As Wohlstetter went about his business, discovering the vulnerabilities of SAC and peering into the most arcane detail, he had Lutz and Harry Rowen—and later, toward the end of the study, another young RAND economist named Frederic S. Hoffman—do the more classical analysis of bombing campaigns: designing bomber-penetration tactics, gathering data on fighter-bomber duels, figuring out precise flight paths from various bases to various targets, timing the attack so that as many SAC planes as possible—coming from several different directions—would show up on Russian radar screens all at once, calculating the attrition rates resulting from estimates of Soviet bombers and air-defense weapons. Rowen and Hoffman, as members of the RAND cost-analysis division, also put price tags on the various systems.
Wohlstetter, meanwhile, was discovering that SAC was sitting in a house of cards. He was looking at the period from 1956 to 1961, four to nine years into the future. By then, SAC would have about 1,600 B-47 bombers with a combat radius of 1,700 miles, 300 vintage B-36 bombers with a 2,950-mile radius, perhaps a wing of 3,060-mile B-52s, and 720 KC-97 tankers. SAC operations involved flying the bombers from thirty bases in the United States to eighty-two bases overseas. (In 1952, when Wohlstetter began looking at the data, SAC had only thirty-two of these overseas bases.) SAC crews would set up replicas of the bases back home, then load the planes with bombs and send them off to the targets from the overseas bases. This operation would take about a week to get off the ground.
The practice came from World War II, when bombing was merely part of an overall campaign and the bomber bases were rarely in danger of being destroyed. But this was the atomic age; the Soviets had exploded an A-bomb in 1949. Yet SAC had not adjusted to the changed situation; the commanders were still, almost unconsciously, planning for a World War II-type operation.
Moreover, most of the bases were very close to the Soviet Union—one-third of them within the combat radius of Soviet lightweight bombers, all of them within reach of their medium bombers. Many were so close that Soviet bombers could attack them with no U.S. radar warning. Also, for economizing measures, the airplanes were parked close to spare parts and repair facilities; most fuel-storage tanks were built above ground; no attempt had been made to disperse the storage of critical items. A few bombs could knock out not only the airplanes but also everything that could fix the planes that were merely bruised and all the supplies and fuel that might allow other bombers from other bases to fly over and to carry out the bombing mission as planned.
In fact, Wohlstetter and his team estimated that a mere 120 bombs—forty kilotons each, with average miss distances of 4,000 feet—could destroy 75 to 85 percent of the B-47 bombers while they casually sat on overseas bases. The Strategic Air Command, seemingly the most powerful strike force in the world, was appearing to be so vulnerable in so many ways that merely putting it into action—moving it overseas in accordance with the official Mobility Plan—created a target so concentrated that it invited preemptive attack from the Soviet Union.
The conclusion, by this point, was clear. But Wohlstetter and company thought that they might not be readily believed. They spent all of 1952 exploring every conceivable way that they might be proved wrong and trying to come up with their own solution to the problem. They examined everything from the effectiveness of air-defense fighters deployed to protect SAC bases from attack to the rate at which hydraulic faucets on SAC bases could pump water. Their conclusions were left essentially unaltered.
They recommended that early-warning radar systems should be improved, so that the bombers could evacuate their bases more quickly; adequate warning alone reduced the percentage of bombers a Soviet attack could wipe out from 75 or 85 to only 20 percent. More drastically, they recommended that overseas bases be used only for refueling, not as a base of complete operations in their own right. This would minimize both the amount of material overseas and the time that each airplane would have to spend on the ground. On each base, furthermore, supplies should be dispersed and, where possible, hardened to resist blast. Air-defense squadrons should be strengthened. Repair facilities should be better protected.
The conclusion that these refueling bases would be the best way to go emerged from a standard systems analysis, asking the Paxson-inspired questions, “Which system more cheaply destroys a given number of targets?” and “For a given budget, which system destroys the most targets?” However, the Wohlstetter analysis carried an unusual twist: instead of comparing different bombers, it compared different basing schemes—an air-refueled system, the Air Force-programmed operating base system, and (the one that Wohlstetter eventually favored) a ground-refueled system. They also compared these three systems under three different assumptions concerning the size of the Soviet bomber force as of 1956—higher than that expected by Air Force Intelligence, the expected number, and some number lower than that expected.
The results of the analysis determined the choice of system. In all cases, the ground-refueled system—using overseas bases only to refuel bombers—was by far the cheapest. Operating from these sorts of bases, B-47s could destroy the same number of targets for nearly one-third the cost of the Air Force system. And for a hypothetical $40 billion budget, they could destroy about twice as many targets—300 to about 1,200 (depending on the size of the Soviet bomber force), compared with fewer than 100 to about 600. On this score, the purely intercontinental operation—refueling the bomber in the air all the way, with no overseas bases at all—would be much too expensive. Since so much of such a plan’s budget would have to pay for more tankers, SAC would have only enough bombers to destroy 350 or so Soviet targets under the best of circumstances and fewer than 50 under the worst.
The results surprised nearly everyone. After Wohlstetter circulated around RAND a draft of the team’s tentative findings, he was approached by Herman Kahn, a nuclear physicist at RAND who was becoming increasingly interested in nuclear strategy. Kahn thought that Wohlstetter had overdone it, that this “programmed system,” against which Wohlstetter’s preferred system looked so good by comparison, was too obviously rigged and exaggerated.
“Albert, this programmed system that you’ve demolished,” he said, “is a straw man.”
“Herman,” Wohlstetter calmly replied, “that is the system that SAC has programmed.” Kahn was aghast.
By the fall of 1952, if there were any doubts left in the Wohlstetter team’s minds about the vulnerability of SAC, they were dispelled at once by a strange twist of fate. On September 1, a tornado unexpectedly swooped down upon Carswell SAC Air Force Base in Texas. Wind gusts approached 125 miles per hour. The storm demolished one B-36 bomber and heavily damaged eighty-one more, putting both wings on the base out of commission for a month. Further damage was done by flying debris from nearby repair docks. One fighter airplane was blown into the air and carried several hundred feet. One hangar collapsed. Gasoline spilled from many of the wrecked airplanes, some of them frighteningly close to broken electrical circuits. If some Carswell crew members had not quickly shut off all electrical current, blazing fires would have caused still greater damage. As it was, no electrical equipment was allowed on the runways until three days after the storm.