The Doomsday Machine

by Daniel Ellsberg

  Far from being accompanied by any offers to resign, there was no evident embarrassment, no shame, apology, or evasion: no apparent awareness of any need for an explanation of this answer to the new president. I thought: this was what the United States had come to, sixteen years after Hiroshima. Plans and preparations, awaiting only presidential order to execute (and, I’d discovered, not requiring even that in some circumstances), for whose foreseen consequences the term “genocidal” was totally inadequate.

  I myself at that time was not a critic of the explicit logic of deterrence or its legitimacy. On the contrary, I had been urgently working with my colleagues at RAND and in the Pentagon to assure a survivable U.S. capability to threaten clearly unacceptable damage to the Soviet Union in response to the most successful Soviet nuclear attack on the United States. But the planned slaughter of hundreds of millions of Soviets (and Chinese), twenty times the staggering deaths of Soviet citizens in World War II, along with an equal number of our allies and citizens of neutral countries! That expected outcome exposed a dizzying irrationality, madness, insanity, at the heart and soul of our nuclear planning and apparatus.

  The fact is that the estimate of fatalities, in terms of what was calculable at that time—even before the discovery of nuclear winter—was a fantastic underestimate. More than forty years later, Dr. Lynn Eden, a scholar at Stanford’s Center for International Security and Cooperation, revealed in Whole World on Fire71 the bizarre fact that the war planners of SAC and the Joint Chiefs—throughout the nuclear era to the present day—have deliberately omitted entirely from their estimates of the destructive effects of U.S. or Russian nuclear attacks the effects of fire.

  They have done so on the questionable grounds that these effects are harder to predict than the effects of blast or fallout, on which their estimates of fatalities are exclusively based, even though, as Eden found, experts including Hal Brode have disputed such conclusions for decades. (A better hypothesis for the tenacious lack of interest is that accounting for fire would reduce the number of USAF warheads and vehicles required to achieve the designated damage levels: which were themselves set high enough to preclude coverage by available Navy submarine-launched missiles.)

  Yet even in the sixties the firestorms caused by thermonuclear weapons were known to be predictably the largest producers of fatalities in a nuclear war. Given that for almost all strategic nuclear weapons, the damage radius of firestorms would be two to five times the radius destroyed by the blast, a more realistic estimate of the fatalities caused directly by the planned U.S. attacks on the Sino-Soviet bloc, even in 1961, would surely have been double the summary in the graph I held in my hand, for a total death toll of a billion or more: a third of the earth’s population, then three billion.

  Moreover, what no one would recognize for another twenty-two years were the indirect effects of our planned first strike that gravely threatened the other two thirds of humanity. These effects arose from another neglected consequence of our attacks on cities: smoke. In effect, in ignoring fire the Chiefs and their planners ignored that where there’s fire there’s smoke. But what is dangerous to our survival is not the smoke from ordinary fires, even very large ones—smoke that remained in the lower atmosphere and would soon be rained out—but smoke propelled into the upper atmosphere from the firestorms that our nuclear weapons were sure to create in the cities we targeted. (See chapter 16.)

  Ferocious updrafts from these multiple firestorms would loft millions of tons of smoke and soot into the stratosphere, where it would not be rained out and would quickly encircle the globe, forming a blanket blocking most sunlight around the earth for a decade or more. This would reduce sunlight and lower temperatures72 worldwide to a point that would eliminate all harvests and starve to death—not all but nearly all—humans (and other animals that depend on vegetation for food). The population of the southern hemisphere—spared nearly all direct effects from nuclear explosions, even from fallout—would be nearly annihilated, as would that of Eurasia (which the Joint Chiefs already foresaw, from direct effects), Africa, and North America.

  In a sense the Chiefs can’t be blamed for their failure to foresee that the firestorms caused by their planned attacks would actually have led to the death by global famine of nearly all humanity—whether that was three billion as in 1960 or the seven billion that exist today; after all, the phenomenon of nuclear winter wasn’t predicted by environmental scientists until decades after the Cuban missile crisis.

  Still, one can ask why they didn’t explore more vigorously the possible environmental consequences of this unprecedented ecological experiment—an all-out thermonuclear war—for which they were preparing. Or why, more than thirty years since scientists first posited these dangers, and more than ten years since scientific uncertainties about their calculations have been put to rest, our plans have continued to include “options” for detonating hundreds of nuclear explosions near cities, which would loft enough soot and smoke into the upper stratosphere to lead to death by starvation of nearly everyone on earth, including, after all, ourselves.

  Yet even if I had known all this in 1961, it would scarcely have affected my reaction to the graph I held in my hand that spring morning. Moreover, that unabashed calculation by the JCS said to me then that any confidence—worse, it seemed, any realistic hope—that the alert forces on either side might never be used was ill-founded. Americans had built this machine, knowing, it turned out, that it would kill more than half a billion people if it were turned on (and they were unhesitant in reporting this to the president). Humans like that would not fail to pull the switch if ordered by a president—or, as I’ve discussed earlier, possibly on the order of a superior other than the president.

  And the presidents themselves? A few months prior, Dwight Eisenhower had secretly endorsed the blueprints of this multi-genocide machine. He had furthermore demanded, largely for budgetary reasons, that there be no other plan for fighting the Russians. He had approved this single strategic operational plan despite reportedly being, for reasons I now understood, privately appalled by its implications. And when the Joint Chiefs responded so promptly to the new president’s question about the human impact of our attacks, they clearly assumed that Kennedy would not, in response, order them to resign or be dishonorably discharged, nor order the machine to be dismantled. (In that, it turned out, they were right.)

  Surely neither of these presidents actually desired ever to order the execution of these plans, nor would any likely successor. But they all must have been aware, or should have been, of the dangers of allowing such a system to exist. They should have reflected on, and trembled before, the array of contingencies—accidents, false alarms, outage of communications, Soviet actions misinterpreted by lower commanders, unauthorized action, foolish Soviet initiatives or failure to comply with U.S. threats, escalation stemming from popular uprisings in East Germany or elsewhere in Eastern Europe—that might release these pent-up forces beyond their control or that, in ways they had not foreseen, could lead them personally to escalate or to initiate a preemptive attack.

  Eisenhower had chosen to accept these risks, to impose them on humanity and all other forms of life. Kennedy—and later, Johnson and Nixon to my direct knowledge—did likewise. There is much evidence that such catastrophic “major attack options”73 were among the choices offered to presidents Carter, Reagan, and George H. W. Bush, i.e., until the end of the Cold War. There is little known publicly about the range of options since then, although four hundred Minuteman missiles remain on high alert, along with a comparable number of Trident sub-launched missiles, each alert force more than enough to cause nuclear winter.

  Moreover, I felt sure in 1961 that the existent potential for moral and physical catastrophe—our government’s readiness to commit multi-genocidal extermination on a hemispheric scale by nuclear blast and fallout—was not only a product of aberrant Americans or a peculiarly American phenomenon. I was right. A few years later, after the humiliation of the Cub
an missile crisis and the ouster of Khrushchev, the Russians set out to imitate our destructive capacity in every detail and surpass it where possible. By the end of the decade, they had succeeded. Ever since then, there have existed two Doomsday Machines, each on high alert and subject to possible false alarms and the temptation to preemption, a situation much more than twice as dangerous as existed in the early sixties.

  To be sure, Americans, and American Air Force planners in particular, were the only people in the world who believed that they had won a war by bombing, and, particularly in Japan, by bombing civilians. But the nuclear era eventually put that demonic temptation—to deter, defeat, or punish an adversary by an operational capability to annihilate most of its civilian population—within the reach of many nations. By the spring of 1961, four states (soon to be five, now nine) had, at great expense, bought themselves that capability. Humans just like these American planners—and presidents—were surely at work in every nuclear weapons state, producing plans like these for nuclear attacks on cities.

  I personally knew many of the American planners, though apparently—from this fatality chart—not quite as well as I had thought. They were not evil, in any ordinary, or extraordinary, sense. They were normal Americans, capable and patriotic. I was sure they were not different, surely not worse, than the people in Russia doing the same work, or the people who would sit at similar desks in later U.S. administrations or other nuclear weapons states.

  I liked most of the planners and analysts I knew: not only the physicists at RAND who designed bombs and the economists who speculated on strategy (like me), but also the colonels who worked on these very plans, whom I consulted with during the workday and drank beer with in the evening. What I was looking at was not simply an American problem or a superpower problem. With the age of warring nation-states persisting into the thermonuclear era, it was a species problem.

  A few years after leaving the White House,74 McGeorge Bundy wrote in Foreign Affairs, “In the real world of real political leaders—whether here or in the Soviet Union—a decision that would bring even one hydrogen bomb on one city of one’s own country would be recognized as a catastrophic blunder; ten bombs on ten cities would be a disaster beyond human history; and a hundred bombs on a hundred cities are unthinkable.”

  In the last year of the Cold War,75 Herbert York cited Bundy’s statement in a talk at Lawrence Livermore National Laboratory (of which he had been the first director), where, along with Los Alamos, all U.S. nuclear weapons had been designed. York posed the question, how many nuclear weapons are needed to deter an adversary rational enough to be deterred? Concurring with Bundy’s judgment—as who would not?—he answered his question, “somewhere in the range of 1, 10, or 100 … closer to 1 than it is to 100.”

  In 1986, the U.S. had 23,317 nuclear warheads and Russia had 40,159, for a total of 63,836 weapons.76

  CHAPTER 10

  Berlin and the Missile Gap

  In early June 1961, just one month after Gilpatric had sent my proposed changes to the existing JCS-SAC war plan to the Joint Chiefs, the possibility that their horrific plan might soon be set in motion loomed abruptly. At the Vienna summit between Khrushchev and Kennedy on June 3–4, Khrushchev renewed an ultimatum77 he had earlier made to President Eisenhower in 1958 (after Sputnik) and then withdrawn from in the face of Eisenhower’s intransigence. Now he again warned that he would turn over all control of access to Berlin to the East Germans by the end of the current year, in connection with signing a peace treaty with East Germany.

  That would mean, presumably, that the East Germans would block our access unless we negotiated terms, including their rights of inspection of cargo. That would amount to recognizing the German Democratic Republic as a separate sovereign state rather than a Soviet puppet—something totally unacceptable to our close ally Konrad Adenauer, West German chancellor of the Federal Republic of Germany, who claimed exclusive legitimacy to represent the German nation.

  If an American military convoy—rather than either meeting East German demands or returning to West Germany—attempted to force its way through, it would be confronting East German troops that would shortly be backed up by the full power of Soviet divisions. Such a confrontation would open the door to general war, a state of affairs for which the United States still had one plan only. Khrushchev was confident that President Kennedy would not let things come to that pass.

  The Berlin threat came at the end of the meeting in Vienna, after Kennedy had made an assertion to Khrushchev that many military officers in the Pentagon regarded as disastrously inappropriate. Kennedy said that for purposes of discussion, the nuclear forces of the two nations could be regarded as “equal.” Khrushchev, indeed, seized eagerly on this acknowledgment, though remarking that in reality his generals told him that Russia was stronger.

  The Air Force itself was still, in the spring of 1961, projecting Russian capabilities in a way that appeared to validate Khrushchev’s statement, not Kennedy’s. However, they were intensely dismayed that he should express openly to Khrushchev that U.S. strategic forces (which they were estimating to be greatly inferior in numbers) were “equal” to Russians rather than superior, since this “admission” obviously strengthened Khrushchev’s hand in bargaining. They felt that this showed a combination of naïveté, with respect to bargaining procedure, and character weakness on the part of Kennedy. Like the rest of the military, the Air Force wanted the president to take a tough line in a Cold War crisis like this one.

  Nevertheless, they could not bring themselves to stiffen his spine in the manner they thought necessary by assuring him that his understanding of the strategic balance was in fact mistaken, and that we were superior. After all, major decisions on the size of the U.S. missile force, which were based on the dimensions of the supposed Soviet threat, had yet to be made. It had actually been leaks from the Air Force about the alleged Soviet superiority in ICBMs that had encouraged Kennedy to campaign for the presidency on a promise to eliminate the “missile gap” by accelerating our own missile program. For the Air Force even to entertain what Army and Navy intelligence had been saying secretly for several years—that the Soviets were actually greatly inferior to the United States in strategic capability and numbers and that they showed no signs of attempting to change that situation—might have undermined the perceived necessity for an increased missile force, and perhaps radically lowered the size of the force that the Kennedy administration would procure.

  This dilemma grew more intense for the Air Force as the crisis atmosphere deepened during the summer. Within the administration, former secretary of state Dean Acheson, acting as a high-level advisor leading a planning group with respect to Berlin, was urging strongly the necessity of standing fast in Berlin, conceding no change in our rights of access. Acheson stressed the desirability of being able to defend those rights militarily, if necessary, initially without using nuclear weapons. However, he emphasized equally strongly that in the face of superior Russian conventional force, the access could ultimately be guaranteed only if the president were willing to threaten the use of nuclear weapons. These threats, furthermore, could not be sufficiently credible, he argued, unless the president was in fact committed in his own mind to the decision to use them if and when necessary.

  When JFK asked Acheson privately78 during the summer, with only Bundy present, at what point he should use nuclear weapons, Acheson replied that the president himself should give that question careful and private consideration well before the time when the choice might present itself, that “he should reach his own clear conclusion in advance as to what he would do, and that he should tell no one at all what that conclusion was.” Evidently, Acheson feared that JFK’s private conclusion, if leaked, would not be such as to deter Khrushchev. Bundy believed, in retrospect, that Acheson’s own answer to the president’s question would have been that “the right final choice might be to accept defeat, and the loss of West Berlin, if the only remaining alternative were to start a nucle
ar war.”

  I would have agreed with that wholeheartedly. I felt, like Acheson did, that it was highly important to maintain our position in Berlin, if possible. But I would never have believed nuclear war in Europe or elsewhere was justified in this effort. And consciously, I recoiled from the policy of relying on a threat that I thought must never be carried out. Along with several of my RAND colleagues, including Harry Rowen and Morton Halperin, a young RAND consultant on arms control, I believed that for the United States to initiate limited or general nuclear war under any circumstances would be catastrophic. We felt very strongly about this, though it was a position that opposed explicit U.S. defense policy and strategy in NATO, which rested centrally on U.S. readiness to carry out its threat and preparations for nuclear first use against a large Soviet conventional attack. And I had been given almost exclusive reason to believe that Robert McNamara secretly agreed with us on this.

  In early July, Alain Enthoven had arranged for me to have a brief luncheon with McNamara, to discuss my work on the guidance to the JCS on the war plan, which he had already approved and sent to the Chiefs. We ate at his desk, in his office. It was scheduled to last only half an hour, but it went on nearly an hour longer. I told him about the astonishing answers the JCS had given to the questions I had drafted in the name of the president, in particular about the effects they anticipated on our own European allies from their planned attacks on the Sino-Soviet bloc. I’d had no prior intention to bring up my own strongly heretical view on first use, but midway through our talk, he raised the issue himself.

  There was no such thing as limited nuclear war in Europe, he said. “It would be total war, total annihilation, for the Europeans!” He said this with great passion, belying his reputation as a cold, computer-like efficiency expert. Moreover, he thought it was absurd to suppose that a supposedly “limited use” would remain limited to Europe, that it would not quickly trigger general nuclear war between the United States and Soviet Union, to disastrous effect.