The term “custody dispute” is one we usually associate with acrimonious divorce proceedings. But there was a very real nuclear custody dispute in the 1950s, one which I’ve often been surprised that even folks fairly well-versed in nuclear history weren’t aware of.1
The issue in a nutshell: When Eisenhower took office in 1953, there were around 1,000 nuclear bombs in the US nuclear arsenal and all but a few dozen were in the hands of the civilian Atomic Energy Commission (AEC). When he left office in 1961, the nuclear stockpile contains some 18,000 weapons (of widely varying size, yield, and delivery mechanism), and 90% of them were under military control.2 That shifting of control — of whether it was a civilian or a military organization that physically had responsibility over the bombs — was the essence of the custody dispute.
The backstory to the dispute is rather simple. As part of the Congressional battle over postwar domestic legislation to regulate atomic energy, the Atomic Energy Act of 1946 (the McMahon Act) assigned all control of the manufacture, development, and possession of nuclear weapons to the civilian Atomic Energy Commission (AEC). This was not accidental — the key push by Senator Brien McMahon was that this was “civilian control” of the atom, as opposed to “military control.” (The bill the McMahon Act was meant to replace, the May-Johnson Act, was castigated as a “military” bill, which was not entirely true.) The fear, in 1946, was that a military-controlled nuclear complex would result in an un-reflective arms race, preempt democracy, and make an unpleasant atmosphere for scientists. (Thank goodness civilian control resulted in none of those ills!)
The Atomic Energy Act allowed for the possibility that the President could, “from time to time,” direct the AEC to give the military access to fissile material and weapons “for such use as he deems necessary in the interest of national defense.” Otherwise, all weapons would be kept by the AEC.
Thus the custody dispute: the military was, over the course of the late 1940s and 1950s, meant to be ramping up its integration of nuclear weapons into its war plans. And yet, they did not physically have access to live nuclear weapons. In fact, they didn’t even have access to weapons casings to practice with, on account of the weapons casings being classified as “restricted data,” and required Q-clearances (with full FBI investigations) to even look at!
For the first year of the AEC’s existence (two years after the first bombs were produced by the Manhattan Project), this was actually somewhat of a null issue because there weren’t any atomic bombs to use anyway. In 1947 there was fissile material for 13 bombs ready to go, and some number of non-nuclear assemblies to go with them, but they weren’t in anything like a usable state — the United States couldn’t have immediately dropped an atomic bomb even if it needed to. It would have taken weeks, at least, to launch an atomic attack against the USSR.
The “civilian” side of this started with the initiative because of the Atomic Energy Act, and because the early AEC was extremely committed to maintaining custody over the weapons. Why? The stated reasons are primarily technical: the early stockpile was comprised of weapons not too different from those developed during the Manhattan Project, and they were complicated, dangerous, and still considered incredibly secret. Norris Bradbury basically thought that they were “too complicated” for the military to maintain effectively, and the facilities on military bases were not yet up to snuff for something as important as an atomic bomb.
But that can’t be the full picture. Obvious in Lilienthal’s journals from the period (which are a great read in any case, and can be picked up for a song one various used-book websites) are two major themes: 1. he thought the military was more or less crazy-eager to use atomic bombs; 2. he felt the AEC needed to have a more-or-less unified front with respect to maintaining atomic autonomy for the “civilian” sphere. The former is an argument from fear, the latter is a mix of both ideology and bureaucratic politics.
What strikes me as most interesting about the custody dispute, and how it worked out, is that there wasn’t one decision that made the change happen. There wasn’t some grand realignment of strategic thinking or an explicit shift in policy. There were a lot of little, tiny decisions, each one adding a little bit towards greater and greater military control, and less and less civilian control:
- In 1948, during the Berlin blockade, Truman sends B-29s to England and West Germany, with the idea that he could ship the bombs later. In June 1950, two weeks before the Korean conflict formally began, Truman authorized the permanent transfer of 90 non-nuclear assemblies to the military for training purposes. Later that year, Truman and the AEC agreed to move the non-nuclear components of a number of bombs to England, Guam, and a few aircraft carriers; the fissile pits would stay in the US, to be transported by a fast plane if necessary.
- In April 1951, Truman authorized — without consulting the AEC first — the transfer of a small number of complete nuclear weapons to the military, to be stationed in Guam. By September 1952, Truman had agreed to a proposal that stated that the President still had total authority over when and how atomic weapons could be used, that the DOD would have “custodial responsibility” for all bombs kept outside of the continent US and any that were needed inside of the US, and that the AEC would retain custody for any other bombs.
- By the summer of 1953, Eisenhower had agreed to give the DOD the pits to correlate with their existing non-nuclear assemblies, with the AEC keeping the rest of the stockpile. By 1954, there were continued deployments of both nuclear and non-nuclear bomb components to a number of countries and carriers.
- By 1955, Eisenhower had agreed to regular “dispersals” of nuclear weapons overseas, but kept high-yield (over 600 kilotons) thermonuclear weapons within the control of the AEC, though some overseas. By 1956, it had been decided that even this was too cumbersome, and a number of high-yield weapons were transferred to the DOD. Eisenhower agreed that there should be automatic transfer of all weapons from the AEC to the DOD in the event of a defense emergency, even one declared unilaterally by military commanders.
- In 1957, high-yield thermonuclears were moved overseas (to the UK) for the first time. Eisenhower also issued directives which allowed for the “pre-delegation” of authority to use nuclear weapons under certain situations, allowing field commanders to make the nuclear call without Eisenhower’s explicit approval.3 More bombs were “dispersed” overseas. By 1959, Eisenhower had granted the DOD full control over all of the “dispersed” weapons, including all high-yield weapons. By 1961, the Joint Chiefs of Staff was in a position to more or less automate the transfer of arbitrary numbers of weapons into DOD custody at will.
What’s behind these slow changes? I’m inclined to synthesize the changes over three intertwined themes:
- Changes in the strategic situation: the Cold War heats up, the Soviets get the bomb, and the idea that you might need to rapidly use the bomb becomes more important to strategic thinking at the Executive level.
- Changes in bureaucratic momentum: the early AEC (esp. under Lilienthal) was committed to carving out a strong position for itself in the postwar regulatory landscape. Lilienthal in particular was a firm believer that the patterns set in the first years of a new agency would determine its later abilities, and fought hard to keep the bomb firmly within the AEC’s control. Later commissioners fought less hard, either because they thought that battle had become less relevant, or because they trusted the military more.
- Changes in weapons technology: the bombs not only become more “GI-proof,” they become more intimately linked to their delivery mechanisms, making separate custody less plausible. In the earlier years, weapons designs with separable fissile capsules were developed, something which enabled custody; in later years, designs with sealed pits and other technical innovations made this more difficult, especially with tactical weapons. I don’t want to overemphasize this: had civilian custody been the primary concern, the weapons designs could have reflected that. The fact that the designs changed in directions that made a custody divide impossible is in part a reflection of the changes that were already taking place.
These changes are, of course, intertwined. The changed possibility of technology changed considerations about strategy and changed the ideas about the proper role of the AEC. The changed considerations about strategy changed ideas about what types of technology should be pursued and so on. Lack of conviction by the AEC on the custody issue, and lack of clout, led towards the AEC relationship being increasingly defined as a “producer” of the bombs and little else. And so on.
Today the military more or less owns all of the nuclear weapons in the stockpile. The chain of command for usage of the weapons is meant to be purely civilian — the President would have to authorize any use of them explicitly — but the ownership is military.
Does the issue matter, today? There’s a tendency, I think, to see the custody issue, or the civilian-military disputes of the 1940s-50s, as somewhat quaint. The civilian AEC didn’t exactly do the best job. They didn’t exactly keep the stockpile to a reasonable amount. They didn’t exactly guarantee that weapons wouldn’t be used in war, or that the military couldn’t have run amok with them if they had wanted to.
That being said, I do think the AEC had a very different managerial atmosphere, even under the later Commissioners. I do think they fostered a very different attitude towards the creation and use of the weapons than the military commanders had. I think in the early years, in particular, Truman and Lilienthal were right to be dubious about the military’s enthusiasm for the atomic bomb, and to treat the bomb as something “special” — even before there was threat of Soviet retaliation. I suspect that if they hadn’t, the results of a more cavalier attitude towards atomic strikes, especially with regards to Soviet actions in Europe and Chinese actions in Asia, could have been disastrous.
- My source for this is a rather classic document for information about US nuclear deployments: History of the Custody and Deployment of Nuclear Weapons, July 1945 through September 1977 (8MB PDF here), prepared by the Office of the Assistant to the Secretary of Defense (Atomic Energy), in February 1978. What I mean by “classic,” here, is that lots of other folks seem to have known about this document for a long time, but I was only made aware of it when I asked a former Harvard colleague of mine, Dan Volmar, who is doing a dissertation on nuclear command and control systems, for something about the custody issue. [↩]
- Peter J. Roman, “Ike’s Hair-Trigger: Nuclear Predelegation, 1953-60,” Security Studies 7, no. 4 (Summer 1998), 121-164, on 121. [↩]
- On this, see the previously cited article by Peter Roman. See also the National Security Archive’s Electronic Briefing Book on the issue. [↩]