Significantly, the only portion of the R-290/Gaither program that the Air Force consistently and successfully resisted was the notion of putting the bombers inside underground hardened shelters. Officers argued that it would be too expensive, maybe $10 billion or more, and that it might not protect, ultimately, the bomber against radiation effects. But the real reason had more to do with Air Force interests. With hardening, the dispersal and airborne-alert programs, so advantageous to the Air Force budget, might be cut back. To spend money on offense, not defense, was practically dogma in Air Force circles.
SAC was even more eager to use intelligence estimates as a method of advancing its own interests. The forceful leaders of SAC’s own intelligence agency at the time were Generals James Walsh and George Keegan. Keegan was the more fiery of the two. He received his first training in intelligence as a member of a small advisory group to the Air Force Assistant Chief of Staff for Intelligence in the early 1950s. Keegan’s boss was Professor Stefan Possony, an extremely right-wing Russophobic Sovietologist with a particular penchant for conspiratorial views of history. This advisory group’s mission was to brainstorm on what kinds of horrifying things the Russians might be doing, and then to find the evidence.
Keegan had learned his job well, and was a full believer in the technique. He was a forceful speaker, a master showman, a superb briefer. Around the late 1950s, as even Air Force Intelligence was giving way on high Soviet ICBM estimates, SAC kept a full steam blowing. Keegan and Walsh had briefings, charts, diagrams, photographs proving that the Russians were already fielding ICBMs but that they were hiding them—in barn silos, medieval monasteries, mysterious-looking buildings out in the middle of nowhere.
With so many Soviet missiles that you could never know precisely how many there were or where they were located, arguments for an enormous SAC force could proceed indefinitely. The military requirements worked both ways: the large number of Soviet ICBMs meant a large number of targets to hit, which required a large number of SAC bombers and missiles; likewise, with so many Soviet missiles that might attack SAC, America needed hundreds and hundreds more to allow for heavy attrition. The Air Force proper finally agreed officially with the view that the Soviets were probably engaging in deceptive practices in their ICBM program. But not even Air Force Intelligence was willing to go as far as Keegan.
Still, with SAC or Air Force or even CIA intelligence estimates on the size of the near-future Soviet ICBM arsenal taken as the truth, the fundamental assumptions about the nature and magnitude of the Soviet threat could still be retained as legitimate.
On the other hand, if the Army and Navy numbers were treated seriously, the Soviets would appear to pose essentially no great threat to SAC. They would not have enough missiles to do so until mid-1963; and by that time, the Navy would have several new Polaris submarines, nuclear-powered, each carrying sixteen nuclear-tipped missiles, based underwater and virtually invulnerable to attack. Moreover, thanks in part to such studies as R-290 and the Gaither Report, the Air Force would have started to field its new Minuteman ICBMs in dispersed and hardened shelters.
The Army-Navy numbers, in other words, said there was no great danger to SAC, and no missile gap.
Throughout this period, nobody in the Senate knew the origins of the missile gap, knew that it sprang from the failure of the bomber gap to materialize. Nobody knew of the wide disagreements among the intelligence agencies as to the number of ICBMs the Soviets might have in place by the early 1960s. Nobody knew about the U-2 flights. Symington and most of the other Democrats, many of whom took their cues on this point from him, heard only about the Air Force Intelligence estimates, which (next to those of SAC Intelligence) were most pessimistic of all. Thus, when they heard Allen Dulles or Defense Secretary Neil McElroy or his successor as of December 1959, Thomas Gates, or even President Eisenhower say that there would be no missile gap, the Democrats and other critics of the Administration felt that these officials must be knaves or fools, that they were deluded or misled, that they were endangering the nation merely by their presence in high office.
But Eisenhower did know the background of the bomber gap and the missile gap. When charges of the missile gap began to circulate widely among political opponents in 1958, he assigned his staff secretary, Brigadier General Andrew Goodpaster, to find out whatever happened to the mysteriously vanished bomber gap. Goodpaster went through all the old NIEs, talked with intelligence officers, and learned how the NIEs had assumed that another Bison plant would be built, how they relied on a host of assumptions concerning production rates that turned out to be false. When he reported his findings to Eisenhower, the President felt secure in resisting all the fuss about a new gap. And when, in 1959 and 1960, the CIA started to back off somewhat, when the year of maximum danger started to recede into the distant future, Eisenhower felt his judgment vindicated.
On August 29, 1958, after he had met twice with Allen Dulles, Symington met with Eisenhower and gave him a letter, telling him that the intelligence community was wrong, that he was being misled. Eisenhower told Symington that whoever his sources were in Air Force Intelligence, they could not possibly know everything that those in the upper levels of the agency knew. Eisenhower never told Symington or anyone else in Congress about the U-2 or the Turkish radar site, but that was what he was talking about.
The missile gap also dominated the discussions of the day at the RAND Corporation. But there it was a more sophisticated conceptualization than the simple bean-count comparisons tossed around by Symington and his followers. The strategists of RAND preferred to call it a “deterrence gap.” The issue was not so much that the Russians had more missiles as it was that SAC was so vulnerable that even the low side of the official intelligence estimates indicated that the Soviets would have enough missiles to knock out America’s power to strike back—in Bernard Brodie’s by-now ancient phrase, “to retaliate in kind”—after an aggressive first-strike. That being so, the nation’s and thus the free world’s ability to deter Soviet aggression was on the verge of being shattered.
Still, this more sophisticated view was the product of quantitative analysis, and the numbers came from the National Intelligence Estimates that foresaw an impending missile gap. And like the Stuart Symingtons of the world, most of the RAND strategists knew much less about those estimates than they thought they did. CIA policy on the distribution of the annual NIEs had changed after 1958: henceforth, no contracting firms—and that included RAND—were to receive copies. By coincidence, the 1958 NIE represented the peak year of the missile gap. It was not until 1959 that the estimated numbers of future Soviet missiles began to go down and the Army and the Navy began to add their footnotes. But almost none of the RAND analysts knew anything about this. They received intelligence estimates only from the Air Force Chief of Staff, and did not know that, from 1959 on, the Air Force numbers were considerably higher than those of the rest of the intelligence community. If RAND got any dissenting data at all, it came from SAC Intelligence, whose officers thought that the Air Force estimates were on the low side.
In quantitative studies, there is a technique known as “sensitivity analysis”: the idea is that in a world of uncertainties, an analyst should test the validity of his conclusions by altering the key assumptions; if within a reasonable range of assumptions the conclusions remain roughly the same, then they could reasonably be considered correct. Having read only intelligence estimates estimating 500 Soviet ICBMs by the early 1960s, the RAND strategists thought they were being more than reasonable to do sensitivity analysis assuming that the Soviets attacked the U.S. with only 150 or even 250 ICBMs. They had no way of knowing that some intelligence agencies were predicting only 50.