Daniel Ellsberg gained notoriety in the early 1970s for leaking the Pentagon Papers, the Defense Department’s top-secret history of the Vietnam War, and then for outspokenly protesting the war and the government’s secrecy which sustained it. Yet few, then or now, are aware that he spent much of the previous decade immersed in highly classified studies of the U.S. nuclear-war machine: how it works, who can launch an attack, and how much devastation it can wreak if someone ever pushed the button.
His new book, The Doomsday Machine: Confessions of a Nuclear War Planner, is his long-gestating memoir of those times and the years since, and it is one of the best books ever written on the subject—certainly the most honest and revealing account by an insider who plunged deep into the nuclear rabbit hole’s mad logic and came out the other side.
One of its biggest surprises comes in the first chapter: In 1969–70, when Ellsberg was spending his nights Xeroxing the Pentagon Papers, a copy of which was in his safe at the RAND Corporation, he was also—he reveals here for the first time to anyone besides a handful of friends and loved ones—making copies of everything in his safe, including the top-secret-and-beyond documents from his nuclear studies. Intending to release these as a follow-up to his Vietnam leaks, he gave them to his brother for safekeeping, who buried them in his yard in upstate New York. Soon after, a hurricane and flood swept them away into a landfill—in retrospect a blessing, Ellsberg now admits, as he would have been locked up for many years if he’d gone through with his plan. (He nearly received a harsh sentence for the Pentagon Papers leak until President Richard Nixon offered a plush federal job to the trial judge, who reacted by declaring a mistrial.)
But Ellsberg must have taken notes on those nuclear papers before they vanished, because much of the book summarizes their contents, the impact they’ve had on his life, and why it should matter to the rest of us, especially now.
Ellsberg frankly notes that in the late 1950s, when he joined the ranks of elite nuclear strategists at RAND, the Air Force–funded think tank in Santa Monica, he was an entrenched Cold Warrior (and before then, an infantry platoon commander in the Marines), fully convinced that the Soviet Union posed an imminent threat and that the best way to stave off its aggression was to threaten to kill at least 20 million of its citizens in response.
Ellsberg was in fact one of the leading scholars of deterrence theory; his widely hailed essays, “The Theory and Practice of Blackmail” and “The Political Uses of Madness,” which he wrote as a professor at Harvard, landed him his perch at RAND, where he fit right in. “I loved RAND,” he writes. His colleagues were “as smart a bunch of men as I had ever encountered,” all of them—Ellsberg very much included—shouldering the “enlivening burden” of “believing we knew more about the dangers ahead, and what might be done about them, than did the generals in the Pentagon … or Congress or the public, or even the president.” He worked 70-hour weeks and didn’t sign up for RAND’s generous retirement plan, figuring—he was 27 at the time—that he, like millions of others, would be killed in a nuclear war long before the premiums paid out.
But his first foray into the nuts and bolts of nuclear warfare—a study of command-and-control procedures that gave him access to top secret documents and chats with top commanders—set off a gradual unraveling of his worldview. It turned out that the nuclear war plan—and there was just one plan, with no room for flexibility—called for the rapid firing of America’s entire arsenal of nuclear weapons in response to any armed conflict, even a small conventional skirmish, with the Soviet Union. And once the orders came down, the bombs would rain down not just on the USSR but also on Communist China, even if the Chinese weren’t involved in the war. (The intelligence at the time viewed the two countries as all but unified.)
Another shock: The president, contrary to popular belief, was not the only person with his finger on the button. Rather, President Eisenhower had signed an order delegating authority to a small group of four-star generals and admirals, all outside Washington, to launch nuclear weapons in case he was incapacitated. Ellsberg accepted the logic of the idea: If the Russians thought they could shut down the U.S. nuclear war machine by launching a surprise nuclear attack on Washington, thus killing the only man who could order a retaliatory strike on Moscow, they might be tempted to do so. But as Ellsberg learned during his study, the delegations “reverberated downward in a widening circle,” to the point where—in case the top generals and admirals were killed—fairly junior commanders onboard ships in the middle of the ocean had the authority to launch nuclear weapons on their own. Finally, once bombers had been given the “GO” order, it was very difficult to call them back to base. (Missiles, which came into the arsenal later, were, of course, impossible to recall after launch.)
As he fully acknowledges, some of this history has been well-documented in other books, many relying on declassified documents and interviews with former officials over the past few decades. (I should note that Ellsberg writes some nice words about my own 1983 book on the subject, The Wizards of Armageddon. I should also note that, as part of my research for that book, I interviewed Ellsberg at some length, along with 159 other characters in the story.) But it’s rare to get the history laid out in such human detail by someone who was so immersed in the scene.
For example, Ellsberg writes of the afternoon when he and a colleague played hooky from work to go see Dr. Strangelove, Stanley Kubrick’s dark comedy about a lunatic general who launches a nuclear attack against the USSR on his own, using the pre-delegated authority that he’d been given in case of an attack on Washington (though, in this case, there hadn’t been such an attack). Walking out of the theater, Ellsberg turned to his friend, another nuclear denizen, and said, “That was a documentary.” Kubrick had dramatized the command-control system that actually existed—a war that could actually happen.
Ellsberg’s title was inspired by Dr. Strangelove. In the film, the Soviets have secretly buried what they call a “doomsday machine”—several very powerful hydrogen bombs hooked up to well-placed seismometers. If the sensors detect the explosion of a nuclear weapon by another power, they will automatically detonate the H-bombs, which will spread radioactive fallout, killing all life on Earth. The idea of this machine was that the prospect of total destruction would deter anyone from starting a nuclear war. (Kubrick got the idea from one of Ellsberg’s RAND colleagues, Herman Kahn, a voluble physicist who described precisely such a doomsday machine as a parody of “mutually assured destruction,” the deterrence strategy at the time.)
By the time of Strangelove’s release, in early 1964, Ellsberg had been working for three years, off and on, as an assistant to Secretary of Defense Robert McNamara. When he first took that job, at the start of John F. Kennedy’s term in office, Ellsberg sent a memo to the Joint Chiefs of Staff, under McNamara’s signature, asking how many people would die if the United States unleashed its full nuclear strike.* The answer came back in a memo for the president’s eyes only, though a few others, including Ellsberg, saw it: between 275 million and 325 million in the USSR and China. “This piece of paper should not exist,” Ellsberg remembers thinking when he looked at the single-sheet memo. “It depicted evil beyond any human project ever. … From that day on, I have had one overriding life purpose: to prevent the execution of any such plan.”
Ellsberg and several other assistants—McNamara’s “whiz kids,” they were called, most of them also from RAND—did much to change the nuclear war plan. They directed the military to institute “limited options,” so the president could attack Russia without also attacking China, or attack Russian military targets without also attacking Russian cities.
But as Ellsberg later discovered, the actual war planners—in the Pentagon and at Strategic Air Command headquarters in Omaha—never took the whiz kids’ directives seriously. Not until the 1970s and ’80s, with the administrations of Gerald Ford, Jimmy Carter, and George H.W. Bush did SAC introduce flexibility and restraint into their nuclear-attack options.
But by then, scientists had discovered “nuclear winter”—an artifact of nuclear explosions (besides blast, heat, and radioactive fallout) that would spread smoke around the globe, blocking sunlight and snuffing out life. Even a nuclear war that might once have been seen as “limited” would have catastrophic effects.
At the time, Ellsberg knew many of these people who drew up the war plans and would have executed an order to kill many millions, as the plans required. “They were not evil in any ordinary, or extraordinary, sense,” he writes. They, like Ellsberg and his fellow whiz kids who tried to tame the bomb with the lasso of rational analysis, were “in the grip of institutionalized madness.” And this is where the title of Ellsberg’s book takes on a broader meaning. The very existence of large nuclear arsenals, poised to attack one another at a moment’s notice, forms almost inherently a “doomsday machine.”
The logic of that machine creates the “catastrophe waiting to happen.” Because ballistic missiles can zoom from blastoff to target in 30 minutes, the decision to launch must be left in the hands of a president. Because a president might be killed in a sneak attack, authority to launch the weapons must be delegated to subordinates. The most powerful and most accurate nuclear weapons—the land-based intercontinental ballistic missiles—are also the most vulnerable, so it’s tempting in a crisis to launch the ICBMs pre-emptively or to launch them on warning of an attack (“use them or lose them”), even though there has been a history of false warnings. All of these propositions are logical—and all of them contain a whiff of what Ellsberg calls “criminal insanity.”
Ellsberg would like to see the abolition of nuclear weapons, but he’s logical, and practical, enough to realize that this isn’t in the offing, given the state of global politics, national politics (in many nations), and possibly something screwy about the human species. He offers some short-term proposals that would go some distance toward dismantling the doomsday machine.
First, dismantle land-based ICBMs, the weapons that make fingers itchy for pre-emption in a crisis. Second, give up plans to attack leaders (“decapitation strikes,” they’re called), in part because they compel those leaders to delegate launch authority to underlings, in part because without a leader, a nation can’t negotiate a cease-fire—as a result of which both sides are more likely to keep firing their nuclear weapons.
His third proposal will set off controversy even among many who otherwise share his views, and he knows it: Declare that the United States will never use nuclear weapons except in response to some other power’s nuclear first strike. No American president has ever done this, though Barack Obama seriously thought about it. Since the beginning, U.S. nuclear war plans have envisioned America using nuclear weapons first—usually in response to an enemy conventional attack against an ally.
By Ellsberg’s count (from his own observations and his reading of the literature), U.S. presidents have threatened to use nuclear weapons at least 25 times since the end of World War II—which, as he sees it, amounts to using nuclear weapons, “in the precise way that a gun is used when you point it at someone’s head in a direct confrontation.” Moreover, Ellsberg acknowledges that in “several” of those 25 cases, “it is at least plausible that the threats were effective”—and, more to the point, “most of these threats were seen by some high administration officials as effective.” In other words, it is plausible that some wars might have erupted—or might erupt still—without the threat of first-use. On the other hand, Ellsberg argues, a nation can bluff only so many times before the bluff is called—and if this leads to a nuclear conflict, the devastation might be far greater than that of all the non-nuclear conflicts that the threat had deterred.
Had Ellsberg finished this book at any time between the end of the Cold War and 2016, it might have been of interest to the dwindling number of nuclear scholars, but few others. Donald Trump’s bravado threats to spill “fire and fury” on North Korea in response to mere tests of nuclear weapons has revivified fears of nuclear war that haven’t been felt for decades.
Trump and the book also come at a time when portions of the U.S. nuclear arsenal are gravely aging—and the Pentagon is calling for a “modernization” program to replace the bombers and missiles and submarines at a cost of $1.2 trillion over the next few decades. Trump is enthusiastic about the program; few outright opponents have stood up in Congress. Even Obama, who famously called for the elimination of nuclear weapons at some point in the future, expressed support for at least some of the upgrade if just to ensure that the arsenal, at its current size, works.
Which leads to Ellsberg’s fourth proposal: a drastic reduction in the number of nuclear weapons—in part to reduce the devastation if a nuclear war broke out, in part because the thousands of weapons that the U.S. and Russia have now (or even the hundreds that China and Israel possess) amount to overkill.
From the onset of the nuclear age, analysts have asked how many weapons are enough to deter an enemy from launching a nuclear attack. In the 1960s, McNamara and his whiz kids came up with a number—enough to kill 20 million Russians, roughly the number that were killed in World War II. But this was always an arbitrary number: a way to put a cap on the military’s appetites. (McNamara requested, and Kennedy granted, the money to build 1,000 Minuteman ICBMs—way more than he knew were needed, but the Joint Chiefs wanted 10,000, and McNamara couldn’t go lower than one-tenth that number without siring insurrection.)
Many national security officials, after they leave office, snap out of their fugue states and resume a more “human” way of thinking. Ellsberg quotes McGeorge Bundy, Kennedy’s national security adviser, as writing years later that the explosion of 10 nuclear bombs would be a “disaster beyond human history.” Herbert York, the Pentagon’s former science director, similarly estimated that somewhere between 1 and 100 nuclear bombs (and probably a number closer to 1 than 100) would be enough to deter an enemy from launching an attack. One can see the same insight in today’s headlines: so many people are spooked by North Korea’s possession of maybe a dozen nuclear weapons. Whether or not Kim Jong-un plans to launch a first strike on California (unlikely, given that we could strike back with thousands of nukes), those dozen bombs seem to be deterring some Americans from launching a first strike on Pyongyang—or anyway, that’s no doubt part of Kim’s calculation. (The U.S. currently has 4,500 long-range nuclear warheads, down from a mid-’60s peak of 31,000 but still more than enough to incinerate all imaginable targets and the areas around them.)
One of Ellsberg’s colleagues from the RAND days, the late William Weed Kaufmann, who was later a special assistant to secretaries of defense under every president from Kennedy to Carter—and also a professor of mine in grad school—once told me, a few years after retiring, “It was easy to get caught up in the whole nuclear business. You would eat and breathe the stuff. … Then you move away from it for a while, look at it from a distance, and think, ‘God, that’s a crazy world.’ ”
The marvel of Ellsberg’s book is that he captures that world from both of those angles—that of the ground-burrower who can’t see past his confining premises and that of the mile-high flyer who views the landscape in its full moral context—without flinching from the fact that he has occupied both of those personas in his lifetime with equal measures of passion.
*Correction, Dec. 4, 2017: This article originally misstated that President John F. Kennedy was still in office in 1964. Kennedy was assassinated in November 1963. (Return.)
The Doomsday Machine: Confessions of a Nuclear War Planner by Daniel Ellsberg. Bloomsbury.