War of Numbers

by Sam Adams

  As already noted, MACV discovered its vast underestimate of Vietcong numbers in late 1966. Westmoreland’s then J-2, General Joseph A. McChristian, although embarrassed, admitted his error, and by early 1967 was pressing for a higher order of battle. At this point, the main resistance against one came from the Pentagon, including the office of the secretary of defense, Robert S. McNamara. As McNamara explained to an aide in late January, he realized the official OB was all wrong, but that he was not yet prepared to tell Congress. He meant what he said. On 6 March 1967, he briefed a Congressional committee using the official numbers, the same ones he knew to be low.

  McChristian’s response to the Pentagon’s foot-dragging was to adopt a second set of books. Kept informally by his OB chief, Colonel Gains B. Hawkins, the second set listed the lesser three of the OB’s four parts. The total for its most important component—the VC’s main battle forces—remained public knowledge. To MACV strength analysts (mostly unaware that a controversy existed) the compromise was satisfactory. None of them felt much pressure to raise or lower their numbers for any reason other than evidence.

  Incredibly, General Westmoreland during this period seems not to have grasped what the full public impact might be of the higher numbers. Furthermore he had neglected to add them all up. He received his first detailed briefing on the second set of books, with their big sums totaled, in May 1967. General McChristian and Colonel Hawkins conducted the briefing in Westmoreland’s private office. Using a flipchart, they reviewed the OB’s four components one by one, and when they reached the bottom line on the flipchart’s last page, Westmoreland—according to my source—“almost fell off his chair.” “What will I tell Congress?” he gasped. “What will I tell the press?” On recovering, some minutes later, he turned to McChristian and said: “General, I want you to take another look at those numbers.”

  McChristian took this as a suggestion to tamper with the second set of books. This he refused to do. He was sent home on 1 June 1967, protesting vigorously. At least some of his chagrin arose from the fact that his replacement was his archrival in the Army, an old West Point classmate, General Philip Davidson.

  Davidson was more amenable than McChristian to manipulating the unofficial books. Among his first acts on 1 June was to lobby with the agency’s DDI representative in Saigon to drop from the OB one of its main subcomponents, the so-called self-defense militia (or tu ve). Davidson’s suggestion flew back and forth between Saigon and Washington for over a month. It came up at a meeting between Westmoreland and McNamara in Vietnam on 9 July, and again when the general and the secretary saw President Johnson at the White House on the thirteenth. Exactly what transpired at these meetings I have yet to find out, but what happened thereafter is clear.

  MACV strength analysts began to suspect that someone was doctoring the order of battle. Among the first to harbor this suspicion was Lieutenant Joseph Gorman, chief analyst for the VC main battle forces in IV Corps, which comprised South Vietnam’s southernmost and most populous quarter. One of his jobs was to warn J-2 headquarters each time he discovered a new VC unit, so that J-2 could add it to the OB. During this period, however, he found the headquarters increasingly reluctant to enter new units on the lists. At first he thought that J-2 had tightened its “acceptance criteria,” but as the summer wore on its reasons for disallowing new units became more and more frivolous. One VC battalion was turned down by J-2 because Gorman’s request form had a typographical error; another because the form’s cover sheet was not centered; a third because the sheet lacked the proper red-pencil markings. A second analyst, Lieutenant Richard McArthur—assigned in June to keep track of VC guerrillas countrywide—wrote his parents on 26 July that he had found that the guerrilla number for II Corps was “completely false,” and that J-2 was “feeding people nonsense figures with no documentary evidence.” He added, “I can’t believe half the things I’m digging up.”

  Meanwhile, pressure continued to build on the order of battle. At the Board of National Estimates conference on Fourteen Three—convened in Langley in June—the CIA was still insisting on higher numbers. By August, its sessions had reached an impasse, and the principals had agreed to meet at an OB conference at Westmoreland’s headquarters in early September—with CIA, DIA, and MACV attending. The purpose of the conference was to come to an agreement over VC strength.

  MACV’s preparations for the conference were both above board and below it. In one of the war’s most unusual messages, dated 20 August, MACV deputy General Creighton Abrams cabled Washington, with Westmoreland’s approval, the old request to drop the self-defense militia from the OB. What made the cable so extraordinary was the frank reason he gave for wanting to do so. To leave the militia on the lists, he explained, would contradict the “image of success” MACV had been lately building, and would provoke the press into drawing “an erroneous and gloomy conclusion” over the progress of the war. The message was widely distributed in official Washington.

  MACV’s below-board measures were also unusual. No longer content to exclude units from the OB, J-2 now began to cut down the size of units already in it. Marshall W. Lynn, a lieutenant charged with keeping tabs on six regiment-sized VC formations near Saigon, has explained how it was done. One morning shortly before the start of the scheduled conference, a colonel from J-2 stopped by Lynn’s desk with the suggestion that the strengths at which Lynn was carrying his six VC units were “way too high.” Lynn denied it, at which the colonel simply picked up Lynn’s strength sheet, crossed out the numbers by each regiment, and penciled in new ones, on the average one-third lower. To Lynn’s amazement, a unit which he had carried with 3,100 men became “1,900” instead. As for J-2’s acceptance criteria for new units, Gorman remarked that by early September, “you could march a VC regiment down the hall, and they wouldn’t put it in the OB.”

  The conference, which I describe at length in Chapter 5, ended with the CIA caving in on the first day, 11 September. And the order of battle, instead of doubling—to reflect the evidence discovered in late 1966—actually fell, from about 290,000 to just over 240,000. As Chapter 5 points out, the drop was accomplished by marching the subcomponent, the self-defense militia, from the lists (as well as another whole category, the so-called political cadres) and by a general acceptance of J-2’s “scaled down” numbers. Among those dismayed by the proceedings was DIA’s chief order-of-battle analyst, Captain Barrie Williams. Captain Williams felt “the whole session was painful. It was clear we were double-dealing.”

  Thinking that the conference had laid the VC numbers dispute to rest, official Washington was pleased with the result. Almost at once, however, new problems replaced the old ones. On his return from Saigon to his desk at the Pentagon, DIA’s Captain Williams got his first whiff of what the new problems were. An order had arrived from Secretary McNamara’s office saying that McNamara wanted the newly agreed-upon OB number of some 240,000 to be “retroactively readjusted.” Apparently the secretary needed the VC strength figures for previous years so he could plot them on the same graph with the latest sum. Completely at a loss on how to “retroactively readjust” a number he thought fake to begin with, Williams went to his boss, Colonel John Lanterman, for advice. Lanterman was sympathetic. He felt that since MACV J-2 had come up with the number of 240,000, it ought to produce the earlier numbers as well; and he dropped the problem in MACV’s lap. Williams flew back to Saigon to monitor what they did with it.

  The “retroactive readjustment” took place in the third week of September in a conference room outside General Davidson’s office. The readjustment was simple. A J-2 officer chalked a curve on a blackboard.* On the right-hand end of the curve he wrote the number agreed upon at the Saigon conference—241,200—and the date of the last official OB, August 1967. Then he wrote “August 1966” at the top of the curve, and “August 1965” at its left-hand end. He stepped back to look at the curve from a distance, then returned to the board to write “285,400” next to “August 1966” and “204,700�
�� next to “August 1965.” This done, someone suggested that McNamara might want the OB’s components (now three in number) readjusted as well. The components were figured out so that they would of course add up to the totals on the curve, the J-2 officer chalked them in, and Captain Williams copied them down for the secretary’s benefit. The numbers from the board were as follows:

  They henceforth became the official U.S. intelligence estimate for enemy forces in South Vietnam for the times mentioned, and became the basic starting point from which later J-2 estimates were derived. The J-2 officer assigned to take them over was a West Pointer, Lieutenant Colonel Everette S. Parkins, head of a branch within the Order of Battle Section called OB Studies. On 25 September (about a week after Captain Williams’ departure from Saigon with the readjusted numbers) Parkins was given a deputy to help him out with his computations. The deputy was a U.S. navy commander by the name of James A. Meacham.

  As might be expected, demands for the official numbers started almost at once. In early October, for example, the American Embassy in Saigon, in preparing a message for the State Department, asked MACV J-2 for the latest figures on VC strength. OB Studies supplied them and the embassy cable—classified “Secret, Limited Distribution Only,” and addressed to Dean Rusk with special instructions that he pass it to Rostow—noted that our effort in Vietnam was making “solid progress.” To back up the assertion, it cited, among other facts, the drop in VC main force strength from “126,000 in August 1966 to 118,000” a year later—statistics clearly gotten from the curve on Davidson’s blackboard. The cable was dated 7 October 1967, and that evening, the man who supplied its numbers, the newly assigned Commander Meacham, wrote his wife: “Once in a while I sit back—mentally—and consider the place and the rules of the game objectively, and then I look around for the Mock Turtle and the March Hare.”

  Meanwhile, pressures were building once again on the order of battle. As usual, they built from opposite directions.

  On one side the pressure was exerted from Washington. It came from a series of meetings convened in October in the Executive Office Building of the White House by President Johnson’s main Vietnam advisor, Walt Rostow—the same Rostow to whom Rusk had passed the embassy’s cable of 7 October. The Gallup Poll readings for Johnson’s handling of the war had reached their lowest point thus far, and the meetings (called Measurements of Progress) were designed to come up with information that showed that the Vietnam struggle had gone better than was generally thought. Thus when the OB question came up at the meetings, the context was not whether the numbers were too low, but whether they were too high. Rostow took the latter position, and as one of his aides later confided to a friend, “we really leaned on the OB.” Contrary information was discouraged. When the CIA’s George Allen, who’d always thought the OB was too low, mentioned unfavorable statistics, Rostow complained, “I’m sorry you won’t support your president.”

  On the other side the pressure came from Vietnam. There the Vietcong, preparing for their Tet Offensive, were engaged in the greatest buildup of the war. Their preparations included the formation of hundreds of new units—battalions, companies and platoons—and the beefing up of units already in existence. Although MACV strength analysts were well aware of the buildup, it was unreflected in the OB. The policy of J-2 to disallow new formations in the OB was in effect.

  In November the pressures increased. President Johnson’s Success Offensive was now in full swing, and there were almost daily briefings on how well the war was going. On 11 November for example, Westmoreland gave Davidson a surprise order to brief the press in Saigon, and Meacham was pulled from lunch to fill Davidson’s briefing folder. Among the papers he put in the folder was a graph showing the curve drawn on the general’s blackboard in September. Davidson called in the press at two in the afternoon, and the next day the New York Times reported on page one that “US Aides Say Foe Is Weakening Fast,” noting in the text that VC strength had fallen from 285,000 to 242,000 men. Westmoreland used the same numbers at a larger briefing at the Pentagon on 22 November with similar results. Earlier in the month, Davidson’s curve was also used to brief President Johnson’s “wise men”*—eleven distinguished citizens brought together to advise him on the war. On hearing the good news, the wise men concluded that the war was proceeding satisfactorily. They advised the president to keep fighting.

  On 27 November, an odd cloud appeared on the horizon. It was a memo sent to Washington from Saigon, composed at the CIA Station by its Collation analyst, Joe Hovey. In his daily reading of captured documents, Hovey had noticed the VC were up to something big, but he couldn’t tell just what. When he finally put the documents together, it became clear. As his memo put it, the Vietcong were planning “a political and military offensive utilizing all VC assets,” its target to include “all major cities” in South Vietnam. The person who typed his memo was the Collation secretary, Betsy Gibson.

  Hovey’s paper created a modest stir. For almost three weeks it travelled around CIA headquarters, gathering comments. It went to the White House on 15 December and President Johnson read it the next day. But as comments on it pointed out, the paper always raised the same question: Why would the VC be planning an offensive “utilizing all their assets” when it was clear from the OB that their “assets” were going rapidly downhill? The most sensible answer, although none too convincing, was that the VC had become suicidal. In any case, five days after reading the memo, President Johnson explained in a briefing that the VC seemed to be planning a “kamikaze attack.”

  New Year’s passed, with MACV’s Order of Battle Section awaiting the go-ahead to finish its year-end OB—the one dated “31 December.” On about 3 January, the OB Section chief, Marine Colonel Paul Weiler (who had replaced Colonel Hawkins in mid-September) received the OK from General Davidson’s office, along with instructions to lower the estimate—which had dropped to 235,852 at the end of November—by 11,000 or so more men. Having already shaved a thousand-odd soldiers from the VC main forces, Weiler had about 10,000 more to go. He informed Meacham—now chief of OB Studies instead of Parkins—and their prearranged plan went into effect. If the estimate was to drop, they had decided, the component to absorb the loss would be the one they thought was the OB’s “softest”—the guerrillas. They had already programmed the main computer to pare down guerrillas by province, to whatever loss total Davidson decreed. The appropriate buttons were pushed. The computer made the cuts: 500 guerrillas from Quang Nam, 50 from Quang Ngai, 400 from Long An, 450 from Vinh Long, and so forth. The results were entered into the order of battle. Thus MACV’s year-end estimate for VC strength became 224,651. It was the last monthly order of battle before the Tet Offensive.

  Among those oblivious to the proceedings was MACV’s guerrilla analyst, Lieutenant Richard McArthur. Never shown the official OB, he kept track of VC guerrillas by province on a plexiglass toteboard next to his desk. He based his entries on reports from the field, and updated them with captured enemy documents. His toteboard showed VC guerrilla strength going up. McArthur’s morale was low, and he was drinking a lot. Many of his OB colleagues were also unhappy, and for many reasons. Not the least of them was that MACV now kept, not two sets of books, but three.

  The first set was the official OB, sent to Washington—and used in Saigon by such nonmilitary analysts as Joe Hovey. The second set belonged to the OB analysts themselves. It tended to be informal, displayed on maps, which showed increasing numbers of battalions that were labeled “unofficial.” The third set, an amalgam of the first two, went to General Westmoreland, and was very tightly held. (One of its keepers was an Army lieutenant colonel named David Morgan. I have yet to approach him, but I know where he is—Bonita, California, retired. He is said to be ashamed of his role and may be willing to talk. But maybe he won’t, fearing the loss of his pension.)

  If January was a bad month for Richard McArthur, it was even worse for Joe Hovey. He had written what he felt was the most important paper of his life in la
te November and he was on pins and needles whether his prediction would come about. He was in an increasing quandary. All the most recent evidence confirmed his forecast of an enormous countrywide offensive—all the evidence, that is, except the order of battle. “How’ll they pull if off?” he kept asking himself; “Where’ll they get the men?”

  The Tet Offensive started on 30 January 1968. It was a big attack all right, but to Hovey’s puzzlement, all the action was in the northern half of the country. The southern half was quiet. That evening most American intelligence officials in Saigon went to bed at their usual times and places: Colonel Weiler and Commander Meacham to their rooms in the senior officers’ BOQ near Tan Son Nhut; Joe Hovey to his lodgings downtown. None was aware that several Vietcong battalions had already stolen into the city or that VC sappers were screwing detonator caps onto bombs they hoped to use to blow their way into the United States Embassy.

  The VC sappers attacked at 2:47 in the morning. The Saigon docks erupted at 3:58. By 5:00 A.M. guns were firing all over the city, Hovey was wondering whether to hide under his bed. Weiler and Meacham—neither of whom had remotely suspected a big offensive was in the offing—were manning a machine gun on the BOQ roof. To his horror, Meacham discovered that the machine gun’s ammunition was all tracer. To shoot tracers was to reveal their firing position. Weiler said: “If we shoot this damn gun, we’ll get a B-40 rocket smack in the puss.” It takes only one man to fire a B-40 rocket. As chief of the MACV Order of Battle Section, Weiler’s scope of interest in enemy strength had considerably narrowed …