The attack that wasn't
How erroneous intelligence reports led to a previous war
The murky events in the Gulf of Tonkin in August 1964 propelled the United States into the Vietnam War. Over three days, the U.S. government reported two attacks on its destroyers. The first assault by North Vietnamese PT boats on August 2 provoked only a protest from Washington. On the cloudy night of August 4, two U.S. ships detected another apparent ambush. They fired their guns in response, claiming to have sunk several enemy boats. The incident prompted Lyndon Johnson to order airstrikes and obtain what amounted to a declaration of war from Congress.
Within a few years, though, public doubts grew about the second incident. There was no physical evidence of an attack, though Secretary of Defense Robert McNamara assured Congress in 1968 that the intelligence was "unimpeachable." He was referring mostly to North Vietnamese naval communications intercepted by the National Security Agency. In 1984, a U.S. News cover story quoted two top intelligence officials from the period expressing strong doubts about those reports (more information at www.usnews.com/tonkin ).
Now, a newly declassified NSA study--and over 140 formerly top secret intelligence reports--confirm just how wrong McNamara was. Indeed, the 55-page report by NSA historian Robert Hanyok effectively rewrites the history of this pivotal event. Hanyok proves not only that there never was a second attack but that the NSA had plenty of contradictory evidence that was actively suppressed. "The overwhelming body of reports, if used, would have told the story that no attack had happened," Hanyok concludes. "So a conscious effort ensued to demonstrate that the attack occurred."
Getting it wrong. Even as the first news of the alleged August 4 ambush reached Washington, there were doubters. One U.S. commander warned that freak weather and faulty sonar readings might have fooled sailors. At the same time, the NSA posted what analysts described as an intercept of a North Vietnamese after-action report of the August 4 battle that acknowledged losing two ships. This intelligence swung the debate in the Johnson administration toward war.
Hanyok systematically reconstructs how the NSA created--and stuck by--this erroneous report. While some of the original source material in Vietnamese vanished, Hanyok shows that the intercepted report was actually written before the supposed August 4 attack--and that it most likely referred to the August 2 attack. Further, the NSA report was plagued by translation errors and based on two unrelated intercepts strung together as one.
Even more damning, NSA analysts relied on only six intercepts to assemble their account. But Hanyok uncovers another 53 related intercepts that were buried for nearly four decades, off limits to senior U.S. officials and congressional investigators. The lost intercepts--90 percent of the NSA's related intelligence--show that the North Vietnamese boats suspected in the August 4 attack were busy salvaging boats damaged two days earlier. And North Vietnamese were actively trying to avoid U.S. ships.
Hanyok concludes officials "deliberately skewed" intelligence to support the claim of a second attack. But he blames lower-level NSA analysts, not senior political leaders. "The conspiracy was always believed to be that President Johnson and Secretary McNamara basically manipulated the intelligence . . ." says intelligence historian Matthew Aid. "Now, we know the mistakes and conspiracy were widespread but at a midlevel." Still, the Johnson administration set the climate. For one thing, there was intense White House pressure to confirm the attack to clear the way for immediate retaliation. Hanyok quotes Ray Cline, the CIA's deputy director for intelligence at the time: "We knew it was bum dope that we were getting . . . but we were told only to give the facts with no elaboration on the nature of the evidence. [Johnson] did not like to deal with uncertainties."
The parallels to the current debate over the Bush administration's case for war in Iraq are unavoidable. Some of the same mistakes, including analytical errors and the exclusion of contradictory evidence, plagued the administration's use of intelligence about Iraq's alleged weapons of mass destruction. Aid spent 18 months pushing the NSA to release the report: "I think there was some concern that this was just the kind of thing the intelligence community does not need at this point, which is another massive intelligence failure that led to another war that did not go well for the United States."
This story appears in the December 12, 2005 print edition of U.S. News & World Report.