Join the conversation
Subscribe to Our Emails
Boston Review is a public space for the discussion of ideas and culture. Sign up for our newsletters and don’t miss a thing.
In 1963, JFK ordered a complete withdrawal from Vietnam.
President and Mrs. Kennedy arrive at Love Field, Dallas, Texas on November 22, 1963. Photograph by Cecil Stoughton, White House, in the John F. Kennedy Presidential Library and Museum, Boston.
Forty years have passed since November 22, 1963, yet painful mysteries remain. What, at the moment of his death, was John F. Kennedy’s policy toward Vietnam?
It’s one of the big questions, alternately evaded and disputed over four decades of historical writing. It bears on Kennedy’s reputation, of course, though not in an unambiguous way.
And today, larger issues are at stake as the United States faces another indefinite military commitment that might have been avoided and that, perhaps, also cannot be won. The story of Vietnam in 1963 illustrates for us the struggle with policy failure. More deeply, appreciating those distant events tests our capacity as a country to look the reality of our own history in the eye.
One may usefully introduce the issue by recalling the furor over Robert McNamara’s 1995 memoir In Retrospect. Reaction then focused mainly on McNamara’s assumption of personal responsibility for the war, notably his declaration that his own actions as the Secretary of Defense responsible for it were “terribly, terribly wrong.” Reviewers paid little attention to the book’s contribution to history. In an editorial on April 12, 1995, the New York Times delivered a harsh judgment: “Perhaps the only value of “In Retrospect” is to remind us never to forget that these were men who in the full hubristic glow of their power would not listen to logical warning or ethical appeal.” And in the New York Times Book Review four days later, Max Frankel wrote that
David Halberstam, who applied that ironic phrase [The Best and the Brightest] to his rendering of the tale 23 years ago, told it better in many ways than Mr. McNamara does now. So too, did the Pentagon Papers, that huge trove of documents assembled at Mr. McNamara’s behest when he first recognized a debt to history.
In view of these criticisms, readers who actually pick up McNamara’s book may experience a shock when they scan the table of contents and sees this summary of Chapter 3, titled “The Fateful Fall of 1963: August 24–November 22, 1963”:
A pivotal period of U.S. involvement in Vietnam, punctuated by three important events: the overthrow and assassination of South Vietnam’s president Ngo Dinh Diem; President Kennedy’s decision on October 2 to begin the withdrawal of U.S. forces; and his assassination fifty days later. (Emphasis added.)
Kennedy’s decision on October 2, 1963, to begin the withdrawal of U.S. forces from South Vietnam? Contrary to Frankel, this is not something you will find in Halberstam. You will not find it in Leslie Gelb’s editorial summary in the Gravel edition of The Pentagon Papers, even though several documents that are important to establishing the case for a Kennedy decision to withdraw were published in that edition. Nor, with just three exceptions prior to last spring’s publication of Howard Jones’s Death of a Generation—a milestone in the search for difficult, ferociously hidden truth—will you find it elsewhere in 30 years of historical writing on Vietnam.
Did John F. Kennedy give the order to withdraw from Vietnam?
* * *
Certainly, most Vietnam historians have said “no”—or would have if they considered the question worth posing. They have asserted continuity between Kennedy’s policy and Lyndon Johnson’s, while usually claiming that neither president liked the war and also that Kennedy especially had expressed to friends his desire to get out sometime after the 1964 election.
The view that Kennedy would have done what Johnson did—stay in Vietnam and gradually escalate the war in 1964 and 1965—is held by left, center, and right, from Noam Chomsky to Kai Bird to William Gibbons. It was promoted forcefully over the years by the late Walt Rostow, beginning in 1967 with a thick compilation for Johnson himself of Kennedy’s public statements on Vietnam policy and continuing into the 1990s. Gibbons’s three-volume study states it this way: “On November 26 , Johnson approved NSAM [National Security Action Memorandum] 273, reaffirming the U.S. commitment to Vietnam and the continuation of Vietnam programs and policies of the Kennedy administration.”
Equally, Stanley Karnow writes in his Vietnam: A History (1983) that Johnson’s pledge “essentially signaled a continuation of Kennedy’s policy.” Patrick Lloyd Hatcher, while writing extensively on the Saigon coup, makes no mention at all of the Washington discussions following Johnson’s accession three weeks later. Gary Hess offers summary judgment on the policy that Johnson inherited: “To Kennedy and his fellow New Frontiersmen, it was a doctrine of faith that the problems of Vietnam lent themselves to an American solution.”
Kai Bird’s 1998 biography of McGeorge and William Bundy briefly reviews the discussions of withdrawal reported to have occurred in late 1963 but accepts the general verdict that Kennedy did not intend to quit. So does Fredrik Logevall, whose substantial 1999 book steadfastly insists that the choices Kennedy faced were either escalation or negotiation and did not include withdrawal without negotiation.
All this (and more) is in spite of evidence to the contrary, advanced over the years by a tiny handful of authors. In 1972 Peter Dale Scott first made the case that Johnson’s NSAM 273—the document that Gibbons relied on in making the case for continuity—was in fact a departure from Kennedy’s policy; his essay appeared in Gravel’s edition of The Pentagon Papers. Arthur M. Schlesinger’s Robert Kennedy and His Times tells in a few tantalizing pages of the “first application” in October 1963 “of Kennedy’s phased withdrawal plan.”
A more thorough treatment appeared in 1992, with the publication of John M. Newman’s JFK and Vietnam.1 Until his retirement in 1994 Newman was a major in the U.S. Army, an intelligence officer last stationed at Fort Meade, headquarters of the National Security Agency. As an historian, his specialty is deciphering declassified records—a talent he later applied to the CIA’s long-hidden archives on Lee Harvey Oswald.
Newman’s argument was not a case of “counterfactual historical reasoning,” as Larry Berman described it in an early response.2 It was not about what might have happened had Kennedy lived. Newman’s argument was stronger: Kennedy, he claims, had decided to begin a phased withdrawal from Vietnam, that he had ordered this withdrawal to begin. Here is the chronology, according to Newman:
(1) On October 2, 1963, Kennedy received the report of a mission to Saigon by McNamara and Maxwell Taylor, chairman of the Joint Chiefs of Staff (JCS). The main recommendations, which appear in Section I(B) of the McNamara-Taylor report, were that a phased withdrawal be completed by the end of 1965 and that the “Defense Department should announce in the very near future presently prepared plans to withdraw 1,000 out of 17,000 U.S. military personnel stationed in Vietnam by the end of 1963.” At Kennedy’s instruction, Press Secretary Pierre Salinger made a public announcement that evening of McNamara’s recommended timetable for withdrawal.
(2) On October 5, Kennedy made his formal decision. Newman quotes the minutes of the meeting that day:
The President also said that our decision to remove 1,000 U.S. advisors by December of this year should not be raised formally with Diem. Instead the action should be carried out routinely as part of our general posture of withdrawing people when they are no longer needed. (Emphasis added.)
The passage illustrates two points: (a) that a decision was in fact made on that day, and (b) that despite the earlier announcement of McNamara’s recommendation, the October 5 decision was not a ruse or pressure tactic to win reforms from Diem (as Richard Reeves, among others, has contended3) but a decision to begin withdrawal irrespective of Diem or his reactions.
(3) On October 11, the White House issued NSAM 263, which states:
The President approved the military recommendations contained in section I B (1-3) of the report, but directed that no formal announcement be made of the implementation of plans to withdraw 1,000 U.S. military personnel by the end of 1963.
In other words, the withdrawal recommended by McNamara on October 2 was embraced in secret by Kennedy on October 5 and implemented by his order on October 11, also in secret. Newman argues that the secrecy after October 2 can be explained by a diplomatic reason. Kennedy did not want Diem or anyone else to interpret the withdrawal as part of any pressure tactic (other steps that were pressure tactics had also been approved). There was also a political reason: JFK had not decided whether he could get away with claiming that the withdrawal was a result of progress toward the goal of a self-sufficient South Vietnam.
The alternative would have been to withdraw the troops while acknowledging failure. And this, Newman argues, Kennedy was prepared to do if it became necessary. He saw no reason, however, to take this step before it became necessary. If the troops could be pulled while the South Vietnamese were still standing, so much the better.4 But from October 11 onward the CIA’s reporting changed drastically. Official optimism was replaced by a searching and comparatively realistic pessimism. Newman believes this pessimism, which involved rewriting assessments as far back as the previous July, was a response to NSAM 263. It represented an effort by the CIA to undermine the ostensible rationale of withdrawal with success, and therefore to obstruct implementation of the plan for withdrawal. Kennedy, needless to say, did not share his full reasoning with the CIA.
(4) On November 1 there came the coup in Saigon and the assassination of Diem and Nhu. At a press conference on November 12, Kennedy publicly restated his Vietnam goals. They were “to intensify the struggle” and “to bring Americans out of there.” Victory, which had figured prominently in a similar statement on September 12, was no longer on the list.
(5) The Honolulu Conference of senior cabinet and military officials on November 20–21 was called to review plans in the wake of the Saigon coup. The military and the CIA, however, planned to use that meeting to pull the rug from under the false optimism which some had used to rationalize NSAM 263. However, Kennedy did not himself believe that we were withdrawing with victory. It follows that the changing image of the military situation would not have changed JFK’s decision.
(6) In Honolulu, McGeorge Bundy prepared a draft of what would eventually be NSAM 273. The plan was to present it to Kennedy after the meeting ended. Dated November 21, this draft reflected the change in military reporting. It speaks, for example, of a need to “turn the tide not only of battle but of belief.” Plans to intensify the struggle, however, do not go beyond what Kennedy would have approved: A paragraph calling for actions against the North underscores the role of Vietnamese forces:
7. With respect to action against North Vietnam, there should be a detailed plan for the development of additional Government of Vietnam resources, especially for sea-going activity, and such planning should indicate the time and investment necessary to achieve a wholly new level of effectiveness in this field of action. (Emphasis added.)
(7) At Honolulu, a preliminary plan, known as CINCPAC OPLAN 34-63 and later implemented as OPLAN 34A, was prepared for presentation. This plan called for intensified sabotage raids against the North, employing Vietnamese commandos under U.S. control—a significant escalation.5 While JCS chief Taylor had approved preparation of this plan, it had not been shown to McNamara. Tab E of the meeting’s briefing book, also approved by Taylor and also not sent in advance to McNamara, showed that the withdrawal ordered by Kennedy in October was already being gutted, by the device of substituting for the withdrawal of full units that of individual soldiers who were being rotated out of Vietnam in any event.
(8) The final version of NSAM 273, signed by Johnson on November 26, differs from the draft in several respects. Most are minor changes of wording. The main change is that the draft paragraph 7 has been struck in its entirety (there are two pencil slashes on the November 21 draft), and replaced with the following:
Planning should include different levels of possible increased activity, and in each instance there be estimates such factors as: A. Resulting damage to North Vietnam; B. The plausibility denial; C. Vietnamese retaliation; D. Other international reaction. Plans submitted promptly for approval by authority.
The new language is incomplete. It does not begin by declaring outright that the subject is attacks on the North. But the thrust is unmistakable, and the restrictive reference to “Government of Vietnam resources” is now missing. Newman concludes that this change effectively provided new authority for U.S.–directed combat actions against North Vietnam. Planning for these actions began therewith, and we now know that an OPLAN 34A raid in August 1964 provoked the North Vietnamese retaliation against the destroyer Maddox, which became the first Gulf of Tonkin incident. And this in turn led to the confused incident a few nights later aboard the Turner Joy, to reports that it too had been attacked, and to Johnson’s overnight decision to seek congressional support for “retaliation” against North Vietnam. From this, of course, the larger war then flowed.
• • •
A reply to Newman’s book appeared very quickly. It came from Noam Chomsky, hardly an apologist for Lyndon Johnson or the war.
Chomsky despises the Kennedy apologists: equally the old insiders and the antiwar nostalgics—Arthur Schlesinger and Oliver Stone—and the historical memory of “the fallen leader who had escalated the attack against Vietnam from terror to aggression.” He reviles efforts to portray Kennedy’s foreign policy views as different from Johnson’s. On this point he may well be fundamentally correct, though for reasons quite different from those that he offers.
Chomsky’s Rethinking Camelot challenges Newman’s main points. First, did Kennedy plan to withdraw without victory? Or, were the plans of NSAM 263 contingent on a continued perception of success in battle? Second, did the change in NSAM 273 between the draft (which was prepared for Kennedy but never seen by him) and the final version (signed by Johnson) represent a change in policy?
Chomsky is categorical on both issues: “Two weeks before Kennedy’s assassination, there is not a phrase in the voluminous internal record that even hints at withdrawal without victory.” Elsewhere he notes that “[t]he withdrawal-without-victory thesis rests on the assumption that Kennedy realized that the optimistic military reports were incorrect. . . . Not a trace of supporting evidence appears in the internal record, or is suggested [by Newman].” And, as for the changes to NSAM 273: “There is no relevant difference between the two documents [draft and final], except that the LBJ version is weaker and more evasive.”
Chomsky denies Newman’s claim that the new version of paragraph 7 in the final draft of NSAM 273 signed by Johnson on November 26 opened the way for OPLAN 34A and the use of U.S.–directed forces in covert operations against North Vietnam. Rather, he reads the Johnson version as applying only to Government of Vietnam forces, even though the language restricting action to those forces is no longer there.
Peter Dale Scott, the former diplomat, professor of English at the University of California, Berkeley, and author of part of the Pentagon Papers, replied to Chomsky on both points almost immediately.
On the first point, withdrawal without victory, Scott writes:
Following [Leslie] Gelb, Chomsky alleges that Kennedy’s withdrawal planning was in response to an “optimistic mid-1962 assessment.” . . . But in fact the planning was first ordered by McNamara in May 1962. This was one month after ambassador Kenneth Galbraith, disenchanted after a presidentially ordered visit to Vietnam, had proposed a “political solution” based in part on a proposal to the Soviets entertaining “phased American withdrawal.”
Scott goes on to point out that it cannot be proven that Galbraith’s recommendation was responsible for McNamara’s order. But there is good reason to believe they were linked, that both reflected Kennedy’s long-term strategy on Vietnam.6 As for the proposition that no evidence hinting at withdrawal without victory exists, Scott argues that Chomsky’s “internal planning record”—for the most part the Pentagon Papers—“is in fact an edited version of the primary documents.” Moreover, “the documentary record is conspicuously defective” for November 1963. “[I]n all three editions of the Pentagon Papers there are no complete documents between the five [coup] cables of October 30 and McNamara’s memorandum of December 21; the 600 pages of documents from the Kennedy Administration end on October 30.”
On the second point, concerning NSAM 273, Scott writes that Chomsky reads “Johnson’s NSAM as if it were as contextless as a Dead Sea Scroll,” dismissing its importance and ignoring “early accounts of it as a ‘major decision,’ a ‘pledge’ that determined ‘all that would follow,’ from journalists as diverse as Tom Wicker, Marvin Kalb, and I. F. Stone.” Scott writes that Chomsky also ignores Taylor’s memo to President Johnson of January 22, 1964, which cites NSAM 273 as authority to “prepare to escalate operations against North Vietnam.”
In the course of this controversy, the ground had narrowed sharply. After Newman’s book, no one seriously disputed that Kennedy was contemplating withdrawal from Vietnam. Instead, the disagreements focused on four questions: Did the withdrawal plans depend on the perception of victory? Did Kennedy act on his plans? Were actions he may have taken noisy but cosmetic, a pressure tactic aimed at Diem or a ploy for the American public, or were they for real? And were the OPLAN 34A operations that got under way following Kennedy’s death a sharp departure from previous U.S. policy or merely a “Government of Vietnam” activity consonant with intensifying the war in the South?
• • •
The publication of McNamara’s In Retrospect sharpened the terms of debate. Some key source materials, including the texts of the McNamara-Taylor report and those of NSAM 263 and 273, have been in the public domain for years. McNamara’s 1995 account of his September 1963 mission to Vietnam makes substantial use of the McNamara-Taylor report and the quotations presented are a study in ambiguity. He quotes General Maxwell Taylor’s apparent conviction that the war could be won by the end of 1965, but then he acknowledges that there were “conflicting reports about military progress and political stability” and describes the impressive doubts of those he spoke with that the South Vietnamese government was capable of the effective actions that military victory required:
The military campaign has made great progress and continues to progress. . . . There are serious political tensions in Saigon. . . . Further repressive actions by Diem and Nhu could change the present favorable military trends. . . . It is not clear that pressures exerted by the U.S. will move Diem and Nhu toward moderation. . . . The prospects that a replacement regime would be an improvement appear to be about 50-50.
The drift seems clear enough: the Diem government is failing and there is no reason to think a replacement would be better. But the references to “great progress” leave room for doubt. Withdrawal with victory or without it?
McNamara then reproduces the precise wording of the military recommendations from Section I(B) of the report:
We recommend that:  General Harkins review with Diem the military changes necessary to complete the military campaign in the Northern and Central areas by the end of 1964, and in the Delta by the end of 1965.  A program be established to train Vietnamese so that essential functions now performed by U.S. military personnel can be carried out by Vietnamese by the end of 1965. It should be possible to withdraw the bulk of U.S. personnel by that time.  In accordance with the program to train progressively Vietnamese to take over military functions, the Defense Department should announce in the very near future presently prepared plans to withdraw 1000 U.S. military personnel by the end of 1963.
The report then went on to make a number of recommendations to “impress upon Diem our disapproval of his political program.” These matters dealt with the repression of the Buddhists and related issues; the recommendation to announce plans to withdraw 1,000 soldiers is not listed under this heading.
The reason for the ambiguity over the military situation, as well as the vague “it should be possible” wording of the second recommendation, becomes clearer when McNamara describes the National Security Council meeting of October 2, 1963, which revealed a “total lack of consensus” over the battlefield situation:
One faction believed military progress had been good and training had progressed to the point where we could begin to withdraw. A second faction did not see the war as progressing well and did not see the South Vietnamese showing evidence of successful training. But they, too, agreed that we should begin to withdraw. . . . The third faction, representing the majority, considered the South Vietnamese trainable but believed our training had not been in place long enough to achieve results and, therefore, should continue at current levels.
As McNamara’s 1986 oral history, on deposit at the Lyndon Baines Johnson Library, makes clear (but his book does not), he was himself in the second group, who favored withdrawal without victory—not necessarily admitting or even predicting defeat, but accepting uncertainty as to what would follow. The denouement came shortly thereafter:
After much debate, the president endorsed our recommendation to withdraw 1,000 men by December 31, 1963. He did so, I recall, without indicating his reasoning. In any event, because objections had been so intense and because I suspected others might try to get him to reverse the decision, I urged him to announce it publicly. That would set it in concrete. . . . The president finally agreed, and the announcement was released by Pierre Salinger after the meeting.
Before a large audience at the LBJ Library on May 1, 1995, McNamara restated his account of this meeting and stressed its importance. He confirmed that President Kennedy’s action had three elements: (1) complete withdrawal “by December 31, 1965,” (2) the first 1,000 out by the end of 1963, and (3) a public announcement, to set these decisions “in concrete,” which was made. McNamara also added the critical information that there exists a tape of this meeting, in the John F. Kennedy Library in Boston, to which he had access and on which his account is based.
The existence of a taping system in JFK’s oval office had become known over the years, particularly through the release of partial transcripts of the historic meeting of the “ExComm” during the Cuban Missile Crisis of October 1962. But the full extent of Kennedy’s taping was not known. And, according to McNamara, access to particular tapes was tightly controlled by representatives of the Kennedy family. When McNamara spoke in Austin, only he and his coauthor, Brian VanDeMark, had been granted the privilege of listening to the actual tape recordings of Kennedy’s White House meetings on Vietnam.
In 1997, however, this situation changed. The Assassination Records Review Board (ARRB), an independent civilian body established under the 1992 JFK Records Act that has already been responsible for the release of millions of pages of official records deemed relevant to Kennedy’s assassination, ruled that his tapes relating to Vietnam decision-making should be released. In July the JFK Library began releasing key tapes, including those of the withdrawal meetings on October 2 and 5, 1963.7
A careful review of the October 2 meeting makes clear that McNamara’s account is essentially accurate and even to some degree understated. One can hear McNamara—the voice is unmistakable—arguing for a firm timetable to withdraw all U.S. forces from Vietnam, whether the war can be won in 1964, which he doubts, or not. McNamara is emphatic: “We need a way to get out of Vietnam, and this is a way of doing it.”
In Retrospect’s discussion of Kennedy’s decision to withdraw ends at this point. McNamara makes no mention of NSAM 263. However, on the tape of the meeting of October 5, 1963, one can clearly hear a voice—it may be Robert McNamara or McGeorge Bundy—asking President John F. Kennedy for “formal approval” of “items one, two, and three” on a paper evidently in front of them. It is clear that one of these items is the recommendation to withdraw 1,000 men by the end of 1963, the rationale being that they are no longer needed. This short exchange is thus unmistakably a request for a formal presidential decision concerning the McNamara-Taylor recommendations. After a short discussion of the possible political effect in Vietnam of announcing this decision, the voice of JFK can be clearly heard: “Let’s go on ahead and do it,” followed by a few words deciphered by historian George Eliades as “without making a public statement about it.”
Unfortunately, the last White House tape from the Kennedy administration is dated November 7, 1963. The archivists at the JFK Library have no information on why the tapings either ended or are unavailable for later dates. McNamara states that he has “no specific memory” of the Honolulu Conference that he was sent to chair on November 20, 1963.
The President of the United States does not make decisions in a vacuum. Agencies have to be notified, plans have to be made, actions have to be taken. Part of the enduring doubt over Kennedy’s decision to withdraw U.S. forces from Vietnam surely stems from the failure of this decision to cast a shadow in the primary record, and particularly in the Pentagon Papers, on which so many historians have relied for so many years. Furthermore, a persistent skeptic can still point to the “it should be possible” language of the McNamara-Taylor Report with respect to the final date of 1965 as leaving an “out” for the case where the military situation might turn sour. In two years and two months, much can happen, as events would prove.
But as Scott already pointed out to Chomsky in 1993, the primary record available to date has been heavily edited. Documents from November 1, 1963, through early December are conspicuously missing. So, we now learn, are many others.
In January 1998, again under the supervision of the ARRB, about 900 pages of new materials were declassified and released from the JCS archives. These include important records from May 1963, from October, and from the period immediately following Kennedy’s death; many had been reviewed for declassification in 1989 but were not declassified at that time. They clarify considerably the nature of the “presently prepared plans” referred to in the McNamara-Taylor third recommendation, and they give the military leadership’s interpretation of the direction they were getting from JFK. Since it is well known that the Pentagon did not favor withdrawal, it is fair to assume that if wiggle room existed in the President’s instructions it would surface in these documents.
Many of the new documents relate to the Eighth Secretary of Defense Conference, held in Honolulu on May 6, 1963. Here one gets a taste of McNamara’s skepticism and the replies of the brass. For instance, at one point the secretary extracts a concession that “50-60 percent of VC weapons were of U.S. origin.” A bit later, we read: “GEN HARKINS stated that for effective control the border should be defined, marked and cleared similar to the Greek boundary with Albania and Bulgaria. However, this cannot be done in the foreseeable future.”
Turning to the development of a “comprehensive plan,” the documents immediately reflect discussions of a phase-down in the U.S. presence. For instance: “SEC MCNAMARA stated that our efforts should be directed toward turning over equipment now in U.S. units supporting the Vietnamese as rapidly as possible. He added that we must avoid creating a situation that now obtains in Korea where we are presently spending almost half a billion dollars per year in foreign aid.” A little later, we find a decision noted: “1. Draw up training plans for the RVNAF that will permit us to start an earlier withdrawal of U.S. personnel than proposed under the plans presented.” And: “d. Plan to withdraw 1000 U.S. military personnel from RVN by December 1963.”
Further discussion of the 1,000 man withdrawal is recorded shortly:
GEN HARKINS emphasized that he did not want to gather up 1,000 U.S. personnel and have them depart with bands playing, flags flying etc. This would have a bad effect on the Vietnamese, to be pulling out just when it appears they are winning. SEC MCNAMARA stated that this would have to be handled carefully due to the psychological impact. However, there should be an intensive training program of RVNAF to allow removal of U.S. units rather than individuals.
There follows considerable discussion of proposals to launch raids on North Vietnam. For Geneva convention reasons, it is agreed that these must be covert. Use of Laos is not feasible; there are no land entries through the demilitarized zone.
As for sea entry, available boats are susceptible to weather and too slow. Sea is the only means of exfiltration. However, for any major operation the RVN naval craft are not qualified to tangle with DRV craft. . . . Build-up in CIA resources by end CY 1963 includes 40 teams in addition to 9 in country. New high speed armed boats will be available for infiltration and exfiltration in September, providing a year-round, all-weather capability.
Thus emerges an answer to one of the critical questions separating Newman and Scott from Chomsky. OPLAN 34A, when it emerged in November, would be a CIA operation. It could not be otherwise, for the Government of Vietnam did not possess the boats.8
Eventually, discussion turns to projected force structures, and a table titled “CPSVN—FORECAST OF PHASE-OUT OF US FORCES” gives precise estimates, by major unit, of the projected American commitment through 1968. McNamara’s reaction to this timetable is recorded clearly:
In connection with this presentation, made by COMUSMACV (attached hereto), the Secretary of Defense stated that the phase-out appears too slow. He directed that training plans be developed for the GVN by CINCPAC which will permit a more rapid phase-out of U.S. forces, stating specifically that we should review our plans for pilot training with the view to accelerating it materially. He made particular point of the desirability of speeding up training of helicopter pilots, so that we may give the Vietnamese our copters and thus be able to move our own forces out. ACTION: Joint Staff (J-3); message directive to CINCPAC, info COMUSMACV. (Emphasis added.)
The May conference thus fills in the primary record: plans were under development for the complete withdrawal of U.S. forces from Vietnam. On October 2, 1963, as we have previously seen, President Kennedy made clear his determination to implement those plans—to withdraw 1,000 troops by the end of 1963, and to get almost all the rest out by the end of 1965. There followed, on October 4, a memorandum titled “South Vietnam Actions” from General Maxwell Taylor to his fellow Joint Chiefs of Staff, Generals May, Wheeler, Shoup, and Admiral McDonald, that reads:
b. The program currently in progress to train Vietnamese forces will be reviewed and accelerated as necessary to insure that all essential functions visualized to be required for the projected operational environment, to include those now performed by U.S. military units and personnel, can be assumed properly by the Vietnamese by the end of calendar year 1965. All planning will be directed towards preparing RVN forces for the withdrawal of all U.S. special assistance units and personnel by the end of calendar year 1965. (Emphasis added.)
“All planning” is an unconditional phrase. There is no contingency here, or elsewhere in this memorandum. The next paragraph reads:
c. Execute the plan to withdraw 1,000 U.S. military personnel by the end of 1963 per your DTG 212201Z July, and as approved for planning by JCS DTG 062042Z September. Previous guidance on the public affairs annex is altered to the extent that the action will now be treated in low key, as the initial increment of U.S. forces whose presence is no longer required because (a) Vietnamese forces have been trained to assume the function involved; or (b) the function for which they came to Vietnam has been completed. (Emphasis added.)
This resolves the question of how the initial withdrawal was to be carried out. It was not to be a noisy or cosmetic affair, designed to please either U.S. opinion or to change policies in Saigon. It was rather to be a low-key, matter-of-fact beginning to a process that would play out over the following two years. The final paragraph of Taylor’s memorandum underlines this point by directing that “specific checkpoints will be established now against which progress can be evaluated on a quarterly basis.” There is much more in the JCS documents to show that Kennedy was well aware of the evidence that South Vietnam was, in fact, losing the war. But it hardly matters. The withdrawal decided on was unconditional, and did not depend on military progress or lack of it.
Four days after Kennedy was killed, NSAM 273 incorporated the new president’s directives into policy. It made clear that the objectives of Johnson’s policy remained the same as Kennedy’s: “to assist the people and government of South Vietnam to win their contest against the externally directed and supported Communist conspiracy” through training support and without the application of overt U.S. military force. But Johnson had also approved intensified planning for covert action against North Vietnam by CIA-supported South Vietnamese forces.
With this, McNamara confirms one of Newman’s central claims: NSAM 273 changed policy. Yes, the “central objectives” remained the same: a Vietnamese war with no “overt U.S. military force.” But covert force is still “U.S. military force.” And that was introduced or at least first approved, as McNamara writes, by NSAM 273 within four days of Kennedy’s assassination.Moreover, McNamara effectively supports Newman on the meaning of NSAM 273’s seventh paragraph, which was inserted in the draft (as we have seen) sometime between November 21 and 26—after the Honolulu meeting had adjourned and probably after Kennedy died.
A final military document is relevant here. Dated December 11, 1963, it is titled “Department of Defense Actions to Implement NSAM No. 273, 26 November 1963.” This document was prepared by Marine Lieutenant Colonel M. C. Dalby; it is from CINCPAC files and is labeled “Group 1—Excluded from Automatic Downgrading and Declassification.” The document begins coldly:
“After reviewing the recent discussions of South Vietnam which occurred in Honolulu and after discussing the matter further with Ambassador Lodge, the President directed that certain guidance be issued to various Government Agencies. This was promulgated in the form of National Security Action Memorandum 273, 26 November 1963.”
There is no reference to the change of commander in chief, which had occurred within the time frame indicated by the opening sentence. The particular importance of this document is its reference to paragraph 7 of NSAM 273.
Planning for intensified action against North Vietnam was directed following the Honolulu Conference (JCS 3697, 26 Nov 1963) in the form of a 12-month program. . . . A deadline of 20 Dec 63 has been set for completion of the plan.
There are then notes that these requirements were communicated to CINCPAC and COMUSMACV on December 2, with a reply from COMUSMACV on December 3. CIA station guidance, however, happened even more rapidly than that:
CIA guidance to Saigon Station for intensified planning was dispatched following the Honolulu Conference (CAS 84972, 25 Nov 63). (Emphasis added.)
In other words, the CIA began developing intensified plans to implement OPLAN 34A, the program of seaborne raids and sabotage against North Vietnam that would lead to the Gulf of Tonkin incident and eventually to the wider war, one day before President Johnson signed the directive authorizing that action. How this happened, and its precise significance, remains to be determined.9
John F. Kennedy had formally decided to withdraw from Vietnam, whether we were winning or not. Robert McNamara, who did not believe we were winning, supported this decision.10 The first stage of withdrawal had been ordered. The final date, two years later, had been specified. These decisions were taken, and even placed, in an oblique and carefully limited way, before the public.
Howard Jones makes two large contributions to this tale. One of them is simply range, depth, and completeness. His recent book Death of a Generation is a full history of how the assassinations of Diem and then of JFK prolonged a war that otherwise might have ended quietly within a few years. Where this essay has presented the story-within-a-story of just a few Washington weeks, Jones goes back to the start of the 1960s, chronicling the struggle for power and policy that marked the whole of Kennedy’s thousand days. And he presents a reasonably complete account of the archival record surrounding the withdrawal decisions of October 1963.
Equally important, Jones’s reach extends to Saigon. In a long and fascinating section he outlines the intrigues that led to the murders of Diem and his brother Ngo Dinh Nhu on November 1, 1963. Here, Kennedy’s White House appears at its worst. It was fractious, disorganized, preoccupied with American politics, ignorant of the forces it faced in Vietnam. Diem’s mistreatment of the Buddhists, which provoked the monk Quang Duc to burn himself on a Saigon street in June 1963, traumatized the White House. And following that incident, Madame Nhu and her remarks about “barbecued bonzes” were an irritant out of proportion to their importance. Thus, in part, the decision to dissociate from Diem.
In August 1963 it was a faction of subordinates (Averell Harriman, Roger Hilsman, Michael Forrestal) who seized the opportunity to foment a Saigon coup, taking advantage of the absence of the most senior officials over a Washington weekend. Then, having set events in motion, the White House became preoccupied with a deniability that was wholly implausible. Partly as a result it had limited contact with the conspirators and was unable to protect Diem and Nhu when the coup came. Diem was indefensible in many ways. But the coup went forward with no alternative in view; and as the French ambassador to Saigon put it at the time: “any other government will be even more dependent on the Americans, will be obedient to them in all things, and so there will be no chance for peace.” Meanwhile, there are tantalizing undercurrents of what might have been. Was Nhu in discussions with intermediaries for Ho Chi Minh, with the possibility that there might have been a deal between North and South to boot the Americans from Vietnam? It appears that he was. And had he succeeded, it would have saved infinite trouble.
U.S. policy over Vietnam changed again in late November1963. The main change was a decision to authorize OPLAN 34-A—minor but fateful commando raids against targets in the North. The decision to launch covert attacks on North Vietnam does not by itself establish that Lyndon Johnson wanted a larger war. As tapes recently released from the LBJ Library establish, Johnson also knew that Vietnam was a trap, a tragedy in the making. He feared that a catastrophe would follow. In this respect, Johnson and Kennedy were similar.
And yet, Johnson could not muster Kennedy’s determination, one might say blind determination, to avoid the disaster. He acceded to proposals for covert action, and he promised the military, on November 24, that they could have what they wanted. And so the sequence of events that led to the Tonkin Gulf, to our retaliation, to the North Vietnamese decision to introduce their own main forces in the South, and to our decision to introduce main forces, played out. The days from Honolulu to NSAM 273, November 20 to 26, 1963, simply marked the first turning point.
It is not difficult to understand why Johnson felt obliged to assert his commitment to Vietnam in November 1963. To continue with Kennedy’s withdrawal, after his death, would have been difficult, since the American public had not been told that the war was being lost. Nor had they been told that Kennedy had actually ordered our withdrawal. To maintain our commitment, therefore, was to maintain the illusion of continuity, and this—in the moment of trauma that followed the assassination—was Johnson’s paramount political objective. Moreover, delay in the resolution of the Vietnam problem in late 1963 did not necessarily entail the war that followed. Our commitment then was still small. Tonkin Gulf and its aftermath lay almost a year into the future. Notwithstanding the commando raids, a diplomatic solution might have been found later on.
Left in charge, Lyndon Johnson temporized, agonized, and cursed the fates. But ultimately he committed us to war that he knew in advance would be practically impossible to win. Nothing can erase this. And yet meanwhile, alongside McNamara, he too prevented any steps that might lead to an invasion of the North, direct conflict with China, and nuclear confrontation. He bided his time, until the trauma of Tet in January of 1968 and his own departure from politics in March liberated him to do what Kennedy had done over Laos in 1961: send Harriman to end it at the negotiating table.
* * *
Why did Johnson do it? He was not misinformed about the prospects for sucess. He was not crazy. His political fate in 1964 did not depend on a show of toughness. But one possibility is that the alternatives, as he saw them, were worse. To appreciate this possibility, one needs to grasp not one but two exceptionally thorny nettles: that of the strategic balance in the early 1960s on the one hand, and that of the assassination of John F. Kennedy on the other. In contemplating Johnson’s dilemma we find ourselves poised between the two black holes of the modern history of the United States.11
Kennedy’s decision to withdraw from Vietnam was, as Jones writes, “unconditional, for he approved a calendar of events that did not necessitate a victory.” It was also part of a larger strategy, of a sequence that included the Laos and Berlin settlements in 1961, the non-invasion of Cuba in 1962, the Test Ban Treaty in 1963. Kennedy subordinated the timing of these events to politics: he was quite prepared to leave soldiers in harm’s way until after his own reelection. His larger goal after that was to settle the Cold War, without either victory or defeat—a strategic vision laid out in JFK’s commencement speech at American University on June 10, 1963.
And that was, partly, a question of atomic survival—a subject that can only be said to have obsessed America’s civilian leadership in those days, and for very good reason. The Soviet Union, which had at that time only four intercontinental rockets capable of hitting the U.S. mainland, was not the danger that rational men most feared. The United States held an overwhelming nuclear advantage in late 1963. Accordingly, our nuclear plans were not actually about deterrence. Rather, then as evidently again now, they envisioned preventive war fought over a pretext.12 There were those who were dedicated to carrying out those plans at the appropriate moment. In July 1961, the nuclear planners had specified that the optimal moment for such an attack would come at the end of 1963.
And yet, standing against them (as Daniel Ellsberg was told at the time), the civilian leaders of the United States were determined never, under any circumstances, to allow U.S. nuclear weapons to be used first—not in Laos or Vietnam, nor against China, not over Cuba or Berlin, nor against the Soviet Union. For political reasons, at a moment when Americans had been propagandized into thinking of the atomic bomb as their best defense, this was the deepest secret of the time.
Was it also a deadly secret? Did LBJ have reason to fear, on the day he took office, that he was facing a nuclear coup d’etat?13 Similar questions have engendered scorn for 40 years. But they are not illegitimate—no more so, let me venture, than the idea that Kennedy really had decided to quit Vietnam. Perhaps someday a historian will answer them as well as Howard Jones has now resolved the Vietnam puzzle. Meanwhile, let us hope that we might learn something about the need to recognize and cope with policy failure. And as for the truth behind the darkest state secrets, let us also hope that the victims of September 11, 2001, don’t have to wait as long. <
James K. Galbraith, a 2003 Carnegie Scholar, holds the Lloyd M. Bentsen, Jr., Chair of Government/Business Relations at the Lyndon B. Johnson School of Public Affairs, University of Texas at Austin.
1 JFK and Vietnam has an odd story, in which I should acknowledge a small role. On release, it received a front-page review by Arthur Schlesinger Jr. in the New York Times Book Review. But of some 32,000 copies printed (in two printings, according to Newman) only about 10,000 were sold before Warner Books abruptly ceased selling the hardcover—a fact I discovered on my own in the fall of 1993, when I attempted to assign it to a graduate class. I met Newman in November 1993, partly through the good offices of the LBJ Library. I carried his grievance personally to an honorable high official of Time Warner, whose intervention secured the return of his rights. Still, the hardback was never reissued, and no paperback has appeared.
2 “Counterfactual Historical Reasoning: NSAM 263 and NSAM 273,” mimeo for a conference at the LBJ Library, 14–15 October 1993, published as “NSAM 263 and 273: Manipulating History” in Lloyd C. Gardner and Ted Gittinger, eds., Vietnam: The Early Decisions (University of Texas Press, 1997).
3 Reeves, author of President Kennedy: Profile of Power, made this argument in a televised lecture at the LBJ Library in early 1995.
4 In a contribution to Vietnam: The Early Decisions, Newman adds a further reason: Kennedy had, on October 2, allowed McNamara and Taylor to announce, as their recommended target date, that the withdrawal be completed by 1965. It would have been awkward to follow just three days later with a presidential decision making clear that the timetable was, in fact, a firm one.
5 The fate of these commandos surfaced in the New York Times of 14 April 1995, where it was reported that after 30 years in prison, many were denied immigration to the United States because of a lack of service records.
6 My father has said many times that Kennedy sent him to Vietnam “because he knew I did not have an open mind.”
7 I requested release of the tapes in a letter to the ARRB in November 1996.
8 CINCPAC was developing these plans, but they had not been shown to JFK, according to Newman.
9 According to Newman, LBJ took a belligerent tone at his first Vietnam meeting as President on November 24, and McGeorge Bundy attributed the escalatory language in NSAM 273 to this. However, by any standard the CIA moved quickly, and by this account it relied on the discussions at Honolulu—which occurred while JFK was still alive.
10 I have in this narrative deliberately underplayed the role of my own father, who was repeatedly called upon by Kennedy to deliver arguments in favor of disengagement from Vietnam, and whose 1962 recommendation for phased withdrawal was probably the basis of the 1963 orders. My father did not know that the actual decision was taken in October 1963, but he is in no doubt as to Kennedy’s determination: he recalls Kennedy in 1962 saying to him privately and unmistakably that withdrawal from Vietnam, as that from Laos and the detachment from Cuba, was a matter of political timing.
11 My father retains a distinct, chilling recollection of LBJ’s words to him, in private, on one of their last meetings before the Vietnam War finally drove them apart: “You may not like what I’m doing in Vietnam, Ken, but you would not believe what would happen if I were not here.”
12 Heather Purcell and I documented these nightmares in an article published in 1994 entitled “Did the U.S. Military Plan a Nuclear First Strike for 1963?” It is still available on the website of the American Prospect. When once I asked the late Walt Rostow if he knew anything about the National Security Council meeting of July 20, 1961 (at which these plans were presented), he responded with no hesitation: “Do you mean the one where they wanted to blow up the world?”
13 There is no doubt that the danger of nuclear war was on Johnson’s mind. It also explains important points about his behavior in those days, including his orders to Earl Warren and Richard Russell (the latter in a phone call, a recording of which has long been available on the C-SPAN website) as to how they would conduct their commission. The point to appreciate is that there is only one way a war could have started at that time: by preemptive attack by the United States against the Soviet Union.
© 1997–2003 by James K. Galbraith. All rights reserved.
Originally published in the October/November 2003 issue of Boston Review
...we need your help. Confronting the many challenges of COVID-19—from the medical to the economic, the social to the political—demands all the moral and deliberative clarity we can muster. In Thinking in a Pandemic, we’ve organized the latest arguments from doctors and epidemiologists, philosophers and economists, legal scholars and historians, activists and citizens, as they think not just through this moment but beyond it. While much remains uncertain, Boston Review’s responsibility to public reason is sure. That’s why you’ll never see a paywall or ads. It also means that we rely on you, our readers, for support. If you like what you read here, pledge your contribution to keep it free for everyone by making a tax-deductible donation.
Vital reading on politics, literature, and more in your inbox
Printing Note: For best printing results try turning on any options your web browser's print dialog makes available for printing backgrounds and background graphics.