of our foreign policy. Our civilian leaders need a means of assessing, in broad terms for each specific situation, the probable adequacy of our deterrent and the strength of the military foundation supporting our diplomatic position. Moreover, some approximate aggregate factor is needed to correct the widespread tendency to equate U.S. and Soviet divisions, despite the fact that ours contain, with their Corps support, at least twice as much manpower, as well as more expensive equipment. I am, therefore, asking the Army to continue its investigations with the aim of reaching the most accurate ICE possible.

If U.S. Mechanized and Armored divisions are compared to their Soviet counterparts, the Motorized Rifle and Tank divisions, on a basis of raw firepower, we find a differential of about 10 per cent in our favor:

[Page 572]

[Here follows a table entitled “Indices of Comparative Firepower: Divisions”. The U.S. Mechanized division advantage was 1.075/1; the Armored division advantage was 1.48/1.]

This calculation does not, however, take account of the firepower contributed by the division’s proportionate share of those non-divisional combat units normally pooled at the next higher tactical level: the U.S. Corps and Soviet Combined Arms Army. When non-divisional combat support is included, the U.S. advantage increases:

[Here follows a table entitled “Indices of Comparative Firepower: Division Fighting-Force”. The U.S. Mechanized division advantage was 1.28/1; the Armored division advantage was 1.69/1.]

However, our mechanized and armored division fighting-forces are about twice as large as their Soviet counterparts in terms of manpower, and it appears that we equip each of our men with about 30 per cent more initial equipment than do the Soviets, measured in U.S. prices. There is substantial uncertainty about comparative expenditures, but a CIA study has estimated the costs of building various Soviet items of equipment in the U.S., with the following results:

[Here follows a table showing comparing the U.S. equipment cost of $9,300 per man to the Soviet equipment cost of $7,100 per man.]

In combination, the manpower and money we invest in our divisions and their immediate combat support should produce an advantage in comparative effectiveness of at least two to one, but firepower ratios give only 1.2—1.7 to 1.10 At issue is the contribution of those men and that equipment which either do not appear in firepower calculations or whose real contribution is out of proportion to their direct firepower scores because their contribution is to mobility, target acquisition, command and control, or essential “M-day onward” staying power. The Secretary and the Chief of Staff of the Army, while considering raw firepower the most reliable available basis for comparison, believe that we do gain a real advantage from these “add-ons.” Their evaluation is:

[Here follows a table entitled “U.S. Organizational and Firepower ‘Add-ons’” that shows the evaluation of the Secretary and the Chief of Staff, U.S. Army.]

In summary, the Secretary and Chief of Staff of the Army conclude that on comparing a U.S. Corps and a Soviet Combined Arms Army we find ourselves stronger, more mobile and able to outshoot the Soviets. The U.S. Corps gets superior reconnaissance and flank protection from its armored cavalry units, better control from its communications units, [Page 573] and from its advantage in engineers, the ability to facilitate its own movements while impeding the enemy’s. With light aviation units organic at lower levels, it benefits in mobility, control, and battlefield surveillance, and its logistics come from a system proven repeatedly in war. In short, the U.S. organization provides its commander a rounded capability for combat, which I believe is at least 1.75 to 2.3 times that of comparable Soviet divisions. The Secretary of the Army and the Chief of Staff feel, however, that at this stage of our investigation of this matter, we should equate one U.S. division to 1.2 to 1.7 Soviet divisions,11 depending of course on the specific tactical situation under consideration. Therefore, I have used these latter figures hereafter, although quite evidently there is much room for further study, and I am directing that the Army pursue this matter.

C. The Special Situation in NATO

At the outset of any discussion of United States strategy in NATO-Europe, it should be clearly recognized that the United States is not a free agent. United States strategy and NATO strategy are inextricably linked and our ability to influence the NATO members is limited. NATO strategies which may be militarily and economically feasible may be politically infeasible or feasible only after extensive psychological preparation.

If Europe is vulnerable to a non-nuclear attack today, as it is, it is not the result of overwhelming Soviet strength, nor failure of the Europeans to devote men and a great deal of money to defense. The immediate problem is much more one of balance and inefficiency in the allocation of resources. In my remarks to the NATO Council last December,12 I pointed out that for Europe to meet virtually its [less than 1 line of source text not declassified] goals would take about 1 per cent more of the growing aggregate European GNP and that this, along with our own contribution, would provide Europe with a solid non-nuclear defense. In the light of our recent studies of NATO and Warsaw Pact tactical air and ground strength it appears that our position was entirely warranted and, in fact, that it should be possible to get the most important gains with increases of less than this amount. In any case, U.S. forces will not by themselves provide an adequate non-nuclear force for the defense of Europe. Our leadership can help, but the basic remedies lie with the other members of NATO.

[Here follows a detailed exposition of NATO and Warsaw Pact force dispositions.]

[Page 574]

Despite NATO’s relatively encouraging position in manpower and units, its forces could not assuredly hold the Soviets at a forward line. Mal-deployments, frictions, and misunderstandings within NATO forces could give the Soviets an opportunity to break through in a decisive way. If one or more of the larger countries in the center were to drop out of the conflict NATO’s position would be untenable. However, the Soviets also have problems—apart from their lively concern that NATO might employ nuclear weapons—in that the reliability of their allies is questionable, especially if the situation is such that the Soviets are not clearly in the ascendancy.

Given anything like the present Russian strength, and our tactical air strength, well-equipped and stocked NATO forces [less than 1 line of source text not declassified] to include first echelon forces, should be able at least to delay to and defend on the line of the Rhine against a Bloc surprise non-nuclear attack without using nuclear weapons. There would remain the possibility of a sudden Soviet nuclear attack catching us by surprise, and there is also a possibility that the Alliance might not hold together. But, for a wide range of circumstances NATO forces should do well, provided they had reached their [less than 1 line of source text not declassified] goals.

Within the past year there has been a net increase in non-U.S. NATO defense budgets of about $1.3 billion (from $18.7 billion to $20.0 billion). About one-half of the increase has come in Germany, and has provided an increment of real power. The “hard core” high priority budget increase that I proposed last December as meeting the most urgent needs of the non-U.S. members of NATO, amounts to $3 billion a year. The prospects for getting this much of an increase in European defense budgets are remote. However, having in mind the existing balance of forces described above, it is becoming increasingly obvious that the problem is not so much getting the Europeans to spend a great deal more, but rather getting them to fill gaps in their efforts and to get a better balance of effort.

Most of our non-nuclear requirements studies look at the problem of defending Europe against a massive attack. I believe it at least equally important that NATO have strong conventional forces for use in contingencies arising over Berlin, or in connection with other contingencies whose course is hard to predict. In either, our ability to put pressure on the Soviets—a crucial element in crises of this sort—may depend on our ability to make limited military moves without using nuclear weapons.

Neither our opponents nor our Allies have their forces at the high readiness of U.S. forces in-place in Europe. France has only two divisions in Germany. [less than 1 line of source text not declassified]. In addition to having to move all of its forces forward in a crisis, the French Army is to be at 450,000 man strength in peacetime, with plans for expansion in a [Page 575] crisis, to a million men. The Germans say that they plan by end-1964 to go from a peacetime force of 260,000 men to over 700,000 within 72 hours. Britain also plans to fill out its three division requirements by airlifting two brigades and a division headquarters from the U.K. in an emergency—a move that would take at least seven days.

If, however, the conventional forces provided [less than 1 line of source text not declassified] are not capable of stopping a surprise non-nuclear Soviet attack reasonably well forward, we are faced with the alternative of resorting to tactical nuclear weapons or increasin. [less than 1 line of source text not declassified] force levels.

Soviet vs. Western Rapid Mobilization Capabilities: Mobilization vs. Instant Readiness

Both the Soviet Union and our Allies rely on rapid mobilization and deployment, and can fill understrength units rapidly. Nevertheless, these mobilization actions cannot be accomplished instantaneously nor, if done on a massive13 scale, in secrecy. Further, once existing, equipped units are filled out it will take several months to create additional combat-ready units, even if stocks of equipment exist.

This practice is at variance with both that of the United States and official NATO doctrine. The U.S., in accordance with agreed NATO requirements, keeps its NATO committed ground forces, including logistic support units, at high readiness. The FRG and the Soviet Union distribute their active duty personnel as follows:

Percent Active Duty Personnel FRG USSR
Divisions (U.S. at 98%) 79-96 30-85
Non-Div. Combat Support 81 40-85
Logistic & Service Support Units 14 0-40

These German and Soviet forces clearly do not have “instant readiness” by U.S. standards, but their practices have some merit. The Germans in particular man their combat units at a relatively high level, counting on reservists to perform less demanding rear activities. They claim to be able to mobilize the necessary reservists in 72 hours, and one estimate is that their mobilized units would reach 70 per cent effectiveness in 10-14 days. For many rear area functions this German estimate may be correct; although it should be noted that German planning has not yet been rigorously tested, and not all the necessary reserve units are organized and equipped. Somewhat similarly, the British plan on reinforcing the BAOR within a period of seven days from the alert, and the French would have to move forces into Germany as well as calling up [Page 576] reservists. CINCEUR may, therefore, be optimistic in estimating that all NATO first echelon reserve units with a peacetime manning level of 65 per cent would be brought to full strength and deployed in 20 days, and that second echelon units at 5 per cent strength could be brought to full strength and deployed in 30-35 days. Certainly, at that stage, these reserve units would be less effective than those active on M-Day.

The Soviets apparently also plan to fill out units rapidly. Although they intermittently call up former active duty people for training, they do not have a systematic program for maintaining their reserve personnel at a high state of training. Therefore, at a minimum, it would probably take the Soviets, as well as our Allies, several weeks to get non-operational ground force units in a state approaching combat readiness and deploy them. For this reason, the JCS estimate that although the Soviets could fill out their total divisions (115-135 in DIA/CIA panel’s view) rapidly, only about 20 Soviet divisions, beyond those now combat-ready (58-75 in DIA/CIA panel’s view) would become operational by M+30.

This disparity between the high readiness of our ground forces and those of our opponents and Allies raises the question: “Are we allocating our manpower resources efficiently?” Although the Communists might launch a sudden attack in any of a number of areas, our principal limiting factor in meeting such a contingency at this time is our deployment capability, not the readiness of our entire Army structure, including its logistic system. Further, in Europe 100 percent manned U.S. units have on their flank non-ready, under-manned Allied units.

D. U.S. Force Requirements for Various Conflicts

Thus far, emphasis has been on NATO. It is necessary also to look at other likely areas and contingencies. A number of studies of possible non-nuclear conflicts in various areas of the world were done last year by the Chairman’s Special Studies Group in the Joint Staff. These studies, which are being brought up-to-date in the light of new intelligence and other factors, cover a wide range of situations, showing the requirements for ground forces in relation to those now available. United States and Allied forces worldwide are summarized in the following table. Communist forces are summarized on the following page. The table after these summarizes U.S. divisional requirements for nine potential conflicts on the Sino-Soviet periphery. (Fatalities in Millions)

[Here follow 3 pages of tables.]

Middle East

In the Middle East Iran is the area most vulnerable to Soviet attack. It is also especially difficult for us to move forces to this country rapidly and to support them there. The Russians maintain 24 divisions, of which [Page 577] 12 are operational, and about 1,000 combat aircraft in the military districts bordering Iran.

Taking account of the severe communications and logistics limitations in the rugged terrain of northern Iran, the net initial Soviet threat is estimated to be about 10 divisions. The bulk of the Soviet aircraft are short-range day fighters unable to reach very far south into Iran.

This attack would present us with a formidable problem. With our programmed airlift capacity for 1968 we could build up our forces as depicted below:

No. of U.S. Divisions in Iran Days After D-Day
10 20 30 40 50 65
Airlift/Sealift only:
Suez Open .5 1.0 2.0 3.0 4.0 5.0
Suez Closed .5 1.0 2.0 2.0 4.0 5.0
Airlift/Sealift plus Forward Pre-Positioning 1.0 2.0 2.0 3.0 4.0 5.0

With its 10 divisions and strong air support the Soviet Union can dominate the situation in northern Iran. Assuming that Iranian forces hold together and withdraw south to the Zagros Mountains, and also assuming that the Russians do not take key centers in the south, including Abadan, by air assault, the U.S. might be able to deploy enough forces (8 divisions and about 1,600 aircraft) to stop the Soviets at the Zagros Mountains, or even mount a counterattack, although this would require an additional six divisions.

This situation obviously is touch and go. Our response time in this case is critically important. For this reason, the Middle East-Indian Ocean area is the one in which pre-positioning of stocks has a great payoff. For example, if we had 30,000 tons of stocks in a depot in the area (possibly a forward floating depot) the time to get two divisions into Southern Iran in 1968 (one airborne, one infantry) would be cut from 40 days to 20 days. The 40 day figure takes into account that heavy equipment which must be moved by sea. This is the equipment which would be pre-positioned. The difference of 20 days could be decisive in certain circumstances. We are investigating this possibility and analyzing its costs.

Another point to be noted is that our ability to hold Southern Iran might depend on moving a relatively small force, along with some air defenses (interceptors or Hawks), very quickly to key ports and airbases in Southern Iran, in order to prevent their seizure by Soviet air attack.

In short, treating this as a single-situation contingency, our main initial limitation is deployment capability, not forces. Programmed additions to our airlift capacity will help, and so would pre-positioned equipment in the area. I will make recommendations to you when the results are in from our investigation of pre-positioning.

[Page 578]

Southeast Asia

The General Purpose Forces Study of the defense of South Vietnam, Thailand, and the southern part of Laos against attack by Communist China and North Vietnam concluded that the U.S. and its SEATO Allies could successfully defend the area against a 21-division Communist attack. The 13 Thai and South Vietnam divisions in the area (used mostly for counter-guerrilla and rear area security tasks), plus five U.S. divisions and one Commonwealth Division, rapidly introduced into the area, along with allied air superiority, could halt the enemy along the general line of the 15th parallel and thence north along the Mekong River. To restore the situation would take four more U.S. divisions, or a total of nine.

In that region we have a major logistic advantage. For example, with air superiority on our part the existing road and rail net combined permits the tonnages shown in the table below.

Comparative Logistic Capability (Tons per Day) U.S. and Allied Forces Communist Forces
Dry Season Wet Season All Season and
Area Road/Rail Air Road/Rail Air All Sources
Northeast Thailand (near Vientiane) 5,000 1,000 4,000 500 Less than 1,000
East Thailand 2,000 —— 1,000 —— Less than 1,000
Totals 8,000 5,500

A Chinese Communist or North Vietnamese division consumes about 75-150 tons per day in conditions of average combat; a U.S. infantry division, approximately 371 tons. The transportation net available to the Communists can support about 7 to 12 of their divisions at their average combat rates. Most of the 21 division Communist force therefore would be capable only intermittently of a high level of activity.

Although the Chinese Communist and North Vietnamese can do a lot with a little, these logistic differentials of at least 8 to 1 and 5.5 to 1 place them at a severe disadvantage—one they must face for years to come (despite Chinese road building endeavors in Northern Laos).

Northeast Asia

This study examined a resumption of the war in Korea under several different assumptions, the most pessimistic of which had the Chinese Communists sending troops covertly into North Korea before a coordinated North Korean and Chinese attack: 36 divisions on D-day and 63 divisions by D-plus 30 days. The Special Studies Group estimated that seven more U.S. divisions in addition to the two in place, or a total of nine, along with the ROK Army, would be required to absorb the blow and regroup on the Han River line. With eventual U.S. air superiority and five additional U.S. divisions, these forces could restore the present position.

[Page 579]

This requirement is fairly sensitive to assumptions about the timing of enemy moves and about the size of his force committed. The following table shows three variants of this situation: (1) Low in which only the North Koreans attack; (2) Medium in which the Chinese join after the outbreak but commit forces only on the scale of the Korean War because of concern about trouble elsewhere should they get over-committed; and (3) High in which the ChiComs commit all their divisions in the Mukden area plus further divisions from their strategic reserve (this is the case referred to above, the most severe looked at by the Special Studies Group).

Chinese Communist and North Korean Ground Strength at D+30 Days Higha Mediumb Low
Divisions in Korea
ChiCom’s 43 27
North Korean 20 18 18
Total 63 45 18
Ground Force Totals
ChiCom 801,000 509,000
North Korean 198,000 178,000 178,000
Total 999,000 687,000 178,000
ROK and U.S. Ground Strength Divisions in Korea At Present Required to Hold High Attack Required toRestore High Attack Situationc
U.S. 2 9 14
ROK 18 20 20
Total 20 29 34
Ground Force Totals
U.S. 56,000 360,000 550,000
ROK 536,000 600,000 600,000
Total 592,000 960,000 1,150,000


a ChiCom’s (a) do not feel compelled to keep major forces opposite Taiwan and the Soviet Far East, and can, therefore, employ all operational divisions now in the Mukden area and commit further divisions from their strategic reserve; (b) possess enough equipment to meet their combat support force needs; and (c) have adequate supplies for support of both own and North Korean forces.

b ChiCom’s (a) commit the bulk of the operational forces from the Mukden area, but feel compelled to hold back their strategic reserve. This is the commitment of forces the JCS estimate would be available to support the North Koreans in view of Chinese requirements to maintain deployments opposite Taiwan, in Tibet, and elsewhere.

In the Medium case, 10 Chinese divisions (approximately 150,000 men) move to the 38th parallel in ten days, and an additional 17 divisions move within 17 days. This force, along with 18 North Korean divisions, in terms of combat effectiveness is over 50 per cent stronger than the presently deployed U.S. and ROK forces. However, an Army study of [Page 580] this situation indicates that with a fortified zone developed in advance, (estimated cost of system approximately $50 million) present forces in Korea would defeat the attack. In effect, this is approximately the scale of the Korean War.14 At their peak, the Chinese Communists and North Koreans had 33 divisions engaged on the line while the U.S. had eight and the ROK’s had 8-10.

E. Tactical Nuclear Posture and Strategy

Though we have conducted several studies of tactical nuclear warfare on the ground in Europe, we have not yet been able to clarify fully the role of tactical nuclear weapons in our over-all strategy. For example, we have yet to resolve important questions about the vulnerability of ground forces to nuclear weapons and the feasibility of sustained operations in a nuclear environment. In general, our understanding of this kind of warfare is far less advanced than our apparent understanding of most other kinds of warfare. So far our studies suggest the following conclusions:

There are reasons for believing that the escalation potential of tactical nuclear warfare, of the type on which present plans are based, is high, particularly in terms of increasing yields. We are looking for ways to reduce this escalation potential.

[2 paragraphs (13 lines of source text) not declassified]

Ultimate Soviet attainment of approximate parity in tactical nuclear weapons may reduce and possibly eliminate whatever military advantages the US or NATO would obtain by resorting to tactical nuclear weapons in fighting the USSR. Though this aspect has not yet been fully explored, there is little evidence to suggest that the use of tactical nuclear weapons under such conditions will result in anything better than a military stalemate between the engaged forces.

Nuclear conflict can be divided into:

Strategic Nuclear warfare, global in scale.
Theater Nuclear warfare; probably as damaging within the theater as global war, and probably extremely difficult to limit to the theater.
Battlefield warfare, with nuclear strikes limited to the forward battle area plus interdiction of lines of communications outside the US and USSR.
Battlefield warfare, limited to just the forward battle area.

[Page 581]

5. Demonstrative nuclear strikes, intended to put pressure on the opponent by threatening escalation.

Although ground forces must be prepared to participate in any of these types of nuclear conflict, the design of our ground force nuclear systems should be oriented primarily, but not exclusively, toward the last three of these cases.


Tactical nuclear warfare presents large areas of uncertainty which have not been resolved. We have had no experience in it and cannot be sure we have the right mix and balance of weapons. The main uncertainties are these:

How Might the War Start?

Troops in barracks areas are extremely vulnerable to nuclear attack without warning. On the other hand, with even a little warning, troops can rapidly disperse to field locations, in which case far more nuclear strikes would be needed to inflict the same casualties. If nuclear weapons are employed from the outset, installed atomic demolitions can be fired as the enemy approaches; if they are not, uncertainties about timing complicate their installation and employment. Similar problems plague other planned nuclear fires and troop deployments. These problems of timing stem primarily from delays inherent in NATO’s decision making machinery, which, of course, is subject to a multitude of political imponderables.


Estimates of civilian casualties in Europe range from several hundred thousand, for a restricted zone battlefield campaign, to from five to fifty million or more for a theater-wide tactical nuclear war. Whether constraints against escalation can prevail over the temptation to employ larger weapons to destroy the enemy’s nuclear delivery means before they are used is a large question. For example, the vulnerability of our nuclear strike aircraft in Europe creates a major instability. This would be a source of major concern to commanders attempting to conduct a limited battlefield nuclear war.

Soviet Weapons and Concepts?

As far as we know, Soviet planning, like NATO planning, has emphasized theater-wide nuclear attack. The Soviet nuclear stockpile is estimated to be much smaller than ours and, perhaps largely for this reason, Soviet programs have not emphasized short range small nuclear warheads as we have. They, of necessity, would be forced to use longer range, higher yield systems. Soviet classified military writings reflect their relative scarcity of warheads, the absence of a concept to limit nuclear engagement to forward areas or to a sector of the battlefield, and an interest in some quarters in employing a small number of high yield weapons to destroy the opposing army. These negative Soviet attitudes [Page 582] toward the possibility of a forward zone, small yield nuclear conflict may change as their stockpile of nuclear weapons increases, but we cannot predict with any confidence when, or if, this change in doctrine will take place. Moreover, a “high-yield-low-cost” faction may come to dominate Soviet thinking on ground warfare.

In this regard a recent study has assessed the impact of a Soviet blanket barrage of 32 20MT weapons along the general area occupied by NATO troops in Central Europe, assuming that the NATO forces are preparing and occupying defensive positions whose exact locations are unknown.15 The NATO force of over 670,000 men in 30 divisions and associated units sustains losses of seven per cent of men in trenches, 33 per cent of troops in buildings and 43 per cent of troops in the open. It also loses 11 per cent of its tanks and artillery and 35 per cent of its vehicles and electrical equipment. If it is assumed that the troops were disposed equally in trenches, buildings and in the open (the study hazards no such guess), then 28 per cent of the entire force become casualties and 15 of the 19 divisions affected by the blast suffer from 40-65 per cent casualties. 1.8 million civilians also become casualties in this attack, even though it avoids large cities.

The Nuclear Battlefields

[1 paragraph (11 lines of source text) not declassified]

Psychological Reactions

A tactical nuclear war would be an entirely new experience for all involved, troops, civilians, commanders, and statesmen, on both sides. It is extremely difficult to predict how these various groups would behave when subjected to such a novel and unpleasant experience as an intensive tactical nuclear engagement. Our studies suggest that in the first several days casualties in engaged combat forces might be as much as 50 per cent. A large number of units would receive 30 per cent or more casualties. Many survivors would have gotten radiation dosages, whose level would be quite difficult to determine. The cumulative effect would be intense, on troops, on intermingled civilians and on decision-makers alike, as they sought to correlate fragmentary reports and take appropriate actions, both military and civil. There would be major problems at all levels in sustaining a rational, controlled effort.

Results of our Studies

These examples from recent studies illustrate the general nature and quantitative impact of some of the principal factors involved in determining how many tactical nuclear weapons are needed for Central Europe.

[Page 583]

Many qualifications must be attached to any conclusions based on these studies. In addition to those pointed out earlier, the figures are highly sensitive to the specific details of the war situations assumed. In particular, the numbers and types of weapons expended reflect the weapons allocations assumed for both sides and apply only to the first day or two of nuclear warfare.

In October 1962 the JCS published the following results of four manual map exercises conducted to develop the first day or two of nuclear exchanges between a 3-1/2 division NATO corps defending 100 kilometers of front and a seven division Soviet combined arms Army in 1967:

Situation NATO Employs Soviet Employs Results

[Table entries (4 rows) not declassified]

Taking a different, more theoretical approach16 to over-all NATO ground warfare tactical nuclear requirements, the October 1962 JCS study also calculated the dynamic effects of the progressive attrition of opposing nuclear weapons systems. Inputs to the model included force relationship, the size of Soviet nuclear stockpile, relative rates of fire, and random fire weapons effects. It revealed that in order to reduce likely Soviet 1967 forces by 50 per cent and exhaust Soviet tactical nuclear weapons, while sustaining no more than 30 per cent losses to NATO forces, NATO Central Region tactical nuclear requirements, depending on the selected mix of weapons by yield, would be as shown below. Note that the number of weapons required is sensitive both to the number available to the Soviets and to the sizes of weapons we choose to use, particularly the latter.

[Table (3 columns, 3 rows) not declassified]

Note: This table is subject to the qualifications which follow:

Though the method employed does take into account NATO weapon losses due to enemy ground action, it does not include a number of other factors which quantitatively affect total NATO stockpile requirements, such as: (1) organizational losses due to enemy air and missile forces; (2) weapon system reliabilities; (3) maldeployment of delivery systems or nuclear weapons; (4) restrictions on weapon deliveries due to constraints; (5) improvement from firing at close-in acquired targets rather than using the random fire involved in the mathematical model. The first three factors create extra requirements; the small percentage of [Page 584] high yield weapons in the mixes considered make it unlikely that constraint restrictions would significantly modify requirements; but the fifth factor promises substantial reductions. JCS studies involving somewhat different troop densities, weapon availabilities, and rates of fire from those discussed above indicate that through the use of acquired target fires, the last alternative above might be modified as follows:

[Table (3 columns, 2 rows) not declassified]

In April 1963, the JCS published the results of further investigations into the effects of varying the mix of tactical nuclear weapons available to the NATO Corps defending the 100 Km of front described above. [2 lines of source text not declassified]

[Table (3 columns, 4 rows) not declassified]

[1 paragraph (10 lines of source text) not declassified]

Allied Views

[1 paragraph (20 lines of source text) not declassified]17

Summary on Tactical Nuclear Weapons

On the basis of the work done so far we have much to learn concerning tactical nuclear warfare, both from the point of view of its military consequences, including the central question, “To whose advantage would it be?” and from the point of view of the political environment in which it would occur. It is clear, however, that we do not have an adequate basis at this time for placing greater dependence on a tactical nuclear strategy of the type described by the British and Germans. We are providing an option for this strategy in the planned buildup of tactical nuclear weapons, and in the event of a sudden Soviet attack in Europe, we may be forced to rely upon it. Against this contingency a serious effort is being made by the JCS to develop plans for limited zone, low yield operations involving the engaged battle zone and its immediate interdiction.

[Here follows a full-page table.]

F. Conclusions on Contingencies, Forces and Deployments

Although it is a major objective of our General Purpose Forces programs to open up as many useful military options as possible, we must decide on the relative emphasis to be put on preparations for non-nuclear versus nuclear conflict. In the nuclear category, resources must be further allocated among programs for short range battlefield systems, longer range interdiction systems, and still longer range theatre systems whose functions are virtually the same as our strategic retaliatory forces. [Page 585] Within both the nuclear and non-nuclear categories we must decide on how much mobility to have, and on the scale and duration of possible conflicts. We must also make decisions concerning the number of conflicts we should be prepared to handle simultaneously and the performance to expect from Allies, and must make assumptions on the circumstances in which war may break out.

Several things are clear in this regard. First, our ability to predict contingencies is poor. Second, while we can and should take out some insurance against low probability contingencies—especially if they are dangerous ones—in the design of our General Purpose Forces we must give great weight to the more probable ones. Third, being able to handle the biggest or the most dangerous contingencies does not necessarily mean that we can manage others equally well. We have learned this in counterinsurgency, and we are learning the greatly differing demands of nuclear and non-nuclear conflict. Fourth, there may or may not be any wars in the 1965-1969 period above the counterinsurgency scale, but there will surely be situations in which our ability to use measured force will be relevant—if not critical—as in Cuba, Berlin, and Thailand.

It would be a mistake to concentrate on any one or a few of these peacetime or wartime objectives and neglect the others. We must give careful consideration to the full range of enemy capabilities. Surprise attack is one of these, against which we must always be alert and prepared. We were surprised at Pearl Harbor and in Korea, and so long as there is an enemy capable of launching a surprise attack we may be surprised again. Logically, surprise attack is less likely where U.S. forces are deployed and, historically, the best deterrent and defense against surprise attack has been forward deployments. The presence of U.S. forces insures that any attack will immediately involve U.S. power, decreasing the enemy’s chance of achieving a decision before U.S. and Allied reinforcements can intervene. Where there are no U.S. forces the temptation is greater and the likelihood of surprise attack is increased. In such areas emphasis must be on early detection of the long build-up required for a major attack—with due consideration to the lag between perception and action—and our general purpose forces must be able to move quickly into the area, applying increasing pressure both to reassure our Allies and to force the enemy to back-off. In Europe, in particular, for both military and political reasons, we should continue to have large forces on the ground.

In the event of surprise attack against areas in which we have ground forces, Europe and Korea, we may find it necessary to use nuclear weapons. Present U.S. and Allied capabilities, however, are strong enough to keep this issue from being forced on us early. In any event, this decision should not be regarded as automatic. And, if nuclear weapons are used, we should try to limit their use to the battlefield, initially [Page 586] to force the opponent to back down or failing that to hold him. Depending upon the method of allocating “dual capable” costs, our battlefield nuclear programs take less than 10 percent of the ground forces budget. This is not too high a price for this insurance.

As to the scale and duration of non-nuclear contingencies for which we should be prepared, at present our largest scale contingency plan involves no more than 10 divisions (five deployed from CONUS to Yugoslavia or the European satellites plus whatever portion of CINCEUR’s five divisions he may make available). In the General Purpose Forces Study, the JCS have examined a 12 Army division single contingency (Iran) and a 14 Army division two-conflict situation (Europe and Korea). This does not mean there is a guarantee against larger conflicts, but simply that we have been unable to foresee them.

It is easy to be very wrong on contingencies. But, it is clear that the U.S. should not, for example, plan on trying to solve the problem of non-nuclear defense of Europe by itself. Despite our assessment that Soviet conventional strength is lower than has been thought, and improvement in German forces, NATO will not be able to fight a large non-nuclear war on a sustained basis during at least the next several years. The conditions for a sustained non-nuclear defense of Europe are now in sight, however, and our planning and negotiations should begin to exploit this forthcoming alternative to nuclear war.

This assessment of contingencies and their implications for the design of our forces can be summarized as follows:

We will continue to be faced with counterinsurgency conflicts and should continue with our counterinsurgency programs.
Crisis deployments and show-of-force operations short of fighting will be necessary to support our foreign policy. This clearly means a continuation of large forces in Europe.
We can expect the more likely conflicts to arise from local confrontations which will call for deployments from the U.S. of forces up to several divisions. We should be capable of rapid reinforcements for such situations.
So long as surprise attack remains an enemy capability we must be alert and prepared to counter it. We should not make preparations for this type of conflict a major design objective of our General Purpose Ground Forces although some insurance programs are appropriate in order to keep the probability of this contingency low, and in order to meet it if it occurs.
The programmed force will enable us to meet this requirement for single-region, non-nuclear conflicts in all but a few cases. In Europe, this would not be so in the worst case, given current non-U.S. force deficiencies. In the medium and low cases, non-U.S. forces would run out of stocks after about four weeks. In each of the two simultaneous conflict [Page 587] situations, and in the Middle East in particular, our main current deficiency is more in deployment than in forces in being.
Our present battlefield nuclear programs are not especially costly and they help to reduce the probability of local nuclear wars and to provide a way of meeting it if it occurs. However, we must recognize that such a war with the Soviet Union would stand a good chance of leading either to a collapse of Alliance support or to a global war.

  1. Source: National Archives and Records Administration, RG 218, JCS Records, JMF 7000 (3 Jan 64) Sec 1A. Top Secret. This draft memorandum is Tab III of “Department of Defense Draft Memoranda for the President: Recommended FY 1965-1969 Defense Programs.” dated December 19. An early version of the draft memorandum, which was circulated to the JCS for comment, is dated October 10. (Ibid., JCS 1800/775, JMF 7000 (10 Oct 63) Sec 1) A later draft dated November 15, not identical to that printed here, is in Department of State, Central Files, DEF 1 US.
  2. I have disqualified myself from participating in this decision because my two minor children each own 300 shares of Ford stock. [Footnote in the source text.]
  3. Entitled “Basis for Particular Force Structure Recommendations,” not printed.
  4. Includes only the cost of the forces listed in Table I.
  5. The JCS have expressed divergent views in commenting on an earlier draft of this section. The Chairman, JCS, the CNO, and the Commandant, USMC, would prefer a substitute section based on JSOP, analyses of studies, and the contingency plans of overseas commanders. Their proposal is attached as Appendix III. The Chief of Staff of the Air Force recommended that the earlier draft be forwarded with certain changes, most of whose essentials have been incorporated in this version. The Secretary and the Chief of Staff, Army, supplied a completely rewritten section which has formed the basis for this draft. [Footnote in the source text. The “earlier draft” referred to is that of October 10, cited in the source note. Appendix III is not printed.]
  6. The source of the quotation is not identified.
  7. Entitled “Soviet Military Capabilities and Policies, 1962-1967,” dated March 22. (Johnson Library, National Security File, National Intelligence Estimates, 11-63)
  8. U.S. Divisions are not considered fully combat-ready unless 90 per cent manned. Our divisions in Germany are programmed for 98 per cent manning in FY 1964. [Footnote in the source text.]
  9. This study has not yet been addressed by the JCS and is still subject to confirmation by the USIB. The full range of uncertainty is estimated by the group to be 1.5 to 2.4 million men in the total Soviet ground forces. [Footnote in the source text.]
  10. The CIA/DIA panel’s highest confidence estimate is that Soviet divisions and division cadres are divided into readiness categories, thus:

    Combat (Near full strength) Reduced (70%) Cadre (Off, NCO, Few Others)
    65 25 25-45

    Assuming an average authorized strength of 10,000, the men actually assigned to divisions would be:

    Combat Reduced Cadre (Assuming 10%/35 Divs) Total
    650,000 175,000 35,000 860,000

    Assuming further that the combat-ready divisions will share in total manpower in the same proportion, two possible slices can be computed. The first uses NIE 11-4-63’s theatre ground forces total of 1.95 million; the second uses the panel’s average high confidence total of 1.95 million in total ground force, less 200,000 for air defense forces. This latter subtraction is only one of several which should be made, but it is typical, and the largest of NIE 11-4-63’s non-theatre ground force drains on the Soviet Army. (1. 1,950,000x.765B65=23,000. 2. 1,750,000x.765B65=20,500.) [Footnote in the source text.]

  11. The Chairman, JCS, also has reservations: “the ‘Index of Combat Effectiveness’ has no practical validity and would fall under sharp criticism if presented for serious consideration in international discussions. While we in the military service have sometimes used coefficients of effectiveness in maneuvers to determine and control combat outcomes, no one seriously believes that the outcome of battle is calculable in mathematical terms through their use. If a mathematical solution to combat outcomes were possible, it would have to resolve an equation of many variables (such as weather, terrain, proximity of the enemy, time of day, status of supply and equipment, physical condition of troops, their courage, morale, leadership), which are all constantly changing. The most important factors, courage, morale and leadership, are not subject to physical measurement. If as most soldiers believe, ‘in war the morale is to the physical as three is to one,’ only about a fourth of the determinants of victory are susceptible to the coefficient approach and they are variables undergoing constant change . . . . There is insufficient recognition of some of the critical aspects of the defense of NATO such as the shallowness of the combat theatre and the decisive advantage accruing to an aggressor in concentrating his forces swiftly for the attack of a critical sector.. [Footnote in the source text. The quotation is from CM- 1009-63, a memorandum from Taylor to McNamara dated November 12, 1963. (National Archives and Records Administration, RG 218, JCS Records, JCS 1800/775-2; JMF 7000 (10 Oct 63) Sec 3) The sentences immediately following the quoted material read: “If these considerations are taken into account, it is difficult to conclude that for the present a successful defense can be effected by NATO forces without the early use of tactical nuclear weapons. I believe this to be the unanimous judgment of the military leaders of NATO, US and European, and I would be loath to have the President receive a different impression.”]
  12. The manpower ratio is approximately two to one, the equipment cost ratio, 1.3 to 1, producing an over-all cost ratio of 2.6 to 1. This ratio should be further increased to reflect the higher percentage of professionals and the higher level of experience and training in our Army. [Footnote in the source text.]
  13. A fuller discussion can be found in Appendix II. [Footnote in the source text. Appendix II, “Comparative Combat Effectiveness,” is not printed.]
  14. See Document 120.
  15. In the sense of exploiting the full capacity of the Soviet mobilization base. [Footnote in the source text.]
  16. c “Required” divisions from report of the Chairman’s Special Study Group includes one Marine Division-Wing Team.
  17. Historical research indicates the Korean Peninsula limits the number of combat forces at the front to something less than 400,000. The Communist Spring Offensive of April-May 1951, for example, involved 30 ChiCom and North Korean Divisions, with about 375,000 men at the front. It was stopped north of Seoul by some 360,000 U.N. forces in terms of combat troops at the front. The U.N. counter-offensive which followed involved 360,000 combat troops versus about 322,000 Communist combat troops—Communist casualties having been about 70,000. This offensive regained the line of the 38th parallel. [Footnote in the source text.]
  18. Data from “Development of Basic Arms Control Research Data for Measuring Collateral Damage from Nuclear Weapons Effects,” dated 30 Nov 1962, by Stanford Research Institute. [Footnote in the source text.]
  19. In commenting on this report, the JCS have declared this theoretical approach to be useful in gaining insights into requirements, but to be invalid for determining specific numbers because the relation of the area fire model to practical delivery techniques and concepts is not realistic and because the delivery of other interacting weapons was not considered. [Footnote in the source text.]
  20. This concept is not, so far as can be determined, a fully coordinated government position. It should not be considered as the FRG’s last words on the matter. [Footnote in the source text.]