Learn about the beta

189. Draft Memorandum From Secretary of Defense McNamara to President Johnson1


  • Courses of Action in Vietnam

We confront major policy decisions with respect to our course of action in Vietnam. This memorandum attempts to frame the substance of the choices and to identify some of the problems we face.

1. US strategy. The February decision to bomb North Vietnam and the July approval of Phase I deployments make sense only if they are in support of a long-run United States policy to contain Communist China. China—like Germany in 1917, like Germany in the West and Japan in the East in the late 30's, and like the USSR in 1947—looms as a major power threatening to undercut our importance and effectiveness in the world and, more remotely but more menacingly, to organize all of Asia against us. The long-run US policy is based upon an instinctive understanding in our country that the peoples and resources of Asia could be effectively mobilized against us by China or by a Chinese coalition and that the potential weight of such a coalition could throw us on the defensive and threaten our security. This understanding of a straightforward security threat is interwoven with another perception—namely, that we have our view of the way the US should be moving and of the need for the majority of the rest of the world to be moving in the same direction if we are to achieve our national objective. We would move toward economic well-being, toward open societies, and toward cooperation between nations; the role we have inherited and have chosen for ourselves for the future is to extend our influence and power to thwart ideologies that are hostile to [Page 515]these aims and to move the world, as best we can, in the direction we prefer. Our ends cannot be achieved and our leadership role cannot be played if some powerful and virulent nation—whether Germany, Japan, Russia or China—is allowed to organize their part of the world according to a philosophy contrary to ours.

There are three fronts to a long-run effort to contain China (realizing that the USSR “contains” China on the north and northwest): (a) the Japan-Korea front; (b) the India-Pakistan front; and (c) the Southeast Asia front. Decisions to make great investments today in men, money and national honor in South Vietnam makes sense only in conjunction with continuing efforts of equivalent effectiveness in the rest of Southeast Asia and on the other two principal fronts. The trends in Asia are running in both directions—for as well as against our interests; there is no reason to be unduly pessimistic about our ability over the next decade or two to fashion alliances and combinations (involving especially Japan and India) which will keep China from achieving her objectives until her zeal wanes. The job, however—even if we can shift some responsibilities to some Asian countries—will continue to require American attention, money, and, from time to time unfortunately, lives.

Any decision to continue the program of bombing North Vietnam and any decision to deploy Phase II forces—involving as they do substantial loss of American lives, risks of further escalation, and greater investment of US prestige—must be predicated on these premises as to our long-run interests in Asia.

2. Estimate of the situation in South Vietnam. The massive infusions of US ideas, aid and manpower have frustrated any Communist design to move into their conventional-warfare “Stage 3,” but the guerrilla war continues at a high pace, the economy of South Vietnam is deteriorating, and there are no convincing indications that the South Vietnamese body-politic is reviving.


Military. The large US deployments have thwarted the VC monsoon offensive that was threatening a serious military defeat for the GVN forces at the time of my July memorandum.2 These “Phase I” deployments, 160,000 of whom are already in place, have enabled the GVN forces to begin restoration of their strength, morale and tactical integrity, which is essential if they are to retain their fighting capability in the months ahead. The US forces have also substantially secured their own bases and the areas immediately adjacent to the bases. (see Map I, attached)3 US ground combat elements deployed offensively have moved rapidly—in conjunction with tactical air support and often following up B-52 strikes—to probe into VC base sanctuaries that had long [Page 516]been untouched. The US forces have also proved to be an effective quick-reaction reserve for GVN forces that has turned potential defeat in several battles into VC retreat or at least a stand-off.

But the pace of the war remains high, with VC last week attacking in regimental strength in Phu Yen province and in battalion strength in Hau Nghia province and against the Marine perimeter outside Danang. There are increased casualties on both sides. VC attacks, terrorism and sabotage show no sign of abating. Desertions from Government Forces remain high, but recruitment has improved, permitting a gradual increase in force levels. Defections from the VC are running at 1000 a month and are increasing slowly.

The present assessment is that our Phase I deployments (bringing the US total to about 210,000 to 225,000 men) when completed will achieve the limit of their military potential between December 1965 and March 1966. As can be seen on Map I, the estimate is that Phase I forces, working with the Vietnamese, will result in “control” of 40 per cent of the population and 5-10 per cent of the area; it will keep open 20-25 per cent of the critical lines of communications; it will permit 10-20 per cent disruption of the VC bases and military logistics infrastructure, and 5 per cent disruption of the VC political/security infrastructure. We will then be in a position where a few enclaves, including perhaps 5,500,000 of South Vietnam's 15,000,000 people (of which only some 1,000,000 are in the Delta) are under US or GVN “control”—and even these will of course be subject to sporadic VC mortar and suicide satchel-charge attacks of the October 27 kind.4 Security conditions within the enclaves will be suitable for reconstruction of political infrastructure, but it is probable that the enclaves will be too small and the per cent of population controlled too little to give any hope that such pacification will be rapid or that it will tend to snowball beyond the enclaves.

The capabilities and intentions of the VC and of North Vietnam are troublesome unknowns complicating our estimate of the situation at this time. There are signs that VC morale is sagging. But, even when confronted by the US force-build-up and by our program of actions against North Vietnam and with increasing casualties and weapons losses, the VC have demonstrated an ability to continue the build-up and supply of their forces in the South, both from in-country sources and by infiltration over improved roads in Laos, by sea and via Cambodia. The indications are that the VC are not reverting to a lesser level of insurgency, but are still striving to build up for the transition to conventional warfare—that [Page 517]they are continuing both to upgrade guerrilla units in the Delta and to infiltrate additional regular army elements from the North in order to restore the balance temporarily upset by US deployments.

The VC forces are now estimated at 71,300 VC “main force and logistics,” 40,000 political cadre, and 110,000 guerrillas—an increase in the estimate since early last summer of some 6,000 (9%), 10,000 (33-1/3%) and 18,200 (20%) in the three categories respectively. In addition, although as of last month there were estimated to be three regiments of North Vietnamese regular forces in the South, early this month two additional North Vietnamese regiments were confirmed and three others were estimated as possible. In total, therefore, the NVN/VC forces are now believed to number about 230,000 men, and there is reason to believe that they will continue to increase.


Political. In Saigon, the Ky “government of generals” has survived, but accomplished little more during the period since last summer. The government, while recognizing many of its problems and working willingly with the US Mission, still lacks any broad, or even developing, base of support among the various political factions. No major faction (e.g., Buddhists, Catholics) is in open opposition, and all appear to be continuing their “wait and see” attitude. This lack of support probably will not be important over the short-run, so long as the generals in power remain united; over the longer term, however, some base of political support must be developed if the non-Communist elements in SVN are to become prepared to develop a viable society as security in the countryside is expanded and after hostilities cease.

In the countryside, the government's political situation is even worse than in Saigon and other urban centers. The VC have savagely and thoroughly destroyed the political structure of the rural areas. The potential leaders who could form the nucleus of counter-VC efforts in each locale have been killed, intimidated or driven out, leaving the old, weak and incompetent behind. Even most of the war veterans are afraid to live in the rural areas and gather in the cities. Absent security (and where there is security, absent willing and able leadership), rural reconstruction (pacification) continues to make little, if any, progress. The very difficult problem of getting pacification moving has not been solved.

Economic. SVN's economic situation today is grave for two reasons: (1) The Communist strategy has been to use some of Vietnam's economy to support the VC's own operations and to withhold the remainder from the Government and from those of the people who support it. Thus, in large part as a consequence of VC operations, Vietnam has become a food-import area, commodities cannot move freely within the country (e.g., rice from the Delta to Saigon), and the VC are in a position to derive much of their own needs (e.g., construction material, food, medical supplies) from the GVN. (2) The rapid and large US/GVN force [Page 518]build-up has introduced a new and strenuous competition for Vietnam's human and material resources. When added to the VC economic warfare efforts, the US/GVN build-up has resulted in a growing inflation in the Vietnamese economy.

3. Objectives in Vietnam. In my July 20 memorandum to you, I stated that in my view a “favorable outcome” for Vietnam has nine fundamental elements:

VC stop attacks and drastically reduce incidents of terror and sabotage.
DRV reduces infiltration to a trickle, with some reasonably reliable method of our obtaining confirmation of this fact.
US/GVN stop bombing of North Vietnam.
GVN stays independent (hopefully pro-US, but possibly genuinely neutral).
GVN exercises governmental functions over substantially all of South Vietnam.
Communists remain quiescent in Laos and Thailand.
DRV withdraws PAVN forces and other North Vietnamese infiltrators (not regroupees) from South Vietnam.
VC/NLF transform from a military to a purely political organization.
US combat forces (not advisors or AID) withdraw.

I added that a “favorable outcome” could include also arrangements regarding elections, relations between North and South Vietnam, participation in peace-keeping by international forces, membership for North and South Vietnam in the UN, and so on. It was, and is, my belief that the nine fundamental elements can evolve with or without an express agreement and, except for what might be negotiated incidental to a cease-fire, are more likely to evolve without an express agreement than with one.

Ambassador Lodge has set down his settlement terms (in Saigon 1377):5

The area around Saigon and all of the Delta (55 to 60 per cent of the population of South Vietnam), the thickly populated northeastern strip along the coast (25 per cent of the population), and all cities and provincial capitals would be pacified; and all principal roads would be open to the Vietnamese military day and night. (“Pacified” is defined as the existence of a state of mind among the people that they have a stake in the government as shown by the holding of local elections; it also means a proper local police force. A “pacified” area is economically, socially and politically a part of the Republic of Vietnam.)
Those areas not pacified would not be safehavens for the VC, but would be contested by energetic offensive forays to prevent consolidation of a Communist base.
The VC would disarm, and their weapons and explosives would be removed from their hands. Their main force units would be broken up. Hardcore VC to go to North Vietnam. Chieu Hoi (“Open Arms”) rehabilitation would be extended to individual VC who are suitable, with plans to resettle them.
North Vietnam would stop its infiltration and direction of the war.
The Government of South Vietnam would approve.

Ambassador Lodge added that “this means that we would not be insisting on the complete elimination of the Viet Cong from all corners of the country although no land or safehaven would be allotted to them. It would mean that we and the GVN would control 80 to 85 per cent of the population and that the Viet Cong would be limited to the jungle and mountainous areas where they would go on as bandits, much as their counterparts do today in Malaya and in Luzon—and where the GVN would have the right to pursue them and try to destroy them.”

The question whether we should be prepared ultimately to settle for a “compromise outcome”—for something less than the terms indicated above—may have to be faced soon. The areas where there may be pressure to ease our terms are these:

Safehavens. The degree to which the VC are permitted to hold safehavens tacitly allotted to them; the extent to which the VC are permitted to tax in “their” areas and to retain a military organization for “defensive” purposes.
Force withdrawals. The extent to which hardcore VC on the one hand and US forces on the other must withdraw from South Vietnam.
Role of NLF. The extent to which the National Liberation Front is permitted to play a role in the political life of the republic—at local, provincial and national level.

Easing of terms in any of these three areas—especially in giving the NLF status in the Saigon government—runs large risks of putting South Vietnam on the skids toward a Communist-controlled government, probably not “Titoist” but rather subservient to China. Furthermore, because of the attitudes of the present Saigon government and because of the importance of the GVN in any compromise gambit, a “compromise outcome” would be very difficult to manage. The GVN would probably support the “favorable outcomes” described above; but the GVN almost certainly would not support any “compromise outcome”—especially one which appears to “cede” territory or political recognition to the NLF.

Our tactical concern in moving toward any compromise settlement, therefore, is the high probability that the Government of South Vietnam [Page 520]will shake apart in the process. That is, so far as tactics are concerned, we seem to have the problem that if we lower our sights to a “compromise outcome,” we lose the support of the GVN, making the compromise impossible. (It follows that an essential part of any scenario to compromise would be the creation of a Saigon government not hostile to the compromise course.)

On the substantive side of the compromise issue: It should be noted that the decisions that have already been taken this year, and the ones proposed here, are consistent with the strategy stated in paragraph 1 and with a striving for a “favorable outcome,” and are probably inconsistent with any lesser objectives or other strategy. Specifically, we may already have passed the “Y” in the road: Our course of action has been and is increasingly becoming inconsistent with any design to settle, through negotiations or otherwise, for a compromise solution—especially one involving Communists in the Saigon government and the consequent high risk of quick Communist take-over.

4. Tools available. We have four tools available to us in working toward a solution to the problem in South Vietnam: (a) Military power in South Vietnam, Laos and in the adjacent waters; (b) military interdiction in and pressure on North Vietnam; (c) non-military effort in South Vietnam; and (d) efforts to negotiate.

5. Military variables. The military variables are (a) a Pause, (b) Rolling Thunder, and (c) Phase II deployments:

Pause. An interruption in the program of bombing North Vietnam. The Pause would stand down all strikes and armed reconnaissance in North Vietnam for approximately four weeks unless its futility became apparent earlier. It would not involve stoppage of other reconnaissance or intelligence operations in North Vietnam or of any operations elsewhere. Particularly, ground and air operations in South Vietnam would continue. The Pause would be publicly acknowledged, but a serious effort would be made to avoid advertising it as an ultimatum to the DRV. Third countries would be encouraged to help produce meaningful negotiations and substantial reductions in DRV/VC activity. With respect to termination of the Pause, our state of mind could be (1) “hard-line” or (2) “soft-line”:
“Hard-line” Pause. Under a “hard-line” Pause, we would be firmly resolved to resume bombing unless the Communists were clearly moving toward what is described above as a “favorable outcome.” Special care would be taken to avoid being trapped in a status-quo cease-fire or in negotiations which, though unaccompanied by real concessions by the VC, made it politically costly for us to terminate the Pause.
“Soft-line” Pause. Under a “soft-line” Pause, we would be willing to feel our way with respect to termination of the Pause, with less insistence on concrete concessions by the Communists. (This kind of Pause [Page 521]makes sense only if we are working toward a possible “compromise outcome.”)
Rolling Thunder. The program of bombing North Vietnam (1) could be intensified abruptly by a sharp, heavy blow, as recommended by the Joint Chiefs of Staff, or (2) could continue to evolve at the present pace (the possibility of reducing the program is omitted):

(1) JCS “sharp blow.” The JCS would initiate immediately, and carry out in a few days, a program of air strikes against

—Phuc Yen, Hanoi/Gia Lam, Hanoi/Bac Mai, Haiphong/Cat Bi, and Haiphong/Kien An airfields, including the aircraft thereon (Phuc Yen to be struck by B-52s).

—Rail, highway and waterway routes and traffic between Hanoi and Haiphong and between Hanoi-Haiphong and South China.

POL storage facilities at Haiphong (or four thermal power plants generating approximately 50 per cent of North Vietnam's thermal power).

SAM installations and other AA defenses which pose a threat to the above air operations.

The JCS would permit normal armed reconnaissance in all of North Vietnam, and would order follow-on strikes as necessary to keep the targets destroyed. They would conduct naval surface force operations against targets near the coast and against North Vietnamese shipping at sea. They would mine the harbors. Every effort would be made to minimize destruction of non-military installations; and the civilian populations, as such, would not be targeted.

(2) Present evolving RT. At the present time, we are carrying out 600 armed reconnaissance sorties a week in the “three quadrants” and striking 2 fixed LOC targets a week in the “northeast quadrant” (but not against targets in the “China strip,” the “Hanoi circle,” or the “Haiphong circle”). Natural evolution of the program over a five-month period would maintain activity in the three quadrants at the present level, and would evolve in the “northeast quadrant” as follows:

  • 1st month. Include Hanoi-Haiphong lines-of-communications targets (outside the Hanoi and Haiphong circles), with 2 fixed LOC targets in the quadrant each week.
  • 2nd and 3rd months. No fixed targets in the quadrant, but “controlled armed reconnaissance” of selected routes at selected times and at the rate of 100 sorties a week.
  • 4th month. EXTEND “controlled armed reconnaissance” to LOCs within the Hanoi and Haiphong circles and ADD strikes on 3 key POL targets a month.
  • 5th month. In the first two weeks, ADD mining of Haiphong harbor; thereafter, change the armed reconnaissance in the northeast quadrant to the kind carried out in the other three quadrants and with no geographical limit except the strip next to China. (Left unstruck would be population targets, power plants, and locks and dams.)
[Page 522]

c. Phase II Deployments. US force deployments (1) could stop at the Phase I 225,000+/-or (2) could continue by deployment of a Phase II 125,000+/-.

Stop at Phase I. Limit US forces to Phase I deployments. Added expenditures attributable to Phase I, through FY 1967, are estimated to be $13 billion.
Deploy Phase II forces. Deploy 28 additional combat battalions and 4 additional tactical air squadrons and associated support forces (125,000 additional men), bringing the totals to 62 US combat battalions (plus 10 third-country battalions) and 26 US tactical air squadrons, or approximately 350,000 Americans, in South Vietnam. These deployments would be essentially completed by the end of 1966. They could be accomplished without calling up the Reserves or extending tours of duty, but, in that case, they would lead to further reductions in the strength of our strategic reserve to meet contingencies elsewhere. (An alternative would be to call up Reserves—not only replenishing the strategic reserves, but also giving a clear demonstration of US power and purposes.) Expenditures attributable to Phase II for its first 18 months are estimated very roughly to total $3 billion.

It is estimated—as can be seen on Map II6—that, by the time Phase II forces are making their full impact in the Spring of 1967, the US/GVN would have “control” of 65 per cent of the population and 20-30 per cent of the land; 60-65 per cent of the critical lines of communications would be kept open; 40-50 per cent of the VC bases and military logistical infrastructure would be subject to disruption; and 30 per cent of the VC political/security infrastructure would be disrupted.

It should be noted that the MACV-proposed Phase II does not allocate any forces to the Delta. To handle the Delta properly, Phase II would have to be augmented by one or two additional divisions—30-60,000 additional men—bringing the Phase II total to 155,000 to 185,000 men.

6. Illustrative scenarios. Here are three illustrative courses of action for the future. It is my recommendation that we follow Course C.

Course A: “Soft-line” Pause, then feel way re RT and force levels toward a “compromise outcome.” This is a compromise course, aimed at reduced objectives described in paragraph 3 above. It would be extremely difficult to gain public acceptance of this course, in view of the size of the US investment already on the line; and it is very doubtful that a scenario for this course can now be written that would support the strategy in paragraph 1. The scenario probably would end up in disintegration in Vietnam, in political humiliation for the US, and in impaired US political effectiveness on the world scene. It also could lead to more costly confrontations with the Chinese later on.

[Page 523]

Course B: No Pause, with evolving RT, with no Phase II. This is a continuation of the present evolving course of action.

This course of action can be supported in this way: (1) It will be January before the weight of our full Phase I effort really comes into play. By then we might see at least a significant adverse morale trend in the VC that would indicate we were getting somewhere. (2) It may be useful to give ourselves time to digest the impact of 225,000 men in terms of their effect on the Vietnamese psychology and economy. This is a factor that some reports, notably a recent cable from Saigon on inflationary and labor division problems,7 suggest may be becoming very serious, even to the point of being a limiting factor on our basic effectiveness in appealing to the people. (3) It would give the GVN additional time, hopefully without serious frictions with us, to get on with its part of the job and to build up a GVN contribution that would not be dwarfed by the US role, a present tendency in Vietnam.

On the drawback side are two factors: (1) Deferring additional US deployments—particularly in the face of continuing increases in VC strength—raises grave problems in our domestic situation. This is not solely a question of the war dragging on—which it can be argued it would be likely to do even with Phase II deployments—but rather, a question of not having clearly done all on the ground in Vietnam that appears wise. It would be still more serious, of course, if the rate of progress slowed down, as is likely to be the case. (2) The second drawback is that, by delaying the added deployments, we shall lose momentum in Vietnam—a momentum displaying a determination and confidence that is crucial to the psychology of the confrontation if we are to emerge with an acceptable solution. That is, the product of this course of action is virtually certain to be a stalemate by March 1966 under the unsatisfactory conditions described in paragraph 2 above.

Course C: “Hard-line” Pause, then evolving RT with Phase II. This course involves all three controversial ingredients—(a) a Pause, (b) an evolving Rolling Thunder, and (c) Phase II deployments.

a. Pause. The arguments for the Pause are four: (1) It would offer the DRV and VC a chance to move toward a solution if they should be so inclined—and we continue to receive hints that such may be the state of their minds. (2) It would demonstrate to domestic and international critics that our efforts to settle the war are genuine. (3) It would probably tend to reduce the dangers of escalation after we had resumed the bombing, at least in so far as the Soviets were concerned. And (4) it would set the stage for another Pause, perhaps in late 1966, which might produce a settlement.

[Page 524]

Arguments against the Pause are not so much that the North Vietnamese might profit from the respite to repair the damage or to increase infiltration. The arguments are related to Saigon, Hanoi and US hard-liner reactions: (1) In Saigon, bringing the GVN aboard would be difficult just in itself, and it could adversely affect GVN solidity. Any major falling out between the GVN and the US, or any overturn in the GVN political structure, could at this stage set us back enormously. (2) In Hanoi, a Pause might at least confuse our message and justify their tendency to think that, despite all our actions, we are really looking for an easy way out. We could offset this to some degree at the time by the kind of message we get through to them and by our continued actions then and later in the South, but there would still be some chance that the hard-liners in Hanoi would read it the other way notwithstanding. In other words, it could conceivably be a step away from getting Hanoi to seek a peaceful solution. (3) In the US, there would be rumbling among the harder-action school of critics. This might be kept under control during the Pause, but it could also mean that the pressures would increase thereafter to hit the DRV harder.

b. Rolling Thunder. Intensification of the Rolling Thunder program, following the Pause, on an “evolving” (not “sharp blow”) basis as described in paragraph 5(b)(2) above, is designed to accomplish the three fundamentals underlying the bombing program. Those are:

  • To promote a settlement. The program was designed (1) to influence the DRV to negotiate (explicitly or otherwise), and (2) to provide us with a bargaining counter within negotiations.
  • To interdict infiltration. The program was calculated to reduce the flow of men and supplies from the North to the South—at the least, to put a ceiling on the size of war that the enemy could wage in the South.
  • To demonstrate US commitment. The program was intended to demonstrate to South Vietnam, North Vietnam and the world the US commitment to see this thing through.

While the gradual intensification runs some risk of “flashing” China or the USSR into some form of escalation, the risks of this are not great, especially after a Pause, and should be taken in order to keep pressure on the North to stop the aggression in the South.

c. Phase II deployments. The predicted results of Phase II deployments are discussed in paragraph 5(c)(2) above.

Note: There are two variants of Course C: (1) Delete the Pause, and carry out the evolving Rolling Thunder at the same time as (in parallel with) an early Phase II; (2) Postpone Phase II, putting Phase II after (in tandem with) a several-month's evolution of Rolling Thunder.

7. South Vietnamese reaction to Course C. The GVN will enthusiastically support the evolving Rolling Thunder portion of Course C; the people of South Vietnam, so long as we avoid population targets per se,[Page 525]will likewise support the program. Ambassador Lodge has advised that the GVN will react well to our Phase II deployments, performing better as a consequence (he states that the South Vietnamese, knowing our good record in the Philippines, Korea and elsewhere, are more afraid of our leaving than of our becoming entrenched as a “colonial” power);8 I agree with that assessment, provided we take appropriate action especially on the economic side to offset the shortages and inflation caused by the added US and GVN (and VC) efforts. The GVN, as indicated in the discussion of Course C in paragraph 6 above, may misconstrue the Pause—thinking that we are on Course A, looking for an easy way out; but this fear we should be able to handle. Absent Course C—or a course similar to it—we are very likely to find ourselves in a stalemate by Spring of next year. The odds are less than even that South Vietnam could hang on for long in that condition; centrifugal forces in the government and society, on which the VC have always counted to make up for their inferiority in numbers, would be likely to take over causing an eventual collapse and disintegration of the nation we have been supporting. While a similar stalemate could follow even Course C, it is considerably less likely and would certainly come later if it happened. The people of South Vietnam will, I believe, support our moves even though our numbers and functions by that time will be approaching those of an occupation force.

8. Communist reactions to Course C. The Soviets would probably seize upon the Pause as an occasion both to persuade the DRV to negotiate (and perhaps to dampen activities in the South) and to maneuver us into a position so that we could not resume the bombing of the North. With respect to the other aspects of Course C, the USSR can be expected only to throw propaganda barbs and to continue material assistance to North Vietnam.

The Chinese have already sent non-combat personnel into North Vietnam to do construction and repair work. This could be expected to continue at an even higher scale. But they would probably not send regular combat forces or aircraft into the war unless we invaded North Vietnam or struck China. As in the case of the USSR, China might react against our ships at sea if she lost a ship to a mine at Haiphong.

The DRV, like the Soviets, would try during the Pause to maneuver us into a position so that we could not resume the bombing. Later on, under the pressure of the bombing and blockade by us (and fearing the influx of men and influence from China) would probably, over time, look more favorably on a settlement in South Vietnam. At the same time, the DRV can be expected to send into Laos and South Vietnam up to several [Page 526]divisions of regular forces to help the VC. As stated in paragraph 2 above, there are indications that this is already under way. They could, by such deployments, match our deployments on a 1-for-3 or 1-for-4 basis, thus effectively canceling the likelihood that our Phase I and Phase II forces will be able to provide security to areas beyond the enclaves shown on the attached maps. If the DRV does this, we could find ourselves in a position where the pressures were great to invade or to bomb the cities in the North—actions which are likely to lead to open war with China if not with the Soviet Union.

The VC can be expected to continue their “Phase II” sabotage, murder and guerrilla activities, while continuing to strive to build up a conventional military capability. They will depend more and more on regular PAVN forces; they will draw harder on the men and material in the areas they control, including the Delta; and they can be expected to try to bring the economy of South Vietnam to a grinding halt. The question, of course, is whether the pressure on the North and the added forces in the South can frustrate these VC designs.

9. Other actions in South Vietnam. The military moves recommended in this memorandum are essential to success in Vietnam, but they are not sufficient. The heart of South Vietnam will begin to beat and the body to breathe only when GVN militia, police, intelligence and administrative personnel have been introduced in sufficient strength to saturate the area, destroy the VC infrastructure and reestablish the agencies of Government. We are repeatedly told that the GVN is “just now starting” to revitalize this program. We must work hard with the government in Saigon, we must bolster the economy (inflation must be brought under control promptly and boldly), we must press on with our psychological warfare campaign, and we should be prepared to take a more active role in administration of the provinces and districts—even though it will have some attributes of a “military occupation.” Ambassador Lodge is driving hard in these most difficult areas.

10. Actions in Laos. The pattern of infiltration and VC force augmentation strongly implies that Laos is being heavily used both as a channel and as a staging base for operations in South Vietnam. We may soon have to expand our air actions in Laos and to initiate large-scale ground operations there. Furthermore, as a part of any Phase III, if the tide of the war has not begun to turn, the time may come to attempt to seal off South Vietnam by land and sea. Contingency plans should therefore be available for the creation of a 175-mile-long anti-infiltration “barrier” to run near the 17th Parallel from the sea to Thailand. Any such barrier would of course have to be complemented by effective measures countering infiltration by sea and from Cambodia. To minimize political costs, Laotian FAR and Thai forces should probably be used at the west end, and Vietnamese and US forces at the east end of the barrier; and Souvanna's [Page 527]approval should of course be obtained if possible. (In the long term, taking into account the unstable nature of any likely settlement in Vietnam, such a barrier, perhaps manned ultimately by an international force, seems to be an essential.)

11. Political moves. Whatever military actions are taken, the US should continue to keep our lines of communication open with friends and enemies alike and should continue to keep our objectives well understood. Specifically, (a) we should continue the dialogue with Moscow, and should keep our ears open for signals from Hanoi or the VC, and (b) we should keep our allies and neutrals informed—at best, bestirring them to help toward a solution and, at worst, keeping their opposition within manageable proportions.

12. Appraisal. In view of the stagnation, and therefore the beginnings of disintegration, that is otherwise likely to occur in South Vietnam in early to mid-1966, we seem to have only two major courses open to us. One is to start immediately and carefully to plot a course toward cutting our losses (as listed under Course A in paragraph 6 above); the other is to increase our investment (via Rolling Thunder, Phase II, or both). The former course, as I indicated earlier, is inconsistent both with our strategy for Asia (see paragraph 1) and with our objectives in Vietnam (see paragraph 3). I therefore have recommended that we increase our investment in an effort to make further progress in Vietnam.

Further progress inside Vietnam, beyond that predicted for Spring 1966 in paragraph 2 above, will have to be based on one or more of the following developments:

  • —Snowballing support by the people of South Vietnam.
  • —Increases in, and better, South Vietnamese forces.
  • —Additional US or Free World forces.

Popular support can be expected to snowball only after there is “hard” security; such security for most of South Vietnam is beyond the immediate horizons. GVN force increases and improvements are under way; but they will be slow and insufficient. We are left therefore with the need to deploy additional US or Free World forces.

An acceptable and early end to the war, however, turns not only on proving that the VC cannot win in the South. It turns also on continuing pressure on the North. Both elements are required.

The situation therefore clearly requires the Phase II deployments and probably requires continued evolution of the Rolling Thunder program. The best timing of these actions is debatable—whether they should proceed together or one before the other, and whether action should be postponed by way of a Pause.

[Page 528]

Taking into account all considerations—likelihood of success, risk of escalation, South Vietnamese reaction, domestic support, etc.—I favor the tandem, one-after-the-other scenario: Pause, then evolving RT, and then Phase II.

I favor this sequence because I believe that there is a finite chance that a Pause will lead to a reduction in DRV infiltration and in VC activity (and possibly to profitable negotiations), and because I believe that a Pause is a prerequisite to US public and international acceptance of the stern actions implicit in the evolving Rolling Thunder and Phase II. I favor “evolution” of Rolling Thunder before Phase II deployments because, here too, I believe that there is a finite chance that added pressure on the North, without Phase II deployments, may be enough to bring the DRV/VC to terms; further, US public acceptance of Phase II deployments will be greater if such a Rolling Thunder program has been given a fair run first.

It must be remembered, however, that none of these actions assures success. There is a small but meaningful risk that the course I have recommended—especially the harbor-mining aspect of Rolling Thunder—will lead the Chinese or Russians to escalate the war. US killed-in-action can be expected to increase to 500-800 a month. And the odds are even that the DRV/VC will hang on doggedly, effectively matching us man-for-man (taking into account the lop-sided guerrilla war ratio advantage), while our efforts may not push the South Vietnamese over the crest of the hill, so that the snowball begins to roll our way. That is, the odds are even that, despite our efforts, we will be faced in early 1967 with stagnation at a higher level and with a need to decide whether to deploy Phase III forces, probably in Laos as well as in South Vietnam. And even if my recommended course of action is successful in moving Vietnam toward a “favorable outcome,” it will be difficult to disengage our forces. Any negotiated agreement would, as was the case in Laos, be little better than a reflection of the power situation on the ground. Unless an international force of some kind can be substituted, large numbers of US forces may be required to stay in Vietnam for some time.

My overall evaluation is that the best chance of achieving our objectives, and of avoiding a costly national political defeat, lies in the combination of political, economic and military steps described in this memorandum. If carried out vigorously, they stand the best chance of achieving an acceptable resolution of the problem within a reasonable time.

Robert S. McNamara9
  1. Source: Washington National Records Center, RG 330, McNamara Files: FRC 71 A 3470, South Vietnam, Statements and Supporting Papers. Top Secret. Designated as “1st rough draft.” McNamara wrote the following note on the source text: “A copy of this was sent to the Pres. by courier through Mac's office on 11/7 and discussed with him by me, Dean, George, and Mac on 11/7. RMcN” According to the President's Daily Diary, the President's meeting with McNamara, Rusk, Ball, and McGeorge Bundy was held at the LBJ Ranch in Texas on November 11, not November 7. See Document 189 regarding the meeting. The President was at the LBJ Ranch October 23-November 14, recovering from surgery. (Johnson Library)

    The copy of this memorandum in the Johnson Library has a covering memorandum by McNaughton, who hand-delivered it to McGeorge Bundy on November 4. McNaughton informed Bundy that Rusk and Ball were also getting copies and related McNamara's desire that Bundy not show this memorandum to anyone else without telephoning McNamara first. (Ibid., National Security File, Country File, Vietnam, Vol. XLII, Memos (B)) In The Vantage Point (pp. 233-234), Johnson describes this memorandum and recalls that his “first reaction” to it was “one of deep skepticism.”

  2. Document 67.
  3. Attached but not printed.
  4. Apparent reference to an October 27 Viet Cong attack on a GVN ranger battalion, 10 miles west of Saigon. (Memorandum for the President by the White House Situation Room briefing officer, November 2; Johnson Library, National Security File, Country File, Vietnam, Vol. XLII, Memos (A))
  5. Document 176.
  6. Attached but not printed.
  7. Not found.
  8. See Documents 170 and 188.
  9. Printed from a copy that bears this typed signature.