A Vicious Entanglement, Part V: The Body Count Myth


Editor’s Note: This is the fifth installment of “A Vicious Entanglement,” a series by Jon Askonas on The Vietnam War, a new documentary series by Ken Burns and Lynn Novick. Read the first four installments: “Elusive Responsibility for the Vietnam War,”  “Idealism Chastened in Vietnam,” “The Asymmetries of Vietnam” and If Only the North Vietnamese Had Read Galula.”

One of the more cogent criticisms of Ken Burns and Lynn Novick’s The Vietnam War is that the filmmakers end up telling a quite conventional story. At times, they obscure more recent historical work that complicates our understanding of the war and rely instead on the first draft embedded in collective memory. Nowhere is this truer than in the brief but revealing treatment of the body count, the measure of how many enemy soldiers American operations had killed. The filmmakers reinforced a false view on the problems of the body count that still influences how military professionals and scholars think about it today. The reality is much more complicated. Neither of the two common criticisms of the body count – that it was inaccurate or that it was the wrong thing to measure – holds up to scrutiny. The humbling lesson we need to learn from the body count is not about picking the right metrics, but about the intrinsic, unavoidable limits of our tools for measuring and managing progress in war.

Resolve (January 1966-June 1967), the fourth episode in the series, only briefly touches on the subject, but it leaves the viewer with a totally misguided understanding of how Military Assistance Command, Vietnam (MACV) used enemy body count, how Gen. William Westmoreland made sense of the war, and why it was so difficult for U.S. military leaders to assess success. The documentary tells us the conventional line, that MACV over-relied on body count as part of a strategy of attrition. The narrator grimly intones that, lacking other good metrics, “MACV fell back more and more on a single grisly measure of supposed success: counting corpses. Body count.” Not only is this a bad measure, we’re told, but the numbers weren’t even right. Joe Galloway (the reporter of Ia Drang fame) tells us, “the numbers are lies, most of ‘em….If body count is your success mark, you’re pushing otherwise honorable men…to become liars.”

It has become fashionable to disparage the body count as evidence of an overly conventional notion of how to fight the war (the solution to which is, according to the logic of this critique, counter-insurgency doctrine). This narrative, which Burns and Novick largely buy into, gravely misrepresents the problems with the body count in two ways. First, Westmoreland was circumspect about the accuracy of the body count and not overly dependent on it for making sense of the war. In total, the American command took in over 180 metrics. None of them sufficiently captured the complexity of the war or helped American leaders see through the fog. Second, the biggest problems with body count – and the greatest inaccuracies – occurred not in MACV headquarters but in the field, where the body count created perverse incentives and where it actually was over-relied upon as a measure of success. Unless otherwise noted, the details below are derived from Greg Daddis’ masterful No Sure Victory.

The Body Count in MACV HQ

First, an in-depth review of MACV practices makes it hard to uphold the idea that Westmoreland was obsessed with body count. Not only was he aware of the limitations of the metric, he also utilized a variety of other measures to understand the war. The body count was far from the only metric MACV or the military leaders in Washington collected or cared about. Later in the episode, Burns and Novick briefly mention the Hamlet Evaluation System (HES), a pioneering system for evaluating pacification at the smallest of scales. They don’t mention that it was one of dozens of systems for collating information about friendly and enemy activity, intelligence on the enemy, assessments of host nation forces training, and popular opinion. IBM 360 mainframes in Saigon churned day and night crunching numbers from the HES and other metrics. The result was a deluge of data that nobody knew how to make sense of, absent a more definitive strategy allowing a prioritization of different metrics.

Headquarters was aware of the limitations imposed by the way body count was collected and were relatively realistic about its accuracy. Westmoreland caught a lot of flak for claiming that body counts were “conservative.” People like Galloway, who had seen how body counts were haphazardly collected at platoon level, rightly rubbished the idea. But, while he was optimistic about accuracy, Westmoreland was more or less right that body counts were reasonably accurate at the MACV level. The plurality of casualties in the body count were not from small infantry patrols (most susceptible to inflation) but from artillery and bombing actions, where assessments were more controlled and conservative. Given what we now know about North Vietnamese casualties, it does not appear that American body count estimates were off by more than fifteen or twenty percent.

The accuracy of body counts at the MACV level was not the issue. When it came to a strategy of attrition and the “crossover point,” the issue was not that Westmoreland had inaccurate information about the enemy forces being destroyed. The crossover point strategy relied on a basic equation: on one side, the number of soldiers the North Vietnamese could (or were willing to) recruit, train, and deploy into South Vietnam, on the other, the number the U.S. military could kill, capture, or deter. With the benefit of hindsight, what appears dramatically off in Westmoreland’s understanding of this equation is not the body count. Rather, the U.S. military, vastly underestimated the casualties the enemy was prepared to take as well as the degree of control the enemy was able to exercise over his losses by choosing when not to engage with the Americans.

Moreover, collecting information on body count was vital in its proper context. The Vietnam War was far from a conventional counter-insurgency. Much of the U.S. effort was focused on combating well-equipped, centrally organized North Vietnamese Army and Viet Cong units. For the most part, these were not part-time insurgents hiding “amongst the people who might be induced to lay down arms or go over to the other side. And so Westmoreland concluded that pacification and counter-insurgency efforts against the part-time, population-embedded “red guerrillas” would be for naught without the aggressive destruction of VC/NVA military units operating in the South. But MACV was never as successful at pacification and political development as it was at search-and-destroy missions, partially because the South Vietnamese units the Americans relied on for the task varied dramatically in quality and commitment. As a result, VC units that were decimated by American action only had to wait for the bulk of American forces to leave, and then they could resume business as usual. One can argue that this was bad counter-insurgency strategy, but it’s evidence of MACV grappling with a complex threat, not a straightforward confusion about the kind of war they faced. In principle, the body count provided useful information about this “big unit war”.

The Body Count in the Field

But, second, while the body count was not the root of the problem at the MACV level, it was a serious problem in the field. As Robert Gard Jr. tells Burns and Novick in an interview, “If body count is your measure of success, then there’s a tendency to count every body as an enemy soldier.” In some ways, body count was a straightforward victim of Goodhart’s Law: “When a measure becomes a target, it ceases to be a good measure.” If MACV had somehow been able to collect information about body count in total secrecy, it would have remained a valuable measure of progress. But the same people who were collecting information about the body count were also those being evaluated by it.

Then, as now, the Army’s organizational structure incentivized officers to rely on and artificially inflate the body count, pumping up numbers by going on unnecessary missions or double-counting enemy dead. Army officers in Vietnam generally served six- or twelve-month field command tours. But undertaking rural construction, increasing political participation, and building support for the government all take a long time, much longer than six months or a year. As field officers sought to prove their mettle to superiors who would conduct their performance evaluations, they could not rely on making progress in these long-term endeavors during their short tenures. The one metric that a unit commander could reliably affect was body count, by focusing on aggressive pursuit of the enemy at the expense of the counter-insurgency/pacification mission.

And so, even as MACV HQ tried to balance the “military war” and the “other war,” an organizational culture inclined towards aggressiveness, combined with incentives to pursue body count, created strong pressure on field-grade officers to pursue the military search-and-destroy mission at the expense of the political mission. In maybe the most egregious example, body count enthusiast Major General Julian Ewell, commanding Operation Speedy Express, led a 9th Division which racked up 10,899 dead VCs and 2579 captured, but only turned up 748 weapons. Pacification leaders were furious at the heavy-handed approach, which undid years of work. Ewell was subsequently promoted and given a higher command by Westmoreland’s successor, Creighton Abrams.

More recent historical work has complicated the idea that Westmoreland and MACV pursued a conventional strategy of attrition, ignoring or downplaying counter-insurgency. But one of the American military’s problems in Vietnam was precisely that it never moved away from the logic of attrition, which applied even across a bevy of variables. The fundamental logic of metrification is also the fundamental logic of attrition: Make the good numbers go up and the bad numbers go down. Even a counter-insurgency, run in this manner, relies on the idea of attrition (as we have generally seen in Afghanistan as well). In contrast, the North Vietnamese, who lacked a sophisticated assessment strategy, ran an almost textbook war of maneuver, focusing on initiative, tempo, surprise, and cultivated asymmetry at every level of warfare.

Burns and Novick do their audience a real disservice by reinforcing two of the most flawed “lessons” from Vietnam: a) the problem with the body count is that it was inaccurate (solution: just use accurate metrics) and b) the use of body count is evidence that the U.S. military pursued an overly conventional strategy of attrition through search-and-destroy (solution: conduct counter-insurgency instead). In reality, the problem is that metrics have inherent constraints as assessment tools and performance targets. The United States has run into many of the same problems in Iraq and Afghanistan, albeit with different inflationary pressures on metrics. While there is a robust and growing literature on how to use metrics in counter-insurgency, you can’t get around the facts that showing progress in a variety of indicators does not amount to a strategy and that anytime you use an indicator as a target, you will damage its usefulness as an indicator. Until the U.S. military grapples with the real problems MACV faced – a bewildering, complex war and unsatisfying metrics – the story of the body count in Vietnam will remain not a “lesson learned” but a calming fairy tale.


Jon Askonas is a predoctoral fellow at the Clements Center for National Security at the University of Texas-Austin and a DPhil candidate in International Relations at the University of Oxford. His current research addresses the impact US Army policies in Vietnam had on knowledge production, learning on the ground, and battlefield effectiveness.

Image: U.S. Army