Wohlstetter was inspired toward this solution by his old friend Paul Weidlinger, the civil engineer who had worked on quickly assembled hangars for the aircraft company owned by Wohlstetter’s brother, and who had hired Albert to work with him at the National Housing Agency after the war. Wohlstetter had talked with Weidlinger about the vulnerability problem, and thought that as a structural engineer, he might have something to offer. Weidlinger discovered, among other things, that the Air Force assumed that objects could be hardened to resist a maximum of twenty psi. Examining the assumptions behind these Air Force calculations, Weidlinger guessed that it was possible to do better. The Air Force, it turned out, had been relying on calculations by very conservative engineers who, noting that normal buildings collapsed at five psi, thought that it was really going out on a limb to speculate that one might build something that could resist twenty.
Twenty psi might indeed have been an upper limit for many objects in civilization: nobody wanted airplanes or bridges that sagged or buildings that were half-collapsed. But with something like a shelter for protecting a missile or a bomber, all that was necessary was for the thing inside to survive and to be used once; aesthetic virtues were utterly unimportant. With this in mind, the hardness of a structure could be extended well beyond twenty psi.
Paul Weidlinger spent that summer at RAND as a consultant, and with some help from Herman Kahn—who, while in the RAND physics division, was becoming increasingly interested in strategic issues and was gradually allying himself with Wohlstetter—concluded that a 200-psi shelter was quite feasible. Indeed Weidlinger designed on paper just such a shelter, suitable for the B-52. Wohlstetter was now satisfied that his and Hoffman’s theoretical observations about hardening shelters were technically feasible.
For nearly nine months, the Wohlstetter-Hoffman team slaved away on calculations, conceiving the most ingenious ways that the Soviet combined bomber and missile threat—based on Air Force Intelligence projections—could do SAC in with as little warning as possible. Detailed map exercises were conducted, with Soviet bombers refueling in midair and going way around U.S. radar-warning nets to maximize surprise. Two-wave attacks were assumed—the first wave against the small number of SAC bases inside the United States (twenty-nine when Wohlstetter and Hoffman began their study, with fifty-five planned by 1962), the second wave nabbing those bombers that managed to survive the initial attack and fly on to dispersed bases. Air Force Intelligence was assuming a ferocious Soviet strategic buildup of 500 ICBMs and 500 Bear and Bison intercontinental bombers by 1960. Wohlstetter and Hoffman thought that they were making a very dramatic point when they illustrated how SAC could be devastated even assuming a very “low” estimate of Soviet forces—a strike of 250 or even 150 heavy bombers and the same number of ICBMs.
As for the American ICBM force, the plans for deploying it involved placing the missiles so close to one another and leaving them so unshielded that a mere 24 Soviet nuclear missiles would destroy 96 of the 120 that might be deployed by the early 1960s.
The Wohlstetter-Hoffman team came up with fifty ways to help solve the problem—among them hardening bombers inside shelters (the Weidlinger solution, which Wohlstetter considered most important), dispersing and hardening critical facilities on each base, adding more recovery bases inside the United States, building up active defenses around SAC bases, extending the perimeter of radar-warning nets, constructing long-range bomb alarms that would alert all SAC bases when one of them is hit by a nuclear bomb, allowing ambiguous warning signs of an attack to provoke preliminary bomber-evacuation measures.
Just as in the overseas-base study, the existing SAC system was compared with the proposed system through Edwin Paxson’s systems-analysis technique—asking which system can destroy a given number of targets more cheaply and which system can destroy the most targets for a given amount of money. And again, the RAND system looked better. To destroy 85 percent of the 270 Soviet urban targets, “in the face of moderately high Soviet offense and defense capabilities” (defined as 500 Soviet ICBMs, 500 Soviet bombers, 295 air-defense fighter regiments, 225 local-defense-missile sites), SAC’s system would cost nearly $35 billion, whereas RAND’s would cost about $10 billion. RAND’s proposal would cost more money than SAC was spending, but in terms of cost per target destroyed, it would be much cheaper. Likewise, in the face of an attack composed of 300 Soviet bombers and 250 ICBMs, with a second wave of 500 additional bombers, SAC’s planned system could kill about 15 percent of the population in the 270 largest Soviet cities, given a $3.4 billion budget, whereas the RAND system could kill about 90 percent.
The exercise was complex but fundamentally mechanicaclass="underline" conceive the cleverest Soviet attack; punch in the Air Force Intelligence estimates on Soviet strategic forces and air defenses; from the yield and accuracy of Soviet offensive weapons and the psi-resistance of U.S. SAC targets, compute the weapons’ “kill probability,” the chance that they destroy the bombers or crater the runways or gut the support facilities. Then make the hypothetical improvements in SAC and do the calculations again. Keep doing the calculations over and over, adjusting them to slight improvements in the force, until nearly the entire SAC force is, on paper, able to survive even the most ingenious of Soviet surprise attacks and deliver the most powerful of retaliatory blows.
The results were presented in a RAND study, Staff Report R-290, entitled Protecting U.S. Power to Strike Back in the 1950s and 1960s. There were a few at RAND, most notably in the social science division, who had some doubts about R-290, who thought it lacked a political dimension, who wondered whether the Kremlin leaders really were prone to such risky maneuverings, who were skeptical that war would come with a bolt-from-the-blue surprise attack, as R-290 assumed. But the social science division was removed from the rest of RAND—literally, 2500 miles away, in Washington, D.C. Most were figuratively removed, too: quantitative analysis had triumphed at RAND, through the spread of systems analysis and game theory and—until the Wohlstetter studies, which put the economics division on top of the strategic business—through the domination over the rest of RAND by the mathematics division. These sorts of studies were scientific, so it was thought; there were numbers, calculations, rigorously checked, sometimes figured on a computer. Maybe the numbers were questionable, but they were tangible, unlike the theorizing, the Kremlinology, the academic historical research and interpretation produced by social science. Wohlstetter snootily denigrated all such works as being in “the essay tradition.”
And within this newly triumphant time for the economics division, Wohlstetter established himself as the self-styled intellectual leader, the man who had calculated the essence and requirements of deterrence in the nuclear age. Few of Wohlstetter’s thoughts were wholly original. The idea of a surviving second-strike force as the essence of deterrence came from Bernard Brodie; the systems analysis he took from Ed Paxson, E. S. Quade and others; the historical parallel of Pearl Harbor was inspired by his wife, Roberta Wohlstetter; much of the actual quantitative work in his two big studies was handled by his colleagues, Fred Hoffman, Harry Rowen, Bob Lutz and dozens of others. But Wohlstetter expanded on that which he borrowed. Brodie, assuming that nuclear weapons would be used against cities, never thought that protecting the second-strike force would take very much imagination. Paxson, creative as he was, failed to include some important factors, such as the location of air bases, in his conception of a “system.” Wohlstetter, coming to the field of national security with no background in it, was able to view things from a novel perspective and to synthesize its elements into a new package. Through Wohlstetter’s own personal influence within RAND, vulnerability began to loom as the preoccupying issue, the virtual obsession, of strategic analysis.