The NSA's own historian concluded in 2001 — declassified 2005 — that the second Gulf of Tonkin attack "almost certainly did not happen" and that NSA analysts selectively filtered intelligence to support a false conclusion. Congress voted 88-2 for the Gulf of Tonkin Resolution anyway. The Vietnam War followed.
The Gulf of Tonkin incident — as it is commonly known — was actually two separate events, two days apart, in August 1964. The first incident on August 2 is well-documented and not seriously disputed: North Vietnamese torpedo boats attacked the USS Maddox, a U.S. destroyer conducting surveillance operations in the Gulf of Tonkin. The attack caused minor damage; the Maddox returned fire and repelled the boats.
The second incident, reported by the U.S. Navy on August 4, was the basis for the Gulf of Tonkin Resolution — which gave President Johnson authorization to use military force in Southeast Asia and effectively launched the Vietnam War. This second attack almost certainly did not happen.
The definitive primary source on the August 4 "attack" is an NSA historical study declassified in 2005 and titled "Tonkin Gulf Intelligence: The Skewed Record that Led to War." The study was written by NSA historian Robert Hanyok and was initially completed in 2001 but withheld from declassification for four years.
Hanyok's analysis of NSA signals intelligence from August 4, 1964 concluded: "It is not simply that there is a different story as to what happened; it is that no attack happened that night." The study found that NSA signals intelligence analysts had selectively reported intercepts that supported an attack while omitting or discarding intercepts that contradicted one. Hanyok described this as an "echo chamber" in which initial reports of an attack were reinforced by selective filtering of subsequent intelligence. His conclusion was that the NSA had provided policymakers with a false picture of events on August 4.
The NSA study's findings are supported by the contemporaneous doubts of the sailors who were supposedly under attack. Captain John Herrick, commander of the Maddox, cabled Washington on the night of August 4 that the attack reports might be the result of "freak weather effects" and "overeager sonarmen." Defense Secretary Robert McNamara knew about Herrick's doubts before he testified to Congress the next day. He did not disclose them.
President Johnson used both the confirmed August 2 attack and the disputed August 4 attack to request war authorization from Congress. The Gulf of Tonkin Resolution passed the Senate 88-2 and the House 416-0 on August 7, 1964 — three days after an attack that the NSA's own internal history now says did not occur.
Johnson privately expressed his own doubts. According to multiple accounts, including that of Undersecretary of State George Ball, Johnson privately acknowledged that the August 4 attack might not have happened. He proceeded with the resolution regardless.
McNamara acknowledged the uncertainty in later years. In the documentary film "The Fog of War" (2003), he said: "We were wrong... but we had in mind... demonstrating the Soviet Union and China that we had the will to resist."
"It is not simply that there is a different story as to what happened; it is that no attack happened that night."
— NSA Historian Robert Hanyok, "Tonkin Gulf Intelligence," declassified 2005The NSA's own internal historical analysis, declassified in 2005, concluded that the August 4, 1964 attack — the primary justification for the Gulf of Tonkin Resolution — almost certainly did not happen and that NSA analysts selectively filtered intelligence to support a conclusion of an attack. The Gulf of Tonkin Resolution, passed 88-2 in the Senate and 416-0 in the House, authorized the use of military force that led to full U.S. involvement in Vietnam — 58,000 American deaths and millions of Vietnamese casualties. The authorization was based on a reported event that the government's own records now establish did not occur.