by David Kaiser
Both authors carefully emphasized caveats and qualifications. DeWitt began both of his books by citing the large professional literature devoted to interpreting Soviet statistics. Both of his books also included detailed appendices on the “perplexities and pitfalls” of working with Soviet statistics. Raw data like enrollment figures or graduation rates never speak for themselves, DeWitt cautioned; such social statistics always require careful interpretive work. All the more so in the Soviet case: mundane gaps in data (which afflicted most social-scientific studies) were compounded by the Soviet government’s penchant for secrecy and for propagandistic massaging of data. Korol similarly urged caution, arguing time and again that it was fruitless to compare graduation rates between the Soviet Union and the United States, since the two nations’ educational systems differed so radically in structure and function. Indeed, Korol refused even to tabulate Soviet and American statistics side by side, in order to avoid “unwarranted implications.”6
DeWitt and Korol urged that Soviet educational trends be seen in the proper light. Although curricula for elite programs of study—such as physics at Moscow State University and at Columbia University or MIT—seemed to be roughly comparable in quality, several mitigating factors stood out. First, they both argued, a large proportion of scientists and engineers in the Soviet Union never practiced their craft, working instead in various bureaucratic or administrative posts. Second, the Soviet system was built around extraordinary specialization: the specialty of nonferrous metals metallurgy, for example, was itself carved up into eleven distinct specializations (copper and alloys metallurgy, precious metals refining, and so on). Students selected only one narrow specialty and devoted the bulk of their studies to it. Well into the late 1950s, meanwhile, Soviet students suffered from widespread shortages of textbooks and poor-quality (or missing) laboratory equipment. Student-to-faculty ratios had ballooned immediately after the war and continued to widen over the 1950s. There were also indications that academic standards were massaged to fit the central planning committee’s “production quotas”: both Korol and DeWitt noted internal Soviet reports of pressure to let mediocre students pass when overall numbers looked low.7
Most important of all, a fast-growing proportion of Soviet students were enrolled in extension-school or correspondence programs. Unlike regular full-time students, these students held full-time jobs away from universities and pursued their studies largely alone, reading textbooks (when these were available) and occasionally sending written assignments to overworked professors, most of whom juggled sixty-five to eighty such students at a time. Even Soviet education officials routinely remarked on the inferior quality of this type of training, especially for hands-on fields like science and engineering. Yet enrollments in extension and correspondence programs were soaring, even as regular full-time enrollments remained flat. By 1955, extension and correspondence students composed about one-third of all engineering enrollments in the Soviet Union; five years later, they accounted for more than half of enrollments in all fields combined.8
Only after delineating each of these factors at length did DeWitt broach numerical comparisons. Focusing on the Soviet five-year “diploma” programs—roughly akin to American-style training at the undergraduate and master’s degree levels—he presented some quantitative findings. Total enrollments in the United States were substantially greater than in the Soviet Union: three times as great as the regular full-time student population in 1953–54, for example, and still one-third larger if one included all the extension and correspondence students in the Soviet tally. Yet the balance of fields was quite different. In the United States, roughly one out of four students majored in scientific or technical fields, while in the Soviet Union it was three out of four. In particular, when DeWitt counted up annual degrees granted in the two countries, it appeared that the Soviets were graduating two to three times more students per year in science and engineering than were American institutions.9
That ratio—“two to three times”—quickly took on a life of its own. DeWitt’s and Korol’s reports had been careful, lengthy, serious affairs. The journalistic coverage, on the other hand, leaned toward the sensationalistic. Major newspapers like the New York Times and the Washington Post splashed the “two to three times” finding across their front pages. Leading spokespeople from the CIA, the Department of Defense, the Joint Congressional Committee on Atomic Energy, and the Atomic Energy Commission trotted out the same stripped-down number in public speeches and congressional testimony, with no trace of DeWitt’s caveats or cautions. Each proclamation elicited further hand-wringing in the newspapers.10 Here, in raw form, was the first step in economist Robert Shiller’s model of a speculative bubble: hype.
Before Sputnik, at least some observers tried to put the numbers in perspective, much as DeWitt had urged all along. In June 1956, Lee DuBridge—former scientific director of the wartime Radiation Laboratory at MIT, which had served as headquarters of Allied efforts in radar, and at the time president of Caltech—addressed the media frenzy when he testified before the newly formed National Committee for the Development of Scientists and Engineers. (Eisenhower had established the elite twenty-one-member committee just two months earlier, in response to congressional hectoring on the scientific manpower issue.) “It is true that in Russia more men and women received degrees in science and engineering last year than in the United States,” DuBridge began. “So what? Maybe that is because in the past 100 years they have so neglected their technical strength that they must now exert strenuous efforts to build it up. If this is true, then our rate of production should not be determined by their weakness—only by our own. Let us ask how many engineers we need to do our job, and not take over their figures from the numbers they require to do their job.” DuBridge might have mentioned another of DeWitt’s findings to bolster the point: even after the Soviets’ recent burst in scientific and technical training, they still lagged behind the United States in accumulated numbers of scientists and engineers available to the workforce.11
More typical, however, was the lesson that Senator Henry “Scoop” Jackson read in DeWitt’s numbers. Jackson released a special report entitled “Trained Manpower for Freedom” on 5 September 1957, just one month before Sputnik revved the rhetoric of manpower still higher. With the Soviets marching forward on the scientific manpower front—he invoked the now-familiar ratio—Jackson urged that nothing was “more precious” than the immediate development of all potential scientific talent in the United States and its NATO allies. More than ten pages spelled out various training programs to address the critical shortfall, including fellowships for high school and college students, special summer study institutes, and awards for students and teachers who excelled in science education. These resources, Jackson explained with a telling metaphor, “should be used as catalytic agents which, so to speak, can initiate educational chain reactions extending over the broadest possible scientific and technological front.”12
Sputnik further galvanized these discussions. DeWitt despaired of the “hysterical” reaction sweeping the country; it must have been especially galling to hear his own statistic echoed over and over again, stripped of all nuance and subtlety. Responding to the satellite, for example, former president Herbert Hoover groaned that “the greatest enemy of all mankind, the Communists, are turning out twice or possibly three times as many” scientists and engineers as the United States. Senator Lyndon Johnson quickly convened hearings before his Senate Defense Preparedness subcommittee within a week of the satellite’s launch, before which General James Doolittle (famous for his Tokyo bombing raid during the Second World War) brandished the same dire figure. During closed sessions of the hearings, CIA director Allen Dulles returned to the “manpower gap.” Details of Dulles’s testimony remained secret, but Johnson alerted the press that Dulles had confirmed that the Soviets were “now outstripping the U.S. in developing a scientific and technological manpower pool.”13 In the frenzied weeks after Sputnik, Korol’s book suf
fered similar misreadings. Reporting on the book’s release, one Washington Post article began by exclaiming, “The free world must radically change its ways to meet the challenge of the Soviet Union’s power to marshal brains and resources for priority projects.” This was an exact inversion of Korol’s point—as he had been at pains to make clear—and yet the reporter attributed this alarmist conclusion to Korol himself. Another Post article interwove coverage of Korol’s book with quotations from Eisenhower’s post-Sputnik addresses, giving the appearance that both had called for the “absolute necessity of increasing our scientific output” in trained personnel.14
Next came Shiller’s second phase: amplification. Enterprising physicists took full advantage of the opportunity afforded by the launch of Sputnik to further flog DeWitt’s number. Quickest to respond was Columbia University’s pugnacious Nobel laureate I. I. Rabi. Eisenhower and Rabi had known each other since the late 1940s, when Eisenhower had served as Columbia’s president; after Eisenhower became president of the United States, Rabi headed his new Science Advisory Committee. Meeting just a week and a half after the launch of Sputnik, Rabi pressed Eisenhower to use the satellite as a pretext for bulking up American scientific training. Soon after that, Elmer Hutchisson, the new director of the American Institute of Physics (AIP), a large umbrella organization that helped to coordinate various professional societies in the field, opined to reporters from Newsweek that the entire American way of life could well be “doomed to rapid extinction” unless the nation’s scientific reserves were expanded quickly. Behind the scenes, Hutchisson alerted his AIP colleagues that they had “the opportunity of influencing public opinion greatly.” He saw “an almost unprecedented opportunity,” he wrote in a memo to the AIP Advisory Committee on Education, “to take advantage of the present public questioning concerning the quality of science instruction in our schools.” Edward Teller, a veteran of the secret H-bomb program who had helped Senator Jackson prepare his “Trained Manpower” report, hit the same theme when talking with the press. “We have suffered a very serious defeat,” he exclaimed, “in a field where at least some of the most important engagements are carried out: in the classroom.” Hans Bethe from Cornell University—Los Alamos veteran and past president of the American Physical Society—found himself repeating DeWitt’s ratio of “two to three times” to journalists and in radio addresses without knowing (as his handwritten notes on typewritten speeches indicate) from whence the number had come or how it had been computed. Eager journalists soaked it all up.15
Lawmakers and their physicist consultants used the launch of Sputnik and the purported “manpower gap” in science and engineering to push through the massive National Defense Education Act, signed into law in September 1958. The act authorized about $1 billion in federal spending on education (nearly $9 billion in today’s currency), restricted to critical “defense” fields of science, mathematics, engineering, and area studies. The act represented the first significant federal aid to education in a century: not since the Morrill Land-Grant Colleges Act of 1862 had the federal government intervened so directly in higher education, which had traditionally been considered the prerogative of state and local governments. One close observer of the legislative wrangling behind the National Defense Education Act concluded that opportunistic policymakers had used the Sputnik scare and the DeWitt and Korol reports as a “Trojan horse”: the act’s proponents had been “willing to strain the evidence to establish a new policy.”16
Figure 7.1. Members of President Eisenhower’s new Science Advisory Committee leaving the White House in October 1957. From left to right: David Z. Beckler, Isidor I. Rabi, Jerome B. Wiesner, and Charles Shutt. (Source: Photograph by Paul Schutzer, The Life Picture Collection, courtesy of Getty Images.)
Passing legislation is usually a messy affair. The effects, in this case, were crystal clear. On the eve of the bill’s passage, US institutions had been producing only 2,500 PhDs per year across all of engineering, mathematics, and the physical sciences. During its first four years, the National Defense Education Act supported 7,000 new graduate fellowships, about 1,750 per year. The huge federal outlay, in other words, amounted to an overnight increase of 70 percent in the nation’s funding capacity to train graduate students in the physical sciences. During that same period, the act funded half a million undergraduate fellowships as well as providing block grants to institutions, with added incentives to states to increase science enrollments.17 Hence the final element of economist Shiller’s model: feedback.
: : :
The feedback loop had an immediate impact on graduate-level training. During the decade after passage of the National Defense Education Act, the number of US institutions granting PhD degrees in physics doubled, driving an exponential rise in the number of young physicists. According to data collected by the National Register of Scientific and Technical Personnel—created during the early 1950s to facilitate the federal government’s mobilization of scientists in times of war—the number of professional physicists employed in the United States grew substantially faster than the number of professionals in any other field between the mid-1950s and 1970: 210 percent faster than Earth scientists, 34 percent faster than chemists, 22 percent faster than mathematicians, and so on.18
Figure 7.2. Number of physics PhDs granted per year by US institutions, 1900–1980. (Source: Figure by Alex Wellerstein, based on data from the US National Science Foundation.)
Yet it was not to last. The conferral of physics PhDs in the United States peaked in 1971. Then, all at once, the curve fell sharply, its rate of descent an eerie reflection of its spectacular rise. A “perfect storm” had triggered the fall. By the late 1960s, internal auditors at the Department of Defense had begun to question whether the postwar policy of funding basic research on university campuses—which had underwritten the education of nearly all physics graduate students since the end of the Second World War—had produced an adequate return on investment. As the Vietnam War raged, meanwhile, campus protesters grew equally dissatisfied with the Pentagon’s presence on American campuses; protesters often targeted physicists’ facilities for their military ties (whether real or merely perceived). To supply troops for the escalation of fighting, military planners began to revoke draft deferments—first for undergraduates in 1967, then for graduate students two years later—reversing a twenty-year policy that had kept science students in their classrooms. Détente with the Soviets and the onset of “stagflation” in the early 1970s exacerbated the situation, inducing steep cuts in federal spending for defense and education.19
No field of study was affected more quickly or more dramatically than physics. While annual conferrals of PhDs across all fields slid by a modest 8 percent between their peak in the early 1970s and 1980, physics PhDs plummeted by half. Several fields experienced sharp downturns—mathematics down 42 percent, history down 39 percent, chemistry down 31 percent, engineering down 30 percent from the early 1970s highs—but physics led the way. Demand for young physicists vanished even more quickly. Whereas more employers than student applicants had registered with the Placement Service of the American Institute of Physics throughout the 1950s and into the mid-1960s, by 1968 young physicists looking for jobs outnumbered advertised positions by nearly four to one—and that included all kinds of positions, in academia, industry, and government laboratories. Three years later, the numbers were even grimmer: 1,053 physicist job seekers registered, competing for just 53 jobs.20
Robert Shiller is quick to note that speculative bubbles can develop without outright chicanery. That was certainly the case here. The influential physicists who used the DeWitt and Korol reports to argue for increased graduate training were doing their job; it was their responsibility to lobby on behalf of the profession. More generally, increasing support for higher education is hardly an evil thing. Yet in the rush to exploit DeWitt’s “two to three times” finding, the cycle of hype, amplification, and feedback came unmoored from any reasonable assessment of the underlying si
tuation.
Even if one set aside the significant caveats that DeWitt and Korol had delineated with such care—uneven quality of training, severe specialization, and the substantial numbers of extension and correspondence students inflating the Soviet statistics—the numbers themselves deserved a closer look. In tabulating numbers of graduates in engineering and applied sciences in the Soviet Union and United States, DeWitt had included three main categories: engineering, agricultural specialists, and health fields; these were the fields which, when tallied, produced the “two to three times” ratio. (As DeWitt explained, the grouping of fields in his tables was an artifact of the Soviet educational system, in which the vast majority of students earned degrees from “technical institutes” rather than universities; the institutes focused largely on these areas, whereas natural sciences and mathematics were mostly taught at universities.) Yet when repeating the “two to three times” number, not one commentator stopped to ask how a superabundance of agricultural specialists might lead to military supremacy—least of all given the Soviet Union’s catastrophic history of collective farming in the 1930s or the fierceness with which the state had backed agronomist Trofim Lysenko’s eccentric biological theories after the Second World War, quashing the study of genetics. Similarly for health professionals: no doubt an important segment of the labor force, but would more nurses and dentists lead to better bombs? Nearly everyone who picked up DeWitt’s numbers used the label “science and engineering,” never pausing to consider just which fields of science or engineering they represented.21