Daniel Ellsberg
Daniel Ellsberg; drawing by Siegfried Woldhek

Daniel Ellsberg in his youth and Daniel Ellsberg in his age are the same man—a born worrier quick to spot trouble, take alarm, and issue warning. He is best known for worrying about the American war in Vietnam, which time in the war zone convinced him was a crime, and for doing what he could to bring it to an end. In that case he copied and illegally released a huge collection of secret documents about the war, first published in June 1971 by The New York Times, which came to be known as the Pentagon Papers.

But Vietnam was not the first or the biggest thing that worried Ellsberg after he went to work in his late twenties as an analyst for the RAND Corporation in 1959. His first and biggest worry was the American effort to defend itself with nuclear weapons. When Ellsberg finally got a look at the plans for such a war he realized immediately that the Strategic Air Command had built a military instrument that not only could but in his view probably would break the back of human civilization.

It was Vietnam that got in the way of his plan to do something about the nuclear war plans. In his new memoir, The Doomsday Machine: Confessions of a Nuclear War Planner, the second about the big things that obsessed Ellsberg in his youth,1 he does not try to explain why he set aside worry about the bomb to tackle America’s hopeless war in Southeast Asia, then in its sixth year. The probable answer is that he had gone to see it. Arguing about nuclear weapons with other supersmart young analysts and Air Force colonels was dismaying but not horrifying in the way of war itself. In Vietnam hundreds of Americans and thousands of Vietnamese were dying every month, and sometimes every week, with no end in sight. The commitment of American policymakers to go on killing peasants rather than confess failure was the crime that Ellsberg felt impelled to expose and denounce.

But he never stopped worrying about nuclear weapons. He was far from alone, of course. The horror of the bombs dropped on the Japanese cities of Hiroshima and Nagasaki was immediately apparent to all who did not refuse to see. What separated Ellsberg from ordinary civilian worriers was his access to the actual war plans for doing it again. By the time he received his first clearances to know official secrets about types and numbers of weapons, the handful of first-generation bombs, assembled one by one by hand at Los Alamos, New Mexico, had been replaced by more and better devices. Fat Man, the fission bomb that destroyed Nagasaki, was blimplike in shape, weighed about 10,000 pounds, and exploded with the energy of 20,000 tons of TNT. By the late 1950s the first few fission bombs had been replaced by ever-expanding numbers (soon to be thousands) of thermonuclear fusion weapons, small enough to fit in the nose cone of a missile or under a jet fighter, and roughly a thousand times more powerful than Fat Man. RAND did many studies for the Pentagon on the best way to defend America with these superweapons, and the best way to fight a war with them.

Ellsberg’s initiation into the secrets did not happen in a day, and it took him awhile to realize that there were many levels of clearances, each more secret, more tightly held, and shared with fewer people than the last. Beyond Top Secret, the highest clearance known to exist by the general public, were the code-word clearances for what is now called “sensitive compartmented information.” These permitted an individual to know certain specific secrets, like the fact that the United States had developed tools—spy planes and reconnaissance satellites—to photograph the Soviet intercontinental ballistic missiles (ICBMs) that could carry thermonuclear warheads. The number of Soviet missiles was not the one hundred argued by Air Force alarmists in the Pentagon or the fifty claimed in a National Intelligence Estimate (NIE) in June 1961.

In September of that year Ellsberg learned that the United States would not find it hard to destroy the Soviet missile force. Only four ICBMs were ready to go and they were all at the missile-testing site in Plesetsk, about five hundred miles north of Moscow and a hundred miles south of the White Sea. The four missiles were liquid-fueled and took a long time to prepare for launch. They were standing up in the open and were close enough together to get all four with a single nearby hit. To know this you had to have code-word clearances for Talent and Keyhole, the systems of overhead reconnaissance that filmed the vulnerable Soviet missile force.

There is a widespread belief, Ellsberg writes, that “everything leaks; it all comes out in the New York Times.” That, he says, “is emphatically not true.” Even analysts at the heart of the secret world are not cleared for many categories of secret information and are not cleared to know that they are not cleared. While Ellsberg was being initiated into these secrets he did not know that his own father had once enjoyed an early version of a code-word clearance, a “Q” clearance that protected the secret work on fusion weapons in the years after World War II. Ellsberg’s father told him this in 1978, when he also confessed that he had resigned in 1949 from a bomb-related engineering job—“the best job he’d ever had,” Ellsberg writes—because he wanted no part in building anything a thousand times more powerful than the bomb that destroyed Hiroshima.


Ellsberg was astonished. Why had he never known about this? “Oh, I couldn’t tell any of this to my family,” answered the senior Ellsberg. “You weren’t cleared.”

The Doomsday Machine addresses three subjects. The first is the history of Ellsberg’s work at RAND on nuclear war planning just before and during the Kennedy administration, when he discovered what Air Force General Curtis LeMay, commander of the Strategic Air Command, had planned and prepared by 1960 to do to the Sino-Soviet bloc in the event of war. The second is how city-destroying attacks became the air strategy of choice during World War II, with the effect of gradually resigning airmen to the efficiency of nuclear weapons, one of which could do what it had taken three hundred B-29 bombers over Japan to do using conventional bombs. The third is how to end the dependence of so many nations on nuclear weapons before a spark creates a conflagration that incinerates the world.

Ellsberg writes briskly in the service of opinions formed by long and sober study. What he means is never in doubt and it is always interesting. But even so it is the first and longest section of his new book that makes the greatest contribution. Many able historians of planning for nuclear war—Gregg Herken and Fred Kaplan thirty years ago, Ron Rosenbaum and Eric Schlosser more recently2—have set the scene for Ellsberg’s narrative, but he adds numerous revealing stories of important figures at the time while providing what is probably the best first-person account of what nuclear war planning was actually like. It was intellectually exciting, but at the same time it fed a sense of dread. Catastrophe did not require a monster running things, just a moment of head-to-head dispute about something hard to give up, a couple of wrong guesses, or a run of bad luck. Nuclear weapons, as Ellsberg describes them, are like the pistol in the bedside drawer of a man subject to wild mood swings—too close to hand in moments of fear or despair.

Ellsberg loved the intellectual energy of RAND, where men like the “enormously fat” Herman Kahn, author of On Thermonuclear War and Thinking About the Unthinkable, never softened their opinions. When Ellsberg at his very first group discussion in the summer of 1958 hazarded a thought, Kahn shot back, “You’re absolutely wrong.” Ellsberg didn’t mind; he welcomed the “gloves-off” debate. But thinking it all over in the sixty years since, he has concluded that if security were really their goal, they were all wrong just about all of the time. The problem was not so much the terrible power of nuclear weapons, but what the Air Force planned to do with them.

The principal author of the Strategic Air Command’s 1960 plan was LeMay, the cigar-chomping former commander of the American bomber fleet that had aimed to destroy Japan’s will to fight in 1945, one burned city at a time. LeMay’s strategy for preserving the peace was to threaten war so terrible that no Soviet leader would ever risk it. How terrible? An answer was given to Ellsberg by a RAND friend, the physicist Sam Cohen, who had been one of the youngest of the bomb-makers who worked on Fat Man, and Little Boy, which destroyed Hiroshima. Later Cohen invented the neutron bomb, which he believed might help save us from big bombs because it would make little ones safer to use. Ellsberg thought Cohen was dead wrong on that point but they were friends anyway.

Cohen told Ellsberg that in the early 1950s he had been part of a planning group sent by General Bernie Schriever to ask General LeMay about nuclear force requirements, starting with the question of how big was the biggest bomb he wanted. LeMay’s answer: “One bomb, for Russia.” He meant one bomb for all of Russia. There were two problems with LeMay’s approach—it ignored the fact that war often comes unwanted and unexpectedly, and it offered no clear guidance for knowing when the moment for the one bomb had arrived.


The sudden and utter destruction of the Soviet Union was the goal of LeMay’s strategic thinking. The SAC’s actual plan never included one bomb big enough to destroy all of Russia, but it promised the same result with many, many bombs. When Ellsberg started to work at RAND the immensely complicated and seldom-changed American plan for nuclear war was spelled out in Annex C of a document called the Joint Strategic Capabilities Plan (JSCP, pronounced Jay-SCAP). Annex C was very closely held by planners and the Joint Chiefs of Staff, so closely that not even the secretary of defense or civilians in his office were ever shown or informed about the plan or told even the name of the document by order of the Joint Chiefs.

That was still the case when the Kennedy administration arrived in 1961. None of them had ever heard of the JSCP, Annex C, or its recent successor, the Single Integrated Operational Plan (SIOP, pronounced Sy-OP). The reasons for this secrecy had to do with service rivalries, technical complexities in executing the plan, and the personality of LeMay, who had made up his mind that he would know and decide when a nuclear attack on Russia was necessary, and what ought to be on the target list. Freedom from meddling was what LeMay wanted, and the Joint Chiefs had helped him to get and keep it.

Ellsberg began to pierce the veil of secrecy while working on a study of war preparations in the Pacific. The plan he discovered was basically the Strategic Air Command’s plan, which was essentially LeMay’s. Herman Kahn’s term for it was “wargasm.” As drawn up by LeMay’s team the first SIOP called for nuclear strikes on just about every city in Russia and in China. Why China, too, if the war was with Russia? The answer, stripped to plain language, had nothing to do with politics: one plan was all the planners could handle at a time.

The first SIOP in December 1960 planned an overwhelming knockout blow. Moscow alone was targeted with at least eighty nuclear weapons, and every Russian city with a population greater than 25,000 would be hit by at least one. China would get the same, for no particular reason. Ellsberg was surprised to discover that the planners had not been afraid to add up the probable number of dead. Over the first six months following the initial strike they estimated that about half the population of Russia and China would die of radiation effects alone—a total of about 380 million people. Three things about this plan convinced Ellsberg to do what he could to stop it: its magnitude, its all-or-nothing character, and the fact that General LeMay had reserved to himself the power to decide when to order the attack.

DOD/Life Picture Collection/Getty Images

Daniel Ellsberg when he was a lieutenant in the Marines, 1954

Some readers may draw up at this point and wonder whether these horrors were really true. The answer is that they were, as the reader may learn from the stout books by Kaplan, Herken, Rosenbaum, and Schlosser. Ellsberg was not the only analyst absorbed in this struggle, but he was in the thick of it. The heart of his plan, he writes, was “moving a few pieces of paper from one level of authority to a higher one”—that is, from the level of the Joint Chiefs of Staff, where everybody agreed that the LeMay approach was fine, up to the level of the president and the secretary of defense, who had been kept out of the loop for fifteen years. Ellsberg believed that Kennedy and Secretary of Defense Robert McNamara would be horrified by the details and would insist on changing the plan, and he was right.

The story of that effort is the meat of the first half of his book. What was at stake was succinctly captured without exaggeration in the title of another book published twenty years later, when things had reached a still more desperate state, by the writer Jonathan Schell—The Fate of the Earth.

“I got China off the automatic target list” was the very first thing Daniel Ellsberg told me in March 1986 when we met at a conference of antinuke groups in San Francisco. That was not the absolute worst moment of East–West nuclear tension, but it was close. In 1986 the Russians and the Americans had thousands of weapons targeted on each other, firing procedures on both sides were on hair-trigger status, and neither side really knew how to stop an “exchange” once it began. On more than one occasion in those years (and since), war was brought suddenly close by improbable things like radar bouncing off a flock of geese or the rising moon, which both mimicked a missile attack.

When I met Ellsberg he had already been thinking about these dangers for twenty-five years. From the conference our conversation moved to his house in the Berkeley Hills, where he talked nonstop about nuclear war planning until the sky turned light. Two days later he did the same again. When I started to read The Doomsday Machine I was struck immediately by the sense that I had heard his stories before and hunted up my old notes. Page after page might have come almost verbatim from his new book.

Ellsberg’s compulsion to share what he knew was rare in my experience. I had encountered something like it only once before, in 1984 with Ellsberg’s friend Sam Cohen. Both men had quit worrying about the rules of secrecy but for different reasons. Cohen was compelled to break his silence by the stupidity of American war plans, which offended his intelligence. He wanted to stop people from talking nonsense. Ellsberg was driven instead by moral horror; what we planned to do struck him as just inexcusably wrong. In 1986 he was on fire to warn the world. I was writing often about nuclear weapons at that time and Ellsberg devoted eight or ten hours to making sure I knew the worst. There was nothing scattershot or erratic about his message; it poured out in crisp paragraphs with dates and full names and a clear narrative structure. Ellsberg told me that he was trying to put it all into a book but something was blocking him; with a ready listener like me he could talk forever, but as soon as he sat down in front of a typewriter, the words froze.

Eventually (it took thirty years) he cleared the jam by dividing his story and writing two books, Secrets: A Memoir of Vietnam and the Pentagon Papers and now at last, in his eighty-seventh year, The Doomsday Machine. Everything that alarmed and horrified him back in 1986 is in the new book, but something important has been added. During the intervening years while Ellsberg struggled, he read widely, talked to people, went back over everything in his own mind, and took his story a big step further. What changed was his slow realization of how close we came in October 1962, and why he had failed to see it at the time.

The first day of the Cuban Missile Crisis caught Ellsberg by surprise, just as it did almost everybody else. What alarmed him most was President Kennedy’s threat to respond to any launch of a missile from Cuba on any country in North or South America with “a full retaliatory response on the Soviet Union.” Ellsberg knew what was in the SIOP. “I wondered if the speechwriter had any idea what he was saying,” he writes. Scores of millions would die in a day, hundreds of millions within six months or a year. Ellsberg called up his friend Harry Rowen in the Pentagon, flew to Washington the next day from California, and joined the analysts and officials trying to think their way through the challenge raised by Nikita Khrushchev’s secret move to base thirty-eight Soviet missiles in Cuba. What follows is Ellsberg’s rich personal account of the crisis, including many new details, to join the others already published.

The big new thing in Ellsberg’s book, the important contribution he makes to our thinking on the danger that never goes away, began with a conversation with Rowen about odds: What had been the real chance that we would go to war in 1962? In the first few days of the crisis Ellsberg had convinced himself that the chance was really quite small. Khrushchev, in Ellsberg’s view, was in a box—if push came to shove in the Caribbean he couldn’t win, and if he chose to fight anyway Russia would be reduced to a vestigial state.

Ellsberg had been arguing about this with Rowen and Herman Kahn and many others for two years, and the logic was clear—you can’t use nuclear weapons if your victim can come back at you, which the United States was prepared to do to the Soviet Union in overwhelming fashion. Khrushchev was facing something like a desperation move in chess; he could push that last piece out there but the American response would be check and mate. So Khrushchev had to back down, in Ellsberg’s view. Rowen thought the same thing, and so did the Joint Chiefs and Paul Nitze, one of the principals on the Executive Committee making the decisions. “At thirty-one,” Ellsberg writes, “I was overconfident that a leader who was outgunned would back down under threat.”

In fact, that’s the way it worked out. Khrushchev backed down. When things were still tense Rowen had remarked that he thought the Executive Committee, which included the president and his top advisers, had been putting the chance of war too high—maybe even ten times too high—not one in a thousand (Rowen’s estimate) but one in a hundred. Then a day after the crisis ended Rowen told Ellsberg he had been way off. Nitze had confided to Rowen that he had been guessing the chance of war at “one in ten,” and he was the optimist on the committee—other members thought the chance was even higher than that.

Ellsberg’s first reaction was “puzzlement.” Nitze knew the facts and he understood the logic of nuclear confrontation. War couldn’t possibly make sense in Khrushchev’s position. But then Ellsberg’s eyes opened to the thing that has obsessed him ever since: the Executive Committee had chosen a course of action that they believed risked a one in ten chance of a nuclear war that would kill hundreds of millions of people.

This point is the crux of The Doomsday Machine, what Ellsberg contributes to our understanding of the danger we continue to face: the knowledge that decent men of courage and intelligence with a personal horror of war were prepared to run a one-in-ten chance of killing hundreds of millions of people—to avoid what? “I’ll be quite frank,” the secretary of defense told President Kennedy at a meeting of the Executive Committee early in the crisis. “I don’t think there is a military problem…. This is a domestic political problem.”

What McNamara meant was that Soviet missiles in Cuba might look bad but did not really change the military balance—might look so bad, he did not have to say, that Kennedy might even lose the next election. The meaning of that fact has grown in Ellsberg’s mind over the decades since 1962. What hope in the long term could there be if presidents or other national leaders were willing to run a one-in-ten chance of killing hundreds of millions of people just to help win the next election?

That defied all the logic that Ellsberg and the rest of the RAND analysts had been counting on to protect the fate of the earth. Over the last fifty years new information has emphasized the real gravity of the confrontation between the US and the USSR. The problem wasn’t in the logic, but in the mammoth military organizations that began to stir at the outset of the crisis. Unplanned events began to happen that could have triggered catastrophe, like the shooting down of an American reconnaissance plane over Cuba. When that happened a furious Robert Kennedy called in the Soviet ambassador, Anatoly Dobrynin, and threatened to attack Cuba right away if they tried it again. Bobby thought Khrushchev had ordered the shootdown to crank up the pressure, but he was wrong; it was Castro and the Cubans who did it on their own.

Another thing Bobby and the Executive Committee didn’t know was how the Soviets in Cuba would respond to the attack the US planned if the missiles weren’t moved. They didn’t know that the Soviets had a military force of 42,000 men in Cuba, not the 7,000 estimated by American intelligence. Nor did they know that the Soviet force had been supplied with more than a hundred tactical nuclear weapons. And they did not know that local Soviet commanders in Cuba had authority to use those weapons to halt an invasion. Some commander of a Cuban antiaircraft unit might have shot at another American reconnaissance plane; the US might have gone ahead with its planned invasion; and a Soviet commander might have used a tactical nuclear weapon to attack American ships drawing close. “And where would it have ended?” asked McNamara when he learned for the first time in 1992 that Russian soldiers had been cleared to use nuclear weapons. “In utter disaster.”

But it was not only Soviet ground troops that had nuclear weapons. Soviet submarines in the Caribbean were also armed with “a special weapon”—nuclear torpedoes. On two occasions Soviet submarine commanders believed they were under attack by American surface ships that were trying to force them to surface. The Americans were “signaling” the submarines by bombarding them with “practice” depth charges, not real depth charges, but the Soviet sub commanders did not know these were practice explosives. They believed the Americans were attacking in earnest, their subs were running out of air, and on the second occasion the commander felt he had only two choices—surface and surrender or use the special weapon, and he did not dismiss the second possibility out of hand. The consequences of that are almost unimaginable, but not quite.

Ellsberg has a great deal else to say about nuclear weapons in The Doomsday Machine, but he avoids the easy and the obvious, which means there are few mentions of the overheated posturing of Kim Jong-un and Donald J. Trump, who before his inauguration tweeted that “the United States must greatly strengthen and expand its nuclear capability until such time as the world comes to its senses regarding nukes.” During a meeting last summer of national security officials, he reportedly called for a tenfold increase in the number of American nuclear weapons. It is not fear of the wrong finger on the red button that makes Ellsberg tremble, but the weapons themselves.

Rousing the comfortable is never easy, but Ellsberg is a vigorous writer with a gift for dramatic tension and the unfolding of events as they cascade toward disaster. His story wakes old concerns. We worried about nukes for decades, but then the Soviet Union collapsed and the cold war ended. Now we have grown used to thinking that the danger is fainter, not the planet-busting wargasm once planned by General LeMay when both nations were armed and wired to shoot off everything in a day. Ellsberg says the danger persists as long as the weapons are there. The great achievement of his new book is to make clear what was hidden fifty years ago—that Khrushchev’s decision to move missiles into Cuba, and Kennedy’s decision to stop him, threatened a war that neither man wanted. The crystal logic prized by analysts, the faith in reason that allowed us to think the unthinkable, evaporated under the pressure of events. We came this close.

This is not a young man’s argument, assured and confident. It is an old man’s warning, the fruit of long reflection and tinged with sorrow, as clear as he can make it: these weapons are too dangerous to have because they are too dangerous to use.