To Ellsberg, the implications were staggering. The military’s nuclear war plan was a single war plan, calling for blowing up all enemy targets in Russia, China and Eastern Europe as quickly as possible, with the chances of restraining or redirecting the attack remote at best. Ellsberg saw that the JSCP, the terms defining the overall war plan, called for carrying out this nuclear attack whenever the U.S. and the U.S.S.R. directly engaged in battle, however tentatively, over however grand or trivial a cause. Ellsberg, at this time, was an extremely hardline Cold Warrior. After taking his doctorate exams at Harvard in 1953, he had enlisted in the Marine Corps (hardly typical for a budding Ivy League Ph.D. candidate), became a company commander and, in 1956, extended his duty when his battalion was sent to the Mediterranean, savoring the prospect of possible combat in the Suez crisis. Ellsberg truly believed, as did most of his comrades at RAND, that some sort of armed conflict between the superpowers would erupt in the coming decade. Seeing that the military planned on automatically escalating the conflict to all-out nuclear war terrified him.
When Ellsberg returned to RAND, Kaufmann was just beginning work on his counterforce/no-cities study. Ellsberg played no role in it, but he recognized its significance. He disagreed with one crucial aspect—the degree to which Kaufmann was emphasizing the counterforce element of the strategy. Ellsberg agreed that the targets hit should be counterforce in nature, but he knew that several Soviet air bases and command centers lay near cities; trying to destroy all the counterforce targets would therefore wreck the attempt to avoid damaging enemy cities. And it was the no-cities aspect of the strategy that caught Ellsberg’s eye. He saw it as the one hope of avoiding the murderous and suicidal policy of the JSCP.
On April 7, 1961, Ellsberg’s thirtieth birthday, and eleven days before Lyman Lemnitzer turned in his dissent against a “controlled response” to Robert NcNamara, Ellsberg finished a draft of the revisions for a basic national security policy that McNamara had requested as Project No. 1 of his 96 Trombones five weeks earlier. Ellsberg’s draft suggested a strategy for achieving U.S. wartime objectives “while limiting the destructiveness of warfare…. Specifically, the United States does not hold all the people of Russia, China or the [East European] Satellite nations responsible for the acts of their governments. Consequently, it is not an objective of the United States to maximize the number of people killed in the Communist Bloc in the event of war.” In fact, “attacks against high governmental and military command centers, or indiscriminate initial attacks on all major urban-industrial centers would fail to inhibit punitive retaliation by surviving enemy units, but would instead eliminate the possibility that enemy responses could be controlled or terminated to U.S. advantage.” Over and again, Ellsberg stressed the need for a durable reserve force and an enduring command-control apparatus, which had no place in the JSCP’s Atomic Annex or in the subsequent SIOP-62. “In particular,” he wrote, in a passage that most reflected the Kaufmann briefing, “alternative options should inelude counterforce operations carefully avoiding major enemy cities while retaining U.S. ready residual forces to threaten those [urban-industrial] targets.”
In May, one month before SIOP-62 was scheduled to go into effect, McNamara signed the Ellsberg memorandum as his own and sent it to the Joint Chiefs as his basic initial guidance for a revised SIOP-63.
Meanwhile, Ellsberg was crusading all over official Washington, telling McGeorge Bundy at the NSC, Walt Rostow at the State Department, Paul Nitze and Roswell Gilpatric at the Pentagon all about the horrors of the JSCP, urging them to change its definition of “general war” so that an armed conflict between the U.S. and the Russians would not inevitably escalate into nuclear war of any sort. With practically every national security adviser in the Administration—and the President himself—keen on building up non-nuclear forces, this effort succeeded and the following year’s JSCP was changed accordingly.
Over the summer, Alain Enthoven and Frank Trinkl took Ellsberg’s revised SIOP guidance and elaborated it into greater detail, ordering that Soviet targets be divided into five separate categories: (1) strategic forces, meaning air bases, missile sites and submarine pens; (2) air-defense sites away from cities; (3) air-defense sites closer to cities; (4) command-control centers; and (5) the all-out strike against Soviet cities. The revised SIOP-63 was to allow for options successively combining these categories. Only the least destructive, most purely counterforce option would be exercised in the initial U.S. strike; then, if the war escalated beyond that level, that option would be combined with the second, and then with the third, then the fourth—and finally, but only if all other options had been exhausted and if the war could not be kept under control, or if the Soviets did not follow our signals of restraint and “controlled response,” Option Five, the all-out strike that dominated the existing SIOP. There were also to be suboptions, providing for a choice on whether the weapons should be airburst or groundburst, high yield or lower yield, “clean” or “dirty.” Finally, Enthoven and Trinkl, under McNamara’s signature, directed that there should be “country withholds,” allowing the President to hit or avoid hitting any targets at all in China or Eastern Europe or both. The watchwords were those that informed the Kaufmann briefing: control, flexibility, discrimination, options.
Some features of certain weapons systems would have to be changed to make all this practical—not only a new emphasis on hardened and mobile missiles rather than more vulnerable bombers, but also fine tuning command-control mechanisms governing the launching of the missiles, making it possible to fire only a few of them, or even a single missile, at a time, and building “options” into each missile’s computer program so that it could be fired at one of several possible targets. Under the existing Minuteman ICBM program, the President would have to launch missiles in multiples of fifty if he wanted to launch any, and each ICBM was inflexibly programmed to strike one preselected target.
Marvin Stern, the Assistant Director for Strategic Systems in the Pentagon’s R&D office, was ordered to see that the Air Force changed the Minuteman ICBM program accordingly. Stern, coauthor of the Hitch-Enthoven-Stern “General War Offensive Package” of the previous February, had worked on the Atlas ICBM program at Convair in the mid-1950s, served on several expert panels dealing with the problems of early warning and accidental nuclear launches in the late 1950s, and accompanied McNamara and the others to SAC for the SIOP-62 briefing. He firmly believed that the changes should be made, and he knew that they were technically feasible. The Air Force, however, wanted no part of this “controlled response” business and simply refused to spend money on it, arguing that it was infeasible. Stern approached McNamara’s general counsel, Cyrus Vance, and asked what could legally be done to prod the Air Force into action.
“Anything you can get away with,” Vance replied.
Against all regulations and laws, Stern canceled Minuteman funding for one month, pending a commitment by the Air Force to improve command-control and develop a rapid retargeting program for the missile. In justifying the action, Stern spoke before a group of high-ranking military officers that Secretary of the Air Force Gene Zuckert and the new Air Force Chief of Staff, General Curtis LeMay, who had moved up when Tommy White retired, assembled to respond to Stern’s fiscal pressure. Stern explained the changes that McNamara was making in the SIOP in a way that he thought would most logically appeal to his audience.