War of Numbers

by Sam Adams

  The Manpower Branch received my critique of Snider’s paper on 27 December 1971. It cavilled once more at the absence of guerrillas, then dwelt on omissions in the main and local forces. The critique noted, for example, that Snider had included some main forces soldiers who belonged to regiments, but none in the myriad smaller units such as battalions and companies. The shortchanging of small units, the critique observed, was a device that the U.S. command in Saigon had used on the Vietcong estimate before Tet. In early January, at Tate’s behest, Snider wrote a short rebuttal to the critique, which never left the building. That ended the matter.

  With nothing else to do, I returned to my history of the Cambodian rebels. I traced them back to 1947 and discovered that in 1954, some three thousand had gone to Hanoi for training. I also found that a full-scale communist-run rebellion had started in Cambodia in 1968, and that the head of the Cambodian rebels was an obscure Khmer named Pol Pot. None of the facts had been known at the time, not even that the VC called the Cambodian rebels the Red K, and that the Cambodian rebels called the VC, Friend Seven. Increasingly glum, I turned in my history in February 1972. Its concluding paragraphs ended with this prediction:

  Finally there is the question of the party’s position in the communist world. Here the evidence strongly suggests that the Cambodian party, although formed in North Vietnam, had a mind of its own. There are already clear indications that its leaders have flirted with China to offset Hanoi … Although the Khmers’ short-term interests are clearly with Hanoi, the party leaders probably regard their dependence as temporary.

  In fact, what evidence there is points to friction in the future. A KC document of late last year was already referring to the “Vietnamese problem.” A more recent document talked of difficulties the Khmer hierarchy was having with Friend Seven … And a COSVN assessment of October 1971 suggested that in some areas relations between the Cambodian and Vietnamese communists had grown “steadily worse.”

  Whatever the problems may become eventually, they seem unlikely to get out hand in the immediate future. The two parties have more pressing near-term goals … to gain Saigon and Phnom Penh, respectively, but it would not be surprising if at some more distant time, the ancient hatred between the Khmers and the Vietnamese publicly reemerges in the trappings of communist dialiectic.

  Those were the last words I wrote that were published by the agency. The events they foretold have since come to pass. After their publication, I was told to work on Communist China.

  On 13 March, however, a fresh CIA memorandum arrived on my desk again fixing the Khmer numbers at 15,000 to 30,000. I called Snider to ask why the range was unchanged. He replied that the Manpower Branch had stopped research on Cambodian numbers after the publication of his paper in November. He also mentioned in passing that as far as he knew no one in CIA headquarters kept a card file on enemy units. Hoping that a new volume of evidence might reopen the issue of size, I fished from my desk a cardboard box and a stack of three-by-fives, and reported to Bud’s cubicle to ask if I could work upstairs for a while on “filing.” Busy as usual at his typewriter, he replied absently that it was OK with him. That afternoon, I sat down at an unoccupied desk in the Cambodian Section of the DDI’s Office of Current Intelligence, filling out index cards on communist units. (Since I still had to do my job downstairs with the staff, the project was slow going, and took upwards of a month.)

  On 31 March, the CIA sent a memorandum to the White House stating that the Vietnamese communists were unlikely to launch a really large-scale offensive in South Vietnam in the next few months. The memo, which followed an earlier CIA brief to the White House that the main Vietcong attack had fizzled in February, once more put the Khmer numbers at 15,000 to 30,000. The top end of the range was now nine months old. The next day Hanoi launched its Easter offensive in South Vietnam, its biggest since Tet 1968. Apparently sure that the Khmer rebel army was large enough to hold its own, the Vietnamese communists began to march their combat units from Cambodia to South Vietnam. As the Vietcong bolted through the Vietnamese countryside, I started to shuffle my three-by-fives in preparation for another study on enemy strength in Cambodia.

  “What are you trying to do—get all three of us fired?” Bud asked, evidently referring to himself, Mr. Pontiac, and me. I had just handed him the study based on the card file. The study listed by title an aggregation of Khmer communist units unaccounted for in the official estimate (and also noted that some forty Vietcong battalions were absent from the lists.) Despite his perturbation, Bud sent the study at once to the DDI front office. It responded shortly by instructing him to stop me from working on Indo-China altogether; and then it killed the memo. That evening I found that it had also packed me off to bureaucratic Coventry. A researcher from the Office of Current Intelligence telephoned my home to tell me—morose already—that Mr. Lehman, the office head, had passed down the official word to OCI analysts—some of whom had helped on the card file—to stop abetting my endeavors.

  *On the track of a particular communist unit, I once called the Defense Intelligence Agency for help. A Captain White answered the phone, saying he worked with Sergeant Reisman. He didn’t know about the unit, but when routinely asked how DIA arrived at its Cambodian estimate, the captain replied, “First they give you the number, then they tell you to prove it.” He did not elaborate.

  9 THE CROSSOVER POINT

  THE EASTER OFFENSIVE was the last straw for me. As usual, it was a complete surprise, and I decided that something, goodness knows what, had to be done about American intelligence. I knew that my old friend, Gains Hawkins, had retired to Mississippi, so I determined to track him down. Eventually, I discovered he had bought a house in the small farming town of West Point. I called him up, he said he’d be glad to see me, and I flew down from Virginia. On a hot June afternoon in his backyard, he told me how, right before the Saigon conference of September 1967, he’d been given a number for the OB—and told not to go above it. He felt that the number he was given was unreasonably low, and that General Westmoreland was probably the person behind it. Hawkins agreed that an investigation was in order, and that Congress ought to do it. He particularly recommended the head of the Senate Armed Services Committee, John Stennis of Mississippi. Hawkins was taking a big risk. In agreeing to talk to the Senate, he was putting his pension on the line. Retirees are subject to the Uniform Code of Military Justice.

  I returned to Virginia, and later in the year gave Stennis’ committee a thirteen-page paper that listed names, dates, and sequences of events. A staff assistant told me it was an interesting document, but he doubted that the Intelligence Subcommittee would take it up because it hadn’t met in over a year and a half. I gave up on the Senate and tried the House. The House was likewise unresponsive.

  Despairing of Congress, I tried the U.S. Army and CIA Inspector General. Neither would investigate my charges.

  With the arrival of additional reports that the Khmer communist army had continued to grow, I called upon the office of the CIA’s Inspector General to lodge a formal complaint about the agency’s research on enemy strength in Cambodia. The date was 4 December 1972. Scott Breckenridge, who had handled some earlier grumbles of mine to the inspector about the mischief over the Vietcong estimate before Tet, scribbled notes as I related my story about Cambodia. Mr. Breckenridge later informed me that William Colby, then the CIA’s executive director, had learned of my complaint, and said, “Let the chips fall where they may.” It was the last I heard of the matter.*

  This curious story of numbers poses several questions, the first of which is whether it really mattered. The answer of course depends on one’s point of view, but for American intelligence it must be yes. Because some time back—in McNamara’s day, I suppose—the U.S. government decided to measure its wars statistically. It follows that when the numbers are cockeyed, the conclusions drawn from them tend to be skewed as well. Had the computers gotten the data of late 1970, for example, that the Khmer rebel army was growing rapid
ly, we might have concluded far sooner than we did that the conflict in Cambodia was fast becoming a civil war, not unlike the one in South Vietnam.

  A second question is why anyone bothered at all to cook the Cambodian books. Although in the case of the Vietcong numbers before Tet everyone concerned knew that the main reason for the adjustments was to keep aglow the light at the end of the tunnel, the motives for doctoring the Khmer estimate are more obscure. Certainly the agency had reason to hide its fifteen-month delay in asking whether our Cambodian foes had recruited an army; but as a sole cause, it seems farfetched. Perhaps the real motive lies hidden in the rubble of the American policy for Cambodia, which held that the problem there was caused by Vietnamese, not native, communists. This was an argument Washington used to justify the bombing of Cambodia.

  A third question is who was responsible. In the absence of a proper inquiry, it’s impossible to say. But clearly the agency officials with whom I normally dealt were anything but sinister. Snider, for example, was a cheerful and usually candid young analyst from southern Virginia, who sometimes slipped me documents supporting higher numbers, and who often complained about what he thought was Congress’s airy neglect of the CIA. Bud was a quiet, careful, Far Eastern scholar whom I considered more a protector than adversary, and who, when adrift from his typing machine, was a sometime poet. (Once, on being asked to display an example of his scholarship in the agency library, he submitted a literary review opened to one of his poems; on the page opposite was some verse by Robert Lowell.) Mr. Pontiac was a dignified, rather kindly man—when not assigning penance on weekends—who in the previous decade had helped write some of the CIA’s darker reports on Vietnam, which made their way into the Pentagon Papers.

  Of the agency’s higher reaches I was scarcely in a position to judge from my post at the bottom of the well. But I knew that Messrs. Helms, Carver, Proctor, Walsh, and Lehman had all involved themselves, one way or another, in the CIA’s acceptance of Westmoreland’s suspect numbers before Tet. So for them to diddle once again with an estimate—this one on Cambodia—would not be out of character. I have no proof, however, and in any case am unpersuaded that they were villains in the grand manner. I doubt, for example, that any one of them snarled, “OK, Tate, go falsify those books.” Rather, there must have been a series of nervous consultations in which numbers popped up and down like soybean futures, broad hints passed, and finally, guilt sloughed downwards.

  So the question of accountability seems to me unresolved. Perhaps the trouble is that the answer is too diffuse, and that the best explanation I’ve heard of the problem came from one of the agency’s burnt-out elders on Indo-China.

  His name was Edward Haskins, and he had a grey crew cut, perhaps left over from World War II when he was with the U.S. mission in the hills of China with Mao. In the first half of the sixties he had run one of the CIA’s research groups that had repeatedly warned of our deepening Vietnam commitment. In 1966, he had despaired of being listened to, stopped working on Indo-China, and gone on to another assignment.

  I forget when it was that he made his remarks about the problem of responsibility. As I recall, it was during a conversation in which he was recounting examples he’d seen in the last decade or so of official humbug. He mentioned the faked progress reports on Diem’s strategic hamlets, the whoppers the Air Force told when it first started bombing the Trail, and the habit that senior editors had of changing modifiers and dropping paragraphs to water down predictions. But what struck him most, he said, was not that there were scoundrels topside (this he assumed) but that it took so many people to practice the deceits. Because like all else in the U.S. government, tampering with evidence is a cumbersome thing to do. The generals have to pass the word to the colonels, the colonels to the majors and lieutenants, the lieutenants to the corporals. Then the supervisors must sort it all out, and the analysts and statisticians somehow glue together the wildly misshapen parts. Finally, the secretaries type it up, and the clerks store it away in the archives. And there it stays until months or years later (always too late to do any good) when some misfit complains aloud.

  At this I joined in and the conversation spun toward its logical end. It seemed to us—I paraphrase—that the whole country, not just the government, had laid aside its normal pursuits and danced off to disport itself in a puddle of flummery, that we had become a nation of pettifoggers, of smalltime tricksters, a padded Lilliput whose citizens had simultaneously forgotten how to tell the truth. That the itch to equivocate had become as widespread and as irresistible as the temptation to fudge on taxes—and so on and so forth.

  At last it must have seemed to Mr. Haskins that the talk had crossed the line from the pompous to the goody-goody, because he leaned back in his chair and burst out laughing. “Maybe,” he said, smiling broadly as if the problem was certain to evaporate, “maybe they’ve put something in the water.”

  By mid-January 1973 I had reached the end of the road. I happened to read a newspaper account of Daniel Ellsberg’s trial in Los Angeles, and I noticed that the government was alleging that Ellsberg had injured the national security by releasing estimates of the enemy force in Vietnam. I looked, and damned if they weren’t from the same order of battle that the military had doctored back in 1967. In late February I went to Los Angeles to testify for the defense. Naturally, it was oveijoyed to see me. When its lawyers heard my story, however, they decided to send for Colonel Hawkins. I knew this was more than the colonel had bargained for when he had agreed to an investigation. I tried vainly to dissuade them.

  When Hawkins arrived in California a short while later, he looked at me as if I had betrayed him. “Stennis,” he said, “not Ellsberg.” I shared the colonel’s distaste for Ellsberg. (The first I had heard of the defendant was from George Allen who had recalled Ellsberg’s emotional pleas in Saigon in 1965 to send more troops to Vietnam. “Every year a new hobbyhorse,” said George.) I felt my position was untenable. In volunteering for Ellsberg, I was now jeopardizing the pension of a person who had gone out on a limb for me.

  The matter came to a head as I was preparing to mount the witness stand. The colonel came up to me and said: “Sam, this is the end of our friendship.”

  I pleaded: “Please, Colonel, not that.”

  He hesitated for perhaps ten seconds and said: “Well, I suppose you know what you’re doing. Go on up there and give ’em hell.” We shook hands. The rest of the trial was an anticlimax.

  When I returned to Washington in March, the CIA once again threatened to fire me. I complained, and, as usual, the agency backed down. After a decent interval, I quit. The date was 17 May 1973. It was the first day of Senator Sam Ervin’s hearings on Watergate.

  *Broadcasting by radio from Hanoi on 9 April 1973, Prince Sihanouk stated that the number of soldiers in the Khmer rebel army’s “offensive” units, presumably meaning those in the main and local forces, had “now reached 120,000 men.” On receiving word of the broadcast, the CIA announced in its daily bulletin on Indo-China that Sihanouk’s claim was clearly exaggerated since the U.S. official estimate then stood at 40,000.

  This estimate of the rebel’s strength clearly sprung from the mold set on 22 June 1971. It excluded guerrillas and service troops, who numbered in the many additional tens of thousands, and failed to account for legions of the main and local forces it purported to include. A study from the field put the number of such combatants as high as 90,000, not all that far from Sihanouk’s claim. But the study (which also omitted guerrillas and logisticians) remained unofficial. The true size of Khmer forces could have exceeded 200,000 by a considerable margin.

  APPENDIX

  Publisher’s note: After several years of research into the military’s side of the OB controversy Adams wrote an outline of what he believed had happened. The text which follows—“A Number to Live With”—was apparently written about 1980 or 1981. It is unsupported by footnotes or other source references, but the factual claims made in it were the substance of the CBS-
Westmoreland trial and are all discussed repeatedly and are abundantly supported in the trial transcript. All are discussed, as well, in one of the books about the trial, Vietnam on Trial: Westmoreland vs. CBS (Atheneum, 1987), a thorough and carefully written account by Bob Brewin and Sydney Shaw. With the exception of the four officials named in the first paragraph, Adams interviewed all those cited in the text of this appendix. Adams believed that the OB was falsified in Saigon for political reasons, since there was no way to change the OB without informing the press, and there was no way to double the Vietcong forces in the OB while continuing to claim the sort of success required to continue the war. But Adams felt it was unfair to lay the blame for this falsification solely on General Westmoreland, whom he grew to like in the course of the trial. In Adams’s view MACV as a whole, Westmoreland included, was coerced by officials in Washington to suppress anything that might be interpreted as bad news. Adams believed that these Washington officials understood what was at stake in the OB controversy and deliberately exerted the pressure which resulted in falsification of the numbers. What mattered to Adams was not identification of those responsible, but reaffirmation of the importance of the integrity of intelligence. The whole sordid web of deceit which followed MACV’s refusal to accept an increase in the OB was the result of trying to make the evidence support the conclusion. In Adams’s view, all might have been avoided—the controversy, the trial, conceivably even the military disaster of the Tet offensive—if only the CIA had insisted on telling, and officials had been willing to listen to, the truth.

  A Number to Live With

  WHAT FOLLOWS IS A DETAILED ACCOUNT of how General Westmoreland’s intelligence staff—with White House encouragement—falsified the Vietcong strength estimates before the communist Tet Offensive of January 1968. It is by far the most heavily researched portion of the book. My sources include forty military and twelve civilian intelligence officials, voluminous files of official reports, and other correspondence, such as letters home. I have yet to interview the four persons still living whom I believe chiefly responsible for the falsification: General Westmoreland himself, General Philip Davidson (Westmoreland’s J-2 after McChristian), Mr. Robert S. McNamara, and Mr. Walt W. Rostow. I plan to approach them before the book goes to press, in the hope that they will shed further light on what happened, including the extent of President Johnson’s involvement.