The Doomsday Machine

by Daniel Ellsberg

  Whether that was true or not, the U.S. Army Air Force came out of the war convinced it had won the war in the Pacific by burning masses of civilians to death. Certainly that was the conclusion of Curtis LeMay. In contrast, his civilian superiors, Truman and Stimson, denied to the end of their lives that the commanders and forces under their authority had ever violated the code of jus in bello by deliberately targeting noncombatants. In LeMay’s eyes, that was something of a semantic question. In a lengthy interview with historian Michael Sherry, he said, “There are no innocent civilians.182 It is their government and you are fighting a people, you are not trying to fight an armed force anymore. So it doesn’t bother me so much to be killing the innocent bystanders.”

  In the early sixties, my RAND colleague and friend Sam Cohen told me he had once been in a meeting at Air Force Systems Command when its commander, General Bernie Schriever (who pressed the development of our ICBM) asked LeMay, “What is your requirement for a large warhead?” That is, what’s the largest yield you need, what would be “large enough”? LeMay answered, “One bomb, for Russia.”

  In the ensuing discussion, Sam told me he had argued for the development of smaller bombs, more usable in limited wars like Korea, that would cause fewer unintended victims. He was a physicist and bomb designer who liked to be known as the “father of the neutron bomb.” LeMay, who had a friendly, fatherly feeling toward Cohen, drew him into an adjoining empty room, just the two of them, put his arm around his shoulders and told him, “Sam, war is killing people. When you kill enough of them, the other guy quits.”

  Whether or not they consciously shared it, General LeMay’s viewpoint was well known to the presidents—Truman, Eisenhower, Kennedy, and Johnson—who placed him and kept him in charge of nuclear war plans and implementing forces that embodied that perspective for fifteen years, as commander of the Strategic Air Command and later chief of staff of the U.S. Air Force.

  CHAPTER 16

  Killing a Nation

  In August 1945 the atom bomb was simply fitted into a long, secret pattern of war making by the massacre of civilians. The atomic attacks seemed to vindicate that pattern by the sudden ending of the war against Japan that followed almost immediately and which, so far as the public and troops knew (in ignorance of our secret intercepts of Japanese communications), had no other way of being achieved. The military service that delivered the bomb had no trouble, after all, winning its independence from the other services soon after the war, and no great resistance to accepting its own subsequent domination by the Strategic Air Command, built up and commanded in turn by Generals Curtis LeMay and Thomas Power, an organization which was committed to the tactics of extermination perfected in the last six months of World War II.

  But against what adversary was it now directed? As World War II came to an end, only one country was left with the population, armed forces, and industrial and scientific strength to challenge the United States militarily: the Soviet Union—despite having suffered virtually unprecedented wartime destruction and casualties. Moreover, it was ruled by a dictator as ruthless as Hitler and by a single party more cohesive and competent than the Nazi Party; it already occupied half of Europe and had the military strength to take over the other half. Increasingly, and with various underlying motives, some high-level members of the Truman administration came to adopt and promote a fear that the Soviet Union intended to do just that.

  This was not a new perspective to General Leslie Groves, who was in charge of all aspects of the atomic Manhattan Project. As early as 1944, the Polish physicist Joseph Rotblat, having dinner with Groves at Los Alamos, had been shocked to hear the fervently anti-Communist general say to him that, in his eyes, the project had always been aimed at confronting the Soviets. The Army Air Corps had a similar view. Looking for a target system that would justify a large postwar force of strategic bombers and thus an independent Air Force, it turned its eyes to the Soviet Union.

  On August 30, 1945, just two weeks after Japan surrendered, Major General Lauris Norstad, assistant chief of Air Staff for plans, sent General Groves a document identifying for possible future atomic attack fifteen “key Soviet cities,”183 headed by Moscow, and twenty-five “leading Soviet cities,” including Leningrad, and specifying the number of atomic bombs needed to destroy each. Moscow and Leningrad would require six apiece.

  But the United States didn’t have six atomic bombs in 1945. At the end of the year it had two. By June 30, 1946 (the end of the fiscal year), nine bombs were in the stockpile. The first official war plan against the Soviet Union, in November 1947, called for hitting twenty-four Soviet cities with thirty-four bombs. But there were only thirteen bombs in the U.S. arsenal at that time, perhaps only seven complete weapons. The war planners didn’t know that. It was a super secret. President Harry Truman himself wasn’t formally briefed184 on the number until April 3, 1947, when he was shocked to find out it was so small.

  Two months earlier the Joint Chiefs of Staff had told the secretaries of war and Navy that the supply of atomic weapons was “inadequate” to meet U.S. security requirements. Until late 1948, all the weapons produced—all Nagasaki-type plutonium implosion bombs—were in many ways hand tooled, considered as “laboratory weapons.”185 The JCS evaluation of the Bikini Atoll tests in the summer of 1946—which had used up two of the nine weapons available that year—“concluded that because of the scarcity of fissionable material, the bomb would have to be used as a ‘strategic’ weapon against urban industrial targets.” But General LeMay, then in charge of Air Force research and development (in which capacity he supported the creation of Project RAND), summarized the report’s main conclusions:

  Atomic bombs in numbers186 conceded to be available in the foreseeable future can nullify any nation’s military effort and demolish its social and economic structures.

  In conjunction with other mass destruction weapons it is possible to depopulate vast areas of the earth’s surface, leaving only vestigial remnants of man’s material works.

  In October 1947, a report on longer-run bomb requirements was sent to the Atomic Energy Commission (AEC), which was now in charge of all aspects of bomb production, from the Joint Chiefs by their de facto Chairman, Admiral William D. Leahy. Two years earlier as Truman’s chief of staff, as Leahy recounted in a memoir, he had privately deplored dropping either the Nagasaki or the Hiroshima bombs on cities, believing that “in being the first to use it, we had adopted an ethical standard common to the barbarians of the Dark Ages. I was not taught to make wars in that fashion, and wars cannot be won by destroying women and children.” Now he reported to the AEC that a “military requirement exists for approximately 400 atomic bombs of destructive power equivalent to the Nagasaki bomb” to be dropped on approximately one hundred urban targets. The target date for achieving that capability for “killing a nation”187—a concept that arose in the Air Staff that prepared the recommendations—was January 1, 1953.

  By mid-1948, Air Force plans were coming into line with the stockpile, though that was far below what the JCS regarded as adequate. The plan at that time was to hit twenty cities with fifty bombs.188 There were fifty bombs in the arsenal189 on June 30, 1948. Moscow would be hit with eight bombs, Leningrad with seven.

  General LeMay became head of SAC in October 1948. He drew up its Emergency War Plan (EWP), which called for SAC to “increase its capability to such an extent that it would be possible to deliver the entire stockpile of atomic bombs, if made available, in a single massive attack.” Primary objectives would be urban industrial concentrations and government control centers. Secondary objectives included petroleum production; two-thirds of that was within sixteen Soviet cities. The plan entailed strikes on seventy Soviet urban areas190 with 133 atomic bombs. Estimates said the plan might kill 2.7 million people191 in the seventy target cities, with four million additional casualties.

  A year later, in October 1949, the target annex for the Emergency War Plan called for attacks with 220 bombs on 104 urban targets
, plus a reattack reserve of 72 weapons. The 292 bombs required for this were available by June 30, 1950. The AEC, with three separate budget increases for bomb production by Truman after the Berlin blockade in 1948–49 and the first Soviet atomic test in August 1949, was now turning out Nagasaki-type bombs on a production line. The era of “nuclear scarcity,” in Pentagon terms, was giving way to “nuclear plenty.” The 400 bombs required for killing a nation were in the stockpile by January 1, 1951,192 two years ahead of schedule. But by this time, the targets requiring atomic attack, in the eyes of Air Force planners, had expanded manyfold.

  For the first four years of the nuclear era, the JCS, the newly independent Air Force, and the newly formed Strategic Air Command had been making plans for attacking a nation that posed no military threat, conventional or atomic, to the homeland of the United States. These were only first-strike plans, in later terms, though not thought of as such at the time because there was no adversary that could strike second.

  America had a monopoly of atomic weapons, which President Truman and General Groves (though not the nuclear scientists, if they had been asked) expected to last for a generation or more. He and Truman foolishly believed that by a highly secret program of purchase and diplomacy, they had succeeded in controlling all the known high-grade sources of uranium. (Groves had overlooked,193 he said later, high-grade supplies of ore in East Germany, occupied by the Soviets.) That program was, in their eyes, the critical “atomic secret.” It was in that mistaken belief that Truman had sought and achieved the consent of the Senate to commit the United States, for the first time, to the defense of Western Europe by NATO.

  Scientists had pressed for international control of uranium supplies, research, and all enrichment and possession of fissile material for energy, predicting in 1945 that otherwise the Soviets would have a bomb in about four years. Four years later, in September 1949, U.S. intelligence flights detected evidence that the Soviets had conducted a test of a Nagasaki-type plutonium implosion bomb. (It was, in fact, a replica of the Nagasaki bomb, based on blueprints supplied by Klaus Fuchs, a Soviet spy at Los Alamos.) Truman, Groves, Congress, the American public, and our NATO allies were shocked.

  The JCS, however, didn’t panic. They soon correctly estimated that it would be years before the Soviets had the means of delivery or sufficient weapons to threaten the United States itself. But in SAC planning, urban-industrial areas now ceded the very highest priority for an attack to a target system related to the future Soviet delivery of atomic weapons on the United States and its allies. That implied an almost unlimited multiplicity of urgent targets for an American atomic attack—above all, airfields, of which there were eleven hundred in the USSR, most in or near cities. By 1953, General LeMay had identified 409 airfields194 that could be used for a nuclear attack, along with nuclear production facilities of all kinds sprawled across the USSR.

  In the fall of 1949, the production of fissile material was again accelerated, to provide warheads for an expanding set of targets and weapons of all kinds to deliver them. When Truman left office in early 1953, a thousand atomic weapons were in or scheduled shortly to be in the U.S. stockpile. At the end of his two terms, President Eisenhower bequeathed to the Kennedy administration eighteen thousand nuclear weapons.

  While the target system remained essentially what it was in the early fifties, the eighteenfold increase in the number of nuclear weapons—many of them now shorter-range “tactical” weapons averaging the yield of the Nagasaki bomb—did not begin to measure the meaning to human survival of the change in the nature of the strategic weapons, more than ten thousand of them, carried by SAC and the Navy. The meaning of “nuclear” had changed, in a way largely hidden, deliberately, from the people of America and the world. The great majority of the weapons in the nuclear arsenal that President John F. Kennedy inherited in 1961 were not “atomic” weapons of the type used on Japan in 1945 and later tested at Bikini and Nevada, based on fission of isotopes of the heavy elements uranium or plutonium. Until the early fifties, those A-bombs had been the only kind of nuclear weapons in existence. But by 1961 virtually all SAC’s weapons were “thermonuclear” weapons—hydrogen bombs, or H-bombs, based on the fusion of heavier isotopes of hydrogen—which were first tested in November 1952.

  It was that change, I discovered in 1961, that explained what had been a striking puzzle to me earlier in the year. In the course of reviewing Top Secret documents associated with various JSCPs in the fifties, as background for drafting guidance for the JSCP under the Kennedy administration, I had seen successive estimates for Soviet casualties in general war that in the early years of the decade seemed surprisingly “low” for the nuclear era: a few million deaths, then ten million, then up to thirteen million or so by 1955. But from that year to the next, 1956, there was a sudden tenfold jump in the estimates—an order-of-magnitude increase, as RAND analysts would put it—to a hundred and fifty million Soviet dead. By 1961, as I had already learned, the JCS forecast was for more than two hundred million in the Soviet bloc alone. Why this increase? Why just then?

  My shock at this, described in the prologue and chapter 9, was accompanied by questions in my mind: How and why had any planners or decision makers proposed this increase? Had someone concluded that “killing a nation” with four hundred atomic bombs that would kill tens of million Russians was not enough devastation for deterrence? Or perhaps that the fulfillment of our commitment to NATO to respond to or preempt a Soviet ground invasion absolutely required this much additional “collateral damage”? On what basis might they have reached either of those judgments?

  The reason for the jump, from one year to the next, in the number of deaths we were preparing to inflict in a war against Russia—from numbers that were huge but still less than Soviet fatalities in World War II (though inflicted in days and months rather than years), to levels that were totally unparalleled in human history—turned out to be neither of the above explanations. It was much simpler.

  There was no new judgment of the necessity for the dramatic change in the planned-for effects of our attack. The war planners were simply assuming, correctly, that SAC meant to replace their atomic weapons of the first decade of the nuclear era with the newly available H-bombs, thermonuclear warheads, against essentially the same ever-expanding target system. That entailed SAC’s preparedness to kill ten times or more the number of people as before. Not tens but hundreds of millions of dead, perhaps a billion, largely from radioactive fallout from hydrogen bombs, of which hundreds in the SAC arsenal were a thousand times larger in yield than the atom bombs of World War II.

  This change was introduced not because it was judged by anyone to be necessary, but because it was simply what the new, more efficient nuclear bombs—cheaper but vastly larger in yield than the old ones—could and would accomplish when launched against the same targets. (One contributing factor in this increase in casualties was the fact that nearly all the nuclear targeting in the late fifties and early sixties planned ground-burst explosions of thermonuclear weapons, deliberately creating greater radioactive fallout and “bonus” casualties in the Sino-Soviet bloc—though, regrettably, also among their neighbors, including our allies and neutrals.)

  These estimates of U.S.-inflicted deaths were so secret and so closely held, even in SAC and the Pentagon, that very few Americans outside or within the government were aware of the drastic change in the meaning of “nuclear war” that had occurred in the late fifties or how it had come about. In the interests of allowing atmospheric tests of thermonuclear weapons to be carried on in the continental United States, despite their predictable effects on “downwinders” in Nevada and Utah, President Eisenhower had done his best to maintain that lack of awareness in the public about the changes in nuclear weapons and their effects. He told Gordon Dean, chairman of the AEC, to leave the terms “thermonuclear,” “fusion,” and “hydrogen” out of press releases and speeches and to “keep them confused195 about ‘fission’ and ‘fusion.’ ” But as I di
scovered to my surprise in the spring of 1961, the JCS, the Joint Staff, and President Eisenhower himself were aware of the horrific potential consequences in Eurasia of their preparations.

  Eisenhower had been “appalled”196 in late 1960 by the prospects of “overkill” (reported to him by his science advisor George Kistiakowsky) in SIOP-62—especially by its uneconomic redundancy of target coverage, but surely not only by that. He told his naval aide that the presentation “frighten[ed] the devil out of me.” Nevertheless, he approved the plan and passed it on to Kennedy. When in July 1961 JFK was briefed on the projected results of a 1963 exchange, he commented in shock on leaving the briefing room, “And we call ourselves the human race!”197 But he made that comment to Secretary of State Dean Rusk, not to the JCS and certainly not to the public; and that all-out “option” remained in the plans throughout his brief time in office and throughout that of Lyndon Johnson.

  President Nixon in 1969198 was reportedly likewise “appalled” to learn in January 1969 in his first briefing on the SIOP that the only available options were for massive nuclear strikes involving thousands of weapons, some killing ninety million Russians in hours. His national security assistant199 Henry Kissinger said that such plans were not the basis for “politically plausible,” sufficiently credible threats. Later in the spring he asked in a meeting, how can “one rationally … make a decision to kill eighty million people?” But his efforts in the next eight years to add less murderous options to the plans (like the efforts of Robert McNamara before him, with my help) came to little or nothing.

  In 1973, midway in his abortive search200 for more limited and credible alternatives, Kissinger asserted in another meeting, “To have the only option that of killing eighty million people is the height of immorality.” (It was not, in fact, the only option in the plan; all the others killed many more.) But his private judgments about morality remained entirely secret from the American public until declassified decades later. Presidents Ford, Carter, and Reagan all multiplied alternatives for “limited nuclear options” that would be less apocalyptic, but as General Lee Butler, the last commander of SAC, has revealed,201 the war planners in Omaha and the Pentagon never took any of these proposals seriously, either in their operational planning or in rehearsals for the war they expected and planned to be “all-out.”