.png)
Operation Menu Cambodia: The Bombing the Records Said Never Happened (Cold Files 022)
On 18 March 1969, sixty B-52 bombers were launched against targets inside South Vietnam. Forty-eight of them were turned across the Cambodian border in flight and bombed a patch of jungle the planners had codenamed Breakfast. The White House chief of staff, H. R. Haldeman, wrote in his diary: "Historic day. K's 'Operation Breakfast' finally came off at 2:00 pm our time."
The post-strike reports said South Vietnam.
That is the whole of this file. Not that the bombing of Cambodia was classified, because plenty of things were. It is that for fourteen months the United States ran a bombing campaign against a neutral country and entered it into its own records as something else, so that the records a congressional committee would ordinarily ask for said the campaign had never occurred.
Part of the Cold Files series.
The System That Wrote Cambodia Into Another Country
The procedure was drawn up by Colonel Ray B. Sitton, a Strategic Air Command bomber officer then working in the Operations Directorate of the Joint Chiefs of Staff, known around the Pentagon as Mr. B-52. Seymour Hersh, in "The Price of Power," calls it the dual reporting system, approved by Kissinger and Haig.
Crews were briefed en masse on targets inside South Vietnam. Selected crews were then taken aside and told that a ground radar station would pass them special instructions shortly before the bomb run.
The real coordinates travelled separately. On the afternoon before a mission, a courier flight from Saigon carried them to Major Hal Knight, who supervised the Combat Skyspot radar site at Bien Hoa. John T. Correll, writing in Air Force Magazine, says they arrived in a plain manila envelope. A contemporaneous report in Time described sealed, unmarked envelopes kept locked until dusk.
Over the target, the radar site effectively flew the aircraft. Combat Skyspot's MSQ-77 ground-directed bombing worked by activating a transponder aboard the bomber, computing the release solution on the ground, passing steering corrections to the lead aircraft by radio and calling the drop on a voice countdown. Most of the crew never knew what country was underneath.
Afterwards, the reports went in against the South Vietnamese coordinates the crews had originally been briefed on.
Then Knight destroyed the evidence. He collected every document showing the actual strike locations, along with the computer input tapes, and burned them. Time reported that he used a barrel outside his hut, and that he then telephoned a number in Saigon and reported the mission complete with an agreed phrase: "The ball game is over."
Wheeler later said the military had devised the dual reporting system itself.
The best measure of how far this went beyond ordinary secrecy is who remained outside it. The Secretary of the Air Force and the Air Force Vice Chief of Staff were not told the Cambodia bombing was happening. Classification keeps information from people without clearance. This kept it from men who held the highest clearances and ran the service flying the missions.
4,247
The target boxes were not empty, and the United States did not think they were.
Six areas were struck under Menu, each named for a meal and each mapped onto a numbered Vietnamese sanctuary area: Breakfast, Lunch, Snack, Dinner, Supper and Dessert. Those sanctuaries were real. Even Prince Norodom Sihanouk, briefing the American Senate Majority Leader Mike Mansfield in August 1969, said the main threat to Cambodia was Vietnamese communism and conceded that North Vietnamese forces were crossing his frontiers.
What the records show is that the Department of Defense counted the civilians in the boxes first. William Shawcross, in "Sideshow," reports a Pentagon estimate of 4,247 non-combatants across the six Menu target areas, drawn from material produced to the Senate in 1973, along with planners' notes that the strikes could increase casualties and that Cambodian houses often had no shelters.
The per-box numbers are worse than the total. Base Area 353, the Breakfast box, covered about 25 square kilometres and was assessed as holding some 1,640 residents. It absorbed 228 B-52 sorties. A single B-52D configured for the war carried up to 108 bombs. In the Supper box, Base Area 740, the estimate was 1,136 non-combatants.
The internal language stayed calm throughout. A Kissinger memorandum to Nixon of 27 April 1970 called the risk to noncombatants in the Breakfast and Dinner boxes "minimal", and in the populous Parrot's Beak "substantial".
Menu's own totals are still not settled. The oldest figures, traceable to the Pentagon's 1973 count, give about 3,630 sorties and 110,000 tons. A Joint Chiefs history gives 3,875. Published tonnages cluster between 100,000 and 110,000. Fourteen months, somewhere near that, on six boxes the planners had already assessed as holding 4,247 non-combatants.
The Story That Could Not Be Corroborated
It is not true that nobody knew.
On 9 May 1969, seven weeks after the first strike, William M. Beecher, the military correspondent of the New York Times, put the Cambodia raids on the paper's front page. The headline carried the problem inside it: "Raids in Cambodia By U.S. Unprotested."
The reaction was immediate and it was aimed at the leak rather than the story. Within hours Kissinger asked the FBI director, J. Edgar Hoover, to find the sources. The next day Alexander Haig told the bureau the matter was so sensitive it demanded handling with no record maintained. The National Security Council aide Morton Halperin was suspected and his telephone was tapped for twenty-one months. Thirteen officials and four journalists were wiretapped in all.
Phnom Penh did not confirm the story, and the reason is more interesting than silence. Sihanouk told Mansfield in August 1969 that Cambodia had not protested bombings which struck Viet Cong forces rather than Cambodian villages and populations, that much of what he knew about strikes in uninhabited areas came from the American press, and that he asked for incidents involving Cambodian lives to be avoided.
That is the documented position, and it is neither of the two things it usually gets turned into. Sihanouk did not authorise Menu. The historian Kenton Clymer argues he was never asked, and that his earlier remark to the American envoy Chester Bowles about not opposing hot pursuit in uninhabited areas does not stretch to cover B-52 raids. But he also did not object to strikes he understood to be falling on Vietnamese troops in empty country. His stated condition was that Cambodians not be harmed, and the United States had tabulated 4,247 of them in the boxes before the first aircraft lifted.
So the honest version is not that the bombing was unknown. It is that it was reported once, could not be corroborated by the country being bombed, met a policy of neither confirming nor denying, and was contradicted by an internally consistent routine record. A front-page story that the files refute is not a disclosure. It is an allegation.
One Letter
What eventually broke it was not a document release. It was a major who could not live with what he had signed.
In December 1972 Hal Knight wrote to Senator William Proxmire, who passed the letter to Senator Harold E. Hughes. Knight's account, as recorded in his Senate statement, was flat and self-implicating. While he was in South Vietnam, he said, the falsification of strike reports was a common practice. He had prepared at least a dozen himself. All the reports he personally faked were B-52 strikes into Cambodia. He believed the practice broke military law, which forbade signing false records with intent to deceive, and he had feared court-martial if he refused. His objections earned him two poor effectiveness ratings. He was passed over for promotion and resigned.
The rest came quickly. On 12 July 1973, General George S. Brown, who had commanded the Seventh Air Force in 1969, was asked at his confirmation hearing to be Air Force Chief of Staff whether there had been strikes before May 1970. He requested a closed session and confirmed there had. He was confirmed anyway. On 16 July the Defense Secretary, James R. Schlesinger, acknowledged the 1969 bombing publicly, and the Senate Armed Services Committee opened eight days of hearings at which Knight was the first witness.
The committee had already been told otherwise. When it asked for air operations records in March 1973, the Pentagon's report came back with no B-52 attacks on Cambodia before May 1970. Hughes, holding a declassified list that showed zero, called it official deception.
Nobody denied the dual reporting. They disputed its name. General Earle Wheeler said there had been no intent to deceive, the element military law required, and that the people who needed the truth had it. Melvin Laird said he had approved a separate reporting procedure but had not authorised any falsification of records. A Pentagon report that August produced the formulation that everyone in the reporting chain received and reported the information for which he had a need to know. The acting Air Force judge advocate general, Brigadier General Harold Vague, had advised that inaccurate reporting to Congress could be permissible for military reasons.
No one was prosecuted. Airmen who had refused to fly the missions remained liable to court-martial. On 30 July 1974 the House Judiciary Committee considered an article of impeachment covering the false statements about the bombing and rejected it, 26 votes to 12. The reason is the hardest objection in this file: a handful of congressional leaders had been briefed on Menu and had not objected, so the committee held that Congress had notice. Representative John Conyers, who proposed the article, told the chairman it was not frivolous.
The Count That Was Never Started
There is no death toll for Operation Menu. Not a disputed one. None.
Published estimates cover 1969 to 1973 or the 1970 to 1975 civil war. Nobody has counted the Menu dead. The best known range, 50,000 to 150,000 deaths, belongs to the historian Ben Kiernan, who published it in Vietnam Generation in 1989 for the whole 1969 to 1973 bombing and built it from survivor and refugee testimony extrapolated outward village by village. It is not a count, and Kiernan has never said it was.
No census followed the bombing, no bomb damage assessment counted Cambodian civilians, and no exhumation programme was aimed at it. In fairness, nobody counts war dead by codename, and the fourteen months are America's unit, not Cambodia's. But no one has counted them under any other name either. The number is bounded and not knowable.
Here the file has to turn on its own side of the argument, because the sharpest correction on this subject came from the critics, not the defenders. In 2006 Taylor Owen and Ben Kiernan published, from the air force data declassified in 2000, a figure of 2,756,941 tons dropped on Cambodia, which supported the familiar claim that Cambodia absorbed more bombs than all of the Second World War. The anthropologist Holly High, working with the computer scientists James R. Curran and Gareth Robinson, found that the database's load weight field had been multiplied by ten. Owen and Kiernan published the correction themselves in 2010. Around 500,000 tons is the figure that stands; High's floor is 472,313 tons. American tonnage in the whole of the Second World War was about two million.
So the headline statistic about the Cambodia bombing is false, and the people who retired it are the ones who had the most to lose by doing so. What survives is still heavy enough: over the whole 1965 to 1973 air war Cambodia took roughly what the American Pacific theatre took.
Sources consulted
Hersh, The Price of Power, Summit, 1983
Correll, Shadow War in Cambodia, Air Force Magazine, 2018
Time, Defense: Bombing Coverup, July 1973
Shawcross, Sideshow, Simon and Schuster, 1979
Kastenberg, Fifty Years On, Southwestern Journal, 2020
Senate Armed Services Committee, Bombing in Cambodia, 1973
Owen and Kiernan, Bombs Over Cambodia, Walrus, 2006
Owen and Kiernan, Making More Enemies, Asia-Pacific Journal, 2015
High, Curran, Robinson, Electronic Records, Vietnamese Studies, 2013
Kiernan, US Bombardment of Kampuchea, Vietnam Generation, 1989
Kiernan, Collateral Damage, Antipode, 2003
Foreign Relations of the United States, volume VI
National Security Archive, Kissinger Declassified Obituary, 2023
Turse, Blood on His Hands, Intercept, 2023
Clymer, The United States and Cambodia, Routledge, 2004
Beecher, Raids in Cambodia Unprotested, New York Times, 1969
Public Law 93-52, section 108, 1973
Eisenberg, This Is Not Frivolous, American Heritage, 2024
US Air Force, Ray Sitton biography, af.mil
National Archives, Southeast Asia sortie records, archives.gov
Karnow, Vietnam: A History, Viking, 1983
What was destroyed, what was falsified and what was kept are three different things, and the difference is the file. The real coordinates burned in a barrel at Bien Hoa. The routine reporting channel carried another country's grid references upward for fourteen months. The true accounting went to a handful of men who already knew. The dead were never in any of the three sets of books, because nobody opened one for them. A toll can be disputed, revised, corrected and fought over for fifty years. Menu's was never started.
Cold Files runs twice a week on Project Leaflet: long reads on the wars and the records other feeds leave alone. For the daily brief and every new file as it lands, follow us on t.me/TheLeaflet.
More Cold Files: East Timor Death Toll and Congo War Death Toll.
Next in Cold Files. In 1994 about two thousand people sheltered at a technical school in Kigali behind Belgian soldiers. The soldiers left. A Brussels court held Belgium liable, and then a higher one took it away again.
Comments