Status: reported by multiple sources, one of whom passed two polygraph examinations, and partly corroborated by satellite imagery. Never confirmed. No American has been identified from these reports.
Laos · Reported after the war · Never resolved
What was reported
On 18 July 1979 the Hmong general Vang Pao, then living in the United States, wrote to Col. Vincent “Budd” Donahue, whose son, Air Force Lt. Morgan J. Donahue, was missing in Laos. Vang Pao said a Lao resistance leader, Koun Ma Phimmachak, had reported that an aircraft flew from Vieng Say in Sam Neua to Khamkeut carrying “18 American pilot and 25 Lao former political,” and that the prisoners were then put “in prison in the cave around Nhom Marat between Khamkeut and Thakket.” The reason given for the move was fear that Chinese troops might free them. “I hope one of the 18 Pilot is your son,” the general wrote.
The letter was passed to the National League of Families and from there to the Defense Intelligence Agency. On 11 October 1979 U.S. officials interviewed Phimmachak in a refugee camp at Ubon, Thailand. He said that in September 1979, on an operation into Khammouane province with a Lao resistance company, his unit captured a Pathet Lao soldier who gave his name as Tong. Under what investigators called vigorous interrogation, Tong said he was one of 13 guards assigned to 18 American prisoners held in a cave at the foot of a mountain Phimmachak called Hou Meut. Tong said the Americans had been moved there from Xiang Khoang and Phong Saly, arriving on 10 March 1979, and that they were bomber crews shot down over North Vietnam as early as 1967. He said the senior prisoner was an Air Force “Lt. Colonel Paul W.”
What the government did with it
Phimmachak was polygraphed twice. Both examinations found no deception, and the reports concluded he had accurately relayed what he was told. A review of existing satellite imagery allowed the Defense Intelligence Agency to confirm a cave area as he had described it, calling his account “generally accurate.” The Joint Casualty Resolution Center went further, finding that his “description of the area and his hand-drawn map agree with the analysis and photography.”
The name did not resolve. In a second interview in November 1979 Phimmachak gave the senior prisoner’s surname as something like “Mercaclan” or “Macalen.” No missing American matched. Analysts concluded the syllables were probably a mispronunciation of the word American. One Air Force officer did fit the rank and the first names: Col. Paul W. Bannon, whose wife had been notified of his presumptive finding of death months earlier. He has never been identified, if the report was accurate.
The CIA’s source
In November 1980 the CIA station chief in Bangkok cabled headquarters with reporting from a sensitive source, described to Senate investigators in 1992 as an elderly Laotian woman with direct access to the Lao leadership. She reported that a “group of POWs consists of approximately 30 Americans, all pilot’s, currently constructing a road near Nhommarath and that the LPDR has attempted to keep their existence a secret.” She said the guards were disaffected over food shortages, that the officer in charge had agreed to engineer the prisoners’ escape and that the entire Lao unit wished to defect to Thailand.
The station chief said he would instruct the case officer to tell the source to stand down on any escape plan unless headquarters had other views, and that “priority must be given to positively confirm the existence of the POWs with photographs, a list of names and the exact location.” He asked headquarters to review the most recent overhead imagery of the Nhommarath area. A later report said Lao officials had told the same source the government was still holding American and Thai prisoners, and that the Minister of Defense and the Vice Minister of Communication had confirmed it to her personally.
The agency was not persuaded. John McMahon, then the CIA deputy director for operations, wrote that it was “unlikely that the LPDR would hold prisoners of such sensitivity” so close to the Thai border, and that no motive for holding them had been offered. He added that the subject was “such as to necessitate a maximum effort to confirm or refute the accuracy of the information.” Imagery later placed the camp about 67 kilometers east of Thakhek, farther inside Laos than his objection assumed.
The rescue that was never launched
By 1980 the reporting — human sources, signals intelligence and imagery — was strong enough that the Joint Chiefs of Staff and the White House authorized the Joint Special Operations Command to plan a rescue. The operation was code-named Pocket Change. On 13 May 1981 planners briefed members of the Joint Chiefs. The task was stated simply: “Rescue POWs Held at detention Camp East of Thakhek, Laos.” Adm. Robert Long summed it up this way: “Bottom Line: We have the forces and will soon have the plan. All we need are domestic and international green lights, good weather — and somebody to rescue.”
A CIA-run indigenous reconnaissance team had reached the area and crossed back into Thailand that same day. It reported that a building in the inner compound was suspected of being a maximum-security facility, that no one held in it could be seen from the team’s vantage point and that “the question of Caucasian prisoners remains open.” At Langley that afternoon the agency told JSOC officers it had found no evidence Americans were in the camp. JSOC’s reading was that there was no evidence they were not. The raid never went.
What remains unresolved
- Whether the cave Phimmachak described and the camp JSOC planned to raid were the same place. Investigators later noted that a refugee’s description of a camp at Ban Naden corresponded to the Nhommarath camp to a remarkable degree, and that they believed the two were the same, but they were never able to prove or disprove it.
- If there really were 18 prisoners, who were they? No name from this reporting has ever been matched to a missing American, and the “Lt. Colonel Paul W.” identification was set aside rather than run to ground.
- What the reconnaissance team actually saw. The team never got close enough to hear language spoken inside the camp, and the full debriefing has not been released.
- Where the overhead imagery is. Coverage of the Nhommarath area was flown and reviewed at the time, and the comparison that produced the “generally accurate” finding was made against it.
Request for information
The POW Investigative Project is seeking any of the following.
- Any record identifying the 18 Americans reported moved to the cave near Nhommarath in March 1979, or the roughly 30 reported building a road there in 1980.
- The full debriefing of the CIA-run reconnaissance team that reached the camp area in May 1981, and any photography it brought back.
- Overhead imagery of the Nhommarath and Ban Naden areas flown between 1979 and 1982, and the imagery analysis compared against Phimmachak’s hand-drawn map.
- Records of the Pocket Change planning held outside the Joint Special Operations Command, including Joint Chiefs and National Security Council files.
- Any account from Lao, Vietnamese or Thai citizens who served at, guarded or worked near the site.
Related on this site
Laos · Americans reported at Camp C-53 · Two Americans alive in Hanoi · Research and sources