War of Numbers (25 page)

All this was interesting, but the time had come to pack. I began loading my files into a shopping cart. As I did so, the “VC Winter-Spring Campaign” folder caught my eye. I skimmed the station cable that had prompted the folder’s start. Its date was 24 November 1967, and out popped the key phrases:
All-out offensive; January to March 1968;
and
urban centers.
⁶
My Lord, the message was almost ten weeks old, but whoever’d written it was right on the button. I asked Tom Becker, who’d been in Vietnam in late November, if he knew the author.

“Joe Hovey,” said Becker. “He wrote it on Thanksgiving Day. I saw it at Collation right before I left Saigon. The Agency should put his name in lights on top of the headquarters building:
Joseph Hovey, The Man Who Predicted The Tet Offensive.
”

“Fat chance,” I replied, recalling that Hovey’s cable had gone to the White House in mid-December under a note by Carver which strongly implied—at OCI’s behest—that Hovey was crying wolf. I was still annoyed at this thought when Theresa handed me the draft that I’d sent Drexel Godfrey the day before about VC units omitted from the order of battle. Godfrey had scribbled on the buckslip: “To Sam Adams. Suggest you hold this until things quiet down. Also, its validity seems a little dubious—at least as of now.”
⁷

That buckslip was too damn much. Godfrey had tried to kill realistic numbers nine months before, and he was still at it. Overcome with disgust, I wrote Carver a letter of resignation, something I hadn’t intended to do. The letter said that the CIA’s failing was to acquiesce to MACV’s half-truths, distortions and sometimes outright falsehoods.” Furthermore, Westmoreland’s order of battle was a “monument of deceit” to which the agency had cravenly bowed in Saigon in September. These were the last sentences I composed for the Office of the Director, and when I slid it in Carver’s in-box on my way out the door, it occurred to me that he was the wrong recipient.
⁸
It wasn’t really Carver’s fault, not even Drexel Godfrey’s. It was Helms’s.

I found myself wheeling the shopping cart back and forth all day. The VC Branch secretary, Beverly, was appalled. “Where am I going to put all this crap?” she asked. It turned out not to be a problem. There was room. The VC Branch was so new that it had few files of its own. Unfortunately, it was also on the bottom of the mail room’s distribution list, so I spent most of the rest of the week back with Major Blascik, looking for VC unit designations. I searched everything in sight, even the newspapers. In fact, there were a couple of interesting stories. On Thursday, for example, Ward Just of the
Washington Post
wrote that “even the toughest pessimists here had not thought that the communists could mount so many offensives with so many men.”
⁹
When I showed it to George Allen, he looked insulted. And in a later
Post
, Peter Arnett reported from Saigon that Allied soldiers rifling the pockets of dead VC sappers after their attack on the U.S. Embassy had found forged curfew passes. The story said that perhaps five communist battalions were now in Saigon, and that many VC troops had entered the city three days before the attack.
¹⁰
Five VC battalions equals two thousand men, likely a fraction of the people involved. I recalled the pre-Tet document which had put the claimed production of a VC forging cell for nine months at 250 fake ID cards, scarcely enough for that big a force. My suspicion heightened that this forging cell was only one of many such cells.

Saturday’s Bulletin observed that VC troops were still in Hue, Hoi An, Phan Thiet, and Dalat, with new assaults on Xuan Loc and Phuoc Le, and that “the effectiveness of the Saigon government was being sorely taxed in the current military crisis.”
¹¹
Eventually, the situation room maps showed that the VC attacked forty of South Vietnam’s forty-four province capitals, almost a hundred of its district seats, and four of the five autonomous cities. In most of these places, they’d broken into town. Other situation-room records showed the VC destroyed or damaged twelve hundred U.S. aircraft.
¹²
The offensive was not only a slam-bang performance, it was a fine piece of evidence. How could the communists have mounted such a big attack with an army of only 225,346 men, the number carried in MACV’s latest Order of Battle?
¹³

I put that question on Monday morning to the VC Branch chief, Ron Smith. A friendly man of ruddy complexion and Maine accent, he said: “It’s obvious they couldn’t. They’ve got double that figure, and probably a lot more. The main question is whether the higher numbers are acceptable to the chain of command.” I was well aware of the problem with the chain of command. The VC Branch was part of the Office of Economic Research, whose boss on Vietnam was Paul Walsh, a pre-Tet supporter of the Saigon agreement. Moreover, as part of the DDI, Walsh reported to its head, R. Jack Smith, who in turn reported to the director, Richard Helms. Looked at from this angle the numbers predicament was the same as ever.

But for two other reasons, it wasn’t. I’ve already mentioned reason number one—the enormous indisputable fact of the Tet Offensive, over which the whole United States press was in a grand halloo that showed few signs of abating. The second reason was less obvious, but in my view almost as important. Finally I had some help. There were four analysts beside myself assigned to work on the Vietcong. Two were alumni of Saigon’s Collation Branch, which meant they knew something about the VC; a third was named Joe Stumpf, who’d already visited Vietnam briefly to look into enemy recruiting; and the fourth was Doug Parry, who had given me the enemy document before Tet about the 150,000
VC guerrillas. A tall, clean-cut Mormon from Salt Lake City, Parry was eager to get to work. That was another problem. Having no direction from up top on what to do, the branch was at almost a dead standstill.

In other words, the machinery was in place; all that remained was to turn it on. I made my first pass at this switch on 5 February, with a memo that began: “The events of the last week may well have changed many of the assumptions on which U.S. intelligence has operated in the past.” It tossed out some suggestions for research projects, most with an eye to upping the numbers, but some on other important-looking topics, such as VC spies. Its final recommendation was speed, because “big decisions are in the offing, such as whether to stay in or get out of Vietnam.”
¹⁴
When I gave it to Ron Smith, he said: “I’ll see if I can get a decision from upstairs.”

Goodness knows how long that would take, so I started to crank out papers by myself. On 7 February, for example, I wrote five: one about VC policy, three on missing regulars, and one about an attack at Tet. The fifth stemmed from new evidence, Vietcong documents having arrived at last concerning the offensive. Among them were the plans for the VC assault on Pleiku, including the formations assigned to do it. They were the Vietcong H-15 Battalion, Unit 90 (part of the VC 407th), and the Pleiku City Unit. Only one of the three, the H-15, was in the MACV Order of Battle.
¹⁵
I showed the memo around the building. Ron Smith said: “I’ll take it upstairs.” George Allen said: “It makes me sick.” George Carver said: “If any more of these show up, I want to see them.”

Over the next few days, I must have visited Carver a dozen times. The situation room crew joined in and so did George Allen. The situation room came up with the T89 and T87 battalions as having been in on the attack on Danang. Allen was the first to spot that the VC had formed extra units to help in the attack. Apparently, provinces with a single infantry battalion had formed two, districts had doubled up on companies, and so on. The new units were manned by late recruits and upgraded guerrilla militia—which helped explain the pre-Tet ferment in the VC home guard. George Allen told me: “One of the biggest
casualties so far is the pacification program. The South Vietnamese have pulled in to protect the cities, and the guerrillas are raising hell in the hamlets.”
¹⁶
He added that Vietcong local forces had borne the brunt of the attack so far, the communists having held in reserve most of their big divisions and regiments. There were exceptions. At least two enemy regiments were holed up in Hue, and the communists were finally moving in on the Marine Corps base at Khe Sanh. Troops of the North Vietnamese 304th Division had just overrun another Special Forces camp to Khe Sanh’s west—Lang Vei—for the first time using tanks.

There was only one type of communist soldier unheard from so far. George Carver brought it up on the afternoon of 11 February, a Sunday. He asked me: “Did you read President Thieu’s speech?”

“No, Sir,” I replied.

“Well I did just a little while ago, and he said that many of the enemy soldiers active in the cities belong to the VC militia.”

“The self-defense militia?”

“Yes,” Carver said. “Apparently the Vietcong are using them as support troops for the attack, and to keep order in occupied parts of the cities.
^*
You’ll remember that MACV read them out of the order of battle last September.”

“I remember.”

“It’s time to bite the bullet,” said Carver. “We plan to send a cable on the quiet to the Saigon Station voicing headquarters’ concern about the troops MACV dropped from its OB, such as the militia. This might be a good opportunity to reopen the whole numbers question, since the recent offensive obviously couldn’t have happened if MACV’s figures were accurate. I don’t see how we can start subtracting losses from the communist force structure until we have a handle on how large the structure is.”

“Yes, sir,” I replied, smiling enigmatically, and went below to write a memo of conversation. This had become habit. To Carver’s remarks, I appended these comments: “It appears to me that the last hand has been
taken from the brake (at least on the quasi-working level), and that we can now plunge forward toward making a realistic estimate of enemy strength. With the political atmosphere being what it now is—a new willingness to hang the rap on MACV—I think we can push forward many ideas that would have been rejected two weeks ago.”
¹⁸
For Ron Smith’s benefit, I suggested the same research projects as the week before. The front office was still sitting on them.

Not having come in over the weekend, Ron Smith didn’t read the memo until Monday morning. “I’ll see what I can do,” he said and disappeared once again, presumably to tug the chain of command. Smith had no sooner gone out the door than Doug Parry observed, “You realize that in most other trades, these people would be in deep trouble. They lied about the OB before Tet, which caught them by complete surprise as a result, and now they’re trying to think up ways to get out of it. There ought to be an investigation. Somebody should be told about this.”

Parry had touched a sore point. As if trying to excuse myself, I explained that although an investigation was a good idea, it was unclear how to get one started. Who should I complain to? The CIA Inspector General? He reported to Helms. The White House? The White House had probably put Helms up to it. Congress? The committees supposed to oversee the agency were well-known patsies. A more important problem was timing. “The CIA’s about to go straight,” I said, “and if I complained now, they’d probably stay crooked.” Parry agreed it was a problem.

The next morning Carver at last sent off a cable to Saigon that indicated that headquarters was thinking of reopening the order-of-battle dispute. Among those to sign on it were Drexel Godfrey of OCI, plus a close colleague of OER’s Vietnam honcho, Paul Walsh. Helms’ name didn’t appear on the message, but I guessed he must have given his OK.
¹⁹
In any case, Ron Smith was optimistic. “The light is no longer red,” he said. “I think it’s amber.”

Six days later, the Saigon Station sent in a message the effect of which was to turn the light green. It said that, far from repenting, MACV had climbed out even further on its limb. According to the DDI representative in Saigon, a “crash” J-2 study entitled “Cost and Impact on the
Enemy of the Tet Offensive” had concluded—after some intricate math involving VC casualties and recruitment rates—that the communists had suffered a net loss during the offensive to date of 24,000 men. OK so far but in order to reflect this mayhem, MACV had slashed its order of battle from 225,000 (the truncated number before the offensive) to 201,000 men.
²⁰
“You see where they’re headed,” I said to Ron Smith, “at this rate there won’t be any VC left by the end of the year.” “Impossible,” he replied, and took off to see Paul Walsh.

A short while later, Walsh called me up to his office. I had mixed feelings about him. Although he’d gone along with the fake OB before Tet, he’d done so reluctantly. Furthermore he’d fought the military over enemy logistics, having argued repeatedly that despite the bombing, the Ho Chi Minh Trail provided the VC with all the munitions they needed throughout the country. He was pale, with pouches under his eyes, which looked balefully over plastic-rimmed glasses.

“This is a travesty,” he said, pointing at the message. “and I don’t think we should let MACV get away with it. I want you to draft a reply to the ‘Cost and Impact’ paper.” I did so, saying we didn’t doubt that the Vietcong had suffered a net loss of 24,000 men—if anything, the losses were even higher, what with the carnage—but it came from a force two or three times bigger than MACV would admit. This put the offensive into a different perspective; for example, belying Westmoreland’s claim that communist casualties were “disastrous,” when actually they were quite reasonable under the circumstance. Walsh signed my cable almost without change, and shortly thereafter the DDI front office gave Ron Smith the go-ahead for the sidelined research projects.
²¹