Frank Stanton sensed trouble. Sitting in his living room on the night of October 30, 1938, the young CBS executive tuned in to catch Orson Welles’s adaptation of H.G. Wells’s War of the Worlds. The program sounded crisp and engaging — but a bit too realistic. Stanton grabbed his coat and headed back to CBS’ headquarters on Madison Avenue. Pushing his way through chaotic hallways jammed with reporters, police, and network employees, he reached his desk and telephoned his friend Paul Lazarsfeld.
Stanton and the sociologist Lazarsfeld set out to measure the panic as quickly and accurately as possible before it subsided. Their basic results would spur a remarkable conversation that reverberates 70 years later in social psychology, media theory, federal regulation, and other fields.
The “War of the Worlds” broadcast remains enshrined in collective memory as a vivid illustration of the madness of crowds and the deeply invasive nature of broadcasting. The program seemingly proved that radio could, in the memorable words of Marshall McLuhan, turn “psyche and society into a single echo chamber.” The audience’s reaction clearly illustrated the perils of modernity. At the time, it cemented a growing suspicion that skillful artists — or incendiary demagogues — could use communications technology to capture the consciousness of the nation. It remains the prime example used by media critics, journalists, and professors to prove the power of the media.
Yet the media are not as powerful as most think, and the real story behind “The War of the Worlds” is a bit more complex. The panic was neither as widespread nor as serious as many have believed at the time or since.
Nobody died of fright or was killed in the panic, nor could any suicides be traced to the broadcast. Hospital emergency-room visits did not spike, nor, surprisingly, did calls to the police outside of a select few jurisdictions. The streets were never flooded with a terrified citizenry. Ben Gross, the radio columnist of the New York Daily News, later remembered a “lack of turmoil in front of CBS” that contrasted notably with the crowded, chaotic scene inside the building. Telephone lines in New York City and a few other cities were jammed, as the primitive infrastructure of the era couldn’t handle the load, but it appears that almost all the panic that evening was as ephemeral as the nationwide broadcast itself, and not nearly as widespread. That iconic image of the farmer with a gun, ready to shoot the aliens? It was staged for Life magazine.
So what accounts for the legend? First — and perhaps most important — the news media loved the story, and Welles loved the news media. The panic became a global story literally overnight. Even the Nazis could not resist commenting, noting the credulity of the American public. Americans certainly appeared gullible, but they were not alone. The news media, handed a sensational story of national scope, reported every detail (including fictional ones) about Welles, the program, and the reaction.
Welles’s greatest performance that evening wasn’t in the studio; it was in a hallway, at the improvised news conference, when he feigned a stunned, apologetic demeanor. In reality, as Paul Heyer notes in The Medium and the Magician, Welles carefully concealed his satisfaction with the hysteria while expressing concern over the rumors of deaths attributed to the program. The threats of investigation coming from the Federal Communications Commission bothered Welles, too, but they were primarily CBS’s problem.
It was the government, and its relationship to CBS, that worried Stanton. While Welles spoke to reporters a few floors away, he and Lazarsfeld created a brief survey instrument to gauge the significance of the panic. Without consulting his bosses, who were occupied at the time, Stanton phoned a trusted survey organization to conduct nationwide interviews as soon as possible. Data were compiled over the following 24 hours and immediately forwarded to Stanton’s CBS office.
Unfortunately, those data, if they still exist, are unavailable to scholars. CBS, unlike NBC, severely restricts access to its archives. But Stanton’s survey has trickled down to us through a classic study in the emerging field of social psychology, Hadley Cantril’s The Invasion From Mars: A Study in the Psychology of Panic (1940). Cantril, a Princeton social psychologist; Stanton; and Lazarsfeld had created the Office of Radio Research, a Rockefeller Foundation-supported project based at Princeton that can be considered the first significant attempt to empirically analyze the effects of mass media.
Cantril’s study, which remains the most enduring source for what we know about that night, combined the CBS data, a second survey conducted six weeks later by the American Institute of Public Opinion, and a series of detailed interviews with 135 people, of which “over 100 were known to have been upset by the broadcast.” Admitting that his interviews did not comprise an accurate sample of either the national population or the radio audience that evening, Cantril nevertheless filled his short volume with narratives of terror and fear. The interview subjects — all from New Jersey “for reasons of finance and supervision” — were found by the “personal inquiry and initiative of the interviewers” hired by Cantril. They were a self-reporting, self-selected cohort. Cantril did attempt to interview people identified in newspapers as frightened, but that effort proved almost entirely futile.
Such reliance on qualitative measures, while using an unrepresentative sample, only begins to hint at Cantril’s methodological problems. Cantril’s estimates of how many people actually heard the broadcast, and how many were frightened, are wildly imprecise. Because CBS’s Mercury Theatre on the Air lacked sponsorship, the C.E. Hooper Company, the commercial ratings service used at the time, did not rate Welles’s program. The American Institute of Public Opinion national survey (taken six weeks after the program, following an avalanche of publicity) found 12 percent of respondents claiming they had heard the broadcast. That represents an audience of almost 12 million Americans — a number that is certainly far too high. Slightly less than four million Americans had tuned into Welles’s Mercury Theatre on the Air the week before “The War of the Worlds.”
From such disparate approximations Cantril offered the “conservative estimate” that six million Americans heard the broadcast. The public-opinion institute’s survey found that 28 percent of the listeners believed the broadcast contained real news bulletins, and of that 28 percent about 70 percent were “frightened or disturbed.” These numbers undercut several of Cantril’s assertions about the scope of the panic; they reveal that about three out of four listeners knew the program was fiction. So Cantril did what many social scientists faced with disagreeable data do: He spun the numbers. The low numbers, he wrote, represent the “very minimum of the total number actually frightened” because “many persons were probably too ashamed of their gullibility to confess it in a cursory interview.” He candidly admitted that “there is the possibility that some people heard so much about the broadcast that they reported actually hearing it.”
In other words, Cantril concluded that many respondents probably lied.
Cantril’s assertions about the data are largely forgotten. His book is cited far more for its tales of panic than for its faulty statistical analysis or sampling anomalies. His study survives because it supplies what many scholars and journalists need: academic proof for what they think they already know. It legitimized the myth of the night of terror as perhaps nothing else could.
Neither Stanton nor Lazarsfeld was satisfied with Cantril’s work. On the personal level, Cantril and Lazarsfeld did not get along. One was a Harvard-trained WASP with the social connections needed to land a prestigious post at Princeton; the other was a thickly accented, chain-smoking, Jewish refugee from Vienna trained at the intersection of economics, mathematics, and applied psychology. Nor did it help that Lazarsfeld once made a pass at Cantril’s wife, a piece of information Stanton relayed to me in an interview.
A few years after the publication of Cantril’s book, Stanton and Lazarsfeld excoriated their colleague in confidential interviews with Rockefeller Foundation officials. Stanton told an interviewer that Cantril’s original manuscript was “completely unsatisfactory,” and he admitted he had “no respect for Cantril’s scholarly standards.” Lazarsfeld was even more brutal, telling the interviewer that some of Cantril’s conclusions were “laughable.” Because Cantril was “pathologically ambitious,” according to Lazarsfeld, he was “a highly dangerous influence in the field.” Stanton told the foundation officials that he and Lazarsfeld essentially rewrote the manuscript and allowed it to be published under Cantril’s name.
That explains why some of the book’s less-emphasized conclusions foreshadowed important findings about the power of the media. The hypodermic model of media effects, which prevailed at the time, posited that the media injected ideas, more or less directly, into the consciousness of the audience. The book’s data seriously undermined that model, demonstrating empirically that each member of the mass audience filters the media’s messages through a matrix of personal variables (education, critical ability, class, etc.). Those data complicated media theory tremendously and intensified the research focus on the complexities of audience reception.
Lazarsfeld surprised many by concluding in The People’s Choice, his classic study of the 1940 election, that the media’s effects are, in general, much more selective and limited than we assume. Other forms of communication, from those in the education system to religious communication to interpersonal communication, were apparently more powerful. The mass media were but one part of a larger web of influence, and as one factor, their actual influence was mediated by several other variables. Thus, the media’s ability to control us was far less pronounced than assumed.
That is the ultimate irony behind “The War of the Worlds.” The discovery that the media are not all-powerful, that they cannot dominate our political consciousness or even our consumer behavior as much as we suppose, was an important one. It may seem like a counterintuitive discovery (especially considering its provenance), but ask yourself this: If we really know how to control people through the media, then why isn’t every advertising campaign a success? Why do advertisements sometimes backfire? If persuasive technique can be scientifically devised, then why do political campaigns pursue different strategies? Why does the candidate with the most media access sometimes lose?
The answer is that humans are not automatons. We might scare easily, we might, at different times and in different places, be susceptible to persuasion, but our behavior remains structured by a complex and dynamic series of interacting factors.
Later media theory, and empirical research, would complicate and refine those earliest findings. But the basic problem of audience reception remains stubbornly resistant, and as long as the mass media exist, we’ll have empirical studies with dueling conclusions concerning effects. Many people, including scholars, will continue to believe something they intuitively suspect: that the media manipulate the great mass of the nation, transforming rational individuals into emotional mobs. But notice how those who believe this never include themselves in the mob. We are, as the Columbia University sociologist W. Phillips Davison once pointed out, very susceptible to the notion that others are more persuadable than ourselves.
Would you have fallen for Welles’s broadcast? If not, why do you assume so many other people did?
Michael J. Socolow is an assistant professor in the department of communication and journalism at the University of Maine at Orono. He is completing a manuscript on regulation and competition in the first two decades of American network radio.
http://chronicle.com Section: The Chronicle Review Volume 55, Issue 9, Page B16