Measuring Industry Concentration, Diversity, and Innovation in Popular Music Author(s): Richard A. Peterson and David G.

Berger Reviewed work(s): Source: American Sociological Review, Vol. 61, No. 1 (Feb., 1996), pp. 175-178 Published by: American Sociological Association Stable URL: http://www.jstor.org/stable/2096413 . Accessed: 16/10/2012 09:50
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

American Sociological Association is collaborating with JSTOR to digitize, preserve and extend access to American Sociological Review.

http://www.jstor.org

COMMENTS AND REPLIES

175

MEASURING INDUSTRY CONCENTRATION, DIVERSITY, AND INNOVATION IN

Richard A. Peterson
VanderbiltUniversity

David G. Berger
TempleUniversity

VY nity to reflect on some important issues raised by researchers concerning our ASR article "Cycles in Symbol Production: The Case of Popular Music" (Peterson and Berger 1975). Contrary to the thinking then prevalent in economics, that innovation is most likely in oligopolistic industries (Schumpeter 1950), we found that the greater the competition in the popular music industry in a giyen year, the greaterthe innovation in the music. Analyses by Alexander (1996, henceforwardAlexander) and other researchers prompt us to comment on the measure of concentration/competition, the measure of musical diversity, the difference between diversity and innovation, and the likely range of the positive relationship between industry competition and product innovation.

XX.,Fe are delighted to have this opportu-

(Carroll 1985) that measures the average market share of all firms (Rothenbuhler and Dimmick 1982; Burnett and Weber 1989; Burnett 1990; Dowd 1992; Lopes 1992; Schulze 1994; Christianen 1995; Alexander 1996). These ratios are accurate measures of the concentration of musical product ownership, but, as we and others show, they are no longer good measures of the concentration of creative control (Lopes 1992; Anand and Peterson 1995; Dowd 1995). This is because the major firms, like Time/Warner,now have autonomous competing divisions (Time/ Warnercurrently has three), which, in turn, release popular music on numerous labels (Time/ Warnercurrently has 62).1 Therefore, a more accurate index of creative control should be based not on the number of financially independent corporations but on the number of creatively independent divisions or labels that successfully compete in the market. MUSICAL DIVERSITY

Like Dowd (1992, 1995), Alexander uses musical characteristics of hit songs to measure diversity. Unlike Dowd, however, Alexandertakes his five measures from sheet music rather than from the hit record itself, arguing that sheet music is a blueprint for what is recorded. This may be an adequate assumption for the era up to 1955 (which MEASURINGCONCENTRATION Alexander does not study). At that time, To measure the degree of control of the mu- bands generally recorded songs, reading sic industry held by a few firms we used the from printed music charts. In the rock era, conventional four-firm concentration ratio, however, pop music increasingly has been which gives the proportion of the market composed while being recorded. If sheet mucontrolled by the top four firms. The concen- sic is published at all, it is produced after the tration ratios depicted in Alexander's Figure song has become a hit (Tagg 1982). 1 clearly show the over-time dynamic we Musicologists of popular music advise projected in 1975. Studies of concentration that any written transcription is an inin the music field continue to use either this adequate representation of a recording measure, or a more sophisticated derivative (Winkler forthcoming) and of the nuances that made it a hit (Keil and Feld 1994). In * Direct all correspondence to Richard A. any case, the commercially available sheet Peterson, Department of Sociology, Box 1635 music of the sort used by Alexander does
Station B, Vanderbilt University, Nashville, TN 37235 (Internet:[email protected]). We are grateful for help and advice from Timothy Dowd, Eric Fine, Larry Griffin, Roger Kern, Barbara Kilbourne, Holly McCammon, Claire Peterson, David Sanjek, Darren Sherkat, and Peter Winkler. Through the mid-1970s the correlation between the number of firms with records in the Top 10 of the weekly Billboard chart and the number of labels reaching the Top 10 is nearly perfect. See Peterson and Berger (1975).
I

176

AMERICAN SOCIOLOGICAL REVIEW

not faithfully represent what is recorded. His claiming that it represents the hit recording is analogous to saying that the Mona Lisa printed on a T-shirt accurately represents the painting on display in the Louvre. Sheet music is a simplified version of the hit, made to be sold to amateur pianists or guitarists and to club-date professionals who are expected to play the latest pop hits at weddings and bar mitzvahs. Consider for example, time and meter, the most objective of Alexander's five diversity variables. From the sheet music, he codes each song as 4/4 or 2/2 versus anything else, but this poorly represents the hit record. Dowd (personal communication) reports that in his sample of number-one songs of the 19551988 period, fully 20 percent of the 105 songs shift meter and do so for an average of 11 percent of the duration of the recording. Thus, measures of the characteristics of music must be generated directly from the hit recording, as Dowd (1992) has done. Moreover, an index of diversity should also include measures for song lyrics because lyrics are vital in making a record into a hit (Frith 1987).

DIVERSITY DOESN'T EQUAL INNOVATIONS

Alexander faults us for using the number of records that in a year reach the Top 10 of the weekly pop Billboard chart, ignoring completely the numerous independent lines of evidence we developed to show that this simple measure of diversity did in fact fairly represent innovation in the period we examined. This measure, Peterson (1994) argues, cannot be used uncritically now because a rapid turnoverof songs "may no longer indicate aesthetic innovation but ratheraesthetic exhaustion, as trivially different songs quickly reach the top of the charts-and as quickly fade because they are derivative" (Peterson 1994:176). There must be diversity if there is innovation, but the opposite is not true, because there can be great diversity with no innovation. For example Christianen (1995) describes a number of streams of music in the Dutch music market that have maintained their distinctiveness over decades. This is a situation of considerable diversity in which the innovativeness may be slight. Nontrivial innovation in music is generally signaled by the wide use of a new name for a style of A SUMMARY MEASURE OF music and an associated group of performDIVERSITY ers. Possible examples of innovation include We applaud Alexander for combining the rave, techno, acid jazz, grunge, rap, house, several measures of songs in a single index New Age, disco, funk, punk, acid rock, of diversity.2 The particular measure that he Motown, big beat, folk-rock, soul, rockacalls entropy does not, however, take into ac- billy, do-wop, bop, torch, swing, etc. In focount the degree of difference between cusing on diversity alone, Alexander, among songs. Thus, for example, if half the songs others, has forgotten that the master question fit in one cell and are like the other half of deriving from Schumpeter (1950) is the relathe sample except for a single characteristic, tionship between industry concentration and the same difference score is obtained as innovation. when the two homogeneous halves are different from each other in every single charLINKING CONCENTRATION AND acteristic! Since Alexander neither provides INNOVATION annual plots of the distributionof songs, nor computes a measure of the distances of the Given our concerns about Alexander's measongs from each other, it is not possible to sures of concentration and diversity, we will estimate the actual diversity of songs from not comment on his testing of their relationwhat he calls entropy.Network analysts have ship. Suffice it to say that we are heartened developed just such measures, and so has that he finds that there is still a strong linear relationship between our measure of diverDowd (1995). sity-as-innovation and music industry concentration as seen in Alexander (1996) col2 It would be useful to know the degree to which the individual items are correlated and how umn 5 of Table 1. The failure of the nonlinear model shown in column 6 adds weight to much each contributes to the diversity index.

COMMENTS AND REPLIES

177

University, Nashville, TN. Unpublished manuour assertion that the relationship between script. concentration and innovation is linear. Burnett, Robert. 1990. "Concentrationand DiverFuture studies that use regression models sity in the International Phonogram Industry." to test the relationshipbetween concentration Ph.D. dissertation, Department of Journalism and innovation (or diversity) will need to be and Mass Communication, University of sensitive to the assumptions underlying the Gothenburg, Sweden. methods of time-ordered analysis of histori- Burnett, Robert and Robert Philip Weber. 1989. "Concentration and Diversity in the Popular cal processes. Burnett and Weber (1989), Music Industry: 1948-1986." Paper presented Peterson (1990), Lopes (1992), and Anand at the annual meeting of the American Socioand Peterson (1995) show that the structure logical Association, August, San Francisco, of the music industry has changed several CA. times since 1948, and as Isaac and Griffin Carroll, Glenn R. 1985. "Concentrationand Spe(1989) suggest, such change necessitates the cialization: Dynamics of Niche Width in Popuhistoricization of quantitativemethodology. lations of Organizations."American Journal of Sociology 90:1262-83. Christianen, Michael. 1995. "Cycles in Symbol PROBABLE LIMITS OF THE Production? A New Model to Explain Concentration, Diversity, and Innovation in the Music GENERALIZATION Industry."Popular Music 14:55-93. Looking beyond popularmusic, a positive re- Dowd, Timothy J. 1992. "The Musical Structure lationship between competition and innovaand Social Context of Number One Songs: An tion has been found in a wide range of fields, Exploratory Analysis." Pp. 130-57 in Vocabularies of Public Life, edited by R. Wuthnow. but this does not mean that the relationship London, England: Routledge. should apply universally. We believe it is . 1995. "The Song Remains the Same? The most likely to hold in regulation-free market Musical Diversity and Industry Context of situations where demand is elastic, barriers Number One Songs: 1955-1990." Department to entry are low, and research and developof Sociology, Emory University: Atlanta, GA. ment costs are not high. Unpublished manuscript. Frith, Simon. 1987. "Why Do Songs Have Richard A. Peterson is Professor of Sociology at Words?"Pp. 77-106 in Lost in the Music: CulVanderbilt University. With Narasimhan Anand, ture, Style, and the Musical Event, Sociological he is researching the role of new forms of Review Monograph 34. London, England. information in restructuring the commercial mu- Isaac, Larry and Larry Griffin. 1989. sic field. With Roger Kern and others, he is ex"Ahistoricism in Time-Series Analysis of Hisploring the changing stratification of taste with torical Process." American Sociological Rethe emergence of omnivores and lower status view 54:873-90. univore taste groups. The next installment in this Keil, Charles and Steven Feld. 1994. Music set (forthcoming in ASR)deals with the displaceGrooves. Chicago, IL: University of Chicago ment of highbrow snobs by omnivores. In addiPress. tion, he is completing a monograph on the fabri- Lopes, Paul D. 1992. "Innovation and Diversity cation of authentic country music in the second in the Popular Music Industry: 1969 to 1990." quarter of the twentieth century. American Sociological Review 57:46-71. Peterson, Richard A. 1990. "Why 1955? ExplainDavid G. Berger is Associate Professor of Sociing the Advent of Rock Music." Popular Muology at Temple University. He is currently resic 9:97-116. searching and shooting a film documentary on . 1994. "CulturalStudies Through the Prothe life and works of Jazz bassist and photograduction Perspective." Pp. 163-89 in The Socipher, Milt Hinton. ology of Culture, edited by D. Crane. Oxford, England: Blackwell. Richard A. and David G. Berger. 1975. Peterson, REFERENCES "Cycles in Symbol Production: The Case of Popular Music." American Sociological Review Alexander, Peter J. 1996. "Entropy and Popular 40:158-73. Culture: Product Diversity in the Popular Music Recording Industry." American Sociologi- Rothenbuhler, Eric and John Dimmick. 1982. "Popular Music: Concentration and Diversity cal Review 61:171 -74. in the Industry, 1974-1980." Journal of Anand, Narasimhan and Richard A. Peterson. 1995. "When Market Information Constitutes Communication32:143-49. Fields: The Music Industry Case." Owen Schulze, Rolf. 1994. "Hit Record Trends on the German Music Market for Popular Music Graduate School of Management, Vanderbilt

178 1975-1993" Paper presented at the 8th International Congress on Cultural Economics, August 24-27, Witten, Germany. Schumpeter, Joseph A. 1950. Capitalism, Socialism, and Democracy. New York: Harper and Row. Tagg, Philip. 1982. "Analyzing Popular Music:

AMERICAN SOCIOLOGICAL REVIEW Theory, Method, and Practice." Popular Music 2:37-68. Winkler, Peter. Forthcoming. "Writing Ghost Notes: The Poetics and Politics of Transcription." In State of the Art: Refiguring Music Studies edited by D. Scharz and A. Hassabian. Richmond, VA: University of Virginia Press.