Fleet Admiral William D. Leahy’s
Estimate of 63,000 American
Casualties in an Invasion of
Japan, and the Use of the Number
by the Smithsonian’s National Air
Space Museum

Excerpt from:  

Casualty Estimates: D. M. Giangreco’s
Rebuttal of Barton J. Bernstein,” July 31, 1998

We [Giangreco and Bernstein] talked for two hours; Bernstein was vocal in calling my attention to a grievous misinterpretation in our changed script [proposed for the Smithsonian’s Enola Gay exhibit]. He claimed that the casualty estimates for an invasion of Japan given in the text hammered out with the American Legion . . . misrepresented Admiral Leahy's June 18, 1945, remarks to President Truman. Bernstein claimed that Leahy's diary entry for that same day stated he had meant that there would be 63,000 casualties - far lower than the "quarter of a million casualties" figure we had attributed to him in our revised casualty label.

Bernstein told a reporter that he had challenged me on these figures, and added that I was unable to refute him. That is quite true. The reason was simple. I was dumbfounded! The figure of a quarter of a million had come from a paper that Bernstein himself had written in 1986. So when he suddenly told me, in the middle of a large meeting, that we had made a grave mistake, I did not know how to respond. I was pretty sure we had based our numbers on his paper but on the spur of the moment I could not swear to that. I don't recall Bernstein stating that his article had been in error. (27)

In a letter to the American Legion on 9 January 1995, Harwit [Director of the National Air and Space Museum Martin Harwit] explained that Bernstein "took us to task" for misunderstanding what he now said he meant in the article. Harwit noted that Bernstein "in the meantime had found Leahy's diary entry" (28) and, on the basis of this discovery, he had decided to modify the text yet again - a modification which now clearly implied deception on Truman's part. The rest, as they say, is history.

Ferrell is also taken to task for not finding Bernstein's discovery to be worthy of inclusion in Truman and the Bomb. Ferrell had several good reasons to exclude it, not the least of which is that the 63,000 figure, reputed by Leahy to have come from Marshall, is inconclusive as to what it actually represents. This is why military historians do not use it and why I did not use it and a variety of other guesstimates in my JMH piece (long before the Enola Gay controversy, it and guesstimates like it were referred to within the hallowed halls of the US Army Command and General Staff College [CGSC] as "junk numbers"). I do not mind so much that it is secondhand, but knowing Admiral Leahy's unfamiliarity with Army methodologies used in producing casualty estimates, the figure may or may not represent just one month of combat in what was anticipated by Marshall to be an extremely lengthy campaign. In any event, there were considerably better and more appropriately documented estimates to run in my long and somewhat dense article. However, if I had been aware during the article's production that Bernstein had been successful at selling the number to Harwit and the Smithsonian as something it was not, I would have made note of it in JMH.

There are several ways Leahy might have arrived at the 63,000 figure, but Leahy's expression "would not cost us" represents a guarantee that Marshall would not give to Leahy or Truman for all the institutional reasons given in my casualties piece. I'm afraid that Bernstein is going to have a tough time finding someone at the CGSC (and I suspect the Naval War College and George C. Marshall Library as well) who believes that, even if it did come from Marshall - and is not just an interpretation of what Leahy thought Marshall meant - that the Army chief would characterize the number as a ceiling which could not be exceeded.

So what does the 63,000 figure represent? Other than a purely 30-day estimate for just the ground force losses there are a couple possibilities:

First, Leahy had made it a point, during the 18 June 1945 meeting with the President, to note that just over one-third of the ground forces employed on Okinawa had become casualties, and the number is almost exactly one-third of a straight head count of the divisions landing at Kyushu. If you exclude the very substantial nondivisional combat units and the service and support units and take a snapshot of the TOE on X-Day (13 divisions and 2 regimental combat teams), you get 190,000 men, and a third of that is your 63,000. Unfortunately, over the course of about three or four months of heavy ground fighting, these units would functionally turn themselves over (100 percent casualties on paper) with the line regiments - what the Brits call "poor bloody infantry" - turning over at two and possibly even three times that rate like they had at Okinawa, Iwo Jima, San Pietro, Normandy's hedgerows, the Hürtgen Forest, etc. As noted in a July 1950 manpower piece in Military Review, "It was not a question of whether [an infantryman] would become a casualty, but rather of when and how." (29)

Second, an apparently old joke that I'd heard twice from instructors at the CGSC's Department of Combat Support in 1985(30) and repeated at least once before the Smithsonian imbroglio was that Leahy had just doubled the Luzon number of 31,000 casualties from the 18 June meeting because of the stated 2-1 Okinawa ratio. The joke goes that Leahy added "a thousand more bodies to be safe," and put it in his diary as his best guess as to what he thought Marshall meant. (Logistician humor; I guess you just had to be there.)

Virtually every serious military historian who works in the World War II-Cold War period is aware of Leahy's comment through its use in his widely read autobiography, I Was There, (31) yet discard it as too vague to be useful. This should have raised warning flags as to the credibility that specialists in this field attach to it. But Bernstein apparently did not notice the danger signs because, in spite of the long lists of books and archives he unscrolls in his essay, he is only vaguely familiar with military aspects of the subject and its literature. He either did not notice the flashing red lights or is convinced that he possesses a superior understanding. And irrespective of the fact that Bernstein does not have a clear idea where Leahy fits into the picture (students of the period will be surprised to find that the president's chief of staff at the White House has been promoted by Bernstein to the position of "wartime chairman of the Joint Chiefs"), it is still impossible to believe that he could have missed Leahy's use of the number in a book that Bernstein himself quotes from. Characterizing it as a discovery to Harwit certainly leaves the impression that Bernstein was taking unfair advantage of Harwit's lack of knowledge of military historiography.

Use of Leahy's 63,000 figure may have started out as simply a rhetorical device intended to shock Harwit into rejecting the criticisms of veterans groups - and who would be the wiser? The group of professors pressing Harwit to show some backbone in his dealings with the veterans were as untutored in this area as the museum director and operated within a back channel well hidden from public view. How was Bernstein to know that Harwit would later write a well-received book (32) that would signal him out as the individual who forcefully and successfully pressed for acceptance of Leahy's vague number in the Enola Gay script - a change which led directly to Harwit's exit from the National Air and Space Museum. Throughout the sorry episode, Bernstein moved from Smithsonian consultant, to critic, and finally to active participant in a short-lived organization he helped found: Historians for Open Debate on Hiroshima.


Bernstein's successful promotion of the idea, repeated in his essay, that "it is highly unlikely that Marshall in June [1945] assumed even up to 100,000 U.S. casualties in Olympic" had far-reaching consequences during the late unpleasantness over the Enola Gay exhibit, but it is also amazing that Harwit apparently didn't contact any of the military specialists working with the Smithsonian in order to get their thoughts on what the 63,000 number actually represented before using it as the basis for a ratcheting down of expected invasion casualties in the exhibit. If he did, neither he nor they ever mentioned it in print. (33) But while it is evident that Harwit was certainly led down a treacherous path, the nearly two full months between Bernstein's bombshell at the November 17 meeting and the 9 January American Legion letter have left the impression that he went perhaps a little too willingly. We will never know how events would have unfolded if, soon after the meeting, Harwit had learned that Bernstein's earth-shaking discovery in Leahy's diary had, in fact, been printed verbatim in the admiral's widely read book - and one which Bernstein quotes from!




27. Martin Harwit An Exhibit Denied: Lobbying the History of the Enola Gay, (New York: Copernicus/Springer-Verlag, 1996), 345-46.

28. Harwit, An Exhibit Denied, 380.

29. Lieutenant Colonel U. P. Williams, "They May Not Die -- But They Wither Fast" Military Review (July 1950): 16.

30. The Department of Combat Support at the CGSC has since changed its name to Department of Logistics and Resource Operations.

31. William D. Leahy, I Was There: The Personal Story of the Chief of Staff to Presidents Roosevelt and Truman Based on His Notes and Diaries Made at the Time (New York: Whittlesey House, 1950), 384.

32. As late as 16 November 1997, the New York Times Book Review ran a review essay on Harwit's 1996 book as its cover story, "Tailspin: The Ex-director of the Air and Space Museum Tells How the Enola Gay Exhibit Crashed."

33. Richard H. Kohn, "History and the Culture Wars: The Case of the Smithsonian Institution's Enola Gay Exhibition," Journal of American History (December 1995): 1036-63; Richard P. Hallion and Herman S. Wolk, "Air and Space Museum guilty, as charged" Bulletin of the Atomic Scientists (July/August, 1995): 75.