The greatest advertising medium in the world is radio.

—Rudy Vallée, 1930

Media is a question of how many seats and how many arses.

—Phil Richardson, head of the first corporate media department in the United States (Procter and Gamble), 1930s

2

The Classes and the Masses in the 1920s and 1930s

“Tell it to the masses and the classes will understand,” wrote the National Broadcasting Company of its program The Clicquot Club Eskimos in 1929.¹ This chapter addresses the question of the audience: how it was investigated and studied by the networks and advertising agencies, and how it was eventually commodified. The writings on the subject of the audience in this period are influential though few in number. They are also mainly theoretical, or concentrate on the rise of audience studies around major figures such as sociologist Paul Lazarsfeld.² It is important, however, to examine the problem historically: How were audience preferences obtained? What were the processes that resulted in the construction of the audience? The central tension for broadcasters and advertisers in this era was to cast a broad net. But at the same time, identifying audiences to target was important, as broadcasters wanted to be able to attract the wealthiest consumers to market expensive goods to, as well as a large middle group of listeners.

Constructing the Mass Audience

Radio played a powerful role in the rise of mass culture in the United States, although Americans had a concept of the “mass” before. Indeed, this was one of the central tensions of late nineteenth- to early twentieth-century thought, as has been studied by T. J. Jackson Lears and others. Lears argues that urban-industrial transformation in the nineteenth century gave rise to changes in Americans’ everyday lives; urban life as anonymous individuals instead of known neighbors in villages and the countryside, as well as modern conveniences that removed people from the daily experience of the outdoors resulted in much hand-wringing about a perceived loss of individuality. Lears quotes a magazine writer from 1909: “We are a mass. As a whole, we have lost the capacity for separate selfhood.”³

The advent of radio did not create a sense of mass; it only emphasized a tendency already under way. Yet, at the same time, radio imparted a feeling of togetherness, since everyone knew that listeners were hearing the same programs at the same time. Millions of people could read popular books and magazines, or see a movie, or listen to a phonograph record, though one commentator observed in 1923 that there is “a lack of immediate contact with the audience” when playing phonograph records.⁴ With radio, however, those millions were listening simultaneously, but not as a group. As one early observer wrote the same year:

How easy it is to close the eyes and imagine the other listeners in little back rooms, in kitchens, dining-rooms, sitting-rooms, attics; in garages, offices, cabins, engine-rooms, bungalows, cottages, mansions, hotels, apartments; one here, two there, a little company around a table away off yonder, each and all sitting and hearing with the same comfort just where they happen to be.⁵

This writer continues to imagine the far-flung audience, united in time and in their rapture at hearing organ music emanating from their radios.

Roy Durstine, one of the founders of BDO (Barton, Durstine and Osborn) in 1919, which founded the first radio department in 1925, wrote in 1928,

Radio comes right into the middle of the family circle. It speaks and plays and sings for millions, but those millions are divided into countless audiences of one or two or perhaps three or four. At most a small group of friends is gathered together.

This is no mass psychology.⁶

By early 1929, according to Charles Henry Stamps (on whom I am relying heavily for this history of audiences), the radio audience was becoming seen as something more complicated than an undifferentiated mass or as a collection of individuals. Merlin H. Aylesworth, the first president of NBC, wrote in June 1929 that to understand radio advertising, one must remember that radio

cannot be an appeal to crowd psychology. We hear so much about the radio audience of many millions that we are inclined to visualize it as a vast throng sitting together, listening to what comes over the air. Of course it’s nothing of the kind. It is almost invariably made up of family groups of not more than four or five people who select the evening’s entertainment with increasing care and discrimination.⁷

Radio was thus an unprecedented technology, making people feel alone and together, simultaneously.

Who Is Listening?

But sentiments such as the above were mainly based on conjecture, for, unlike print advertising, whose sales could be measured in magazines and newspapers sold, broadcasters and advertisers had little idea who was listening, or how they were listening. A 1923 article in Radio Broadcast said, “It would be illuminating to have on record not the number of the possible radio audience, but the actual number of the audience listening, on a particular evening, to any particular programme.”⁸ From the beginning of radio, broadcasters, advertisers, and advertising agencies were desperate to discover everything they could about their audiences; people attempted to extrapolate from figures of radios sold, numbers of fan letters received, and so on.

In the absence of hard data about audiences, the most common assumption was that if one person liked it, everybody could like it. But who was this everybody? Most influential was the notion that the “average mental age” of an American was somewhere between the ages of twelve and fourteen. This information was derived from a pool of one hundred thousand American soldiers who were tested for their IQ in World War I and found to have an average mental age of thirteen, a figure that was taken as gospel by people in broadcasting and advertising for much of the period before World War II and which played a powerful role in the rise of discourses about the “mass” nature of broadcasting for several decades.⁹ An exasperated editorial in Fortune in 1942 said, “For nearly a generation, the people of this country have been suffering patiently from a curious mental illness, a peculiar delusion about themselves. That disease is the wide acceptance of a fallacious axiom: that they have an average mental age of a thirteen-year-old child.”¹⁰

Before the establishment of a system to rate listenership in 1929, advertisers and sponsors gauged the success of programs by the quality and volume of listener response. They often offered samples in order to gather information about listeners who mailed in their responses; and they sponsored contests whose main purpose was to solicit listener feedback.¹¹ One contest, sponsored by Gilmore Oil Company Ltd., manufacturer of Blu-Green Gas, solicited verses to its program’s theme song, entitled, unimaginatively, “Blu-Green Gas.” In 1931, the vice president of Gilmore’s advertising agency wrote an article for the trade press, reporting, “To date more than 25,000 verses have been received, which we believe justifies our calling the Blu-Green Gas Song the longest song in the world.”¹²

Another tactic was to hint that a program was about to be canceled in order to generate fan mail, which seems to have worked; I have seen a number of sets of fan mail in various archives complaining about the announced cancellation of a program. A variant on this strategy was mentioned in a 1931 letter in the NBC archives, which says, “Sometimes, we will take a program off the air in order to find if public opinion expresses itself in the form of a protest.”¹³

But the main measure of a program’s success in the early days was to read listener mail, however it was solicited. Roy Durstine wrote of fan mail in 1930,

It is a part of the agency’s work in connection with radio to keep a record of this fan mail. This record is not merely a count of replies analyzed by localities. It is a careful study of the type of stationery, the sex and apparent age of the writer, a rough classification of good, fair, and poor responses, and a thoughtful reading of the letters for suggestions in building future programs.¹⁴

Mail was read not just for content but, as it were, ethnographically: Frederick Lumley wrote of CBS studying letters and searching for clues to social group by examining the type of paper, grammar, spelling, punctuation, and sentence structure.¹⁵

Early radio listeners seem to have been avid letter writers. A letter from E. P. H. James, sales promotion manager at NBC, to Harcourt Parrish of Ivy Lee and Associates, a public relations firm in New York City, in July of 1931 said that in 1930, NBC received 2,180,000 letters from the public with comments on programs, some with criticisms. “These criticisms are carefully read and every attempt made to offer programs which will please as many members of the general listening public as possible,” he observed, and went on to note that there are a variety of tastes. Tastes changed quickly, which means that NBC had to offer a variety of programs and “watch the reception of each type of program very closely.”¹⁶ Parrish had evidently asked about inducements in order to solicit mail response, and James responded with a tally of what 143 NBC advertisers had done, as of 1 May 1931. The majority (34 percent) made no offer; 19 percent offered a booklet, 14 percent a sample, 9 percent recipes, 6 percent novelties, 4 percent miscellaneous, 4 percent dealer’s name, 3 percent photographs, 3 percent club membership, 2 percent songs, 1 percent program suggestions, and 1 percent comic newspaper.¹⁷

In January 1930, the J. Walter Thompson Company began airing the Davey Tree Hour, a weekly program featuring light classics and sentimental songs, including a lecture on trees by Martin L. Davey, the son of the company’s founder. Davey published an article in 1932 on the success of his program, saying that he believed the public’s appetite for jazz to be satiated, and that “a very small percent of the people really understand and appreciate classical music of the more difficult sort.”¹⁸ Because of this low number, Davey believed that familiar melodies would offer the most appeal though variety was also necessary. To that end, “We made sure that every other piece was fast and every other piece was slow. We made an effort to have a variety of songs of racial origin, including English, Scotch, Irish, Welsh, German, Italian, etc.”¹⁹

This program generated a good deal of mail that pleased John Reber, head of the J. Walter Thompson Company Radio Department, and his colleagues. Reber said at the staff meeting of 14 January 1930 that the company received about six hundred unsolicited letters in praise of the program. “Never before . . . have we had letters of such a high quality; they were even considerably higher quality than those received as a result of one classical music program that we had. The stationery, the English and the evidences of culture on the part of the writers were decidedly higher than the average.”²⁰ Davey himself wrote that the letters the company received regarding its program were almost never rude, and that “a majority of them come from people of education and culture.” Yet, he noted, “We have received a very large number of letters written on cheap paper and in poor English.”²¹ Davey nonetheless maintained that these writers were intelligent people.²²

Despite the broad audience that Davey sought with his program, however, he still felt that the program appealed to a higher class of people (than, perhaps, jazz programs, though he didn’t say this). Davey, echoing the worries of advertisers and agencies in this period, also admonished the reader that since radio programs come into people’s homes, broadcasters had an obligation “to respect the feelings and taste of the better class of citizens.”²³

Attention to fan mail dropped with the rise of more sophisticated means of measuring audiences through polling; there may also have been suspicions about fan mail. Merrill Denison wrote in 1934, in an assertion that I have not encountered anywhere else, “Fan mail was highly regarded until the discovery was made that supplying it had become a racket,—that any interested person could buy letters to New York at so much a ton.”²⁴ Accurate or not, by the mid- to late 1930s, fan mail was less important as a way of measuring audience reactions. “To-day,” wrote one commentator in 1937, “executives read only exceptional letters, depending upon audience-mail tabulators to record, file and analyze letters and their contents. Favorable and unfavorable reactions are recorded on punched file-cards, enabling one to see statistically what the fans are saying. Machines, each with a daily capacity of 5,000 cards, compile the records.”²⁵

Polling

Countless polls and surveys were conducted in order to construct a profile of the average audience member and what he or she wanted. Advertisers and sponsors also used polls about listener preferences to conceive programs.²⁶ Depending on how the poll was worded and the kinds of musical categories employed, the poll results usually favored dance music, though classical music fared better in some polls.

Taking polls and encouraging listener responses was the first step away from simple assertions of advertisers’ and sponsors’ musical tastes in programming and reliance on fan mail. In a broadcasting environment in which advertising paid for programming, obtaining accurate information on listeners’ preferences was crucial, and polling and audience surveys quickly became prevalent, and increasingly scientistic.²⁷ These polls and listener surveys helped advertisers and advertising agencies tailor programs to fit the target audience, based on the kind of music used.

Table 2.1 The “ideal program,” 1923

Source: “Radio Audience Decides Programs,” Wireless Age, August 1923, 29.

The earliest polls were mainly designed to assess audience desires about broadcasting generally, as in this 1923 poll about the ideal program, with the following responses tabulated in table 2.1. Note how classical music and jazz are the only two music categories, as this is the generic fault line of the debate over what kind of music should be aired, a debate that lasted for much of the 1920s.

Later polls addressed more specific listener preferences, especially about music. The categories are revealing of the tastes of the time, with categories such as “old-time songs,” “Hawaiian music,” “mixed quartettes,” “saxophone,” “pipe organ,” “religious music,” “Mexican music,” and others.²⁸

What was reported as the first survey of the radio audience was conducted in 1928 by Daniel Starch, a Harvard-trained psychologist turned market researcher. It was less about program preference than a simple attempt to ascertain the number of radios in American homes, how much people listen, when people listen the most, and the like; programming was almost an afterthought. But the study was innovative in that it broke out the data based on geography, with categories such as “farm families.” Starch concluded:

Preferences for most of the different types of programs are practically the same on the part of farm, town, and city families. There are significant differences, however, with respect to several types of programs—semi-classical and classical music, and grand opera are preferred less by farm and small town families than by city families, whereas religious services, crops and market reports, and children’s programs are preferred more by farm than by city families.²⁹

An exhaustive study in 1929 by Archibald M. Crossley for the Association of National Advertisers Inc., drawing on a poll of radio editors east of the Rockies, six thousand personal interviews in twenty-five large cities, twelve hundred interviews in twelve other cities, and fifteen thousand interviews east of the Mississippi, revealed that, of the seventeen different programs named, most were musical (almost all popular music of various kinds) except for five. In this era, 33 percent of programs were produced by advertising agencies, 28 percent by the network for its sponsors, 20 percent by the sponsors themselves, and 19 percent by special program builders; these numbers gradually shifted toward the advertising agencies.³⁰

In this same period, data were appearing that attempted to index listening preferences to income and age, such as these from a 1931 article by advertising man P. H. Pumphrey, who drew on a study commissioned by the Universal Broadcasting Company in Philadelphia. This study charted program types with income, revealing the unsurprising fact that high-income groups tended to prefer classical or “semi-classical” music over “Sacred” music or “Old-fashioned melodies,” and that low-income groups overwhelmingly preferred dance music.³¹ Pumphrey concluded that the popularity of dance music makes it the most desirable music for advertising (others had stated that it was youth who preferred dance music).³² But good bands are very hard to find, Pumphrey wrote, and unless there is a “personality” such as early superstar Rudy Vallée, most programs could not be supplied. Thus, many advertisers found that the light classical or semipopular musical program, employing some “dance music, some musical comedy and light opera, some ballads, and occasionally some heavier music, is hailed by many as the golden mean.”³³ This format “permits the development of a number of distinctive personalities,” and Pumphrey continued by listing such programs, which “seem to have struck the least common denominator of popular taste, and appeal to everyone.”³⁴

Polling and audience research were to become even more important with a shift in the business model of the broadcasting industry. According to a historian of the N. W. Ayer agency, Ayer reversed the method by which advertising agencies worked. Before 1930, the strategy was to seek attractive programs and then attempt to find advertisers that would pay for them. But after the novelty and glamour of radio advertising wore off, advertising agencies realized that they had to begin with the client’s needs and attempt to devise a program that would address them.³⁵ This shift, according to Stamps, required more “scientific” knowledge of the audience. Enter Archibald Crossley in 1929, who was asked by the American Association of Advertising Agencies and the Association of National Advertisers to head the new Cooperative Analysis of Broadcasters, which began rating program popularity in 1930. Crossley employed the “recall” method of learning what listeners preferred, which entailed calling them several days after a program had aired.

Learning how to attract a particular market segment became an increasingly important aspect of broadcast advertising as the 1930s progressed.³⁶ Listener polls attempted to be comprehensive both about programs and about the social class of the listener. A table published in Broadcast Advertising in 1930, for example, divided Boston listeners by occupations into three classes:

Class A—8 percent of the total—includes the families of merchants, professional men, executives, manufacturers, etc. Class B—73 percent of the listeners interviewed—is made up of families of skilled and clerical workers, salesmen, city living farmers, retired, and small merchants. Class C—19 percent—takes in the families of laborers, domestics, clerks, and non-employed.³⁷

Polling, and the idea of attempting to capture a particular audience for a particular product with particular music, spelled the beginning of the end of the goodwill concept, also hastened by the onset of the Great Depression. Advertising agencies became bolder about incorporating advertising messages into the programs and targeting particular audiences. As Fortune magazine observed in September 1932, “The sponsors have shaken off all the old network inhibitions except the one about specifying prices during evening hours.”³⁸

Commodifying the Mass Audience

The National Broadcasting Company had originally conceived of programs as a means to sell hardware, since, because it was owned by the Radio Corporation of America, its main profits were realized through sales of radios and related equipment. But in 1932, government antitrust action forced the General Electric Company and the Westinghouse Electric and Manufacturing Company to relinquish their holdings in the Radio Corporation of America, NBC’s owner. Since General Electric and Westinghouse manufactured radio components, NBC lost this incentive.

But even before this, NBC had begun to shift its business model, thanks to competition from the Columbia Broadcasting System, which did not profit from radio sales. William Paley, the network’s founder, understood what no one else did at the time—that selling what one would now call “content” was not the way to make money. This was old-think: if Tin Pan Alley music publishers sold songs, if publishers sold books, then broadcasters logically sold programs. Paley realized that what broadcasters really sold were audiences. That is, in a system in which programs were funded by viewers indirectly through advertising rather than directly through government subsidies as in Europe, the broadcaster’s function was not to produce programs but to sell audiences to advertisers. The idea, as summarized succinctly by Dallas W. Smythe, is that advertisers purchase the anticipated attention of audiences of known demographic content who will attend to particular programs in predictable numbers at particular times in particular geographical markets.³⁹ Or, as Paley put it plainly in 1934,

We start with the premise that the advertiser makes our 16 daily hours of radio service possible, just as the advertiser makes possible the daily newspaper and the national magazine. In order . . . to have something to sell this advertiser, we must render a very definite service to the radio audience. We must have listener attention—just as a newspaper must have reader circulation—before we can secure advertising revenue.

This means the first problem of our business is to win an audience, hold an audience, interest an audience.⁴⁰

Advertising agencies understood this perfectly well. As Roy Durstine wrote in 1930, “The public wants entertainment. The advertiser wants the public’s attention and is willing to pay for it. Therefore let the advertiser provide the entertainment.”⁴¹

Paley wrote in his autobiography that when he first acquired the majority share in United Independent Broadcasters, the original 1927 contract was a severe drain on the company’s resources, for it obligated the network to purchase ten hours a week from each of its affiliates for fifty dollars an hour. The network was paying out about seven thousand dollars per week, regardless of whether it had sold time to sponsors to cover its expenses.⁴² Paley’s predecessor had devised a new contract, which required the stations to pay the network for sustaining programs that the network developed. But this system did not staunch the drain on resources, for the network didn’t have that many sustaining programs in this period.

When Paley took over, he revised the contract again to make it more attractive to the network’s affiliates. The sustaining programs would henceforth be free to affiliates, and the network would guarantee twenty hours of programming per week, pay the stations fifty dollars an hour for the commercial hours used, but the network wouldn’t pay the stations for the first five hours of commercial programming time. The main change, however, was the crucial one: the network would have exclusive rights for network broadcasting through the affiliate, and affiliates would have to identify programs using the CBS name.⁴³ The affiliates agreed to the proposition in November 1928, and the fledgling network attracted more affiliates.

As Fortune magazine put it in 1935 in an analysis of the economics of broadcasting, “Falling down the rabbit hole of the broadcasting studio he is in a land of Mad Hatters and White Knights who sell time, an invisible commodity, to fictitious beings called corporations for the purpose of influencing an audience that no one can see.”⁴⁴ This was the problem, as Fortune put it: Radio stations have to fill the air with sixteen hours of programming per day, from 8:00 a.m. to midnight (sometimes the networks would broadcast more). How? If, say, an independent station could cover three hours a day through advertising revenue, thirteen hours still remained. If the station got cheap and free talent and used prerecorded programs, it wouldn’t cost that much (Fortune estimated it worked out to a talent cost per hour of just thirty-five dollars). Selling this airtime would generate significant revenue, but in the end, the independent station would pretty much break even unless it could manage to sell more than three hours per day.

Because of this risk, many independent stations elected to become part of a network, even though they had to give up their best evening hours, which they could have sold to local advertisers. In exchange, however, they received much better talent in the form of free sustaining programs provided by the network. Better talent brought greater prestige for the local station. With a network contract, local affiliates received prestige and security; the network received exclusive access to the affiliates’ best evening time slots, which it could then guarantee to advertisers.⁴⁵

Fortune provided a case study, a Cleveland radio station that derived 70 percent of its revenue from local advertisers and national advertising targeting that city. The affiliate’s contract with CBS gave the affiliate

an option on its time to the network, which may buy any hour it wants, even the choice evening hours. On the other hand, the network pays the affiliate something less than $100 an hour . . . for handling network commercial programs and at the same time sends out sustaining programs for the local station to use free of charge during those hours that have not been sold to advertisers.⁴⁶

According to Fortune, “The motive for all Mr. Paley’s toil has been to sell time. From 1929 to 1934 the number of hours Mr. Paley sold more than doubled.”⁴⁷

Paley later had the idea of enticing advertisers to purchase time on a greater number of affiliates, for in that period advertisers didn’t always buy time on the entire range of affiliates. Paley realized that if advertisers didn’t want the entire network, he could offer them a discount if they took the entire network. This would mean that the network-sponsored programs would be much more widely heard and raise the network’s overall income.⁴⁸

NBC, by contrast, charged its affiliates for sustaining programs, which gave them no incentive to air them, and it didn’t reserve any hours for itself. This meant that local stations could run their own programs during the best time slots, which might have been desired by an NBC advertiser. NBC’s comparative inability to guarantee audiences was discussed at a J. Walter Thompson Company staff meeting in 1931, in which a Mr. Spencer reported,

N.B.C. is getting its network into better shape, and we hope that before very long they will have their stations lined up, either by purchase or by more firmly bound contracts so that we will know just what stations we are going to get when we put in an order for them. As you know, there are some stations on the network now that take a program or leave it, just as they like.⁴⁹

Obviously, such a system was unsatisfactory to both advertisers and advertising agencies, but it was dominant until Paley’s new contracts at CBS. NBC ultimately had to change its contracts, essentially duplicating CBS’s.

A document in the NBC archives sheds some light on CBS’s strategy in this period. Somehow (the memo attached to the report at NBC says, “A CBS program—how did NBC get it?”), NBC acquired a CBS publicity campaign report, dated 3 December 1931, for a program entitled Music That Satisfies with Nat Shilkret. After an introductory paragraph, the report states matter- of-factly, “The purpose of the campaign, in general, will be that of attracting listeners to the program rather than advertising the product. Columbia’s publicity thus will center upon Nat Shilkret, conductor, and Alex Gray, Baritone soloist.” Even though this program was sponsored by Liggett and Myers Tobacco Company, makers of Chesterfield cigarettes, the star of the show was to be Shilkret, not Chesterfield. This document mentions Columbia’s “exploitation division,” which was “assigned to the building up of artists in ways other than the dissemination of news and pictures designed primarily for radio pages.” This division was essentially tasked with putting Shilkret or Gray in every conceivable venue for the publicizing of the show, including attempting to get one of the major railroads to name one of its new trains “The Chesterfield,” the formal dedication of which would be attended by Gray and other cast members. The proposed publicity campaign was multipronged and extensive: placing stories in the radio press (“Columbia has taken the lead in offering tailor-made copy,” it claimed); biographies of the stars to be distributed nationally, but also “localized,” sent to “home town papers,” stories not just of the stars but of the “production men and control engineer assigned to the program”; setting up interviews with the stars with radio editors at local papers; and “the fan mail received by the artists will be examined carefully for possible stories.” “In short, the publicity department will put special emphasis upon Mr. Shilkret and Mr. Gray, but will attempt to cover the programs from every angle in a concerted effort to build an ever-increasing audience for ‘Music That Satisfies.’ ”⁵⁰

By 1933, as Stamps notes, both NBC and CBS saw that radio was primarily a sales medium.⁵¹ Only fifteen years later, NBC could brag to a potential client, “Through what medium is it possible for an advertiser virtually to own a specific half-hour of millions of families’ LIVES, with a week-after-week, month-after-month regularity?”⁵²

Ears and Incomes

Learning more and more about audiences was not simply a way for broadcasters, advertising agencies, and advertisers to design programs that audiences might like; it was increasingly, as Paley grasped, a way to ascertain what kind of audience each program attracted in order to sell that demographic to advertisers.

In 1933, CBS published a study that tackled the question of income and listening preferences, Vertical Study of Radio Ownership, 1930–1933. The following year, it published Markets in Radio Homes, by Income Levels and Price Levels, and Ears and Incomes.⁵³ It hired Daniel Starch to attempt to discover how much of an audience a highbrow radio program might reach beyond its targeted audience. “Can a specific network program penetrate those upper income levels—in actual audience—as deeply as it does the lower? Even more deeply? And what about the income levels in between? Can it hold the three-room ‘Smiths’ while it wins the ten-room ‘Smythes’?”⁵⁴

To address this question, Starch examined the audiences for four different programs: The March of Time, a news program associated with Time magazine; The Chesterfield Program, a musical show featuring André Kostelantez and his orchestra and classical music guests; The Philco Radio Program with news commentaries by Boake Carter; and The Fletcher’s Castoria Program with an orchestra led by Don Voorhees with classical music guests. Starch’s research strategy was touted by CBS as more accurate than the ratings system used at the time, in which people were simply telephoned and queried about what program they were listening to at that moment. Starch’s researchers went door-to-door to 7,490 homes across the country and asked specifically about each of the programs, in order to tackle “the question of the whole job which the whole advertising investment has done.”⁵⁵ The income level of each family was coded as A (incomes over $5,000 per year), BB (from $3,000 to $5,000), B ($2,000 to $3,000), C ($1,000 to $2,000), or D (under $1,000). Income level was determined by the husband’s occupation and the size and character of the home. It should be remembered that this was not a study that set out to find correlations between income group and radio listening preferences but rather one that was designed to see how far down the economic ladder highbrow programs penetrated, and whether or not an advertiser that aimed at an upper-income audience actually captured that audience. The CBS study concluded that it was possible to attract upper-income audiences as readily as those with lower incomes, and that programs designed to appeal to upper-income audiences could garner attention from lower-income listeners; and maintained that “even with a ‘selective’ program keyed to a selective audience, it is possible for an advertiser, in a few months of CBS broadcasting . . . to reach 2 out of 3 of all radio homes in all income levels, 36,000,000 listeners (while he reaches 3 out of 4 of all radio homes in the upper income levels).”⁵⁶

A later study, by H. M. Beville Jr., research manager at NBC, further investigated the “social stratification of the radio audience,” as his 1939 report was called. Beville, who used data from the Cooperative Analysis of Broadcasting, divided social class into four groups: over $5,000 (6.7 percent of the total sample), $3,000–$5,000 (13.3 percent), $2,000–$3,000 (26.7 percent), and under $2,000 (53.3 percent). Beville found that the link between income and interest in classical music was very strong; in the words of the foreword by the Princeton Radio Research Office, “When the audience of a serious musical program is analyzed, good music is shown to be the monopoly of the upper income classes. The audience decreases markedly with decreasing income.”⁵⁷ Beville’s data on this are clearer in a subsequent article, with data that generally trend downward from the most serious classical music to less, with the exception of a program featuring the Philadelphia Orchestra, but it’s evident that this program was the highest rated, with 14.2 percent of its audience coming from the highest income group.⁵⁸

Results

How did all this research play out? Money spent on radio advertising generally skyrocketed as broadcasters competed to hire the most attractive stars. The Variety Radio Directory wrote that in 1930, talent expenditures were about 30 percent of the total radio budget, but less than a decade later, the figures were much higher, varying from about 11 percent to 60 percent, though averaging around 40 percent.⁵⁹

Greater pay combined with better data on audiences meant less freedom for performers. The renowned Russian violinist Jascha Heifetz wrote in the late 1930s of the difference between state-funded European broadcasters and those in America:

In Europe an artist selects his material and submits it to the broadcasting station. If he intends to play the “Habañera” and someone else has just played it, he may be asked to substitute another number; but that is the most the station asks of him in the way of alteration. Anything further would be regarded as effrontery. Good manners alone forbid it; for the artist is an expert in his field and it is assumed that he knows his business.

In the United States the artist is shown into a luxurious suite and a contract is handed across the desk for his signature. Then he is asked: “What are you going to play?”

He mentions three or four pieces of music in which for weeks past he has been rigorously drilling himself.

“Oh, no,” he is told. “They won’t do at all.”

“They won’t?” he asks in bewilderment.

“We have to give the public what it wants, you see.”

At this point Heifetz’s hypothetical artist, whom he dubbed Petrov, wonders if the audience wants Petrov, and, if so, why not Petrov’s pieces? He attempted to make the American radio representatives understand.

“Oh, now, you mustn’t feel that way about it, Mr. Petrov,” he is deftly placated. “You see, you don’t know radio. We’re radio experts; we know what the public wants.”

Whereupon Petrov reflects a little wistfully that radio is barely fifteen years old. He has been pleasing the public for twenty, perhaps thirty years. He would not think of telling the technicians in the control booth how to adjust their dials. He would not think of telling the advertising expert at the desk how to prepare a layout. Yet no one has any compunction about telling him how to please the public with his music, the thing in which he has a special and expert proficiency.

Heifetz’s concluding sentence was a strong condemnation of the commodification of the audience and the rationalized ways in which it was calculated in America: “I have played—quite recently—in Italy, Germany and Russia. I had to come to the United States to find a dictatorship.”⁶⁰ Strong words in the late 1930s.

Programs were devised to appeal to what the audience wanted, usually divided demographically, and sponsors were pitched programs based on these demographics. For example, the World Broadcasting System Inc., based in New York City, a large producer of syndicated programs, employed a promotion department to generate documents outlining the programs it produced, documents that are quite revealing of the ways that this company and the networks conceptualized their audiences and the kinds of programming that a particular audience might respond to.

Broadcasters, sponsors, and advertisers worried endlessly (as they still do) over how best to reach an audience, and many of these debates focused on music. Phil Spitalny, who directed the Hour of Charm program, sponsored by General Electric, said in 1938:

Experience showed that the vast majority of our people are music lovers at heart, provided they can be given music which means something to them. . . .

To my mind, the answer is light music: melodic, rhythmic, well played tunes which will satisfy the ear and the emotions, without overtaxing an intellect which has not been trained so that it may grasp the beauties of the greater classics.⁶¹

This is the music of the masses, wrote Warren B. Dygert. But there is also music for the “classes,” music with class appeal for those who know something about music. Dygert listed Ford’s Sunday Evening Hour as a good example, a “radio program designed to select this limited but profitable audience.”⁶² But one couldn’t go too far in presenting the classics, because, he says, “moderism [sic], i.e. extremism, is offensive to many listeners.”⁶³

The Fleischmann’s Yeast Hour

I’ve heard so much of Rudy Vallee that I think he’s more wonderful than Beethoven’s Sonatas.

—Fan letter to the National Broadcasting Company, ca. 1937

Still, many national advertisers, then as now, wanted to attempt to reach as broad an audience as possible. One program that straddled the “classes and the masses” with great and long-lasting success was The Fleischmann’s Yeast Hour, starring crooner Rudy Vallée, and I will spend some time discussing this program as a way of pulling together the various strands of argument in this chapter.

Vallée’s was a variety show, one of the main ways that advertisers sought to cultivate a broad market: “Starting with the premise that radio is a ‘mass’ medium, advertisers have set out in assiduous pursuit of that will-o’-the-wisp, ‘universal appeal,’ ” wrote one commentator in 1933.⁶⁴ And as two observers said in 1935,

Figure 2.1 Rudy Vallée, ca. 1930. (Author’s collection.)

In order that his program may appeal to all classes of people and to all members of the family, the sponsor often tries to include within the same period a considerable variety of entertainment. Instead of turning to one program to listen to a comedian, another to hear a drama, a third to hear jazz, or a fourth to enjoy a symphony, we may turn to the Canyon Tobacco Hour and hear a little of everything. Perhaps no member of the family enjoys the whole hour, but the chances are that each will like a certain fraction of it. The variety program is the broadcaster’s ingenious creation to appeal to the greatest number of people during one period.⁶⁵

In many ways, The Fleischmann’s Yeast Hour was an innovative and ingenious variety program. In a 1931 speech before the League of Advertising Women at the Advertising Club in New York City, Daniel P. Woolley, vice president of Standard Brands, Fleischmann’s parent company, discussed the notion of “tempo” (a common word in radio advertising in this era)—that is, matching a product to a particular kind of program and entertainment—in the period before advertisers had much concrete knowledge of their audiences.⁶⁶ His main example was The Fleischmann’s Yeast Hour. In those days, yeast was sold not simply for baking bread but for health, as a health supplement. So Fleischmann’s yeast was marketed as more of a medicine than a foodstuff in this era, a difficult undertaking since its taste wasn’t particularly appealing. Woolley’s own words on the subject are worth quoting at length:

Yeast for health is a very delicate subject to handle. It has much to do with good health, so when we started to look around for a radio program we said, “What is the audience we have to deal with?”

We decided that, probably, now-a-days they wanted “It” more than anything else. Who had “It” the most of anybody we could find? We found a young crooner, Rudy Vallee, and we found the young ladies panting over him and even some of the old ladies. He also has a great many men admirers. So we engaged Rudy Vallee as the star of this great thing called Health. We wanted him to croon but also we wanted more in the program. We wanted athletics or robust health to play a part. We went through the list—Jack Dempsey and all. Finally we said, “Graham MacNamee [sic] as the noted sports announcer stands for sports!”⁶⁷

(“It” refers to sex appeal, the pronoun having been made famous by its attachment to movie star Clara Bow, the “It” girl.)

Woolley said that the company wanted to make sure it had a well-rounded program and so it should get some “ladies with deep and soulful voices and a soloist.” Dr. R. E. Lee, head of the Fleischmann Health Research Department, was also included on the program to talk about health issues: “Now, I might tell you that that combination of MacNamee [sic] for virility, and Vallee for crooning, Dr. Lee to give the advice of the old family physician, plus a lady who sings, has been a very successful radio program.”⁶⁸ It should be noted here that in this year, crooning was widely decried as an effeminate mode of singing, since the recently invented electrical microphone permitted singers to croon rather than belt, and so balancing Vallée with McNamee was an important consideration for Standard Brands.⁶⁹ Yet, as Roland Marchand has argued, this intimate form of singing became a way that advertisers and advertising agencies understood their relationship to their audiences, for intimacy and personalized address to audience members was a way both to sound modern and to attempt to reassure them in an era of increasing uncertainty and complexity.⁷⁰

So, on 5 September 1929, J. Walter Thompson began producing and airing a program sponsored by Fleischmann’s Yeast, with Vallée as the main attraction (example 2.1).⁷¹ John Reber told his colleagues at a staff meeting on 26 August that the audition of the program went well. An “audition” was a hearing, in Reber’s words, “merely a parading before you of the talent and of the general idea to be developed. We get it for nothing.” Auditions were probably put on for sponsors to see before their program aired. Reber reported that the star was “really marvelous.”⁷²

Vallée did not enjoy much of a spotlight on his program at first. The early broadcasts featured Vallée mainly as bandleader, not as personality. The announcer, the renowned Graham McNamee, did virtually all of the talking, except for Dr. Lee, who told the listeners of the health benefits of ingesting three Fleischmann’s yeast cakes per day. By the program that aired on 14 January 1932, the musical variety show format had started to solidify, reducing the announcer to little more than a commercial spokesman for the yeast; in the past, he had been the main speaker, as on most programs. But Vallée, who previously had only announced the numbers he was playing, began introducing the guests as well, at the behest of J. Walter Thompson. “This kind of personality opportunity became a JWT program characteristic,” according to the brief, undated history of the program written by an anonymous employee of the company.⁷³ By the end of April 1932, Vallée told his listeners that he was both “announcing and directing” Fleischmann’s Yeast Hour. Vallée says in his second autobiography that the show by 1932 had become “a program that was to discover and develop more personalities and stars than any radio show before or since.”⁷⁴

Vallée wasn’t a trouble-free personality himself, however. At the height of his popularity in the summer of 1931, Vallée was the subject of a number of memos at NBC. He was causing consternation for “advertising anything and anybody,” complained John F. Royal, director of programs, changing musical numbers at the last minute, and not having his programs cleared before broadcast, as was the norm.⁷⁵ Despite his employers’ occasional misgivings, Vallée could behave this way because of his massive popularity as the first musical superstar in the new medium of radio—and one who was widely written about.⁷⁶ The Fleischmann’s Yeast Hour remained on the air for over a decade, though as the practice of eating yeast for health waned, the program was rechristened The Royal Gelatin Hour in 1936 (Royal Gelatin was also owned by Standard Brands).⁷⁷

Vallée continued to find work in radio because he was also thought to possess that quality most desired by advertisers—showmanship; William B. Benton, chairman of the board of the Benton and Bowles agency, called him “one of the greatest showmen in radio.”⁷⁸ The audience might have been a commodity, but selling goods was no less important than the commodification of the audience. Vallée’s first autobiography, written not long after he had started broadcasting (and just as his stint on the Fleischmann program was beginning), includes an entire chapter on showmanship. Vallée was more concerned with showmanship in live performance than on radio, though he admitted that showmanship on the radio was extremely difficult since the audience can’t see the performers. He later offers a characterization of showmanship that approximates those promoted by radio writers and producers: “The arrangement of a well-balanced program with every number calculated to please someone in the average mixed audience—this is just as much an act of showmanship as is the presentation of the numbers.”⁷⁹

At the J. Walter Thompson staff meeting on 12 August 1930, George Faulkner, the writer of dramatic sketches for The Fleischmann’s Yeast Hour, gave a presentation on showmanship in radio. He began by asking what a showman was, saying that the term had connotations of someone undignified, cheap; “it has a vaguely Semitic, Barnumish, Broadway air to it.” This kind of showmanship is exploitation, said Faulkner. Showmanship in radio means the technique of show building, the craftsmanship of production. Showmanship in radio compelled people not only to listen but also to pay attention. Faulkner quoted an article from Variety: “One agency has as its radio slogan, ‘while money will put a radio program on the air, only showmanship will put it in the ear.’ ” Faulkner elaborated on this problem:

Radio has become to a much too large extent background music for the American home . . . a background for bridge or the evening paper or the evening meal or even for conversation. If you want an audience really to listen to an advertising appeal you must give that audience something which will catch and hold its attention . . . the audience must listen actively, not passively.⁸⁰

The answer to this problem was “showmanship,” which for the J. Walter Thompson Company in the early 1930s meant variety, unity, pace, punch, and, sometimes, humor.⁸¹ William L. Bird Jr. has written of the conception of showmanship in this era, arguing that the term captured the business community’s show business–ization of business.⁸²

With the arrival of the variety show, advertisers, networks, and advertising agencies had found a way to appeal to the broadest audience possible, their main goal in this era, even as they were devising ever more sophisticated means of ascertaining information about their audiences. Demographic information began to drive the selection of programs, leading to the commodification of the audience as we now recognize it. Advertising agency workers were beginning to understand their power in creating meanings for goods, meanings that could be proffered to consumers in appealing musical packages, inviting them to consume even an unappetizing bar of yeast.