Charles Hamm:
Some Thoughts on the
Measurement of Popularity in Music

Paper delivered at the first conference of IASPM (International Association for the Study of Popular Music), Amsterdam, June 1981.
Published in Popular Music Perspectives, eds. P Tagg & D Horn (Göteborg & Exeter: IASPM, 1982;
ISBN 91-7260-610-X), pp. 3-15. Page turns in original edition are rendered in this text as follows: |3-4|.


Go to top of document*****

We begin this conference under a severe handicap: we're not sure what we're talking about. This statement is not intended as a prior judgement of the individual papers to follow, which I am confident will be uniformly interesting, valuable and informative, but as a commentary on the fact that there is no general agreement on just what is encompassed by the term "popular music".

Some take it to include a wide spectrum of music, including rock 'n' roll, Tin Pan Alley song, disco, Highlife, urban blues, country-western music, punk rock, minstrel songs of the 19th century, country blues, New Wave, oral-tradition Irish song, Western Swing, jazz-rock fusion, sentimental parlour songs of the previous century — and much more.

Others use it in a more restricted way.

This confusion is perhaps best epitomized by the attitude of the present conference toward jazz. Is jazz "popular music"? Is some of it "popular music" and some not? Will some of the participants assume that jazz is "popular music" and others assume that it is not?

This is not just a question of semantics. There are practical considerations, important ones. A new yearbook to be published by the Cambridge University Press, entitled Popular Music , will probably not have the distribution it deserves in the United States because its title does not suggest to many people just what its range of subject matter will be. I suspect that many people did not think to attend the present conference for the same reason, and I predict that some of you in attendance will be disappointed by the range of subject matter of the papers and discussions. A more specific example of the mischief resulting from disagreement on the meaning of "popular music": my recent book, Yesterdays. Popular Song in America (New York: W.W. Norton & Company, 1979) was just reviewed in the journal of the Society for Ethnomusicology — surprising in itself, perhaps — by Charles Keil, author of an excellent book on.the urban blues; unfortunately, his areas of interest and expertise intersected with the subject matter of my book in the most minimal |3-4| way, and the resulting review proved to be an embarrassment to all concerned.

I will not presume to offer a definition of popular music myself, but will rather offer some comments that may help define the problem.

In America, prior to the 20th century, there were two quite distinct bodies of music with some claim on the title of "popular":

  1. composed songs, notated and disseminated in the form of sheet music, aimed at a wide audience of amateur music makers (the songs of Stephen Foster, for instance);
  2. oral-tradition music stemming from various non-literate sub-cultures in the United States: Child ballads; other Anglo-American, Celtic-American and American ballads and songs; most Afro-American music; dance music of both white and black Americans; the music of other ethnic and racial minorities.

Though there was some overlap between the two, there were important differences in both musical style and function in society between the first, usually referred to as "popular song", and the second, which most scholars have thought of as "folk music".

In the first decades of the 20th century, dissemination of the first category by means of the phonograph disc became increasingly important; in the 1920s and '30s, dissemination of this music on the radio and in sound films also became common. MacLuhan would say, I believe, that these new media became extensions of the human voice, extensions of the practice of performing this music on the stage or in the parlour. They were still composed, notated songs, still disseminated most importantly in the form of printed sheet music.

Dissemination of the second type of "popular" music via the phonograph disc and radio did not begin until the 1920s. These two media represented a logical and appropriate extension of the performance means of this music (voices and instruments), which operated within the oral tradition; for the first time, such music was able to reach a mass audience, to reach out beyond private performance for a circle of family, friends and neighbours and public performance in medicine shows and sometimes on the minstrel stage. This first mass dissemination of oral-tradition music via the phonograph disc brought vast proliferation and rapid stylistic development to jazz, the blues, and hillbilly (later country-western) music.

In 1955, with the advent of rock 'n' roll, what had been known as "popular music" to the music industry became — for the first time — an oral music, disseminated almost exclusively in live performance and its extensions: radio, the phonograph disc, and to a more limited extent television and film. Printed sheet music, created after the fact (written down after the music had been conceived and popularized orally) played virtually no role. |4-5|

At just the same time, music of the second category ("folk music") began moving slowly but inexorably in the direction of composed, notated, arranged music, a trend that would soon yield Country Pop and Black Pop.

Distinctions between the two categories eventually became matters of musical style and the socio-economic status of their performers and audiences and not — as had been the case formerly — methods of dissemination. All varieties of music in the two categories became products and users of the several mass media, and I applaud the emphasis on the term "mass media" by the organizers of the present conference as an attempt to head us all in the same general direction.

However, it must also be said that classical music is also disseminated via the mass media in the 20th century (phonograph disc, radio, television, and film), and I sense no attempt to include the music of Beethoven, Vivaldi, Obrecht, or Stravinsky on our agenda. This is not a facetious point. Returning to my opening remarks, I fear we are not in agreement as to whether the "popular music" which is the subject of our conference is a matter of musical style — or a means of dissemination.

Go to top of document*****

Certain bodies of music comprise such a limited number of compositions that a scholar may be familiar with the entire corpus: the 13th-century motet, for instance, or operas by Russian composers written before 1870, or string quartets by American composers of the 19th century. Other genres comprise so many individual pieces as to make it impossible for any single person to examine all of them: the symphony of the 18th century; the Romantic Lied.

Whatever definition one accepts for "popular music", it clearly belongs to the second situation. In studying it, one must devise some method of restriction to limit the sample to a reasonable number of pieces.

An approach which seems logical to me is to let the term define the method: to deal with the pieces which are demonstrably the most popular items of "popular music", with the most widely disseminated items of music disseminated in the mass media.

Rather than developing this notion abstractly and philosophically, I'd like to offer some comments on practical means and problems that one must deal with in proceeding along this line with the mass media in America.

First, the phonograph disc, considered by most rock journalists and cultural historians to be the most important medium for the dissemination of "popular music".

Billboard,1 published weekly, carries a number of popularity charts in each issue, the most important being:|5-6|

Other charts offer rankings of jazz, disco, Latin, "spiritual", "easy listening", and international discs; some of these appear weekly, others more sporadically.

These charts have a profound impact on the mass media. A syndicated weekly radio program, "American Top Forty", plays the first 40 songs on Billboard 's "Hot 100" chart in reverse order; commercial stations all over America, hundreds of them, pick up this program. In addition, other hundreds of radio stations (mostly AM) concentrate their airplay on the "Top 40" repertory all week, again taking the Billboard chart as a guide to their selection of pieces. A syndicated television program, "Solid Gold", offers the "Top 10" songs from the Billboard chart every week. More generally, both the music industry and most people who write about popular music (journalists, critics, historians) give complete credence to these charts; placing a song at or near the top of Billboard 's "Hot 100" list (or one of the other major charts) gives a tremendous boost to a performer, and generates additional sales of the disc in question; and many writers on popular music base their arguments and conclusions on Billboard rankings.

An enterprising organization in Wisconsin, Record Research Inc., has compiled and published a series of books and pamphlets in which the information from the various Billboard charts, over the years, has been organized and summarized.2 Armed with these volumes, the scholar or the merely curious can quickly locate details concerning every piece that has appeared on any of the Billboard charts since their inception.

One can determine from a recent issue of Billboard , the one brought out on 23 May 1981, that "Bette Davis Eyes", as sung by Kim Carnes, was the top item on the "Hot 100" charts; "A Woman Needs Love", performed by Ray Parker Jr. & Raydio, was the #1 item on the "Soul" charts; the album High Infidelity by the rock group REO Speedwagon was at the top of the album chart; and so on. And by resorting to Whitburn's several compilations, it is equally simple to find out that Little Richard's "Good Golly, Miss Molly" made its first appearance on the "Hot 100" charts on 8 February 1958, that it stayed on the charts for 15 weeks, and its highest position was #10; the same song first appeared on the "Rhythm & Blues" chart on 15 February of that year, persisted for only 8 weeks, and reached a top of #6.

The easy availability of such data has vastly simplified the task of the critic or |6-7| historian of popular music. But it is disconcerting to note that these charts, which have become such a powerful force in the world of music, offer no supporting statistics for their rankings. The "Hot 100" chart carries only the simple explanation (or disclaimer?) that it is "a reflection of National Sales and programming activity by selected dealers, one-stops and radio stations as compiled by the Charts Dept. of Billboard ", and the other charts are headed by the statement that they have been "compiled from national retail stores by the Music Popularity Chart Dept. and the Record Market Research Dept. of Billboard . The only hard information offered is the identification of those discs which have "gone gold" (i.e., sold a million copies, for singles, or 500,000 for LPs) or "gone platinum" (sales of two million, for singles, or one million for LPs). It is impossible to avoid the impression that Billboard 's ranking techniques are not refined enough to differentiate between sales and radio play time of, say, the song ranked #5 for a given week and that in the 10th position — or, for that matter, that the various Billboard charts function to generate sales and air play as much as to measure it.

Go to top of document*****

Radio is a quite different matter.

Virtually every American community of more than a few thousand inhabitants has at least one commercial radio station. Many of these are totally independent, others are affiliated with one or another of the 10 national radio networks. Even the latter pick up only a small fraction of their programming from nationally syndicated sources. A typical American radio station broadcasts mostly recorded music, selected by staff members, interspersed with brief news spots and weather reports, with syndicated programs from time to time. Thus the programming for each of the thousands of stations in the United States is unique, and there is no way for anyone to determine with any accuracy how many times a given piece of music is broadcast on any given day, week, or any other period of time to the approximately 477,800,000 radio receiving sets in the country.3

However, approximations of what kinds of music are sent out over American airwaves are possible, and the size of the audience for each kind. Almost all American radio stations base their programming on a certain type of music — Top Forty pop songs, harder rock, music aimed at black listeners, country-western music, whatever. Listeners tend to identify with one or more stations favouring the type(s) of music they prefer, and to stay with these stations, once identified, rather than constantly redialling in search of specific songs or types of music.

Measurement of radio audiences is done by several commercial concerns, most importantly the Arbitron Company. Their technique is to distribute diaries to |7-8| randomly-selected households, with the request that such information as what radio stations are listened to, at what hours of the day, by which members of the family, for how long, be entered in these diaries. Arbitron concerns itself only with persons in this group; 95.2% of them listen to the radio at least once a week. For purposes of testing and rating, Arbitron has divided the country into some 250 market survey areas, each centred around the largest city or town in that region, the idea being that every listener in a given area has access to the same cluster of broadcasting stations. Arbitron releases statistics of audience size and composition, for each market survey area, from one to four times each year; these figures are supplied to all radio stations which subscribe to Arbitron's services, and to commercial firms which advertise over the air or may have an interest in doing so, to inform them of how many people they might expect to reach if they advertise on a given station. Since Arbitron is a service concern, and its service consists of statistics issued to its customers and subscribers, it does not make its finding public. Radio stations may release partial or full figures concerning the distribution of the listening audience once they have received these from Arbitron, if they choose; otherwise, information is available only from Arbitron, which has expressed willingness to cooperate with scholarly studies dealing with the mass media.4

While it is impossible to determine statistics on the nation-wide air play of individual pieces of music, it would be possible to study the dissemination of various types of music within any or all of the market survey areas established by Arbitron. This would involve establishing the musical profile of each radio station within an area (what type of music it played), obtaining audience statistics from Arbitron, converting these into numbers of listeners, and thus drawing a profile of what mu. sic goes over the air in that area, in what proportions.

For instance, the market survey area centred in Boston encompassed a population (over 12 years of age) of 2,922,100 in early 1981. There were 30 radio stations on the air in the first quarter of that year — 17 AM, 13 FM. The following table lists the 10 stations with the largest audience, according to Arbitron's statistics for March of 1981; the percentage of the listening public which preferred each station is given, with this figure translated into numbers of listeners; each of these most successful stations is further identified as to the type of music programming it stresses.

WCOZ—FM 11.1% (330,000) hard rock-heavy metal album play
WHDH—AM 10.3% (300,000) "adult contemporary"
WBZ—AM 8.4% (250,000) Top Forty (based on Billboard)
WEEI—AM 6.5% (200,000) all-talk: news, weather, sports
WJIB—FM 5.3% (150,000) "easy listening"
WXKS—FM 5.0% (150,000) disco and "dance-rock" |8-9|
WBCN-FM 4.8% (140,000) New Wave, rock
WROR-FM 45% (130,000) new and old "rock standards"
WEEI—FM 3.7% (120,000) soft rock and jazz
WVBF—FM 3.5% (105,000) "mainstream rock"

The profile reveals a heavy emphasis on the several varieties of rock music, from old "standards" through Top Forty music all the way to New Wave and Heavy Metal bands; "easy listening" styles are a distant second; none of the 10 top stations offer programming aimed at black or Hispanic listeners, nor do any of them program country-western music. This picture matches the socioeconomic character of Boston, of course. The city's multitude of institutions of higher education (Harvard University, Massachusetts Institute of Technology, Boston University, Boston College, Northeastern University, Tufts University, the New England Conservatory of Music, the Berklee College of Music, and many other smaller schools) give it the highest ratio of college-age persons of any American city and account for the high proportion of youth-orientated music offered on the radio; the Boston market survey area encompasses only urban and suburban areas, with no history whatever of large-scale immigration from the South, Midwest or Southwest, and hence virtually no market for country-western music; and though there are sizable black and Hispanic populations, Boston remains one of the most racist spots in the United States, these minority groups play almost no role in the cultural and political life of the city, and are largely disregarded in radio programming.

The problems faced in studying the dissemination of music by radio in the United States are immense. One would have to proceed by examining each of the more than 250 market survey areas in ways suggested by the above remarks: determining the musical profile of each radio station, obtaining statistics on each area from Arbitron, understanding the ethnic and cultural mix found within each area. But such studies must be undertaken if one is to obtain a comprehensive grasp of the mass dissemination of music in America. Surely more people, many more, are exposed to music on the radio than by listening to private phonograph equipment. Surely Billboard's charts give an inaccurate picture of the types of music actually heard in such areas as New York City (with its unique mix of black, Hispanic, Jewish, Irish, and professional and highly-educated population), or Miami (with its curious blend of retired, elderly people, its Cuban refugees, its working class drawn partly from the rural South), or Dallas, or Los Angeles, or Chicago. It would be a difficult and complex job to consider all these factors. But it could be done. |9-10|

Go to top of document*****

American television is yet another different proposition.

All commercial TV stations offer programming composed of three elements:

  1. Networks shows carried during prime time (several hours in the morning; early afternoon; the evening, from 7 or 8 p.m. until 11; some weekend slots), picked up from whichever of the three national commercial networks — ABC, CBS, NBC — the local station is affiliated with. Thus in television, unlike radio, audiences in all parts of the country are exposed to exactly the same material at the same time.
  2. Syndicated programs (including news shows, movies, and re-runs of older television programs) picked up by local stations at their discretion, and shown in time slots chosen by each station. Much of this material is seen in various parts of the United States during a given week, but not necessarily at the same time.
  3. Local programming, consisting mostly of regional newscasts and a scattering of locally-produced shows concerned with civic and community matters. These are seen only in the areas in which they are produced.
Of the several firms which make statistical measurement of the size and constitution of the audience for television programs, the oldest and most prestigious is the A.C. Nielsen Company, in operation since 1952 and so important to American television programming that the term 'Nielsen rating' has crept into popular usage. Their methodology relies on:
  1. The Sample : statistical determination of how many American households must be tested in order to yield reliable figures for their ratings.
  2. The Storage Instantaneous Audimeter (SIA): an electronic meter installed in the necessary number of homes, connected to the household television set(s), which automatically registers which television programs are tuned in, for how long. All SIAs are connected by telephone line to a central computer, which may be used to retrieve information at any time.
  3. The TV Diary : each household fitted with an SIA is also given a diary, with the request that individuals watching any program be identified by age and sex.

The SIA and the TV diaries, taken together, yield information not only on the total television audience, but also on details of the make-up of that audience. And from calculations based on the geographical areas in which selected households are located and the socio-economic-ethnic status of participating families, the Nielsen Company can determine with some precision which age groups and which economic-cultural and ethnic groups are most responsive to various programs. |10-11|

The results of all this are published every two weeks, as the Nielsen National TV Ratings , a detailed breakdown of the size and constitution of the audience for every network program telecast. This publication is not available to the general public; it is sent only to paid subscribers — the television networks and stations, and commercial concerns which advertise on network shows or may be considering doing so. General information is released to the press: periodic reports on the relative rankings of the three networks, information on audience size of certain programs. This publication has become the 'Bible' of the commercial television industry; shows are dropped or shifted to new time slots if their Nielsen ratings are low, new shows for each television season (starting in the early fall) are planned largely on the basis of which types of programs attracted the largest audiences the previous season; directors and actors are often let go — or given more important new assignments — on the basis of the success or failure of the shows with which they have been involved.

The Nielsen Company offers other services as well, including ratings of syndicated programs, more refined breakdowns of audiences for various shows and for a price quicker reports on audience sizes, for television networks and advertisers who want more immediate reports on the number of people viewing specific programs.

There are an estimated 142,000,000 television sets in the United States.5 More importantly, the Nielsen Company places "television penetration" into the country's some 75,000,000 households at 98%, with little fluctuation by geographical area or economic status. Even in such regions as the rural South and the worst urban slum areas, the percentage of Americans reached by television almost never drops below 95. Putting it another way, almost 200,000,000 Americans reside in households with at least one television set. Thus far more people are reached by television than by any other mass medium; and given the saturation of every element of the population by television, surely network television programs are presently the most important common denominator of American life and culture. It is absolutely urgent that anyone wishing to understand the impact of mass-disseminated music on our life pay attention to commercial television's attitudes towards music.

In order to get some idea of this factor, I watched every network television program for the week of 17—23 May, 1981, made up mostly of music or including substantial amounts of music. I noted types of music, and specific pieces, on each program. Mr. Bob Bregenzer of the Nielsen Company, with whom I talked about my project, kindly gave me the Nielsen ratings of each show by telephone, before they were available in print.

Now, as we used to say in American slang, let me lay some figures on you.

Other programs this week devoted largely or entirely to music were syndicated — not network — programs, and thus attracted smaller audiences; the tabulation of their share of the audience is much more complicated, and the results are not available until 6 weeks later, hence the Nielsen Company was not able to furnish me with figures in time for this presentation. These syndicated shows included NBC's "Solid Gold", hosted by Dionne Warwick and concerned mostly with heavily orchestrated versions of some portion of the current Top Forty repertory; CBS's "Hee Haw", a mixture of Country Pop and comedy; and the "Lawrence Welk Show", with its distinctive brand of middle-of-the-road classics drawn from the popular repertory of the past century or so, aimed squarely at elderly and middle-class viewers. Interestingly, in the area in which I watched TV for this week (eastern Vermont), these three shows were scheduled in direct competition with one another, at 7 p.m. on 23 May.

To put some of this in perspective: during this week, when Kim Carnes's "Bette Davis Eyes" reached the #1 position on Billboard 's "Hot 100" charts and was thus regarded as the most popular song in America by journalists and critics who take the sale of phonograph discs as the most important measure of popularity, this recording had not yet sold a million copies — while some 45,000,000 Americans were exposed to a number of songs performed by Donnie Osmond which were nowhere to be seen on any of the Billboard charts; and the REO Speedwagon LP High Infidelity , at the top of the album charts, had just passed the million mark in sales —but as many as 30,000,000 people heard "Oh, What a Beautiful Morning" from Oklahoma , a song written in 1943.

If we are to deal with the dissemination of music by the mass media, we must deal with all media. The above remarks have been offered as nothing more than suggestions as to how one can begin to proceed along these lines, what resources are available, what problems are encountered, and what conclusions might emerge from such a study. Needless to say, what I have done is merely suggestive, a first small step in the direction of determining what sorts of music are heard by the largest |12-13| number of people in the United States. And I would be the first to point out that comparisons between number of phonograph discs sold and the number of people hearing a given song as performed on television approach a comparison of apples and oranges; once a disc has been purchased, there is no way to determine how many times it will be played, for how many people, in a given period of time. But one must start somewhere.

Go to top of document*****

My remarks have been wide-ranging, as I hoped would be appropriate for an opening paper. Let me end by touching on another dimension of popular music, perhaps the most important one.

Mass-disseminated music has enormous potential for influencing the ways in which a population perceives and thinks about the world. This potential has been realized in the past for both commercial and political gain, and will be again.

If one depends on data concerning the dissemination of phonograph discs alone, one must conclude that the United States is still in the Age of Rock. The several varieties of rock music have been associated since the mid-1950s with rebellious youth and radical politics; surely this music played a critical role in helping to reverse American policy on Vietnam a decade ago, and in creating a climate which made it impossible for Richard Nixon to continue in the presidency.

If one considers the dissemination of music via other media, however, a quite different picture emerges.

One would not suspect that America is living through the rock era from sitting in front of a television set. The vast majority of music emanating from this medium is country-western music (mostly Country Pop) and music from the Tin Pan Alley era or newer music in this same style — styles associated with older and more conservative layers of American society.

There is not time here to consider whether television's neglect of rock is an honest reflection of current American taste, or whether television management has consciously and deliberately allied itself with those forces in America determined to resist and ignore the Age of Rock, the counter-culture which emerged in the 1960s, and all that these trends imply.

My point is this: if one had been truly attentive in recent years to trends in the mass dissemination of music in America via television, the medium which reaches by far the largest number of people in the country, and had refused to be blinded by the much greater attention paid to the products of a more limited medium — the phonograph — one could easily have predicted the outcome of last fall's presidential election and anticipated other recent events in the United States signalling a massive swing to the right, politically and socially. |14-15|

The mass media can tell us a great deal about the world in which we live, and the ways in which attitudes and opinions are being formed and moulded. But they can tell us such things only if we learn to look and listen in a truly objective way.

Go to top of documentEndnotes

1. Billboard. The International Music-Record-Tape Newsweekly (Los Angeles: Billboard Publications, 1894 — present).

2. Joel Whitburn. Top Pop Artists & Singles, 1955-1978. (Menomonee Falls: Record Research Inc., 1979). Whitburn has also compiled and published similar compilations of other popular repertoires: country-western music; rhythm & blues, and soul; LP albums; "easy listening" music; etc.

3. This figure was taken from the 35th edition of World Radio TV Handbook, edited by J.M. Frost (London: Billboard Limited, 1981).

4. The information concerning Arbitron was kindly and generously supplied to me by Mr. Jay Billie of the Washington office of Arbitron Radio, located in Laurel, Maryland.

5. According to World Radio TV Handbook.

Go to top of document