Phoenix Program

by Douglas Valentine

  The Nixon Doctrine as applied in Vietnam was called Vietnamization, and the man upon whom the mantle of Vietnamization fell was William Colby, godfather of the Covert Action program that had set the stage for American intervention ten years earlier. In November 1968 Colby was appointed DEPCORDS, replacing Democratic party loyalist Robert Komer, whom President Johnson had named U.S. ambassador to Turkey. Colby reported to Henry Kissinger, who supported Colby’s ambitious pacification program, geared to facilitate Vietnamization.

  Colby subdivided his pacification plan into three main categories, beginning with military security, which he called “the first step in the pacification and development process”—in other words, borrowed from Nelson Brickham, “shielding the population from the Communist main forces,” a job which “is the task of the Vietnamese regular forces.”1

  Often generated by Phoenix intelligence, the resulting air raids, artillery barrages, and search and destroy operations were an integral part of pacification, insofar as they created defectors, prevented guerrillas from assembling in large concentrations, and, by creating refugees, separated the fish from the water.

  Part II of Colby’s strategy was territorial security, the 1969 manifestation of Revolutionary Development, in which the Regional and Popular Forces—thereafter called Territorial Security Forces—were advised by U.S. Army mobile advisory teams (MATs) under the auspices of CORDS. In combating VC guerrilla units and the VCI, Territorial Security Forces were assisted by the People’s Self-Defense Forces.

  In a Defense Department report titled A Systems Analysis of the Vietnam War 1965-1972, Thomas Thayer says that as of 1968, “The Revolutionary Development program had significant problems in recruiting and retaining high quality personnel.” The RD Cadre desertion rate was over 20 percent, “higher than for any GVN military force, perhaps because they have a 30% better chance of being killed than the military forces.” Thayer notes that in response, the RD ministry had directed its cadre “to concentrate on building hamlet security and to defer, at least temporarily, the hamlet development projects which formerly constituted six of the teams’ eleven RD tasks.”2

  Under these revised guidelines, providing intelligence to Phoenix replaced “nation building” as the RD program’s top priority. Reflecting this change, the RD Cadre program was incorporated within the CORDS Pacification Security Coordination Division in November 1968, at which point MACV officers and USAID employees moved in to manage the program, bringing about, according to Robert Peartt and Jim Ward, a decline in performance and morale. In line with Lou Lapham’s redirection of the station away from paramilitary operations back toward classic intelligence functions, the CIA’s role in RD diminished, although it continued to skim off whatever strategic intelligence was produced. As Peartt noted, the station was “interested in going after region people, and would get involved at that point in RDC/O operations.”3 To a lesser degree, the CIA’s PRU program was also affected.

  “The agency made a decision,” John Wilbur said, “to get their ass out of Vietnam as fast as they could, for all the reasons Kinloch Bull foretold. It was losing control … diluting its cadre … being misdirected. It had become the sponsoring agency for a hodgepodge thing, and Phoenix was going to be the mechanism by which it was going to withdraw its control and sponsorship … and transition it over to the military. And that … meant that the PRU were no longer going to be the CIA’s exclusive boys, which foretold a real human crisis in the units.”4 Their “elan and morale had been carefully nurtured,” Wilbur explained. “We protected them from the dilution of control … from the province chiefs and battalion commanders. We insulated them from being used for whatever multiple good and bad reasons other people wanted to use them for. We would pay them a little better, we would take care of their dependents, and we would provide them with the best military support there was.” That, according to Wilbur, motivated them to “go out and do the things they did.”

  But, he added, “they had incurred a lot of resentment by the Vietnamese to whom they had previously been untouchable…. The leadership levels were marked men among many Vietnamese political forces.” And as soon as the Vietnamese got control in the summer of 1968, “everybody started messing with them.” The PRU began to be used as bagmen.

  “I was hurt in the last attack on Can Tho,” Wilbur continued, “and when I got back [from the hospital], my replacement had already arrived … and I spent most of the next six weeks introducing Chuck [Lieutenant Commander Charles Lemoyne] to the provinces, to all the hundreds of people he would have to deal with.” At that point Wilbur went home, where he remained until May 1968, when the CIA asked him to return to Vietnam to help Bill Redel “develop a national PRU unit which was targeted to recover American POWs in South Vietnam. It was the only thing that seemed worth fighting for,” Wilbur said, so he accepted the job. He was transferred to a naval security group, assigned to MACV, given an office (formerly occupied by Joe Vacarro) on the second floor of USAID II, and went to work for Redel.

  “We were going to set up a unit that would go around the provinces and try to collate whatever extant information there was, and in the event there was something that indicated [a POW camp] was there, we would try to put an in-place person, or try to develop … somebody to deal with an agent in place, and then gather the intelligence sufficiently to mount some sort of rescue operation.”

  But the rescue program was scuttled, and Wilbur instead got the job of transferring management of the PRU to the Vietnamese. He was introduced to Special Forces Mayor Nguyen Van Lang,* the first PRU national commander, and they began traveling around the country together. “And it became very apparent when I showed up with a Vietnamese colonel … what was going to happen. It meant the military, and that meant that the leadership elements of the PRU were in jeopardy of maintaining allegiance—they weren’t colonels and majors and captains.”

  Wilbur sighed and said forlornly, “The fact that there was no national overlay allowed the CIA to maintain autonomy over the PRU program longer than they would have otherwise.” But by the summer of 1968 “The official word had to go out that the PRU was becoming part of the Phoenix program: ‘We’re going to lose control. Get ready for the transition.’

  “It was the dissolution of American protection of the units that was mandated in our withdrawal,” Wilbur explained, “that corrupted the quality of control, which in turn allowed the PRU to be turned into a department store. And I became an agent of that. I was going to try to convince people to give up control of the PRU, after I had spent all this time arguing for its insulation and control and independence.”

  To effect territorial security, Colby intended “to get weapons into the hands of the Vietnamese villagers, so they could participate in their own defense” and to provide “funds to the elected village leaders to carry out local development programs.”5 The mechanism for this was Ralph Johnson’s village chief program at Vung Tau, about which Professor Huy writes: “[A]fter 1968, when Thieu succeeded to restore security in the countryside, several province and district chiefs used fraud and threats to put their men in the village and hamlet councils. These men were often the children of rich people living in cities. They needed the title of ‘elected representatives of the population’ to enjoy a temporary exemption from military service, and their parents were ready to pay a high price for their selection as village councilors. Thus, even the fiercely anti-Communist groups became bitter and resentful against Thieu.”6

  That brings us to Part III of Colby’s plan, internal security, otherwise known as Phoenix, the two-track CIA program to destroy the VCI and ensure the political stability of the Thieu regime by insulating him from the backlash of his repressive policies. As it was in the beginning, the pacification purpose of Phoenix was to weaken the link between the “people” and the VCI, while the political-level Phoenix was designed to exploit that link.

  To implement his plan, Colby forged ahead with a three-month startup program dubbed the accelerated pacificat
ion campaign (APC). Begun in November 1968, APC was designed to bolster Kissinger’s negotiating position in Paris by boosting the GVN presence in the hamlets, and was expected to show its effect by Tet [of 1969]. The goal was to add twelve hundred hamlets to the five thousand already classified under the Hamlet Evaluation System (HES) as “relatively secure.” Afterward APC was to be followed by an annual “full year pacification and development program.” To facilitate this process, Colby created the Central Pacification and Development Council as his personal staff and private conduit to Tran Thien Khiem, who replaced Tran Van Huong as prime minister in August 1969.

  Said Evan Parker about his patron William Colby: “The interesting thing was his relationship with Khiem … they would travel around the countryside in the same plane, each sitting there with his briefcase and a stack of working papers, writing like mad, answering memoranda, writing memoranda, passing memorandum back and forth…. There’s your coordination on this stuff—one of them or both would use his authority to support what I was asking the Vietnamese to do.”

  To assist him on the council, Colby hired Clayton McManaway as program manager; Tony Allito for HES reports; Harry “Buzz” Johnson for territorial security; and Ev Bumgartner and Frank Scotton for political liaison. With his personal staff in tow, Colby spent two days each week canvassing the provinces, bringing pressure to bear on people in the field, and promoting the accelerated pacification campaign.

  Phoenix adviser John Cook describes the accelerated pacification campaign as “an all out nationwide effort to put as many hamlets under government control as soon as possible. The Viet Cong violently opposed this action, since its primary purpose was to eliminate them and their control. It involved large military operations coupled with psychological operations, resulting in increased emphasis on the pacification program.” Insofar as the attack on the VCI strengthened Henry Kissinger’s bargaining position, Cook writes, “Pressure was placed on the Intelligence and Operations Coordinating Centers to provide more valid information about the enemy’s location. This required more of an effort from all of us, which meant an increase in the number of raids, ambushes and operations.”7

  The hour of Phoenix was at hand. With American troops withdrawing and emphasis being shifted from military to political operations, the pressure began to mount on Phoenix advisers, who were expected to eliminate any vestiges of revolutionary activity in South Vietnam. Reasons why they failed to accomplish this goal are offered by Jeffrey Race in his book War Comes to Long An.

  Blaming “overcentralization,” Race observes that the district, where the DIOCCs were located, “was the lowest operational level” of Phoenix, “one having no significance in terms of social or living patterns, and staffed by outsiders whose interests bore no necessary connection to the districts. By contrast, the revolutionary organization was the essence of simplicity … and intimately familiar with the local population and terrain.” Race traces the lack of “security” at the village level to the GVN’s disdain for the common people and its “failure to develop a highly motivated and trained local apparatus.”8

  Operational as well as organizational errors also factored into the equation. Forces under the Phoenix program, Race explains, “operated in the manner of a conventional war combat organization—independently of their environment—and so they did not have the enormous advantage enjoyed by the party apparatus of operating continuously in their home area through a personally responsive network of friends and relatives. This in turn severely handicapped their ability to locate intended targets and to recognize fortuitous ones. The program was also handicapped in developing a sympathetic environment by the use by the Saigon authorities of foreign troops and by the program’s intended purpose of maintaining a distributive system perceived as unfavorable to their interests by much of the rural population.”9

  Responding to the grievances of the rural population and taking steps to correct social injustices might have enabled the GVN to collect intelligence and contest the VCI in the villages. But acknowledging the nature of the conflict would have undermined the reason for fighting the war in the first place. And rather than do that, Race says, “attention was turned to the use of such new devices as starlight scopes, ground surveillance radar, and remote listening devices, as well as the previously employed infrared and radio transmission detection devices.”10

  In August 1968, concurrent with Robert Komer imposing, as “a management tool,” a nationwide quota of eighteen hundred VCI neutralizations per month, the science fiction aspect of Phoenix was enhanced with the advent of the Viet Cong Infrastructure Information System. VCIIS climaxed a process begun in February 1966, when Secretary of Defense Robert McNamara established the Defense Department’s Southeast Asia Programs Division. The process was carried forward in Saigon in January 1967, when the Combined Intelligence Staff fed the names of three thousand VCI (assembled by hand at area coverage desks) into the IBM 1401 computer at the Combined Intelligence Center’s political order of battle section. At that point the era of the computerized blacklist began.

  As the attack against the VCI exploded across South Vietnam in 1968, reports on the results poured into the Phoenix Directorate, inundating its analysts with reams of unreliable information on individual VCI and anti-VCI operations. In DIOCCs the data could be processed manually, but in Saigon it required machines. Hence, with input from the Defense Intelligence Agency, the FBI and the CIA—all of which had an interest in analyzing the finished product—VCIIS became the first of a series of computer programs designed to absolve the war effort of human error and war managers of individual responsibility.

  The cerebellum of Phoenix, VCIIS compiled information gathered from all U.S. and free world field units on VCI boundaries, locations, structures, strengths, personalities, and activities. The end product, a monthly summary report, was a statistical summary of Phoenix operational results by province, region, and the country as a whole and showed the levels and methods of neutralizations at each echelon within the VC infrastructure. A monthly activity listing listed each “neutralized” VCI by name. In July 1970 the Vietnamese were invited to contribute to the program and started key punching at the National Police Interrogation Center. Until then the computerized blacklist was a unilateral American operation.

  In January 1969 VCIIS was renamed the Phung Hoang Management Information System. The PHMIS file included summary data on each recorded VCI in the following categories: name and aliases; whether or not he or she was “at large”; sex, birth date, and place of birth; area of operations; party position; source of information; arrest date; how neutralized; term of sentence; where detained; release date; and other biographical and statistical information, including photographs and fingerprints, if available. All confirmed and suspected VCI members were recorded in this manner, enabling Phoenix analysts instantly to access and cross-reference data, then decide who was to be erased. All of this added up to hard times for NLF sympathizers, Thieu opponents, and those unfortunate enough to be creditors or rivals of Phoenix agents.

  As a management tool PHMIS was used by Komer and Colby to measure and compare the performance of Phoenix officers—unless one believes those like Tom McCoy, who claims that Komer was a fraud who went to Vietnam “not to do pacification but to prove that it was being done.”11 In that case the numbers game was computerized prestidigitation—an Orwellian manipulation of statistics to shape public opinion.

  According to McCoy’s scenario, PHMIS was part of a larger hoax begun in January 1967, when Robert Komer introduced the Hamlet Evaluation System (HES)—eighteen factors subject to computer analysis for each of South Vietnam’s fifteen thousand hamlets. These factors included data on VC military activity, GVN security capabilities, the strength of the VCI, Revolutionary Development activities, etc. The data were assembled by MACV district advisers, with the computer then putting the hamlets into one of three classes: A, secure; B, contested; or C, controlled by the VC.

  On the verge of Tet in December 19
67, nearly half of South Vietnam’s hamlets were rated A. One year later more than half were rated A. As Public Safety chief Frank Walton told me, “We would get reports of provinces being eighty-five percent pacified and ninety percent pacified, and then, when it got to the point that they were near a hundred percent, figures had to be revised downward. It was done with computers, and that’s where I first heard the term ‘GIGO’ for ‘garbage in; garbage out.’ “12

  The Hamlet Evaluation System also included input on “the known strengths of the 319 currently identified, upper-level VCI organizations at COSVN, region, province and district levels.” The HES guesstimate of VCI strength in January 1969 was 75,500.

  Statistics on the VCI; definitions of the VCI; attitudes toward the VCI—all were subjective. Yet despite his own admission that “we knew there was a VCI, but we could not be said to know very much about it,” William Colby set about attacking it. Armed with technology that rendered due process obsolete, he “set up standards and procedures by which to weed out the false from the correct information.” To ensure that Phoenix operations were mounted on factual information, “The general rule was established that three separate sources must have reported a suspect before he could be put on the rolls.” Thus, the VCI was put into three classes of offenders: A, for leaders and party members; B, for holders of other responsible jobs; and C, for rank-and-file members and followers. “And the decision was taken that those in the ‘C category should be ignored, since Phoenix was directed against the VCI command and control structure and not the occasional adherent or supporter.”13