War of Numbers (17 page)

Casualties from small arms and other causes are omitted.

^*
The U.S. Ninth Division later captured a Vietcong report that gave the size of the Long An home guard in 1966. The report listed 1,321 village guerrillas, 2,029 hamlet guerrillas, 198 secret guerrillas, and 3,363 self-defense militiamen for a total of 6,911. You will recall the equivalent numbers in the MACV Order of Battle were 100 guerrillas and 60 militiamen for a total of 160. You may also remember my puzzlement in Long An on 9 February on noting that almost as many guerrilla-militia had defected as were in the official estimate. (See
Chapter 3
) Obviously my confusion stemmed from the fact that the MACV Order of Battle listing for the province home guard was forty-three times too low.
¹⁹

^*
The 600,000 was broken down as follows: Regulars, about 100,000; guerrilla-militia, about 300,000; service troops, about 100,000; political cadres, about 100,000.

^*
Allen’s exact wording. It predated the Salem cigarette ad.

5   FOURTEEN THREE

THE CHIEF HUN opposing the upward revision of the OB turned out to be George Fowler. A gray-haired, heavyset, chain-smoker of Chesterfields, Fowler was the Pentagon factotum who had carried my guerrilla-militia memo to Saigon when the CIA had first published it in September 1966. His battle cry was “Harrumph.”

He had arrived with thirty or forty other intelligence people at just before 10:00
A.M.
on 23 June 1967 at the Board of National Estimates’ seventh-floor conference room, to attend the first session of Fourteen Three.
^*
Windowless, furnished with leather chairs and maps, the room had a large conference table with Bobby Layton, the estimate’s drafter, at one end, and General Collins, Fourteen Three’s chairman, at the other. The rest—from the State Department, the Defense Intelligence Agency (namely, Fowler), the three services, the National Security Agency, and the CIA—arranged themselves either at the table or along the walls. I sat next to a wall. A board member since his retirement from the Army, General Collins—stocky, bullnecked, with flared eyebrows—opened the proceedings:

“Gentlemen, as you all know, the guts of this year’s Fourteen Three is the numbers game, so we might as well face up to it before we do
anything else. I want all positions put clearly, so there’ll be no misunderstandings. George, you look like you’ve got ants in your pants, why don’t you lead off?” The general was addressing George Fowler.

Fowler said: “Harrumph! DIA cannot agree to this estimate as currently written. What we object to is the numbers. We feel we should continue with the official order of battle.” And he read off the OB’s four components. As usual, the first one, the regulars, was somewhat higher than a month earlier but the other three—the service troops, the guerrilla-militia, and the political cadres—were the same as ever. When Fowler announced their total, now 296,000, Collins turned to Layton: “Ok, Bobby, it’s your turn. Do it succinctly and without vituperation.”

Layton said: “The notion that we should use numbers just because they’re ‘official’ simply won’t wash. I would like to remind everybody, in case they need reminding, that what we’re here for is to decide which numbers
are
official—the ones in MACV’s order of battle, which in three of four cases are several years old, or those in the draft, which are based on evidence, some good, some not so good. The nub of the question is
evidence.
” And Layton ticked off the draft’s four components, the same ones as Fowler’s, but three of them much larger. Layton’s total came to well over 500,000. Collins scribbled on a pad of paper, and said:

“If my math’s correct, you gentlemen are close to a quarter of million apart. That’s a heap of folks. George, the ball’s in your court. What’s wrong with Bobby’s numbers?”

The only thing
right
with Layton’s numbers, Fowler declared, was the regulars. There the estimate agreed with the OB. However, the draft’s service-troop figure is a “wild guess”; its guerrilla-militia number came from a speech by Nguyen Chi Thanh which was “pure propaganda”; and the political cadres were based on “extrapolations from documents.” Layton answered Fowler point by point, always admitting where the evidence was weak, but also asking where DIA’s component had come from. The response was unvaried: “The OB, harrumph, it’s official.” An hour and a half later, the onlookers were glassy-eyed, and General Collins interrupted:

“That’s enough, George, you’ve made your point, but Bobby’s right. We can’t insert numbers in the estimate just because they’re official.” Collins paused. “Now gentlemen, it seems to me that we have two alternatives. Bearing in mind that counting enemy soldiers is normally a military prerogative, our first alternative is to put the OB numbers in the estimate, but with caveats that they’re way too low, and MACV’s trying to come up with something better. The second is to use our own numbers, or rather Bobby’s—a ‘best estimate,’ so to speak. I want a round-robin on the choices. OK, what’ll it be? The best estimate or the order of battle? George, I’ll begin with you. I know what you’re going to say, but say it anyway.”

Fowler obliged: “The OB.”

“Navy?” Collins asked.

“Pass. At last report the VC were still ashore. When they put to sea, the Navy’ll count ’em.”

“Air Force?”

“Also pass. Same line of reasoning as my colleague in blue.”

“Army?”

“Frankly, general, I’d as soon not get involved, but I guess I have to go with the OB.”

“NSA? Where do the code-breakers stand?”

“As you know, sir, the National Security Agency doesn’t take positions on how to interpret evidence.
^*
We have observed, however, that the communications net serving the communist army has increased severalfold in the last three years. Perhaps this bears on the problem.”

“You’re damn right it bears on the problem,” said Collins. “State Department, what about the diplomats?”

“Best estimate.”

Collins then polled a half-dozen CIA-ers, myself included, who all said: “Best estimate.”
¹

The meeting broke for lunch, and I went down to the sixth floor to tell Carver about George Fowler’s astonishing performance.

Carver laughed. “Don’t pay mind to that old curmudgeon. He’ll repeat any whopper the Pentagon gives him and do it with a straight face. He’s DIA’s permanent fixture at the board meetings, but fortunately, he’s predictable.
²
It’s the unpredictable ones who make all the trouble. Now look, on VC strength, it’s time to bite the bullet. You go back up there and do the best you can.”

Three more sessions of Fourteen Three occurred in late June. The numbers continued as the main issues, and I took over from Bobby Layton most of the job of defending the agency’s position. There was some minor slippage in a couple of categories, but otherwise the CIA held firm. Serenely puffing on Chesterfields, Fowler stuck with the official OB. After each meeting, I reported back to Carver to tell him what had happened. One day in early July, his desk was empty. As was her custom, his secretary, Mary Ellen, wouldn’t let on where he’d gone. I found out on 10 July, however, when a “secret” cable arrived from Vietnam signed “Funaro,” which was Carver’s cover name.
³
Evidently, Helms had sent him to Saigon to break what the message termed “the current Washington impasse” on VC strength. Accordingly, Funaro had met the day before with Gains Hawkins—still MACV’s Order of Battle chief—and General Philip Davidson, who not long before had succeeded General McChristian as Westmoreland’s head of intelligence.
⁴

Carver first relayed Colonel Hawkins’ assurances that—George Fowler notwithstanding—the MACV Order of Battle Section basically agreed with the CIA: about 100,000 was OK for the guerrillas, the colonel had said, although our guess of 75,000 service troops was perhaps “a little” too high. General Davidson fingered the real difficulty. “The chief problem,” Carver quoted Davidson as saying, “is the political and presentational one of coming out with a brand new set of figures showing a much larger force at a time when the press knows that MACV is seeking more troops.”

No doubt the press
was
a problem, a big one, and Carver’s message
suggested a solution. It was to break the order of battle into two parts: “military,” to include regulars, service troops, and guerrillas, totaling about 300,000 (or approximately the number in the old OB); and “nonmilitary,” to include self-defense militia and political cadres, coming to around 200,000. The total would be half a million, and then MACV could hold a press conference to explain the transaction. “Some elements in the press would always carp,” Carver said, “but the air would be cleared … and a valid baseline established for future … analysis.” Davidson said that it sounded like a good idea to him and that he’d try it out on General Westmoreland. I tried it out on George Allen.

“Balls,” said George. “This business of fooling around with the OB for a few dumb reporters makes me sick. God knows what it’ll lead to. Maybe stowing the political cadres up on a nonmilitary shelf is legitimate, but not the militia, goddamnit. The militia are part of the VC army, and always have been. That’s the reason Bill Benedict and I put them in the OB in 1962, and that’s where they damn well belong.”

“But George,” I said, “Carver’s only slicing the pie in a different way. It still adds up to half a million. What difference does it make so long as they show what we’re trying to demonstrate: that the war’s a lot bigger than we thought it was.”

“Balls,” he repeated. I left him fuming.

Frankly, the cable was fine by me: I was not necessarily pleased by its “solution,” which as far as I could see didn’t “solve” Westmoreland’s problem of how to explain the higher numbers to the press, but rather by a brief statement in its last paragraph. That said that the next session of Fourteen Three would have to wait until August, at which time Gains Hawkins would come to Langley to take over from George Fowler as champion of the military’s numbers. This meant the hanky-panky would stop; I shared the general opinion that Colonel Hawkins was scrupulously honest. But there was an even more pressing consideration. On my return from Saigon in May I had promised Bill Johnson of the counterintelligence staff that the study on the Vietcong police would be ready by the end of July. The delay in Fourteen Three would
allow me to meet this deadline. A dividend to the postponement was the fact that the VC police were the political cadres, the OB category about which I knew least. I could bone up on them now.

In fact the study was well underway. A chapter on the COSVN Security Section, which Commander Siller and I had started at the Collation Branch, was already in final draft. A second chapter, concerning communist police operation in and around Saigon, had been farmed out to still another helper, Manny Roth, an eager young lieutenant whom Bill Johnson had wangled from the Army. Roth loved the work. “Binh Tanh!” he whooped at my desk one morning: “I got Binh Tanh’s!” He had discovered the VC cover designation of Binh Tanh Subregion, one of the six surrounding Saigon. It was “A23,” he said, and its police headquarters was “A536.” By mid-July he had turned up the cover designation of all six subregions, and for four of their six police headquarters. More important, he had begun to figure out which ones operated where. For example, Binh Tanh’s police ran secret agents into Saigon’s Second and Fourth precincts.

I was also immersed in cover designations. Having determined through the VC interrogator in Soc Trang that the “83” subsection of their police apparatus was diep bao, (which the Vietnamese sergeant had translated as “espionage”) I went through my boxful of VC documents looking for reports marked “83.” The policy spy network fell rapidly into place. Subsection 83, it turned out, had three parts: A1, A2, and A3. The first consisted of agent-handlers—who made a special effort to recruit the Vietnamese interpreters working for the Americans. I wondered about that sergeant in Soc Trang. The second section made blacklists of victims targeted for various woes: assassinations (which the VC called “executions”); kidnappings (“arrests”); or other “disciplinary measures” which would surely take place, the documents claimed, when the communists “liberated the south.” The third part, A3, were the thugs who actually did the work. Hitmen from A3 wielded silencer pistols, hustled Saigon officials into waiting cars, and set off explosions in city streets. Once again I recalled the
kaboom
that had sent me under the bed on my first Saturday in Vietnam. If the bomber were a VC policeman, I
thought, he would have belonged to the A3 component of the B3 subsection of a local subregion police headquarters.
Which
subregion, I didn’t know, because I was unsure which of Saigon’s precincts the bomb had gone off in. That was easy to find out. I could look at a map.

When the completed draft landed on Bill Johnson’s desk on 28 July, he whistled, saying: “I’d love to see the VC damage assessment when they finally read this thing.” It was a high compliment. A damage assessment is what intelligence organizations write after they’ve been found out. Perhaps “found out” was a little strong. For although the study presented an intricate wiring diagram of the Communist police network from its headquarters at the Ministry of Public Security in Hanoi down to the smallest hamlet in the south, and described at length how the police went about their work, it lacked important details. I mentioned the most prominent of these to Bill Johnson: “Names. There were a thousand documents, but not a true name in the lot. We don’t know who these people are.”