The Pentagon Papers
Gravel Edition
Volume 4
Chapter I, "The Air War in North Vietnam, 1965-1968," pp. 1-276.
(Boston: Beacon Press, 1971)

Section 2, pp. 58-112


1. Background

When the 37-day bombing pause was terminated at the end of January 1966, the principal issue before decision-makers was not whether to intensify the bombing but whether the intensification should be gradual as before or be sharply accelerated.

Some kind of escalation if the bombing pause failed, i.e., if the North Vietnamese did not give "concrete evidence of a willingness to come to terms," was foreshadowed by the October paper from State recommending the pause:

We would have to convey our intent to reinstitute the bombing if the North Vietnamese refused to negotiate or if their willingness to negotiate is not accompanied by a manifest reduction of VC aggression in the South. If it is necessary to reinstitute bombing, we should be prepared to consider increasing the pressure, e.g. through striking industrial targets, to make clear our continuing, firm resolve.

According to this thinking, failure of the pause would indicate that the bombing had not exerted enough pressure; greater effort was needed to convince Hanoi that the U.S. intended not only to continue the bombing but to do so on an increasing scale. Moreover, the pause had improved the political atmosphere for escalation. U.S. willingness to negotiate and NVN's unreasonableness had been amply and dramatically displayed for all the world to see. If the U.S. now decided to intensify the bombing, the decision could at least be presented as one that was made reluctantly after trying to find a more peaceful alternative.

The debate over the form of escalation in early 1966 was a continuation of the debate over bombing policy which had surfaced again in the fall of 1965, and which had mixed into the debate over the long pause. Regardless of any pause, it was clear by November that even the gradual rate of escalation of 1965 was approaching a point at which any further increase would be possible only by attacking the sensitive targets in the Hanoi/Haiphong sanctuaries and the China buffer zone. As of the end of October, 126 of the 240 existing JCS targets had been struck; and of the remaining 114, two thirds (75) were in the off-limits areas, and 29 of the other 39 remaining were in the touchy northeast quadrant. As the debate gathered momentum in the winter of 1965 without a clear decision to begin attacking "the hostage," the bombing actually levelled off. During November and December only 8 more JCS targets were struck and armed reconnaissance missions were held to a sortie ceiling of 1200 per two-week period.

Apart from general cautiousness about the next obvious escalatory step, one of the reasons for the Administration's hesitancy was apparently the fear that the timing might not be right. As the bombing drew closer to Hanoi and Haiphong, some officials felt forcing the pace might oblige NVN to confront the issue of negotiations versus greater Chinese and/or Soviet involvement prematurely, i.e. before NVN was sufficiently convinced that it could not outlast the U.S. and win in the South. The theory was that so long as Hanoi was hopeful there was a greater risk that it would opt for escalation rather than a compromise settlement. As the October paper from State put it:

We may be able to recognize the optimum time for exerting further pressure by increasing the level of our bombing, but an increase in our bombing of the North at the present time may bring matters to a head too soon.

In addition, of course, there was good reason to hold off any escalation until a substantial bombing pause was undertaken, both to test Hanoi's intentions and to disarm critics on the dovish side who felt that the Administration had not gone far enough to meet Hanoi halfway.

a. JCS Recommendations

Dissatisfied with the measured pace of the bombing program from the start, they again began advocating a sharp intensification of the bombing in early November. Diplomatic and political considerations were secondary. Their position was that ROLLING THUNDER had succeeded in making it "substantially" more costly and difficult for NVN to support the insurgents in Laos and SVN, and had "substantially" degraded NVN's capability to conduct a conventional invasion of the South, but they agreed that the campaign had not materially reduced NVN's other military capabilities, damaged its economy, deterred it from supporting the war in the South, or brought it closer to the conference table. It was not because of any difficulty in applying pressure on Hanoi by bombing or in interdicting support South that the program had not been more successful, however; it was because numerous "self-imposed restraints" had limited the potential effectiveness of the program:

. . . we shall continue to achieve only limited success in air operations in DRV/Laos if required to operate within the constraints presently imposed. The establishment and observance of de facto sanctuaries within the DRV, coupled with a denial of operations against the most important military and war supporting targets, precludes attainment of the objectives of the air campaign. . . . Thus far, the DRV has been able and willing to absorb damage and destruction at the slow rate. Now required is an immediate and sharply accelerated program which will leave no doubt that the US intends to win and achieve a level of destruction which they will not be able to overcome. Following such a sudden attack, a follow-on program of increasing pressures is necessary, but at a rate of increase significantly higher than the present rate.

The JCS accordingly recommended an immediate acceleration in the scale, scope, and intensity of the bombing, beginning with heavy strikes against POL targets and power plants in the Hanoi/Haiphong area and continuing with aerial mining of NVN ports and air strikes against the remaining "military and warsupporting" targets. Specifically, the JCS proposed an immediate sharp blow against the remaining 9 of the original 13 major POL tank farms, most of them in the Hanoi/Haiphong area, and against 5 key power plants, 2 in Hanoi and others at Uong Bi, Thai Nguyen, and Hon Gai, in order to "materially reduce enemy military capabilities." These strikes would be followed by an accelerated program of fixed target and armed reconnaissance strikes to cut down NVN's ability to direct and support the war in the South. The follow-on program would attack first the major airfields in the Hanoi/Haiphong area; then the rail, road, and waterway LOCs throughout NVN, including the major LOC targets in the Hanoi/Haiphong area, "at a rate of destruction that would exceed the recuperability rate"; then the ports at Haiphong, Hon Gai, and Cam Pha; and finally military installations and other targets of military significance, such as the Ministry of Defense, the Radio Transmitter Station, and the Machine Tool Plant in Hanoi; the Ammunition Depot at Haiphong; and the Iron-Steel Combine and Army Supply Depot at Thai Nguyen. SAM installations and other antiaircraft defenses would be attacked in order to keep friendly losses down. According to the proposal, most of the significant fixed targets in NVN would be destroyed within three or four months. Thereafter, the effort would concentrate on keeping the targets inoperative and maintaining the pressure on LOCs.

The JCS proposal to escalate all aspects of the bombing was largely oriented toward greatly increasing the pressure on Hanoi's will. On the same day, however, in a separate memorandum, the JCS made a strong pitch for an immediate attack on the NVN POL system as an interdiction measure:

Attack on this system would be more damaging to the DRV capability to move war-supporting resources within country and along the infiltration routes to SVN than an attack against any other single target system.

It is not surprising that the JCS singled out the POL target system for special attention. NVN had no oil fields or refineries, and had to import all of its petroleum products, in refined form. During 1965, it imported about 170,000 metric tons, valued at about $4.8 million. Nearly all of it came from the Black Sea area of the USSR and arrived by sea at Haiphong, the only port capable of conveniently receiving and handling bulk POL brought in by large tankers. From large tank farms at Haiphong with a capacity of about one-fourth of the annual imports, the POL was transported by road, rail, and water to other large storage sites at Hanoi and elsewhere in the country. Ninety-seven percent of the NVN POL storage capacity was concentrated in 13 sites, 4 of which had already been hit. The other 9 were still off limits. They were, of course, highly vulnerable to air attack.

In making the recommendation, the JCS emphasized the interdiction effects. They pointed out that the strikes would not hurt the industrial base or the civilian economy very much. They would directly affect the military establishment, which consumed some 60 percent of all POL, and the "government transportation system," which consumed nearly all the rest. Supplying the armed forces in NVN as well as in Laos and SVN depended heavily on POL-powered vehicles, and this dependence had if anything increased as a result of air attacks on the railroads:

The flow of supplies to all communist military forces, both in and through the country to SVN and Laos, would be greatly impeded since POL-fueled carriers are the principal vehicles for this transport. Further, the interdiction of rail lines and destruction of railroad rolling stock has resulted in the need to move increased tonnages by alternate means, primarily trucks and motor driven water craft. Thus, the most effective way to compound the current interdiction of DRV LOCs, and to offset the introduction and use of substitute modes and routes, is to reduce drastically the available supply of POL.

The JCS also suggested that POL in NVN was becoming increasingly important to the effort in the South. There were now 5 confirmed and 2 suspected NVA regiments in SVN, increasing the load on the supply lines through Laos, and the roads there were being improved, indicating that NVN planned to rely more heavily on trucks to handle the load. Significantly, the importation of trucks was increasing, and despite losses inflicted by ROLLING THUNDER strikes, the size of the truck fleet was growing.

The JCS recommended hitting the most important target, Haiphong POL storage, first, followed closely by attack on the remaining 8 targets. The weight of effort required was 336 strike and 80 flak suppression aircraft, with not more than 10 losses predicted. All POL targets could be destroyed with only light damage to surrounding areas and few civilian casualties (less than 50).

According to the JCS, the destruction of the Haiphong target "would drastically reduce the capability to receive and distribute the major portions of DRV bulk POL imports." Destruction of the others would "force reliance upon dispersed POL storages and improvised distribution methods." Recovery would be difficult and time-consuming. As stated in an annex to the JCSM:

Recuperability of the DRV POL system from the effects of an attack is very poor. Loss of the receiving and distribution point at Haiphong would present many problems. It would probably require several months for the DRV, with foreign assistance, to establish an alternate method for importing bulk POL, in the quantities required. An alternative to bulk importation would be the packaging of POL at some point for shipment into NVN and subsequent handling and distribution by cumbersome and costly methods over interdicted LOCs. Loss of bulk storage facilities would necessitate the use of small drums and dispersed storage areas and further compound the POL distribution problem.

Any further delay in carrying out the strikes, on the other hand, "will permit further strengthening of DRV active defenses of the POL, as well as the improvement of countermeasures, such as dispersed and underground storages." On the latter point, the appendix to the JCSM added detailed intelligence information that boded ill for any procrastination:

Current evidence shows that the DRV has in progress an extensive program of installing groups of small POL tanks in somewhat isolated locations and throughout the Hanoi area. Photographs reveal groups of tanks ranging in number of 16 to 120 tanks per group. The facilities are generally set into shallow excavations and are then earth-covered leaving only the vents and filling apparatus exposed. This construction was observed at several places in the Hanoi area in August and appeared to be an around-the-clock activity. . . . In addition, considerable drum storage has been identified.

It appeared that NVN had already begun a crash program to drastically reduce the vulnerability of its POL storage and handling system. As in other instances, NVN expected further escalation of the bombing, and was preparing for it.

b. The Intelligence Community Demurs

There was no immediate action on the November 1965 JCS recommendations, but they were taken under study. Secretary McNamara asked for intelligence evaluations, and on 27 November and 3 December, respectively, he received special reports from the Board of National Estimates on (a) U.S. air attacks on NVN petroleum storage facilities, and (b) a generally stepped-up effort involving doubling or tripling U.S. troop commitments, bombing military and industrial targets in the Hanoi/Haiphong area, and mining NVN harbors.

The Board reported that strikes against POL targets in the Hanoi/Haiphong area would represent "a conspicuous change in the ground rules" which the U.S. had hitherto observed, but would not appreciably change the course of the war:

. . . the Communists would unquestionably regard the proposed US attacks as opening a new stage in the war, and as a signal of US intention to escalate the scale of conflict. . . . We do not believe, however, that the attacks in themselves would lead to a major change of policy on the Communist side, either toward negotiations or toward enlarging the war.

The strikes would cause strains and embarrassment but would not have a major military or economic impact:

Hanoi would not be greatly surprised by the attacks. Indeed . . . it has already taken steps to reduce their impact. It has developed some underground storage facilities, and some capacity for dispersed storage in drums. . . .We believe that the DRV is prepared to accept for some time at least the strains and difficulties which loss of the major POL facilities would mean for its military and economic activity. It is unlikely that this loss would
cripple the Communist military operations in the South, though it would certainly embarrass them.

NVN might possibly ask the Chinese to intervene with fighter aircraft to help defend the targets but would probably not ask for ground troops. The Chinese would probably decline to intervene in the air and would not volunteer ground forces, though they would urge NVN to continue the war. The Soviets would be "concerned" at the prospect of a further escalation of the bombing:

The Soviets would find their difficulties and frustrations increased. . . . They are committed to provide defense for North Vietnam, and . . . their inability to do so effectively would be dramatized. . . . We believe that they would not change their basic policy of avoiding overt involvement in combat while giving extensive military equipment and economic assistance to NVN. But their relations with the US would almost certainly deteriorate, for it is the bombing of North Vietnam which is, for Moscow, the most nearly intolerable aspect of [the War-]

In its estimate of the likely reactions to the wider course of substantially expanding the U.S. effort in the South, together with the bombing and aerial mining of the North, the Board similarly offered little hope that the escalation would produce any marked improvement in the situation. They characterized NVN's will to resist in the North and to persevere in the South as virtually unshakeable in the short run and extremely tough even in the long run:

Present Communist policy is to continue to prosecute the war vigorously in the South. The Communists recognize that the US reinforcements of 1965 signify a determination to avoid defeat. They expect more US droops and probably anticipate that targets in the Hanoi-Haiphong area will come under air attack. Nevertheless, they remain unwilling to damp down the conflict or move toward negotiation. They expect a long war, but they continue to believe that time is their ally and that their own staying power is superior.

Heavier air attacks by themselves would not budge them:

The DRV would not decide to quit; PAVN infiltration southward would continue. Damage from the strikes would make it considerably more difficult to support the war in the South, but these difficulties would neither be immediate nor insurmountable.

Aerial mining would create serious problems, but NVN would keep supplies moving by resorting to shallow-draft coastal shipping and intensive efforts to keep the rail lines open. As for the South, NVN would accept the challenge:

Rather than conclude in advance that the tide of battle would turn permanently against them, the Communists would choose to boost their own commitment and to test US capabilities and will to persevere at a higher level of conflict and casualties. Thus the DRV reaction would probably be a larger program of PAVN infiltration.

The Board's picture of Hanoi was one of almost unbelievably strong commitment and dogged determination, by contrast with previous estimates. Thus, if the U.S. committed enough forces in the South to prevent NVA/VC forces from sustaining the conflict at a significant level--and the Board would not estimate how many U.S. forces were "enough"--

. . . . they might believe it necessary to make a more fundamental choice between resorting to political tactics or enlarging the war. [But] We believe that it would take a prolonged period of military discouragement to convince the DRY and the VC, persuaded as they are of their inherent advantages, that they had reached such a pass.

Even if it found itself in such straits, however, the chances were close to 50-50 that NVN would bring in Chinese forces rather than quit:

If this point were reached. . . . Prudence would seem to dictate that Hanoi . . . should choose . . . to reduce the effort in the South, perhaps negotiate, and salvage their resources for another day. We think that the chances are a little better than even that this is what they would do. But their ideological and emotional commitment, and the high political stakes involved, persuade us that there is an almost equal chance that they would do the opposite, that is, enlarge the war and bring in large numbers of Chinese forces.

The two CIA intelligence estimates of the probable consequences of the proposed escalatory measures were apparently closely held, but the available documentary evidence does not reveal how influential they may have been. Secretary McNamara's response to the JCS was merely that he was considering their recommendations "carefully" in connection with "decisions that must be taken on other related aspects of the conflict in Vietnam." He was apparently not satisfied with the estimate of reactions to the POL strikes, however, which was largely confined to an estimate of political reactions, and asked CIA for another estimate, this time related to two options: (a) attack on the storage and handling facilities at Haiphong, and (b) attack on the facilities at Haiphong together with the other bulk storage sites.

The new estimate was submitted by Richard Helms, then Acting Director of CIA, on 28 December (with the comment that it had been drafted without reference to any pause in the bombing "such as is now the subject of various speculative press articles." The estimate spelled out with greater force than before what "strains" the POL strikes might create in the North and how they might "embarrass" NVA/VC military operations in the South, and its tone was much more favorable to carrying out the strikes.

The estimate made little distinction between the two options. Haiphong was by far the most important and most sensitive of the targets and the closest to a major city; the attacks on the others were of secondary importance. Neither option was likely to bring about a change in NVN policy, either toward negotiations or toward sharply enlarging the war, but either option would substantially increase NVN's economic difficulties in the North and logistics problems in the South.

First, the estimate said, NVN would have to resort to much less efficient methods of receiving, storing and handling POL:

Destruction of the storage tanks and bulk unloading equipment at Haiphong would substantially increase the Communists' logistic problems and force them to improvise alternate POL import and distribution channels. These could include, subject to the hazards of interdiction, the use of rail or highway tankers and the transport of POL in drums by road, rail, or coastal shipping. The DRV is already increasing its use of drums because this facilitates dispersal and concealment. However, handling POL this way also requires greater expenditures of time and effort, and very large numbers of drums. Resort to these methods would necessitate transshipping through Chinese ports or transport directly across China by rail, which would in turn not only involve physical delays and difficulties but also increase the DRV's political problems in arranging for the passage of Soviet supplies through China.

This in turn would interfere with the production and distribution of goods in NVN:

The economy would suffer appreciably from the resultant disruption of transportation. This . . . would somewhat curtail the output of the DRV's modest industrial establishment and complicate the problems of internal distribution.

And make it more difficult to support the war in the South (although it would not force a reduction in such support):

The loss of stored POL and the dislocation of the distribution system would add appreciably to the DRY's difficulties in supplying the Communist forces in the South. However, we have estimated that the Communist effort in South Vietnam, at present levels of combat, does not depend on imports of POL into the South and requires only relatively small tonnages of other supplies (say 12 tons per day, on an annual basis). Accordingly, we believe that adequate quantities of supplies would continue to move by one means or another to the Communist forces in South Vietnam, though the supplies would not move as fast and it would hence require more to keep the pipeline filled. . . .

But was not likely to break Hanoi's will:

Although there presumably is a point at which one more turn of the screw would crack the enemy resistance to negotiations, past experience indicates that we are unlikely to have clear evidence when that point has been reached. . . . Though granting that each increase of pressure on the DRV bears with it the possibility that it may be decisive, we do not believe the bombing of the Haiphong facility is likely to have such an effect.

With the exception of State's INR, other intelligence agencies appeared to look with favor upon escalating the bombing. In a SNIE issued on 10 December, they agreed that intensified air attacks, beginning with POL facilities and key power plants and extending to other targets in the Hanoi/Haiphong area and mining the harbors, would not bring about any basic change in NVN policy but would in time hamper NVN's operations and set a lid on the war in the South:

We believe that Hanoi's leaders would not decide to quit and that PAVN infiltration southward would continue. Though damage from the strikes would make it considerably more difficult to support the war in South Vietnam, these difficulties would not be immediate. Over the long run, the sustained damage inflicted upon North Vietnam might impose significant limitations on the numbers of PAVN and VC main force units which could be actively supported in South Vietnam from North Vietnam.

Mining the ports, despite the dilemma created for the Soviets, would probably succeed in blocking all deep-water shipping:

The difficulty of clearing such mine fields and the ease of resowing would virtually rule out efforts to reopen the ports. The Soviets would protest vigorously and might try for some kind of action in the UN. We do not believe, however, that the Soviets would risk their ships in mined Vietnamese harbors. Peking and Hanoi would try to compensate by keeping supplies moving in shallow-draft coastal shipping and overland.

DIA, NSA, and the 3 Service intelligence even recorded a judgment that the intensified air strikes, combined with the projected build-up of U.S. ground forces in SVN to about 350,000 troops by the fall of 1966, might ultimately result in a change of heart in Hanoi. In a footnote to the SNIE they said they believed:

. . . . that as time goes on and as the impact of sustained bombing in NVN merges with the adverse effects of the other courses of action as they begin to unfold, the DRV would become clearly aware of the extent of US determination and thus might reconsider its position and seek a means to achieve a cessation of the hostilities.

INR dissented. Its Director, Thomas L. Hughes, wrote that the escalation would evoke stronger reactions than indicated in the SNIE, "because it would be widely assumed that we were initiating an effort to destroy the DRV's modest industrial establishment":

The distinction between such operations and all-out war would appear increasingly tenuous. As these attacks expanded, Hanoi would be less and less likely to soften its opposition to negotiations and at some point it would come to feel that it had little left to lose by continuing the fighting.

2. The Issue Focuses

a. POL and the Pause

Meanwhile, the flow of JCS papers urging POL strikes as the next step continued. Secretary McNamara sent the Chairman, General Wheeler, the 27 November CIA estimate which had suggested that the strikes would not have great impact on the war (they would only "embarrass" operations in the South). General Wheeler commented that the loss of POL storage would do much more:

It would, in fact, have a substantial impact not only on their military operations but also would significantly impede their efforts to support the anticipated build-up of VC/PAVN forces in South Vietnam during the coming months.

General Wheeler also forwarded a Joint Staff-DIA study of the POL target system, with the comment that destruction of the system would force NVN to curtail all but the most vital POL-powered activities and resort to "more extensive use of porters, animal transport, and nonpowered water craft." The net result would be to considerably reduce NVN's capability to move large units or quantities of equipment, an important consideration in view of the fact that motorable segments of the Ho Chi Minh trail were being extended.

The Joint Staff-DIA study showed that NVN's bulk POL storage capacity was greatly in excess of what NVN required to sustain current consumption levels- 179,000 metric tons available as compared with 32,000 metric tons needed- indicating that the strikes would have to be very damaging in order to cause NVN any major difficulties. The study also hinted that an adequate substitute system could be improvised, with lighterage from ocean tankers and dispersed storage, but it nonetheless concluded that the strikes would result in "a reduction of essential transport capabilities for military logistic and infiltration support operations," i.e., as a result of a deprivation of necessary POL.

As already noted, during the 37-day Pause, the JCS continued to recommend not only the resumption of the bombing but resumption with a dramatic sharp blow on major targets, including POL, followed by uninterrupted, increasing "pressure" bombing. They wished, in short, to turn the limited bombing program into a major strategic assault on NVN. In mid-January 1966 they sent Secretary McNamara a memo reiterating old arguments that the current ROLLING THUNDER program would not cause NVN to stop supporting the war in the South, and that the piecemeal nature of the attacks left NVN free to replenish and disperse its supplies and contend with interdictions. The way to achieve U.S. objectives, the JCS said, was to implement the bombing program they had recommended long ago, in JCSM-982-64 of 23 November 1964 which called for the rapid destruction of the entire NVN target system. In order to get the program started, the JCS recommended extending armed reconnaissance to all areas of NVN except the sanctuaries, which they would shrink (to a 10-mile radius around Hanoi and Phuc Yen airfield, a 4-mile radius around Haiphong, and a strip 20 miles along the Chinese border); lifting the sortie ceiling on armed reconnaissance; and removing "tactical restrictions" on the execution of specific strikes. The strikes would be heavy enough to deny NVN external assistance, destroy in-country resources contributing to the war, destroy all military facilities, md harass, disrupt, and impede movement into SVN.

The idea of resuming the bombing with a large and dramatic bang did not appeal much to decision-makers. Apart from the old problem of triggering an unwanted Chinese reaction, the Administration was interested in giving the lie to NVN and Chinese claims that the Pause was a cynical prelude to escalation. Although it was possible that resuming merely where the bombing left off (following as it would an extended pause and a display of great eagerness for peace) might signal too much irresolution and uncertainty, there was good reason to put off any escalatory acts for a while. As Assistant Secretary of State William Bundy wrote:

For a period of two-three weeks at least, while the world is digesting and assessing the Pause, we should do as little as possible to lend fuel to the charge--which will doubtless be the main theme of Communist propaganda--that the Pause was intended all along merely as a prelude to more drastic action.

Bundy in fact suggested resuming at a lesser level, opening with strikes below the 20th parallel, and only after a few weeks again moving northward. McNaughton

No consideration argues for a "noisy" resumption. . . The program at first should be at the level and against the kinds of targets involved prior to the Pause (only two weeks later should the program begin . . . to escalate).

He also suggested that criticism would be less if the first strikes were clearly identified with the effort to stop the southward flow of men and supplies, which had been greatly increased during the Pause.

The decisions went against ending the Pause with a bang. When the bombing was resumed on 31 January (Saigon time) it was limited "until further notice" to armed reconnaissance. No new major targets were authorized. The former sanctuary restrictions and the sortie ceilings were maintained.

It was also decided to postpone any serious escalation for the time being. Secretary McNamara informed the JCS that their proposals for rapid escalation were being considered, and on 24 January he sent the President a momorandum on the overall Vietnam program which sidestepped the issue. For 1966, the memorandum said, the bombing program against NVN should include 4000 attack sorties per month "at a minimum." It should consist of day and night armed reconnaissance against rail and road targets and POL storage sites. The present sanctuaries should be preserved. There should be more intense bombing of targets in Laos, along the Bassac and Mekong Rivers running into SVN from Cambodia, and better surveillance of the sea approaches.

The use of interdiction rather than pressure terms in the Presidential memorandum, and the emphasis on bombing infiltration routes into SVN, rather than the flow of supplies into or within NVN, indicates that the Secretary was still interested in keeping the objectives of the bombing limited and any escalation in check. The memorandum said that the bombing had already achieved the objective of raising the cost of infiltration, and was reducing the amount of enemy supplies reaching the South. In NVN it had also diverted manpower to air defense and repair work, interfered with mobility, and forced the decentralization of many activities. It could further reduce the flow of supplies to NVA/VC forces in the South, and limit their "flexibility" to defend themselves adequately or undertake frequent offensive action, but it was doubtful that even heavier bombing would put a "tight ceiling" on the NVN effort in the South.

Despite the application of the brake on ROLLING THUNDER operations, the debate over escalation wore on. Further proposals were made and further studies and reviews were requested. DIA was asked to conduct a special analysis of the NVN POL system. The study said that the exceptionally high ratio of storage capacity to consumption allowed the system to "absorb a high degree of degradation," and noted that the dispersed sites in the system were "relatively invulnerable," but concluded nonetheless that (a) the loss of storage at Haiphong would be "critical to the entire bulk distribution system" and would require either a "modification" in the handling of marine imports or a switch to importation by rail or truck through China, and (b) the loss of the other facilities would produce local POL shortages and transportation bottlenecks until substitutes and alternatives could be devised.

b. The February Debate

In February a SNIE was published, estimating how NVN's physical capabilities (not its will) to support the war in the South would be affected by increasing the scope and intensity of ROLLING THUNDER. The enlarged program which the estimate considered included attacks to destroy all known POL facilities, destroy all large military facilities except airfields and SAM sites (unless they seriously interfered with our operations), interdict the land LOCs from China, (a) with or (b) without closing the ports, put and keep electric power plants out of action, and restrict the use of LOCs throughout NVN but especially south of Hanoi.

The SNIE concluded that although the increased bombing might set a limit somewhere on the expansion of NVA/VC forces and their operations in SVN, it would not prevent their support at substantially higher levels than in 1965. The destruction of electric power facilities would practically "paralyze" NVN's industry, but

. . . . because so little of what is sent south is produced in the DRV, an industrial shutdown would not very seriously reduce the regime's capability to support the insurgency.

Destruction of POL storage facilities would force NVN to almost complete dependence on current imports, but N\TN could manage. Destruction of military facilities would mean the loss of some stockpiled munitions, "although most such storage is now well dispersed and concealed." Closing the ports and interdicting the LOCs from China would reduce the level of imports--leaving the ports open would not--but NVN could continue to bring in enough supplies that were critical to the survival of the regime and essential military tasks, including the "small quantities" necessary for transshipment to SVN.

Importation of POL would be a key problem, but would be surmountable in a comparatively short time, probably a few weeks, since quantities involved would not be large, even if increased somewhat over previous levels. Soviet POL could be unloaded from tankers at Chan-chiang in South China, moved thence by rail to the DRV border and from there to the Hanoi area by truck. It could also move from the USSR by rail directly across China, or down the coast from Chan-chiang in shallow-draft shipping.

Restricting the LOCs south of the Hanoi region would create logistical problems for NVN military forces in Military Region IV south of the 20th parallel, but would not stop the relatively small amounts of material forwarded to SVN.

The cumulative effect of the proposed bombing program would make life difficult for NVN, therefore, but it would not force it to curtail the war in the South:

The combined impact of destroying in-country stockpiles, restricting import capabilities, and attacking the southward LOCs would greatly complicate the DRV war effort. The cumulative drain on material resources and human energy would be severe. The postulated bombing and interdiction campaign would harass, disrupt, and impede the movement of men and material into South Vietnam and impose great overall difficulty on the DRV. However, we believe that, with a determined effort, the DRV could still move substantially greater amounts than in 1965.

The bombing program would not prevent NVN from further expanding NVA/ VC forces in the South at the projected reinforcement rate of 4500 men per month and from further providing them with heavier weapons, but it might set some limit on their size and their operations:

. . . . an attempt by the Communists to increase their strength . . . to intensify hostilities . . .or . . . to meet expanded US/GVN offensive operations . . . will use up supplies at a higher rate . . . [This] might raise supply requirements to a level beyond the practical ceiling imposed on their logistic capabilities by the bombing campaign. . . . There are, however, too many uncertainties to permit estimating at just what level the limit on expansion would be.

Also in February, Secretary McNamara asked the JCS to develop an optimum air interdiction program "to reduce to the maximum extent the support in men and materiel being provided by North Vietnam to the Viet Cong and PAVN forces in South Vietnam." The study, forwarded to the Secretary on 14 April, managed to frame an interdiction program which embraced virtually everything the JCS had been recommending. It pointed out that less than half of the JCS targets, "the most critical to North Vietnam's support of the insurgency, military capabilities, and industrial output," had been hit, "due to self-imposed restraints":

These restraints have caused a piecemealing of air operations which has allowed the enemy a latitude of freedom to select and use methods that significantly increase his combat effectiveness. It has permitted him to receive war supporting materiel from external sources through routes of ingress which for the most part have been immune from attack and then to disperse and store this materiel in politically assured sanctuaries. From these sanctuaries the enemy then infiltrates this materiel to SVN/Laos. . . . Throughout the entire movement, maximum use is made of villages and towns as sanctuaries. These and the Hanoi, Haiphong, and China border buffer areas cloak and protect his forces and materiel, provide him a military training and staging area free from attack, and permit him to mass his air defense weapons.

. . . The less than optimum air campaign, and the relatively unmolested receipt of supplies from Russia, China, satellite countries, and certain elements of the Free World have undoubtedly contributed to Hanoi's belief in ultimate victory. Therefore, it is essential that an intensified air campaign be promptly initiated against specific target systems critical to North Vietnam's capability for continued aggression and support of insurgency.

The study went on to outline an intensified bombing campaign to cause NVN to stop supporting the insurgency in the South

by making it difficult and costly for North Vietnam to continue effective support of the NVN/VC forces in South Vietnam and to impose progressively increasing penalties on NVN for continuing to support insurgency in Southeast Asia.

Its language left no doubt that while the strikes were intended "to restrict NVN capability to support and conduct armed aggression in SEAsia," the ultimate purpose was to apply pressure against Hanoi's will:

The strategy of this plan requires initial application of air attacks over a widespread area against the NVN military base structure and war supporting resources. The intensity of air operations and the number of targets to be attacked gradually increase. Under such pressure of attack, NVN must
further disperse or face destruction in depth of its military base and resources. The dispersal will increase the stresses on command, control, and logistic support and should cause some concern in the Military Command of the wisdom of further aggression. . . . The combined effects of reducing and restricting external assistance to NVN, the progressive attacks against NVN military and war supporting resources, the interdiction of infiltration routes in NVN and Laos, and the destruction of NVN/VC forces and bases in SVN and Laos should cause a reappraisal in Hanoi as to NVN's military capability to continue aggression.

The plan, which was merely "noted" and not red-striped by the JCS, called for the "controlled and phased intensification of air strikes" and a "modest adjustment" in the sanctuaries (to 10 miles around Hanoi, 4 around Haiphong, and 20 from the Chinese border, as previously recommended by the JCS). A first phase extended armed reconnaissance to the northeast, and struck 11 more JCSlisted bridges, the Thai Nguyen railroad yards and shops, 14 headquarters/barracks, 4 ammunition and 2 supply depots, 5 POL storage areas, 1 airfield, 2 naval bases, and 1 radar site, all outside the (reduced) sanctuaries. The second phase attacked 12 "military and war supporting installations" within the Hanoi and Haiphong sanctuaries: 2 bridges, 3 POL storage areas, 2 railroad shops and yards, 3 supply depots, 1 machine tool plant, and 1 airfield. The third phase attacked the 43 remaining JCS targets, including 6 bridges, 7 ports and naval bases, 6 industrial plants, 7 locks, 10 power plants, the NVN ministries of national and air defense, and assorted railroad, supply, radio, and transformer stations.

The plan also provided for three special attack options for execution during any of the phases "as a counter to enemy moves or when strong political and military action is desired." The options were: attack on the POL center at Haiphong; aerial mining of the channel approaches to Haiphong, Hon Gal, and Cam Pha, the three principal maritime ports; and strikes against the major jet airfields at Hanoi, Haiphong, and Phuc Yen.

The JCS were apparently not in complete sympathy with the gradual phasing of stronger attacks over several months, as proposed in the study. In their formal memoranda to the SecDef they continued to restate their mid-January recommendations for the sharp blows with maximum shock effect as "the soundest program from a military standpoint" which offered "the greatest return for the air effort expended." Apparently sensing that this was more than the traffic would bear, however, they began to push for early strikes against POL as "one of the highest priority actions not yet approved." They pointed out that NVN was busily expanding and improving its LOCs, and its "offensive and defensive" air capabilities; it was expediting its import of trucks. POL was becoming increasingly significant to NVN's war effort, and its destruction would have an "immediate effect on the military movement of war supporting materials."

c. The CIA Recommends Escalation

While the JCS kept up its barrage of recommendations during March, CIA broke into the debate with an apparently very influential report on the past accomplishments and future prospects of the bombing. The report virtually wrote off the bombing results to date as insignificant, in terms of either interdiction or pressure; blamed "the highly restrictive ground rules" under which the program operated; and took the bold step, for an intelligence document, of explicitly recommending a preferred bombing program of greater intensity, redirected largely against "the will of the regime as a target system."

The report held that the economic and military damage sustained by NVN had been moderate and the cost had been passed along to the USSR and China. The major effect of the bombing had been to disrupt normal activity, particularly in transportation and distribution, but with considerable external help the regime had been singularly successful in overcoming any serious problems. It had been able to strengthen its defenses, keep its economy going, and increase the flow of men and supplies South. Most of the direct damage so far had been to facilities which NVN did not need to sustain the military effort, and which the regime merely did without. It had been able to maintain the overall performance of the transportation system at the levels of 1964 or better. It had increased the capacity of the LOCs to the South and made them less vulnerable to air attack by increasing the number of routes and bypasses. Despite the bombing, truck movement through Laos, with larger vehicles and heavier loads, had doubled.

The program had not been able to accomplish more because it had been handicapped by severe operational restrictions:

Self-imposed restrictions have limited both the choice of targets and the areas to be bombed. Consequently, almost 80 percent of North Vietnam's limited modern, industrial economy, 75 percent of the nation's population and the most lucrative military supply and LOC targets have been effectively insulated from air attack. Moreover, the authorizations for each of the ROLLING THUNDER programs often have imposed additional restrictions, such as limiting the number of strikes against approved fixed targets. The policy decision to avoid suburban casualties to the extent possible has proved to be a major constraint.

The overall effect of those area and operational restrictions has been to grant a large measure of immunity to the military, political, and economic assets used in Hanoi's support of the war in the South and to insure an ample flow of military supplies from North Vietnam's allies. Among North Vietnam's target systems, not one has been attacked either intensively or extensively enough to provide a critical reduction in national capacity. No target system can be reduced to its critical point under existing rules.

Moreover, the bombing had been too light, fragmented, and slowly paced:

The ROLLING THUNDER program has spread bomb tonnage over a great variety of military and economic targets systems, but the unattacked targets of any one system have consistently left more than adequate capacity to meet all essential requirements. Furthermore, the attacks on major targets have often been phased over such long periods of time that adequate readjustment to meet the disruption could be accomplished.

What was required was a basic reorientation of the program:

Fundamental changes must be made if the effectiveness of the campaign is to be raised significantly. First, the constraints upon the air attack must be reduced. Secondly, target selection must be placed on a more rational basis militarily.

Putting the program on a "more rational" military basis apparently involved abandoning interdiction as a primary goal. The report held out little promise that any acceptable bombing program could physically interfere with the flow of supplies to the South. The NVN economy, it stated, was not "an indigenous economic base heavily committed to the support of military operations in the South," but rather a "logistic funnel" through which supplies from the USSR and China flowed. As such, it was a hard target, easy to maintain in operation and quite large for the load. This was particularly the case in the lower half of the "funnel," where the bombing had been concentrated:

. . . . the rudimentary nature of the logistic targets in the southern part of North Vietnam, the small volume of traffic moving over them in relation to route capacities, the relative ease and speed with which they are repaired, the extremely high frequency with which they would have to be restruck--once every three days--all combine to make the logistic network in this region a relatively unattractive target system, except as a supplement to a larger program. A significant lesson from the ROLLING THUNDER program to date is that the goals of sustained interdictions of the rudimentary road and trail networks in southern North Vietnam and Laos will be extremely difficult and probably impossible to obtain in 1966, given the conventional ordnance and strike capabilities likely to exist.

The upper half of the "funnel" was a much more lucrative target--not, however, because attacking it would choke the volume of supplies flowing into the South, but because it would inflict more pain on the regime in the North.

The flow of military logistics supplies from the USSR and China cannot be cut off, but the movement could be made considerably more expensive and unreliable if authorization is granted to attack intensively the rail connections to Communist China and if the three major ports are effectively mined. About 2/3 of North Vietnam's imports are carried by sea transport and the remainder move principally over the rail connections from Communist China. Mining the entrances to the three major ports would effectively transfer all imports to rail transport, including the flow of imports needed to maintain economic activity. The rail connections to Communist China would then become a more lucrative target and the disruptive effect of interdiction would then be more immediately felt. Sustained interdiction would then force Hanoi to allocate considerable amounts of manpower and materials to maintain the line.

Bombing the supplies and supply facilities at the top of the "funnel" was therefore a "preferred LOC target system." It was not advanced as an interdiction measure, however, but as a means of increasing the penalty to Hanoi (and its allies), in terms of economic, social, and political consequences, of supporting the war in the South, and thus presumably to reduce the desire to continue it. Other targets which might be attacked in order to similarly influence the will of the regime were: 26 military barracks and/or supply facilities on the JCS list, the neutralization of which would "impede the flow of military supplies and disrupt the military training programs of NVN"; 8 major POL storage facilities, which had a "direct bearing" on the regime's ability to support the war in the South, but which had to be hit almost simultaneously in order to reduce NVN to the critical point in meeting essential requirements; the Haiphong cement plant, the loss of which would "create a major impediment to reconstruction and repair programs" until cement could be imported; 3 major and 11 minor industrial plants which, though they made "no direct or significant contribution to the war effort" and "only a limited contribution" to the economy, were "highly prized and nominally lucrative" targets; or, as an alternative method of knocking out industrial production, the main electric power facilities.

As for other potential targets in NVN--the command and control system, agriculture, and manpower--

Attacks on these targets are not recommended at this time. In each case the effects are debatable and are likely to provoke hostile reactions in world capitals.

The March CIA report, with its obvious bid to turn ROLLING THUNDER into a punitive bombing campaign and its nearly obvious promise of real payoff, strengthened JCS proposals to intensify the bombing. In particular, however, the report gave a substantial boost to the proposal to hit the POL targets. The POL system appeared to be the one target system in NVN to which, what the report called, "the principle of concentration" might be applied; that is, in which enough of the system could be brought under simultaneous attack to cut through any cushion of excess capacity, and in which a concentrated attack might be able to overwhelm the other side's ability to reconstruct, repair, or disperse its capacity.

The POL targets had other qualities to commend them as the next escalatory step in ROLLING THUNDER. They really were pressure targets, but they could be plausibly sold as interdiction targets. The main ones were in the Hanoi/Haiphong sanctuaries, so that over and above any economic or military impact, strikes against them would signal that the last sanctuaries were going and the industrial and other targets there were now at risk. They fit the image of "war-supporting" facilities which strategic bombing doctrine and ample military precedent had decreed to be fair game in bringing a war machine to a standstill. They had, in fact, been struck before in other parts of NVN without any unusual political repercussions. They were situated in the arbitrarily-defined urban/industrial centers, but somewhat set apart from the densest civilian housing areas, and thus might not entail as many civilian casualties as other targets in those areas.

Moreover, even if the impact of POL strikes would be within NVN itself--because NVN supplied no POL at all to NVN/VC forces in the South and used next to none in transporting other goods there--POL was at least relevant as an interdiction target. It did power trucks and boats which were involved in carrying men and supplies South. If any truck in the NVN fleet was an acceptable interdiction target, wherever it was and whatever its cargo, why not any POL?

d. McNamara Endorses POL, the President Defers It

Resumption of ROLLING THUNDER, as initiation of the pause, did not, of course, constitute a final decision on escalation. The views of CINCPAC and the JCS remained unaltered, and Secretary McNamara stood committed, unless he reversed himself, to enlarging the area and intensity of interdiction bombing and to destroying North Vietnamese POL. Neither in OSD nor the White House had anyone opposed these measures on other than prudential grounds--the risk of alienating allies or provoking Chinese or Russian intervention or uncertainty that results would justify either the risks or the costs. Everyone seemed agreed that, were it not for these factors, intensified bombing of the North would help to accomplish American objectives. Nevertheless, the position of the decision-makers can best be characterized as hesitant.

The services naturally undertook to tip the balance toward the rapid and extensive escalation they had all along advocated. To McNamara's memorandum to the President, the JCS had attached a dissent. They felt that the Secretary underrated the "cumulative effect of our air campaign against the DRV on morale and DRV capabilities" and overestimated the "constancy of will of the Hanoi leaders to continue a struggle which they realize they cannot win in the face of progressively greater destruction of their country."

When McNamara reported to the Chairman the President's ruling on ROLLING THUNDER, he apparently spoke of the difficulty of making out a convincing case that air operations against North Vietnam could seriously affect PAVN/VC operations in the South. In any event, following a conversation with the Secretary, General Wheeler ordered formation of a special study group to devise a bombing effort "redirected for optimum military effect." He explained, "the primary objective should be to reduce to the maximum extent the support in men and materiel being provided by North Viet-Nam to the Viet Cong and PAVN forces in South Viet-Nam." Headed by a Brigadier General from SAC, composed of five Air Force, three Navy, two Army, and one Marine Corps officers, and making extensive use of CINCPAC assistance, this study group went to work in early February, with an assignment to produce at least an interim report by 1 March and a final report no later than 1 August.

Meanwhile, routine continued, with CINCPAC recommending programs thirteen days prior to the beginning of a month and the JCS acting on these recommendations two days later. In consequence, McNamara received from the Chiefs on 19 February the same advice that had been given during the pause. He and the President responded much as before, though now permitting armed reconnaissance within the geographical limits fixed just before the pause and authorizing a significant increase--to above 5,000--in numbers of sorties.

On 1 March, when this slightly enlarged campaign opened, the Chiefs filed a memorandum stressing the special importance of an early attack on North Vietnamese POL. They had singled out POL somewhat earlier, writing McNamara in November, 1965, that attack on this target "would be more damaging to the DRV capability to move war-supporting resources within country and along infiltration routes to SVN than an attack against any other single target system." While causing relatively little damage to the civilian economy, it would, they reasoned, force a sharp reduction in truck and other road traffic carrying men and suppies southward. They held also that the attack should be made soon, before North Vietnam succeeded in improving air defenses and in dispersing POL storage.

McNamara had rejected this recommendation, not only because of the planned pause but also because CIA sources questioned some of the Chiefs' reasoning and stressed counterarguments which they tended to minimize. Assessing the probable results of not only taking out North Vietnamese POL, but also mining harbors and bombing military and industrial targets in the northeast quadrant, the Board of National Estimates said, "Damage from the strikes would make it considerably more difficult to support the war in the South, but these difficulties would neither be immediate nor insurmountable." With regard to the POL system alone, the Board observed, "It is unlikely that this loss would cripple the Communist military operations in the South, though it would certainly embarrass them." Pointing out that the bulk of storage facilities stood near Haiphong and Hanoi, the Board went on to say that "the Communists would unquestionably regard the proposed U.S. attacks as opening a new stage in the war, and as a signal of U.S. intention to escalate the scale of conflict." This appraisal did not encourage adoption of the JCS recommendation.

The Chiefs continued nevertheless to press for a favorable decision. Before and during the pause, they presented fresh memoranda to McNamara. A more detailed CIA study, obtained just after Christmas, provided somewhat more backing for their view. It conceded that the Communists were dispersing POL facilities and that an early attack on those at Hanoi and Haiphong "would add appreciably to the DRV's difficulties in supplying the Communist forces in the South." Nevertheless, it forecast that "adequate quantities of supplies would continue to move by one means or another to the Communist forces in South Vietnam."

In mid-January, the DIA prepared an estimate considerably more favorable to the scheme. But in early February appeared a SNIE estimating effects on "DRV physical capabilities to support the insurgency in the South" of the various measures, including attacks on POL, previously recommended by CINCPAC and the JCS. Its conclusion, subscribed to by all intelligence services except that of the Air Force, was that, even with a campaign extended to port facilities, power plants, and land LOC's from China, "with a determined effort, the DRV could still move substantially greater amounts than in 1965."

In renewing their recommendation on 1 March, and again on 10 March, the JCS once more disputed such assessments. In an appendix to their long March 1 memorandum to the Secretary, the Chiefs outlined a concept of operations upon which they proposed to base future deployments. With respect to the air war, they urged that it be expanded to include POL and the aerial mining of ports and attacks on Hanoi and Haiphong. Their rationale was as follows:

To cause . . . NVN to cease its control, direction, and support of the communist insurgency in SVN and Laos, air strikes are conducted against military and war-sustaining targets in all areas, including the Hanoi/ Haiphong complex and areas to the north and northeast. Armed reconnaissance within NVN and its coastal waters is conducted to interdict LOCs, harass, destroy and disrupt military operations and the movement of men and materials from NVN into Laos and SVN. Aerial mining of ports and interdiction of inland waterways and coastal waters, harbors and water LOCs are conducted to reduce the flow of war resources. Air reconnaissance and special air operations are conducted in support of the overall effort.

Ten days later the Chiefs again requested attacks on the POL together with authorization to mine the approaches to Haiphong. This time they noted that Ambassador Lodge and Admiral Sharp had each recently endorsed such measures (no documents so indicating are available to the writer). Supporting their request they cited recent intelligence reports of North Vietnamese orders for expedited delivery of additional trucks. With the arrival of more trucks, POL would become even more critical to the North Vietnamese logistical effort. Once POL reserves were initially destroyed, however, the mining of Haiphong harbor would be the next immediate priority to prevent resupply by North Vietnam's allies. The Chiefs argued that the elimination as a package of these high value targets would significantly damage the DRV's war-sustaining capability.

This time, moreover, the Chiefs possessed support in the intelligence community. A study by CIA addressed the question which had been deliberately omitted from the terms of reference for the 4 February SNIE, i.e., what effect bombing might produce on the will of the North Vietnamese regime. Judging from a summary with some extracts, preserved in Task Force files, it made a strong case for almost unlimited bombing such as CINCPAC and the JCS had steadily advocated. It accepted previous judgments that "the goals of sustained interdictions of the rudimentary road and trail networks in southern North Vietnam and Laos will be extremely difficult and probably impossible to obtain in 1966, given the conventional ordnance and strike capabilities likely to exist." Though arguing that more payoff could result from regarding North Vietnam as a "logistic funnel" and attempting to stop what went into it rather than what came out, it conceded that the "flow of military logistics supplies from the USSR and China cannot be cut off." But the report contended that such measures as mining harbors, maintaining steady pressure on LOC's with China, and destroying militarily insignificant but "highly prized" industrial plants would not only reduce North Vietnam's capacity to support the insurgency in the South but would influence her leaders' willingness to continue doing so. "Fundamental changes must be made if the effectiveness of the campaign is to be raised significantly," said the report. "First, the constraints upon the air attack must be reduced. Secondly, target selection must be placed on a more rational basis militarily." One point stressed was the importance of taking out all remaining POL storage facilities simultaneously and at an early date.

With memoranda from the JCS now reinforced by this CIA report, Secretary McNamara had to reconsider the POL issue. Conferring with Wheeler on 23 March, he put several specific questions, among them whether destruction of POL storage facilities would produce significant results if not coupled with mining of North Vietnamese ports, what exact targets were to be hit, and with how many sorties. Responding with the requested details, the Chiefs said that they attached the highest importance to the operation, even if enemy harbors remained open. They strongly recommended, in addition, attacks on adjoining industrial targets and LOC's, in order to enhance the effect of destroying POL facilities.

In a memorandum for the President on bombing operations for April, McNamara endorsed most of these JCS recommendations. He proposed authorizing attacks on seven of the nine POL storage facilities in the Hanoi-Haiphong area. Of the two he omitted, one lay near the center of Hanoi. In addition, McNamara recommended attacks on the Haiphong cement plant and on roads, bridges, and railroads connecting Haiphong and Hanoi and leading from the two cities to the Chinese border, and asked that the military commanders be permitted to run up to 900 sorties into the northeast quadrant, at their discretion.

For this marked stepping-up of the air war, McNamara put on paper a much more forceful presentation than that in his January memorandum. Using as a point of departure the general estimate that bombing could neither interdict supply of the South nor halt flow from China and Russia into the North, he argued that:

. . . .The movement can be made considerably more expansive and unreliable (a) by taking action to overload the roads and railroads (e.g., by destroying the domestic source of cement), (b) by attacking the key roads, railroads and bridge between Hanoi on the one hand and Haiphong and China on the other, and (c) by pinching the supply of POL, which is critical to ground movement and air operations.

Amplifying one of these recommendations, McNamara commented that destruction of the plant, which produced 50% of North Vietnam's cement, would make bridge and road rebuilding difficult. As for POL, he observed that the facilities targeted represented 70-80% of those in the country. Though the North Vietnamese possessed reserves and had probably already built up some in the South, their transportation system depended on a continuous supply. They were known to have recently doubled their orders for imported Soviet POL. Eventually, though not necessarily in the short run, he said, they were bound to suffer a shortage.

While McNamara conceded that he did not expect the proposed program to yield quick results in South Vietnam, he predicted that it would ultimately have some effect. Addressing some political issues that had influenced the previous hesitancy, he asserted that the South would probably do nothing more than adopt "a somewhat harsher diplomatic and propaganda line" and that the Chinese "would not react to these attacks by active entry-by ground or air," unless the United States took further steps, the decisions on which "at each point would be largely within our own control." And offsetting such risks stood the possibility of favorable political effects. McNamara ventured no promises. He said, "We would not expect Hanoi to change its basic policy until and unless it concluded that its chances of winning the fight in the South had become so slim that they could no longer justify the damage being inflicted upon the North." Nevertheless, he commented that destruction of POL facilities "should cause concern in Hanoi about their ability to support troops in South Vietnam" and concluded his memorandum by writing:

In the longer term, the recommended bombing program . . . . can be expected to create a substantial added burden on North Vietnam's manpower supply for defense and logistics tasks and to engender popular alienation from the regions should shortages become widespread. While we do not predict that the regime's control would be appreciably weakened, there might eventually be an aggravation of any differences which may exist within the regime as [to] the policies to be followed.

Reading this memorandum, one might conclude that the Secretary, after passing through a season of uncertainty, had finally made up his mind-that he now felt the right action to be sharp escalation such as CINCPAC, the JCS, and McNaughton had advocated during the pause. But even now, despite the comparatively vigorous language of the memorandum, one cannot be sure that McNamara expected or wanted the President to approve his recommendations.

The memorandum was probably brought up at the White House Tuesday luncheon on 28 March. Just sixteen days earlier, in response to Marshal Ky's removal of General Nguyen Chanh Thi from Command of the I Corps Area, Buddhist monks had initiated anti-Ky demonstrations in DaNang and Hue. Soon, with other groups joining in, dissidents dominated the northern and central part of the country. Many not only attacked the Ky regime but denounced the American presence in Vietnam and called for negotiation with the NLF. Controlling the Hue radio and having easy access to foreign newsmen, these dissidents won wide publicity in the United States. As a result, Americans previously counted as supporters of administration policy began to ask why the United States should expend its resources on people who apparently did not want or appreciate help. Such questioning was heard from both Democrats and Republicans in Congress. Quite probably, the political situation in Vietnam and its repercussions in America stood uppermost in the President's mind. Equally probably, McNamara recognized this fact. If so, it should not have surprised him to find the President taking much the same position as that which they had both taken, and recorded in NSAM 288 in March, 1964, when the Khanh government trembled--that it was imprudent to mount new offensives "from an extremely weak base which might at any moment collapse and leave the posture of political confrontation worsened rather than improved."

In any event, the principal outcome of White House meetings at the end of March was a string of urgent cables from Rusk to Lodge, suggesting steps which might be urged on the Ky government and saying, among other things,

. . . . We are deeply distressed by the seeming unwillingness or inability of the South Vietnamese to put aside their lesser quarrels in the interest of meeting the threat from the Viet Cong. Unless that succeeds, they will have no country to quarrel about. . . . We face the fact that we ourselves cannot succeed except in support of the South Vietnamese. Unless they are able to mobilize reasonable solidarity, the prospects are very grim.

As for McNamara's proposals, the President approved only giving commanders discretion to launch 900 sorties into the northeast quadrant during April and permission to stroke roads, railroads, and bridges outside or just on the fringe of the prohibited circles around Hanoi and Haiphong. He did not consent to measures involving more visible escalation of the air war. McNamara returned to the Pentagon to inform the Chiefs that, while these operations had not been vetoed, they were not yet authorized.

The President had authorized the extension of armed reconnaissance into the northeast quadrant and strikes on 4 of the 5 bridges recommended by McNamara but deferred any decision on the crucial portion, the strikes against the 5th bridge, the cement plant, the radar, and above all the 7 POL targets. The JCS execution message for ROLLING THUNDER 50, which was sent out on 1 April, directed implementation of what had been approved. In addition, it ordered CINCPAC to "plan for and be prepared to execute when directed attacks during April" against the 5th bridge, the cement plant, the radar, and the 7 POL sites. A pencilled notation by Secretary McNamara with reference to these targets also mentions April: "Defer . . . until specifically authorized but develop specific plans to carry out in April."

3. April and May--Delay and Deliberation

a. Reasons to Wait

Although the President's reasons for postponing the POL decision are not known, and although the initial postponement seemed short, a matter of weeks, it is evident from the indirect evidence available that the proposal to strike the POL targets ran into stiffening opposition within the Administration, presumably at State but perhaps in other quarters as well. Before the question was settled it had assumed the proportions of a strategic issue, fraught with military danger and political risk, requiring thorough examination and careful appraisal, difficult to come to grips with and hotly contended. The question remained on the agenda of senior officials for close to three months, repeatedly brought up for discussion and repeatedly set aside inconclusively. Before it was resolved a crisis atmosphere was generated, requiring the continuing personal attention of all the principals.

There can be little doubt that the POL proposal instigated a major policy dispute. The explanation seems to be two-fold. One, those who saw the bombing program, whatever its merits, as seriously risking war with China or the USSR, decided to seize the occasion as perhaps the last occasion to establish a firebreak against expanding the bombing to the "flash points." Two, those who saw the bombing program as incurring severe political penalties saw this as the last position up to which those penalties were acceptable and beyond which they were not. Both points no doubt merged into a single position. Both turned the POL question into an argument over breaching the Hanoi/Haiphong sanctuaries in any major way.

McNamara's Memorandum for the President, which had treated the POL strikes as a logical extension of the previous interdiction program into an area in which it might be more remunerative, did not address these questions of sanctuaries. No other single document has been located in the available files which does. Pieced together and deduced from the fragmentary evidence, however, it appears that the view that POL strikes ran too great a risk of counter-escalation involved several propositions. One was that the strikes might trigger a tit-for-tat reprisal (presumably by the VC) against the vulnerable POL stores near Saigon. The Secretary of Defense had himself made this point as early as mid-1965 in holding off Congressional and other proponents of Hanoi/Haiphong area POL strikes, citing the endorsement of General Westmoreland. The JCS had recognized the possibility in their November 1965 paper on POL strikes, although they considered it "of relatively small potential consequence, minor in comparison to the value of destruction of the DRV POL system." General Wheeler had also gone out of his way to allude to it. Under Secretary of State Ball, in a January 1966 memorandum, saw the possibility of an enemy reprisal in SVN as only the first act of a measure-countermeasure scenario which could go spiralling out of control: a VC reprisal against POL and SVN would put unbearable pressure on the U.S. to counter-retaliate against the North in some dangerous manner, which in turn would force the other side to react to that, and so on.

More important than the fear of a VC reprisal, one assumes, was the belief that the POL sites were the first of the "vital" targets, high-value per se but also generally co-located with and fronting for NVN's other high-value targets. NVN, with its "vital" targets attacked and its economic life at stake, would at a minimum defend itself strenuously (again, provoking us to attack its airfields in our defense, which in turn might set off an escalatory sequence); or, at the other extreme, NVN might throw caution to the winds and call on its allies to intervene. This might be only a limited intervention at first, e.g. use of Chinese fighters from Chinese bases to protect NVN targets, but even this could go escalating upward into a full-scale collision with China. On the other hand, the strikes at the "vital" targets might be the Southeast Asian equivalent of the march to the Yalu, convince the other side that the U.S. was embarked on a course intolerable to its own interests, such as the obliteration of the NVN regime, and cause it to intervene directly.

These arguments were not new, of course; they were arguments which could be, and no doubt were, used against any bombing at all. They gained force, however, as the bombing became more intense and the more the bombing was thought to really hurt Hanoi. (It was an irony of the original concept of the air war North that the more pressure it really applied and hence the more successful it was, the more difficult it was to prosecute.)

The belief that POL strikes would overload the negative side of the scale on political grounds had to do with the possibility that, since the targets were situated in relatively populated "urban" areas (even though outside of the center cities), the strikes would be construed as no less than the beginning of an attack on civilian targets and/or population centers. This possibility, too, could widen the war if it were taken by NVN and its allies as indicating a U.S. decision to commence "all-out" bombing aimed at an "unlimited" objective. But even if it did not widen the war, it could cause a storm of protest world-wide and turn even our friends against us. The world had been told repeatedly that the U.S. sought a peaceful settlement, not a total military victory; that the U.S. objectives were limited to safeguarding SVN; that bombing NVN was confined to legitimate military targets related to the aggression against SVN; and that great care was taken to avoid civilian casualties. Any or all of this could be called into question by the POL strikes, according to the argument, and the U.S. could be portrayed as embarking on a course of ruthless brutality against a poor defenseless population.

The argument about the escalatory implications of the proposed POL strikes was difficult to deal with. Official intelligence estimates were available which said, on balance, that Chinese or Soviet intervention in the war was unlikely, but no estimate could say that such intervention was positively out of the question, and of course intelligence estimates could misjudge the threshold of intervention, it was said, as they had in Korea.

The argument about the political repercussions made some headway, however. Progress became possible because of the development of military plans to execute the strikes with "surgical" precision, thus minimizing the risk of civilian casualties, and because of the development of a "scenario" for the strikes in which military, diplomatic, and public affairs factors were coordinated in an effort to contain adverse reactions. There slowly unfolded a remarkable exercise in "crisis management."

b. The April Policy Review

Though McNamara's memorandum, and the President's indication that he might later approve POL, brought the Administration somewhat nearer to a decision for escalation, there was as yet no new consensus on how the air war against the North might be tailored to serve American objectives or, indeed, on what those objectives were or ought to be. The study group in the Joint Staff, completing its work early in April, offered a straightforward answer: "The overall objective is to cause NVN to cease supporting, directing, and controlling the insurgencies in South Vietnam and Laos." With his understanding, they could recommend a three phase campaign leading to destruction of between 90 and 100% of all POL storage, bridges, airfields, rail facilities, power plants, communications, port structures, and industry in North Vietnam. Whether the Chiefs reasoned similarly is not apparent from the papers available. Although they came out with comparable recommendations, they merely "noted" this study.

Certainly, in spite of McNamara's memorandum recommending escalation, no clear view prevailed within OSD or among civilians elsewhere in the government occupied with Vietnam policy. Among the papers left behind by McNaughton are some fragments relating to an attempt early in April, 1966, to rethink the question of what the United States sought in Vietnam. These fragments suggest an evolution between winter, 1965-66, and spring, 1966, from hesitancy to perplexity.

The political situation in South Vietnam became increasingly explosive. On March 31, 10,000 Buddhists had demonstrated in Saigon against the government and the demonstrations had spread to other cities in the next several days. On April 5, Premier Ky flew to Danang to quell the rebellion and threatened to use troops if necessary. In this context, a meeting was convened at the White House on Friday, 9 April. Vance and McNaughton represented Defense; Ball, Bundy, and Leonard Unger the State Department; and George Carver the CIA. Walt Rostow, who had just replaced McGeorge Bundy, took part. So did Robert Komer and Bill Moyers.

In preparation for this meeting, McNaughton, Ball, Unger, and Carver undertook to prepare memoranda outlining the broad alternatives open. Carver would make the case for continuing as is, Unger and McNaughton for continuing but pressing for a compromise settlement-Unger to take an optimistic and McNaughton a pessimistic view and Ball to argue for disengagement. Then four options were labelled respectively, A, B-O, B-P, and C.

Carver, advocating Option A, wrote:


I. Description of the Course of Action

1. Option A involves essentially persevering in our present policies and programs, adhering to the objectives of

a. Preventing a North Vietnamese takeover of South Vietnam by insurrectionary warfare, thus

(1) Checking Communist expansion in Southeast Asia
(2) Demonstrating U.S. ability to provide support which will enable indigenous non-Communist elements to cope with "wars of national liberation" and, hence,
(3) Demonstrating the sterile futility of the militant and aggressive expansionist policy advocated by the present rulers of Communist China.

b. Aiding the development of a non-Communist political structure within South Vietnam capable of extending its writ over most of the country and acquiring sufficient internal strength and self-generated momentum to be able to survive without the support of U.S. combat forces whenever North Vietnam ceases its present campaign of intensive military pressure.

To adopt this option, Carver reasoned, required, on the political side, work with all non-Communist Vietnamese factions "to insure that the transition to civilian rule is as orderly as possible and effected with a minimum disruption of current programs." The United States would have to make plain in Saigon that continued support was "contingent upon some modicum of responsible political behavior" and would have to "initiate the Vietnamese in the techniques of developing political institutions such as constitutions and parties." An "intensive endeavor at provincial and district levels" would have to complement efforts in the capital.

On the military side, Carver judged the demands of Option A to be as follows:

a. Current U.S. force deployments in Vietnam will have to be maintained and additional deployments already authorized should be made.
b. Efforts to hamper Communist use of Laos as a corridor for infiltrating troops and supplies into South Vietnam should be continued and in some respects intensified. There should be further employment of B-52's against selected choke points vulnerable to this type of attack. Additional programs should be developed to make our interdiction attacks more effective.
c. The aerial pressure campaign on North Vietnam should be sustained for both military and psychological purposes. Attacks should not be mounted against population centers such as Hanoi or Haiphong, but major POL storage depots should be destroyed and, probably, Haiphong harbor should be mined.
d. Within South Vietnam we must recognize that the period of political transition now in train--even if it evolves in the most favorable fashion possible--will produce some diminution in the effectiveness of central authority and some disruption in current programs. At best, we will be in for a situation like that of late 1963. It is essential that the Communists be prevented from making major military gains during this time of transition or scoring military successes which would generate an aura of invincibility or seriously damage the morale of our South Vietnamese allies. Therefore, it is essential that during this period, Communist forces be constantly harried, kept off balance, and not permitted to press their advantage. The bulk of this task will have to be borne by U.S. and allied forces during the immediate future and these forces must be aggressively and offensively employed.

Option B-O, as developed by Unger, assumed a "policy decision that we will undertake to find a way to bring to an end by negotiation the military contest in South Viet-Nam." (This paper, dated "4/14/66," was prepared after the April 9 meeting but was filed with the other papers of that date.) It was the optimistic version of this option because Unger assumed the possibility of reaching a settlement "on terms which preserve South Vietnam intact and in a condition which offers at least a 60-40 chance of its successfully resisting Communist attempts at political takeover."

In pursuit of this option the United States would persuade the GVN to negotiate with the NLF, offering amnesty and a coalition government, though not one giving the NLF control of the military, the police, or the treasury. The United States would withdraw troops "in return for the withdrawal of North Vietnamese military forces and political cadre." Perhaps, agreements between South Vietnam and North Vietnam would provide for economic intercourse and mutual recognition.

It would not be easy to persuade the GVN, Unger conceded. Doing so might require not only words but withholding of funds or withdrawal of some American forces. And once the GVN appreciated that the United States was in earnest, there would be danger of its collapse. Even if these problems were surmounted, there would remain the difficulty of pressing the negotiations to conclusion. "There is no assurance," Unger wrote, "that a negotiated settlement can pass successfully between the upper millstone of excessively dangerous concessions to the VC/NLF and the nether millstone of terms insufficiently attractive to make the VC/NLF consider it worthwhile to negotiate."

Militarily, Unger reasoned, Option B-O would call for continuation of current efforts, perhaps with a modest increase in ground forces but with no step-up in the air war. Total refusal to talk on the part of the Communists would, however, Unger wrote,

. . . . leave us with a question of what kind of stick we have to substitute for the proffered carrot and this might bring us up against the judgment of whether intensification and extension of our bombing in North Viet-Nam, coupled with whatever greater military efforts could be made in the South would bring the Communists to the table.

McNaughton's papers do not contain his original memorandum setting forth the pessimistic version of Option B. One can, however, infer its outlines from various other pieces in the McNaughton collection.

The difference between McNaughton and Unger presumably did not concern the objective--negotiating out. It lay in McNaughton's expressing less confidence in an outcome not involving Communist control of South Vietnam. On the first Monday in April, he had talked with Michael Deutch, freshly back from Saigon. His notes read:

1. Place (VN) in unholy mess.
2. We control next to no territory.
3. Fears economic collapse.
4. People would not vote for "our ride."
5. Wants to carry out economic warfare in VC.
6. This is incorruptible and popular. Chieu [sic] is best successor for Ky.
7. Militarily will be same place year from now.
8. Pacification won't get off ground for a year.

If McNaughton himself accepted anything like this estimate, he would have been pessimistic indeed about prospects for the GVN's survival. Even if he did not take quite so gloomy a view, he probably felt, as he had intimated in one of his January memoranda, that the United States should prepare to accept something less than the conditions which Unger sketched. What practical consequences followed from this difference in view, one can only guess.

Option C, as stated by Ball, rested on the assumption that "the South Vietnamese people will not be able to put together a government capable of maintaining an adequate civil and military effort or--if anything resembling actual independence is ever achieved--running the country." On this premise, he concluded, much as in earlier memoranda, "we should concentrate our attention on cutting our losses." Specifically, he recommended official declarations that United States support depended on a representative government which desired American aid and which demonstrated its ability to create "the necessary unity of action to assure the effective prosecution of the war and the peace." Seizing upon the next political crisis in South Vietnam, the United States should, said Ball, "halt the deployment of additional forces, reduce the level of air attacks on the North, and maintain ground activity at the minimum level required to prevent the substantial improvement of the Viet Cong position."

Ball described two alternative outcomes from Option C. One was that the South Vietnamese might unify and "face reality," the other, far more likely in Ball's estimation, was that South Vietnam would fragment still further, "leading to a situation in which a settlement would be reached that contemplated our departure." He closed:

Let us face the fact that there are no really attractive options open to us. To continue to fight the war with the present murky political base is, in my judgment, both dangerous and futile. It can lead only to increasing commitments, heavier losses, and mounting risks of dangerous escalation.

In McNaughton's files are pencil notes which may relate either to his own missing memorandum or to a conversation that took place among some of the officials concerned. Despite its cryptic nature, it is worth reproducing in its entirety, in part because it gives a clue to thoughts passing at this time through McNamara's mind:

Do we press VNese or do they move themselves[?]
What the point of probes if (w[oul]d be counterproductive otherwise)


l. No more US forces unless better gov
2. Reemph[asis] of cond[itions]

(a) Rep govt ask[ed]
(b) Performance

3. Fashion govt unified and stable govt. Give time. Protect selves. Defend selves.
4. Effect

(a) Nationalist
(b) VC deal by GVN

If squeeze GVN first, and go to [Ball's position] later, have contaminated Course C. Better to claim we want to win and they rush out to settle.
Timing critical. 10 days ago. Not today. Will have new chance when advisors decide how election set up. Unless elections rigged, Budhists to streets. Need Pres. statements re (a) cond[itio]ns and (b) optimism VNese moving that way.
W[oul]dn't the SVNese just comply and knuckle down and not do any better[?] How do we move them toward compromise[?] Maybe second time, we do throw in the towel and they make deal.
Lodge more likely to go for Ball ultimatum than B.
Anti-US govt likely to follow. How handle actual departure[?] Do we want to precipitate anti-US[?}
Must we condition US and world public for 6 mos before "ultimatum." Pres. to press, ans. qn. giving bases of our help.
BUT, why not get better deal for SVN by RSM approach? Give them choice now between (1) chaos 6 mos from now (via Ball) and VC govt. and (2) chance at compromise now with even chance of something better. If we followed RSM approach, ruin our image (pushing for deal) and cause demoralization. Tri Quang may even say we selling out.
We chilled bids earlier.
Could there be an independent Delta? Already accommodation.

As McNaughton's notes reveal, the group that met at the White House on April 9 was preoccupied with the immediate political crisis in South Vietnam. Eary that morning, Walt Rostow had addressed a memo to Secretaries Rusk and McNamara suggesting a course of action for "breaking Tri Quang's momentum." His proposal--which was the form the subsequent solution took--called for giving substantial tactical concessions to the Buddhists on the issue of the Constituent Assembly in order to bring the regime-threatening demonstrations to an end. At the White House meeting later that day several participants were called on to prepare papers on the crisis.

Leonard Unger of the State Department drafted a paper outlining five possible outcomes of the crisis, the last two of which were a secession of neutralist northern provinces and/or a complete collapse of Saigon political machinery with the VC moving into the vacuum. His paper was probably considered at a meeting on Monday, April 12, as suggested by McNaughton's handwritten notes. At the same meeting, a long memorandum prepared by George Carver of CIA in response to a request at the Friday meeting, and entitled "Consequences of Buddhist Political Victory in South Vietnam," was also considered. Carver argued that while a Buddhist government would have been difficult for us to deal with it would not have been impossible and, given the evident political strength of the Buddhists, might even work to our long range advantage. The three American options in such a contingency were: (1) trying to throw out the new government; (2) attempting to work with it; or (3) withdrawing from South Vietnam. Clearly, he argued, the second was the best in view of our commitments.

That same day, Maxwell Taylor sent the President a detailed memo with recommendations for dealing with the Buddhist uprising. In essence he recommended that the U.S. take a tough line in support of Ky and against the Buddhists. In his words,

. . . . we must prevent Tri Quang from overthrowing the Directorate (with or without Ky who personally is expendable) and support a conservative, feasible schedule for a transition to constitutional government. In execution of such a program, the GVN (Ky, for the present) should be encouraged to use the necessary force to restore and maintain order, short of attempting to reimpose government rule by bayonets on Danang-Hue which, for the time being, should be merely contained and isolated.

These recommendations, however, had been overtaken by events. The GVN had already found a formula for restoring order and appeasing the Buddhists. In a three day "National Political Congress" in Saigon from April 12-14, the GVN adopted a program promising to move rapidly toward constitutional government which placated the main Buddhist demands. For a few weeks the demonstrations ceased and South Vietnam returned to relative political quiet. While not unusual as policy problems go, this political crisis in South Vietnam intervened temporarily to divert official attention from the broader issues of the war and indirectly contributed to the deferral of any decision to authorize attacks on the POL in North Vietnam. Other issues and problems would continue to defer the POL decision, both directly and indirectly, for another two months.

With some semblance of calm restored momentarily to South Vietnamese politics, the second-level Washington policy officials could turn their attention once again to the broader issues of U.S. policy direction. On April 14, Walt Rostow sent McNaughton a memo entitled "Headings for Decision and Action: Vietnam, April 14, 1966," (implying topics for discussion at a meeting later that day?). Item one on Rostow's agenda was a proposed high-level U.S. statement endorsing the recent evolution of events in South Vietnam and stipulating that continued U.S. assistance and support would be contingent on South Vietnamese demonstration of unity, movement toward constitutional government, effective prosecution of the war, and maintenance of order. His second topic was the bombing of the North, and subheading "b" re-opened the POL debate with the simple question, "Is this the time for oil?" Other issues which he listed for consideration included:
accelerating the campaign against main force units, economic stabilization, revolutionary construction, Vietnamese politics (including constitution-making), and negotiations between the GVN and the VC (if only for political warfare purposes).

On the same day, the JCS forwarded to the Secretary the previously mentioned "ROLLING THUNDER Study Group Report: Air Operations Against NVN" with a cover memo noting that its recommendations for a stepped up bombing campaign were "in consonance with the general concept recommended in JCSM4 1-66 The voluminous study itself recommended a general expansion of the bombing with provision for three special attack options, one against the Haiphong POL center; the second for the aerial mining of the sea approaches to Haiphong, Hon Gai, and Cam Pha; and the third for strikes at the major airfields of Hanoi, Haiphong, and Phuc Yen. In offering these options, the report stated that, "Military considerations would require that two of the special attack options, POL and mining, be conducted now. However, appreciation of the sensitivity of such attacks is recognized and the precise time of execution must take into account political factors." Somewhat optimistically, the report estimated that the POL strike would involve only 13 civilian casualties, and the mining would cause none. While there is no specific record of the Secretary's reaction to this full-blown presentation of the arguments for expanded bombing, he had sent a curt memo to the Chiefs the previous day in reply to their JCSM-189-66 of March 26, in which they had again urged attacking the POL. Tersely reflecting the President's failure to adopt their (and his) recommendation, he stated, "I have received JCSM-189-66. Your recommendations were considered in connection with the decision on ROLLING THUNDER 50."

As the second-echelon policy group returned to its consideration of the four options for U.S. policy (previously known as A, B-O, B-P, and C), the weight of recent political instability shifted its focus somewhat. When the group met again on Friday, April 16, at least three papers were offered for deliberation. William Bundy's draft was titled, "Basic Choices in Viet-Nam"; George Carver of CIA contributed "How We Should Move"; and a third paper called "Politics in Vietnam: A 'Worst' Outcome" was probably written by John McNaughton.

Bundy began with a sober appraisal of the situation:

The political crisis in South Viet-Nam has avoided outright disaster up to this point, but the temporary equilibrium appears to be uneasy and the crisis has meant at the very least a serious setback of the essential nonmilitary programs.

But the closeness with which political disaster had been averted in the South in the preceding week, "forces us to look hard at our basic position and policy in South Viet-Nam. We must now recognize that three contingencies of the utmost gravity are in some degree, more likely than our previous planning had recognized. . . ." The three contingencies Bundy had in mind were: (1) a state of total political chaos and paralysis resulting from an uprising by the Buddhists countered by the Catholics, Army, etc.; (2) the emergence of a neutralist government with wide support that would seek an end to the war on almost any basis and ask for a U.S. withdrawal; and (3) a continuation of the present GVN but in an enfeebled condition unable to effectively prosecute the war, especially the vital nonmilitary aspects of it. Bundy's estimate was that the third contingency was the most likely at that moment, and that even the most optimistic scenario for political and constitutional evolution could not foresee a change within the succeeding three to four months. Nevertheless, he outlined the four possible U.S. lines of action much as they had been presented before:

Option A: To continue roughly along present lines, but to hope that the setback is temporary.
Option B: To continue roughly along present lines, but to move more actively to stimulate a negotiated solution, specifically through contact between the Saigon government and elements in the Viet Cong and Liberation Front. This option [lined out in McNaughton] could be approached on an "optimistic" [underlined in McNaughton] or "lesser risk" [lined out in McNaughton with "harder" penciled in above and question marks in the margin] basis, or on a "pessimistic" [McNaughton underline] or "greater risk" [lined out in McNaughton with "softer" pencilled in] basis. The opening moves might be the same in both options, but more drastic indications of the U.S. position would ["be involved" penned in by McNaughton] in the "pessimistic" approach ["which shades into option C below." penned by McNaughton].

Option C: To decide now that the chances of bringing about an independent (and non-Communist) [parentheses added by McNaughtonl South Viet-Nam have shrunk to the point where, on an over-all basis, the US effort is no longer warranted [lined out by McNaughton and replaced in pencil with "should be directed at a minimum-cost disengagement." Stet pencilled in the margin.] This would mean setting the stage rapidly [circled by McNaughton] for US disengagement and withdrawal irrespective of whether any kind of negotiation would work or not." [question marks in the margin.]

Bundy did not identify in the paper his preferred option. The tone of his paper, however, suggested a worried preference for "A". In a concluding section he listed a number of "broader factors" which "cut, as they always have, in deeply contradictory directions." The first was the level of support for the Vietnam policy within the U.S. While it was adequate for the moment, continued GVN weakness and political unrest could seriously undermine it. With an eye on the 1968 Presidential elections, Bundy prophetically summed up the problem:

As we look a year or two ahead, with a military program that would require major further budget costs-with all their implications for taxes and domestic programs-and with steady or probably rising casualties, the war could well become an albatross around the Administration's neck, at least equal to what Korea was for President Truman in 1952.

Moreover, if the prevailing malaise about the war among our non-SEATO allies degenerated into open criticism, a far wider range of world issues on which their cooperation was required might be seriously affected. With respect to the Soviet Union, no movement on disarmament or other matters of detente could be expected while the war continued. But since no significant change in Chinese or North Vietnamese attitudes had been expected in any circumstances, continuing the war under more adverse conditions in South Vietnam would hardly worsen them. Bundy ended his paper with an analysis of the impact of a U.S. failure in South Vietnam on the rest of non-communist Asia, even if the failure resulted from a political collapse in Saigon.

5. Vis-a-vis the threatened nations of Asia, we must ask ourselves whether failure in Viet-Nam because of clearly visible political difficulties not under our control would be any less serious than failure by our own choice [lined out in McNaughton] without this factor. The question comes down, as it always has, to whether there is any tenable line of defense in Southeast Asia if Viet-Nam falls. Here we must recognize that the anti-Communist regime in Indonesia has been a tremendous "break" for us, both for in [McNaughton] removing the possibility of a Communist pincer movement, which appeared
irresistible almost certain [McNaughton] a year ago, and in [McNaughton] opening up the possibility that over a period of some years Indonesia may become a constructive force. But for the next year or two any chance of holding the rest of Southeast Asia hinges on the same factors assessed a year ago, whether Thailand and Laos in the first instance and Malaysia, Singapore, and Burma close behind, would--in the face of a US failure for any reason in Viet-Nam--have any significant remaining will to resist the Chinese Communist pressures that would probably then be applied. Taking the case of Thailand as the next key point, it must be our present conclusion that--even if sophisticated leaders understood the Vietnamese [McNaughton] political weaknesses and our inability to control them--to the mass of the Thai people the failure would remain a US failure and a proof that Communism from the north was the decisive force in the area. Faced with this reaction, we must still conclude that Thailand simply could not be held in these circumstances, and that the rest of Southeast Asia would probably follow in due course. In other words, the strategic stakes in Southeast Asia are fundamentally unchanged by the possible political nature of the causes for failure in Viet-Nam. The same is almost certainly true of the shockwaves that would arise against other free nations--Korea, Taiwan, Japan, and the Philippines--in the wider area of East Asia. Perhaps these shockwaves can be countered, but they would not [McNaughton] be mitigated by the fact that the failure arose from internal political [sic] causes rather than any US major error or omission.

Once again, the domino theory, albeit in a refined case by case presentation, was offered by this key member of the Administration as a fundamental argument for the continuing U.S. involvement in Vietnam. Bundy rejected even the subtle argument, offered by some longtime Asian experts, that the uniqueness of the Vietnamese case, particularly its extraordinary lack of political structure, invalidated any generalization of our experience there to the rest of Asia. Thus, he argued the American commitment was both open-ended and irreversible.

George Carver of CIA argued quite a different point of view. His paper began, "The nature and basis of the U.S. commitment in Vietnam is widely misunderstood within the United States, throughout the world, and in Vietnam itself." Placing himself squarely in opposition to the kind of analysis presented by Bundy, Carver argued that we had allowed control over our policy to slip from our grasp into the "sometimes irresponsible and occasionally unidentifiable hands of South Vietnamese over whom we have no effective control. This is an intolerable position for a great power. By inferring that our commitment was irreversible and open-ended, Carver maintained we permitted the Vietnamese to exercise leverage over us rather than vice versa. To correct this mistaken view of our commitment and get our own priorities straight, Carver proposed a reformulation of objectives:

Whatever course of policy on Vietnam we eventually decide to adopt, it is essential that we first clarify the nature 'of our commitment in that country and present it in a manner which gives us maximum leverage over our Vietnamese allies and maximum freedom of unilateral action. What we need to do, in effect, is return to the original 1954 Eisenhower position and make it abundantly clear that our continued presence in Vietnam in support of the South Vietnamese struggle against the aggressive incursions of their northern compatriots is contingent on the fulfillment of both of two necessary conditions:

(a) A continued desire by the South Vietnamese for our assistance and physical presence.
(b) Some measure of responsible political behavior on the part of the South Vietnamese themselves including, but not limited to, their establishment of a reasonably effective government with which we can work.

Carver was careful to state, however, that two to three months would be required to prepare the ground for this kind of clarification so as not to have it appear we were reversing directions on Vietnam or presenting the GVN with an ultimatum. Effectively carried out, such a clarification would broaden the range of available options for the U.S. and place us in a much better position to effect desired changes. The mechanics of his proposal called for a Presidential speech in the near future along the lines suggested earlier that week by Walt Rostow. The President should express satisfaction at the evolution of political events in South Vietnam toward constitutional government and indicate "that our capacity to assist South Vietnam is dependent on a continued desire for our assistance and on the demonstration of unity and responsibility in the widening circle of those who will now engage in politics in South Vietnam." Other speeches by the Vice President and members of Congress in the succeeding weeks might stress the contingency of our commitment, and press stories conveying the new message could be stimulated. Finally, three or four months in the future, the President would complete this process by making our position and commitment crystal clear, possibly in response to a planted press conference question. This public effort would be supplemented by private diplomatic communication of the new message to South Vietnamese leaders by the Embassy.

Carver argued that putting the U.S. in a position to condition its commitment would considerably enhance U.S. flexibility in an uncertain policy environment.

Once the U.S. position is clear we can then see whether our word to the Vietnamese stimulates better and more responsible political behavior. If it does, we will have improved Option A's chances for success. If it does not, or if South Vietnam descends into chaos and anarchy, we will have laid the groundwork essential to the successful adoption of Option C with minimal political cost.

Questions which remained to be answered included: (1) whether to continue with scheduled troop deployment; (2) whether to give the GVN a specific list of actions on which we expected action and then rate their performance, or rely on a more general evaluation; (3) whether the U.S. should continue to probe the DRV/NLF on the possibility of negotiations; (4) whether to encourage the GVN to make negotiation overtures to the VC.

The third paper, Politics in Vietnam: A "Worst" Outcome (presumably by McNaughton), dealt with the unsavory possibility of a fall of the current government and its replacement by a "neutralist" successor that sought negotiations, a ceasefire, and a coalition with the VC. After considering a variety of possible, although equally unpromising, courses of action, the paper argued that in such a case the U.S. would have "little choice but to get out of Vietnam. . . . Governing objectives should be: minimizing the inevitable loss of face and protecting U.S. forces, allied forces, and those South Vietnamese who appeal to us for political refuge." An intriguing tab to the same paper considered the impact on the U.S. position in the Pacific and East Asia in the event of a withdrawal from Vietnam. Unlike the Bundy paper this analysis eschewed pure domino theorizing for a careful country by country examination. The overall evaluation was that, "Except for its psychological impact, withdrawal from Vietnam would not affect the present line of containment from its Korean anchor down the Japan-RyukyusTaiwan-Philippine Island chain." Four possible alternate defense lines in Southeast Asia were considered: (1) the Thai border; (2) the Isthmus of Kra on the Malay peninsula; (3) the "Water Line" from the Strait of Malacca to the North of Borneo; and (4) an "Interrupted Line" across the gap between the Philippines and Australia. Like other analyses of the strategic problem in Southeast Asia, this paper rejected any in-depth defense of Thailand as militarily untenable. The best alternatives were either the Isthmus of Kra or the Strait of Malacca; alternative four was to be considered only as a fall back position. The paper stands as a terse and effective refutation of the full-blown domino theory, offering as it does cool-headed alternatives that should have evoked more clear thinking than they apparently did about the irrevocability of our commitment to South Vietnam.

What the exact outcome of the deliberations on these papers was is not clear from the available documents. Nor is there any clear indication of the influence the documents or the ideas contained in them might have had on the Principals or the President. Judgments on this score must be by inference. A scenario drafted by Leonard Unger and included by McNaughton with Carver's paper suggests that some consensus was reached within the group reflecting mostly the ideas contained in Carver's draft. Its second point stated:

On U.S. scene and internationally we will develop in public statements and otherwise the dual theme that the U.S. has gone into South Viet-Nam to help on the assumption that (a) the Government is representative of the people who do want our help (b) the Government is sufficiently competent to hold the country together, to maintain the necessary programs and use our help. President will elaborate this at opportune moment in constructive tone but with monetary overtones if there is any political turmoil or if Government unwilling to do what we consider essential in such fields as countering inflation, allocating manpower to essential tasks and the like.

In fact, however, while we did attempt to steer the South Vietnamese toward constitutional government on a democratic model, when the President spoke out in succeeding weeks it was to reiterate the firmness of our commitment and the quality of our patience, not to condition them. At a Medal of Honor ceremony at the White House on April 21, he said:

There are times when Viet-Nam must seem to many a thousand contradictions, and the pursuit of freedom there an almost unrealizable dream.

But there are also times--and for me this is one of them--when the mist of confusion lifts and the basic principles emerge:

--that South Viet-Nam, however young and frail, has the right to develop as a nation, free from the interference of any other power, no matter how mighty or strong;
--that the normal processes of political action, if given time and patience and freedom to work, will someday, some way create in South Viet-Nam a society that is responsive to the people and consistent with their traditions. . .

The third point in the Unger scenario was to encourage the GVN to establish contacts with the VC in order to promote defections and/or to explore the possibilities of "negotiated arrangements." This emphasis on contacts between the GVN and the VC may well have reflected the flurry of highly public international activity to bring about negotiations between the U.S. and the DRV that was taking place at that time (considered in more detail below). In any event, this entire effort at option-generation came to an inconclusive end around April 20.

The last paper to circulate was a much revised redraft of Course B that reflected the aforementioned ideas about GVN/VC contacts. It was, moreover, a recapitulation of ideas circulating in the spring of 1966 at the second-level of the government. That they were considerably out of touch with reality would shortly be revealed by the renewed I Corps-Buddhist political problem in May. The paper began with a paragraph discussing the "Essential element" of the course of action
--i.e. ". . . our decision now to press the GVN to expand and exploit its contacts with the VC/NLF." The point of these contacts was to determine what basis, if any, might exist for bringing the insurgency to an end.

The proposed approach to the GVN was to be made with three considerations in mind. The first was the dual theme that U.S. assistance in South Vietnam depended on a representative and effective GVN and the genuine desire of the people for our help. Continued political turmoil in South Vietnam would force us to state this policy with increasing sharpness. The second consideration was the U.S. military effort. McNaughton specifically bifurcated this section in his revision to include two alternatives, as follows:

(b) Continuation of the military program including U.S. deployments and air sorties.

(1) Alternative A. Forces increased by the end of the year to 385,000 men and to attacks on the key military targets outside heavily populated areas in all of North Vietnam except the strip near China.
(2) Alternative B. Forces increased in modest amounts by the end of the year to about 300,000 (with the possibility of halting even the deployments implicit in that figure in case of signal failure by the GVN to perform) and air attacks in the northeast quadrant of North Vietnam kept to present levels in terms of intensity and type of target.

The third consideration was a continuation of U.S. support for GVN revolutionary development and inflation control.

Two alternative GVN tactics for establishing contact with the NLF were offered. The first alternative would be an overt, highly publicized GVN appeal to the VC/NLF to meet with representatives of the GVN to work out arrangements for peace. Alternative two foresaw the initiation of the first contacts through covert channels with public negotiations to follow if the covert talks revealed a basis for agreement. All of this would produce, the paper argued, one of the following outcomes:

(a) If things were going passably for our side but the VC/NLF showed no readiness to settle on terms providing reasonable assurances for the continuation of a non-Communist regime in SVN, we might agree to plod on with present programs (with or without intensified military activity) until the VC/NLF showed more give.
(b) If things were going badly for our side we might feel obliged to insist on the GVN's coming to the best terms it could get with the VC/NLF, with our continuing military and other support conditioned on the GVN moving along those lines.
(c) If things were going well for our side, the VC/NLF might accede to terms which entailed no serious risks for a continuing non-Communist orientation of the GVN in the short term. It would probably have to be assumed that this would represent no more than a tactical retreat of the VC/NLF.

c. Exogenous Factors

No precise reason can be adduced for the termination of this interdepartmental effort to refine options for American action. In a general way, as the preceding paper shows, the effort had lost some touch with the situation; the GVN was far too fragile a structure at that point (and about to be challenged again in May by I Corps Commander General Thi and his Buddhist allies) to seriously contemplate contacts or negotiations with the VC. In Washington, the President and his key advisors Rusk and McNamara were preoccupied with a host of additional immediate concerns as well. The President had a newly appointed Special Assistant, Robert Komer, who had recently returned from a trip to Vietnam urging greater attention to the non-military, nation-building aspects of the struggle. In addition, the President was increasingly aware of the importance of the war, its costs, and its public relations to the upcoming Congressional elections. McNamara and the JCS were struggling to reach agreement on force deployment schedules and requirements; and Rusk was managing the public U.S. response to a major international effort to bring about U.S. negotiations with Hanoi. These concerns, as we shall see, served to continue the deferral of any implementation of strikes against North Vietnamese POL reserves.

On April 19, about the time the option drafting exercise was ending, Robert Komer addressed a lengthy memo to the President (plus the Principals and their assistants) reporting on his trip to Vietnam to review the non-military aspects of the war. Presidential concern with what was to be called "pacification" had been piqued during the Honolulu Conference in February. Upon his return to Washington, President Johnson named Komer to become Special Assistant within the White House to oversee the Washington coordination of the program. To emphasize the importance attached to this domain, Komer's appointment was announced in a National Security Action Memorandum on March 28. As a "new boy" to the Vietnam problem, Komer betook himself to Saigon in mid-April to have a first-hand look. His eleven page report represents more a catalogue of the well-known problems than any very startling suggestion for their resolution. Nevertheless, it did provide the President with a detailed review of the specific difficulties in the RD effort, an effort that the President repeatedly stressed in his public remarks in this period.

At Defense, problems of deployment phasing for Vietnam occupied a good portion of McNamara's time during the spring of 1966. On March 1, the JCS had forwarded a recommendation for meeting planned deployments that envisaged extending tours of service for selected specialties and calling up some reserve units. Whatever McNamara's own views on calling the reserves, the President was clearly unprepared to contemplate such seemingly drastic measures at that juncture. Like attacks on North Vietnamese POL, a reserve callup would have been seen as a complete rejection of the international efforts to get negotiations started and as a decisive escalation of the war. Moreover, to consider such an action at a time when South Vietnam was in the throes of a protracted political crisis would have run counter to the views of even some of the strongest supporters of the war. So, on March 10, the Secretary asked the Chiefs to redo their proposal in order to meet the stipulated deployment schedule, stating that it was imperative that, ". . . all necessary actions . . . be taken to meet these deployment dates without callup of reserves or extension of terms of service." The JCS replied on April 4 that it would be impossible to meet the deployment deadlines because of shortages of critical skills. They proposed a stretch-out of the deployments as the only remedy if reserve callups and extension of duty tours were ruled out. Not satisfied, the Secretary asked the Chiefs to explain in detail why they could not meet the requirements within the given time schedule. The Chiefs replied on April 28 with a listing of the personnel problems that were the source of their difficulty, but promised to take "extraordinary measures" in an effort to conform as closely as possible to the desired closure schedule. The total troop figure for Vietnam for end CY 66 on which agreement was then reached was some 276,000 men. This constituted Program 2-AR.

These modifications and adjustments to the troop deployment schedules, of course, had implications for the supporting forces as well. The Chiefs also addressed a series of memos to the Secretary on required modifications in the deployment plans for tactica' aircraft to support ground forces, and for increases in air munitions requirements. These force expansions generated a requirement for additional airfields. When these matters are added to the problems created for McNamara and his staff by the French decision that spring to request the withdrawal of all NATO forces from French soil, it is not hard to understand why escalating the war was momentarily set aside.

Another possible explanation for delaying the POL strikes can be added to those already discussed. The spring of 1966 saw one of the most determined and most public efforts by the international community to bring the U.S. and North Vietnam to the negotiating table. While at no time during this peace initiative was there any evidence, public or private, of give in either side's uncompromising position and hence real possibility of talks, the widespread publicity of the effort meant that the Administration was constrained from any military actions that might be construed as "worsening the atmosphere" or rebuking the peace efforts. Air strikes against DRV POL reserves would obviously have fallen into this category.

In February, after the resumption of the bombing, Nkrumah and Nasser unsuccessfully attempted to get negotiations started, the former touring several capitals including Moscow to further the effort. DeGaulle replied to a letter from Ho Chi Minh with an offer to play a role in settling the dispute, but no response was forthcoming. Prime Minister Wilson met with Premier Kosygin in Moscow from Feb. 22-24 and urged reconvening the Geneva Conference; the Soviets countered by saying the U.S. and DRV must arrange a conference since the conflict was theirs. Early in March, Hanoi reportedly rejected a suggestion by Indian President Radharrishnon for an Asian-African force to replace American troops in South Vietnam. Later that month Canadian Ambassador Chester Ronning went to Hanoi to test for areas in which negotiations might be possible. He returned with little hope, other than a vague belief the ICC could eventually play a role.

Early in April, UN Secretary General U Thant advocated Security Council involvement in Vietnam if Communist China and North Vietnam agreed, and he reiterated his three point proposal for getting the parties together (cessation of bombing; scaling down of all military activity; and willingness of both sides to meet). No response was forthcoming from the DRV, but later that month during meetings of the "Third National Assembly" Ho and Premier Pham Van Dong reiterated the unyielding North Vietnamese position that the U.S. must accept the four points as the basis for solving the war before negotiations could start. On April 29, Canadian Prime Minister Pearson proposed a ceasefire and a gradual withdrawal of troops as steps toward peace. The ceasefire was seen as the first part of peace negotiations without prior conditions Phased withdrawals would begin as the negotiations proceeded. The U.S. endorsed the Pearson proposal which was probably enough at that stage to insure its rejection by Hanoi. On the same day, Danish PM Krag urged the US to accept a transitional coalition government as a realistic step toward peace.

In May, Netherlands Foreign Minister Luns proposed a mutual reduction in the hostilities as a step toward a ceasefire and to prevent any further escalation. Neither side made any direct response. On May 22, Guinea and Algeria called for an end to the bombing and a strict respect for the Geneva Agreements as the basis of peace in Vietnam. In a major speech on May 25, U Thant called for a reduction of hostilities, but rejected the notion that the UN had prime responsibility for finding a settlement. Early in June press attention was focused on apparent Romanian efforts to bring Hanoi to the negotiating table. Romanian intermediaries made soundings in Hanoi and Peking but turned up no new sentiment for talks. In mid-June Canadian Ambassador Ronning made a second trip, to Hanoi but found no signs of give in the DRV portion (detailed discussion below). Near the end of June a French official, Jean Sainteny, reported from Hanoi and Peking through Agence France-Presse that the DRV had left him with the impression that negotiations might be possible if the U.S. committed itself in advance to a timetable for the withdrawal of forces from South Vietnam. With pressure again mounting for additional U.S. measures against the North and the failure of the Ronning mission, the State Department closed out this international effort on June 23 (the day after the original POL execute order), stating that neither oral reports nor public statements indicated any change in the basic elements of Hanoi's position. On June 27, Secretary Rusk told the SEATO Conference in Canberra, "I see no prospect of peace at the present moment." The bombing of the POL storage areas in Hanoi and Haiphong began on June 29.

The seriousness with which these international efforts were being treated within the U.S. Government is reflected in two memos from the period of late April and early May. On April 27, Maxwell Taylor, in his capacity as military advisor to the President, sent a memo to the President entitled, "Assessment and Uses of Negotiation Blue Chips." The heart of his analysis was that bombing was a "blue chip" like ceasefire, withdrawal of forces, amnesty for VC/NVA, etc., to be given away at the negotiation table for something concrete in return, not abandoned beforehand merely to get negotiations started. The path to negotiations would be filled with pitfalls, he argued,

Any day, Hanoi may indicate a willingness to negotiate provided we stop permanently our bombing attacks against the north. In this case, our Government would be under great pressure at home and abroad to accept this precondition whereas to do so would seriously prejudice the success of subsequent negotiations.

To avoid this dilemma, Taylor urged the President to clearly indicate to our friends as well as the enemy that we were not prepared to end the bombing except in negotiated exchange for a reciprocal concession from the North Vietnamese. His analysis proceeded like this:

To avoid such pitfalls, we need to consider what we will want from the Communist side and what they will want from us in the course of negotiating a cease-fire or a final settlement. What are our negotiating assets, what is their value, and how should they be employed? As I see them, the following
are the blue chips in our pile representing what Hanoi would or could like from us and what we might consider giving under certain conditions.

a. Cessation of bombing in North Viet-Nam.
b. Cessation of military operations against Viet Cong units.
c. Cessation of increase of U.S. forces in South Viet-Nam.
d. Withdrawal of U.S. forces from South Viet-Nam.
e. Amnesty and civic rights for Viet Cong.
f. Economic aid to North Viet-Nam.

The Viet Cong/Hanoi have a similar stack of chips representing actions we would like from them.

a. Cessation of Viet Cong incidents in South Viet-Nam.
b. Cessation of guerrilla military operations.
c. Cessation of further infiltration of men and supplies from North VietNam to South Viet-Nam.
d. Withdrawal of infiltrated North Vietnamese Army units and cadres.
e. Dissolution or repatriation of Viet Cong.

Continuing his argument, Taylor outlined his views about which "blue chips" we should trade in negotiations for concessions from the DRV.

If these are the chips, how should we play them to get theirs at minimum cost? Our big chips are a and d, the cessation of bombing and the withdrawal of U.S. forces; their big ones are c and e, the stopping of infiltration and dissolution of the Viet Cong. We might consider trading even, our a and d for their c and e except for the fact that all will require a certain amount of verification and inspection except our bombing which is an overt, visible fact. Even if Hanoi would accept inspection, infiltration is so elusive that I would doubt the feasibility of an effective detection system. Troop withdrawals, on the other hand, are comparatively easy to check. Hence, I would be inclined to accept as an absolute minimum a cessation of Viet Cong incidents and military operations (their a and b), which are readily verifiable in exchange for the stopping of our bombing and of offensive military operations against Viet Cong units (our a and b). If Viet Cong performance under the agreement were less than perfect, we can resume our activities on a scale related to the volume of enemy action. This is not a particularly good deal since we give up one of our big chips, bombing, and get neither of Hanoi's two big ones. However, it would achieve a cease-fire under conditions which are subject to verification and, on the whole, acceptable. We would not have surrendered the right to use our weapons in protection of the civil population outside of Viet Cong-controlled territory.

Summing up, Taylor argued against an unconditional bombing halt in these words:

Such a tabulation of negotiating blue chips and their purchasing power emphasized the folly of giving up any one in advance as a precondition for negotiations. Thus, if we gave up bombing in order to start discussions, we would not have the coins necessary to pay for all the concessions required for a satisfactory terminal settlement. My estimate of assets and values may be challenged, but I feel that it is important for us to go through some such exercise and make up our collective minds as to the value of our holdings and how to play them. We need such an analysis to guide our own thoughts and actions and possibly for communication to some of the third parties who, from time to time, try to get negotiations started. Some day we may be embarrassed if some country like India should express the view to Hanoi that the Americans would probably stop their bombing to get discussions started and then have Hanoi pick up the proposal as a formal offer. To prepare our own people as well as to guide our friends, we need to make public explanation of some of the points discussed above.

In conclusion he sounded a sharp warning about allowing ourselves to become embroiled in a repetition of our Korean negotiating experience, where casualties increased during the actual bargaining phase itself. It is hard to assess how much influence this memo had on the President's and the Administration's attitudes toward negotiations, but in hindsight it is clear that thinking of this kind prevailed within the U.S. Government until the early spring of 1968.

Taylor's memo attracted attention both at State and Defense at least down to the Assistant Secretary level. William Bundy at State sent a memo to Secretary Rusk the following week commenting on Taylor's ideas with his own assessment of the bargaining value and timing of a permanent cessation of the bombing. Since they represent views on the bombing which were to prevail for nearly two years, Bundy's memo is reproduced in substantial portions below. Recapitulating Taylor's analysis and his own position, Bundy began,

Essentially, the issue has always been whether we would trade a cessation of bombing in the North for some degree of reduction or elimination of Viet Cong and new North Vietnamese activity in the South, or a cessation of infiltration from the North, or a combination of both.

Worried that Taylor's willingness to trade a cessation of US/GVN bombing and offensive operations for a cessation of VC/NVA activity might be prejudicial to the GVN, Bundy outlined his own concept of what would be a reciprocal concession from the DRV:

I have myself been more inclined to an asking price, at least, that would include both a declared cessation of infiltration and a sharp reduction in VC/NVA military operations in the South. Even though we could not truly verify the cessation of infiltration, the present volume and routes are such that we could readily ascertain whether there was any significant movement, using our own air. Moreover, DRV action concerning infiltration would be a tremendous psychological blow to the VC and would constitute an admission which they have always declined really to make.

Whichever form of trade might be pursued if the issue even arose--as it conceivably might through such nibbles as the present Ronning effort--I fully agree with General Taylor that we should do all we can to avoid the pitfalls of ceasing bombing in return simply for a willingness to talk.

Concerned that the current spate of international peace moves might entice the Administration in another bombing pause, Bundy reminded the Secretary that,

. . . . during our long pause in January, we pretty much agreed among ourselves that as a practical matter, if Hanoi started to play negotiating games that even seemed to be serious, we would have great difficulty in resuming bombing for some time. This was and is a built-in weakness of the "pause" approach. It does not apply to informal talks with the DRV, directly or indirectly, on the conditions under which we would stop bombing, nor does it apply to possible third country suggestions. As to the latter, I myself believe that our past record sufficiently stresses that we could stop the bombing only if the other side did something in response. Thus, I would not at this moment favor any additional public statement by us, which might simply highlight the issue and bring about the very pressures we seek to avoid.

Hence, he concluded,

As you can see, these reactions are tentative as to the form of the trade, but quite firm that there must in fact be a trade and that we should not consider another "pause" under existing circumstances. If we agree merely to these points, I think we will have made some progress.

Bombing was thus seen from within the Administration as a counter to be traded during negotiations, a perception not shared by large segments of the international community where bombing was always regarded as an impediment to any such negotiations. Hanoi, however, had always clearly seen the bombing as the focal point in the test of wills with the U.S.

While Secretary Rusk was fending off this international pressure for an end to the bombing and de-escalation of the war as a means to peace, the President was having increasing trouble with war-dissenters within his own party. The US had scarcely resumed the bombing of the North after the extended December- January pause when Senator Fulbright opened hearings by his Senate Foreign Relations Committee into the Vietnam war. Witnesses who took varying degrees of exception to U.S. policy as they testified in early February included former Ambassador George Kennan and retired General James Gavin. Secretary Rusk appeared on February 18 and defended U.S. involvement as a fulfillment of our SEATO obligations. In a stormy confrontation with Fulbright the Secretary repeatedly reminded the Senator of his support for the 1964 Tonkin Gulf Resolution. The next day, Senator Robert Kennedy stated that the NLF should be included in any postwar South Vietnamese government. Three days later, he clarified his position by saying that he had meant the NLF should not be "automatically excluded" from power in an interim government pending elections. Speaking no doubt for the President and the Administration, the Vice President pointedly rejected Kennedy's suggestion on February 21. On the other side of the political spectrum, Senator Russell, otherwise a hawk on the war, reacted in April to the continuing political turmoil in South Vietnam by suggesting a poll be taken in all large Vietnamese cities to determine whether our assistance was still desired by the Vietnamese. If the answer was no, he asserted, the U.S. should pull out of Vietnam.

The President was also regularly reminded by the press of the possible implications for the November Congressional elections of a continuing large effort in South Vietnam that did not produce results. Editorial writers were often even more pointed. On May 17, James Reston wrote:

President Johnson has been confronted for some time with a moral question in Vietnam, but he keeps evading it. The question is this: What justifies more and more killing in Vietnam when the President's own conditions for an effective war effort--a government that can govern and fight in Saigon--are not met?

By his own definition, this struggle cannot succeed without a regime that commands the respect of the South Vietnamese people and a Vietnamese army that can pacify the country. Yet though the fighting qualities of the South Vietnamese are now being demonstrated more and more against one another, the President's orders are sending more and more Americans into the battle to replace the Vietnamese who are fighting among themselves.

Public reaction to the simmering political crisis in South Vietnam was reflected in declining popular approval of the President's performance. In March, 68% of those polled had approved the President's conduct in office, but by May, his support had declined sharply to only 54%.

Some indication of the concern being generated by these adverse U.S. political effects of the governmental crisis in South Vietnam is offered by the fact that State, on May 21, sent the Embassy in Saigon the results of a Gallup Poll on whether the U.S. should continue its support for the war. These were the questions and the distribution of the responses:

1. Suppose South Vietnamese start fighting on big scale among themselves. Do you think we should continue help them, or should we withdraw our troops? (A) Continue to help 28 percent; (B) Withdraw 54 percent; (C) No opinion 18 percent.
2. If GVN decides stop fighting (discontinue war), what should US do--continue war by itself, or should we withdraw? (A) Continue 16 percent; (B) Withdraw 72 percent; (C) No opinion 12 percent. Comparison August 1965 is 19, 63 and 18 percent.
3. Do you think South Vietnamese will be able to establish stable government or not? (A) Yes 32 percent; (B) No 48 percent; (C) No opinion 20 percent. Comparison January 1965 is 25, 42 and 33 percent.

Lodge, struggling with fast moving political events in Hue and DaNang, replied to these poll results on May 23 in a harsh and unsympathetic tone,

We are in Viet-Nam because it cannot ward off external aggression by itself, and is, therefore, in trouble. If it were not in trouble, we would not have to be here. The time for us to leave is when the trouble is over--not when it is changing its character. It makes no sense for us here to help them against military violence and to leave them in the lurch to be defeated by criminal violence operating under political, economic and social guise.

It is obviously true that the Vietnamese are not today ready for self-government, and that the French actively tried to unfit them for self-government. One of the implications of the phrase "internal squabbling" is this unfitness. But if we are going to adopt the policy of turning every country that is unfit for self-government over to the communists, there won't be much of the world left.

Lodge rejected the implications of these opinion polls in the strongest possible terms, reaffirming his belief in the correctness of the U.S. course,

The idea that we are here simply because the Vietnamese want us to be here--which is another implication of the phrase "internal squabbling"--; that we have no national interest in being here ourselves; and that if some of them don't want us to stay, we ought to get out is to me fallacious. In fact, I doubt whether we would have the moral right to make the commitment we have made here solely as a matter of charity towards the Vietnamese and without the existence of a strong United States interest. For one thing, the U.S. interest in avoiding World War III is very direct and strong. Some day we may have to decide how much it is worth to us to deny Viet-Nam to Hanoi and Peking--regardless of what the Vietnamese may think.

Apparently unable to get the matter off his mind, Lodge brought it up again in his weekly NODIS to the President on May 25,

I have been mulling over the state of American opinion as I observed it when I was at home. I have also been reading the recent Gallup polls. As I commented in my EMBTEL 4880, I am quite certain that the number of those who want us to leave Viet-Nam because of current "internal squabbling" does not reflect deep conviction but a superficial impulse based on inadequate information.

In fact, I think one television fireside chat by you personally--with all your intelligence and compassion--could tip that figure over in one evening. I am thinking of a speech, the general tenor of which would be: "we are involved in a vital struggle of great difficulty and complexity on which much depends. I need your help."

I am sure you would get much help from the very people in the Gallup poll who said we ought to leave Viet-Nam-as soon as they understood what you want them to support.

Lodge's reassurances, however, while welcome bipartisan political support from a critical member of the team, could not mitigate the legitimate Presidential concerns about the domestic base for an uncertain policy. Thus, assailed on many sides, the President attempted to steer what he must have regarded as a middle course.

The President's unwillingness to proceed with the bombing of the POL storage facilities in North Vietnam continued in May in spite of the near consensus among his top advisors on its desirability. As already noted, the JCS recommendation that POL be included in Program 50 of the ROLLING THUNDER strikes for the month of May had been disapproved. An effort was made to have the strikes included in the ROLLING THUNDER series for the month of May, which ordinarily would have been ROLLING THUNDER 51, but the decision was to extend ROLLING THUNDER 50 until further notice, holding the POL question in abeyance. On May 3, McNaughton sent Walt Rostow a belated list of questions, "to put into the 'ask-Lodge' hopper." The first set of proposed queries had to do with the bombing program and included specific questions attacking POL. Whether Rostow did, in fact, query Lodge on the matter is not clear from the available cables, but in any case, Rostow took up the matter of the POL attacks himself in an important memorandum to Rusk and McNamara on May 6. Rostow developed his argument for striking the petroleum reserves on the basis of U.S. experience in the World War II attacks on German oil supplies and storage facilities. His reasoning was as follows:

From the moment that serious and systematic oil attacks started, front line single engine fighter strength and tank mobility were affected. The
reason was this: it proved much more difficult, in the face of general oil shortage, to allocate from less important to more important uses than the simple arithmetic of the problem would suggest. Oil moves in various logistical channels from central sources. When the central sources began to dry up the effects proved fairly prompt and widespread. What look like reserves statistically are rather inflexible commitments to logistical pipelines.

The same results might be expected from heavy and sustained attacks on the North Vietnamese oil reserves,

With an understanding that simple analogies are dangerous, I nevertheless feel it is quite possible the military effects of a systematic and sustained bombing of POL in North Vietnam may be more prompt and direct than conventional intelligence analysis would suggest.

I would underline, however, the adjectives "systematic and sustained." If we take this step we must cut clean through the POL system--and hold
the cut--if we are looking for decisive results.

On May 9, recalling that the VC had recently attacked three South Vietnamese textile factories, Westmoreland suggested that to deter further assaults against South Vietnamese industry, the U.S. should strike a North Vietnamese industrial target with considerable military significance such as the Thai Nguyen iron and steel plant. Concurring with the basic intent of the proposal, CINCPAC recommended that the target be the North Vietnamese POL system instead. "Initiation of strikes against NVN POL system and subsequent completed destruction, would be more meaningful and further deny NVN essential war making resources.

Lending further support to these military and civilian recommendations was a study completed on May 4 by the Air Staff which suggested that civilian casualties and collateral damage could be minimized in POL strikes if only the most experienced pilots, with thorough briefing were used; if the raids were executed only under favorable visual flight conditions with maximum use of sophisticated navigational aids; and if weapons and tactics were selected for their pinpoint accuracy rather than area coverage. On May 22, COMUSMACV sent CINCPAC yet another recommendation for retaliatory air strikes against North Vietnamese industrial and military targets. He called for plans that would permit the U.S. to respond to any VC terror attacks by an air strike against a similar target in the North. In particular, the Hanoi and Haiphong oil storage sites were recommended as reprisal targets for VC attacks against U.S. or South Vietnamese POL.

Intervening again in mid-May, however, was yet another round of the continuing South Vietnamese political crisis. It is not clear whether or not a decision on the strikes against Hanoi/Haiphong POL was deferred by the President for this reason, but it is plausible to think that it was a factor. In brief, the Buddhists in Hue and DaNang, with the active support and later leadership of General Thi, the I Corps commander, defied the central government. Thi refused to return to Saigon when ordered and only when Ky flew to DaNang and intervened with troops and police to recapture control of the two cities was GVN authority restored to the area. The crisis temporarily put the constitutional processes off the track and diverted high level American attention from other issues. The effect of this dispute on public support for the U.S. involvement in the war has already been discussed. Concern with bringing an end to this internal strife in South Vietnam and with pushing a reluctant GVN steadily along the road to constitutional and democratic government preoccupied the highest levels of the U.S. Government throughout May. These concerns momentarily contributed to forcing the military aspects of the war into the background for harried U.S. leaders whose time is always insufficient to the range of problems to be dealt with.

4. The Decision to Strike

The POL decision was rapidly coming to a head. On May 31, a slight relaxation of the restrictions against attacking POL was made when six minor storage areas in relatively unpopulated areas were approved for attack. Apparently sometime in late May, possibly at the time of the approval of the six minor targets, the President decided that attacks on the entire North Vietnamese POL network could not be delayed much longer. In any case, sometime near the end of the month he informed British Prime Minister Wilson of his intentions. When Wilson protested, McNamara arranged a special briefing by an American officer for Wilson and Foreign Minister Michael Stewart on June 2. The following day, Wilson cabled his appreciation to the President for his courtesy, but expressed his own feeling of obligation to urge the President not to make these new raids. Thus, he stated:

I was most grateful to you for asking Bob McNamara to arrange the very full briefing about the two oil targets near Hanoi and Haiphong that Cot. Rogers gave me yesterday. . . .

I know you will not feel that I am either unsympathetic or uncomprehending of the dilemma that this problem presents for you. In particular, I wholly understand the deep concern you must feel at the need to do anything possible to reduce the losses of young Americans in and over Vietnam; and Cot. Rogers made it clear to us what care has been taken to plan this operation so as to keep civilian casualties to the minimum.

However, . . . I am bound to say that, as seen from here, the possible military benefits that may result from this bombing do not appear to outweigh the political disadvantages that would seem the inevitable consequence. If you and the South Vietnamese Government were conducting a declared war on the conventional pattern . . . this operation would clearly be necessary and right. But since you have made it abundantly clear-and you know how much we have welcomed and supported this-that your purpose is to achieve a negotiated settlement, and that you are not striving for total military victory in the field, I remain convinced that the bombing of these targets, without producing decisive military advantage, may only increase the difficulty of reaching an eventual settlement. . . .

The last thing I wish is to add to your difficulties, but, as I warned you in my previous message, if this action is taken we shall have to dissociate ourselves from it, and in doing so I should have to say that you had given me advance warning and that I had made my position clear to you. . . .

Nevertheless I want to repeat . . . that our reservations about this operation will not affect our continuing support for your policy over Vietnam, as you and your people have made it clear from your [April 1965] Baltimore speech onwards. But, while this will remain the Government's position, I know that the effect on public opinion in this country--and I believe throughout Western Europe--is likely to be such as to reinforce the existing disquiet and criticism that we have to deal with.

The failure of the special effort to obtain Wilson's support must have been disappointing, but it did not stop the onward flow of events. Available information leaves unclear exactly how firmly the President had decided to act and gives no specific indication of the intended date for the strikes. A package of staff papers prepared by McNaughton suggests that the original date was to have been June 10. A scenario contained in the package proposes a list of actions for the period 8-30 June and begins with strike-day minus 2. The suggested scenario was as follows:

S-[Strike] day minus 2: Inform UK, Australia, Japan
S-day minus 1: Notify Canada, New Zealand, Thailand, Laos, Philippines (Marcos only), GRC (Chiang only), Korea
S-hour minus 1: Inform GVN
S-hour: Strike Hanoi, Haiphong
S-hour plus 2: Announce simultaneously in Washington and Saigon
S-hour plus 3-5: SecDef press backgrounder (depends on strike timing and completeness of post-strike reports)

The package also included a draft JCS execute message, a draft State cable to the field on notifying third countries, a draft public announcement, a talking paper for a McNamara press conference, a list of anticipated press questions, and maps and photographs of the targets.

The circle of those privy to this tentative Presidential decision probably did not include more than a half dozen of the key Washington advisers. Certainly the military commanders in the field had not been informed. On June 5, West-moreland urged that strikes be made against POL at the "earliest possible" moment, noting that ongoing North Vietnamese dispersal efforts would make later attacks less effective. Admiral Sharp took the occasion to reiterate to Washington that the strikes, besides underscoring the US resolve to support SVN and increase the pressure against NVN, would make it difficult for Hanoi to disperse POL, complicate off-loading from tankers, necessitate new methods of trans-shipment, "temporarily" halt the flow to dispersed areas, and have a "direct effect" on the movement of trucks and watercraft-perhaps (if imports were inadequate) limiting truck use. Sharp called the POL targets the most lucrative available in terms of impairing NVN's military logistics capabilities. Two days later, in reporting the results of a review of the armed reecce program, CINCPAC again urged that POL be attacked. He particularly noted the importance of,

. . . the effort being made by the NVN to disperse, camouflage and package things into ever smaller increments. This is particularly true of POL. . . . This again emphasizes the importance of souce [sic] targets such as ports and major POL installations.

It is hoped that June will see a modification to the RT [ROLLING THUNDER] rules with authorization to syrike [sic] key POL targets, selected targets in the Hon Gai and Cam Pha compleses [sic], and relaxation of the restrictions against coastal armed recce in the NE. In addition, reduction in the size of the Hanoi,/Haiphong restricted areas would be helpful. . .

The CIA, however, remained skeptical of these expectations for strikes against POL. On June 8, they produced a special assessment of the likely effects of such an attack, probably in response to a request from the Principals for a last minute evaluation. The report emphasized that "neutralization" of POL would not in itself stop North Vietnamese support of the war, although it would have an adverse general effect on the economy.

It is estimated that the neutralization of the bulk petroleum storage facilities in NVN will not in itself preclude Hanoi's continued support of essential war activities. The immediate impact in NVN will be felt in the need to convert to an alternative system of supply and distribution. The conversion program will be costly and create additional burdens for the regime. It is estimated, however, that the infiltration of men and supplies into SVN can be sustained. The impact on normal economic activity, however, would be more severe. New strains on an already burdened economic control structure and managerial talent would cause reductions in economic activity, compound existing distribution problems, and further strain manpower resources. The attacks on petroleum storage facilities in conjunction with continued attacks on transportation targets and armed reconnaissance against lines of communications will increase the burden and costs of supporting the war.

The sequence of events in the POL scenario drawn up by McNaughton was interrupted on June 7 by yet another international diplomatic effort to get negotiations started, or at least to test Hanoi's attitudes toward such a possibility. Canadian Ambassador Chester Ronning had been planning a second visit to Hanoi for June 14-18 with State Department approval. Thus, when Rusk, who was travelling in Europe, learned on June 7 of the possibility of strikes before Ronning's trip, he urgently cabled the President to defer them.

. . . Regarding special operation in Vietnam we have had under consideration, I sincerely hope that timing can be postponed until my return. A major question in my mind is Ronning mission to Hanoi occurring June 14 through 18. This is not merely political question involving a mission with which we have fully concurred. It also involves importance of our knowing whether there is any change in the thus far harsh and unyielding attitude of Hanoi.

Much on his mind in making the request, as he revealed in a separate cable to McNamara the following day, was the likelihood of ". . . general international revulsion toward an act that might sabotage Ronning's efforts.

. . . I am deeply disturbed by general international revulsion, and perhaps a great deal at home, if it becomes known that we took an action which sabotaged the Ronning mission to which we had given our agreement. I recognize the agony of this problem for all concerned. We could make arrangements to get an immediate report from Ronning. If he has a negative report, as we expect, that provides a firmer base for the action we contemplate and would make a difference to people like Wilson and Pearson. If, on the other hand, he learns that there is any serious breakthrough toward peace, the President would surely want to know of that before an action which would knock such a possibility off the tracks. I strongly recommend, therefore, against ninth or tenth. I regret this because of my maximum desire to support you and your colleagues in your tough job.

The President responded to the Secretary's request and suspended action until Ronning returned. When Ronning did return, William Bundy flew to Ottawa and met with him on June 21. Bundy reported that he was "markedly more sober and subdued" and had found no opening or flexibility in the North Vietnamese position.

While these diplomatic efforts were underway, McNamara had informed CINCPAC of the high level consideration for the POL strikes, but stated:

Final decision for or against will be influenced by extent they can be carried out without significant civilian casualties. What preliminary steps to minimize would you recommend and if taken what number of casualties
do you believe would result?

CINCPAC replied eagerly listing the conditions and safeguards for the attack that the Air Staff study had suggested in early May. He would execute only under favorable weather conditions, with good visibility and no cloud cover, in order to assure positive identification of the targets and improved strike accuracy; select the best axis of attack to avoid populated areas; select weapons with optimum ballistic characteristics for precision; make maximum use of ECM support in order to hamper SA-2 and AAA radars and reduce "pilot distraction" during the strikes; and employ the most experienced pilots, thoroughly briefed. He added that NVN had an excellent alert system, which would provide ample time for people to take cover. In all, he expected "under 50" civilian casualties. (This was the Joint Staff estimate, too, but CIA in its 8 June report estimated that civilian casualties might run to 200-300.)

McNamara cabled his approval of the measures suggested and indicated that they would be included in the execute message. He stressed that the President's final decision would be greatly influenced by the ability to minimize civilian casualties and inquired about restrictions against flak and SAM suppression that might endanger populated areas. On June 16, CINCPAC offered further assurances that all possible measures would be taken to avoid striking civilians and that flak and SAM suppression would be under the tightest of restrictions.

The stage was thus set, and when the feedback from the Ronning mission revealed no change in Hanoi's position, events moved quickly.

On 22 June the execution message was released. It authorized strikes on the 7 POL targets plus the Kep radar, beginning with attacks on the Hanoi and Haiphong sites, effective first light on 24 June Saigon time.

The execution message is a remarkable document, attesting in detail to the political sensitivity of the strikes and for some reason ending in a "never on Sunday" injunction. The gist of the message was as follows:

Strikes to commence with initial attacks against Haiphong and Hanoi POL on same day if operationally feasible. Make maximum effort to attain operational surprise. Do not conduct initiating attacks under marginal weather conditions but reschedule when weather assures success. Follow-on attacks authorized as operational and weather factors dictate.

At Haiphong, avoid damage to merchant shipping. No attacks authorized on craft unless US aircraft are first fired on and then only if clearly North Vietnamese. Piers servicing target will not be attacked if tanker is berthed off end of pier.

Decision made after SecDef and CJCS were assured every feasible step would be taken to minimize civilian casualties would be small. If you do not believe you can accomplish objective while destroying targets and protecting crews, do not initiate program. Taking the following measures: maximum use of most experienced ROLLING THUNDER personnel, detailed briefing of pilots stressing need to avoid civilians, execute only when weather permits visual identification of targets and improved strike accuracy, select best axis of attack to avoid populated areas, maximum use of ECM to hamper SAM and AAA fire control, in order to limit pilot distraction and improve accuracy, maximum use of weapons of high precision delivery consistent with mission objectives, and limit SAM and AAA suppression to sites located outside populated areas.

Take special precautions to insure security. If weather or operational considerations delay initiation of strikes, do not initiate on Sunday, 26 June.

The emphasis on striking Hanoi and Haiphong POL targets on the same day and trying to achieve operational surprise reflected an acute concern that these targets were in well-defended areas and U.S. losses might be high. The concern about merchant shipping, especially tankers which might be in the act of off-loading into the storage tanks, reflected anxiety over sparking an international incident, especially one with the USSR.

With the execute message out, high-level interest turned to the weather in the Hanoi/Haiphong area. The NMCC began to send Secretary McNamara written forecasts every few hours. These indicated that the weather was not promising. Twice the strikes were scheduled but had to be postponed. Then, on 24 June, Philip Geyelin of the Wall Street Journal got hold of a story that the President had decided to bomb the POL at Haiphong, and the essential details appeared in a Dow Jones news wire that evening. This was an extremely serious leak, because of the high risk of U.S. losses if NVN defenses were fully prepared. The next day an order was issued cancelling the strikes.

The weather watch continued, however, under special security precautions. The weather reports, plus other messages relating to the strikes, continued, handled as Top Secret Special Category (SpeCat) Exclusive for the SecDef, CJCS, and CINCPAC. (It is not known whether the diplomatic scenario which involved informing some countries about the strikes ahead of time was responsible for the press leak; in any case, the classification and handling of these messages kept them out of State Department channels.) The continued activity suggests that the cancellation of the strikes on the 25th may have been only a cover for security purposes.

On the 28th Admiral Sharp cabled General Wheeler that his forces were ready and the weather was favorable for the strikes; he requested authority to initiate them on the 29th. General Wheeler responded with a message rescinding the previous cancellation, reinstating the original execution order, and approving the recommendation to execute on the 29th. The message informed Admiral Sharp that preliminary and planning messages should continue as SpeCat Exclusive for himself and the SecDef.

The strikes were launched on 29 June, reportedly with great success. The large Hanoi tank farm was apparently completely knocked out; the Haiphong facility looked about 80 percent destroyed. One U.S. aircraft was lost to ground fire. Four MIGs were encountered and one was probably shot down. The Deputy Commander of the 7th Air Force in Saigon called the operation "the most significant, the most important strike of the War."


The attack on North Vietnam's POL system was the last major escalation of the air war recommended by Secretary McNamara. Its eventual failure to produce a significant decrease in infiltration or cripple North Vietnamese logistical support of the war in the South, when added to the cumulative failure of the rest of ROLLING THUNDER, appears to have tipped the balance in his mind against any further escalation of air attacks on the DRV. As we shall see, a major factor in this reversal of position was the report and recommendation submitted at the end of the summer by an important study group of America's top scientists. Another consideration weighing in his mind must have been the growing antagonism, both domestic and international, to the bombing, which was identified as the pincipal impediment to the opening of negotiations. But disillusionment with the bombing alone might not have been enough to produce a recommendation for change had an alternative method of impeding infiltration not been proposed at the same time. Thus, in October when McNamara recommended a stabilization of the air war at prevailing levels, he was also able to recommend the imposition of a multi-system anti-infiltration barrier across the DMZ and the Laos panhandle. The story of this momentous policy shift is the most important element in the evolution of the air war in the summer and fall of 1966.

1. Results of the POL Attacks

a. Initial Success

Official Washington reacted with mild jubilation to the reported success of the POL strikes and took satisfaction in the relatively mild reaction of the international community to the escalation. Secretary McNamara described the execution of the raids as "a superb professional job," and sent a message of personal congratulation to the field commanders involved in the planning and execution of the attacks shortly after the results were in.

In a press conference the next day, the Secretary justified the strikes "to counter a mounting reliance by NVN on the use of trucks and powered junks to facilitate the infiltration of men and equipment from North Vietnam to South Vietnam." He explained that truck movement in the first half of 1966 had doubled, and that daily supply tonnage and troop infiltration on the Ho Chi Minh trail were up 150 and 120 percent. respectively, over 1965. The enemy had built new roads and its truck inventory by the end of the year was expected to be double that of January 1965, an increase which would require 50-70 percent more POL.

The Department of State issued instructions to embassies abroad to explain the strikes to foreign governments in counter-infiltration terms. The guidance was to the effect that since the Pause, the bombing of NVN had been carefully restricted to actual routes of infiltration and supply; there had been no response whatever from Hanoi suggesting any willingness to engage in discussions or move in any way toward peace; on the contrary, during the Pause and since, NVN had continued to increase the infiltration of regular NVN forces South, and to develop and enlarge supply routes; it was relying more heavily on trucking and had sharply increased the importation and use of POL. The U.S. could no longer afford to overlook this threat. Major POL storage sites in the vicinity of Hanoi and Haiphong were military targets that needed to be attacked.

The targets, the guidance continued, were located away from the centers of both cities. Strike forces had been instructed to observe every precaution to confine the strikes to military targets and there had been no change in the policy of not carrying out attacks against civilian targets or population centers. There was no intention of widening the war. The U.S. still desired to meet Hanoi for discussions without conditions or take any other steps which might lead toward peace.

The strikes made spectacular headlines everywhere. Hanoi charged that U.S. planes had indiscriminately bombed and strafed residential and economic areas in the outskirts of Hanoi and Haiphong, and called this "a new and extremely serious step." The USSR called it a step toward further escalation. The UK, France, and several other European countries expressed official disapproval. India expressed "deep regret and sorrow," and Japan was understanding but warned that there was a limit to its support of the bombing of NVN. Nevertheless, according to the State Department's scoreboard, some 26 Free World nations indicated either full approval or "understanding" of the strikes, and 12 indicated disapproval. Press reaction to the attacks was short-lived, however, and within' a week or so they were accepted as just another facet of the war.

Meanwhile in the U.S., following a familiar pattern of the Vietnam war, in which escalations of the air war served as preludes to additional increments of combat troops, Secretary McNamara informed the Joint Chiefs of Staff, the Service Secretaries and the Assistant Secretaries of Defense on July 2 that the latest revision of the troop deployment schedule had been approved as Program #3. The troop increases were not major as program changes have gone in the Vietnam war, an increase in authorized year-end strength from 383,500 approved in April to 391,000 and an increase of the final troop ceiling from 425,-100 to 431,000. But McNamara had personally rewritten the draft memo submitted to him by Systems Analysis inserting as its title, "Program #3." His handwritten changes also included a closing sentence which read, "Requests for changes in the Program may be submitted by the Service Secretaries or JCS whenever these appear appropriate." This language clearly reflected the following instruction that McNamara had received from the President on June 28:

As you know, we have been moving our men to Viet Nam on a schedule determined by General Westmoreland's requirements.

As I have stated orally several times this year, I should like this schedule to be accelerated as much as possible so that General Westmoreland can feel assured that he has all the men he needs as soon as possible.

Would you meet with the Joint Chiefs and give me at your early convenience an indication of what acceleration is possible for the balance of this year.

While the Chiefs were unable to promise any further speed-up in the deployment schedule, the Secretary assured the President on July 15 that all possible steps were being taken. But as in the air war, so also in the question of troop deployments a turning point was being reached. By the fall of 1966 when Program #4 was under consideration, the President would no longer be instructing McNamara to honor all of General Westmoreland's troop requests as fully and rapidly as possible.


In the air campaign strikes continued on the other major POL storage sites, and were soon accepted as a routine part of the bombing program. On 8 July, at a Honolulu conference, Secretary McNamara was given a complete briefing on the POL program. He informed CINCPAC that the President wished that first priority in the air war be given to the complete "strangulation" of NVN's POL system, and he must not feel that there were sortie limitations for this purpose. (He also stressed the need for increased interdiction of the railroad lines to China.) As a result, ROLLING THUNDER program No. 51, which went into effect the next day, specified a "strangulation" program of armed reconnaissance against the POL system, including dispersed sites. The ceiling for attack sorties on NVN and Laos was raised from 8100 to 10,100 per month.

McNamara left CINCPAC with instructions to develop a comprehensive plan to accomplish the maximum feasible POL destruction while maintaining a balanced effort against other priority targets. On July 24, CINCPAC forwarded his concept for the operation to Washington. In addition to the fixed and dispersed sites already under attack, he recommended strikes against the storage facilities at Phuc Yen and Kep airfields; against the DRV's importation facilities (i.e., foreign ships in Haiphong harbor, destruction of harbor dredges, destruction of docks, etc.); and the expansion of the reconnaissance effort to provide more and better information on the overall POL system. Also recommended was a step-up in attacks on rolling stock of all kinds carrying POL, and strikes on the Xom Trung Hoa lock and dam. In spite of this recommendation and a follow-up on August 8, ROLLING THUNDER 51 was only authorized to strike previously approved targets plus some new bridges and a bypass as outlined in the July 8 execute order.

While CINCPAC and his subordinates were making every effort to hamstring the DRV logistical operation through the POL attacks, the Secretary of Defense was keeping tabs on results through specially commissioned reports from DIA. These continued through July and into August. By July 20, DIA reported that 59.9% of North Vietnam's original POL capacity had been destroyed. By the end of July, DIA reported that 70% of NVN's large bulk (JCS-targeted) POL storage capacity had been destroyed, together with 7% of the capacity of known dispersed sites. The residual POL storage capacity was down from some 185,000 metric tons to about 75,000 tons, about 2/3 still in relatively vulnerable large storage centers--two of them, those at the airfields, still off limits--and 1/3 in smaller dispersed sites. This still provided, however, a fat cushion over NVN's requirements. What became clearer and clearer as the summer wore on was that while we had destroyed a major portion of North Vietnam's storage capacity, she retained enough dispersed capacity, supplemented by continuing imports (increasingly in easily dispersible drums, not bulk), to meet her on-going requirements. The greater invulnerability of dispersed POL meant an ever mounting U.S. cost in munitions, fuel, aircraft losses, and men. By August we were reaching the point at which these costs were prohibitive. It was simply impractical and infeasible to attempt any further constriction of North Vietnam's POL storage capacity.

As the POL campaign continued, the lucrative POL targets disappeared and the effort was confined more and more to the small scattered sites. Finally, on September 4, CINCPAC (probably acting by direction although no instructions appear in the available documents) directed a shift in the primary emphasis of
ROLLING THUNDER strikes. Henceforth they were to be aimed at, ". . . attrition of men, supplies, equipment and . . . POL Stressing the new set of priorities CINCPAC instructed, "POL will also receive emphasis on a selective basis." By mid-October, even PACAF reported that the campaign had reached the point of diminishing returns.

c. POL--Strategic Failure

It was clear in retrospect that the POL strikes had been a failure. Apart from the possibility of inconveniences, interruptions, and local shortages of a temporary nature, there was no evidence that NVN had at any time been pinched for POL. NVN's dependence on the unloading facilities at Haiphong and large storage sites in the rest of the country had been greatly overestimated. Bulk imports via oceangoing tanker continued at Haiphong despite the great damage to POL docks and storage there. Tankers merely stood offshore and unloaded into barges and other shallow-draft boats, usually at night, and the POL was transported to hundreds of concealed locations along internal waterways. More POL was also brought in already drummed, convenient for dispersed storage and handling and virtually immune from interdiction.

The difficulties of switching to a much less vulnerable but perfectly workable storage and distribution system, not an unbearable strain when the volume to be handled was not really very great, had also been overestimated. Typically, also, NVN's adaptability and resourcefulness had been greatly underestimated. As early as the summer of 1965, about six months after the initiation of ROLLING THUNDER, NVN had begun to import more POL, build additional small, dispersed, underground tank storage sites, and store more POL in drums along LOCs and at consumption points. It had anticipated the strikes and taken out insurance against them; by the time the strikes came, long after the decision had been telegraphed by open speculation in the public media, NVN was in good position to ride them out. Thus, by the end of 1966, after six months of POL attacks, it was estimated that NVN still had about 26,000 metric tons storage capacity in the large sites, about 30-40,000 tons capacity in medium-sized dispersed sites, and about 28,000 tons capacity in smaller tank and drum sites.

One of the unanticipated results of the POL strikes, which further offset their effectiveness, was the skillful way in which Ho Chi Minh used them in his negotiations with the Soviets and Chinese to extract larger commitments of economic, military and financial assistance from them. Thus, on July 17 he made a major appeal to the Chinese based on the American POL escalation. Since North Vietnam is essentially a logistical funnel for supplies originating in the USSR and China, this increase in their support as a direct result of the POL strikes must also be discounted against whatever effect they may have had on hampering North Vietnam's transportation.

The real and immediate failure of the POL strikes was reflected, however, in the undiminished flow of men and supplies down the Ho Chi Minh trail to the war in the South. In early July, the intelligence community had indicated that POL could become a factor in constricting the truck traffic to the South. The statement was, however, qualified,

The POL requirement for trucks involved in the infiltration movement has not been large enough to present significant supply problems. But local shortages have occurred from time to time and may become significant as a result of attacks on the POL distribution system.

By the end of the month, however, the CIA at least was more pessimistic:

Hanoi appears to believe that its transportation system will be able to withstand increased air attacks and still maintain an adequate flow of men and supplies to the South.

. . . Recent strikes against North Vietnam's POL storage facilities have destroyed over 50 percent of the nation's petroleum storage capacity. However, it is estimated that substantial stocks still survive and that the DRV can continue to import sufficient fuel to keep at least essential military and economic traffic moving.

DIA continued to focus its assessments on the narrower effectiveness of the strikes in destruction of some percentage of North Vietnamese POL storage capacity without directly relating this to needs and import potential. By September, the two intelligence agencies were in general agreement as to the failure of the POL strikes. In an evaluation of the entire bombing effort they stated, "There is no evidence yet of any shortage of POL in North Vietnam and stocks on hand, with recent imports, have been adequate to sustain necessary operations." The report went even further and stated that there was no evidence of insurmountable transport difficulties from the bombing, no significant economic dislocation and no weakening of popular morale.

Powerful reinforcement about the ineffectiveness of the strikes came at the end of August when a special summer study group of top American scientists submitted a series of reports through the JASON Division of the Institute for Defense Analyses (treated comprehensively below). One of their papers dealt in considerable detail with the entire bombing program, generally concluding that bombing had failed in all its specified goals. With respect to the recent petroleum attacks to disrupt North Vietnamese transportation, the scientists offered the following summary conclusions:

In view of the nature of the North Vietnamese POL system, the relatively small quantities of POL it requires, and the options available for overcoming the effects of U.S. air strikes thus far, it seems doubtful that any critical denial of essential POL has resulted, apart from temporary and local shortages. It also seems doubtful that any such denial need result if China and/or the USSR are willing to pay greater costs in delivering it.

Maintaining the flow of POL to consumers within North Vietnam will be more difficult, costly, and hazardous, depending primarily on the effectiveness of the U.S. armed reconnaissance effort against the transportation system. Temporary interruptions and shortages have probably been and can no doubt continue to be inflicted, but it does not seem likely that North Vietnam will have to curtail its higher priority POL-powered activities as a result.

Since less than 5 percent of North Vietnamese POL requirements are utilized in supporting truck operations in Laos, it seems unlikely that infiltration South will have to be curtailed because of POL shortages; and since North Vietnamese and VC forces in South Vietnam do not require POL supplied from the North, their POL-powered activities need not suffer, either.

Coming as they did from a highly prestigious and respected group of policy-supporting but independent-thinking scientists and scholars, and coming at the end of a long and frustrating summer in the air war, these views must have exercised a powerful influence on McNamara's thinking. His prompt adoption of the "infiltration barrier" concept they recommended as an alternative to the bombing (see below) gives evidence of the overall weight these reports carried.

McNamara, for his part, made no effort to conceal his dissatisfaction and disappointment at the failure of the POL attacks. He pointed out to the Air Force and the Navy the glaring discrepancy between the optimistic estimates of results their pre-strike POL studies had postulated and the actual failure of the raids to significantly decrease infiltration. The Secretary was already in the process of rethinking the role of the entire air campaign in the U.S. effort in Southeast Asia. He was painfully aware of its inability to pinch off the infiltration to the South and had seen no evidence of its ability to break Hanoi's will, demoralize its population, or bring it to the negotiation table. The full articulation of his disillusionment would not come until the following January, however, when he appeared before a joint session of the Senate Armed Services and Appropriations Committees to argue against any further extension of the bombing. To illustrate the ineffectualness of bombing he cited our experience with the POL strikes:

There is no question but what petroleum in the North is an essential material for the movement, under present circumstances, of men and equipment to their borders. But neither is there any doubt that with, in effect, an unrestricted bombing campaign against petroleum, we were not able to dry up the supply.

The bombing of the POL system was carried out with as much skill, effort, and attention as we could devote to it, starting on June 29, and we haven't been able to dry up those supplies.

We in effect took out the Haiphong docks for unloading of POL and we have had very little effect on the importation level at the present time. I would think it is about as high today as it would have been if we had never struck the Haiphong docks. And I think the same thing would be true if we took out the cargo docks in Haiphong for dry cargo.

I don't believe that the bombing up to the present has significantly reduced, nor any bombing that I could contemplate in the future would significantly reduce, actual flow of men and materiel to the South.

Thus disenthralled with air power's ability to turn the tide of the war in our favor, McNamara would increasingly in the months ahead recommend against any further escalation of the bombing and turn his attention to alternative methods of shutting off the infiltration and bringing the war to an end.

Go to the Next Section of Volume 4, Chapter 1, "The Air War in North Vietnam, 1965-1968," pp. 1-276.

Glossary of Acronyms and Terms

Go to Volume 1, Chapter 1 of the Pentagon Papers, "Background to the Conflict, 1940-1950." pp. 1-52

Go to Volume 1, Chapter 2 of the Pentagon Papers, "U.S. Involvement in the Franco-Viet Minh War, 1950-1954," pp. 53-107

Go to Volume 1, Chapter 3 of the Pentagon Papers, "The Geneva Conference, May-July, 1954," pp. 108-178.

Go to Volume 1, Chapter 4 of the Pentagon Papers, "U.S. and France in Indochina, 1950-56," pp. 179-241

Go to Volume 1, Chapter 5 of the Pentagon Papers, "Origins of the Insurgency in South Vietnam, 1954-1960," pp. 242-314

Go to Volume 2, Chapter 1 of the Pentagon Papers, "The Kennedy Commitments and Programs, 1961,"pp. 1-127

Go to Volume 2, Chapter 2 of the Pentagon Papers, "The Strategic Hamlet Program, 1961-1963," pp. 128-159.

Go to Volume 2, Chapter 3, of the Pentagon Papers, "Phased Withdrawal of U.S. Forces, 1962-1964," pp. 160-200.

Go to Volume 2, Chapter 4, of the Pentagon Papers, "The Overthrow of Ngo Dinh Diem, May-November, 1963," pp. 201-276.

Go to Volume 2, Chapter 5 of the Pentagon Papers, "US-GVN Relations, 1964-1967," pp. 277-407.

Go to Volume 2, Chapter 6 of the Pentagon Papers, "The Advisory Build-up, 1961-67," pp. 408-514

Go to Volume 2, Chapter 7 of the Pentagon Papers, "Re-Emphasis on Pacification: 1965-1967," pp. 515-623.

Go to Volume 3, Chapter 1 of the Pentagon Papers, "U.S. Programs in South Vietnam, Nov. 1963-Apr. 1965," pp. 1-105.

Go to Volume 3, Chapter 2 of the Pentagon Papers, "Military Pressures Against North Vietnam, February 1964-January 1965," pp. 106-268.

Go to Volume 3, Chapter 3, of the Pentagon Papers, "The Air War in North Vietnam: Rolling Thunder Begins, February-June, 1965," pp. 269-388

Go to Volume 3, Chapter 4, of the Pentagon Papers, "American Troops Enter the Ground War, March-July 1965," pp. 389-485

Go to Volume 4, Chapter 1, of the Pentagon Papers, "The Air War in North Vietnam, 1965-1968," pp. 1-276.

Go to Volume 4, Chapter 2, of the Pentagon Papers, "U.S. Ground Strategy and Force Deployments, 1965-1968," pp. 277-604.

Return to Vinnie's Home Page

Return to Vietnam War Page