(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Turning on and tuning in : public participation in the arts via media in the United States"

URNiNG On and Tuning In 

Ledia Participation in the Arts 



[arles M. Gray 

search Division Report #33 






looo o ooooooo 3 : 
laaooooaaaoac: 
aaaoaoaaoeaac 



*3\ 



% *■ 




-a 



v 



1 



t 



i 



'National Endowment for the Arts 



Turning On and Tuning In 

Public Participation in the Arts via 
Media in the United States 



Turning On and Tuning In 

Public Participation in the Arts via 
Media in the United States 



Charles M. Gray 



Research Division Report #33 



National Endowment for the Arts 
Seven Locks Press 
Carson, California 



Turning On and Tuning In: Public Participation in the Arts via Media in the United 
States is #33 in a series on matters of interest to the arts community commissioned by 
the Research Division of the National Endowment for the Arts. 

Cover: Inside a mobile television truck during a Live from Lincoln Center performance, 
"Luciano Pavarotti in Concert," transmitted nationally via the Public Broadcasting 
Service (PBS). Photo by Susanne Faulkner Stevens. 

First printed 1995 

Library of Congress Cataloging-in-Publication Data 
Gray, Charles M. (Charles Melvin), 1944- 

Turning on and tuning in : public participation in the arts via media in the 
United States / Charles M. Gray. 

p. cm. — (Research Division report : 33) 
Includes bibliographical references. 
ISBN 0-929765-39-7 (paperback) 

1. Arts audiences — United States. 2. Arts surveys — United States. I. Title. 
II. Series: Research Division report (National Endowment for the Arts. 
Research Division) ; 33. 
NX220.G73 1995 

700'. 1 '030973— dc20 95-34806 

CIP 

Manufactured in the United States of America 

Seven Locks Press 
Carson, California 
1-800-354-5348 



Table of Contents 



List of Tables 
List of Figures 
Preface 
Executive Summary 

PART I: Overview, Background, and Hypotheses 

Media Arts as Economic Goods 

"Supply" Factors 

Factors Influencing Arts Participation: Hypotheses 

PART II: Participation Patterns and Changes, 1 982-1 992 

Overview of Public Participation, 1992 
Participation Patterns 
Summary and Conclusions 

PART III: Multivariate Statistical Results 

Logistic Models 

Overall Logistic Regression Results 

Jazz 

Classical Music 

Opera 

Musicals 

Theater (Plays) 

Dance 

Art 

Summary 

PART IV: The Cross Effects of Media and Live Participation 

Jazz 

Classical Music 

Opera 

Musicals 

Theater (Plays) 

Ballet/Dance 

Art 

Summary and Conclusions 



vn 

x 

xi 

1 

5 
6 

7 
10 

15 
15 
15 

47 

48 

49 
50 
52 
55 
57 
60 
62 
64 
66 
67 

70 

71 
72 
73 
74 
75 
76 
77 
78 



vi I Turning On and Tuning In 



PART V: Summary and Conclusions 79 

Major Findings 79 

Policy Implications 80 

Future Research 80 

APPENDIX A: 1 992 Survey of Public Participation in the Arts 82 

APPENDIX B: Technical Discussion of Logistic Regression 89 

Notes 91 

Bibliography 94 

About the Author 95 

Other Reports on the 1 992 SPPA 96 



Table of Contents I vii 



TABLES 

Table 1 . Public Television Programming (Percentages of 

Broadcast Hours) 8 

Table 2. Number of U.S. Radio Stations Offering Various 

Formats, 1992 9 

Table 3. Manufacturers' Shipments of Recorded Music 

(Millions of Units) 9 

Table 4. Sales of Recorded Music by Genre (Percentage of 

U.S. Dollar Value) 10 

Table 5. Live Arts Participation Rates by Place, 1992 

(Percentages of Adult Population) 1 1 

Table 6. Arts Participation Rates by Income Group, 1992 

(Percentages of Adult Population) 13 

Table 7. Rates of Public Participation in the Arts via Media, 1992 

(Percentages of Adult Population) 16 

Table 8. Jazz Participation Rates by Variable, 1982 and 1992 

(Percentages of Adult Population) 17 

Table 9. Change in Jazz Participation Rates by Age, 1982-1992 18 

Table 10. Change in Jazz Participation Rates by Education, 

1982-1992 19 

Table 1 1 . Change in Jazz Participation Rates by Race, 

1982-1992 20 

Table 12. Classical Music Participation Rates by Variable, 

1982 and 1992 (Percentages of Adult Population) 21 

Table 13. Change in Classical Music Participation Rates by Age, 

1982-1992 22 

Table 14. Change in Classical Music Participation Rates by Education, 

1982-1992 23 

Table 15. Change in Classical Music Participation Rates by Race, 

1982-1992 24 

Table 16. Opera Participation Rates by Variable, 1982 and 1992 

(Percentages of Adult Population) 26 

Table 17. Change in Opera Participation Rates by Age, 1982-1992 27 

Table 18. Change in Opera Participation Rates by Education, 

1982-1992 28 

Table 19. Change in Opera Participation Rates by Race, 1982-1992 29 
Table 20. Musical Participation Rates by Variable, 1982 and 1992 

(Percentages of Adult Population) 30 

Table 21. Change in Musical Participation Rates by Age, 1982-1992 31 



viii I Turning On and Tuning In 



Table 22. Change in Musical Participation Rates by Education, 

1982-1992 32 
Table 23. Change in Musical Participation Rates by Race, 1982-1992 33 
Table 24. Theater (Plays) Participation Rates by Variable, 1982 

and 1992 (Percentages of Adult Population) 34 
Table 25. Change in Theater (Plays) Participation Rates by Age, 

1982-1992 35 
Table 26. Change in Theater (Plays) Participation Rates by 

Education, 1982-1992 36 
Table 27. Change in Theater (Plays) Participation Rates by Race, 

1982-1992 37 
Table 28. Ballet/Dance Participation Rates by Variable, 

1982 and 1992 (Percentages of Adult Population) 39 
Table 29. Change in Ballet/Dance Participation Rates by Age, 

1982-1992 40 
Table 30. Change in Ballet/Dance Participation Rates by Education, 

1982-1992 41 
Table 3 1 . Change in Ballet/Dance Participation Rates by Race, 

1982-1992 42 
Table 32. Art Participation Rates by Variable, 1982 and 1992 

(Percentages of Adult Population) 43 

Table 33. Change in Art Participation Rates by Age, 1982-1992 44 
Table 34. Change in Art Participation Rates by Education, 

1982-1992 45 

Table 35. Change in Art Participation Rates by Race, 1982-1992 46 

Table 36. Variables, Definitions, and References 50 

Table 37. Logistic Regression, Jazz Participation, 1982 52 

Table 38. Logistic Regression, Jazz Participation, 1992 53 

Table 39. Logistic Regression, Classical Music Participation, 1982 55 

Table 40. Logistic Regression, Classical Music Participation, 1 992 56 

Table 41. Logistic Regression, Opera Participation, 1982 58 

Table 42. Logistic Regression, Opera Participation, 1992 59 

Table 43. Logistic Regression, Musical Participation, 1982 60 

Table 44. Logistic Regression, Musical Participation, 1992 61 

Table 45. Logistic Regression, Theater (Plays) Participation, 1982 63 

Table 46. Logistic Regression, Theater (Plays) Participation, 1992 64 

Table 47. Logistic Regression, Ballet Participation, 1982 65 

Table 48. Logistic Regression, Dance Participation, 1992 66 

Table 49. Logistic Regression, Art Participation, 1982 67 

Table 50. Logistic Regression, Art Participation, 1992 68 

Table 5 1 . Hypothetical Table Entries 70 



Table of Contents I ix 



Table 52. Jazz Cross-Participation, 1982 71 

Table 53. Jazz Cross-Participation, 1992 72 

Table 54. Classical Music Cross-Participation, 1982 72 

Table 55. Classical Music Cross-Participation, 1992 73 

Table 56. Opera Cross-Participation, 1982 73 

Table 57. Opera Cross-Participation, 1992 74 

Table 58. Musical Cross-Participation, 1982 75 

Table 59. Musical Cross-Participation, 1992 75 

Table 60. Theater (Plays) Cross-Participation, 1982 76 

Table 61. Theater (Plays) Cross-Participation, 1992 76 

Table 62. Ballet Cross-Participation, 1982 77 

Table 63. Dance Cross-Participation, 1992 77 

Table 64. Art Cross-Participation, 1982 77 

Table 65. Art Cross-Participation, 1992 77 



x I Turning On and Tuning In 



FIGURES 



Figure 1. Jazz Participation Rates by Age, 1982 and 1992 (Percent) 18 

Figure 2. Jazz Participation Rates by Education, 1982 and 1992 

(Percent) 19 

Figure 3. Jazz Participation Rates by Race, 1982 and 1992 (Percent) 20 

Figure 4. Classical Music Participation Rates by Age, 1982 and 1992 

(Percent) 22 

Figure 5. Classical Music Participation Rates by Education, 

1982 and 1992 (Percent) 23 

Figure 6. Classical Music Participation Rates by Race, 1982 and 1992 

(Percent) 24 

Figure 7. Opera Participation Rates by Age, 1982 and 1992 (Percent) 27 
Figure 8. Opera Participation Rates by Education, 1982 and 1992 

(Percent) 28 

Figure 9. Opera Participation Rates by Race, 1982 and 1992 (Percent) 29 
Figure 10. Musical Participation Rates by Age, 1982 and 1992 (Percent) 31 
Figure 11. Musical Participation Rates by Education, 1982 and 1992 

(Percent) 32 

Figure 12. Musical Participation Rates by Race, 1982 and 1992 

(Percent) 33 

Figure 13. Theater (Plays) Participation Rates by Age, 1982 and 1992 

(Percent) . 35 

Figure 14. Theater (Plays) Participation Rates by Education, 

1982 and 1992 (Percent) 36 

Figure 15. Theater (Plays) Participation Rates by Race, 1982 and 1992 

(Percent) 37 

Figure 16. Ballet/Dance Participation Rates by Age, 1982 and 1992 

(Percent) 40 

Figure 17. Ballet/Dance Participation Rates by Education, 

1982 and 1992 (Percent) 41 

Figure 18. Ballet/Dance Participation Rates by Race, 1982 and 1992 

(Percent) 42 

Figure 19. Art Participation Rates by Age, 1982 and 1992 (Percent) 44 

Figure 20. Art Participation Rates by Education, 1982 and 1992 

(Percent) 45 

Figure 21. Art Participation Rates by Race, 1982 and 1992 (Percent) 46 



Preface 



Much of the data analysis and writing of this monograph was conducted 
while I was a visiting professor in the Department of Political Science — 
the organizational home of the Administration of Justice and Master of Public 
Administration programs — at the University of North Carolina-Chapel Hill. 
I am deeply indebted to a number of departmental colleagues and university 
staff members who overlooked the fact that I am an economist and helped 
create a welcoming atmosphere. Chairman David Lowery and Professor 
George Rabinowicz were especially supportive. Candy Terrell, department 
administrator, simply made things happen. Sue Dodd and Josie Marsh of the 
Institute for Research in the Social Sciences were instrumental in arranging to 
have the data tape uploaded and cataloged; and Jim Cassell provided answers 
to technical questions. One of my co-researchers, Jeffrey Love of the National 
Assembly of State Arts Agencies, provided reassurances regarding data, analyti- 
cal techniques, and results. Tom Bradshaw of the National Endowment for 
the Arts (NEA) was supportive throughout this process, as my ideas evolved 
and as Murphy's Law brought down disks both hard and floppy, disabled entire 
computing systems, uncovered previously unknown software glitches, and 
otherwise impeded progress. 

This research was completed under contract for the National Endowment 
for the Arts. Ideas and opinions expressed are those of the author and do not 
reflect an official position of the NEA or any other public or private agency. 
Any errors or omissions are solely the responsibility of the author. 



Executive Summary 



This monograph examines and interprets survey data pertaining to public 
participation in the arts via television, radio, and sound recordings. The data 
derive from 1982 and 1992 surveys of leisure activities and public participation 
in the arts conducted by the U.S. Bureau of the Census and funded by the 
National Endowment for the Arts (NEA). The seven "core" or "benchmark" 
arts, listed in their order on the survey instruments, are jazz, classical music, 
opera, musicals and operettas, plays, dance, and the visual arts. 

Participation via media is especially important for at least two reasons: media 
may provide arts access to those who are otherwise unable to participate through 
live attendance — because of location, income, or other factors; and the media 
may be influential in cultivating a taste for the arts, which could lead to higher 
overall participation. 

While the primary focus of this monograph is participation by media, 
appropriate comparisons dictate that the analyses be extended to participation 
in live performances and live attendance as well. Many of the hypothesis tests 
necessarily include such live alternatives. 

Although we can speak meaningfully of "arts markets," the arts are widely 
regarded as entailing market failures, one of which is especially important for 
the purposes of this research. 1 The so-called exclusion principle, by which those 
who do not purchase a good or service are denied use, fails in the case of the 
broadcast media. Anyone with a receiver can listen to or view all programming, 
including arts programming. This is an important factor in providing arts access 
to lower income groups, and it informs the hypotheses developed below. 

Public participation in the arts, like consumption of any other good or 
service, is influenced by willingness and ability to pay. Willingness to pay 
encompasses consumer tastes and preferences, which are themselves very likely 
linked to identifiable demographic characteristics. Ability to pay is influenced 
largely by household income and purchase price. 

Simple bivariate techniques are used to test — and confirm — the hypotheses 
put forth in this monograph, but the conclusions rely primarily on logistic 
regression, a multivariate statistical technique that is especially well suited for 
this kind of data. (Logistic regression is further explained in Part III, as well as 
Appendix B.) The hypotheses tested in this monograph, and the results ob- 
tained, are as follows: 



I Turning On and Tuning In 

Arts participation increases with age. The taste for art is an acquired or 
cultivated taste, and both acquisition and cultivation are time-dependent 
processes. It is said that one must learn how to enjoy the arts. Potentially 
offsetting this influence for older persons is their reduced mobility in 
attending live performances and the special challenge of adapting to new 
technology, which would influence media participation. The results for the 
most part support the hypothesis. The major exception is jazz participation, 
which seems to be a young persons' art form. For other art forms and media, 
participation rates rise at least through the middle-age group and sometimes 
decline for the oldest group. 

An aging population will increase arts participation. This is really a 
corollary of the first hypothesis. It would follow that if participation rises 
with age, then an overall older population, other things being equal, would 
participate in the arts at a higher rate. Unfortunately, other things have not 
been equal. Specifically, participation rates among young adults and "30- 
somethings" fell over the decade for many of the art forms and media 
categories, partially or totally offsetting the impact of an aging population. 
This suggests another matter deserving consideration: If arts audiences are 
aging and no efforts are made to increase participation among younger 
groups, the audiences will not be replenished. 

Arts participation increases with education. Students are exposed to the 
arts at every level of education, with the possible exception of postgraduate 
studies. From primary school rhythm bands through college art appreciation 
classes, students have the opportunity to gain increasingly sophisticated 
participation skills. With negligible exceptions, the results bear out the 
hypothesis. Participation rates for the college educated often are double 
those of less educated groups. The downside is that participation rates for 
the more highly educated segments have declined over time. 
Arts participation rises with income. At the very least, those in higher 
income categories are better able to afford arts participation. A relatively 
large segment of the population may regard the prices of symphony tickets 
and compact discs, for example, as high compared with their incomes. The 
statistical results are consistent with the hypothesis. 

Income plays a reduced role in participation via the broadcast media as 
compared with participation via live and recorded performances. The 
broadcast media entail no explicit user charge beyond possession of a 
receiving unit; subsequent participation requires no additional ability to 
pay. Live participation typically involves an admission fee, and listening to 
a recording follows a purchase by the listener or someone else. The tabular 
presentations and statistical results are consistent with this hypothesis. 
Urban residents are more likely to participate in the arts than are rural 



Executive Summary I 3 

residents. Most live participation opportunities are located in urban areas. 
It is likely that the availability of live arts creates a taste that can also be 
served through participation via media. The results bear out this expectation 
for each art form and for every medium, as well as for live performances. 
The ratios of urban-to-rural participation rates are declining, however, 
especially for the broadcast media, indicating that urban location was not 
so important a relative determinant of participation in 1992 as in 1982. 
This is further supported by the logistic regressions, where the urban 
coefficients in the broadcast equations declined from 1982 to 1992. 
Urban residence is not so important a determinant of participation via the 
broadcast media and recordings as it is for participation via live atten- 
dance. While this outcome is not so clear in the simple tabular displays, the 
logistic regression coefficients linking urban residence to arts participation 
are consistently higher for live attendance in both survey periods. 
Men are less likely to participate in the arts than are women. This 
hypothesis is premised on greater early exposure to the arts through lessons 
and other opportunities for personal involvement for girls as compared to 
boys. To the extent that this is the case, and to the extent that such exposure 
is instrumental in forming tastes and preferences, men are less likely to have 
developed such taste. The results confirm this hypothesis for all art forms 
except jazz. 

Whites are more likely to participate in the arts than are other racial 
groups. This hypothesis arises from the understanding that much of the art 
and culture of the United States has European roots that may not be 
immediately accessible to minority cultures. Accordingly, one would expect 
nonwhite participation to be lower. The results are generally consistent with 
this hypothesis, the exceptions being greater black participation in jazz and 
occasionally greater participation by respondents of Asian heritage. 



Conclusions 

The data generally support every hypothesis. Among the policy implications 
supported by the findings are the following: 

■ Some television channels can be dedicated to arts programming. This is 
similar to radio stations adopting a format, which means that, for example, 
televised opera — live or prerecorded — could be available to viewers 24 hours 
a day. The advent of cable access to literally hundreds of channels can 
facilitate the development of dedicated programming. This supply factor 
may well be an important explanation for the higher radio participation in 



I Turning On and Tuning In 

jazz and classical music, compared with television participation in those art 
forms. 

Arts organizations may explore greater use of videotapes — music videos for 
the cultured. Just as many symphony orchestras have found CDs to be a 
source of revenue and exposure, the development of integrated audio-video 
stereophonic systems may offer opportunities for videotape marketing. 
Educators can further enhance exposure to the arts at all levels. Arts 
education may have been at least a partial victim of the educational fads of 
the last several years, including elimination of required college courses in 
the 1970s. The threatened diminution of our society's cultural core may 
justify restoration and expansion of such courses. Many arts organizations 
already collaborate with local school systems to offer arts exposure, and this 
process could be expanded. The arts could also be directly involved in the 
adoption of new learning technology, including distance learning via tele- 
vision, computer linkages, and CD-ROMS. 



Overview, Background, 
and Hypotheses 




Historically, people who wanted to experience the performing or visual arts 
were largely dependent upon geographic proximity. It was necessary to 
attend the theater or the opera or to visit an art museum or gallery. Such 
participation was therefore limited to those who resided in or visited urban areas, 
or who benefited from a visit by a touring company. Since most of the 
population before the 20th century was rural, the opportunities for such artistic 
experiences were severely limited. The advent of recorded and broadcast media, 
however, altered the accessibility of artistic participation, although the form of 
participation was altered as well. One could, of course, argue that a televised 
concert, dramatic performance, or other arts presentation is a fundamentally 
different experience from live attendance. While no one would seriously con- 
tend that a two-dimensional, small-screen viewing of the American Ballet 
Theater is identical to live attendance, some may regard the televised version as 
a viable alternative. This is especially true for those who enjoy dance but who 
have no easy access to performing venues. 

This monograph examines and interprets survey data related to participation 
in the arts via the media. The data derive from 1982 and 1992 surveys of leisure 
activities and public participation in the arts, conducted by the U.S. Bureau of 
the Census and funded by the National Endowment for the Arts. Respondents 
were queried about their participation in seven "core," or "benchmark," arts 
activities via electronic media, including both broadcast and recorded media. 
The broadcast media are radio and television, and their recorded counterparts 
include compact discs (CDs), cassette tapes, vinyl recordings, and videotapes. 
(See Appendix A for the 1992 Survey of Public Participation in the Arts.) 

This monograph gives special attention to the factors that distinguish those 
survey respondents who participate only through the media and those for whom 
the media constitute a complementary means of participation. And although 
the survey questions changed from 1982 to 1992 (for example, "dance" replaced 
the narrower category of "ballet") and technology has evolved (neither VCRs 
nor CDs were widely used in the earlier period), the monograph also discusses 
changes in participation over time. 

The emphasis here is upon each medium as a means of transmitting art rather 
than as an art form in itself. Under certain circumstances, of course, the media 
can become "plastic art": television screens and radio broadcasts have been used, 



6 I Turning On and Tuning In 

for example, in dance performances and music concerts. But here the references 
are quite specific: jazz, classical music, opera, musical plays and operettas 
(musicals hereafter), theater (plays), dance and ballet (dance, usually), and 
museums and the visual arts (art), as conveyed by media. 

As a means of transmitting art, the broadcast media are perhaps especially 
influential in cultivating the taste for art as well. 4 Although the evidence to date 
is inconclusive, the media may play a significant role in creating arts markets. 
The challenge facing the media in this latter regard is apparent from the fact 
that neither broadcast nor cable television offers significant arts programming, 
having yielded most such offerings to the Public Broadcasting System (PBS). 5 



Media Arts as Economic Goods 

In most markets, the forces of supply and demand interact to bring about 
a reasonably efficient allocation of goods and services. Demand reflects the 
willingness and ability of consumers to purchase the good in question. The 
willingness to purchase reflects tastes and preferences, which are indicated to a 
large extent by such demographic characteristics as education, age, gender, race, 
and other taste-determining variables. Ability to purchase is typically indicated 
by household income. Likewise, supply reflects the willingness and ability of a 
seller to make the good available and is largely determined by the costs of 
production. Since the focus here is on the arts participant, those elements that 
influence demand will bear closescscrutiny. 

Matters are complicated somewhat by the fact that all of the media consid- 
ered in this monograph entail one or another type of market failure, which 
means that the forces of supply and demand encounter some sort of interference. 
One of the primary examples in the arts generally is the existence of collective 
benefits, whereby society as a whole gains some benefits in excess of those 
accruing to the actual purchaser of a good or service. For example, a person who 
purchases and wears a pair of shoes enjoys the benefits of doing so; few if any 
benefits accrue to other individuals. But if enough individuals purchase season 
tickets to the opera, the local community may be enriched in ways that exceed 
the benefits to ticket purchasers. 6 

Other types of market failure are of more immediate interest, however. The 
person who refuses to pay the price for a pair of shoes will not ordinarily be able 
to acquire the shoes. He or she is excluded from acquiring and subsequently 
enjoying the goods in question. This exclusion principle applies not only in the 
shoes example, but to any other purely private goods as well. The broadcast 
media are, however, very different. Once an individual or a household has 
purchased a receiving unit, any subsequent broadcasts can be received at no 



Overview, Background, and Hypotheses I 7 

additional charge. Short of employing some sort of scrambling and unscram- 
bling devices, the broadcaster is unable to exclude anyone from receiving the 
broadcasts. This suggests that household income, as a measure of ability to pay, 
may not be so important in determining who participates in the arts via 
broadcast media. The same is true of cable television, albeit perhaps to a lesser 
extent. Most channels are available in a variety of packages for a corresponding 
variety of monthly fees. Except for "pay-per-view" programming, all programs 
in the selected channel package are available at no additional user charge. Again, 
income may not be an important determinant of participation. 

Live performances and recordings have elements of both private goods and 
public goods. In the case of live performances, those who refuse to purchase a 
ticket can be excluded, assuming the performance venue has the means of 
physical exclusion. On the other hand, such performances entail nonrival 
consumption, at least up to the capacity of the performing space. Not only does 
one person's participation not detract from that of another, their mutual 
attendance may, in fact, enhance each other's experience; for example, the sound 
in most concert halls improves when seats are filled. 

Although listening to a recording entails nonrival consumption (numerous 
individuals can, after all, simultaneously listen to a CD on a stereo system), the 
actual purchase of the CD adheres to the exclusion principle. If someone 
purchases a specific CD, no one else can purchase that one, although, of course, 
precise replicas are likely available at the same price. Nevertheless it is the 
existence of a price that is the key, and both a live performance admission fee 
and a CD price mean that ability to pay — income — may be a more significant 
determinant of participation than in the broadcast media. 



"Supply" Factors 

Market sales data typically portray a market outcome influenced by both 
supply and demand forces. This would certainly be the case for CDs or 
videotapes. This is not so obviously the case for the broadcast media, where 
supply and demand do not interact in the usual fashion. 7 Nonetheless, the data 
discussed below indicate the general availability of selected media alternatives. 

Television 

Portraying one measure of the supply of televised arts, Table 1 indicates the 
distribution of cultural programming on public television from 1 974 through 
1992. Although the term cultural is not defined precisely, it very likely extends 
beyond the benchmark arts discussed here to embrace literature, motion 



8 I Turning On and Tuning In 



TABLE 1 . Public Television Programming (Percentages of 
Broadcast Hours) 



Program 
Content 



1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 



General 

News and 
public affairs 3 

Information 
and skills 

Cultural 

General children 
and youth 

"Sesame Street" 

Other general 

Instructional 

Children 
and youth 

Adult 



12.6 11.9 11.0 12.2 12.4 14.1 16.4 16.3 17.6 17.4 

15.9 19.9 23.6 22.8 24.5 25.5 29.5 31.7 31.5 28.7 
17.9 20.9 22.1 21.9 22.6 20.1 20.5 17.9 19.1 17.5 



10.7 10.0 

21.2 17.8 

4.4 3.8 



8.7 8.9 7.5 

16.1 15.5 14.8 

5.3 5!5 4.8 



7.9 6.5 5.8 6.0 14.6 

14.8 11.4 11.7 11.2 11.0 

5.5 1.6 1.2 0.9 0.9 



15.2 15.2 13.7 13.7 12.9 12.4 
1.9 1.4 1.2 1.0 1.4 0.6 



8.7 
2.9 



Note: Figures do not add up to 1 00 due to rounding. 1 974 and 1 976 are calendar 
years; 1978 to 1990 are fiscal years; 1988 through 1992 surveys used October 
through September seasons. From 1988 through 1992, only broadcasters in the 50 
U.S. states were surveyed. 

a. From 1986 through 1992, the News and Public Affairs category included "Business 
or Consumer." 

b. Not including "Sesame Street," which is reported separately. 

c. After 1974, some general audience programs with instructional applications were 
double counted if aired during school hours when school was in session. 

Source: Unpublished data provided by Janice Jones, media research analyst, 
Corporation for Public Broadcasting 



pictures, and popular culture. The rise that took place in the early 1980s in the 
percentage of broadcast hours devoted to cultural programming had diminished 
by the early 1990s. The 1992 percentage of 17.5, in fact, was the lowest of all 
the years covered. To be sure, some of the children and youth programming 
may include cultural aspects, and that may be a particularly propitious inclusion, 
to the extent that media exposure helps shape tastes. But it is clear, based on the 
data in Table 1 , that cultural programming directed specifically to adults, that 
is, arts via television, has been waning. 



Overview, Background, and Hypotheses 



Radio 

Similarly, Table 2 offers some informa- 
tion on the availability of the benchmark 
arts via radio. Lacking the visual aspect of 
television, radio, of course, presents only 
the audible arts: jazz, classical music, opera, 
and musicals. Clearly, far more stations 
offer rock, country, and other musical for- 
mats than either classical or jazz. 



Recordings 

Table 3 displays trends in unit ship- 
ments of recorded music from 1983 to 
1993. The change in technology over that 
period is apparent in the appearance and 
rapid expansion of compact discs (CDs) 
and music videos and the near demise of 
the long-playing vinyl albums (LP/EP) and 
singles. Unit shipments of CDs grew at an 
average annual rate of 90.2 percent, while 
CD singles, introduced somewhat later, 
expanded at a 34.3 percent annual pace. 

Shipments of cassettes peaked in 1988 
and declined thereafter, for an average growth rate over the entire period of only 
3.7 percent. The declines in shipments of long-playing and single vinyl record- 



TABLE 2. Number of 


U.S. Radio Stations 


Offering Various 




Formats, 1 992 






Total 


Format 


Stations 


Adult contemporary 


2,288 


Beautiful music 


271 


Top-40 


767 


Country 


2,651 


MOR 


528 


News 


596 


Oldies 


1,031 


Religious 


1,144 


Rock 


584 


Talk 


388 


Classical 


442 


Jazz 


367 


Note: Table data cover only 


those stations providing 


! format 


information. 




Source: Broadcasting and Cable 


Yearbook, 1993 





TABLE 3. Manufactui 


rers' 


Shipments 


of Recorded Music 






(Millions of Units) 














Item 


1983 


1984 


1985 


1986 


1987 


1988 


1989 


1990 


1991 


1992 


1993 


CD 


0.8 


5.8 


22.8 


53.0 


102.1 


149.7 


207.2 


286.5 


333.3 


407.5 


495.4 


CD 
single 


NA 


NA 


NA 


NA 


NA 


1.6 


-0.1 


1.1 


5.7 


7.3 


7.0 


Cassette 


236.8 


332.0 


339.1 


344.5 


410.0 


450.1 


446.2 


442.2 


360.1 


366.4 


339.5 


LP/EP 


209.6 


204.8 


167.0 


125.2 


107.0 


72.4 


34.6 


11.7 


4.8 


2.3 


1.2 


Single 


124.6 


131.5 


120.7 


93.9 


82.0 


65.6 


36.6 


27.6 


22.0 


19.8 


15.1 


Music 
video 


NA 


NA 


NA 


NA 


NA 


NA 


6.1 


9.2 


8.1 


7.6 


11.0 


Note: F 
Source: 


gures refer to net number of units after returns. 
Recording Industry Association of America 











1 I Turning On and Tuning In 

ings were at average annual rates of -40.3 and -19.0 percent, respectively. 
Compared with CDs, music video shipments grew at a relatively modest average 
annual pace of 15.9 percent. 

The relative importance of two of the benchmark arts on recordings is 
indicated by the percentage of U.S. dollar sales accounted for by jazz and classical 
music. As Table 4 shows, rock music sales are by far the largest category, with 
country and pop a distant second and third, respectively. Classical and jazz are 
much further down the list. Both show a decline over the period covered, 
although the classical pattern is mixed and the jazz trend may reflect the 1991 
appearance of the "Urban Contemporary" category. 



Factors Influencing Arts Participation: Hypotheses 

Before the data presentations and statistical analyses of Parts II and III of 
this monograph, it is appropriate to consider some of the factors expected to 
influence arts participation. These give rise to the hypotheses that provide a 
framework or focus for the analyses. 

The Impact of Place 

The existence of economies of agglomeration in the art and culture industry 
now seems well established. Such economies occur when firms are able to share 
one or more common productive factors in a given geographic region. For 
example, domestic automobile firms historically chose to locate in or near 
Detroit because they could share a trained labor force, transportation infrastruc- 
ture, and access to raw materials. Similarly, radio and television programming 



TABLE 4. 


Sales of Recorded Music 


by Genre (Percentage of 




U.S. Dollar Value) 








Genre 


1989 


1990 


1991 


1992 


1993 


Rock 


42.9 


37.4 


36.3 


33.2 


32.6 


Country 


6.8 


8.8 


12.5 


16.5 


17.5 


Pop 


14.4 


13.6 


11.7 


11.4 


11.7 


Urban contemporary NA 


NA 


9.3 


8.8 


9.9 


Rap 


NA 


NA 


8.9 


7.9 


7.8 


Classical 


4.3 


4.1 


3.9 


4.4 


4.0 


Jazz 


5.7 


5.2 


4.3 


4.0 


3.3 


Source: Recording Industry Association of America, 1993 Cor 


lsumer Profile 


^ 



Overview, Background, and Hypotheses I 1 1 



were drawn to New York City and to Hollywood because of those locations' 
respective talent pools generated by Broadway and the motion picture industry. 

Another factor causing the arts to play a disproportionately large role in 
places with larger populations is the existence of threshold market sizes. Whereas 
a small community could scarcely support a symphony orchestra or a major art 
museum, most major metropolitan areas are able to support both, and additional 
cultural organizations as well. 

Accordingly, one might expect that residents of urban regions would 
generally exhibit a greater likelihood of attending the live arts if for no other 
reason than because of greater availability. The data in Table 5 demonstrate that 
pattern. It is not so obvious that urban location is as conducive to participation 
via the media. Indeed, one could hypothesize in either direction: (1) urban 
location, which facilitates live attendance, facilitates development of a taste for 
the arts, which can also be partially 
satisfied by media participation; or 
(2) rural residents, denied ready ac- 
cess to live arts attendance, com- 
pensate through heavier media 
participation. 

One might reasonably assume 
that media participation is not so 
constrained as live participation. 
Television and radio signals reach 
into the hinterlands, and recorded 
music is widely available. One need 
not reside in a major metropolitan 
area to tune in to public radio or 
television or to purchase a CD. Yet, 
as we shall see in subsequent chap- 
ters, media participation is also greater in urban areas than in rural areas. For 
example, participation through television is greater in urban than in rural areas 
for all the benchmark arts, although the relative proportions differ somewhat. 



TABLE 5. Live 


! Arts 


Participation 


Rates by Place, 


1 992 (Percentages of Adult 


Population) 






Percentage 


Art Form 


Urban Rural 


Jazz 


12.2 6.3 


Classical music 


13.8 9.1 


Opera 


3.7 2.1 


Musicals 


19.2 12.8 


Plays 


15.3 8.5 


Dance 


7.6 5.9 


Art museum 


29.8 18.5 



Age 

Culture, it is said, is an acquired taste, and acquisition of taste takes time. 
It would follow then that participation would increase with age. The adult 
population has been classified into four groupings for the purposes of analysis 
in this monograph: young adults (18 through 29), "30-somethings" (30 through 
44), middle-age (45 through 64), and "retired" (65 and above). These groupings 
are not entirely arbitrary. The youngest range from traditional college age to 



12 I Turning On and Tuning In 

those who, although in their late 20s, are still in the early stages of their adult 
lives. The 30-somethings, a category that actually extends into the mid-40s, 
have had opportunities to become established in their chosen lifestyles, to incur 
the obligations of adulthood, and to have made a career move or two. The 
middle-age range encompasses a variety of life experiences. The retired group, 
many of whom may not be retired at all, are experiencing the culmination of 
life's experiences. 

Education 

One way we acquire a taste for culture and the arts is exposure through 
education. It has been argued that appreciation of the more complex arts 
requires investment in "consumption skills," that is, learning to understand, say, 
opera. Elementary and secondary school curricula typically offer art and music 
classes, and most colleges offer — and many require — art or music appreciation 
courses. Accordingly, we would reasonably expect participation to rise as 
education level rises. 

For purposes of subsequent analyses, reported years of education have been 
aggregated into four groups: (1) none (respondents indicated zero years of 
education); (2) elementary (education did not proceed beyond grades 1 through 
8); (3) high school (grades 9 through 12); and (4) college (including college and 
postgraduate education). 

Income 

The most obvious means by which income would impact arts participation 
would be ability to pay. Clearly, persons with higher incomes are more likely to 
be able to afford to attend a performing arts activity, to visit a museum, or to 
purchase a CD. They are also more likely to purchase one or more television 
sets and radios; but in this latter case, once the device is purchased, the 
incremental costs of viewing or listening to the arts are likely to be very close to 
zero, especially for broadcasts or the more widely available cable channels. 

As the data in Table 6 clearly indicate, both live and media participation 
rise with income. The difference is unlikely to be attributable to television access; 
in 1992, 93.7 percent of persons in households with incomes of less than 
$10,000 watched television, as did 95.2 percent of those in households with 
incomes between $10,000 and $20,000. n Interestingly enough, radio listening 
rises significantly with income category, from 11 A percent of those individuals 
in the lowest income category to 91.6 percent of those in the highest. 12 But 
given the low prices of the simplest radios it seems unlikely that this reflects 
ability to pay. It would follow that the relationship between household income 



Overview, Background, and Hypotheses I 13 



TABLE 6. 


Arts 


Participation Rates by Income Group, 


1992 




(Percentages of Adult 


Populatior 







Art Form/Medium 




Income 


Level 




Poverty 


Low 


Moderate 


High 


jazz 












Live 




5.2 


8.1 


12.4 


22.1 


Television 




12.5 


17.2 


19.5 


27.9 


Radio 




18.8 


24.3 


31.9 


44.4 


Recorded 




11.4 


16.6 


23.9 


37.5 


Classical music 










Live 




5.8 


9.3 


14.4 


28.6 


Television 




15.2 


19.5 


23.2 


38.1 


Radio 




19.8 


24.7 


34.9 


53.4 


Recorded 




12.7 


17.7 


27.8 


45.2 


Opera 












Live 




1.8 


1.7 


3.3 


12.2 


Television 




7.7 


9.0 


11.4 


18.0 


Radio 




5.4 


6.3 


9.6 


17.9 


Recorded 




3.1 


3.8 


7.9 


16.0 


Musicals 












Live 




7.6 


12.6 


20.2 


42.3 


Television 




9.1 


11.9 


13.2 


18.7 


Radio 




2.7 


2.9 


3.6 


6.4 


Recorded 




3.0 


3.8 


6.3 


13.1 


Plays 












Live 




6.8 


9.9 


15.2 


31.8 


Television 




10.0 


12.2 


16.8 


24.7 


Radio 




2.4 


2.3 


3.1 


4.2 


Dance 












Live 




4.6 


6.3 


8.2 


9.5 


Television 




14.2 


17.0 


17.3 


25.0 


Art 












Live 




13.4 


20.6 


31.6 


49.8 


Television 




21.8 


27.4 


33.0 


44.1 



and broadcast media participation reflects influences of taste rather than pur- 
chasing power. 

For analytical purposes, the several reported income bands were aggregated 
into four groups. These have been characterized as (1) "poverty," the lowest 
group (which does not necessarily correspond to census definitions of poverty 



14 I Turning On and Tuning In 

level); (2) "low"; (3) "moderate," which may be regarded as "middle class"; and 
(4) "high." Because of changes in both price levels and income groupings, the 
four groups are not comparable between the two time periods. 

Gender 

No obvious or intrinsic reason leads one to expect differential participation 
rates between men and women. Yet the differences are well known: women 
participate in the arts at higher rates than men. Perhaps this is rooted in early 
acculturation processes, when boys were often channeled into sports to the 
exclusion of arts. 

Race 

As with gender, there is no obvious source of race-based differentials in arts 
participation. Accounting for other factors that are correlated with race, such as 
income and education, may not eliminate racial differences. Racial or ethnic 
groups that are not European in origin may not be strongly attracted to such 
art forms as symphonic music and traditional opera, which are firmly rooted in 
Western artistic traditions. Minority groups may feel excluded from "main- 
stream" arts. The racial groupings of the survey are collapsed into white, black 
(African American), Asian, and Indian (including all Native Americans). Re- 
spondents of Hispanic origin are identified separately and do not identify 
themselves in only one of the racial groups. 



Participation Patterns and 
Changes, 1 982-1 992 




This part describes participation patterns in the arts via media, with particular 
focus on the roles of selected population demographic characteristics as 
suggested by the hypotheses of Part I. Tables and figures accompanying the text 
help to illustrate audience patterns and trends. The first section offers an 
overview of public participation for 1992, the most recent survey period. The 
second section contains more detailed participation data, and is organized by 
benchmark arts activity, beginning with jazz. The participation patterns are 
examined and discussed, including variations over time. The final section 
summarizes the findings. 



Overview of Public Participation, 1992 

Table 7 summarizes public participation data for each of the benchmark 
activities and each of the media alternatives for 1992. The last column of the 
table includes live participation for comparison purposes. With the exception 
of videocassette viewing, media participation tends to exceed live participation 
across the board. Musicals and plays, however, with their significant visual 
components, do not attract larger audio media audiences in the way that the 
other benchmark activities do. These data reflect the fact that the household 
penetration of television enables quick, easy, and relatively low cost access to 
televised arts. Participation by radio is higher for jazz and classical music, but 
television participation is higher for the performing arts — opera, musicals, and 
plays — with a notable visual component. 

The levels of videocassette participation attest to the novelty of this medium 
as a means of arts participation. These comparatively modest rates suggest 
further analysis to be of limited benefit, but the data provide at least a baseline 
for future comparisons. 



Participation Patterns 

The following subsections examine participation rates for various demo- 
graphic factors for each of the benchmark arts for the years 1982 and 1992. 
Each subsection includes a comprehensive table depicting participation rates by 



15 



16 I Turning On and Tuning In 



TABLE 7. 


Rates of Public 


Parti< 


cipation in 


the Arts via Media, 




1992 (Percentages of Adult Population) 




Art Form 


Television 


Video 


Radio 


Recordings 


Performance 


Jazz 


17.9 


1.0 


28.2 


20.6 


10.6 


Classical music 21 .9 


1.2 


30.8 


23.8 


12.5 


Opera 


10.7 


0.5 


8.7 


6.9 


3.3 


Musicals 


12.5 


1.8 


3.5 


5.7 


17.5 


Plays 


14.8 


1.3 


2.8 


NA 


13.5 


Dance 


17.1 


0.8 


NA 


NA 


7.1 


Art 


30.1 


NA 


NA 


NA 


26.7 



the variables already indicated: urban versus rural, gender, education, age, race, 
and income. The tables offer a summary of participation rates in each year as 
well as a comparison of rates over time. In addition to the tables, charts and 
associated tables offer a quick overview of participation rates and changes over 
time and by medium for age, education, and racial groups. Income is not 
accorded a separate chart since the data are not comparable over time. Those 
variables — location and gender — that take on only two values also do not merit 
separate charts. The tables and charts do not include the relatively minuscule 
"none" group of the education variable. 

Two statistical tests were applied to each component of the tables: a Pearson 
chi-square test of independence and a phi-coefficient, which indicates strength of 
association. In every instance, the Pearson test indicates that the variables (e.g., 
participation by race, participation by income, etc.) are not independent of each 
other — that is, the observed patterns are not simply random events. Further- 
more, the phi-coefficient indicates in every case that the associations are 
statistically significant. A major shortcoming of this approach — measuring 
association by two variables at a time — is that it could be misleading. For 
example, an apparent association between participation and income may actu- 
ally reflect an association between participation and education, which is strongly 
correlated with income. This limits the inferences that can be drawn from the 
tests. Part III draws upon statistical techniques that control for the separate 
effects of variables that are themselves correlated. 



Jazz 

As indicated in Table 8, overall jazz participation via media is higher than 
live participation in both 1982 and 1992. Television viewing and listening to 
recordings show little change over the decade, while radio listening rose by 



Participation Patterns and Changes, 1982-1992 I 17 



TABLE 8. 


Jazz Participation Rates by 


Variable, 1982 and 






1992 (Percentages of Adult 


Popu 


lation) 






Variable 


Live 


Television 


Radio 


Recorc 

1982 


ing 
1992 


1982 


1992 


1982 


1992 


1982 


1992 


Total 


9.6 


10.6 


18.0 


17.9 


18.1 


28.2 


20.2 


20.6 


Location 


















Urban 


11.0 


12.2 


20.3 


19.9 


20.8 


31.5 


23.5 


23.5 


Rural 


6.5 


6.3 


12.6 


12.7 


11.8 


19.4 


12.1 


12.7 


Gender 


















Male 


10.3 


11.9 


19.6 


19.4 


20.5 


30.9 


21.3 


22.6 


Female 


9.0 


9.4 


16.7 


16.5 


16.0 


25.7 


19.2 


18.7 


Education 


















Elementary 


1.3 


0.8 


4.5 


6.7 


6.2 


9.4 


4.7 


4.6 


High school 


6.1 


4.9 


15.2 


12.8 


15.1 


19.6 


16.0 


12.5 


College 


17.0 


18.4 


25.9 


25.3 


25.6 


40.5 


30.5 


31.8 


Age 


















Young adult 


17.0 


12.4 


21.0 


14.8 


27.4 


30.1 


29.5 


23.2 


30-something 


9.7 


12.9 


19.0 


20.1 


17.9 


34.5 


20.4 


25.5 


Middle age 


6.0 


9.7 


19.1 


19.6 


15.3 


26.0 


17.6 


18.4 


Retired 


1.8 


4.8 


9.1 


15.1 


6.0 


15.9 


6.6 


10.1 


Race 


















White 


8.9 


10.0 


16.7 


16.7 


15.8 


26.0 


18.2 


18.8 


Black 


15.5 


16.1 


28.1 


27.8 


35.9 


45.4 


36.5 


35.5 


Other 


8.3 


6.5 


21.7 


14.4 


23.4 


25.1 


20.3 


15.1 


Income 3 


















Poverty 


7.6 


5.2 


12.2 


12.5 


16.9 


18.8 


13.7 


11.4 


Low 


8.8 


8.1 


17.9 


17.2 


18.0 


24.3 


20.4 


16.6 


Moderate 


11.8 


12.4 


21.6 


19.5 


19.0 


31.9 


24.2 


23.9 


High 


16.7 


22.1 


26.0 


27.9 


19.6 


44.4 


30.2 


37.5 


a Groupings not comparable between periods 













slightly more than 10 percentage points. Participation by all modes is higher in 
urban than in rural areas, again with little change over time except for radio. 
Jazz is the only benchmark art with greater male and black participation rates. 
Many jazz artists are black, and much of jazz is rooted in the African American 
cultural experience. 

As hypothesized, participation rises with education and income, but, some- 
what surprisingly, falls with age. Jazz seems to be a young persons' art form. 
Some patterns of change over time are noteworthy. Except for radio, media 
participation by the high school group decreased noticeably over time. Among 
the age groupings, participation rose over time except among young adults. 
Although this group's radio listening rose, other forms of media participation 
by young adults declined, suggesting that the jazz audience is aging. 



18 I Turning On and Tuning In 



FIGURE 1. Jazz Participation Rates by Age, 
1982 and 1992 (Percent) 



Young Adult 30-something Middle Age Retired 




1982 1992 1982 1992 1982 1992 1982 1992 
Live Television Radio Recordings 



TABLE 9. Change 


in Jazz 


Participation 


Rates by Age, 


1982-1992 








Age Group 


Live 


Television 


Radio 


Recordings 


Young adult 


-4.6 


-6.2 


2.7 


-6.3 


30-something 


3.2 


1.1 


16.6 


5.1 


Middle age 


3.7 


0.5 


10.7 


0.8 


Retired 


3.0 


6.0 


9.9 


3.5 


Note: Numbers indicate increase or 


decrease in percentage points. 





Participation Patterns and Changes, 1 982-1 992 I 1 9 



FIGURE 2. Jazz Participation Rates by Education, 
1982 and 1992 (Percent) 



50- 



40- 



Elementary 



High School 



College 




1982 1992 
Live 



1982 1992 1982 1992 
Television Radio 



1982 1992 
Recordings 



TABLE 10. Change in Jazz Participation Rates by Education, 
1982-1992 



Education' 



Live Television Radio Recordings 



Elementary 
High school 
College 



-0.5 

-1.2 

1.4 



2.2 


3.2 


-0.1 


2.4 


4.5 


-3.5 


0.6 


14.9 


1.3 



Note: Numbers indicate increase or decrease in percentage points. 
a "None" not included 



20 I Turning On and Tuning In 



FIGURE 3. Jazz Participation Rates by Race, 
1982 and 1992 (Percent) 



50- 



Black White |_J Other 




1982 1992 1982 1992 
Live Television 



1982 1992 1982 1992 
Radio Recordings 



TABLE 1 1 . Change in Jazz 


Participation 


Rates by 


Race, 


1982-1992 








Race Live 


Television 


Radio 


Recordings 


White 1.1 


0.0 


10.2 


0.6 


Black 0.6 


-0.3 


9.5 


-1.0 


Other -1 .8 


-7.3 


1.7 


-5.2 


Note: Numbers indicate increase or 


decrease in percentage points. 





Participation Patterns and Changes, 1982-1992 I 21 



Classical Music 



As indicated in Table 12, participation in classical music via media exceeds 
live participation. Live and television participation have fallen slightly over time, 
while radio increased substantially and recordings a modest amount. The rise 
in radio participation is true for all of the demographic groupings. Participation 
is higher among urban and female respondents, but male participation rates rose 
over the decade at least slightly in three of the four categories. As expected, 
participation rises with education, age (though most categories show a lower 
level of participation among the "retired" age group), and income. White 
participation is higher than black, while the "other" category shows a mixed 
pattern. 



TABLE 12. 


Classical Music 


Participation 


Rates 


by Variable, 






1982 and 1992 


(Percentages 


of Ad 


ult Population) 


Variable 


Live 


Television 


Rad 


io 


Record 

1982 


ings 
1992 


1982 


1992 


1982 


1992 


1982 


1992 


Total 


13.0 


12.5 


24.7 


21.9 


19.9 


30.8 


22.1 


23.8 


Location 


















Urban 


14.7 


13.8 


26.5 


23.6 


21.5 


33.0 


23.1 


25.7 


Rural 


9.3 


9.1 


20.4 


17.3 


16.2 


24.8 


19.8 


18.9 


Gender 


















Male 


11.3 


11.5 


23.3 


20.3 


20.5 


30.7 


21.2 


23.1 


Female 


14.6 


13.4 


25.9 


23.3 


19.4 


30.9 


22.9 


24.5 


Education 


















Elementary 


1.7 


1.9 


10.1 


9.8 


8.3 


11.8 


6.2 


5.8 


High school 


6.6 


5.8 


19.6 


15.8 


13.0 


19.6 


14.5 


13.2 


College 


25.2 


21.4 


36.1 


30.5 


32.6 


46.0 


37.1 


38.0 


Age 
Young adult 


11.5 


9.9 


17.2 


14.0 


16.5 


24.3 


19.8 


22.0 


30-something 


15.3 


11.8 


25.3 


19.7 


23.4 


32.9 


25.6 


25.6 


Middle age 


13.8 


16.1 


33.2 


27.1 


22.7 


36.5 


25.7 


26.5 


Retired 


10.2 


11.8 


23.3 


29.2 


15.5 


26.6 


14.2 


18.5 


Race 


















White 


13.9 


13.2 


25.6 


22.5 


20.2 


32.1 


22.9 


25.2 


Black 


6.7 


6.9 


15.9 


16.1 


15.4 


19.6 


13.2 


12.6 


Other 


9.4 


12.8 


30.9 


25.3 


29.2 


36.4 


31.4 


25.9 


Income 3 


















Poverty 


8.2 


5.8 


15.1 


15.2 


13.4 


19.8 


12.8 


12.7 


Low 


10.5 


9.3 


25.1 


19.5 


18.0 


24.7 


21.4 


17.7 


Moderate 


18.3 


14.4 


29.8 


23.2 


25.3 


34.9 


27.7 


27.8 


High 


30.1 


28.6 


47.6 


38.1 


40.0 


53.4 


38.9 


45.2 


a Groupings not comparable 


betweer 


periods 













22 I Turning On and Tuning In 



FIGURE 4. Classical Music Participation Rates by Age, 
1982 and 1992 (Percent) 



Young Adult 30-something Middle Age Retired 




1982 1992 1982 1992 1982 1992 1982 1992 
Live Television Radio Recordings 



TABLE 13. Change in Classical Music Participation Rates by 
Age, 1982-1992 



Age Group 



Live 


Television 


Radk 


-1.6 


-3.2 


7.8 


-3.5 


-5.6 


9.5 


2.3 


-6.1 


13.8 


1.6 


5.9 


11.1 



Recordings 



Young adult 
30-something 
Middle age 
Retired 



Note: Numbers indicate increase or decrease in percentage points. 



2.2 
0.0 
0.8 
4.3 



Participation Patterns and Changes, 1982-1992 I 23 



FIGURE 5. Classical Music Participation Rates by Education, 
1982 and 1992 (Percent) 



50- 



40- 



30- 



20- 



10- 







Elementary 



High School 







iiu 




College 



ill 




1982 1992 1982 1992 1982 1992 1982 1992 
Live Television Radio Recordings 



TABLE 14. Change in Classical Music Participation Rates by 
Education, 1982-1992 



Education' 



Live 



Television Radio Recordings 



Elementary 
High school 
College 



0.2 
-0.8 
-3.8 



-0.3 
-3.8 
-5.6 



3.5 

6.6 

13.4 



-0.4 

-1.3 

0.9 



Note: Numbers indicate increase or decrease in percentage points. 
a "None" not included 



24 I Turning On and Tuning In 



FIGURE 6. Classical Music Participation Rates by Race, 
1982 and 1992 (Percent) 



50- 



40- 



30- 



20- 



10- 



Black 



White 



I 



IB III 
Warn 

■ 11 

,1 1 - 



Other 




1982 1992 1982 1992 
Live Television 



1982 1992 1982 1992 
Radio Recordings 



Table 15. Change in Classical Music Participation Rates by 
Race, 1982-1992 



Race 



Live Television Radio Recordings 



White 
Black 
Other 



-0.7 
0.2 
3.4 



-3.1 
0.2 
-5.6 



11.9 
4.2 
7.2 



2.3 
-0.6 
-5.5 



Note: Numbers indicate increase or decrease in percentage points. 



Participation Patterns and Changes, 1982-1992 I 25 

For the most part, live and television participation fell over the decade, even 
for the college-educated group, suggesting that education is not so strong a 
determinant of tastes as it once might have been. These categories of participa- 
tion also declined for the younger age groups, an indicator that the classical 
music audience is aging and not replenishing itself. 

The dramatic increase in radio participation deserves further mention. The 
wide reach of radio stations devoted exclusively to classical music and the growth 
of both National Public Radio and Public Radio International mean that anyone 
with a radio has access to classical music virtually any time of day. This is not 
true of television. As discussed in Part I, few broadcast or cable networks other 
than public television offer classical music, and public television devotes only 
some 17 percent of its programming to cultural events. Interestingly, this 
popularity of radio participation has not carried over to recordings. Perhaps the 
sound quality available on compact discs is offset by the relatively high price of 
CD players, including automobile stereo systems. 

Opera 

In 1992 opera ranked lowest among the benchmark arts for live and 
television participation, and it was a distant third in both radio and recordings. 
As shown in Table 16, participation rates were higher in both 1982 and 1992 
for urban than for rural respondents and for female than for male respondents. 
Interestingly, television participation exceeded participation rates for live per- 
formances and the other media. This may reflect the visual nature of opera. 
Viewers can easily consult television listings to learn when a performance is 
scheduled, which is not so true of radio. 

Participation increases with education, age (except for the "retired" group), 
and income. Except for radio, participation by the high school and college- 
educated groups has not increased over the decade. 

Live and radio participation in general rose over the decade. It is interesting 
to note that black participation rose for live performances and all media except 
recordings. Indeed, black participation even rose for television, which declined 
for nearly every other demographic grouping. 



26 I Turning On and Tuning In 



TABLE 16. 


Opera I 


Participation 


Rates by Variable, 1982 and 




1992 (Percentages of Adult 


Population) 






Variable 


i 


Live 


Television 


Rad 


io 


Record 
1982 


ings 
1992 


1982 


1992 


1982 


1992 


1982 


1992 


Total 


3.0 


3.3 


12.0 


10.7 


7.1 


8.7 


7.5 


6.9 


Location 
















- 


Urban 


3.7 


3.7 


13.4 


12.0 


7.9 


9.4 


8.3 


7.8 


Rural 


1.5 


2.1 


8.7 


7.4 


5.2 


6.6 


5.4 


4.7 


Gender 


















Male 


2.7 


3.1 


10.5 


10.0 


7.2 


8.4 


7.1 


6.6 


Female 


3.3 


3.5 


13.4 


11.4 


7.0 


8.9 


7.7 


7.3 


Education 


















Elementary 


0.5 


0.6 


4.3 


4.9 


3.8 


3.6 


2.8 


2.1 


High school 


1.6 


1.3 


9.7 


7.0 


4.1 


4.9 


5.4 


3.4 


College 


5.8 


5.8 


. 17.6 


15.7 


11.9 


13.6 


11.7 


11.4 


Age 








• 










Young adult 


2.3 


2.5 


6.2 


5.7 


4.9 


4.0 


3.4 


4.3 


30-something 


3.2 


3.3 


11.7 


9.2 


5.9 


7.4 


7.7 


6.3 


Middle age 


3.7 


4.1 


17.9 


14.7 


10.3 


12.7 


11.7 


9.9 


Retired 


2.9 


3.1 


13.5 


14.8 


7.7 


11.5 


7.2 


7.2 


Race 


















White 


3.2 


3.4 


12.1 


10.7 


7.2 


9.0 


7.9 


7.3 


Black 


1.3 


1.8 


9.3 


10.4 


5.3 


6.0 


3.9 


3.7 


Other 


3.0 


5.3 


20.1 


12.3 


12.0 


8.3 


8.3 


9.4 


Income 3 


















Poverty 


1.7 


1.8 


7.5 


7.7 


5.1 


5.4 


3.8 


3.1 


Low 


2.0 


1.7 


11.6 


9.0 


6.1 


6.3 


7.1 


4.5 


Moderate 


4.1 


3.3 


15.3 


11.4 


8.1 


9.6 


9.2 


7.9 


High 


10.0 


12.2 


24.4 


18.0 


15.5 


17.9 


13.5 


16.0 


a Groupings not 


comparable I 


between per 


ods 













Participation Patterns and Changes, 1982-1992 I 27 



FIGURE 7. Opera Participation Rates by Age, 
1982 and 1992 (Percent) 



Young Adult 30-something Middle Age Retired 




1982 1992 
Live 



1982 1992 
Television 



1982 1992 
Radio 



1982 1992 
Recordings 



TABLE 17. Change in Opera Participation Rates by Age, 
1982-1992 



Age Group 



Live 



Television Radio Recordings 



Young adult 
30-something 
Middle age 
Retired 



0.2 
0.1 
0.4 
0.2 



-0.5 

-2.5 

-3.2 

1.3 



-0.9 
1.5 

2.4 
3.8 



0.9 
-1.4 
-1.8 

0.0 



Note: Numbers indicate increase or decrease in percentage points. 



28 I Turning On and Tuning In 



FIGURE 8. Opera Participation Rates by Education, 
1982 and 1992 (Percent) 



Elementary High School College 




1982 1992 1982 1992 1982 1992 1982 1992 
Live Television Radio Recordings 



TABLE 18. Change in Opera Participation Rates by 
Education, 1982-1992 



Education' 



Live 



Television Radio Recordings 



Elementary 
High school 
College 



0.1 

-0.3 

0.0 



0.6 

-2.7 
-1.9 



-0.2 
0.8 

1.7 



-0.7 
-2.0 
-0.3 



Note: Numbers indicate increase or decrease in percentage points. 
a "None" not included 



Participation Patterns and Changes, 1 982-1 992 I 29 



FIGURE 9. Opera Participation Rates by Race, 
1982 and 1992 (Percent) 



25- 



20- 



15 



10- 







Black 



_ White 



1982 1992 
Live 



Other 




■ 






— 



MI 






mm 

111 



1982 1992 
Television 



1982 1992 
Radio 



1982 1992 
Recordings 



TABLE 19. Change in Opera Participation Rates by Race, 


1982-1992 






Race Live 


Television 


Radio Recordings 


White 0.2 


-1.4 


1.8 -0.6 


Black 0.5 


1.1 


0.7 -0.2 


Other 2.3 


-7.8 


-3.7 1.1 


Note: Numbers indicate increase or 


decrease in percentage points. 



30 I Turning On and Tuning In 



Musicals 



Table 20 indicates that, with few exceptions, live and media participation 
rates fell from 1982 to 1992. Live and television participation far exceed 
participation via both radio and recordings, due undoubtedly to the visual 
nature of musicals. Urban participation exceeds rural, and female participation 
generally exceeds male. Participation rises with education, age (except for the 
"retired" group), and income. 

The rather dramatic fall in television participation probably reflects a 
supply-side phenomenon, which exacerbated the overall decline. Participation 
in musicals via media declined for all age groups except the retired, for high 
school and college groups, and generally for all racial groups. 



TABLE 20. 


Musical 


Participation 


Rates 


by Variable, 


1982 and 




1992 (Percentages of Adult 


Population) 






Variable 


Live 


Television 


Rad 


io 


Record 

1982 


ings 
1992 


1982 


1992 


1982 


1992 


1982 


1992 


Total 


18.6 


17.4 


20.4 


12.5 


4.3 


3.5 


8.4 


5.7 


Location 


















Urban 


21.2 


19.2 


22.9 


13.6 


5.1 


3.8 


9.9 


6.5 


Rural 


13.1 


12.8 


14.3 


9.7 


2.6 


2.6 


4.7 


3.5 


Gender 




_ 














Male 


16.6 


15.1 


18.8 


11.4 


4.6 


3.4 


7.2 


5.3 


Female 


20.5 


19.6 


21.7 


13.6 


4.1 


3.6 


9.3 


6.0 


Education 


















Elementary 


4.1 


3.1 


9.0 


6.7 


1.1 


2.4 


1.9 


2.1 


High school 


11.5 


10.4 


15.7 


9.4 


2.8 


1.9 


4.4 


2.4 


College 


32.9 


27.5 


30.3 


17.0 


7.3 


5.4 


15.6 


9.8 


Age 


















Young adult 


17.6 


15.3 


18.1 


7.4 


3.6 


2.0 


7.0 


4.8 


30-something 


22.1 


18.0 


20.7 


11.4 


4.2 


3.1 


9.4 


5.9 


Middle age 


20.0 


21.0 


23.2 


15.6 


5.9 


4.6 


11.5 


7.2 


Retired 


12.0 


13.6 


19.1 


17.2 


3.5 


4.7 


3.8 


4.1 


Race 


















White 


19.8 


18.1 


20.7 


12.7 


4.2 


3.6 


9.1 


6.0 


Black 


10.0 


14.1 


17.4 


11.3 


4.4 


2.5 


1.9 


2.9 


Other 


13.1 


11.7 


19.8 


12.9 


8.1 


5.1 


10.2 


7.3 


Income 3 


















Poverty 


9.2 


7.6 


12.9 


9.1 


2.7 


2.7 


4.2 


3.0 


Low 


15.6 


12.6 


19.8 


11.9 


3.7 


2.9 


6.7 


3.8 


Moderate 


28.0 


20.2 


26.3 


13.2 


4.5 


3.6 


12.6 


6.3 


High 


43.4 


42.3 


35.6 


18.7 


11.3 


6.4 


20.6 


13.1 


a Groupings not comparabl 


e between periods 













Participation Patterns and Changes, 1982-1992 I 31 



FIGURE 10. Musical Participation Rates by Age, 
1982 and 1992 (Percent) 



Young Adult 30-something Middle Age Retired 




1982 1992 1982 1992 1982 1992 1982 1992 
Live Television Radio Recordings 



TABLE 21. Change in Musical Participation Rates by Age, 
1982-1992 



Age Group 



Live Television Radio Recordings 



Young adult 
30-something 
Middle age 
Retired 



-2.4 

-4.1 

1.0 

1.6 



-10.7 
-9.3 
-7.6 
-1.9 



-1.6 

-1.1 

-1.3 

1.2 



Note: Numbers indicate increase or decrease in percentage points. 



-2.2 

-3.5 

-4.3 

0.3 



32 I Turning On and Tuning In 



FIGURE 11. Musical Participation Rates by Education, 
1982 and 1992 (Percent) 



Elementary High School College 




1982 1992 1982 , 1992 1982 1992 1982 1992 
Live Television Radio Recordings 



TABLE 22. Change in Musical Participation Rates 
Education, 1982-1992 

Education 3 Live Television Radio 


by 

Recordings 


Elementary -1.0 -2.3 1.3 
High school -1.1 -6.3 -0.9 
College -5.4 -13.3 -1.9 

Note: Numbers indicate increase or decrease in percentage points. 
a "None" not included 


0.2 
-2.0 
-5.8 



Participation Patterns and Changes, 1982-1992 I 33 



FIGURE 12. Musical Participation Rates by Race, 
1982 and 1992 (Percent) 



25 



20- 



Black 



□ White 



1982 1992 
Live 




■ Other 




1982 1992 
Television 



1982 1992 
Radio 



1982 1992 
Recordings 



TABLE 23. Change in Musical Participation Rates by Race, 
1982-1992 



Race 



Live 



Television Radio Recordings 



White 
Black 
Other 



-1.7 
4.1 
-1.4 



-8.0 
-6.1 
-6.9 



-0.6 
-1.9 
-3.0 



Note: Numbers indicate increase or decrease in percentage points. 



-3.1 

1.0 

-2.9 



34 I Turning On and Tuning In 

Theater (Plays) 

Perhaps the most dramatic pair of numbers in Table 24 is the one that shows 
the total decline in the viewing of plays on television, from a rate of 25.9 percent 
in 1982 to 14.8 percent in 1992. The decline is consistent over all demographic 
groups and to some extent reflects the reduced availability of theater per se on 
television and the stiff competition from other dramatic offerings on cable 
television. 

Participation rises with education, age (except for the "retired" group), and 
income. Although participation by whites is generally higher than participation 



TABLE 24. 


Theater (Plays) Participation 


Rates by Variable, 




1982 and 1992 (Percentages 


of Adult Population) 


Variable 




Live 


Television Radi 


o 


1982 


1992 


1982 


1992 1982 


1992 


Total 


11.9 


13.5 


25.9 


14.8 3.8 


2.8 


Location 












Urban 


13.5 


15.3 


28.4 


16.4 4.3 


3.1 


Rural 


8.5 


8.5 


20.0 


10.6 2.6 


2.0 


Gender 












Male 


10.8 


12.3 


25.2 


14.0 3.7 


3.1 


Female 


12.9 


'14.6 


26.6 


15.6 3.9 


2.6 


Education 












Elementary 


1.7 


1.8 


7.2 


6.4 0.7 


1.5 


High school 


6.1 


7.0 


20.1 


11.0 2.9 


1.8 


College 


23.0 


22.4 


39.5 


20.4 6.0 


4.2 


Age 












Young adult 


11.1 


12.5 


25.3 


8.7 5.3 


2.0 


30-somethin 


g 14.3 


13.5 


27.2 


13.6 3.3 


3.4 


Middle age 


12.5 


16.2 


28.4 


19.5 3.6 


3.0 


Retired 


8.2 


10.6 


20.3 


18.6 2.4 


2.5 


Race 












White 


12.8 


13.7 


27.0 


15.2 4.0 


2.6 


Black 


5.8 


12.2 


18.2 


12.7 2.7 


4.1 


Other 


7.9 


10.5 


21.3 


12.1 2.2 


3.7 


Income 3 












Poverty 


6.1 


6.8 


16.0 


10.0 3.8 


2.4 


Low 


9.5 


9.9 


25.0 


12.2 3.8 


2.3 


Moderate 


17.8 


15.2 


33.5 


16.8 4.6 


3.1 


High 


33.1 


31.8 


34.0 


24.7 2.1 


4.2 


a Groupings not comparable between periods 







Participation Patterns and Changes, 1982-1992 I 35 



FIGURE 13. Theater (Plays) Participation Rates by Age, 
1982 and 1992 (Percent) 



30- 



20- 



Young Adult 30-something (] Middle Age Retired 



1982 



1992 



Live 




1982 1992 

Television 



1982 1992 

Radio 



B-1B1 _Tm^ 

In s in 

, m E I 1 , 111 I I L 



TABLE 25. Change in Theater (Plays) Participation Rates by 
Age, 1982-1992 



Age Group 



Live 



Television 



Young adult 
30-something 
Middle age 
Retired 



1.4 
-0.8 

3.7 
2.4 



-16.6 

-13.6 
-8.9 
-1.7 



Note: Numbers indicate increase or decrease in percentage points. 



Radio 



-3.3 
0.1 

-0.6 
0.1 



36 I Turning On and Tuning In 



FIGURE 14. Theater (Plays) Participation Rates by Education, 
1 982 and 1 992 (Percent) 



50- 



40- 



30- 



20- 



10- 







Elementary [H High School 



i 






,-- 






College 




1982 1992 . 1982 1992 

Live Television 



1982 1992 

Radio 



TABLE 26. Change in Theater (Plays) Participation Rates by 
Education, 1982-1992 



Education' 



Live 



Television 



Radio 



Elementary 
High school 
College 



0.1 
0.9 

-0.6 



-0.8 

-9.1 

-19.1 



0.8 
-1.1 
-1.8 



Note: Numbers indicate increase or decrease in percentage points. 
a "None" not included 



Participation Patterns and Changes, 1982-1992 I 37 



FIGURE 15. Theater (Plays) Participation Rates by Race, 
1982 and 1992 (Percent) 



Black 



White 



Other 




1982 1992 1982 1992 1982 1992 

Live Television Radio 



TABLE 27. Change in Theater (Plays) Participation Rates by 
Race, 1982-1992 



Race 



Live 



Television 



White 
Black 
Other 



0.9 
6.4 
2.6 



-11.8 
-5.5 
-9.2 



Note: Numbers indicate increase or decrease in percentage points. 



Radio 



-1.4 
1.4 
1.5 



38 I Turning On and Tuning In 



by other groups, black attendance at live performances and participation via 
radio rose from 1982 to 1992. Furthermore, the decline in black participation 
via television was not so pronounced as declines in this category overall. It is 
tempting to attribute this to the growth of black theater, the development of 
black acting talent, and the success of such playwrights as August Wilson. 

Ballet/Dance 

The 1982 survey respondents were asked about participation in ballet, 
excluding such other dance forms as modern, jazz, Eastern, and so forth. This 
was partially rectified in the 1992 survey, when a question regarding live 
participation in dance, broadly defined, was added. Unfortunately, that means 
that while live ballet participation can be compared over time, media participa- 
tion cannot be strictly compared over time, nor can media participation be 
strictly compared with live participation for 1992. Nevertheless, some generali- 
zations may be possible. 

In both years, participation rises with education and income, as shown in 
Table 28. Live participation is slightly higher within the two younger age groups, 
while participation by television tends to peak in the middle-age group. 

Over the decade, overall ballet participation rose slightly, and participation 
within most individual groupings rose as well. Not surprisingly, participation 
in dance, more broadly defined, was higher than participation in just ballet in 
1992. The only anomaly was in the highest income grouping, where ballet 
participation exceeded dance participation. Interestingly, nonwhite live partici- 
pation in dance exceeds white participation in 1992, suggesting that the broader 
category captures ethnic exposure that is obscured by a focus on ballet. Media 
participation rates in dance in 1992 are very similar across the racial groupings 
depicted. Dance Theater of Harlem and the Alvin Ailey Company are among 
the best-known examples of primarily black performing groups, and other black 
choreographers, such as Bill T. Jones, have become very popular. One would 
presume that these companies and artists have found an audience. 



Participation Patterns and Changes, 1 982-1 992 I 39 



TABLE 28. Ballet/Dance Participation Rates by Variable, 

1982 and 1992 (Percentages of Adult Population) 



Variable 



Live 



Television' 



1982 Ballet 1992 Ballet 1992 Dance 1982 



1992 



4.2 



4.6 



Total 

Location 
Urban 
Rural 

Gender 
Male 
Female 

Education 
Elementary 
High school 
College 

Age 
Young adult 
30-something 
Middle age 
Reti red 

Race 

White 
Black 
Other 

Income 
Poverty 
Low 

Moderate 
High 



'Data are not strictly comparable over time. 
Groupings are not comparable between periods. 



4.9 


5.4 


2.6 


2.6 


2.7 


3.6 


5.6 


5.6 


0.4 


0.7 


2.0 


2.0 


8.4 


8.1 


4.1 


4.9 


5.7 


5.0 


3.7 


4.9 


2.5 


3.2 


4.5 


4.9 


1.8 


2.6 


3.4 


5.8 


2.2 


2.6 


3.5 


2.5 


5.9 


5.5 


10.7 


12.3 



7.1 



16.3 



17.2 



7.6 


17.7 


18.5 


5.9 


13.1 


13.7 


6.7 


12.1 


14.8 


7.5 


20.1 


19.4 


2.5 


7.6 


11.0 


4.3 


11.6 


13.1 


10.9 


25.4 


22.8 


7.3 


12.7 


12.6 


7.8 


17.3 


16.5 


7.2 


19.8 


20.3 


5.4 


15.4 


20.3 


7.0 


16.8 


17.2 


7.3 


10.3 


16.5 


9.9 


26.3 


18.8 


4.6 


10.8 


14.2 


6.3 


15.3 


17.0 


8.2 


21.4 


17.3 


9.5 


28.8 


25.0 



40 I Turning On and Tuning In 



FIGURE 16. Ballet/Dance Participation Rates by Age, 
1982 and 1992 (Percent) 



Young Adult 30-something I Middle Age Retired 




1 982 Ballet 1 992 Ballet 1 992 Dance 1 982 

Live 



1992 



Television 



TABLE 29. Change in Ballet/Dance Participation Rates by 
Age, 1982-1 992 a 

Live 



Age Group 



Ballet Only Ballet/Dance 



Television 3 
(Ballet/Dance) 



Young adult 
30-something 
Middle age 
Retired 



0.8 

-0.7 

1.2 

0.7 



3.2 
2.1 
3.5 
2.9 



-0.1 

-0.8 

0.5 

4.9 



Note: Numbers indicate increase or decrease in percentage points. 
a Data are not strictly comparable over time. 



Participation Patterns and Changes, 1982-1992 I 41 



FIGURE 17. Ballet/Dance Participation Rates by Education, 
1982 and 1992 (Percent) 



30- 



25 



20- 



15 



10- 



Elementary High School I College 



- 




; 


J 
II 






';""""" " """""" 

'■:'■ . : ■■'■ ' ■ ,■■ 



1982 Ballet 1992 Ballet 1992 Dance 1982 1992 

Live Television 



Table 30. Change in Ballet/Dance Participation Rates by 
Education, 1982-1 992 a 



Live 



Education 


Ballet 


Dance a 


Television 
(Dance Only) 


Elementary 
High school 
College 


0.3 

0.0 

-0.3 


2.1 
2.3 
2.5 


3.4 

1.5 

-2.6 



Note: Numbers indicate increase or decrease in percentage points. 
a Data are not strictly comparable over time. 
"None" is not included. 



42 I Turning On and Tuning In 



FIGURE 18. Ballet/Dance Participation Rates by Race, 
1 982 and 1 992 (Percent) 



30- 



25- 



20- 



15- 



10- 







Black 



QJ White 



Other 





1 982 Ballet 1 992 Ballet 1 992 Dance 
Live 



1982 1992 

Television 



TABLE 31. Change in Ballet/Dance Participation Rates by 
Race, 1982-1992 

Live 



Race 



Ballet Only Ballet/Dance' 



Television 3 
(Ballet/Dance) 



White 
Black 
Other 



0.4 
0.8 
2.4 



2.5 
5.5 
6.5 



0.4 
6.2 

-7.5 



Note: Numbers indicate increase or decrease in percentage points. 
a Data are not strictly comparable over time. 



Participation Patterns and Changes, 1982-1992 I 43 



Art 



In contrast to the preceding performing arts data and discussion, both 
attendance at art museums and participation via television (watching programs 
about art, artists, or art museums and galleries) have risen over time for virtually 
every demographic category (see Table 32). Participation rates are high and 
getting higher. Even male and female participation rates are nearly on a par. 



TABLE 32. Art 


Participation Rates by 


Variable, 


1982 


and 




1 992 (Percentages 


» of Adu 


It Population) 






Variable 


Live 




Television 




1982 


1992 


1982 




1992 


Total 


22.1 


26.7 


22.8 




30.1 




Location 














Urban 


24.6 


29.8 


24.7 




31.2 




Rural 


16.8 


18.5 


18.4 




27.3 




Gender 














Male 


21.0 


26.4 


23.1 




30.1 




Female 


23.1 


26.9 


22.6 




30.1 




Education 














Elementary 


2.5 


3.8 


4.6 




13.3 




High school 


13.5 


14.4 


18.5 




23.3 




College 


40.0 


43.6 


34.0 




40.2 




Age 














Young adult 


24.3 


29.1 


22.0 




26.8 




30-something 


26.9 


29.6 


24.5 




31.7 




Middle age 


20.4 


27.3 


25.7 




34.0 




Retired 


12.3 


16.3 


16.5 




25.1 




Race 














White 


23.2 


27.6 


23.2 




31.3 




Black 


12.3 


19.0 


19.4 




23.3 




Other 


27.5 


28.4 


24.6 




21.5 




Income 3 














Poverty 


12.8 


13.4 


14.4 




21.8 




Low 


19.6 


20.6 


20.9 




27.4 




Moderate 


30.7 


31.6 


32.8 




33.0 




High 


47.2 


49.8 


42.9 




44.1 




a Groupings not com 


parable between periods 











44 I Turning On and Tuning In 



FIGURE 19. Art Participation Rates by Age, 
1982 and 1992 (Percent) 



Young Adult 30-something Middle Age Retired 




1982 1992 

Live 



1982 



1992 



Television 



TABLE 33. Change in Art Participation Rates by Age, 
1982-1992 



Age Group 



Live 



Television 



Young adult 
30-something 
Middle age 
Retired 



4.8 
2.7 
6.9 
4.0 



4.8 
7.2 
8.3 
8.6 



Note: Numbers indicate increase in percentage points. 



Participation Patterns and Changes, 1982-1992 I 45 



FIGURE 20. Art Participation Rates by Education, 
1982 and 1992 (Percent) 



50- 



40- 



30- 



20- 



10- 










'mm-. 



■ 



1982 



Live 



Elementary High School College 



wmwm 






1992 




.':■■'.' 



1982 1992 

Television 



TABLE 34. Change in Art Participation Rates by Education, 
1982-1992 



Education' 



Live 



Television 



Elementary 
High school 
College 



1.3 
0.9 
3.6 



8.7 
4.8 
6.2 



Note: Numbers indicate increase in percentage points. 
a// None" is not included. 



46 I Turning On and Tuning In 



FIGURE 21. Art Participation Rates by Race, 
1 982 and 1 992 (Percent) 



Black 



|_1 White 



1982 1992 

Live 



Other 




1982 1992 

Television 



TABLE 35. Change in Art Participation Rates by Race, 
1982-1992 



Race 



Live 



Television 



White 
Black 
Other 



4.4 
6.7 
0.9 



8.1 

3.9 

-3.1 



Note: Numbers indicate increase or decrease in percentage points. 



Participation Patterns and Changes, 1982-1992 I 47 



Participation rates rise with education and income, while participation by 
age peaks in the 30-something to middle-age range. 



Summary and Conclusions 

The findings in this chapter lend credence to a number of conclusions. 
Participation in all art forms and media tend to rise with income, education, 
and age. Media participation rates, although subject to the same general 
influences, are by and large higher than live attendance rates, probably reflecting 
the lower costs of participating from, say, one's living room. Except for jazz, 
female and white participation rates exceed the alternatives, while urban par- 
ticipation uniformly exceeds rural. 

The measures of association, which suggest that observed relationships are 
not simply random, leave some questions unanswered. It is not clear, for 
example, whether the high association between education and participation is 
due to well-developed tastes for art and culture, or due to the higher incomes — 
and ability to pay — of those with higher educations. Part III resolves this 
difficulty. 



Multivariate 
Statistical Results 




A 



mong the questions posed — and answered — in this part are the following: 



■ What is the impact of any one variable on arts participation via media, 
controlling for all other influences? As indicated in the last chapter, failure 
to control for possible simultaneous influences could lead to erroneous 
inferences. 

■ Do multivariate methods further indicate that patterns of participation via 
media differ significantly from those for live participation? 

■ Have multivariate results altered significantly from 1982 to 1992? 

Multivariate statistical techniques are most appropriate to address questions 
such as these. Such techniques generate estimates of the parameters, or relation- 
ships, among a number of variables simultaneously. The specific technique 
used here is logistic regression, which deals with categorical response variables. 
An example of a categorical response variable is gender, where a respondent can 
be either male or female, i.e., fall into one category or the other. This contrasts 
with more nearly continuous variables, such as household income, which can 
range in value from zero to millions of dollars or higher. But even these latter 
can be grouped into categories for statistical purposes, as has been done in this 
document. Statistical results, literal interpretations, and inferences are indicated 
in the sections that follow. For a more technical discussion of the techniques, 
including indicators of statistical significance, see Appendix B. 

The estimating equation for multiple regression takes the general form 

P t =f{PpG;D) 

where Pi is the live (media) participation measure for activity i (opera, for 
example); Pj is the participation measure for a corresponding media (live) 
activity j; G is a geographic identifier (e.g., urban); and D is a vector of 
demographic characteristics, including those of particular interest for this 
project (income, education, etc.). A positive sign for the estimated coefficient 
of Pj indicates complementarity, while a negative sign would indicate the two 
are substitutes. 

Separate regressions for each of the survey periods will support further 



48 



Multivariate Statistical Results I 49 



analyses of changes over time. 1 A review of general patterns will offer some 
indications as to whether participation patterns have altered over time. 



Logistic Models 

Logistic regression estimates the probability of an event occurring, for 
example, of an individual attending a live performance or participating via some 
media form. The logistic model can be written in terms of the "log of the odds," 
which is called a logit: 

. Prob {event) __ __ 

lo g U // ^r) = %> + B ^ + ■ • ■ + BnK n 

Pro b {no event) 

An equivalent expression is 

Prob {event) = 



I + fiBa + BiXi . . .) 



which we can use to construct an illustrative example. Borrowing some values 
from the third column of Table 37, we seek to determine the probability that 
a 30-something black male with moderate income and a high school education 
would view jazz on television in the survey period. The calculation is 

Prob (jazz on TV) = — - - — = .3253 

v ' , -(-3.5851 + .4623(1) + .1407(1) + .3419(1) + 1.1449(1) -.1172(1) + .8830(1)) 

which means that there is about a one-third, or one out of three, chance that 
the person described above would have watched jazz on television during the 
survey period. In contrast, the corresponding probability for an otherwise 
similar white respondent is slightly less than 25 percent. 

A positive value of Bi increases the log of the odds, while a negative value 
decreases the log of the odds. Put less precisely but more simply, a positive 
coefficient indicates that the variable increases the probability of participation, 
while a negative coefficient decreases the probability. Furthermore, as the 
absolute value of a coefficient increases, the impact of that variable likewise 
increases. For example, a higher value of the coefficient of one age group 
indicator relative to another indicates that someone in the first age group is more 
likely to participate. 



50 I Turning On and Tuning In 



Overall Logistic Regression Results 

Table 36 lists and briefly defines the variables included in the logistic 
regressions — those variables that may influence consumer tastes or consumer 
willingness to pay with respect to each of the benchmark arts. Such variables 
include demographic descriptors of respondents as well as characteristics of their 
locations. Gender, race, and education levels are among the variables that may 
be taken to indicate tastes, while income reflects both tastes and the ability to 
pay for arts participation. 17 The third column in Table 36, "Reference," simply 



TABLE 36. Variables, Definitions, and References 



Variable 


Brief Description 


Reference 


Urban 


Resides in urban area 


Not urban 


Male 


Self-explanatory 


Female 


Low income 


Household income range just 
above poverty 


Poverty income 


Moderate 
income 


Household income range below 
the highest 


Poverty income 


High income 


Household income over $75,000 
(1982) or $100,000 (1992) 


Poverty income 


High school 


Between 9 and 1 2 years of 
education - 


Elementary 


College 


More than 1 2 years of education 


Elementary 


30-something 


Age between 30 and 44 


Young adult (18-29) 


Middle age 


Age between 45 and 64 


Young adult 


Retired 


Age 65 or greater 


Young adult 


Black 


Self-explanatory 


White 


Asian 


All Asian 


White 


Indian 


American Indian, Eskimo, or Aleut 


White 


Hispanic 


Of Hispanic origin 


Not Hispanic 


Live 


Participant via live performance 
of art form in question 


Not live 


Television 


Participant via television in art 
form in question 


Not television 
participant 


Radio 


Participant via radio in art form 
in question 


Not radio 
participant 


Recording 


Participant via audio recording 
in art form in question 


Not recording 
participant 



Multivariate Statistical Results 51 



identifies the variable category against which the categories in the first column 
are compared to determine the coefficients. 

One might assume, further informed by the results reported in Part II, the 
following: urban respondents would be more likely to take advantage of live 
participation simply because of the greater likelihood of availability; those with 
higher education levels may have developed tastes more consistent with complex 
preferences; older respondents would have more time to develop complex tastes 
(or to move from less complex to more complex music). The expected impact 
of race would for the most part be less clear, although African Americans may 
display a disproportionate interest in jazz, traditionally a black art form. 
Ordinarily, individuals with higher income levels would participate more 
because of ability to pay; certainly this might be especially true in the case of 
live performances or purchase of recordings, both of which entail an explicit 
price or user charge. But both television (broadcast as well as many cable) and 
radio are known as "collective goods," wherein after some initial outlay (pur- 
chase of a receiving set, installation of cable access), subsequent viewing entails 
no explicit user charge. In these instances, income may not be quite so important 
a determinant. With those not too unreasonable expectations, we move to the 
statistical results. 

Significance of individual coefficients is based on the Wald statistic, which 
permits a test of whether a coefficient is 0, i.e., has no influence. 1 If the test is 
significant at, say, the 5 percent level, this means that there is only a 5 percent 
chance of the observed outcome occurring randomly. Put another way, there is 
a 95 percent chance of a nonrandom relationship between the variables. 

Several tests or indicators of overall significance of models as specified are 
reported in the following tables. 19 These include the likelihood test (—2 log 
likelihood), which tests how well the model classifies the data in comparison to 
a "perfect" model. A model that classifies especially well would be described as 
not significantly different from the perfect model. As it happens, all of the results 
reported here do differ significantly from the perfect model. This means that 
many factors other than those accounted for in this study influence arts 
participation. 

A second test, the model chi-square, indicates whether the included variables 
as a group are significantly different from zero. Without exception, this test 
indicates that the included variables are significant. Finally, the third indicator, 
classification percentage, reveals the overall classification accuracy of the models. 
A value of 83.6, for example, means that 83.6 percent of the respondents are 
correctly classified as participants or nonparticipants by the model. The values 
for this indicator for most of the models reported here are generally in the 80s 
and 90s. 



52 I Turning On and Tuning In 



Jazz 



Tables 37 and 38 contain the results of logistic regressions for live and media 
participation in jazz for 1982 and 1992, respectively. As indicated by the 
summary statistics, the overall influence of the included variables is statistically 
significant. 



TABLE 37. 


Logistic Regression, Jazz Participation, 1982 


Independent 










Variable 


Live 


Television 


Radio 


Recording 


Urban 


.4576*** 


.4623** 


.4563*** 


.6214*** 


Male 


.0224 


.1407 


.2854*** 


.0289 


Low income 


- .0894 


.2511* 


- .1413 


.3373*** 


Moderate income .0908 


.3419** 


- .1559 


.4715*** 


High income 


.4871 *** 


.4562** 


- .1129 


.7456*** 


High school 


1 .0946***' a 


1 i44g*** 


.8926*** 


1.1475*** 


College 


2 11 ^Q***' 3 


1.7933***' 3 


1 .5900*** 


1.9566*** 


30-something 


- .6907***' a 


- .1172 


- .4750***' a 


- .4961***' a 


Middle age 


-1.0330***' a 


.051 8 a 


- .5473***' a 


— .481 1 *** 


Retired 


-1 .8895***' a 


- .4462*** 


-1.3526***' 3 


—1 2431 *** ,a 


Black 


.7179*** 


.8830*** 


1.2195*** 


1.2675***' 8 


Indian 


- .15301*** 


.2177 


.2814 


- .2216 


Constant 


-3.6688*** 


-3.5851*** 


-2.8000*** 


-3.4061*** 


-2 Log likelihood 8706.671 b 


3158.659 b 


3041 .379 b 


3173.020 b 


Model chi-square 1134.959*** 


215.488*** 


323.466*** 


405.486*** 


Classification 










percentage 


90.51 


82.74 


83.03 


80.88 


***Significance 


greater than .01 








**Significance 


greater than .05 








*Significance 


greater than .10 








a Significantly different from 1 992 coefficient at 5 percent level 




Significantly different from the "perfect 


" model 







Live Participation 

Several factors stand out in the equations for live attendance. The first is 
that an urban location enhances the likelihood of attendance, in part surely due 
to convenience. Urban areas are simply more likely to offer opportunities for 
live attendance. Men are more likely to participate than women, and this 
tendency became more pronounced in 1992. Income, as measured here, seems 
to be a significant factor; each of the included groups participate significantly 
more than the reference (poverty level) group. This is especially true in 1992, 



Multivariate Statistical Results I 53 



TABLE 38. 


Logistic Regressioi 


i, Jazz Participation, 1992 


Independent 
Variable 


Live 


Television 


Radio 


Recording 


Urban 


.4569*** 


.3552*** 


.4165*** 


.4611*** 


Male 


.1278** 


.1478*** 


.1 748*** 


.1490*** 


Low income 


.3718*** 


2782*** 


.2346*** 


.3445*** 


Moderate income 4791*** 


2298*** 


.3615*** 


.5386*** 


High income 


.8732*** 


.4892***' b 


.7131*** 


1.0112*** 


High school 


1.5803***' a 


.7304*** 


.6453***' b 


.8146*** 


College 
30-something 


2.8647***' a 
- .1529*' a 


1 rr^oo***' 3 ' 

33 43***,b 


1.5518***' b 
1379 **,a,b 


1.8546***' b 
- .0302 a ' b 


Middle age 


- .4083***' a 


4290***' a 


- .1507**' a ' b 


- .3084*** 


Retired 


- .8181***' a 


4231 ***' a 'b 


_ 4434***' a '' 3 


- .6617***' a 


Black 


.6677*** 


.8035*** 


.9604***' b 


9787***' a 


Asian 


- .8751*** 


- .2665 b 


- .1982 b 


- .4839*** 


Indian 


- .0628 


.0694 


- .1917 


.1783 


Hispanic 


— .4710*** 


.0306 b 


.1042 b 


- .1335 b 


Constant 


-5.0870*** 


-3.5984*** 


-2.7788*** 


-3.5469*** 


-2 Log likelihood 6784.577 c 1 021 1 .246 c 


12351.924 c 


10095.600° 


Model chi-square 844.977*** 


601 .602*** 


1199.478*** 


1233.858*** 


Classification 










percentage 


89.72 


82.13 


73.02 


80.37 


***Significance 


greater than .01 








**Significance 


greater than .05 








*Significance 


greater than .10 








a Significantly different from 1992 coefficient at 5 percent level 
Significantly different from corresponding "live" coefficient at 5 percent level 
c Significantly different from the "perfect" model 



but this may be an artifact of the income groups chosen for the various 
categories. The likelihood of attendance increases with education; both the high 
school and college groups are more likely to attend than those in the reference 
(elementary) group, with the college coefficient larger than the high school 
coefficient. The education coefficients rose significantly from 1982 to 1992, 
suggesting education's increased importance as a factor in attendance at jazz 
concerts. 

Attendance at jazz performances also seems to be a young person's enter- 
tainment. Attendance is less likely among those in older age groups, with the 
negative coefficient progressively rising from the 30-somethings to the retirees. 
Interestingly, age declined in importance, as indicated by size of coefficient, 
from 1982 to 1992. Race indicators are mixed; black respondents participate 
significantly more than whites; Asians participate significantly less; and Hispan- 
ics participate less than non-Hispanics, the majority of whom are white. 



54 I Turning On and Tuning In 



Television 



Participation in jazz via television is also higher in urban areas than in rural 
areas; gender, not significant in 1982, is significant a decade later; and higher 
incomes increase participation. In 1992, the income coefficients for television 
participation are much lower than for live participation, indicating that higher 
incomes are more likely to influence attendance at a performance than they are 
to influence television viewing of jazz. Another rather stark contrast is the 
positive and significant impact of higher age on television participation in 1992. 
One might suppose on this basis that the taste for jazz rises with age, but that 
this is in part offset by the reduced willingness to travel to live venues. The results 
also support the prior finding that modes of participation are complements as 
opposed to substitutes. Black respondents are more likely to participate via 
television than are whites, but other racial categories are insignificant. 

Radio 

Participation via radio reflects some similar influences: urban residents are 
more likely to listen to jazz than are nonurban residents; males are more likely 
to listen than are females. The role of income on participation via radio is 
insignificant in 1982, but participation clearly rises with income level in 1992. 
Education contributes strongly to participation in both time periods. The age 
measures show a mixed pattern; participation declines with age in 1 982 but does 
so only for the two older groups in 1992. The 30-somethings are slightly more 
likely to participate than the young adults in the latter period. 

Recordings 

Although jazz recordings are ubiquitously available — that is, not limited to 
urban areas — urban respondents are more likely than nonurban respondents to 
participate via this means. High-income groups are more likely to participate 
by this means as well, reflecting the fact that recordings are not a collective good. 
Participation also generally rises with education but decreases with age, similar 
to the trends for live participation. Black respondents are more likely than whites 
to participate via recordings, while other race variables show a mixed response. 

Jazz Summary 

The results for jazz participation are generally consistent with earlier find- 
ings. Participation is higher for younger, black, urban males with higher incomes 
and education than for other groups. Income is not so important for the 
collective goods (television and radio) as for the more nearly private goods 
(recordings). 



Multivariate Statistical Results I 55 



Classical Music 

Tables 39 and 40 show the logistic regression results for classical music 
participation. As indicated by the summary statistics, the set of included 
variables is significant in both time periods. 

Live Participation 

Attendance at live classical music performances is higher in urban areas and 
among females. It rises with income, education, and age, and is generally lower 
for nonwhites. The urban and male coefficients have declined in absolute value 
over time, as have the education measures. The age coefficients have risen 
in absolute value, however, indicating an aging of the audience for live 
performances. 



TABLE 39. 


Logistic Regression, Classical Music Participation, 




1982 








Independent 










Variable 


Live 


Television 


Radio 


Recording 


Urban 


.4643***' a 


.3973*** 


.3619*** 


.1806* 


Male 


- .5046***' a 


— 2592*** 


- .0952 


- .2559*** 


Low income 


- .0345 


.4604*** 


.1037 


.2267** 


Moderate income .2528*** 


.4584*** 


.2409* 


.2829** 


High income 


.7160*** 


5977*** 


.6429*** 


.4345** 


High school 


1.4159*** 


.8842*** 


.4059** 


.8024*** 


College 


2.9238***' a 


1.8330*** 


1 .6040*** 


2.1356*** 


30-something 


.3045***' a 


.3819*** 


.4624*** 


OTQ1 ***/3 


Middle age 


.4371*** 


111 95*** 


.6633*** 


.6046***' a 


Retired 


.4976*** 


Q Cj7t ***/3 


.4449*** 


.1168 


Black 


- .5752*** 


- .3212** 


- .1419' a 


- .4815*** 


Indian 


- .7764*** 


.2825 


.2186 


.1494 


Constant 


4.4824*** 


-3.4511*** 


-3.1031*** 


-3.1406*** 


-2 Log likelihood 1 0596.901 b 


3744.989 b 


3298.71 3 b 


3400.933 b 


Model chi-square 1619.738*** 


379.726*** 


299.813*** 


392.946*** 


Classification 










percentage 


86.91 


75.87 


80.51 


78.67 


***Significance 


greater than .01 








**Significance 


greater than .05 








*Significance 


greater than .10 








a Significantly different from correspond 
Significantly different from the "perfecl 


ng 1 992 coefficient at 5 percent 


level 


" model 







I I 



56 I Turning On and Tuning In 



TABLE 40. Logistic Regression, Classical Music Participation, 


1992 








Independent Variable Live 


Television 


Radio 


Recording 


Urban 2 971***' a 


.2810*** 


.2500*** 


.2403*** 


Male - ,3224***' a 


- .2105*** 


_ i 3 -| 6 ***'b 


- .1904***' b 


Low income .3312*** 


.2462*** 


.0896 


.2387*** 


Moderate income .5969*** 


.3496***' b 


.3104***' b 


.5637*** 


High income 1.0581*** 


.7460***' b 


.7073***' b 


1.0005*** 


High school 1.0906*** 


.8481*** 


.6299*** 


.8935*** 


College 2.4932***' a 


1.6813***' b 


1.7815***' b 


2.1262*** 


30-something .071 1 a 


.3541***' b 


3784***' b 


.0674 a 


Middle age .5792*** 


.8815***' b 


.6507*** 


2171 ***> a /D 


Retired .6897*** 


1 31 95***' a ' 


.5091*** 


1 5gg**/b 


Black - .5042*** 


- .1816**' b 


- .5413***' a 


- .7512*** 


Asian - .1443 


.2017 


.0577- 


.0289 


Indian - .1252 


.5377* 


.2599 


- .1578 


Hispanic - .4664*** 


.0031 b 


- .03623' b 


- .1618 


Constant -4.6564*** 


-3.5044*** 


-2.6868*** 


-3.2770*** 


-2 Log likelihood 771 1 .036 c 1 1 351 .607 c 


12877.276 c 


10871.908*** 


Model chi-square 1 01 9.939*** 


893.472*** 


1368.775*** 


1357.308*** 


Classification 








percentage 87.39 


77.50 


70.52 


76.33 


***Significance greater than .01 








**Significance greater than .05 








*Significance greater than .1 








a Significantly different from correspondi 


ng 1 982 coefficient at 5 percent 


level 


Significantly different from correspondi 


ng 'live" coeffi 


cient at 5 percent 


level 


Significantly different from the "perfect 


" model 







Television 

Television participation patterns are generally similar to those for live 
participation. The urban coefficient is lower in both time periods than in the 
corresponding equations for live participation (and falls from 1982 to 1992), 
reflecting the fact that television signals are not restricted to urban areas. Men 
are not so unlikely to watch classical music on television as they are to attend a 
live performance. The income variables are not so influential as compared with 
live participation in 1992, reflecting the public-goods nature of television. Age 
patterns are similar across periods except for the retired group, where a signifi- 
cantly higher value in 1992 may reflect an aging audience. 



Multivariate Statistical Results 57 



Radio 

Participation by radio is quite similar to participation by television. Income 
coefficients are lower than those in corresponding equations for live and 
television participation, a result consistent with lower user costs of radio. Urban 
residence, while significant in both periods, has become less important over 
time, although the difference is not significant. Radio participation peaks in the 
middle-age group. Black participation relative to white fell over the decade, 
suggesting that this art form is not addressing the tastes of blacks via this 
medium. 

Recordings 

Participation in classical music via recordings shows familiar patterns, with 
the likely participant being urban, female, higher income, better educated, 
white, and somewhat older — except for the retired group. Few significant 
changes have occurred over time, the exceptions being smaller age coefficients. 

Classical Music Summary 

Overall, the urban and gender factors have diminished over the period from 
1982 to 1992. Attendance at live performances increases with income, educa- 
tion, and age; participation via television, radio, and recordings shows generally 
similar patterns. Of special note is the relatively small negative constant term 
for the 1992 radio equation. This means that the base level of participation in 
classical music via radio is higher than in 1982 and is also higher than live and 
other media participation rates. 



Opera 

Tables 41 and 42 contain the logistic regression results for opera. 

Live Participation 

With the exception of the urban coefficient, which fell by half from 1982 
to 1992, opera participation patterns are generally stable over time. Males do 
not participate as much as females; participation rises with income, education, 
and age; and nonwhites do not participate as much as whites. The lower urban 
coefficient in 1 992 suggests that an urban setting is not so powerful in affecting 
relative participation as it once was. 



58 I Turning On and Tuning In 



TABLE 41. 


Logistic Regressi 


on, Opera Participation, 


1982 


Independent 










Variable 


Live 


Television 


Radio 


Recording 


Urban 


g-j gg***,a 


5297*** 


.3452** 


.5273*** 


Male 


- .4112*** 


- .3002*** 


- .0110 


- .1810 


Low income 


- .0448 


.4595*** 


.1242 


.4381** 


Moderate income .2946* 


.5560*** 


.1482 


.4137* 


High income 


1.0805*** 


.7678*** 


.5067* 


.5005 


High school 


1.1090*** 


1 .0070*** 


.1440 


.6265** 


College 


2.4182*** 


1.8099*** 


1.3479*** 


1.6312*** 


30-something 


.2861** 


.6698*** 


.2109 


8986***' a 


Middle age 


.6486*** 


1 .4242*** 


9901 *** 


1.5850***' 3 


Retired 


.8679*** 


1.4908*** 


9451 ***' a 


1.2958*** 


Black 


- .6815*** 


- .1275 


- .2651 


- .3314 


Indian 


- .6340* 


.5532* 


.5493 


.1399 


Constant 


-6.2451*** 


4.841 9*** 


-4.2640*** 


-5.2711*** 


-2 Log likelihood 3745.31 7 b 


2494.523 b 


1675.963 b 


1727.608 b 


Model chi-square 430.128*** 


247.983*** 


125.651*** 


159.043*** 


Classification 










percentage 


97.05 


87.62 


93.25 


92.84 


***Significance 


greater than .01 








**Significance 


greater than .05 








*Significance 


greater than .10 








a Significantly different from correspor 


iding 1992 coefficient at 5 percent 


level 


Significantly different from the "perfect" model 


- 





Television 

Participation in opera via television is, by and large, similar to live partici- 
pation. The age coefficients for television are significantly higher than those for 
live participation. This indicates that, controlling for the other influences, older 
groups are much more active participants than the young adults, whose televi- 
sion tastes may run in other directions. Television participation does not change 
significantly over time. 



Radio 

The major distinction in opera participation via radio is that the income 
variables are not so significant, reflecting perhaps the greater relative accessibility 
of radio broadcasting. The age coefficients are higher than those for live 
attendance, indicating that young adults likely have alternative radio listening 
habits. 



Multivariate Statistical Results I 59 



TABLE 42. Logistic Regression, Opera Participation, 1992 



Independent 
Variable 



Live 



Television 



Radio 



Recording 



Urban 


4093***' a 


.3744*** 


.2144** 


.2558*** 


Male 


- .2195** 


- .1998*** 


- .1415** 


- .2315*** 


Low income 


.0283 


.0924 


- .0028 


.2105* 


Moderate income 


.3441* 


.2080** 


.2986*** 


.6083*** 


High income 


1.3420*** 


.3800***' b 


.6434***' b 


1.0155*** 


High school 


.6250 


7228*** 


.6804*** 


.6558** 


College 


1.8462*** 


1 .6444*** 


1.7754*** 


1.7693*** 


30-something 


.1067 


.4603*** 


.6052***' b 


.2905**' a 


Middle age 


cr o I 7*** 


1.1165***' b 


1.3589***' b 


8990***' 


Reti red 


.6598*** 


1 .41 82***' b 


1.6100***' a ' b 


.9703*** 


Black 


- .5346** 


.0599 b 


- .1999 


- .5127*** 


Asian 


.3332 


.0002 


- .301 3 b 


.2734 


Indian 


.2954 


.7045* 


.0210 


- .3933 


Hispanic 


- .2668 


.3448***' b 


.1417 


.0983 


Constant 


-5.5861*** 


-4.4598*** 


-4.8690*** 


-4.971 7*** 


-2 Log likelihood 


3064.574° 


7402.027 c 


6256.631 c 


5205.818*** 


Model chi-square 


315.855*** 


521.259*** 


589.950*** 


469.359*** 


Classification 










percentage 


96.65 


89.09 


91.18 


93.00 



^Significance greater than .01 

**Significance greater than .05 
*Significance greater than .1 
Significantly different from corresponding 1 982 coefficient at 5 percent level 
^Significantly different from corresponding "live" coefficient at 5 percent level 

Significantly different from the "perfect" model 



Recordings 

The higher user cost of recordings accounts for the higher coefficient values 
for the income variables, compared with other media alternatives. Otherwise, 
participation patterns as reflected in coefficient signs and magnitudes are similar 
to those for other forms of opera participation. The 30-something and middle- 
age coefficients are significantly smaller in 1992, indicating a reduced influence 
of aging relative to the young adults. 

Opera Summary 

Opera participation patterns are very similar to those for classical music. 
Participation rises with age, income, and education for both live and media 
alternatives. The urban coefficients fell from 1982 to 1992, suggesting that 
proximity may have declined in importance over the decade. Income is not so 



60 I Turning On and Tuning In 



important for the broadcast (television and radio) media as it is for live and 
recorded media. 



Musicals 

Tables 43 and 44 display the results of the logistic regressions for musical 
participation. 

Live Participation 

Attendance at musicals is higher in urban areas and among females. Atten- 
dance also rises with income, education, and age up to the retired years, when 
it falls off a bit. One major distinction between 1982 and 1992 is that in 1992 
blacks are less likely to stay away compared with whites. 



TABLE 43. 


Logistic Regression, Musical 


Participation; 


1982 


Independent 










Variable 


Live 


Television 


Radio 


Recording 


Urban 


5928***' a 


594-] ***' a 


.5132** 


.7276*** 


Male 


- .4882*** 


— 2722*** 


.0633 


- .5135*** 


Low income 


.3745*** 


.4576*** 


.2311 


.1731 


Moderate income .8308*** 


.6654*** 


.1757 


.5564*** 


High income 


1.3870*** 


.9291 *** 


.8872** 


.7904*** 


High school 


.8800*** 


.5881*** 


1.0183** 


.6911* 


College 


2.0884*** 


1.4188*** 


1.9627*** 


2.0047*** 


30-something 


1 832***' a 


.1019 a 


.0932 


.3195* 


Middle age 


oil t*** 


,4846***' a 


.7480*** 


.8087*** 


Retired 


.1892** 


.6165***' a 


.7293** 


.0067 


Black 


- .5642***' a 


.0081 


.2656 


-1.3796*** 


Indian 


- .9396*** 


- .1681 


.7062* 


- .6464 


Constant 


-3.7177*** 


-3.2848*** 


-5.6813*** 


-4.6338*** 


-2 Log likelihood 131 72.1 62 b 


3464.1 57 b 


1130.026 b 


1847.202 b 


Model chi-square 1987.318*** 


244.974*** 


77.233*** 


259.908*** 


Classification 










percentage 


81.37 


79.37 


96.09 


91.58 


***Significance 


greater than .01 








**Significance 


greater than .05 








*Significance 


greater than .10 








a Significantly different from correspond 


ng 1 992 coeff 


cient at 5 percent 


eve I 


Significantly different from the "perfect 


" model 







Multivariate Statistical Results 61 



TABLE 44. 


Logistic Regression, Musical 


Participation, 


1992 


Independent 










Variable 


Live 


Television 


Radio 


Recording 


Urban 


.3355***' a 


.2678***' a 


.1805 b 


.5005***' b 


Male 


- .4570*** 


— 1 83?*** 


- .1 51 3 b 


— 2693*** 


Low income 


.4733*** 


.1715*' b 


.0153 b 


.0446 b 


Moderate income .8235*** 


.1943**' b 


.0993 b 


.2728*' b 


High income 


1.5603*** 


.3258***' b 


.3082 b 


.6438***' b 


High school 


9202*** 


.6334*** 


.4395 b 


.2658 b 


College 


1.8835*** 


1.3496*** 


1.5587*** 


1.6190*** 


30-something 


- .0032 a 


.3853***' a ' b 


.4395** b 


.1556 b 


Middle age 


.3085*** 


oyn -^***/3/D 


.9808*** b 


.4583*** 


Retired 


.2206** 


1 .2461 *** ,a '' D 


1 .2471 *** b 


.1966 


Black 


- .1238 a 


- .0318 


- .1991 


- .6784***' b 


Asian 


- .8654*** 


- .1198 b 


.5136*' b 


.21 1 5 b 


Indian 


.1063 


9059***' b 


.1791 


- .0524 


Hispanic 


- .5397*** 


- .0526 b 


.4279**' b 


- .1995 b 


Constant 


-3.7749*** 


-3.7960*** 


-5.2673*** 


-4.5590*** 


-2 Log likelihood 9634.895 c 


8422.515° 


3288.470 c 


4518.503° 


Model chi-square 1 1 08.81 2*** 


421.855*** 


182.875*** 


407.621*** 


Classification 










percentage 


82.15 


87.07 


96.52 


94.30 


***Significance 


greater than .01 








**Significance 


greater than .05 








*Significance 


greater than .10 








Significantly different from correspond 


ng 1 982 coeffi 


cient at 5 percent 


evel 


Significantly different from correspond 


ng "live" coeff 


icient at 5 percent 


level 


c Significantly different from the "perfect 


" model 







Television 

Of special interest in television participation for musicals are the smaller 
coefficients for income, compared with those for live participation, and the 
apparent aging of the audience, as indicated by the higher age coefficient for 
retired persons in 1992. 

Radio 

Radio participation offers no surprises beyond the result that male partici- 
pation is not significantly different from female participation. Income is gener- 
ally insignificant in both time periods, and black participation is not 
significantly different from white. 



62 I Turning On and Tuning In 

Recordings 

Participation via recordings shows little change over time. In 1992, several 
coefficients — especially those for income — are quite different from those for 
live participation. This is a bit surprising, as both forms of participation entail 
explicit user charges. 

Musical Summary 

Overall, patterns of participation in musicals are familiar ones. The urban 
and male coefficients have declined over time. The base levels of participation 
via radio and recordings, as indicated by their large negative constants, were 
lower than those for live and television participation. 



Theater (Plays) 

Results of the logistic regressions for participation in theater are displayed 
in Tables 45 and 46. 

Live Participation 

Results for both 1982 and 1992 show patterns similar to those noted for 
the other benchmark arts — higher rates of participation among urban dwellers 
and females, and rates rising with increased income, education, and age, except 
for a slight drop among retired persons. A notable exception to the familiar 
pattern, however, occurs with black participation. In 1992 the rate of partici- 
pation among blacks is not significantly different from that of whites. This 
perhaps reflects greater availability of drama that speaks to the black experience, 
including the works of such contemporary black playwrights as August Wilson, 
among others. 

Television 

The major divergence from the 1982 participation patterns in the case of 
theater on television is the 1 992 results for age, where the coefficients are much 
higher than those for live participation. This likely reflects the continuing 
development over the decade of alternative viewing habits of young adults. 

Radio 

Income is not a significant influence on participation in drama via radio in 
either 1982 or 1992, but the influence of increased age is higher and significant 



Multivariate Statistical Results I 63 



TABLE 45. Logistic Regression, Theater (Plays) Participation, 
1982 



Independent 
Variable 



Live 



Television 



Radio 



Urban 


.4408*** 


.5247*** 


.5298** 


Male 


— .4199*** 


- .2156*** 


- .1241 


Low income 


.1206 


.3667*** 


- .3207 


Moderate income 


.4980*** 


.5699*** 


- .2707 


High income 


1.1940*** 


1 1 999*** 


-1 .4546** 


High school 


1 291 7*** 


.9908*** 


.8621* 


College 


2.7063*** 


1 .9496***' a 


1.7326*** 


30-something 


2i on***' 3 


.0093 a 


- .3907*' a 


Middle age 


.3289*** 


.3643**' a 


- .0746 a 


Retired 


.2865*** 


.2648*' a 


- .5899*' a 


Black 


- .5964***' a 


- .2590* 


- .3027 a 


Indian 


- .8909*** 


- .4638 


- .4995 


Constant 


-4.5275*** 


—3 1 71 Q*** 


-4.241 9*** 


-2 Log likelihood 


1 0049.071 b 


3818.382 b 


1178.053 b 


Model chi-square 


1541.749*** 


380.843*** 


60.209*** 


Classification percentage 


87.89 


75.04 


95.94 



***Significance greater than .01 

**Significance greater than .05 
*Significance greater than .1 
a Significantly different from corresponding 1992 coefficient at 5 percent level 

Significantly different from the "perfect" model 



in 1992. The 1992 rate of participation by blacks is significantly higher than 
that of whites. 



Theater Summary 

The theater results are all consistent with the stated hypotheses: participa- 
tion generally is greater among females and urban dwellers, and in most cases 
it increases with income, education, and age. Age coefficients increased in value 
and became more generally significant between 1982 and 1992, suggesting a 
wider gap in behaviors between young adults and older groups. Also notable is 
the insignificant difference in rates of live participation by blacks and whites in 
1992, and blacks' higher rate of participation via radio in that year. 



64 I Turning On and Tuning In 



TABLE 46. Logistic Regression, Theater (Plays) Participation, 
1992 



Independent Variable 



Live 



Television 



Radio 



Urban 


.4254*** 


Male 


- .3347*** 


Low income 


2i QO*** 


Moderate income 


.4590*** 


High income 


1.0827*** 


High school 


1.2816*** 


College 


2.4997*** 


30-something 


- .0477 a 


Middle age 


.2803*** 


Retired 


.2643*** 


Black 


- .0135 a 


Asian 


— 9241 *** 


Indian 


.4047 


Hispanic 


- .2849** 


Constant 


-4.4736*** 


-2 Log likelihood 


8119.192° 


Model chi-square 


925.738*** 


Classification percentage 


86.66 



.3915*** 
.1551***' b 
.1058 
.3785*** 



.5572*** 

.7519*** 

1.4245*** 



,b 

,a,b 

4263***' 

9894***' 

1 21 75*** ,a ' 

- .0986 

- .3571*' b 
.2758 
.1031 b 

-3.9266*** 



.2336 c 
.1462 t 

- .2479 1 

- .1328 t 
- .1283 k 

.6565** 



,b 



1.5099***' b 
.481 8***' a ' b 
.4824***' a,b 
.5598***' a ' b 
4732***' 
.341 2 b 
.3126 
.4784**' b 

-5.2953*** 



9300.952 c 
570.747*** 
84.59 



2753.414 c 
85.180*** 
97.29 



***Significance greater than .01 

**Significance greater than .05 
*Significance greater than .10 
a Significantly different from corresponding 1982 coefficient at 5 percent level 

Significantly different from corresponding "live" coefficient at 5 percent level 
c Significantly different from the "perfect" model 



Dance 

Tables 47 and 48 show the results of the logistic regressions for ballet and 
dance. As stated earlier, the 1982 survey included questions pertaining only to 
ballet, while the 1992 survey added questions on dance more broadly defined. 
Furthermore, the 1992 media participation questions pertained only to dance. 
While this permits a comparison of live ballet attendance over time, neither 
dance over time nor media participation can be compared, because of differences 
in definitions. 



Live Participation 

Of special note is the generally insignificant impact of age on live attendance 
for both ballet in 1982 and dance in 1992, suggesting. that older groups do not 
differ markedly from the young adult group in their attendance rates. Age was 



Multivariate Statistical Results 65 



TABLE 47. Logistic Regression, Ballet 


Participation, 1982 


Independent Variable 


Live 


Television 


Urban 


.5359*** 


.3610*** 


Male 


-1.0136***' 3 


- .7816*** 


Low income 


.1403 


.2897** 


Moderate income 


.3637*** 


.5012*** 


High income 


8291 *** 


.5650** 


High school 


1 2474*** 


.4280** 


College 


2.6958*** 


1.4651*** 


30-something 


.2876***' a 


.3807*** 


Middle age 


.0814 


7981 *** 


Retired 


.0813 


7977*** 


Black 


- .7651***' a 


- .3077 


Indian 


- .6083** 


.3966 


Constant 


-5.4316*** 


-3.1942*** 


-2 Log Likelihood 


4917.686° 


3004. 546 c 


Model chi-square 


664.556*** 


276.695*** 


Classification percentage 


95.69 


83.46 


***Significance greater than 


.01 




**Significance greater than 


.05 




^Significance greater than 


.10 




a Significantly different from 


corresponding 1992 coefficient at 5 percent level 


c Significantly different from the "perfect" model 





significant for live ballet in 1992, but the coefficients are small, indicating that 
impacts on probabilities are slight. Age was a significant determinant of media 
(television) participation in ballet in 1982 and 1992, peaking with the middle 
age group in the earlier year and with the retired group in the latter year. 

By 1992, the negative coefficient of black participation in live ballet declined 
significantly, consistent with earlier observations regarding growing popularity 
of primarily black ballet companies. Also in 1992, participation by Indian 
respondents in live dance was significantly higher than that by whites, perhaps 
reflecting the existence of dance as an indigenous art form among Native 
Americans. This would not have been revealed by 1982 questions pertaining 
only to ballet. 

Television 



In contrast to live participation in dance forms, age is a significant determi- 
nant of television participation. Older respondents are considerably more likely 
to participate via television than are 30-somethings. Again, this may reflect the 



66 I Turning On and Tuning In 



TABLE 48. Logistic Regression, Dance Participation, 1992 


Independent Variable 


Live Ballet 


Live Dance 


Television 


Urban 


.4938*** 


.1786* 


.2439*** 


Male 


— ^7f)Q***' a 


— 21 42*** 


- .3708*** 


Low income 


- .1680*** 


.2754** 


.1494* 


Moderate income 


.3535*** 


.4004*** 


.1028 b 


High income 


.9302*** 


.3574** 


.3372*** ' 


High school 


1.1300*** 


.4595** 


.4880*** 


College 


2.3871*** 


1.3781*** 


1.1758*** 


30-something 


- .0052***' a 


.0065 


2952*** 


Middle age 


.0583*** 


.1004 


.6908***' b 


Retired 


.0697*** 


.1060 


.8359***' b 


Black 


— 171 Q*** ,a 


.2039 


.0724 


Asian 


-1.5828*** 


- .0346 


.1270 


Indian 


- .6469*** 


1.3626*** 


.1207 b 


Hispanic 


.0130*** 


.1919 


.4507*** 


Constant 


-3.952.4*** 


-3.9524*** 


-2.9960*** 


-2 Log Likelihood 


5.74E07 C 


5674.1 17 c 


10251.272° 


Model chi-square 


5.76E07*** 


237.472*** 


417.945*** 


Classification percentage 


95.38 


92.87 


82.40 


***Significance greater than 


.01 






**Significance greater than 


.05 






*Significance greater than 


.10 






a Significantly different from 


corresponding 1 982 coefficient at 5 


percent level 


Significantly different from corresponding "live" 


coefficient at 5 


percent level 


Significantly different from the "perfect" model 







differentiated viewing habits of young adults. In neither year was race a 
significant determinant of television participation in ballet and dance. 



Art 

Tables 49 and 50 contain the results of the logistic regressions for art 
participation in 1982 and 1992, respectively. 

Live participation 

From 1982 to 1992, coefficients of the education variables diminished in 
value. In both years, older age groups are less likely to participate than are young 
adults. 



Multivariate Statistical Results 67 



TABLE 49. Logistic Regression, Art 


Participation, 1982 


Independent Variable 


Live 




Television 


Urban 


.4533*** 




.3240*** 


Male 


- .3153***' a 




- .0852 


Low income 


.1174* 




.2123* 


Moderate income 


.3980*** 




.6451*** 


High income 


.9513*** 




9208***' a 


High school 


1 .4741 ***' a 




1.5315***' a 


College 


2.8006***' a 




2.2505*** 


30-something 


.0791 




.0576 


Middle age 


- .0566 




.3828*** 


Retired 


- .1813**' a 




.2899** 


Black 


- .6015***' a 




.0733 a 


Indian 


- .0837 




- .0351 


Constant 


-3.6478*** 




-3.6384 


-2 Log Likelihood 


14320.459 b 




3657. 835 b 


Model chi-square 


2277.911*** 




297.159*** 


Classification percentage 


78.01 




76.80 


***Significance greater than 


.01 






**Significance greater than 


.05 






*Significance greater than 


.10 






a Significantly different from 


corresponding 1992 coefficient at 


5 percent level 


Significantly different from the "perfect" model 







Television 

As for live participation, coefficients of the education variables declined 
between 1982 and 1992. Male participation is not significantly different from 
female participation via this medium, and age variables for television show 
somewhat more significance than those for live participation in both periods. 



Summary 

The results of the logistic regressions generally confirm the findings reported 
in Part II, even when controlling for other influences. Except for jazz, partici- 
pation rises with age, income, and education for all art forms and all media, 
thereby confirming the proffered hypotheses. Other noteworthy findings in- 
clude the following: 

■ Education coefficients generally are lower for media participation than for 
live participation. This suggests that media are something of an equalizer, 



68 I Turning On and Tuning In 



TABLE 50. Logistic Regression, Art Participation, 1992 



Independent Variable 



Live 



Television 



Urban 


.4495*** 


Male 


- .1590***' a 


Low income 


.4274*** 


Moderate income 


.6864*** 


High income 


1.0195*** 


High school 


1 0449***' a 


College 


2.3690***' 3 


30-something 


- .1280** 


Middle age 


- .1172* 


Retired 


- .3566***' a 


Black 


- .3945***' a 


Asian 


- .2123 


Indian 


.2506 


Hispanic 


- .3741*** 


Constant 


-3.3717*** 


-2 Log likelihood 


11514.743° 


Model chi-square 


1758.139*** 


Classification percentage 


74.06 



.1210**'" 

- .0664 
.1442**'* 3 
.1886***' b 

,a 



.3863= 
.7598***' 
1 .4654***' a ' b 

,b 



.1632 

0171 *** 

.1 246***' b 

- .4058***' a 

- .6519***' b 
.1965 
.1868** 

-2.2076*** 



13461 .81 6 e 
646.599*** 
69.54 



K **Significance greater than .01 
**Significance greater than .05 
*Significance greater than .10 
Significantly different from corresponding 1 982 coefficient at 5 percent level 
^Significantly different from corresponding "live" coefficient at 5 percent level 



Significantly different from the "perfect" model 



diminishing the distinctions in participation patterns by education 
groupings. 

Similarly, income coefficients are lower for the media equations than for 
live participation. This indicates that the media also diminish distinctions 
in participation by income group. 

The urban coefficient is lower for media participation than for live partici- 
pation. Although the differences are not significant, the nearly uniform 
pattern indicates that location is not so important for media participation. 
The age coefficients generally are larger for media participation than for live 
participation, indicating that older groups have relatively greater access to 
the arts via media as compared with live attendance. 

While the income and education coefficients generally declined between 
1982 and 1992, indicating a relative decline in participation distinctions, 
age coefficients rose, further indicating an aging audience. 



Multivariate Statistical Results 69 



Although the explanatory variables, both singly and in concert, are signifi- 
cant and generally have the expected signs, the overall equations lack persuasive 
explanatory power. Apparently numerous additional factors — or simply a great 
deal of randomness — influence participation in the arts, both live and via media. 
In summary, the media seem to be ensuring greater relative access to the arts for 
various demographic groups. 



The Cross Effects of Media 
and Live Participation 




What is the nature of the cross effects of live and media participation? Are 
the media complements to live participation in the arts, or are they 
substitutes? This part of the monograph offers a partial answer, and the future 
research agenda points to a fuller answer yet to come. 

The tables in the following sections contain the percentages of respondents 
by column and by row. The top number in each cell is the percentage of those 
who gave the row response and also gave the column response. The bottom 
number in each cell is the percentage of those who gave the column response 
who also gave the row response. For example, in Table 52, the first entry under 
television, No, is 85.6. This is the percentage of respondents who did not 
participate in live jazz who also did not participate in televised jazz in 1 982. The 
second number, 94.0, is the percentage of those who did not participate in 
televised jazz who also did not participate in live jazz. 

Table 5 1 shows hypothetical tabular entries for each of two cases. In the 
first instance, "Medium X," 100 percent of those who do not participate via the 
medium also do not attend live performances (the "northwest" quadrant), while 
100 percent of those who do participate via the medium also attend live 
performances (the "southeast quadrant"). The same is true of those who attend 
live performances; they all participate by media. In this case, the two types of 
participation are perfect complements. 



TABLE 51. 


Hypothetical 


Table Entries 








Perfect Complements 


Perfect Substitutes 


Live 


Med 


ium X 


Medium 


Y 


No 


Yes 


No 


Yes 


No 


100.0 


0.0 


0.0 


100.0 




100.0 


0.0 


0.0 


100.0 


Yes 


0.0 


100.0 


1 00.0 


0.0 




0.0 


100.0 


100.0 


0.0 



70 



The Cross Effects of Media and Live Participation I 71 



For "Medium Y," the opposite is true. None of those who participate via 
the medium attend live performances, while none of those who attend live 
performances participate via the medium. Here, the medium and live perform- 
ances are perfect substitutes. 

The reality is not this simple. 



Jazz 

In general, Tables 52 and 53 indicate that those who did not attend a jazz 
performance were unlikely to participate in a televised, radio, or recorded 
performance in either 1982 or 1992. Nor were those who did not participate 
by media likely to attend live performances. 

In 1982, 51.1 percent of respondents who attended a live jazz performance 
also viewed a performance on television. This percentage had fallen to 43.8 in 
1992. By the same token, 28.3 percent who saw jazz on television in 1982 also 
attended a live performance, but this had fallen to 25.9 percent in 1 992. Overall, 
then, we could state that those who attend live jazz are more likely to watch jazz 
on television than the converse. In contrast, the percentage of those viewing jazz 
on television, but not attending a live performance, rose from 71.7 percent in 
1982 to 74.1 percent in 1992, indicating that televised jazz may have become 
a more significant substitute for live attendance over the decade. 

Participation by radio showed much the same pattern. Over 58 percent of 
those who attended a live performance also listened to jazz on the radio in 1982, 
and this percentage rose to more than 75 percent in 1992. However, the 
percentage of those who listened to jazz on the radio but did not attend a live 
performance rose from 67.7 percent to 71.5 percent. 

Finally, the percentage who attended live jazz and also listened to a recorded 
performance rose from 65.5 percent to 67.9 percent over the decade, while the 



TABLE 52. 


Jazz Cross-Participation, 


1982 






Live 


Television 


Radio 


Recorc 

No 


ling 
Yes 


No 


Yes 


No 


Yes 


No 


85.6 


14.4 


86.4 


13.6 


84.9 


15.1 




94.0 


71.7 


94.9 


67.7 


95.6 


67.2 


Yes 


48.9 


51.1 


41.8 


58.2 


34.5 


65.5 




6.0 


28.3 


5.1 


32.3 


4.4 


32.8 



72 I Turning On and Tuning In 



TABLE 53. 


Jazz( 


Cross-Participation, 


1992 




Live 


Television 


Radio 


Recording 

No Yes 


No 


Yes 


No 


Yes 


No 


85.2 


14.8 


77.5 


22.5 


85.1 14.9 




92.7 


74.1 


96.4 


71.5 


95.7 64.6 


Yes 


56.2 


43.8 


24.3 


75.7 


32.1 67.9 




7.3 


25.9 


3.6 


28.5 


4.3 35.4 



percentage who listened to a jazz recording and also attended a live performance 
also rose, from 32.8 to 35.4. 

Overall, the pattern indicates the media provide access for those who choose 
not to, or are unable to, attend a live performance. Approximately two-thirds 
to three-fourths of those who indicate media participation in jazz do not attend 
live performances. 



Classical Music 

From 1982 to 1992, the percentage of respondents who attended a classical 
music concert and also viewed classical music on television dropped precipi- 
tously, from 62.3 to 51.7 percent, as shown in Tables 54 and 55. As indicated 
earlier, this trend may well reflect reduced television broadcasting of classical 
music. The percentage who watched classical music on television and also 
attended a live performance remained essentially unchanged at just over 29 
percent. 



TABLE 54. 


Classical Music Cross-Participation, 


1982 




Live 


Television 


Rad 


io 


Recording 

No Yes 


No 


Yes 


No 


Yes 


No 


80.3 


19.7 


84.8 


15.2 


83.7 


16.3 




94.1 


70.5 


93.5 


67.3 


94.8 


65.1 


Yes 


37.7 


62.3 


44.4 


55.6 


34.4 


65.6 




5.9 


29.5 


6.5 


32.7 


5.2 


34.9 



The Cross Effects of Media and Live Participation I 73 



TABLE 55. 


Classical Music Ci 


ross-Participation, 


1992 


Live 


Television 


Rad 


io 


Recording 
No Yes 


No 


Yes 


No 


Yes 


No 


82.4 


17.6 


75.3 


24.7 


81.9 18.1 




92.3 


70.6 


95.2 


70.3 


93.8 66.2 


Yes 


48.3 


51.7 


26.6 


73.4 


36.8 63.2 




7.7 


29.4 


4.8 


29.7 


6.2 33.8 



Radio and live performance showed a very different pattern. The percentage 
who attended and also listened on radio rose from 55.6 percent to 73.4 percent 
over the decade, while the percentage of those who listened and also attended 
fell slightly, from 32.7 to 29.7 percent. It seems very likely that this shift reflects 
the nearly ubiquitous availability of classical music on radio by 1992. 

Finally, the percentage of those who attended live performances and also 
listened to recorded performances fell from 65.6 percent to 63.2 percent. This 
occurred despite the greater sound fidelity afforded by compact discs, widely 
available in 1992. One can only hazard the guess that relatively higher prices of 
CDs discouraged potential purchasers/listeners. 



Opera 

The pattern of cross-participation in opera is quite similar to that for classical 
music. As shown in Tables 56 and 57, in 1982, well over half of those who 
attended live performances also watched opera on television; but by 1 992 this 



TABLE 56. 


Opera 


Cross- 


Participation, 


1982 




Live 


Television 


Radi 


o 


Recording 

No Yes 


No 


Yes 


No 


Yes 


No 


89.0 


11.0 


93.8 


6.2 


93.4 6.6 




98.8 


89.6 


98.7 


85.0 


98.6 86.2 


Yes 


45.2 


54.8 


53.8 


46.2 


54.9 45.1 




1.2 


10.4 


1.3 


15.0 


1.4 13.8 



74 I Turning On and Tuning In 



TABLE 57. 


Opera 


Cross- 


Participation, 


1992 




Live 


Television 


Radi 





Recording 
No Yes 


No 


Yes 


No 


Yes 


No 


90.2 


9.8 


92.7 


7.5 


94.2 5.8 




97.7 


88.6 


97.9 


84.0 


97.9 80.2 


Yes 


62.3 


37.7 


57.7 


42.3 


58.2 41.8 




2.3 


11.4 


2.1 


16.0 


2.1 19.8 



percentage had fallen to 37.7 percent, probably reflecting reduced availability 
of televised opera. The percentage of those who watched on television and also 
attended live opera performances remained low, around 1 1 percent. It may well 
be that a taste for opera, as indicated by television participation, is not sufficient 
to overcome relatively high ticket prices. 

The percentage of attenders of live performances who listened to opera on 
radio fell a bit, from 46.2 to 42.3 percent. Perhaps the fact that less than half 
of the performance attenders listen on the radio reflects the visual nature of 
opera, a dimension that is lost on radio. 

Finally, the percentage of attenders of live performances who listened to a 
recorded performance fell from 45.1 percent to 41.8 percent, consistent with 
the expected impact of higher-priced CDs. (There may also be an aversion 
among opera aficionados to CD technology, although this remains conjectural.) 
Interestingly, the percentage of those who listened to a recording and also 
attended a live performance, while remaining low overall, rose from 13.8 to 
19.8 percent. Since live performances did not get less expensive over the decade, 
this may reflect the larger number of opera companies in 1 992, as well as greater 
audiences for existing companies. 



Musicals 

The percentage of respondents who attended live musical performances and 
also watched on television fell substantially — from 43.8 to 24.7 percent — over 
the decade, as shown in Tables 58 and 59. This very likely reflects a decline in 
the availability of televised musicals during that period. The reason for the 
decline — not so stark — in the percentage of television viewers of musicals who 
also attended live performances is not so clear. 

Very few — between 8 and 9 percent in both survey periods — of those who 



The Cross Effects of Media and Live Participation I 75 



TABLE 58. 


Musical Cross- 


■Participation, 1 982 




Live 


Television 


Radi 


o 


Recording 

No Yes 


No 


Yes 


No 


Yes 


No 


85.1 


14.9 


96.7 


3.3 


95.0 5.0 




86.7 


59.5 


82.0 


62.3 


84.1 48.2 


Yes 


56.2 


43.8 


91.4 


8.6 


77.1 22.9 




13.3 


40.5 


18.0 


37.7 


15.9 51.8 



TABLE 59. 


Musical Cross- 


■Participation, 1992 




Live 


Television 


Rad 


o 


Recording 

No Yes 


No 


Yes 


No 


Yes 


No 


90.0 


10.0 


97.6 


2.4 


96.9 3.1 




85.1 


65.8 


83.5 


58.0 


84.7 44.6 


Yes 


75.3 


24.7 


91.6 


8.4 


82.1 17.9 




14.9 


34.2 


16.5 


42.0 


15.3 55.4 



attended a live musical performance also listened on radio, probably because the 
medium does not lend itself well to this art form. Memorable and listenable 
songs are interspersed with spoken dialogue. Furthermore, the visual aspects of 
musicals are lost to the radio listener. Of those who do choose to listen to 
musicals on the radio, however, the percentage who attend live performances 
rose over the decade, from 37.7 to 42 percent. While this may reflect growth 
of new performing venues and the resurgence of Broadway touring companies, 
these phenomena did not show up in the television figures. 

The decline in the percentage of attenders at live performances who also 
listened to recorded performances (from 22.9 percent in 1982 to 17.9 percent 
in 1992) may reflect both the increased prices of CDs and the failure of 
recordings to convey adequately the visual aspects of musicals. 



Theater (Plays) 

Among the media, only television lends itself well to theater participation. 
Even so, among those who attended live performances, the percentage who also 



76 I Turning On and Tuning In 



TABLE 60. 


Theater (Plays) 


Cross-Participation, 


1982 




Live 


Televis 


ion 




Radio 




No 


Yes 


No 




Yes 


No 


78.1 


21.9 


97.0 




3.0 




93.6 


75.3 


89.6 




68.9 


Yes 


42.6 


57.4 


89.4 




10.6 




6.4 


24.7 


10.4 




31.1 



TABLE 61. 


Theater (Plays) C 


ross-Participation, 


1992 




Live 


Television 




Radio 




No 


Yes 


No 




Yes 


No 


88.0 


12.0 


97.9 




2.1 




89.4 


70.0 


87.1 




65.7 


Yes 


66.9 


33.1 


92.8 




7.2 




10.6 


30.0 


12.9 




34.3 



viewed plays on television fell from 57.4 percent in 1982 to 33.1 percent in 
1992, as shown in Tables 60 and 61. Once again we may speculate that this is 
due to reduced availability on broadcast television. On the other hand, those 
who do watch drama on television were more likely to attend live performances 
in 1992 than in 1982. 

Attenders of live theater performances are not very likely to listen to plays 
on the radio. In 1982, only 10.6 percent did so, and this fell to 7.2 percent in 
1992. But the percentage of those who do listen to radio broadcasts and also 
attend live performances rose from 31.1 to 34.3 percent. 



Ballet/Dance 

Changes over the decade in ballet and dance participation must be inter- 
preted in light of the fact that the dance forms reported here are not strictly 
comparable. Given that caveat, it is notable that while 57 percent of those who 
attended a live ballet performance in 1982 also viewed a televised performance, 



The Cross Effects of Media and Live Participation I 77 



TABLE 62. Ballet Cross- 




TABLE 63. Dance Cross- 


Participation, 1982 




Participation, 1992 


Television 




Television 


Live No Yes 


Live No Yes 


No 85.4 14.6 


No 84.6 15.4 


97.9 85.6 




94.9 83.2 


Yes 43.0 57.0 




Yes 59.5 40.5 


2.1 14.4 




5.1 16.8 



only 40.5 percent did so with respect to dance, more broadly defined, in 1992 
(see Tables 62 and 63). In both time periods, the vast majority of those who 
did watch ballet or dance on television did not attend a live performance. 



Art 

As indicated in Tables 64 and 65, the proportion of those who attended an 
art museum and also saw a televised program on art rose from just under half 
in 1982 to just over half 10 years later. Likewise, the proportion of those who 
watched a program and also visited a museum rose. Among all the art forms, 
this is the sole instance of increased cross-participation in both directions. 



TABLE 64 


. Art Cross 




Participat 


ion, 1982 




Live 


Television 


No 


Yes 


No 


84.1 


15.9 




85.0 


54.4 


Yes 


52.8 


47.2 




15.0 


45.6 



TABLE 65 


. Art Cross 


_ 


Participat 


ion, 1992 




Live 


Television 


No 


Yes 


No 


78.0 


21.5 




82.3 


52.7 


Yes 


46.3 


53.2 




17.7 


47.3 



78 I Turning On and Tuning In 

Summary and Conclusions 

The statistics related to cross-participation in the arts via media alternatives 
and live attendance indicate that those who shunned live attendance also 
shunned media participation and vice versa; but very often, a majority of those 
who participated by media did /^attend live performances or showings. A more 
positive slant is that respondents who did not attend live performances or 
showings did participate by media. This further suggests that the media consti- 
tute a more-or-less readily available alternative to live attendance. While in many 
cases this percentage declined over the 1 0-year period between surveys, we can 
only speculate that media participation may have encouraged greater attendance 
at the arts. 



Summary and 
Conclusions 




This monograph examines patterns of public participation in seven core, or 
"benchmark," arts via selected electronic media: television, radio, and 
recordings. More specifically, it examines the demographic determinants of 
participation via media and how the demographics may differ between live and 
media participation. Testable hypotheses derived from the theory of consumer 
demand were tested with data from the 1982 and 1992 Surveys of Public 
Participation in the Arts. 

This part of the monograph reviews the hypotheses and findings, ventures 
some policy implications, and issues the standard call for further research. After 
all, any exploration of this nature should uncover new questions even as it strives 
to answer existing ones. 



Major Findings 

The results of the statistical analysis are consistent with the hypotheses and 
with previously existing evidence that arts participation increases with age. This 
is true of media participation as well as live participation. While this means that 
an aging population will, other things being equal, enhance arts participation, 
it also bodes less well for the future, when the smaller, post— baby boom 
generation is in the ascendancy. Even if participation rates remain high, the 
numbers may decline and imperil arts providers. 

The results reported here also confirm numerous prior analyses of the impact 
of education. A more educated populace is more "arts friendly," and rates for 
both live and media participation rise with education. 

Arts participation also rises as income increases. But of special note here is 
that participation via media — especially the broadcast media — is not so income 
dependent. This reflects the likelihood that additional out-of-pocket costs of 
listening to the radio or viewing television are negligible, so that broadcast arts 
are accessible to those without high incomes. 

While urban residents are more likely than rural residents to participate in 
the arts via media as well as to attend live performances, the difference is not so 
strong for media participation. This suggests that the media provide relatively 
greater access to the arts for nonurban residents. 



79 



80 I Turning On and Tuning In 



Other demographic variables, including race and gender of respondent, 
were employed here primarily for control purposes. Findings confirm that 
whites are more likely to participate than nonwhites, and women more than 
men, except in the case of jazz. 



Policy Implications 

With the increasing availability of cable and direct satellite access television, 
literally hundreds of channels will be available to viewers. Some of these could 
surely be dedicated to the arts, and even to specific art forms: the opera channel, 
the symphony channel, the Schubert channel, the post-modern dance channel, 
and so forth. This will promote both variety and access. 

Since education exposure seems to be consistent with development of taste 
for the arts, educators and educational institutions at all levels may take 
advantage of both live and media access to the arts and develop arts curricula 
that incorporate instructional technology. 

Special arts events — "The Three Tenors," for example — represent oppor- 
tunities for "blockbuster" broadcasts and subsequent audio and video record- 
ings. Improving technology and increasing competition continue to drive 
down the prices of such recordings, making them available to potential buyers 
with limited incomes. 

The rapid development of distance learning techniques and instructional 
technologies offer opportunities for arts programming to reach audiences via 
the Internet, CD-ROM, interactive television, and means as yet unforeseen. 
One can imagine, for example, a CD-ROM that plays a symphony, portrays 
the orchestra, and tracks the score simultaneously, allowing the viewer to 
suspend the performance at any time for an interactive help session or to access 
expert commentary. Public seed monies may be useful to support such 
innovations. 



Future Research 

Many additional questions remain. What, for example, might be the cross 
effects revealed by a simultaneous logit model, whereby live and media partici- 
pation are hypothesized to affect each other? How sensitive are these results to 
the choice of educational, income, and age groupings? 

What information might be gained from more frequent surveys as well as 
more specific geographic breakdowns of data? It is easy to imagine that annual 
or biannual data for, say, a metropolitan area would provide feedback on the 



Summary and Conclusions I 81 

effect of pricing policies on participation, the role of changes in amount or 
structure of subsidies, the impact of new technologies, and similar time- 
dependent changes. 

Even as the data gleaned from these surveys continue to undergo scrutiny, 
planning and implementation of successive efforts should be underway. The 
arts and arts policy are too important to be left to interpolation and other forms 
of informed guesswork. 



Appendix A: 

1992 Survey of Public Participation 
in the Arts 



INTRODUCTION - Now I have some questions about your leisure activities. The Bureau of the 
Census is collecting this information for the National Endowment for the Arts. The survey is 
authorized by Title 20, United States Code, section 954 and Title 13, United States Code, section 
8. Your participation in this interview is voluntary and there are no penalties for not answering 
some or all of the questions. (If PERSONAL INTERVIEW, hand respondent the Privacy Act Statement, 
SPPA-13.) i 



1. The following questions are about YOUR 
activities during the LAST 12 months- 

between 1,19 , and 

19 



3. 



With the exception of elementary or high 
school performances, did YOU go to a live 
jazz performance during the LAST 12 
MONTHS? 

oDNo 

Yes - About how many times did you do 
this during the LAST 12 MONTHS? 



Number of times 



(With the exception of elementary or high 
school performances,) Did you go to a live 
classical music performance such as 
symphony, chamber, or choral music 
during the LAST 12 MONTHS? 

oDNo 

Yes - About how many times did you do 
this during the LAST 12 MONTHS? 



Number of times 



(With the exception of elementary or high 
school performances,) Did you go to a live 
opera during the LAST 12 MONTHS? 

oDNo 

Yes - About how many times did you do 
this during the LAST 12 MONTHS? 



Number of times 



(With the exception of elementary or high 
school performances,) Did you go to a live 
musical stage play or an operetta during 
the LAST 12 MONTHS? 

oDNo 

Yes - About how many times did you do 
this during the LAST 12 MONTHS? 



Number of times 



5. 



(With the exception of elementary or high 
school performances,) Did you go to a live 
performance of a non-musical stage play 
during the LAST 12 MONTHS? 

oDNo 

Yes - About how many times did you do 
this during the LAST 12 MONTHS? 



Number of times 



(With the exception of elementary or high 
school performances,) Did you go to a live 
ballet performance during the LAST 12 
MONTHS? 

oDNo 

Yes - About how many times did you do 
this during the LAST 12 MONTHS? 



Number of times 



(With the exception of elementary or high 
school performances,) Did you go to a live 
dance performance other than ballet, such 
as modern, folk, or tap during the LAST 12 
MONTHS? 

oDNo 

Yes - About how many times did you do 
this during the LAST 12 MONTHS? 



Number of times 



8. (During the LAST 12 MONTHS,) Did you 
visit an ART museum or gallery? 

i»n odno 

Yes - About how many times did you do 
this during the LAST 12 MONTHS? 



Number of times 



(During the LAST 12 MONTHS,) Did you 
visit an ART fair or festival, or a CRAFT fair 
or festival? 

oDNo 

Yes - About how many times did you do 
this during the LAST 12 MONTHS? 



Number of times 



82 



1 992 Survey of Public Participation in the Arts I 83 



10. (During the LAST 12 MONTHS,) Did you 
visit an historic park or monument, or 
tour buildings, or neighborhoods for their 
historic or design value? 



oDNo 

Yes - About how many times did you do 
this during the LAST 12 MONTHS? 



Number of times 



11. With the exception of books required for 
work or school, did you read any books 
during the LAST 12 MONTHS? 



oDNo 

Yes - About how many books did you 

read during the LAST 12 MONTHS? 



Number of books 



12. (During the LAST 12 MONTHS,) Did you 
read any - 

Read answer categories 



a. Plays? 



I 021 I 1DN0 2DYes 



b. Poetry? 



I 022 I iDNo aDYes 



Novels or short stories? I 023 I iDNo 2D Yes 



13. (During the LAST 12 MONTHS,) Did you 
listen to - 

a. A reading of poetry. 



either live or recorded? I 024 I iDNo 2D Yes 



b. A reading of novels or 
books either live or 
recorded? 



I 025 I iDNo 2 DYes 



14a. (During the LAST 12 MONTHS,) Did you 
watch a jazz performance on television or 
a video (VCR) tape? 



026 I 1 □ No - Skip to item 14c 

Yes - Was that on TV, VCR, or both? 

2DTV 
3D VCR 
4 □ Both 



b. About how many times did you do this in 
the LAST 12 MONTHS? 



Number of times 



c. (During the LAST 12 MONTHS,) Did you 
listen to jazz on radio? 



fiU iDNo 
2[DYes 



d. (During the LAST 12 MONTHS,) Did you 
listen to jazz records, tapes, or compact 
discs? 



iDNo 
2D Yes 



Page 2 



15a. (During the LAST 12 MONTHS,) Did you 
watch a classical music performance on 
television or a video (VCR) tape? 



030 I 1 □ No - Skip to item 15c 

Yes - Was that on TV, VCR, or both? 

2DTV 
aDVCR 
4 □ Both 



b. About how many times did you do this (in 
the LAST 12 MONTHS)? 



Number of times 



c. (During the LAST 12 MONTHS,) Did you 
listen to classical music on radio? 



iDNo 
2D Yes 



d. (During the LAST 12 MONTHS,) Did you 
listen to classical music records, tapes or 
compact discs? 



iDNo 
2DYes 



16a. (During the LAST 12 MONTHS,) Did you 
watch an opera on television or a video 
(VCR) tape? 



i DNo - Skip to item 16c 

Yes - Was that on TV, VCR, or both? 

2DTV 
3 D VCR 
4 □ Both 



b. About how many times did you do this (in 
the LAST 12 MONTHS)? 



Number of times 



c. (During the LAST 12 MONTHS,) Did you 
listen to opera music on radio? 



iDNo 
2DYes 



d. (During the LAST 12 MONTHS,) Did you 
listen to opera music records, tapes, or 
compact discs? 



iDNo 
zDYes 



17a. With the exception of movies, did you 

watch a musical stage play or an operetta 
on television or a video (VCR) tape during 
the LAST 12 MONTHS? 



iED 1 DNo - Skip to item 17c 

Yes - Was that on TV, VCR, or both? 

2DTV 
sDVCR 
4 □ Both 



b. About how many times did you do this (in 
the LAST 12 MONTHS)? 



Number of times 



(During the LAST 12 MONTHS,) Did you 
listen to a musical stage play or an operetta 
on radio? 

iDNo 
2DYes 



(During the LAST 12 MONTHS,) Did you 
listen to a musical stage play or an operetta 
on records, tapes, or compact discs? 

iDNo 
2D Yes 

FORM SPPA 2 (4 9 92) 



84 I Turning On and Tuning In 



18a. With the exception of movies, situation 
comedies, or TV series, did you watch a 
non-musical stage play on television or a video 
(VCR) tape during the LAST 12 MONTHS? 



~2*E} 1 DNo - Skip to item 18c 

Yes - Was that on TV, VCR, or both? 

2DTV 
sDVCR 
4 □ Both 



b. About how many times did you do this (in the 
LAST 12 MONTHS)? 



Number of times 



c. (During the LAST 12 MONTHS,) Did you listen 
to a radio performance of a non-musical stage 
play? 



iDNo 

2D Yes 



19a. With the exception of music videos, did you 
watch on television or a video (VCR) tape 
dance such as ballet, modern, folk, or tap 
during the LAST 12 MONTHS? 



i □ No - Skip to item 20a 

Yes - Was that on TV, VCR, or both? 

2DTV 
3D VCR 
4 □ Both 



b. About how many times did you do this (in 
the LAST 12 MONTHS)? 



Number of times 



20a. (During the LAST 12 MONTHS,) Did you watch 
a program about artists, art works, or art 
museums on television or a video (VCR) tape? 



fED 1 □ No - Skip to item 21a 

Yes - Was that on TV, VCR, or both? 

2 DTV 
sDVCR 
4 □ Both 



22a. The following questions are about your 
participation in other leisure activities. 

Approximately how many hours of television 
do you watch on an average day? 



Number of hours 



b. During the LAST 12 MONTHS, did YOU go 
out to the movies? 



iDNo 
2 D Yes 



c. With the exception of youth sports, did you 
go to any amateur or professional sports 
events during the LAST 12 MONTHS? 



°E3 1DN0 

2D Yes 



d. During the LAST 12 MONTHS, did you go to 
an amusement or theme park, a carnival, or 
a similar place of entertainment? 



1DN0 
2D Yes 



e. During the LAST 12 MONTHS, did you jog, 
lift weights, walk, or participate in any other 
exercise program? 



iDNo 
2D Yes 



During the LAST 12 MONTHS, did you 
participate in any sports activity, such as 
Softball, basketball, golf, bowling, skiing, or 
tennis? 



iDNo 
2D Yes 



g. Did you participate in any outdoor activities, 
such as camping, hiking, or canoeing during 
the LAST 12 MONTHS? 



b. About how many times did you do this (in 
the LAST 12 MONTHS)? 



Number of times 



21a. 



I'm going to read a list of events that some 
people like to attend. If you could go to any of 
these events as often as you wanted, which 
ones would you go to MORE OFTEN than you 
do now? I'll read the list. Go to - 

Mark (X) all that apply. 

iDJazz music performances 

2D Classical music performances 

3D Operas 

4 D Musical plays or operettas 

5 D Non-musical plays 

6 D Ballet performances 

7 D Dance performances other than ballet 

sD Art museums or galleries 

9 D None of these - Skip to item 22a 



If only one is chosen, skip to item 22a. 
If more than one is chosen, ask - 

b. Which of these would you like to do most? 



Category number 
00DN0 one thing most 



J2SLJ 1DN0 
2D Yes 



h. Did you do volunteer or charity work during 
the LAST 12 MONTHS? 



1DN0 
2DYes 



i. Did you make repairs or improvements on 
your own home during the LAST 12 
MONTHS? 



1DN0 
2D Yes 



Did you work with indoor plants or do any 
gardening for pleasure during the LAST 12 
MONTHS? 



1DN0 
2D Yes 



23a. (During the LAST 12 MONTHS,) Did you work 
with pottery, ceramics, jewelry, or do any 
leatherwork or metalwork? 



1 D No - Skip to item 24a 

2DYes 



b. Did you publicly display any of your works? 

HD 1DN0 

2D Yes 



FORM SPPA-2 (4-9-92) 



Page 3 



1 992 Survey of Public Participation in the Arts I 85 



24a 



(During the LAST 12 MONTHS,) Did you do 
any weaving, crocheting, quilting, 
needlepoint, or sewing? 

id No - Skip to item 25a 
sD Yes 



b. Did you publicly display any of your works? 



iDNo 
2D Yes 



25a 



(During the LAST 12 MONTHS,) Did you 
make photographs, movies, or video tapes 
as an artistic activity? 

1 □ No - Skip to item 26a 

2D Yes 



Did you publicly display any of your works? 

1DN0 
2D Yes 



26a 



(During the LAST 12 MONTHS,) Did you do 
any painting, drawing, sculpture, or 
printmaking activities? 

1 □ No - Skip to item 27a 

2D Yes 



b. Did you publicly display any of your works? 



1DN0 
2D Yes 



27a. With the exception of work or school, did you 
do any creative writing such as stories, poems, 
or plays during the LAST 12 MONTHS? 



jED i □ No - Skip to item 28a 
2D Yes 



b.Were any of your writings published? 



1DN0 
2D Yes 



28a. Did you write or compose any music during 
the LAST 12 MONTHS? 



075 I 1 D No - Skip to item 29a 
2D Yes 



b. Was your musical composition played in a 
public performance or rehearsed for a public 
performance? 



ID 1ON0 
2D Yes 



29a. Do you own any original pieces of art, such 
as paintings, drawings, sculpture, prints, or 
lithographs? 



077 I 1 D No - Skip to item 30a 
2D Yes 



30b. Did you play any jazz in a public performance 
or rehearse for a public performance? 



_080j ^Mq 

2D Yes 



31a. During the LAST 12 MONTHS, did you play 
any classical music? 



o^D iDNo- Skip to item 32a 

2D Yes 



b. Did you play classical music in a public 
performance or rehearse for a public 
performance? 



1DN0 
2D Yes 



32a. During the LAST 12 MONTHS, did you sing any 
music from an opera? 



J^D 1 D No - Skip to item 33a 
2D Yes 



b. Did you sing in a public opera performance 
or rehearse for a public performance? 



1DN0 
2D Yes 



33a. During the LAST 12 MONTHS, did you sing 
music from a musical play or operetta? 



1 D No - Skip to item 33c 
2D Yes 



b. Did you sing in a public performance of a 
musical play or operetta or rehearse for a 
public performance? 



1DN0 
2D Yes 



c. During the LAST 12 MONTHS, did you sing in 
a public performance with a chorale, choir, 
or glee club or other type of vocal group, or 
rehearse for a public performance? 



1DN0 
2D Yes 



b. Did you purchase or acquire any of these 
pieces during the LAST 12 MONTHS? 



1DN0 
2D Yes 



30a. During the LAST 12 MONTHS, did you 
perform or rehearse any jazz music? 



J! 7 !] 1 D No - Skip to item 31a 
2D Yes 

Page 4 



34. (During the LAST 12 MONTHS,) Did you act in a 
public performance of a non-musical play or 
rehearse for a public performance? 



1DN0 
2D Yes 



35a. (During the LAST 12 MONTHS,) Did you dance 
any ballet? 



j 1 D No - Skip to item 36a 
2D Yes 



b. Did you dance ballet in a public performance 
or rehearse for a public performance? 



1DN0 
2D Yes 



36a. (During the LAST 12 MONTHS,) Did you do any 
dancing other than ballet such as modern, folk, 
or tap? 



_££LJ 1 DNo - Skip to item 37a 
2D Yes 



b. Did you dance modern, folk, or tap in a 
public performance? 



i!2LJ 1DN0 
2DYes 



FORM SPPA-2 14-9-921 



86 I Turning On and Tuning In 



37a. I'm going to read a list of some types of 
music. As I read the list, tell me which of 
these types of music you like to listen to? 

Mark (X) all that apply. 



□ Classical/Chamber music 
2D Opera 
3D Operetta/Broadway musicals/Show tunes 

4 D Jazz 

5 D Reggae (Reg gay) 

6 [2 Rap music 
I 7 D Soul 

8 D Blues/Rhythm and blues 

9 D Latin/Spanish/Salsa 
I io D Big band 

n D Parade/Marching band 
12 D Country-western 
oDBIuegrass 
uDRock 

sLjThe music of a particular Ethnic/ 
National tradition 
| e[ZiContemporary folk music 
i-D Mood/Easy listening 
isDNew age music 
igDChora I/Glee club 
20 D Hymns/Gospel 
21 D All 
22 D None/Don't like to listen to music - Skip to item 38a 



39a. (Have you EVER taken lessons or 

classes) in visual arts such as sculpture, 
painting, print making, photography, or 
film making? 



104 I 1 D No - Skip to item 40a 
2D Yes 



b. Did you take these lessons when you were 

Read categories. (Do not read category 4 if 
respondent is under 25 years old.) 
Mark (X) all that apply. 



iDLess than 12 years old 
2D 12-1 7 years old 
3D 18-24 years old 
4D25 or older 



CHECK 
ITEMC 



b. If only one category is marked in 37a, enter code in 
37b without asking. Which of these do you like 
best? 



Refer to item 39b 

Is box 1 or 2 marked in item 39b? 

D No - Skip to Check Item D 



DYes - Ask item 39c 



39c. Were these lessons or classes offered by the 
elementary or high school you were 
attending or did you take these lessons 
elsewhere? 



1D Elementary/high school 
2D Elsewhere 
3D Both 



Category number 



00D No one type best 



CHECK 
ITEM D 



38a. Have you EVER taken lessons or classes in 
music - either voice training or playing an 
instrument? 



1 D No - Skip to item 39a 

2 DYes 



b. Did you take these lessons when you were 

Read categories. (Do not read category 4 if 
respondent is under 25 years old.) 
Mark (X) all that apply. 



12L1 1D Less than 12 years old 
2 D 1 2-1 7 years old 
3D 18-24 years old 
4 D 25 or older 



CHECK 
ITEM A 



Refer to item 38b 

Is box 1 or 2 marked in item 38b? 

D No - Skip to Check Item B 
D Yes - Ask item 38c 



38c. Were these lessons or classes offered by the 
elementary or high school you were 
attending or did you take these lessons 
elsewhere? 



102 I 1D Elementary/high school 
2D Elsewhere 
3D Both 



Refer to item 39b 

If box 4 is marked in item 39b, ASK item 39d. 

If not - Is box 2 or 3 marked in item 39b AND 
the respondent is under 25 years old? 

D No - Skip to item 40a 
DYes - Ask item 39d 



39d.Did you take any of these lessons or classes 
in the past year? 



1DN0 
2 DYes 



40a. (Have you EVER taken lessons or classes) in 
acting or theater? 



J°l] 1 D No - Skip to item 4 1a 
2 DYes 



b. Did you take these lessons when you were 

Read categories. (Do not read category 4 if 
respondent is under 25 years old.) 
Mark (X) all that apply. 



iDLess than 12 years old 
2D 12-1 7 years old 
3D 18-24 years old 
4 D 25 or older 



CHECK 
ITEM B 



Refer to item 38b 

If box 4 is marked in item 38b, ASK item 38d. 

If not - Is box 2 or 3 marked in item 38b AND 
the respondent is under 25 years old? 

D No - Skip to item 39a 
DYes - Ask item 38d 



CHECK 
ITEM E 



SSd.Did you take any of these lessons or 
classes in the past year? 



1DN0 
2 DYes 



Refer to item 40b 
Is box 1 or 2 marked in item 40b? 
DNo - Skip to Check Item F 
DYes - Ask item 40c 



40c. Were these lessons or classes offered by the 
elementary or high school you were 
attending or did you take these lessons 
elsewhere? 



1 D Elementary/high school 
2D Elsewhere 
3D Both 



FORM SPPA-2 (4-9-92I 



Page 5 



1 992 Survey of Public Participation in the Arts I 87 



CHECK 
ITEM F 



Refer to item 40b 

If box 4 is marked in item 40b, ASK item 40d. 

If not - Is box 2 or 3 marked in item 40b AND 
the respondent is under 25 years old? 

□ No - Skip to item 4 1a 

□ Yes - Ask item 40d 



40d.Did you take any of these lessons or classes 
in the past year? 



iDNo 

2D Yes 



41 a. (Have you EVER taken lessons or classes) in 
ballet? 



i □ No - Skip to item 42a 
2DYes 



b. Did you take these lessons when you were 

Read categories. (Do not read category 4 if 
respondent is under 25 years old.) 
Mark (X) all that apply. 



1 □ Less than 1 2 years old 

2 □ 1 2-1 7 years old 
3D 18-24 years old 
4 □ 25 or older 



CHECK 
ITEM G 



Refer to item 41b 

Is box 1 or 2 marked in item 41b? 

□ No - Skip to Check Item H 

□ Yes - Ask item 41c 



41c. Were these lessons or classes offered by the 
elementary or high school you were 
attending or did you take these lessons 
elsewhere? 



iD Elementary/high school 
2 □ Elsewhere 
3D Both 



CHECK 
ITEM H 



Refer to item 41b 

If box 4 is marked in item 41b, ASK item 41d. 

If not - Is box 2 or 3 marked in item 41b AND 
the respondent is under 25 years old? 

□ No - Skip to item 42a 

□ Yes - Ask item 41 d 



41 d. Did you take any of these lessons or classes 
in the past year? 



™D iDNo 
2D Yes 



42a. (Have you EVER taken lessons or classes) in 
dance, other than ballet such as modern, folk 
or tap? 



1 □ No - Skip to item 43a 
2D Yes 



b. Did you take these lessons when you were - 

Read categories. (Do not read category 4 if 
respondent is under 25 years old.) 
Mark (X) all that apply. 

Ll < □ Less than 12 years old 
2D 12-1 7 years old 
3D 18-24 years old 
4 □ 25 or older 



CHECK 
ITEM I 



Page 6 



Refer to item 42b 

Is box 1 or 2 marked in item 42b? 

□ No - Skip to Check Item J 

□ Yes - Ask item 42c 



42c. Were these lessons or classes offered by the 
elementary or high school you were 
attending or did you take these lessons 
elsewhere? 



iD Elementary/high school 
2D Elsewhere 
3D Both 



CHECK 
ITEM J 



Refer to item 42b 

If box 4 is marked in item 42b, ASK item 42d. 

If not - Is box 2 or 3 marked in item 42b AND 
the respondent is under 25 years old? 

□ No - Skip to item 43a 

□ Yes - Ask item 42d 



42 d. Did you take any of these lessons or classes 
in the past year? 



jZEl iDNo 

zD Yes 



43a. Have you EVER taken lessons or classes in 
creative writing? 

120 I 1 □ No - Skip to item 44a 
2 0Yes 



b. Did you take these lessons when you were 

Read categories. (Do not read category 4 if 
respondent is under 25 years old) 
Mark (X) all that apply. 



i □ Less than 1 2 years old 
2~12-17 years old 
3^1 8-24 years old 
i □ 25 or older 



CHECK 
ITEM K 



Refer to item 43b 

Is box 1 or 2 marked in item 43b? 

□ No - Skip to Check Item L 

□ Yes - Ask item 43c 



43c. Were these lessons or classes offered by the 
elementary or high school you were 
attending or did you take these lessons 
elsewhere? 



1 □ Elementary/high school 
2D Elsewhere 
3D Both 



CHECK 
ITEM L 



Refer to item 43b 

If box 4 is marked in item 43b, ASK item 43d. 

If not - Is box 2 or 3 marked in item 43b AND 
the respondent is under 25 years old? 

□ No - Skip to item 44a 

□ Yes - Ask item 43d 



43d. Did you take any of these lessons or classes 
in the past year? 



iDNo 
2QYes 



44a. (Have you EVER taken a class) in art 
appreciation or art history? 



124 I 1 □ No - Skip to item 45a 
2D Yes 



b. Did you take this class when you were - 

Read categories. (Do not read category 4 if 
respondent is under 25 years old.) 
Mark (X) all that apply. 



[I Less than 12 years old 
2D 12-1 7 years old 
jd 8-24 years old 
4 □ 25 or older 



FORM SPPA-2 (4-9-92) 



88 I Turning On and Tuning In 



CHECK 
ITEM M 



Refer to item 44b 

is box 1 or 2 marked in item 44b? 

□ No - Skip to Check Item N 

□ Yes - Ask item 44c 



44c. Was this class offered by the elementary or 
high school you were attending or did you 
take this class elsewhere? 



45c. Was this class offered by the elementary or 
high school you were attending or did you 
take this class elsewhere? 



i □Elementary/high school 
2D Elsewhere 
sD Both 



CHECK 
ITEM P 



iD Elementary/high school 
2D Elsewhere 
3D Both 



CHECK 
ITEM N 



Refer to item 44b 

If box 4 is marked in item 44b, ASK item 44d. 

If not - Is box 2 or 3 marked in item 44b AND 
the respondent is under 25 years old? 

□ No - Skip to item 45a 

□ Yes - Ask item 44d 



44d.Did you take any of these lessons or classes 
in the past year? 



iDNo 
2D Yes 



45a. (Have you EVER taken a class) in music 
appreciation? 



iH] 1 □ No - Skip to item 46a 
2D Yes 



b. Did you take this class when you were ■ 

Read categories. (Do not read category 4 if 
respondent is under 25 years old.) 
Mark (X) all that apply. 

129 I 1 □ Less than 1 2 years old 
2 □ 1 2-1 7 years old 
sD 18-24 years old 
-Z25 or older 



Refer to item 45b 

If box 4 is marked in item 45b, ASK item 45d. 

If not - Is box 2 or 3 marked in item 45b AND 
the respondent is under 25 years old? 

□ No - Skip to item 46a 

□ Yes - Ask item 45d 



45d.Did you take this class in the past year? 



iDNo 
2 0Yes 



46a. What is the highest grade (or year) of regular 
school your FATHER completed? 



Ll 01 D7th grade or less 
02 08th grade 
03D9th-11th grades 
04 □ 12th grade 

05 □ College (did not complete) 
06 □ Completed college (4+ years) 
07DPost graduate degree (M.A., Ph.D., M.D., J.D., etc.) 
08 □ Don't know 

b. What is the highest grade (or year) of regular 
school your MOTHER completed? 



133 1 01 D7th grade or less 
02 □ 8th grade 
osD9th-11th grades 
04 □ 12th grade 

os □ College (did not complete) 
06 □ Completed college (4+ years) 
07DPost graduate degree (M.A., Ph.D. 
08 □ Don't know' 



M.D., J.D., etc.) 



CHECK 
ITEM O 



Refer to item 45b 

Is box 1 or 2 marked in item 45b? 

□ No - Skip to Check Item P 

□ Yes - Ask item 45c 



CHECK 
ITEM Q 



Is this the LAST household member to be 
interviewed? 

□ No - Go back to the NCS-1 and interview the 

next eligible NCS household member 

□ Yes - END INTERVIEW 




FORM SPPA-2 (2-9-92) 



Page 7 



Appendix B: 

Technical Discussion 
of Logistic Regression 



Logistic regression is useful in explaining why an event occurs or in predicting 
whether or not an event will occur, e.g. whether or not a person watches 
televised symphony concerts. The response alternatives in this case are "yes" or 
"no." The usual multiple regression model is written in linear form as 

Y= Po + plX] + . . . + $„x n 

In such a case, the dependent variable, Y, is ordinarily a continuous variable, 
and certain additional conditions must be met to have fullest confidence in the 
utility of the model. 22 

When the dependent variable takes only two values, as is often the case in 
survey data, the linear regression model fails, and an alternative model is 
required. The logistic regression model is one such alternative, and it can be 
written as 

Prob (event) = 

1 + / 

or 

1 



Prob (event) = 

\ + e~ z 

where .Zis the linear combination 

Z= po + Pl-Xi + 02*2 + • • • + $nXn 

for n explanatory variables. 

An alternative and equivalent expression is 



, Prob (event) 

log ( ) = Z 

Prob (no event) 



89 



90 I Turning On and Tuning In 



which is the log of the odds. The dependent variable, Z, may be attendance at 
a classical music concert, with a "yes" response coded as 1 and "no" as 0. One 
explanatory variable, sayX/, maybe gender, with "male" coded as 1 and "female" 
as 0, and another, say X2, may be household income, measured in dollars. A 
positive value of (3/ would indicate higher attendance by males than females, 
and a positive value of (32 would indicate higher participation by higher income 
families, as compared with lower income families. In logistic regression, the 
model parameters p(/) are estimated using the maximum-likelihood method. 

For large sample sizes, the test that a coefficient is (i.e., that an explanatory 
variable has no influence) is based on the Wald statistic, which has a chi-square 
distribution. The goodness of fit of the model can be assessed by a variety of 
indicators. Those used in this monograph are the log-likelihood, which com- 
pares the statistical results to a "perfect" model, the model chi-square, which 
tests the hypothesis that the included variables jointly are significant, and the 
percent of cases correctly classified by the model. 



Notes 



1 . The three commonly recognized types of market failure are monopoly power, 
positive or negative "externalities" (e.g., pollution), and collective goods (e.g., 
national defense). Broadcasting media are an example of the last of these; the arts 
in general are said to have characteristics of positive externalities. For a more 
complete discussion, see James Heilbrun and Charles M. Gray, The Economics of 
Art and Culture (New York: Cambridge University Press, 1993), especially chap- 
ter 11. 

2. See Jack Faucett and Associates, Arts Participation in America: 1982—1992, Na- 
tional Endowment for the Arts Research Division Report #27, 1993, Appendix 
C.4. 

3. A third, intermediate survey was conducted in 1985. Preliminary analyses of that 
data indicate that they offer insufficient new information to merit analysis. The 
trend patterns are inferior to those offered by the 1982 results, and the 1992 survey 
offers a superior account of current patterns. All three are included in Arts 
Participation in America: 1982—1992, Research Division Report #27, National 
Endowment for the Arts (October 1993). 

4. See the discussion in James Heilbrun and Charles M. Gray, The Economics of Art 
and Culture: An American Perspective (Cambridge: Cambridge University Press, 
1993), chapter 16. 

5. For a lively and entertaining account of the early history of cultural programming 
on commercial television, see Brian G. Rose, Television and the Performing Arts 
(Westport, CT: Greenwood Press, 1986). For an account of programming sup- 
ported by the National Endowment for the Arts, see NEA, The Arts on Television, 
1976—1990, compiled by Rebeccah Krafft (Washington, DC: U.S. Government 
Printing Office, 1991). 

6. For a more extensive discussion of these points, see James Heilbrun and Charles 
M. Gray, The Economics of Art and Culture: An American Perspective (Cambridge: 
Cambridge University Press, 1993), pp. 199-216, and the citations therein. 

7. In the broadcast case, we can think of a supply curve that is horizontal ("perfectly 
elastic") at a zero price. 

8. For a general discussion, see James Heilbrun and Charles M. Gray, The Economics 
of Art and Culture: An American Perspective (Cambridge: Cambridge University 
Press, 1993), pp. 302-306. Two studies by James Heilbrun offer additional 
insights. See his "Growth and Geographic Distribution of the Arts in the U.S.," in 
Douglas V. Shaw et al., eds., Artists and Cultural Consumers (Akron: Association 
for Cultural Economics, 1987), pp. 24-36, and "The Distribution of Arts Activity 
Among U.S. Metropolitan Areas," in Douglas V. Shaw et al., eds., Cultural 
Economics, 88: An American Perspective (Akron: Association for Cultural Econom- 
ics, 1989), pp. 33-40. 



91 



92 I Turning On and Tuning In 



9. This is due largely to Tibor Scitovsky, The Joyless Economy (Oxford University 
Press, 1976), ch. 11, "Our Disdain for Culture," pp. 224-247. 

10. We should note that the survey instrument asks about "listening" to a recording, 
not "buying" one. Yet someone must have purchased a CD or other sound recording, 
and it seems reasonable to assume that the respondent may, in fact, have been the 
purchaser. Hence we treat listening as roughly synonymous with purchasing for 
purposes of hypothesis development. 

11. U.S. Bureau of the Census, Statistical Abstract of the United States: 1993 (Washing- 
ton, DC: U.S. Government Printing Office, 1993), Table 901, p. 561. 

12. Ibid. 

13. A technical discussion of econometric techniques, including regression, is far 
beyond the scope of this monograph. The interested reader can pursue any of a 
number of elementary treatments, including Peter Kennedy, A Guide to Economet- 
rics, 3rd. ed. (Cambridge, MA: MIT Press, 1993). A brief exposition of the 
technique used here is included in Appendix B. 

14. Perhaps the classic reference is G. S. Maddala, Limited-Dependent and Qualitative 
Variables in Econometrics (Cambridge: Cambridge University Press, 1 983). See also 
Marija J. Norusis, SPSS Advanced Statistics Users Guide (Chicago: SPSS Inc., 
1990), pp. 45-69. 

15. Technically, substitutability and complementarity would be indicated by the signs 
of estimated cross-price elasticity of participation. The data do not support such 
determination in this case. 

16. Ideally, tests of equality of coefficients would inform hypotheses pertaining to 
changes in behavior and/or structure over time. Such tests would not be wholly 
applicable in this instance because of changes in the survey instrument (e.g., 
"dance," more broadly defined, replacing "ballet") and variable measures (e.g., 
incomparability of income groupings due to inflation and category changes). 

17. As has been suggested earlier, inclusion of media participation indicators as 
explanatory variables in the live participation equation, and vice versa, would yield 
information on cross-participation. Unfortunately, such inclusion also creates 
econometric difficulties, the correction of which requires simultaneous estimation 
techniques. Accordingly, what is reported in the tables that follow are the simpler 
results. A more technical treatment of this issue, beyond the scope of this mono- 
graph, will be forthcoming as a research note. 

18. For a more complete description of the Wald statistic, see Marija J. Norusis, SPSS 
Advanced Statistics Users Guide (Chicago: SPSS Inc., 1990), p. 48. 

19. For a more complete description of these tests and indicators, see Ibid., pp. 50-53. 

20. See Joni Maya Cherbo and Monnie Peters, American Participation in Opera and 
Musical Theater — 1992, National Endowment for the Arts Research Division 
Report #32. 

2 1 . The blockbuster phenomenon is discussed in James Heilbrun and Charles M . Gray, 
The Economics of Art and Culture (New York: Cambridge University Press, 1 992), 
pp. 185-187. 

22. See any standard econometrics textbook, e.g. Robert S. Pinduyck and Daniel L. 
Rubinfeld, Econometric Models and Economic Forecasts, 3rd ed. (New York: 
McGraw-Hill, Inc., 1991). 



Notes I 93 



23. Ibid., ch. 10, "Models of Qualitative Choice." 

24. For a more complete discussion, see Marija J. Norusis, SPSS Advanced Statistics 
User's Guide (Chicago: SPSS, Inc., 1990), especially pp. 48-51. 



Bibliography 



Arts and Cultural Programs on Radio and Television. Research Division Report 
#4. Washington, DC: National Endowment for the Arts, 1977. 

The Arts on Television, 1976-1990. Compiled by Rebeccah Krafft. Washington, 
DC: National Endowment for the Arts, 1991. 

Beck, Kirsten. Cultivating the Wasteland. New York: American Council for the 
Arts, 1983. 

Cornwell, Terri Lynn. Democracy and the Arts: The Role of Participation. New 
York: Praeger, 1990. 

Heilbrun, James, and Charles M. Gray. The Economics of Art and Culture: An 
American Perspective. New York: Cambridge University Press, 1993. 

Jack Faucett Associates. Arts Participation in America: 1982—1992. Compiled 
by John P. Robinson. Research Division Report #27. Washington, DC: 
National Endowment for the Arts, 1993. 

Kubey, Robert, and Mihaly Csikszentmihalyi. Television and the Quality of Life: 
How Viewing Shapes Everyday Experience. Hillsdale, NJ: Lawrence Erlbaum 
Associates, 1990. 

Lynes, Russell. The Lively Audience. New York: Harper and Row, 1985. 

Maddala, G. S. Limited-Dependent and Qualitative Variables in Econometrics. 
Cambridge: Cambridge University Press, 1983. 

Rose, Brian G. Television and the Pe forming Arts. New York: Greenwood Press, 
1986. 

Waterman, David. "Arts and Cultural Programming on Cable Television: 
Economic Analysis of the U.S. Experience," in Economic Efficiency and the 
Performing Arts. Edited by Nancy K. Grant et al. Akron, OH: Association 
for Cultural Economics, 1987. 



94 



About the Author 



Charles M. Gray is Professor of Economics in the Graduate School of 
Business at the University of St. Thomas (Minnesota). In addition to 
numerous articles and book chapters on various aspects of the economics of the 
arts, he has authored (with James Heilbrun) The Economics of Art and Culture: 
An American Perspective (Cambridge University Press, 1993). He is an active 
member of the Association for Cultural Economics, International. He has 
chaired the board of a small dance company and regularly consults and conducts 
management training seminars for nonprofit organizations. 



Other Reports on the 1 992 SPPA 



The following publications report on various aspects of the 1 992 Survey of 
Public Participation in the Arts. Information regarding availability may be 
obtained by writing to the National Endowment for the Arts, Research Division, 
1100 Pennsylvania Avenue, NW, Washington, DC, 20506. 

Age Factors in Arts Participation, Richard A. Peterson and Darren E. Sherkat 

American Participation in Dance, Jack Lemon/Jack Faucett Associates 

American Participation in Opera and Musical Theater — 1992, Joni Maya Cherbo 
and Monnie Peters 

American Participation in Theater, Chris Shrum/AMS Planning and Research 

Americans' Personal Participation in the Arts, Monnie Peters and Joni Maya 
Cherbo 

Arts Participation and Race I Ethnicity, Jeffrey Love and Bramble C. Klipple 

Arts Participation by the Baby Boomers, Judith Huggins Balfe and Rolf 
Meyersohn 

Cross-Over Patterns in Arts Participation, Richard J. Orend and Carol Keegan 

Effects of Education and Arts Education on Americans' Participation in the Arts, 
Louis Bergonzi and Julia Smith 

Hold the Funeral March: The State of Classical Music Appreciation in the U.S., 
Nicholas Zill 

Jazz in America — Who s Listening?, Scott DeVeaux 

Patterns of Multiple Arts Participation, Jeffrey Love 

Reading in the 1990s: Turning a Page or Closing the Books?, Nicholas Zill 

Socialization in the Arts — 1992, Richard J. Orend and Carol Keegan 



96 



national 
endowment 

forW^the 

ARTS 



Seven Locks Press 
Carson, California 



ISBN 0-929765-39-7 




9 780929V65396 



51